European Competition Law Annual 2002: Constructing the EU Network of Competition Authorities 9781472559661, 9781841133663

The European Competition Law Annual 2002 is the seventh in a series of volumes following the annual workshops on EU Comp

227 63 2MB

English Pages [523] Year 2004

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

European Competition Law Annual 2002: Constructing the EU Network of Competition Authorities
 9781472559661, 9781841133663

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

LIST OF SPONSORS 2002

Akin, Gump, Strauss, Hauer & Feld LLP Contact: Professor Jacques H H Bourgeois Brussels Office Avenue Louise 65 PB No 7 B-1050 Brussels Tel: (32 2) 535 29 11 Fax: (32 2) 535 29 00 E-mail: [email protected] Blake, Cassels & Graydon LLP Contact: Calvin S Goldman, QC Commerce Court West 199 Bay Street Toronto, Ontario Canada M5L 1A9 Tel: 416 863 2280 Fax: 416 863 2653 E-mail: [email protected] Cleary, Gottlieb, Steen & Hamilton Contact: Professor Mario Siragusa Rome Office Piazza di Spagna 15 I-00187 Rome Tel: (06) 695 221 Fax: (06) 692 00 665 E-mail: [email protected]

vi

List of Sponsors

Howrey Simon Arnold & White Contact: James Rill, Esq 1299 Pennsylvania Ave NW Washington, DC 20004 Tel: (001) 202 383 6562 E-mail: [email protected] Fax: (001) 202 383 6610 Martinez Lage & Asociados Contact: Santiago Martínez Lage Claudio Coello 37 – 28001 Madrid Tel: (34) 91 426 44 70 Fax: (34) 91 577 37 74 E-mail: [email protected] Skadden, Arps, Slate, Meagher & Flom LLP Contact: Professor Barry Hawk Brussels Office 523 Avenue Louise B-1050 Brussels Tel: (32 2) 639 03 00 Fax: (32 2) 639 03 39 E-mail: [email protected] White & Case/Forrester Norral & Sutton Contact: Professor Ian Forrester Brussels Office 62, rue de la Loi B-1040 Brussels Tel: (32 2) 219 16 20 Fax: (32 2) 219 16 26 E-mail: [email protected] Wilmer, Cutler, Pickering, Hale & Dorr LLP Contact: John Ratliff Brussels Office Bastion Tower Place du champ 5 B-1040 Brussels Tel: (32 2) 285 49 08 Fax: (32 2) 285 49 49 E-mail: [email protected]

Table of Cases I.

EC Cases

ABB v Commission, Case T–31/99, [2002] ECR II–1881 ................................476 Adams, Case 145/83, [1985] ECR 3539 ..........................................................478 AKZO Chemie v Commission, Case 53/85, [1986] ECR 1965 ....362, 375, 413–5 AM&S Europe v Commission, Case 155/79, [1982] ECR 1575 ..............297, 415 Annibaldi, Case C–309/96, [1997] ECR I–7493 ..............................................441 Automec v Commission, Case T–24/90, [1992] ECR II–2223..........................196 Baustahlgewebe v Commission, Case C–185/95 P, [1998] ECR I–8417..........................................................................................441, 449 BECTU, Case C–173/99 [2001] ECR I–4881 ..................................................444 Belgische Radio en Televisie et Société belge des auteurs, compositeurs et éditeurs v SV SABAM et NV Fonior (BRT-I), Case 127/73, [1974] ECR 51........................................................190, 372, 461 Booker Acquaculture and Hydro Seafood, Joined Cases C–20/00 and C–64/00, [2003] ECR I–7411................................................................444 Bostock, Case C–2/92, [1994] ECR I–955........................................................416 British-American Tobacco and RJ Reynolds, Joined Cases 142/84 and 156/84, [1987] ECR 487................................................................240, 362 Cimenteries CBR, Case T–25/95, [2000] ECR II–491 ....................................385 Der Grüne Punkt — Duales System Deutschland AG v Commission (DSD System), Case T–151/01 R 1, Court Order of 15.11.2001 ........168, 177 Dirección General de Defensa de la Competencia v Asociación Española de Banca Privada (AEB), Case C–67/91, [1991] ECR I–4785 ..................................290–2, 297, 343, 356, 369, 374–5, 378–82, 387, 394, 400, 405, 407, 411, 417, 452–3, 477 Dow Benelux v Commission, Case 85/87, [1989] ECR 3137 ..........279, 291, 297, 374–5, 382, 407, 411, 413, 477 DSR Senator Lines v Commission, Case T–191/98 R, [1999] ECR II–2531 ................................................................................................481 DSR-Senator Lines v Commission, Case C–364/99 P (R), [1999] ECR I–8733..................................................................................................481 ERT, Case C–260/89, [1991] ECR I–2925 ..............................................196, 482 Foto-Frost, Case 314/85, [1987] ECR 419................................................190, 241 Friedrich Kremzow v Republik Österreich, Case C–299/95, [1997] ECR I–2629..................................................................................................196 Heylens, Case C–222/86, [1987] ECR 4097 ....................................................387 Hilti v Commission, Case T–30/89, [1990] ECR II–163 ..........................297, 415

x

Table of Cases

Hoechst v Commission, Joined Cases 46/87 and 227/88, [1989] ECR 2859 ........................................................382–3, 453, 473–4, 482 Hubert Wachauf v Federal Republic of Germany, Case 5/88, [1989] ECR 2609 ........................................................................................416 Hüls v Commission, Case C–199/92 P, [1999] ECR I–4287 ....................449, 479 Hüseyin Gözütok and Klaus Brügge, Joined Cases C–187/01 and C–385/01, judgment of 11 February 2003 ..................................................451 IMS Health, Case T–184/2001 R, [2001] ECR II–0000 ..................................467 Irish Sugar plc, Commission Decision, Case IV/34.621, 35.059/F-3, OJ L 258 of 22.09.1997 …………………………………………………………286 Limburgse Vinyl Maatschappij (LVM) v Commission and Others, Joined Cases C–238/99 P, C–244/99 P, C–245/99 P, C–247/99 P, C–250/099 P to C–252/99 P and C–254/99), [2002] ECR I–8375 ..............453 Mannesmannröhren-Werke v Commission, Case T–112/98, [2001] ECR II–729 ..................................................410, 441, 453, 475–6, 482 Masterfoods Ltd v HB Ice Cream Ltd, Case C–344/98, [2000] ECR I–1412 ......................................................190, 238, 240, 241, 467 Matra v Commission, Case T–17/93, [1994] ECR II–595................................375 Michelin, Commission Decision of 20 June 2001, Case COMP/E-2/36041 ................................................................................305, 466 Ministère public v Asjes, Joined Cases 209–213/84, [1986] ECR 1425 ....................................................................................................461 Nouvelles Frontières, Joined Cases 209–213/84, [1986] ECR–1425 ..............372 Orkem v Commission, Case 374/87, [1989] ECR 3283 ....................................410 Postbank v Commission, Case T–353/94, [1996] ECR II–921 ..........................................................................292, 293, 376, 414 Rhône- Poulenc v Commission, Case T–1/89, [1991] ECR II–867 ..................449 Roquette Frères SA v DGCCRF, Case C–94/00, [2002] ECR I–9011..................................................................................................474 SEP v Commission, Case T–39/90, [1991] ECR II–1497 ........375, 377, 407, 412 SEP v Commission, Case C–36/92 P, [1994] ECR I–1911 ..........363, 375, 412–5 Stauder v Stadt Ulm, Case 29/69, [1969] ECR 419 ..........................................441 Stergios Delimitis v Henninger Bräu, Case C–234/89, [1991] ECR I–935............................................................................................190, 241 Visa International — Multilateral Interchage Fee, Commission Decision of 24 July 2002, COMP 29.373 ............................................305, 466 The Queen v Secretary of State for Transport ex parte Factortame (No. 2), Case C–213/89, [1990] ECR I–2433 ..............................................414 Walt Wilhelm v Bundeskartellamt, Case 14/68, [1969] ECR 1 ................................................................................49, 302, 446, 470–1 Warner Lambert/Gilette and Others and BIC/Gillette and Others, Commission Decision of 10 November 1992 in Cases No IV/33.440 and No. IV/33.486, OJ L 116/21 of 12.5.1993 ............................................240 Woodpulp, Case C–89/85, [1993] ECR I–1307 ................................................385

Table of Cases

II. 1.

xi

Other Cases United Kingdom

A v National Blood Authority [2001] 3 All ER 289–416 (UK) ........................352

2.

United States

Blockburger v US, 284 US 299 (1932)......................................................302, 446 Eastman Kodak Co v Image Technical Services, 504 US 451 (1992) ..............104 Matsushita Electric Industrial Co v Zenith Radio Corp 471 US 1002 (1985) ..................................................................................................104 Microsoft — case documents (judicial decisions and settlement) available at http://www.naag.org/issues/microsoft ........................................................17 US v Dixon, 113 S.Ct. 2849, 2856 (1993) ........................................................447 US v Koon, 34 F.3d 1416, 1438 (9th Cir.’94) ....................................................447 US v Lanza, 260 US 377, 382 (1922) ..............................................................447 US v Federal Communications Commission, 652 F.2d 72, 88 (DC Cir. 1980)........................................................................................................86

3.

European Court for Human Rights

ECHR Beaumartin v France, judgment of 24 November 1994, Series A no. 296-B ..............................................................................191, 462 ECHR Belilos v Switzerland, judgment of 29 April 1988, Series A no. 132 ..........................................................................................462 ECHR Bendenoun v France, judgment of 24 February 1994, Series A no. 284 ..................................................................................448, 459 ECHR Borgers v Belgium, judgment of 30 October 1991, Series A no. 214-B ......................................................................................458 ECHR Brumarescu v Romania, Case 28342/95, judgment of 28 October 1999 ................................................................................................457 ECHR Cantoni v France, judgment of 15 November 1996..............................468 ECHR De Cubber v Belgium, judgment of 26 October 1984, Series A no. 86 ....................................................................................304, 459 ECHR DSR Senator Lines v 15 Member States of the European Union, Case 56672/00 (judgment pending) ................................................445 ECHR Findlay v United Kingdom, judgment of 25 February 1997, Reports 1997–I ..........................................................................304, 459 ECHR Fischer v Austria, judgment of 29 August 2001, Case 37950/97 ..............................................................................302, 446, 448, 491

xii

Table of Cases

ECHR Franz Fischer v Austria, judgment of 29 May 2001......................471, 472 ECHR Fredin v Sweden, judgment of 18 February 1991, Series A no. 192..........................................................................................................457 ECHR Funke v France, judgment of 25 February 1993, Series A no. 256-A......................................................................................................455 ECHR Gasus Dosier — und Fördertechnik GmbH v the Netherlands, judgment of 23 February 1995, Series A no. 306-B ....................................456 ECHR Gradinger v Austria case, judgment of 23 October 1995, Series A no. 328-C ......................................................................................446, 448, 471 ECHR Halford v United Kingdom, judgment of 25 June 1997, Reports 1997–III ........................................................................................................456 ECHR Iatridis v Greece, Case 31107/96, judgment of 25 March 1999 ..........456 ECHR JB v Switzerland, judgment of 3 May 2001, Case 31827/96 ................455 ECHR JJ v the Netherlands and KDB v the Netherlands, judgment of 27 March 1998, Reports 1998–II ................................................................458 ECHR John Murray v United Kingdom, judgment of 8 February 1996, Reports of Judgments and Decisions 1996–I ..............................................455 ECHR Kokkinakis v Greece case, judgment of 25 May 1993, Series A, no. 260-A, para. 52 ......................................................................................468 ECHR Kress v France, Case 39594/98, judgment of 7 June 2001 ..................458 ECHR Le Compte, Van Leuven and De Meyere v Belgium, judgment of 23 June 1981, Series A, no. 43 ............................................................462, 480 ECHR Lobo Machado v Portugal, judgment of 20 February 1996, Reports 1996–I ............................................................................................458 ECHR Lutz v Germany, judgment of 25 August 1987, Series A, no. 123 ........448 ECHR M & Co v Germany, judgment of 9 February 1990, Case 13258/87 ..............................................................................................445 ECHR Matthews v United Kingdom, Case 24833/94, judgment of 18 February 1999 ................................................................................445, 484 ECHR Niemitz v Germany, judgment of 16 December 1992, Series A, no. 251-B ..............................................................................................456, 473 ECHR Oliveira v Switzerland, judgment of 30 July 1998, Reports of Judgments and Decisions 1998–V ......................................................446, 471 ECHR Öztürk v Germany, judgment of 21 February 1984, Series A, no. 73............................................................................................192, 448, 459 ECHR PG and JH v United Kingdom, Case 44787/98, judgment of 25 September 2001 ......................................................................................456 ECHR Ponsetti and Chesnel v France, judgment of 14 September 1999, ECHR 1999–VI ............................................................................................471 ECHR Reinhardt and Slimane-Kaïd v France, judgment of 31 March 1998, Reports 1998–II ................................................................458 ECHR Ringei v Austria, judgment of 16 July 1971, Series A, no. 13 ..............462 ECHR Ruiz-Mateos v Spain, judgment of 23 June 1993, Series A, no. 262 ....458

Table of Cases

xiii

ECHR Saunders v United Kingdom, judgment of 17 December 1996, Reports 1996–VI. b ..........................................................................455, 475–6 ECHR Servès v France, judgment of 20 October 1997, Reports 1997–VI........................................................................................................455 ECHR Sporrong and Lönnroth v Sweden, judgment of 23 September 1982, Series A, no. 52 ................................................................457 ECHR Tre Traktörer AB v Sweden, judgment of 7 July 1989, Series A, no. 159, para. 53 ..........................................................................................457 ECHR Van Orshoven v Belgium, judgment of 25 June 1997, Reports 1997–III ........................................................................................................458 ECHR Vermeulen v Belgium, judgment of 20 February 1996, Reports 1996–I ............................................................................................458 ECHR Z v Finland, judgment of 25 February 1997, Reports 1997–I ..............457

INTRODUCTION

This volume includes the transcripts of the discussion at, and the written contributions to, the seventh edition of the Annual EU Competition Law and Policy Workshop, held on 12–13 June 2002 at the European University Institute in Florence. The Annual EU Competition Law and Policy Workshop is a programme set up in 1996 by professors Giuliano Amato and Claus-Dieter Ehlermann at the Robert Schuman Centre of the European University Institute. Every year the Workshop brings together top-level EU and international policy-makers, academics and legal practitioners, to discuss, in an informal environment, critical issues of EC competition law and policy. The seventh edition of the Workshop (2002), which makes up the present publication, concluded a series of three consecutive annual roundtables devoted to the reform of EC antitrust enforcement. The previous two editions were devoted respectively to discussing the Commission’s reform proposal (2000), and to examining the pre-conditions for an effective private enforcement of EC antitrust rules in the decentralised enforcement system about to be introduced by the reform (2001).1 The 2002 edition analyzes issues related to the

1 In particular, the fifth edition of the Workshop (2000) examined in detail the European Commission’s

ideas with respect to the decentralisation of EC antitrust enforcement, as published in the ‘White Paper on Modernisation of the Rules implementing Articles 85 and 86 of the EC Treaty’ of May 1999, and the public reactions provoked by it (see Claus D. Ehlermann and Isabela Atanasiu, eds. (2001): European Competition Law Annual 2000: The Modernization of EC Antitrust Policy, Hart Publishing, Oxford and Portland Oregon). The sixth (2001) carried further the debate on the modernization project by examining the conditions for an effective private enforcement of EC competition law. In particular, it examined whether, and under what conditions, private action before national courts and arbitration tribunals can effectively play a greater role in the enforcement of EC antitrust rules (see Claus D. Ehlermann and Isabela Atanasiu, eds. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland Oregon). A few comments about other editions of the Annual EU Competition Law and Policy Workshop: the first (1996) examined problems of implementation of competition law and policy in a ‘federal’ context (see Claus D. Ehlermann and Laraine L. Laudati, eds. (1997): The Robert Schuman Centre Annual on European Competition Law 1996, Kluwer Law International, London). The second (1997) discussed the objectives of competition law and policy (see Claus D. Ehlermann and Laraine L. Laudati, eds. (1998): European Competition Law Annual 1997: Objectives of Competition Policy, Hart Publishing, Oxford and Portland Oregon). The third (1998) concentrated on the application of competition policy in the evolving communication and information markets (see Claus D. Ehlermann and Louisa Gosling, eds. (2000): European Competition Law Annual 1998: Regulating Telecommunications, Hart Publishing, Oxford and Portland Oregon). The fourth (1999) studied three groups of problems in the field of EU state aid control: the economic justifications for granting state

Introduction

xvi

setup and functioning of an EU network of competition authorities, composed of the European Commission and the competition authorities of the Member States. Background. The concept of an EU competition network surfaced in the early nineties, in the context of discussions about the decentralised application of EC competition rules by the competition authorities of the Member States. The Commission’s modernization project turned the spotlight on the network concept, given its importance for the success of the reform. The core of the modernization project is the replacement of the centralized prior authorisation system (which was based on the Commission’s monopoly over the application of the exemption provisions in Article 81(3) EC) by a directly applicable exception system, allowing national competition authorities and national courts to apply in full the EC competition provisions. The decentralized application of EC competition rules in a legal exception system increases the enforcement responsibilities of the national competition authorities and implicitly modifies the relationship between public enforcers in Europe, both on the vertical dimension (between the Commission and national competition authorities) and on the horizontal dimension (between the national competition authorities). In the new enforcement system, the effective application of EC competition rules will depend in good part on cooperation between the members of the EC competition network, as on the existence and effectiveness of mechanisms for ensuring coherence and consistency of enforcement. The timing of the seventh edition of the Workshop (2002) is significant for understanding the background against which the discussions took place, as well as the significance of the emerging conclusions. The Commission’s reform proposal underlined the importance of the EC competition network in the new enforcement system, but did not cover all the aspects related to its establishment and practical functioning. In the Modernization White Paper of May 1999 and the formal reform proposal of September 2000,2 the Commission had addressed in aid, state aid in the banking sector, and the possibilities for a decentralised approach to the control of state aid in the EU (see Claus D. Ehlermann and Michelle Everson, eds. (2001): European Competition Law Annual 1999: State Aid Control in the European Union — Selected Problems, Hart Publishing, Oxford and Portland Oregon). The eighth (2003) discussed the concept of abuse of a dominant position in EC law, its equivalents in other major international jurisdictions, and the legal, economic and practical issues surrounding their application (See Claus-Dieter Ehlermann and Isabela Atanasiu, eds., What Is an Abuse of a Dominant Position?, Hart Publishing, Oxford and Portland, Oregon, forthcoming). The ninth edition of the Workshop took place on 11–12 June 2004, examining the relationship between competition law and the regulation of (liberal) professions. Non-edited versions of the written contributions prepared for the annual EUI Workshops (from the 2000 edition onwards) are also available at http://www.iue.it/RSCAS/Research/Competition/CompetitionLaw Policy.shtml. 2 European Commission (1999): White Paper on the Modernization of the Rules Implementing Articles 85 and 86 of the EC Treaty — Commission Programme No. 99/027, COM(1999) 101 final; Commission Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 82 and 82 of the Treaty and amending Regulations (EEC) No 1017/68, (EEC) No 2988/74, (EEC) No 4056/86 and (EEC) No 3975/87 (‘Regulation implementing Articles 81 and 82 of the Treaty’), COM(2000) 582 final of 28 September 2000, published in OJ C 365E/284 of 19.12.2000.

Introduction

xvii

some detail its vertical relationship with the national competition authorities, and its own central role in policy-making and ensuring the consistent application of EC competition rules, but remained somewhat reluctant to regulate, in detail, the technical aspects related to the functioning of the EC competition network. The 2002 Workshop took place some months before the adoption at the Council of Regulation 1/2003, replacing Regulation 17/62.3 At the time, discussions at the Council on the reform proposal were fairly advanced, but some of its key elements — such as the relationship between EC and national competition law, the exact scope for consultation and cooperation within the EC network of competition authorities, the functions of the Advisory Committee, the adoption of block exemption regulations — were still under debate. While, by that time, a general consensus had emerged on the fact that a range of aspects related to the structure and functioning of the EC competition network needed to be further clarified before the implementation of the reform, a tension could still be perceived between the Commission’s desire to leave room for flexibility and adjustment ‘on the road’ and requests for the adoption of clear rules on a number of key issues related to the functioning of the network. As it turned out, a good part of the latter aspects, several of which were debated at the Workshop, were tackled later on in the so-called ‘modernization package’, containing a series of Commission documents complementing Regulation 1/2003,4 including a Joint Statement by the Council and the Commission on the functioning of the network and a Commission Notice on cooperation within the network.5 The objective of the 2002 Workshop was to contribute with a structured and comprehensive analysis, performed from a fresh and inter-disciplinary perspective, of the various issues related to the setup and functioning of the EC network of competition authorities that remained unclear at the time of the debate. A group of 30 top-level participants, including high-ranking officials from the EC Commission and the national competition authorities of several Member States, outstanding international academics from the fields of law, political sciences and economics, and respected international legal practitioners, were invited to discuss, in an informal, environment, issues ranging from the broad concept of a regulatory network, its characteristics, functionality and outputs, to the very specific legal and technical issues related to the setup and functioning of the EC competition network. The debate proceeded from the more general to the more specific. Panel 1 examined the network concept in general, bringing in the perspective of political scientists on the subject. The participants also reflected on 3 Council

Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ L 1/1 of 4.1.2003. 4 Regulation 1/2003 and the complementary Commission documents came into force on the 1 May 2004. The full ‘modernization package’, as well as the commentaries submitted by interested parties during the public debate preceding its adoption by the Commission, are available on DG COMP’s website at http://europa.eu.int/comm/competition/antitrust/legislation/#new_reg. 5 Commission Notice on Cooperation Within the Network of Competition Authorities, OJ C 101/43 of 27.4.2004.

xviii

Introduction

the EC competition network in light of the historical evolution of antitrust enforcement in Europe, and sought to extract lessons useful for the EC competition network from the functioning of infra-national competition networks (the National Association of Attorneys General in the US, the network between the Bundeskartellamt and the competition authorities of the Länder in Germany) and other EU networks of regulatory bodies (telecoms, electricity and gas). Panel 2 extrapolated the broad systemic issues identified in Panel 1 to the specific context of the EC competition network. The participants thus examined three main groups of issues: a) whether the effectiveness of the EC competition network would depend on a degree of homogeneity among its members; b) the position and role of the Commission in the network; c) the horizontal relationship between the national competition authorities as members of the network, and respectively with their national governments. Panel 3 focused on four technical problem areas that are specific to the functioning of the EC competition network: a) the allocation and re-allocation of cases among the members; b) the scope for, and limits to, the exchange of confidential information within the network; c) issues related to consultation on individual cases and general policy among the members of the network; and d) problems arising in relation to the exchange of information among the members of the network from the perspective of the European Convention on Human Rights and the EU Charter of Fundamental Rights. The discussions brought to light a series of important considerations on the nature and functioning of the EC competition network, on the implications of participating in the network for the national competition authorities, and the role of the national competition authorities in the network. Several of the proposals that emerged during the debate, such as those concerning the need to give to the national competition authorities a more active role in terms of policy-making in the new system, and the related proposals regarding the status and functions of the Advisory Committee in the network, later found confirmation in the so-called ‘modernization package’ (which includes the political principles expressed in the Joint Statement of the Council and the Commission on the network, adopted in parallel with Regulation 1/2003,6 and the more technical rules and criteria formulated in the Commission Notice on cooperation within the network7). The Proceedings. The Workshop was opened by Professor Mario Monti, the EC Commissioner responsible for competition, with a brief statement followed by an exchange of views with the participants. In his intervention Professor Monti commented on some of the topical issues at that stage of the debate on the reform proposals related to the EC competition network. In reasoning the need for a network of public enforcers in the new enforcement system, the Commissioner emphasised that the EC competition network 6 Joint

Statement of the Council and the Commission on the functioning of the network of competition authorities, Interinstitutional file 2000/0243 (CNS), of 10 December 2002, available on the website of the Council. 7 See supra note no. 6.

Introduction

xix

would not be created ex novo, but rather built on the existing links between the Commission and the national competition authorities. The Commission’s view was that the network should maintain its informal character in the future, while the reform would provide opportunities for enhancing cooperation among its members, such as the power for the national competition authorities to apply EC competition rules in their entirety, to exchange confidential information, and to assist each other in case-solving. As to criteria for case allocation among the members of the network, the Commission considered that there was a need to maintain a balance between the objectives of flexibility of action and predictability of outcomes. To be more specific, in a system of parallel enforcement competencies, case allocation criteria should, on the one hand, be flexible and of a qualitative nature, and on the other hand, allow the network members to deal efficiently with individual cases. For these reasons, the Commission was not in favour of including rigid rules on case allocation in the new procedural regulation, but rather favoured their being published in a Commission notice, which could more easily be amended following experience gained in the implementation of the reform. On the substance of the criteria for case allocation, the general principle supported by the Commission was that individual cases should be allocated to the authority best placed to deal with them, taking into account the scope of the suspected infringement, the effects of the decision required to bring it to an end, the location of evidence, and the competition authorities’ enforcement resources and priorities. The Commission itself would deal with infringement-producing effects in more than three Member States. It was considered that, with such a system, problematic case allocation issues would arise only in respect of cases producing effects in more than one and less than three Member States. For such situations, the business community had pleaded for a ‘one-stop-shop’ allocation principle, but the Commission argued that the national competition authorities should not be prevented from acting in parallel if wishing to do so. Consultation among the members of the network on case allocation would have to take place only when the case had not been directed to the best placed authority in the first instance, and should have an informal character, implying inter alia, that decisions on case allocation taken within the network should not be open to appeal by the undertakings. (The Commission Notice on cooperation within the network,8 published in April 2004, takes over the criteria for case allocation and principles of cooperation in the allocation of cases mentioned before.) On the controversial issues of the Commission’s power to take up cases from the national competition authorities (Article 11 (6) of the draft regulation), the Commission clarified that it intended to use this procedure only as a necessary ‘safety valve’ for the system, in relation to situations where there was a risk of inconsistency in the application of the EC competition rules (such as, for example, cases where two national competition authorities dealing with the same case 8 See

supra note no. 6.

xx

Introduction

or subject-matter might reach conflicting decisions, or cases where a national competition authority might reach a decision contradicting the EC case law, as well as cases raising important policy issues). As to the debate on whether a certain degree of harmonization of national procedures and sanctions would be necessary for the success of the reform, the Commission’s view was that, at least during the initial stages of implementation of the reform, harmonization should not go beyond those elements strictly necessary for the functioning of the new system, for example the provisions in the new regulation empowering national competition authorities to assist each other in fact-finding and to reject complaints on grounds that another authority is dealing with the case. Panel 1 laid out the conceptual framework for the debate on the EC competition network. The debate was structured along two main lines. First, the participants reviewed the body of political science literature on policy networks (definitions, structural elements, dynamics, typologies of policy networks) with the purpose of shaping a theoretical framework relevant to the context of the EC competition network. Second, they sought to draw lessons relevant for the establishment of the EC competition network from the analysis of other international and infra-national networks of regulators, such as the EC networks of regulators operating in the telecom, gas and electricity sectors, the US inter-state antitrust network (the National Association of Attorney-Generals — NAAG), and the German infranational competition network comprising the Bundeskartellamt and the competition authorities of the Länder. In addition, the participants considered some aspects related to the interaction between competition authorities and sectoral regulators. For the political scientists, the notion of ‘policy network’ generally describes a method of governance involving the coordination of activity, as opposed to hierarchical or formalised relationships. The basic features of a policy network are: interdependency of the actors composing it, diversity of their goals, continuity of interaction, and agreement on norms. Policy networks are, as a general rule, created on the basis of self-interest, and maintained through a combination of self-interest realisation, reduction of transaction costs, familiarity and membership closure. Somehow distinct from the ‘typical’ policy networks are the so-called ‘epistemic communities’, whose particularity is that they are not driven by the conventional network pressures of functional goal achievement, familiarity and resource dependency, but rather by common causal beliefs and policy objectives. Until the nineties, the political science literature concentrated on the dynamics of policy networks, focusing on understanding policy, rather than on designing networks. More recently, the literature re-oriented towards understanding the political bias embedded in institutions by looking at their historical context and evolution. Both these approaches are useful for the debate on the EC competition network, insofar as they bring some conceptual clarity and identifying the preconditions for an effective policy-implementation network. Furthermore, an analysis of the historical evolution of EC competition policy from the perspective of the policy network literature demonstrates that the ground elements for a

Introduction

xxi

network of EC competition authorities featuring the elements of an epistemic community were already in place before the reform. Another useful approach based on the historical evolution of EC competition policy was to examine the relationship between the Commission and the national competition authorities from the perspective of how it affected the decision-making process. From a methodological point of view, this involved distinguishing between the EC and national decision-making spheres. Following this line of analysis, the concept of an EC network of competition authorities has evolved in three stages. The first corresponds to the early stages of implementation of the EC competition rules, up until the eighties, when the EC and the national decision-making spheres were clearly separated. There was no involvement of the national competition authorities in the EC decision-making sphere, and their relationship with the Commission was very formal. This early model, described as ‘foundational’, was followed in the eighties and nineties by a ‘solar model’, whereby the Commission continued to play the central role, and the national competition authorities became more involved in the EC decision-making sphere, although still in a ‘passive’ manner (following the decision lines of the Commission, and passing on information). The modernization project opens the way to a ‘centralized interactive model’, whereby EC decision-making is decentralized, but according to the reform proposal, the Commission would have still retained a leading role with respect to policy-setting and ensuring consistency in the application of the EC competition rules. It was underlined that the new model will have a significant impact on the institutional status of the national competition authorities, which now become part of the EC decision-making sphere. The analysis of the EC regulatory networks operating in the telecom, gas and electricity sectors revealed above all quite a number of differences with respect to the context of the EC competition network, but also a number of interesting lessons to learn. In the telecom sector, EC regulation has traditionally been implemented by national regulators, and allowed ample room for discretion to the latter. The EC regulatory network co-exists with national regulatory networks, comprising the national telecom regulator and a variety of other interdependent actors. The situation at the national level is quite heterogeneous, given the differences between national telecom regulators in terms of institutional status, structure and powers, etc. The principal effect of these arrangements is that the Member States all followed a similar direction in the liberalization of the telecom sector, but with varying outcomes. Until recently, the only instruments used in order to ensure some degree of consistency throughout the EC were bench-marking and crosscountry comparisons. This regulatory approach has also generated an interesting phenomenon: the creation of new regulatory networks at the EC level (such as the High-Level Regulators Group, or the European Regulators Group for Electronic Communications) in order to deal with the coordination of the already-existing ones. This generated the need to strengthen the EC regulatory framework as a modality of reducing the margin of discretion for the national regulators: the 2002 EC telecom regulatory package contained a series of instruments in this sense.

xxii

Introduction

By contrast to the situation in the telecom sector, in the areas of electricity and gas there is less consensus among the EC Member States on the scope for liberalization, due to the inherent tension between the objectives of free competition and the need to ensure security and reliability of supply. The electricity and gas sectors were traditionally characterized by slower technological progress, and so far were less affected by globalization. The EU electricity and gas regulatory networks are composed of the EC institutions, the national regulators, the national competition authorities, and representatives of the market players. Their members meet twice a year, in the so-called Florence Forum (electricity) and Madrid Forum (gas) to discuss objectives and modalities of market liberalization. These innovative voluntary regulatory networks have so far been quite successful in encouraging the Member States to liberalize even further than the objectives established through EC directives. At the same time, they have some shortcomings, one of them being that they give to the incumbent monopolists the opportunity to slow down the liberalization process. The discussion on National Association of Attorneys-General (NAAG — the US infra-national antitrust network) brought to the surface quite a number of differences with respect to the European context. First and most significant, the NAAG is a network of public prosecutors who have no powers to decide on individual cases. In the NAAG there are no issues of case allocation similar to those that arise in the context of the EC competition network. Second, and related to the first, private actions are not only theoretically possible, but a reality in the US antitrust system, and therefore complainants do not have to rely almost exclusively on public enforcement, as in the EU. Third, in the US, federal antitrust law does not pre-empt state antitrust law, and issues of legal harmonization do not arise. The practical attributions of the NAAG are related to: a) conducting interstate investigations and the filing of inter-state actions (noting, however, that the Attorneys General are not authorised to exchange information protected by confidentiality or exchange information received from the federal antitrust authorities); b) the writing of amicus briefs to the courts; c) lobbying for legislation at the state and federal level; and d) the discussion of common antitrust policy guidelines. Having considered the differences between the NAAG and the EC competition network, there are however some important lessons to be drawn from the experience of the NAAG, mostly related to the importance of the human factor and what is usually called a ‘common competition culture’ for the effective functioning of the network. The second infra-national competition network discussed under this Panel was the German network composed of the Bundeskartellamt (the federal competition authority), the competition authorities of the Länder, and occasionally the Federal Ministry of Economics and Technology. As with the case of the EC competition network, the functioning of the German infra-national competition network is also characterised by an inherent tension between the objective of securing coherence and consistency in the application of the law and the need to safeguard the independence of the network members. However, this tension can be more easily

Introduction

xxiii

contrasted with the German system, first and foremost because the members of the network apply the same law, both substantially and procedurally, namely the provisions of the German Act Against Restraints of Competition. Second, coherence and consistency in the application of the law is ensured through judicial review of the decisions taken by the administrative authorities. The Federal Supreme Court and the High Courts of the Länder have specialised competition chambers. The German competition network observes strict criteria for the allocation of cases: competence to deal with an individual case is determined on the basis of the geographical impact of the infringement, and the Bundeskartellamt is competent to deal with cases having impact in more than one Länder. The Länder competition authorities are obliged to inform the Bundeskartellamt when starting proceedings, and conversely, the Bundeskartellamt must inform the authorities of the interested Länder when initiating cases against companies located on their territory. The German experience is relevant, only in part, to the EC context. The EC competition network could not adopt similar rigid rules on case allocation, as the reform creates a system of parallel enforcement competencies. Moreover, the EC competition network must take into account the variety of legal and institutional settings in which its members are called to operate. At the same time, the German experience calls attention to the importance of the judicial review system as a mechanism for ensuring coherence and consistency in the application of the law by the competition network, as well as to the idea of the competition network as a forum for debating conceptual and policy issues. Indeed, one of the most interesting aspects of the analysis of the German competition network is the spontaneous emergence of mechanisms for consultation among its members on specific policy and enforcement issues related to the liberalization of the energy sector. Complementary to the discussion on the context for the establishment of the EC competition network, the participants also examined the specific problems that arise from the interaction between competition enforcers and sectoral regulators. An analysis of the US experience in this domain shows that giving competence to apply competition rules to the sectoral regulators may sometimes lead to distortions in decision-making, either in the sense of adopting decisions which impose conditions that are too strict by competition policy standards, or approving agreements with clear anticompetitive effects, both as a result of allowing other economic policy considerations to prevail. Such distortions often gain an international dimension, in the perspective of globalization. An approach to solving such problems was suggested in a Report issued by the International Competition Advisory Committee of the US Department of Justice: to split the decision-making procedure, allowing the competition authorities to decide initially on the competition issues, and the regulatory agencies to decide on the other aspects at a later stage. This solution may not find political support in many jurisdictions, including the US, whereas a more feasible alternative would be to have the competition authorities decide on cases involving competition issues taking into account expert opinions submitted by the regulatory agencies.

xxiv

Introduction

Some of the most important conclusions that emerged from the discussions under this Panel centred on the subject of incentives to cooperate within the network. To put it in a nutshell, the effective functioning of EC competition depends to a great extent on the existence of adequate incentives for its members to cooperate. Several of the participants argued that the Commission’s reform proposal had not given sufficient weight to this aspect: in the absence of a day-to-day need to cooperate, the EC competition network would remain a legally-driven concept. It was generally accepted, however, that the reform would have a notable impact on the national competition authorities from this perspective. First, being empowered to apply EC competition rules in their entirety would strengthen the position of the national competition authorities vis-à-vis their national governments. Second, the reform expands the national competition authorities’ investigative powers. Third, the national competition authorities will draw direct benefits from the exchange of information within the network. Fourth, participation in the network will have an educational and support role for the competition authorities of the new member States from Central and Eastern Europe, which, although having made undeniable progress in the context of preparations for enlargement, are undoubtedly less familiar with the modes of cooperation and consultation on policy-making that had already been established among the Commission and the competition authorities of the older Member States. Last but not least, there was little doubt that the members of the EC competition network had by now achieved a certain ‘common competition culture’ level — in other words, a sharing of understandings, values and objectives — that is considered an important premise for the functioning of the network. Panel 2 examined in more depth some of the systemic issues specific to the EC competition network, as follows: a) whether the effectiveness of the network depends on a certain degree of homogeneity of its membership with respect to institutional status, structure, powers and resources; b) whether a certain degree of harmonization of national procedural rules was indispensable for the effectiveness of the network; c) the position and role of the Commission, and respectively of the national competition authorities, within the network; d) the institutional impact of participation in the network on its members; and e) the effects of participation in the network on the relationship between national competition authorities and their national governments. From a political science perspective, the question of whether the effectiveness of the EC competition network depends on some degree of homogeneity among its members, and in particular the national competition authorities, can be translated as follows: are the current institutional endowments and enforcement strategies of the national competition authorities adequate for the model of ‘voluntary cooperation’ among the members of the EC competition network envisaged by the Commission in its reform proposal? In terms of institutional endowments, the elements to be considered include: independence from the national governments, institutional status (including aspects such as the physical location of the authority, procedures for the recruitment of staff, the status of the head of the authority,

Introduction

xxv

etc) and resources (buildings, personnel, budget), enforcement powers (including both decision-making and competition advocacy) and constraints (dependence on the administration and judiciary for imposing and enforcing sanctions, whether the competition policy objectives are guaranteed by the constitution, ideological constraints, etc). The outcomes of such inquiry among the competition authorities of the EC Member States bring to the surface differences on several accounts. The most visible variations relate to institutional settings (i.e., while some Member States, like Germany or Italy, have integrated competition authorities, others, like Spain, France and Belgium, have bipolar competition enforcement systems, and in Ireland and Sweden the competition authorities have only prosecutorial functions, while competition cases are decided by courts), but examples arise on other accounts. Moreover, from a purely legal perspective, in the new enforcement system the national competition authorities will apply EC competition rules according to national procedures, and will apply the sanctions available in their national legal systems. The Commission’s reform proposal took into account the different structures of the competition systems in the Member States, and did not seek to impose elements of harmonization in this respect. For instance, the mechanisms stipulated in the reform proposal for allowing the Commission to secure consistency in the application of the EC competition rules by the national competition authorities respect the existing differences between the competition systems of the Member States. The obligation to inform the Commission before adopting a prohibition decision, as stipulated in Article 11(4) of the draft regulation, referred to the stage of statement of objections, which is prepared by the national competition authorities, and therefore did not interfere with the independence of the judiciary in systems where competition cases are decided by courts. As to the Commission’s power to take up cases according to Article 11(6) of the draft regulation, the participants discussed whether this provision should also be applicable with respect to courts in the jurisdictions that are deciding on the competition cases. The representative of the Commission argued that it would not be advisable to allow differences of approach among the Member States on this account. All participants agreed that participation in the network and the power to apply EC competition rules in full will increase the independence of national competition authorities with respect to their national governments, and will also implicitly produce in time a sort of ‘convergence’ impact in terms of their institutional endowments. Therefore, the decision of the Commission not to introduce in the reform proposal further elements of harmonization in this respect was considered adequate by most of the participants. Having accepted as a starting basis in the implementation of the reform the existing differences between national enforcement systems and the dependency of the members of the network on their national environments, the issue considered by the participants was whether there are modalities of enhancing cooperation within the network other than, and complementary to, the legal safeguards already envisaged in the reform proposal. From this perspective, the approach to

xxvi

Introduction

cooperation within the EC competition network proposed by the Commission finds an analogy in the broader EU notion of ‘open coordination’ of governance. To the political scientist, this notion describes a way of ensuring convergence of practices through a combination of peer review, surveillance, benchmarking and development of best practices, all methods that are fit for use in the context of the EC competition network. On the issue of whether the effective functioning of the EC competition network required a certain degree of harmonization of national procedures and sanctions, the views of the participants were varied. The debate centred on the question whether the reform proposal contained sufficient elements for ensuring the consistent application of EC competition rules in a system of parallel enforcement competencies relying on the procedural laws of the Member States. The representatives of the Commission argued that, at least during the early stages of implementation of the reform, it was not necessary to harmonize national procedures beyond a minimum necessary for ensuring the effective application of the EC competition rules within the network. In their view, the elements of harmonization contained in the reform proposal offered sufficient guarantees for a uniform application of EC competition rules by the national competition authorities. In particular, the obligation established by Article 3 of the draft regulation for the national competition authorities to apply EC competition rules to practices and agreements affecting trade between the Member States guaranteed a level playing field with respect to the body of law to be applied within the network. In addition, the reform proposal contained a number of provisions ensuring harmonization on the points essential for the functioning of the network, e.g. exchange of information (Article 12), the carrying out of investigations on behalf of another national competition authority (Article 22), the protection of information covered by an EC notion of professional secrecy (Article 27). Other participants emphasised, however, that at the time the Workshop took place, the outcome of the debate at the Council on the central tenet of the Commission’s argument, namely the provision establishing the primacy of EC competition law over the competition laws of the Member States (Article 3) was still uncertain. As it turned out, in the version of the procedural regulation eventually approved by the Council (Regulation 1/2003), the provision on the relationship between EC and national competition laws was modified to allow their parallel application while respecting the common minimum standards established by EC competition law (Article 3). In particular, the application of national competition law should not lead to the prohibition of practices or agreements not covered by Article 81(1) EC, covered by block exemption regulations or fulfilling the conditions for obtaining an individual exemption under Article 81(3) EC. At the same time, the Member States are allowed to apply stricter standards than those resulting from Article 82 EC on abuse of a dominant position. As anticipated at the Workshop, the Commission had become in time more flexible by comparison to its original position on the primacy of EC competition rules in the new enforcement system, accepting the possibility for the Member States to apply stricter national standards on abuse of dominance — as requested principally by the German representatives.

Introduction

xxvii

Some of the participants, while agreeing in principle with the Commission that full harmonization of national procedures was not indispensable at this stage, pointed to the fact that the practical implications of existing differences between national procedural systems had not been fully considered in the reform proposal. For example, procedural differences among the Member States related to the initiation and conclusion of proceedings (such as limitation periods), the ability of the national competition authorities to select only those cases they wished to deal with, or the nature of judicial review available internally with respect to the decision taken by the administrative authorities, may have a less than negligible impact both on cooperation within the EC competition network as on decisionmaking outputs. Most participants agreed, however, that the harmonization of national procedures raises issues that should be handled in a wider EU context, and not exclusively in the field of EC competition law enforcement. The debate on procedural harmonization should not delay the implementation of the reform, especially since the experience of the OECD working group on harmonization of international antitrust procedures, for example, had shown how difficult it is to reach agreement on these issues. Related to the question of harmonization of national procedures, the participants also discussed whether a certain degree of harmonization with respect to the availability, nature and severity of the sanctions for competition law infringements in the national systems was necessary for the functioning of the EC competition network. One of the problems discussed in this context was whether the availability of sanctions of a criminal law nature (such as fines or imprisonment applicable to the individuals responsible for the competition infringement) in some jurisdictions would raise problems for the exchange of information within the EC competition network, considering that in such cases the level of protection of defence rights must be higher than in administrative proceedings. In the Commission’s reform proposal, this issue was tackled by stipulating that the information exchanged within the network could only be used for the application of EC competition rules and of financial penalties (Article 12(2) of the draft regulation). During discussions, the idea emerged that the Commission’s reform proposal had perhaps been too conservative on this aspect, whereas it might have gone further and established that the information exchanged could also be used for applying sanctions to the individual responsible for the competition law infringement, as long as such sanctions were available in the legal system of both the requesting and the receiving competition authority. This suggestion eventually materialised in the final text of the new procedural regulation: Article 12(3) of Regulation 1/2003 stipulates that the information exchanged within the network can also be applied when imposing sanctions on natural persons, when the national laws of the transmitting and receiving authorities foresee such sanctions, or, if that is not the case, when the transmitting authority has respected the same level of protection of the rights of defence established by the legal system of the receiving authority. The debate on the position and role of the Commission, and respectively of the national competition authorities within the EC competition network, was quite

xxviii

Introduction

animated. Overall, there was a certain sense of contradiction between some critiques addressed to the Commission for arguably using the reform to strengthen its own powers, and fears that the Commission would withdraw from its enforcement and policy-making responsibilities. The Commission representatives emphasised that the reform was not about relieving the Commission of its workload and passing the burden to the national competition authorities, but represented a response to valid and increasing criticism about the shortcomings of the prior administrative authorization system. The Commission will continue to perform its role of law enforcer, policy-maker and guardian of the Treaty in the new enforcement system. On the other hand, the impression that the Commission intended to use the reform and the EC competition network in order to strengthen its own position and powers probably originated from an incomplete understanding of how the Commission intended to perform certain functions relating to maintaining consistency in the application of EC competition rules in the new system. For instance, in situations where matters of case allocation or re-allocation among national competition authorities arise, the Commission would act as a referee, intervening only if the discussions among the former did not lead to satisfactory solutions, and from the perspective of a moderator (in other words, without seeking to impose decisions). Likewise, the provision allowing the Commission to withdraw cases from the national competition authorities (Article 11(6) of the reform proposal, and respectively of Regulation 1/2003) is exclusively destined to allow the Commission to ensure a consistent application of the EC competition rules. The Commission intended to use this power only in situations where several national competition authorities are about to arrive at divergent conclusions on the same case/legal matter, or adopt decisions contradicting the existing EC case law, or where the individual case in question raises new and significant policy issues. Moreover, it would initiate proceedings only after consulting with the national competition authorities in question. It was also reminded that this mechanism also existed in the previous enforcement system, and was occasionally exercised by the Commission, always with the agreement of the national competition authority concerned. Not all participants were entirely convinced by these arguments. In their view, the fact that reform will considerably increase the number of cases in which national competition authorities apply Articles 81 or 82 EC should also change the perspective on the Commission’s already existent power to withdraw cases, generating a need to adopt specific rules framing the Commission’s modes of action. (Rules on the application of Article 11(6) in Regulation 1/2003 were eventually laid down in the Commission Notice on cooperation within the network, at paragraphs 50–57). Concerning the role of the national competition authorities in the network, the reform proposal was criticised for increasing the enforcement responsibilities of the former without allowing them to play a more active role with respect to policy-making. This situation was considered to be in contradiction with the very

Introduction

xxix

spirit of a network. The national competition authorities would perceive their purpose within the network was mainly to support the Commission’s performance of its tasks, and this would diminish their incentives to cooperate. To remedy this inbalance, some participants proposed to give the Advisory Committee a more active role to play in the network, in particular by empowering it to issue opinions on certain Commission decisions and on case allocation, as well as using it as a forum for consultation on block exemption regulations. As it turned out, these suggestions were assimilated and even extended in Regulation 1/2003 (Article 14) and the Commission Notice on cooperation within the network (paragraphs 58–68). The Notice establishes an obligation upon the Commission to consult the Advisory Committee on block exemption regulations, notices and guidelines (paragraphs 63–64). Panel 3 examined in more depth some of the technical problems related to the functioning of the EC competition network: a) issues of case allocation and reallocation among the members of the network; b) mechanisms of consultation among the network members; c) the scope for, and limits to, the exchange of information within the network; d) the compatibility of the reform proposal with the European Convention on Human Rights and the EU Charter of Fundamental Rights. The first group of issues discussed in this Panel concerned the allocation of cases between the members of the EC competition network. As already mentioned, the Commission’s reform proposal had covered in some detail the vertical relationship between the Commission and the national competition authorities, but was less explicit with regard to the horizontal relationship between the national competition authorities, including matters related to the allocation of individual cases among them. During the public debate following the publication of the reform proposal, requests for a clarification of the case allocation criteria and procedures emerged with some force from the national competition authorities as from the legal and business communities, against the more general background of demands for transparency and predictability of procedures and outcomes in the system of parallel enforcement competencies. At the time when the Workshop took place, the Commission had already shaped a proposal on the basic principles of case allocation between the national competition authorities — as presented by Commissioner Monti in his introductory intervention. The debate continued, however, on the legal instrument through which case allocation principles should be laid down. The representatives of the Commission argued that the criteria for and procedures on case allocation should not be included in the text of the new procedural regulation, but rather published in a Commission notice, which could be more easily amended on the basis of experience gained in the implementation of the reform. Other participants considered instead that a Commission notice would not be the most adequate legal instrument for establishing criteria on case allocation, because the national competition authorities had limited opportunities for influencing the substance of such notices. This problem was eventually overcome in the aftermath of the Workshop: following a proposal put forward by the German representatives, the

xxx

Introduction

general principles of case allocation were discussed with the Member States and published in a Joint Statement of the Council and the Commission (annex to Regulation 1/2003). In addition, paragraphs 63–64 of the Commission Notice on cooperation within the network foresee that, for the future, the national competition authorities will be consulted through the Advisory Committee on block exemption regulations, notices and guidelines. The participants also discussed whether undertakings should have the right to appeal decisions taken within the network on case allocation and/or re-allocation. Such decisions have different practical implications for the undertakings concerned according to their procedural position: while claimants are more interested in the procedural implications of the choice of the authority competent to deal with their case, the defendants are more interested in consequences related to the sanctions applicable. Therefore the issue of the legal status of decisions on case allocation should be considered in the wider perspective of the debate on harmonization of national procedures and sanctions. However, for all practical purposes, decisions on case allocation were considered by the majority of the participants to be informal acts of the network, not open to appeal by the undertakings. Also related to the subject of case allocation, the discussions uncovered some other interesting considerations. One is that in some national jurisdictions market substitutability is not taken into account in the definition of the relevant product market, and this may have significant implications in terms of case allocation within the network. The participants therefore recommended the adoption of a Commission notice on market definition, establishing minimum common standards in this domain. Also, the new enforcement system and the effects-based criteria for case allocation will reinforce cooperation at the regional level, creating ‘regional clusters’ of national competition authorities working in closer contact than with the rest of the members of the network. On the subject of consultations between the members of the EC competition network, the participants distinguished between consultations regarding the allocation of individual cases, consultations prior to decision-making, and requests for assistance in the gathering of information. The discussions focused on consultations between the national competition authorities prior to decision-making on individual cases, an aspect that had not been covered by the reform proposal, although the latter contained quite detailed provisions on ‘vertical’ mechanisms of consultation between the Commission and the national competition authorities. Such provisions included: the obligation for the national competition authorities to inform the Commission in cases involving the application of EC competition rules at the initial stage of the proceedings (Article 11 (3) of the draft regulation); the obligation to consult the Commission prior to the adoption of prohibition decisions, decisions accepting commitments or withdrawing the benefit of block exemption regulations (Article 11(4)); the possibility of consulting the Commission prior to the adoption of any other kind of decisions (Article 11(5)); and the consultation of the national competition authorities through the Advisory Committee on decisions to be taken by the Commission

Introduction

xxxi

(Article 14). The participants expressed different views regarding the legal effects of ‘horizontal’ consultations between the national competition authorities. Some considered that the outcome of such consultations should not be binding on the authorities dealing with the individual cases in question, inter alia because the reform proposal offered to the Commission the possibility of intervening so as to ensure consistency in the application of EC competition rules. Others argued instead that the outcomes of the consultation process could be assimilated to some form of ‘soft regulation’, in the sense that the national competition authorities concerned would find it difficult to disregard the conclusions emerging from the consultation process. Everybody agreed however that it was not necessary to insert in the procedural regulation specific provisions in this domain. In terms of the scope for, and limitations to, the exchange of information between the members of the EC competition network, the debate revolved around the following aspects: a) the type of information that may be exchanged within the network; b) the purposes for which the national competition authorities could use the information obtained through the exchange; c) whether the undertakings should have the possibility of challenging the exchange of information before it takes place; and d) obstacles to the exchange of information arising from differences between jurisdictions in terms of standards of protection of the defence rights. As to the type of information that may be exchanged within the network, the participants discussed first whether the existing differences between jurisdictions in terms of level of protection of confidential information could limit the scope of the exchange, given that the reform proposal had not envisaged harmonization in this respect. The representatives of the Commission argued that the reform proposal contained sufficient elements for ensuring the minimum necessary level of harmonization on the matter of protecting the interests of undertakings and individuals related to confidential information: Article 27 of the draft regulation established that the right of access to file did not apply to confidential information and internal documents of the Commission or national competition authorities, and Article 28 of the draft regulation prevented the members of the network from disclosing to the undertakings information protected by the obligation of professional secrecy that has been obtained through exchange. Other participants stressed however that these limitations applied only in relation to information exchanged on the vertical dimension, between the Commission and national competition authorities, and did not cover the exchange of information on the horizontal dimension, between the national competition authorities. Therefore some of the participants pleaded for a minimum level of harmonization between jurisdictions in terms of protection of confidentiality, or at least the establishment of a sort of ‘mutual recognition’ principle between national jurisdictions, according to which information classified as confidential in the legal system of the transmitting authority should be treated as such by the receiving authority. Also in relation to the exchange of confidential information within the network, some of the participants considered that the undertakings concerned should have a right to express their position with respect to the exchange before it

xxxii

Introduction

actually takes place. In the Dutch legal system, for example, the competition authority is obliged to inform the parties before transmitting information to another authority, although this obligation is without prejudice to the right of the competition authority to transmit confidential information. Other participants disagreed: establishing a right for the undertakings to appeal decisions to transmit information would render the exchange of information within the network ineffective, especially in cases where it was necessary to transmit information rapidly. The Commission Notice on cooperation within the network (at paragraph 27) clarifies that the rights of undertakings related to the exchange of information between the national competition authorities are governed by the national laws of the Member States, and in this perspective, establishes that the transmitting authority may inform the receiving authority as to whether the gathering of information is contested, or might still be contested. The participants also discussed whether differences between jurisdictions in terms of the investigative powers of the national competition authorities would impose limits on the exchange of information. The practical question addressed in this context was whether Article 12 of the draft regulation allowed the national competition authorities to use information of the kind obtained through exchange of the kind that they would not have been entitled to obtain in their own jurisdictions. This question is to be distinguished from the one whether the national competition authorities are entitled to use information obtained through exchange in order to impose sanctions that are not available in the legal system of the transmitting authority, in particular sanctions of a criminal law nature, to be imposed on the individuals responsible for the infringement of EC competition rules. With respect to the former question, the participants expressed different views. Some considered that Article 12 of the draft regulation did not impose any limitations other than those related to the nature of the sanctions to be imposed. Others argued that the national competition authorities should not be allowed to take advantage of the weaknesses of other legal systems by using information obtained from jurisdictions where the level of protection of defence rights was lower. As to the second question, Article 12 of the reform proposal had limited the application of information exchanged through the network to the application of financial penalties. This provision was considered by several participants to be unduly conservative, and indeed it was modified in the aftermath of the Workshop. Article 12(3) of Regulation 1/2003 allows the national competition authorities to use information obtained through exchange in order to impose sanctions on the individuals responsible for the infringement, with the exception of imprisonment, when such sanctions are available in the legal systems of both the transmitting and receiving authorities, or the information was collected respecting the same level of protection of the defence rights as in the legal system of the receiving authority. Another unclear aspect in the reform proposal was whether the information exchanged in the context of one case could be used by the receiving authority in another case, including where the latter involves the application of national

Introduction

xxxiii

competition rules. The participants emphasised that the reform proposal had not incorporated the principles emerging from the ECJ jurisprudence on this matter, namely that information gathered in the context of one proceeding could not be used as evidence in another proceeding, although it could be used as a reason to initiate new proceedings. Moreover, the ECJ jurisprudence was relevant only for circumstances relating to the exchange of information between the Commission and the national competition authorities. Indeed, Article 12 of the reform proposal stipulated only that the information exchanged within the network could be used exclusively for the purpose of applying EC competition rules — a provision to be interpreted together with Article 3, establishing the primacy of EC competition law in the new enforcement system. The adoption of Regulation 1/2003 and of the Commission notice on cooperation within the network sheds a new light upon this matter: Article 3 of the Regulation allows for the application of national competition law in parallel with EC competition law insofar as this does not lead to different outcomes, while Article 12(2) of the Regulation takes over the principles resulting from the Dow Benelux jurisprudence, namely that information exchanged can only be used in evidence for the application of EC competition rules and for the subject-matter for which it was collected. According to the same Article, taken over at paragraph 28(b) of the Commission Notice on cooperation within the network, the information obtained through the exchange may also be used for the purposes of applying national competition law on the same subject matter in parallel with EC competition rules. Under the same Panel, the participants sought to determine whether there were potential problem areas in reform proposal from the point of view of its compatibility with the provisions of the European Convention on Human Rights (ECHR) and the EU Charter of Fundamental Rights (CFR). The debate focused on the following main issues: a) the risk of ‘double jeopardy’, or in other words, of infringement of the right not to be prosecuted and sanctioned twice for the same offence (as established under Article 4 Protocol 7 to the ECHR and Article 50 CFR) in the case of parallel action by several network members; b) possible problems of incompatibility with the right to a fair trial (as established under Article 6 ECHR) and the right to respect for private life (as established under Article 8 ECHR and Article 7 CFR) arising in the context of the exchange of information within the network; c) for the national jurisdictions where decisions on competition cases are taken by courts, possible issues of incompatibility with the conditions of court impartiality established under Article 6 ECHR; d) problems related to the existence of different levels of protection of defence rights in the national jurisdictions; and e) possible problems of incompatibility with the principle of legality and predictability of criminal sanctions (established under Article 7 ECHR) emerging from the existing differences between jurisdictions in terms of procedure and sanctions. This overview suggested that the one aspect of the reform proposal creating a real risk of incompatibility with the provisions of the ECHR and CFR concerned the possibility of parallel action by several members of the network. Under the

xxxiv

Introduction

old enforcement system, the Walt Wilhelm jurisprudence had eliminated problems of incompatibility with the ECHR by establishing that, when the Commission and a national competition authority acted on the same infringement, there were no problems of incompatibility with the principle of ‘ne bis in idem’ (or the right not to be tried and punished twice for the same offence) as long as each authority applied a different law (i.e., EC competition rules in the case of the former, and national competition rules in the case of the latter). The Walt Wilhelm principle was, however, overtaken by recent developments in the jurisprudence of the European Court of Human Rights, which clarified that two proceedings would be considered to apply to the same offence where the essential elements of the latter coincide. Some participants considered that the principle of ‘ne bis in idem’ does not preclude the possibility of parallel action by several national competition authorities in the new enforcement system, because in such cases each of the national competition authorities would take into account the effects of the same infringement that are felt in their own jurisdiction. Others underlined, however, that the principle of ‘ne bis in idem’ regards the offence it its substance, and not only its effects. At any rate, it was suggested that the risk of double jeopardy could be eliminated without renouncing the possibility of parallel action by several national competition authorities during the investigation, as long as the decision was adopted by one single authority. As it turned out, Regulation 1/2003 and the Commission Notice on cooperation within the network allow parallel action by several national competition authorities, but do not contain a direct provision on decisions being taken by a single authority. The only guarantees in this respect are Recital 18 in the preamble to Regulation 1/2003, which mentions that each case should be handled by a single authority, and Recital 17 of the same preamble, which establishes that Regulation 1/2003b should be interpreted and applied in conformity with the principles resulting from the EU Charter of Fundamental Rights. Conclusions. The Workshop discussions produced a number of important conclusions on the nature of cooperation within the EC competition network and the impact of the reform at the level of national jurisdictions. The first is that participation in the network, coupled with the power to apply EC competition rules in full, will reinforce the independence of the national competition authorities and strengthen their position with respect to the national governments. The second relates to the importance of the human factor and the existence of an already-consolidated ‘common competition culture’ for good cooperation within the network. The third was the need to give to the national competition authorities a more active role to play in the shaping of EC competition policy, and in relation to this, the proposal to upgrade the attributions of the Advisory Committee within the network, so as to transform it into a forum for consultations among the network members on case allocation, on decisions to be adopted by the Commission, as well as on EC competition regulation (block exemption regulations, Commission guidelines and notices). The fourth concerned the harmonization of national procedures and sanctions: while everybody agreed that

Introduction

xxxv

the implementation of the reform should not be delayed by political debates on harmonization issues, it was nevertheless suggested that in the future a certain degree of convergence between the legal systems of the Member States in terms of competition law procedures and sanctions would be necessary. The fifth related to the importance of the judicial review system as a modality for ensuring consistency and coherence in the application of EC competition rules in the new enforcement system.

Isabela Atanasiu European University Institute, Florence June 2004

Panel Discussions

3

Panel One: The Network Concept, Competition Authority Networks and Other Regulatory Networks

䉴 KAREL VAN MIERT — On behalf of Claus-Dieter Ehlermann and myself, I would like to welcome you all at the eighth edition of the Annual Competition Workshop held at the European University Institute in Florence. It has become a tradition for the acting Commissioner for competition to participate at this workshop, and I would like to welcome again Commissioner Mario Monti. This year’s subject is particularly timely, as by now it is certain that the new regulation replacing Regulation 17/62 will be adopted sometime in the course of this year. The application of the new regulation will depend in good part on how the competition authorities of the Member States will ‘network’ among themselves and with the Commission. By the way, it was during Claus-Dieter Ehlermann’s mandate as Director General of DG IV (now DG Competition) that the Commission started to bring national competition authorities to Brussels a few times every year in order to have an exchange of views on competition developments at the national and European level. This sort of consultation and cooperation proved to be very effective. Yet the modernisation project goes further than that, it is about developing a real network of competition authorities, and establishing the rules of the game.

MARIO MONTI — I am delighted to be here once again, and I wish to thank Claus-Dieter Ehlermann for having organised this workshop and for offering me once again the privilege to open the proceedings with a few introductory remarks. Earlier editions of this workshop have been extremely useful contributions to the ongoing debate about the reform of Regulation 17/62. I am certain that this edition will continue the tradition. Three years after the publication of the Modernisation White Paper,1 the difficult debate of the beginning — when the business community feared that the Commission was retreating from an important European policy area — was replaced by intense consultations over the details of the reform proposal. Yet, as we well know, the devil lies in the detail, so there are still many open questions regarding the functionality of the future network of competition authorities. This is why this workshop is very timely. The draft regulation sets up a system of parallel enforcement competencies, in which a network of competition enforcers comprising the Commission and the competition authorities of the Member States will apply Articles 81 and 82 EC. The network will be a forum for the discussion and coordination of EC competition 䉴

1 European Commission (1999): White Paper on Modernisation of the Rules Implementing Articles 85

and 86 of the EC Treaty, OJ C 132 of 12.05.1999, pp. 1–47.

4

The Network Concept, Competition Authority Networks

policy, and a basis for the maintenance and development of a common European competition culture. Since our previous meeting, which was devoted to the modernisation initiative,2 ‘the network concept developed a dynamic of its own’ — to borrow an expression from Claus-Dieter Ehlermann. Indeed, the network concept has developed on several dimensions. First, the concept was closely examined in the context of the discussions at the Council on the draft regulation3 during the Belgian and Spanish Presidencies. Secondly, in 2001 the European national competition authorities established ECA,4 which is an informal forum for discussing the application of competition rules and how to improve the cooperation among them. Thirdly, again in 2001, representatives of US, EU and other international competition authorities established the International Competition Network, a forum for the debate of international antitrust enforcement issues.5 Finally, I believe that the conclusions of this workshop will be of particular interest, not in the least because the network concept will be discussed in a real interdisciplinary exercise involving the academia. The purpose of my opening exposé is to peg out the Commission’s conceptual view of the network. There are four main issues that need to be discussed: 1. Why do we need a network of public enforcers? 2. How to divide the work between the members of the network? 3. How can we ensure consistent application of the Treaty rules on competition? 4. Should the reform lead to a harmonisation of national procedures and sanctions? Why do we need a network of public enforcers? The Modernisation White Paper put forward the concept of a network of public enforcers as one of the key features in the new enforcement system. However, the network will not be established ex novo by the new regulation, but will build upon the already existing links between the Commission and the national competition authorities of the Member States (NCAs). After the adoption of Regulation 17/62, the Member States which did not have a competition law enforcement system of their own

2 Sixth Annual EU Competition Law and Policy Workshop: The Modernisation of EC Antitrust Policy, 2–3 June 2000, EUI Florence. See C-D. Ehlermann and I. Atanasiu, eds. (2001): European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, Oxford and Portland, Oregon. 3 European Commission (2000): Proposal for a Council Regulation on the Implementation of the Rules on Competition Laid Down in Articles 81 and 82 of the Treaty and Amending Regulations (EEC) No 1017/68, (EEC) No 2988/74, (EEC) No 4056/86 and (EEC) No 3975/87 (‘Regulation Implementing Articles 81 and 82 of the Treaty’), COM(2000) 582 final, OJ C 365E/284 [2000]. 4 The association of competition authorities operating within the EEA (European Economic Area). 5 For details about the International Competition Network, see http://www.internationalcompetition network.org/about_hx.html.

Panel Discussions

5

have set up one. The competition authorities of the Member States apply national competition law, and by now most of them are closely involved in the application of EC competition law. In particular, NCAs advise the Commission on solving individual cases, and, within the framework of the Advisory Committee, they are consulted on the adoption of block exemption regulations. The competition authorities of some Member States have even been empowered to exchange information and cooperate between themselves in the solving of individual cases. These forms of consultation and cooperation will constitute the basis for the set-up of the future EC network of competition authorities. In the new enforcement system, the network will maintain its informal character. At the same time, a number of elements necessary for enhancing cooperation will be put in place. These include: the power for all members of the network to apply EC competition law; the power to exchange confidential information, to share the work and assist each other in the solving of individual cases. The network will be at the centre of the public enforcement pillar in the new system. It will have a three-fold purpose: 1) to promote the effective enforcement of EC competition rules by ensuring that the enforcement resources available are used optimally; 2) to contribute to the consistent application of EC competition rules; and 3) to participate in the development of EC competition policy. In a system where public enforcement of the rules is shared by several authorities, it is essential to ensure that the rules are not only applied consistently, but also that these authorities share the same competition culture. The network will contribute to that. How to divide the work between the members of the network? This is certainly not an easy issue, at least apparently. The cornerstone of the draft regulation is the creation of a system of parallel enforcement competencies. Unlike the merger regulation,6 which builds spheres of exclusive enforcement competencies, the draft regulation envisages full competence to apply the EC competition rules for all the enforcers. This raises the question of how to divide the work between the members of the network. In order to protect competition efficiently, and to allocate enforcement resources properly, individual cases have to be dealt with by the authority best placed to solve the competition problem identified. In order to enable the system to work swiftly and effectively, appropriate criteria for the allocation of cases will have to be defined and made public. This will guide complainants as to where to lodge their complaints and assist the authorities themselves in determining who should deal with a given case. We have to do here with an enforcement system that combines two principal features: it is a system of parallel enforcement competencies, and it is a prohibition system. In a system of parallel enforcement competencies, case allocation 6 See Council Regulation (EEC) No 4064/89 of 21 December 1989 on the control of concentrations between undertakings, OJ L 395, 30.12.1989, pp. 1–12 and Council Regulation (EC) No 1310/97 of 30 June 1997 amending Regulation (EEC) No 4064/89 on the control of concentrations between undertakings, OJ L 180, 09.07.1997, pp. 1–6.

6

The Network Concept, Competition Authority Networks

criteria have to be flexible and of a qualitative nature. In a prohibition system, they have to be based on considerations of enforcement efficiency. For these reasons, the case allocation criteria should not be included in the new regulation, but published in a notice that can be updated in a flexible way as often as necessary. It goes without saying that the Member States will be closely involved in the drafting and the revision of such a notice. However, I do understand the concerns expressed by Dr Ulf Böge in his written contribution for this workshop vis-à-vis this procedure, and I agree that the basic principles on the functioning of the network could be enshrined in a joint statement by the Council and the Commission. Among the elements to be taken into account for the allocation of cases, the following are of most relevance: the scope of the suspected infringement, the effects of the decision required to put an end to the infringement, and the location of evidence. In order to avoid enforcement lacunae, it is also important to take into account the authorities’ resources and enforcement priorities. The place where the effects of the infringement are felt is a primary criterion for any intervention by a public body: competition authorities always intervene in order to protect competition in their own jurisdiction — that is, to protect the consumers in their own jurisdiction from the effects of anticompetitive behaviour. The suspected infringement must therefore have detrimental effects within the jurisdiction of the acting competition authority. Furthermore, in order to reduce the number of authorities competent to deal with a given case, only the direct effects of the suspected infringement should be taken into account. This means, for example, that in the case of foreclosure of a national market by a dominant company, the loss suffered by a competitor outside the foreclosure market should not be considered for the purpose of case allocation — the authority in whose jurisdiction the competitor is located should not become active on the case, since it is not its market to be affected by the behaviour of the dominant company, but the foreclosed market. The authority of the latter market should act in order to solve the case. The second criterion for the allocation of cases, i.e., that of the effect of the decision bringing an end to the infringement, will generally point towards the same authority as the first. For example, a cease-and-desist order will have to be enforceable in the market where the infringement produces its effects. Finally, given that the new enforcement system will rely on prohibitions, it is important for the competition authorities to be able to collect the evidence necessary to prove infringements. Unlike in the ambit of the prior authorisation system, where companies have an interest in bringing the available evidence before the competition authority, prohibition proceedings require investigative powers that allow the competition authorities to collect information even without the cooperation of the companies. The draft regulation contains an important provision designed to facilitate the collection of evidence by NCAs: Article 21 enables them to assist each other in fact-finding. This means that an NCA can carry out fact-finding proceedings on behalf of another NCA and transmit to it the evidence thus collected.

Panel Discussions

7

The application of these three criteria will generally lead to the following conclusion: the NCA best placed to handle an individual case is the one of the Member State where the main effects of the infringement are felt. The same authority may also be able to deal with cases having effects in other Member States, particularly if the prohibition of an anticompetitive agreement in one Member State leads companies to abandon the practice as a whole. The Commission, on the other hand, will be well placed to deal with infringements having direct effects in several Member States, say more than three. As a guardian of the Treaty, and the Community institution entrusted by the Treaty with the task of defining EC competition policy, the Commission will also have to deal with certain cases in order to set Community policy and ensure consistency in the application of the EC competition rules. The remaining sensitive issue is the allocation of individual cases producing effects in more than one, but less than three Member States. The business community has expressed preference for a ‘one-stop shop’ principle for the allocation of such cases, that is, for the allocation of each case to a single NCA. This is certainly desirable: as often as possible, cases would have to be dealt with by a single authority. However, we should not exclude parallel action by NCAs as a matter of principle. In an enlarged Community of up to 28 Member States it might sometimes be necessary to prosecute regional cartels through the coordinated action of several NCAs. This would obviously require close cooperation between the NCAs, and appropriate safeguards destined to ensure consistency both in the course of proceedings and with respect to the outcome of cases. At any rate, not all individual cases producing effects in more than one Member State should be systematically referred to the Commission. In practice the allocation of cases will be automatic: complainants will bring their case to the NCA considered best placed to deal with the matter, and in most cases that authority will effectively deal with the case. The network should not be seen as a clearing house for each and every individual case: consultations within its framework will take place only when a matter of case re-allocation arises, that is, when the case was not directed to the right authority in the first place. Not being an institution, the network will not take decisions. Matters of case allocation will be discussed informally between the members of the network directly concerned. Only in rare cases, when no agreement could be found, would the Commission have to intervene based on Article 11(5) or (6) of the draft regulation. To answer the important question raised by John Fingleton in his written contribution for this Workshop, I think that case allocation should not, and will not, be decided through a formal decision that could afterwards be challenged by undertakings. Undertakings do not need to challenge decisions on case allocation, because all members of the network apply the same rules, namely EC competition rules, and respect a minimum standard of defence rights. The allocation of a case between the members of the network does not pose problems for the undertakings when compared to the present system of parallel application of national and EC law.

8

The Network Concept, Competition Authority Networks

Another objective of the network is to ensure a consistent application of the EC competition rules. To that effect, Article 11(4) of the draft regulation provides for a consultation mechanism between the Commission and the NCAs, namely, before adopting certain types of decisions, the latter will have to consult with the Commission. The other members of the network will also be informed and have the opportunity to discuss and comment on the case. If need be, a meeting of the Advisory Committee can be convened. In the new enforcement system, consistency in the application of the EC rules will be the daily bread and butter of all the members of the network. In most cases bilateral contacts between the members of the network and informal discussions will be sufficient. However, it cannot be excluded that, despite intense discussion, sometimes there will also be disagreements. For these hopefully rare cases, the draft regulation foresees the possibility for the Commission to take over cases from the NCAs. Article 11(6) thus creates the necessary safety valve for the system. The Commission will apply this provision when two NCAs act in parallel on a case and arrive at different conclusions, when an NCA envisages taking a decision that contradicts the caselaw, or when the case raises an important Community policy issue requiring clarification. In such cases it is indeed more effective for the Commission to adopt the policy line rather than running the risk that several NCAs develop divergent policy lines. Normally, such cases will be identified at the beginning of proceedings, and the Commission will take them up at an initial stage. It cannot be excluded though that the importance of a case or an issue be perceived only at a later stage. Yet such situations will no doubt be exceptional. In the context of discussions on the draft regulation the Commission has been warned about the heavy administrative burden resulting from the task of maintaining consistency in the application of EC competition rules. It has even been argued that it would be more demanding to perform such a task, in terms of time and resources, than maintaining the current notification system. I must say we are rather optimistic on this point: in most cases there will hardly be any need for the Commission to intervene for ensuring consistency. To begin with, the adoption of decisions prohibiting hard core infringements does not require an important effort of coordination at the level of the network, as it involves rather a careful examination of the facts than sophisticated legal analysis. As for the more complex cases, requiring coordination, it is obvious that time and resources would be better spent on such cases rather than on reviewing notifications of harmless agreements. Finally, should the reform lead to the harmonisation of national antitrust procedures and sanctions? Article 3 of the draft regulation, which seeks to ensure a level playing field by establishing the primacy of EC competition law in cases where trade between the Member States is affected, was welcomed by most commentators and by the European Parliament. However, some commentators also wondered why the Commission had not gone further, and had not provided for a fully-fledged harmonisation of national procedures and sanctions. It was pointed out in particular that it would be unfair for the undertakings to be exposed to different sanctions for the same infringement depending on which competition

Panel Discussions

9

authority they would deal with. In addition, the rights conferred to undertakings under national procedures differ from one EC Member State to another. The application of one set of substantive rules, i.e., the EC competition rules, brings into light the already existent differences between the EC Member States in terms of national procedures and sanctions. If one is allowed to make a comparison, this reform is a bit like the introduction of the Euro, which renders more evident the already existing, but so far little noticed, distortions in the functioning of the Single Market. The Commission did not propose a complete harmonisation of national procedures and sanctions because it is not absolutely necessary for the functioning of the new enforcement system. The reform will harmonise only those elements that we believe are crucial to the functioning of the new enforcement system. For example, in order to allocate cases within the network, all NCAs should be able to reject a complaint on the ground that another NCA has dealt, or is dealing, with the case. Without such a provision, the complainants would be able to oppose the allocation of cases to the NCAs of those EC Member States where they do not have a right to obtain a decision on the substance of their complaint. Another example of this extremely careful and sparing research for the minimum necessary in terms of harmonisation is the mutual assistance that NCAs can provide to each other in order to deal with cases where some of the evidence is located outside their jurisdiction. It was important to ensure that all NCAs are entitled to carry out investigations on behalf of another member of the network and to exchange the information resulting from their investigations. In sum, the draft regulation contains some elements of harmonisation, but it was not considered that the functioning of the new enforcement system required going beyond that. The members of the network can cooperate even if applying different national procedures and sanctions as long as they protect and respect the fundamental rights of the undertakings. As far as sanctions are concerned, differences between national systems will certainly not prevent NCAs from cooperating, nor will they lead to violating the rights of the undertakings. I agree with the view expressed by Margaret Bloom in her written contribution for this workshop, that there would be strong arguments for reducing the existent differences in terms of national procedures and sanctions, but before taking any steps towards approximating national laws in this respect we need to accumulate some experience with the functioning of the new enforcement system. The Commission will study this issue very carefully once the new enforcement system is in place. Constructing a network of competition authorities applying EC competition rules could have been seen as technical exercise, but neither the Commission, in putting forward the reform proposal, nor the European Parliament and the EC Member States, in debating it, took that view. Setting up the network implies creating closer links between the EC Member States and the Commission in order to achieve a common goal: the protection of competition on the Single Market. The network will be an example of close cooperation between administrations in Europe in order to achieve a goal set out by the Treaty. Each NCA, while maintaining its traditions and legal cultural background, will become part of an enforcement

10

The Network Concept, Competition Authority Networks

mechanism that requires a high degree of coordination and cooperation. In this sense, this reform brings about a new form of government in the area of EC competition law enforcement and beyond that. I am convinced that this reform, and the EC network of competition authorities, will be a source of inspiration for other areas of Community policy. Indeed, in the White Paper on European Governance7 the Commission did refer to the modernisation initiative and to the creation of the network. Just as in the 1950s and 1960s competition policy served in establishing the first model of European integration, today the modernisation initiative and the creation of the EC network of competition authorities will set another example in a new historical phase of the European integration. As the Community becomes broader geographically, national and local traditions and sensitivities are given more and more importance. At the same time, the need for ensuring a smooth and undistorted functioning of the Single Market becomes more acute. Perhaps the modernisation initiative and the network will provide an example of how to combine coherence with decentralisation. KAREL VAN MIERT — I was particularly pleased with what you said at the end of this introductory exposé. Indeed, from the times when I was still in DG Competition (back then DG IV) we already felt the need to reconsider in a critical manner the functioning of the system, particularly in view of the upcoming enlargement of the EU. By then the NCAs were consolidating their role in the application of EC competition rules, and most of them were already very effective, so the need to reform the system became evident. To a certain extent, this project can be seen as an example of reform initiative coming from the centre itself. The challenge of the reform is to develop a balanced enforcement system, ensuring coherence and effectiveness in the application of the law on one hand, and on the other, enhancing the role of NCAs through the set-up of a network. Obviously, we knew from the very beginning that the reform was going to trigger debate on a number of sensitive issues, such as coherence in the application of EC rules, the allocation of cases, etc. However, like Professor Monti said before, the Commission has made every effort to discuss these issues with all the parties interested: NCAs, legal practitioners, the business community, and academics. Still, some further problems may arise when the new enforcement system will be in place, but the idea is to allow sufficient flexibility for eventual adjustments.



MARIO SIRAGUSA — I have a question for Professor Monti related to Article 3 of the draft regulation, which obliges the NCAs to apply EC competition rules in cases where trade among the Member States is affected. Like many others, I consider this provision to be the pillar of the new enforcement system, and therefore I am very worried that discussions on the draft regulation might water down



7 ‘European Governance — A White paper’, COM(2001) 428 final, available at http://www.europa. eu.int/comm/ governance/white_paper/index_en.htm.

Panel Discussions

11

this provision. Could you report on the state of play in the discussion on this aspect? 䉴 MARIO MONTI — This is clearly a fundamental point, in particular from the perspective of undertakings, who want to be able to enter into agreements that may have an impact on trade between the Member States without running into the problem of being obliged to comply with different national competition law standards. This is the main purpose of Article 3 in the draft regulation, and we (the Commission) are not prepared to compromise it. At the Council meeting of December 2001 a large number of national delegations have endorsed this fundamental objective, and therefore we are confident that we move in the direction of a satisfactory solution. But, as you know, the situation is more nuanced with regard to the treatment of abuse of a dominant position under Article 82 EC. The prohibition of an abuse of a dominant position in one jurisdiction would, as a general rule, have less effect on the overall business strategy of an undertaking. Moreover, the scope of a prohibition decision will often be local or regional, as opposed to the case of agreements between undertakings. Therefore the possibility for the Member States to apply stricter standards as resulting from their national law should not be excluded. This will allow the Member States to maintain a range of national provisions that address specific issues of unilateral behaviour, which so far have not been addressed by Article 82 EC. There are also other aspects under discussion, in the Working Group and in the Council, concerning Article 3 of the draft regulation. We (the Commission) are open to suggestions of improvements, particularly so in relation to Article 82 EC, as mentioned before. At the same time, personally I have seen the importance that companies and their associations attribute to this pillar of the reform. Also the European Parliament, in expressing its favourable opinion on the draft regulation, has stressed the importance of Article 3. Therefore I am confident that it is possible to reach an agreement at the Council on Article 3 before the end of this year. 䉴 MARK THATCHER — Commissioner Monti, you presented a very consensual picture of the reform. Yet I wonder how realistic this picture is — as a political scientist, I am interested in issues of power, conflict and organisational roles. For instance, will the NCAs be happy to see the Commission take away their more interesting cases? Will companies be happy to realise that case allocation affects their rights, the kind of sanctions to which they are exposed, and the speed and efficiency with which their case is dealt with? How would the Commission react when there are conflicts of views between it and NCAs? 䉴 MARIO MONTI — In my view, your questions arise from a comparison between the present and the future enforcement systems. Or rather, analysts tend to evaluate the future enforcement system against a theoretical model whereby the powers of each player (enforcer) cannot be limited by the others. I think that, in a decentralisation exercise of this kind — i.e., which is, as economists would say,

12

The Network Concept, Competition Authority Networks

‘supply-pushed’ by the Commission, and not ‘demand-pulled’ by the national enforcers — this kind of intriguing questions would arise inevitably. It is however my deep conviction that the power of the NCAs, and thereby their ability to deal with interesting cases, will emerge as considerably consolidated from this exercise. As far as companies are concerned, I am confident that they will soon have confirmation that the new enforcement system offers consistency in the application of the EC competition rules and prompter, less costly enforcement proceedings. DAVID GERBER — My task today is to set the background of the debate by looking at the historical development of the EC competition enforcement system. A network is, in the end, a set of personal relationships, and most personal relationships are built over time. So we need to look at how the relationship between the members of the future network has developed over time in order to understand its future functioning. There has always been a network of some kind, and the question is what kind of changes have taken place in this network, and what kind of changes are envisioned for the future. By looking at the network from this perspective I think we may get some insights that otherwise are not so easily revealed. As to the analytical framework, I focus on the relationship among the competition authorities. In other circumstances I would have included the courts in the network concept, but my task here is to look at the network as defined by the Commission, that is, a network of public enforcers. One of the problems with this definition is that it is static, it does not tell us anything about the dynamic of the relationships within the network. This drawback can be remedied, however, by focusing on how decisions are taken. The question to be addressed is how the individual relationships among the members of the network affect the decision-making process. From this perspective it becomes useful to consider the legal systems that are involved in the decision-making process. Here we need to distinguish between the EU sphere of decision-making, which applies EC competition rules, and the national spheres of decision-making, which is related to the application of the Member States’ competition laws. The next step is to ask what is the structure of these distinct spheres, how do they operate, and how do they fit together. When we talk about the interaction between these decision-making spheres, I think that we should make another important distinction, namely between the transmission of information within the network and the normative effects of this interaction — in other words, what passes through the network, and what sort of effects this circulation has. All this considered, I believe we can talk about the EC network evolving in three stages, or through three models. From the 1960s to the mid 1980s, the network was based on a foundational model, in which the EC and the national decision-making spheres were clearly separated. The EC sphere involved mainly DG IV of the European Commission, although occasionally representatives of the Member States were brought in for particular reasons and based on particular kinds of expertise. The sphere of EC decision-making was very formal, except



Panel Discussions

13

for these specific occasions when national representatives were brought in. The information was transmitted as a matter of personal communication. There was very little interaction between the EC and the respective Member States’ spheres of decision-making. The authorities of the Member States had to take into account what the Commission did, but that was a rather formal sort of interaction. There was some less formal interaction at the meetings of the Advisory Committees, but these meetings were quite ‘didactic’, in the sense that the Commission used them to ‘teach’ the NCAs. This model began to break down in the mid-1980s, and in the 1990s, and it turned into what I would call a ‘solar model’. Here there is a central (EU) sphere of operation, and everybody else (the authorities of the Member States) must follow. In this model, the authorities of the Member States have to give information to the centre, but do not receive much information back, and they certainly do not receive much in the way of normative influence. The role of the Member States’ authorities is essentially passive. In this model, the EU sphere of decision-making is, of course, much larger than the spheres of the Member States. The network has more members — more people in the Member State authorities belong to it, and more of them do things related to the application of EU law. The process of transmission is still very limited and formal, which makes quite a contrast to what is envisaged for the future network. The Member States’ spheres of decisionmaking have grown in size and importance, but at the same time, the national systems have adapted to the EU rules and way of doing things. The institutions do have a relationship, but the sphere of decision-making is still quite primitive for most purposes, even though there is community of interests. The modernisation project sets up what I call a ‘centralised interactive model’. I think of this model as a large transatlantic ship, moving slowly but in a determinate direction. Decisions are taken in a group, but there is still scope for guidance, and the Commission still steers the wheel. The network is, in this respect, a reinforcement of this feature. The role of the authorities of the Member States will, of course, become more prominent. One of the interesting questions is how the structure of the network will evolve, considering the differences between the authorities of the existent and new Member States. As to the spheres of decision-making, assuming that Article 3 in the draft regulation goes through as proposed, EU competition law is the main sphere, and the Member States’ authorities become part of the EU sphere. This also means that the institutional structure of the Member State authorities, their budgetary limits, relationships with their governments, etc., will play a role with respect to the application of EC competition law, which was not the case. The institutional status factor will become increasingly important, especially with regard to the eastward enlargement. The fact is that some people know more, have more power and experience, and so on. Once again, there will be one single sphere of operation — the EU — but undoubtedly some sort of regional and other kinds of distinctions are going to develop. The network will be based on a more informal communication of information and ideas, so quite naturally language, culture

14

The Network Concept, Competition Authority Networks

and regional considerations will play an increasing role in the development of EC law. The transmission of information through the network becomes now multidirectional, and this can lead to conflict. We need to think of this in decisional analysis terms: up until now most of the information that has been passed within the network referred to specific needs. By contrast, in the new system a considerable amount of information will pass around without anyone really knowing how it is being used. I think that this aspect has not been given sufficient consideration. Much of the information to be transmitted through the network will not be targeted to a specific recipient or destined to a specific operational purpose, but will be sorted and put into a database. This changes the role of the information and its transmission in ways that need to be considered. Another aspect to think about is the fact that some basic characteristics of the network have existed for a long time, but many are changing fairly dramatically. As the decisional spheres integrate, national institutions become part of the EU decision-making structure. Now, we are told that the network is not going to be a decision-making institution, as it will not have the power to act in itself. Yet there will be expectations as to its role, and thereby pressure with respect to decision-making outcomes. I think this is a fairly significant change. Even if decisions will be made by the individual members of the network, each of these decisions will constitute part of an outcome process, so we will want to know who is doing what and under what kind of external influence. Also in terms of conditioning factors, the degree of participation of the members in the network and the organisation of the information transmitted become critical elements for the functioning of the network. I outlined this in my paper, and Dr Böge has also pointed this out with respect to some specific areas of EC competition law enforcement. STEPHEN WILKS — My paper is designed to bring in the perspective of the political scientist on the EC network of competition authorities. In this sense, it is rather broad, and less technical than a number of papers prepared for this Workshop. It is also based on a much broader definition of the network than the one proposed by the European Commission: I look at policy networks in a broader sense, and not only at the rather special and unique network that we are talking about here. There were five issues that I wanted to touch upon in this presentation. First, in the definition of policy network I stress the importance of the interdependence criterion, like my colleague David Gerber did before. Network members will have to be able to depend one on the other; if they are autonomous, the network will not function. Also as part of the definition, I distinguish between policy-making and policy implementation. The second question that I addressed in my paper is whether such policy networks do exist already in the EU. I argue that policy-making networks do exist: they are elite, small and innovative, and political science theorists call them



Panel Discussions

15

‘epistemic communities’. (The concept is discussed in my paper for this workshop.) The challenge, however, is to turn such small, elite policy-making networks into policy implementation networks. In my paper I argue that that it will actually be very difficult to create an effective policy implementation network. Thirdly, and as a consequence of this, I ask what would be necessary in order to create robust policy implementation networks in Europe. Political scientists believe that overall networks emerge from interdependency, and they tend to emerge informally. If the process is formal, legislated, then the outcome is a hierarchy, and hierarchies and networks have actually rather different modes of coordination. So, what sort of characteristics would one expect to see in a robust policy implementation network? The first is the realisation of self-interest: every actor has to something to obtain from participating in the network, even if this involves making some compromise. The second is a reduction in transaction costs — and I noticed that some of the other participants at this roundatble think that transaction costs may actually increase, rather than decrease, as a result of the set-up of the network. The third is something that Professor Ehlermann has stressed in his earlier comments on the Modernisation White Paper:8 familiarity, which breeds trust among the members of the network. The last is closure: networks are defined by who is excluded and who is included. Fourthly, one aspect that is easily overlooked, although absolutely crucial for the development of the network, is what the Commission calls a ‘competition culture’. This involves the consolidation of shared understandings, methods of working, and a common language — or what social scientists call a ‘discourse’ — in this policy area. However, as my colleague Professor Sturm points out in his paper for this workshop, this is not easy to achieve: in Europe there are different national traditions and very different conceptualisations of the goals of competition policy. Yet it is fundamentally important to come to some agreement across the network about the normative basis on which it will operate. Finally, networks are political, as our Chairman rightly said before. They define who has influence — which is why political scientists study them. From this point of view, a powerful, robust network, could also be dangerous: it could become elitist, exclusionary, even closed to new ideas and critical perspectives. The process of creating the network is intensely political, and our discussion today and tomorrow will bring out some of the tensions that this network proposal will inevitably generate. 䉴 BARRY HAWK — My task here is to consider what lessons there are to be drawn for the EC network of competition authorities from the functioning of the US NAAG. This ugly acronym stands for the National Association of Attorneys General, the US inter-state antitrust network. After having researched the subject and read most of the papers prepared for this workshop, frankly, I am not sure

8 Ehlermann

C. D. (2000): ‘The Modernization of EC Antitrust Policy: A Legal and Cultural Revolution’, 37 Common Market Law Review 3, pp. 537–590.

16

The Network Concept, Competition Authority Networks

that the NAAG can be relevant to the European network. But perhaps I am wrong. US Attorneys General are prosecutors; they are the chief legal enforcers in the States. So they are not administrative authorities, and do not decide cases. This is a fundamental difference with respect to the European network, and this difference means that a lot of the issues that are topical here do not apply to the NAAG. The NAAG has over 50 members. It is a large group, and there is no participation closure, but importantly, there is no forced participation either. The NAAG has various Committees; here we are interested only in the Antitrust Committee. In the US, federal antitrust enforcement is tied in with State antitrust law enforcement, because the States have the power to enforce both federal and State antitrust law. The NAAG became very visible in the 1980s, when the States felt that the federal antitrust authority had got too relaxed, so they decided to get tougher. Since then the States have remained active in this area of law enforcement, but there is no longer antagonism between the federal government and the States — to the contrary, they actually perform joint investigations. What does the NAAG actually do? There are four general areas of activity. First, they conduct multi-State investigations and file multi-State actions. So, if Pennsylvania, Iowa and California, for example, think that somebody is fixing the prices of shoes, and they could better prove their case by getting together, they do so. Sometimes they even get the Department of Justice or one of its regional offices to join in. Secondly, they write amicus curiae briefs to courts. (I do not know how important this activity is, since in the US everybody can file an amicus brief in a court.) Thirdly, they lobby for legislation, both at the State level and in Washington. Fourthly, they prepare antitrust policy guidelines (e.g., merger guidelines, vertical restraints guidelines — the latter were issued in the mid-1980s, in conflict with the federal government’s vertical restraints guidelines; and were never really enforced). Among the four types of activities of the NAAG, multi-State enforcement is by far the most important. Let us take the example of a price-fixing case: one State will issue a subpoena or civil investigative demand and the response will be shared with the other States that are investigating the case. If enough proof is gathered, the States choose a court where to bring their lawsuit. Sometimes the federal government comes in as well. This sort of multi-State investigations, with or without the participation of the federal government, has been very successful. Now, there are significant differences between the US and the European antitrust enforcement systems. First, the attorneys general are prosecutors, and not administrative authorities. They do not decide cases, they only prosecute them. Secondly, the courts have no involvement in the functioning of the NAAG. Thirdly, private actions are not just a theoretical possibility, but a reality: private complainants can obtain damages and injunctive relief. Fourthly, issues like case selection and allocation do not arise in the US. Complainants do not have to rely on State authorities to solve their complaint, they can bring an action before a State or federal court. Fifthly, federal antitrust law does not pre-empt State antitrust law. The harmonisation of State antitrust law is not an issue — to the

Panel Discussions

17

contrary, we think that the more sets of rules, the better. There is no equivalent of the European debate about harmonisation and consistency in the application of the law. We just learned to live with this ‘legal anarchy’. What is then to be learned from the functioning of the NAAG? Possibly only the human factor: you want to bring a bunch of people together in a network, and you want them to respect each other and be open-minded. I mentioned before that in the 1980s there was disagreement between the States and the federal government on the approach to the application of antitrust law. The States were accusing the federal government of withdrawing from its task of protecting consumers, etc. When James Rill came in at the Department of Justice in 1988, without changing significantly the federal antitrust policy, in just three months all the hostility dissipated. He just sat down and talked with the State Attorneys General, made them liaise with the ABA Antitrust Section. They became a respected part of ‘the establishment’, or however you may want to put it. The result, as I said before, was that in the 1990s we had more joint multi-State and federal investigations. This is the importance of the human factor. I also believe that in areas like cartel enforcement it is in the interest of State Attorneys General to cooperate. The Microsoft case is a wonderful example of federal-State cooperation in enforcement: the Department of Justice and 19 States worked together for the prosecution in this case. Now, when the case was brought before a court, that became a different story … . The Department of Justice and nine of the States are arguing that the case should be settled, and the other nine States plus the District of Columbia are arguing against the settlement. But this is America, right? So, I do not know if the Microsoft case is an example of success or failure in multiState and federal enforcement cooperation (Let us not forget that the other meaning of the acronym ‘AG’ is ‘Aspiring Governor’… — the American political tradition is that if you are a successful State Attorney General you have good possibilities of moving up politically. So price-fixing cases are a lot more interesting politically speaking than cases of access to essential facilities or vertical restraints, which are not so interesting to the voters). 䉴 JAMES RILL — First of all, Barry, I wanted to thank you for your remarks, but not everybody has applauded so vigorously my endorsement of State cooperation in the antitrust field… The subject of my presentation is another sort of network, the one that links competition authorities and sectoral regulators. The issue of competition authorities ‘invading’ sectors and of sectoral regulators ‘intruding’ into the area of competition is very much a shared problem between the US, Europe and, in fact, the rest of the world. Allow me to make several preliminary points. First of all, I would like to suggest that, in the US at least, there is probably more impact in terms of volume of commerce in the intersection between competition authorities and sectoral regulators who deal in competition issues than there is with respect to those instances where there is an overlap between State and federal competition enforcement.

18

The Network Concept, Competition Authority Networks

Secondly, I think that the problems are, if anything, more complex, because the intersection occurs with direct dual regulation under different statutes, and with different policy foundations, whereas State and federal competition authorities essentially enforce the same kind of law. This issue was considered in some detail by the Department of Justice’s International Competition Policy Advisory Committee — the ICPAC, which commissioned a paper on the sectoral/competition overlap to William Kovacik,9 now General Council at the Federal Trade Commission. This paper highlights the international complexities that are created by the intra-national networks of regulatory and competition authorities. Consider a global merger transaction with now 60 plus — and I think that is probably an understatement — national competition authorities having some sort of review attributions. What happens if, hypothetically or actually, Deutsche Telecoms makes an investment in telecoms in the United States, the merger passes without much trouble with the competition authority, and then runs into public interest concerns at the Federal Trade Commission, which, under some influence by Congress, is very involved in what could be considered broadly ‘consumer welfare issues affecting competition’? The same holds true the other way. The outcome can cut both ways. In many instances in the US the result is perhaps more permissive than a competition authority would prefer it, and in some instances more intrusive than the competition authority would prefer. I will give you a couple of examples in this sense. One striking example of a permissive outcome is the treatment by the Surface Transportation Board of the Union Pacific-Southern Pacific merger. I personally had some role in persuading the then-Assistant Attorney General of the Antitrust Division, Anne Bingaman, to appear before the Surface Transportation Board and oppose this merger. She went to the hearing, and one of the commissioners of the Surface Transportation Board said, ‘Well, you just care about that competition stuff. You don’t know anything about how to run a railroad’. And they decided on that basis that the competition issues raised by the Antitrust Division of the Department of Justice were irrelevant — the statute gave them the authority to do so. The result was the ‘wonderful’ UPSP merger, which took some effort to undo. I am not going to go now through all the other examples mentioned in my written contribution. Perhaps just a few words about the Bell Atlantic/NYNEX merger in the telecomsmunications sector: in this case the issue was one of potential competition. After a thorough evaluation of potential competition, the Department of Justice decided not to oppose the merger, as in the standard legal doctrine on cases affecting potential competition there was no basis for a challenge. I think that from an antitrust law standpoint that was probably a correct decision. Then the

9 William

E. Kovacic (1999): ‘The Impact of Domestic Institutional Complexity on the Development of International Competition Policy Standards’. The paper is reported in Annex 3–B of the Final Report issued by the Department of Justice’s International Competition Policy Advisory Committee (ICPAC Final Report) — available at http://www.usdoj.gov/atr/icpac/3b.pdf.

Panel Discussions

19

Federal Communications Commission came up with the principle of precluded competition — established on competition grounds, and not on the grounds of the agency’s own jurisdiction. That is, the FCC found that the franchising and licensing rules were such that Bell Atlantic could have entered NYNEX’s ‘territory’ even though regulatory provisions would have had to be changed in order to permit it. Therefore the FCC established that Atlantic Bell and NYNEX were potential precluded competitors, and approved the merger with some conditions, which in reality had nothing whatsoever to do with competition policy, but rather with the social policy concerns within the FCC’s own jurisdiction: investment in infrastructure, universal services, safeguards, etc. Finally, a couple of remarks on the agriculture sector. One would hardly think of agriculture as a regulated sector, but one would always think of agriculture in the US, as perhaps across the Atlantic, as a highly politicised structure. There have been congressional hearings on a proposal that antitrust authority over mergers in the agri-business sector be taken away from the Department of Justice and the Federal Trade Commission and put into the hands of the Department of Agriculture. Fortunately that proposal has not passed, but it did require some micro-management — micro-meddling, one might suggest — by Congress to create a permanent office in the Antitrust Division of the Department of Justice to regulate the agriculture sector. So, the question is, what can be done in order to eliminate such problems? In academic writings we always have the right to come up with solutions that we know probably won’t fly. The solution that I propose, and that probably won’t fly, at least in the US, is a follow-up on the ICPAC Report:10 take away the authority of the regulatory agencies to deal with antitrust issues, have the antitrust authorities move in first to review the antitrust issues, and then let the regulatory agencies act over other public interest issues. Interestingly, this view has some support from the Chairman of the Federal Communications Commission (FCC), and from the former Chief of Staff to Assistant Attorney General of the Antitrust Division, Joel Klein. The support on the ‘hill’, for political reasons, is somewhat underwhelming. This is a point for the political scientists: as regulatory agencies lose their powers, the Chairman of the Congressional Committee that supervises the regulatory agency also loses some of his power. Another possibility would be the so-called ‘structured sectoral input’: give the antitrust authorities the power to review the competition issues of the merger but take into account some required input from the regulatory agencies. Should that be discarded as well, we could have concurrent review of the merger by the two kinds of authorities, with the antitrust authorities moving in if there is a competition problem. The Bank Merger Act provides a useful model to follow in that regard: the antitrust authorities are entitled to put an automatic stay on a bank

10 See

supra note no 10.

20

The Network Concept, Competition Authority Networks

merger even it has already been approved by the Federal Reserve. I would propose however to reverse this order, and, if the power to decide on the merger of the regulatory agency is pre-emptive, introduce the requirement to have a formal opinion on the merger from the antitrust authorities, to which the regulatory agency must formally respond. This would open the possibility for the antitrust authorities to bring an action in court seeking to overturn an overly-permissive decision by the regulatory agency. Finally, I think, it would be a good idea to allow the antitrust authorities to bring a structured input into the deliberations of the executive branch on policies affecting consumer welfare and competition. KAREL VAN MIERT — Thank you very much indeed. This was very informative, and it obviously triggers questions that are relevant to the European situation as well. I remember, for instance, when we (the Commission) had to deal with the American Airlines/British Airways merger, and the US Department of Transportation took a very different view from that of the US Department of Justice. Political considerations prevailed over competition considerations and concerns. Similar situations arise from time to time in the telecoms sector.



ULF BÖGE — After listening to Barry Hawk’s intervention about the NAAG, I got the impression that it is very similar to the network between the Bundeskartellamt and the competition authorities of the German Länder. As a premise for my intervention, I would like to say that, in my view, it makes sense to establish a network of competition authorities only if this will help enforce competition law more effectively than if each competition authority were to act on its own. Back to the subject of my intervention, are there any lessons to be drawn from the functioning of the German network of competition authorities for the set-up of the EU network? The German network of competition authorities comprises the Bundeskartellamt, the state competition authorities and, in exceptional cases, the Federal Ministry of Economics and Technology. The independence of the members of this network is in contrast with the objective of consistent application of the law, which requires uniform action in similar cases. There are two modalities of contrasting this tension. The first is legal: Germany has one Competition Act, and the competencies to apply it are divided among the members of the network according to clear geographical and substantive law criteria. The members of the network are under a legal obligation to exchange information. The second is the judicial review system. The Federal Supreme Court and the higher courts of the states have specialised antitrust sections. The judicial review system ultimately ensures that the law is applied consistently, but this also means that reaching a final solution involves various decision-making bodies. Consequently, different interpretations of the law by the competition authorities are consciously allowed. A clear delimitation of competencies is indispensable for the functioning of the network. The question is, however, whether this is sufficient for an efficient 䉴

Panel Discussions

21

functioning. Legal provisions generally leave some room for interpretation in their application. What if new developments, such as the liberalisation of former monopolies, require the competition authorities to apply a coordinated strategy? Moreover, it is conceivable that the members of the network of competition authorities work against each other. Therefore, in my view, the legal approach to solving the functional problem of the network need to be complemented by practical cooperation. In Germany, the representatives of the competition authorities meet twice a year to exchange experience and discuss all kinds of legal and factual enforcement problems. Any participant at these meetings may propose specific issues to be discussed. For instance, a special Working Committee deals with the electricity, gas and water sectors. Common work is done on specific projects, such as the application of rules on abuse of a dominant position in the energy sector. The Bundeskartellamt and the state competition authorities co-authored a special policy paper dealing in particular with questions of comparative markets and the control of costs. Although this paper has no legal character, the German competition authorities use it as a guideline. This is possible because principles contained in this paper were worked out together. To sum up, the German network of competition authorities is ultimately a partnership for the joint application of the law. This idea is embodied in the rules governing the relationship between the members of the network, for example the delimitation of competencies and the information obligation, and reflected in the practical cooperation between the members of the network. Consistency in the application of the law has been achieved through the delimitation of competencies. Can the German experience be transposed at the EU level? Certainly not on a one-to-one basis. The allocation of enforcement competencies is not as clearly regulated at the European level as in Germany. Moreover, the members of the EU network come from different legal and cultural backgrounds. Finally, the role of the European Commission cannot be compared to that of the Bundeskartellamt. At the same time, the EU network will also have to be based on a spirit of cooperation and partnership. This is not only the best guarantee for an effective enforcement of the law, but it can even solve many practical enforcement problems. However, this spirit of cooperation can be achieved only if all the members of the network are allowed to have a say on the interpretation of substantive rules and on the forms of cooperation within the network. 䉴 MARK THATCHER — The regulatory networks acting in the European telecoms sector offer an interesting point of comparison with the EU network of authorities about to be created in the competition field. On the one hand, there are differences between these two areas of EC regulation: EC telecoms regulation is implemented mostly by the authorities of the EC Member States. If the reform of the EC competition law enforcement system goes ahead as proposed, the competition authorities of the EC and the Member States will form a network of enforcers for the EC competition rules. In the telecoms sector, however, there are two sets

22

The Network Concept, Competition Authority Networks

of networks: one that operates at the European and the national level, and others that operate at the national level only. On the other hand, there are similarities: in both of cases there are two levels of regulation — the European and the national one — that are interdependent. In my presentation I intend to first set out the shape of EC telecoms regulation and explain why it leaves room for quite some discretion for the national regulatory authorities. Secondly, I will briefly discuss the shape of national regulatory networks. Thirdly, I intend to look at some of the regulatory outcomes in the area of telecoms in terms of decisions taken. Finally, I will talk about some of the feedback effects of reliance on national implementation of EC regulation. Telecoms is one of the most advanced areas of EC regulation, and yet this has not meant the end of national telecoms regulation. Why? First of all, EC telecoms regulation is incomplete: some crucial areas, such as ownership, are not covered. Secondly, it allows for different methods of compliance with its requirements: EC directives are binding upon the Member States as to the effects, but not as to the means to achieve them. Thirdly, although EC regulation is implemented mostly by national authorities, many of their institutional features, such as institutional status and resources, appointment and dismissal of staff, etc., are not covered. Let me turn now to the national telecoms regulatory networks. In my written contribution for this workshop I attempted a comparison across six Member States — I am not going to be able to go through it in much detail here, but I would like to make a couple of observations. One is that the national networks comprise a variety of interdependent actors: regulatory agencies, governments, legislatures, competition authorities, courts, public operators. The second is that we can observe considerable structure variations across the Member States. Who are the key actors? Perhaps the most important are the regulatory authorities. All six Member States under observation have set up regulatory agencies, but we observe variations in terms of their number (from one in the UK to nine in Italy and France), institutional status (independent or not), the terms of their office (length and modality of appointment) and institutional resources. Also, we can see that their range of objectives is quite broad, so there is a lot of discretion in terms of organisation. Finally, we can see that, even if the regulatory agencies are statutorily independent, in practical terms they are still very much dependent on elected politicians — which are the second group of actors in the national regulatory networks. Elected politicians retain powers over the appointment of the independent regulators, over the determination of their financial and staff resources, and can also use legislation to change the organisational position of regulators. What is more, if there is public ownership in the sector, governments are both regulators and owners. Governments also retain discretion over important regulatory areas, such as the allocation of licences. At the same time, governments are dependent on other actors, and in particular on the sectoral regulators, for information and expertise. The third group of actors is constituted by the telecoms operators, and in particular the incumbents, who still dominate national markets. When we look at the incumbents, there is also a lot of variation: at one extreme we find British

Panel Discussions

23

Telecoms, which is practically private (the State did not retain a ‘golden share’), and at the other extreme there is Deutsche Telecoms, where public ownership is still majoritary. The fourth important category of actors is constituted by the national competition authorities. The competition authorities are important for two reasons. First, they offer an alternative to the sectoral regulators, and often they have a longer institutional history and greater political power. Secondly, they offer an alternative method of regulation. Sectoral regulators typically intervene ex ante, whereas competition authorities also intervene ex post, especially when it comes to firms with market power. The legal arrangements as to the division of responsibilities between competition authorities and sectoral regulators differ from one Member State to another. In practice, however, it is much more difficult to determine who should deal with a specific matter. Thus far the relationship between the competition authorities and sectoral regulators seems to have worked well, but it will be interesting to see how it develops in the future. The fifth category of actors are the courts. These have not been so far particularly active in all Member States, but they are always there as a threat to regulators. Let me turn now to the effects of having these different network arrangements. In my written contribution I looked at various operation areas: licensing, universal services, retail prices. First of all, we see that the Member States have, basically, all followed a similar direction of change towards liberalisation, greater competition, and attempts to ensure a level playing field. At the same time, there are significant variations of outcome. Take the example of third generation mobiles (UMTS): one can see significant differences from one Member State to another in terms of how licenses were allocated — and in particular whether an auction was held or not — and the licensing prices — from virtually nothing in Spain to billions of Euros in Britain and Germany. Another example is that of universal services: in the UK, British Telecoms has an obligation to provide universal services, but OFTEL believes that it does not cost it anything to do so, so it does not get any compensation for the service. This is not the case in other countries, like France. Retail price is another example: the nature of regulation on retail prices varies a great deal from one Member State to another in terms of which kind of services are included, what form of price cap is used, how often that price cap can be revised, who revises it, and so on. Finally, what have been the feedback effects at the European level? The question arises because over time one set of regulatory decisions feeds back into the next set of regulatory design choices. We saw that in the telecoms sector there is a combination of many categories of actors taking part in the regulation process, EC regulation is implemented by the Member States, and there are regulatory gaps in the European framework. As we have seen, this has led to diversity in institutional arrangements and regulatory outcomes at the national level. I think that there are four possible European reactions to these developments. First, one can simply continue to accept cross-national differences. I think that this was basically the underlying agreement for the development of EC telecoms

24

The Network Concept, Competition Authority Networks

regulation: the Member States operated in partnership with the European Commission, and the quid pro quo was that EC regulation would allow a lot of discretion to the Member States. The trouble is that over time such an arrangement can generate problems in terms of fairness of competition, or transaction costs to companies, who find they have to obey different sets of regulations from one Member State to another, and not in the least, undesirable races to the bottom, whereby some member States try to offer much more attractive regulatory conditions with the result of undermining public policy goals such as the provision of universal services. A second approach is benchmarking and cross-national comparisons. This approach has also been used by the European Commission, who publishes regularly comparisons of the regulatory arrangements in the Member States. This slides into a form of benchmarking and informal pressure on the Member States, as the finger is pointed at those Member States doing worse than others. It is a form of soft harmonisation. A third strategy is to create new regulatory networks in order to deal with the coordination of the existing ones. (This obviously leads to a proliferation of networks.) Take for instance the High Level Regulators Group, which under the new regulatory package of 2002 is going to become more formalised, as the European Regulators Group for Electronic Communications. The independent national regulators have also got together and formed the Independent Regulators Group. These are examples of European-wide formal/informal networks trying to deal with cross-country regulatory variation. I think it has not surprised the political scientists that there has been a move towards the formalisation of such crosscountry networks. I am not sure though how sustainable this third approach is. The fourth strategy is strengthening the EC legal framework. For telecoms this has sometimes been done with ‘soft law’ (notices, recommendations, etc.), but recently the most significant instrument has been the so-called ‘2002 regulatory package’. The package contains a series of directives dealing with a wide range of issues regarding the status and behaviour of the sectoral regulatory authorities in the Member States: authorisations, the right to appeal decisions of the sectoral regulators, etc. The objective of these rules is to reduce the margin of discretion of the national regulators. To conclude, I think there is an inherent tension between, on the one hand, the objective of decentralising regulation, and on the other hand, ensuring fair competition. The decentralisation of law enforcement inevitably leads to crossnational differences. Those differences in turn create pressure for stronger regulation at the European level. 䉴 KAREL VAN MIERT — The fact that national regulators still have so much discretion led to a situation whereby in some of the EC Member States monopoly situations were re-created. In Germany, for instance, Deutsche Telecoms re-established a monopoly in certain areas of the telecoms services because the national regulators were not in a position to react as swiftly as needed. Obviously in such

Panel Discussions

25

situations the European Commission can intervene on the basis of EC competition rules regarding abuse of a dominant position, but in practice such interventions take some time, and in the meantime the competitors have already been eliminated. Similar situations have arisen in the electricity sector. So there is indeed a good case for the EC to intervene with regulatory instruments, although this often creates political tensions with the national governments. PETER CAMERON — The regulatory network that operates in the electricity and gas sectors provides a very interesting contrast to the competition network about to be established. In particular, the European Commission’s role in the former is relatively weak. Moreover, by contrast to the situation in the telecoms sector, there is a striking absence of consensus on the aims, and indeed, even the scope of competition in EU energy markets. Competition is to be balanced against other goals, such as security and reliability of electricity and gas supply. And, in the understanding of many EC Member States — notably France — the advent of competition in this rather conservative sector of the EU economy is also associated with sharp reductions in the staff of public utilities. In addition, the unfortunate chaos that resulted from energy liberalisation in California has had a constraining influence upon most programmes of energy liberalisation around the world. Indeed, recently in the EU there have been disturbances or near accidents reported in the Northern Electricity System in Scandinavia, in Belgium, in the Netherlands, in Ireland, in Spain. This has encouraged a spirit of caution towards the very idea of competition in this sector. Allow me first to underline some key features of the energy sector. By contrast to telecoms, the energy sector is characterised by slower technological innovation, a limited — almost negligible — impact of globalisation — at the present time only about 7 or 8 per cent of electricity is traded across borders — and limited consensus among the EC Member States on the aims and extent of market liberalisation. As a result, we find that in this sector liberalisation is very much a ‘work in progress’, and the incumbent monopolists, be they private or publicly owned, are still very influential. The European energy regulatory network was established following adoption of two extremely important EC directives on electricity and gas, in 1997 and 1998, respectively. It took six to seven years for those two directives to be adopted, and this very long period is quite indicative of the problems that especially the European Commission has had in getting its way in this sector. What are the main features of this network, what was the legal basis for its setup, and how is it going to evolve in the future (as currently there are proposals for its amendment)? (As a preliminary parenthesis, many of the energy competition cases that have been dealt with since the adoption of the two EC directives are discussed in my last book.11) The EU energy regulatory network comprises the



11 Peter

Cameron (2002): Competition in Energy Markets: Law and Regulation in the European Union, Oxford University Press.

26

The Network Concept, Competition Authority Networks

EC institutions, the national sectoral regulators and the national competition authorities. But what is most significant is that many of the market players are involved in making the system work. I do not know if that is unique, but this is certainly a very overt co-involvement of transmission system operators, traders, power exchanges and consumers. Many of the national sectoral regulators are, in fact, organised in a body called the Council of European Energy Regulators, which has some similarities to ECA (the network of European national competition authorities). The main characteristic of this network is that the members get together twice a year for meetings held in Florence and Madrid where they attempt to further develop the very general rules set out in the two EC directives. These two fora — the electricity one in Florence, and the gas one in Madrid — have been important ways of bringing all the market players together and getting them to contribute to the further development of what the Commission calls ‘the deepening of the internal energy market process’. The two EC energy directives need to be reformed partly because they have been quite successful in encouraging many EC Member States to go beyond the minimum of market opening and change required by them. The two energy fora have been described as ‘new instruments of governance within the European Union’ and ‘a form of regulation by cooperation’. They are very innovative, voluntary networks, where market players are given an important role to play. The discussions at these meetings are often highly technical in character. It is worth noting that the regulators have been happy to admit that they have learned a great deal from the market players and other members of the network about the particularities of competition in this sector. Indeed, there are certain constraints in the electricity sector that have everything to do with the laws of physics and not the laws of the market, and in designing the competition regime for this sector one must take that into account. In turn, the transmission system operators have felt encouraged to work together for deepening the reform process. At the same time, the voluntary network has its limits: if the incumbent market players decide there is something in the regulatory status quo that they do not want changed, they can put a brake on the reform process. This is very much the reason why we are having now new Commission proposals to amend the existing electricity directives and introduce a new regulation that deals with cross-border electricity trade. If the market players do not want to go ahead with further reforms, the Commission is obliged to make its own proposals. An important element of the EC electricity directives was the establishment of a timetable for the gradual liberalisation of this sector. How was EC competition law applied over the liberalisation period? In my book I have shown how the Commission adopted a dynamic approach to the application of EC competition rules to this sector. In other words, the European Commission’s Competition DG sought to support and further support the reform of the sector, as well as help the regulators through the application of EC competition rules, for example

Panel Discussions

27

through merger decisions that conditioned approval upon following the rules agreed in the two energy fora. So there has been an interplay between the application of EC competition law and the regulatory process that I think has been quite successful. To conclude, progress with the adoption of secondary legislation for this sector is linked to the liberalisation timetable; in the meantime, however, the competition authorities may find themselves hampered in the application of the competition rules by a need not to place obstacles in the way of the liberalisation process, which proceeds according to a timetable. In the future, the role of the network of competition authorities is likely to grow as the internal energy market process moves ahead. The energy sector is already much better integrated in the framework of EC competition law than it was six or seven years ago. There were a lot of competition cases in the energy sector over the last few years that show tremendous creativity on the part of the European Commission’s Competition DG in pushing to promote competition in this deeply conservative sector. The pace of change is by now quite fast, so I am hoping that the subject is one that we will be able to revisit in two or three years’ time. KAREL VAN MIERT — I could not agree more with your concluding remarks. I recall that some seven or eight years ago it was extremely difficult for the Competition DG to step in this sector, as most EC Member States still had monopoly situations in this sector. Gradually, the application of competition rules in this sector became more and more prominent, and nowadays it becomes even crucial, because there still might be a very big difference between liberalisation on paper and the real world.



SANTIAGO MARTÍNEZ LAGE — First, I would like to make a brief comment about the Spanish infra-national competition network, which is about to be established. Two months ago, our Competition Act was amended along the lines of a judgment delivered by the Spanish Constitutional Court, according to which regional governments had to be given competence to apply national competition law. According to this amendment, the Spanish regional governments are competent to prosecute collusive agreements and abuses of dominant position, and to grant individual exemptions. As in Germany, they are not competent to authorise concentrations or adopt block exemptions — these areas of enforcement remain within the competence of the central competition authority. Like in Germany, we now have in Spain strict jurisdictional rules, so there is no room for case allocation. If the effects of an infringement do not go beyond the regional borders, the regional authority is competent to deal with the case. So case allocation rules are supposed to be quite clear, but if the authorities do not come to an agreement on the allocation of a case, a constitutional conflict will arise, and in any case individuals could always appeal a decision taken by the central or regional authority on grounds of lack of jurisdiction.



28

The Network Concept, Competition Authority Networks

Next, I would like to underline two aspects of the amendment which I find particularly interesting. One is to allow the central authority to intervene in the course of regional proceedings not only as amicus curiae — in the same way as the Commission would be entitled to do in the course of national proceedings according to the draft regulation — but also as an interested party. This means that the central authority could even appeal decisions taken by the regional authorities. The other is that the regional governments which do not want to establish their own competition authority have the possibility to just instruct the case and pass it over to the central authority for a final decision. In my opinion these are the only positive aspects of the amendment. This reform was introduced on the background of serious mistrust between the central and the regional authorities. This is why, for instance, the amendment says nothing about mutual assistance with respect to carrying out investigations between the central and regional authorities, or between the regional authorities themselves. Moreover, the amendment says nothing about exchange of information. This situation is not far from that existing years ago between the European Commission and the national competition authorities concerning the application Article 81(3) EC, and we saw that, after a certain time, there is a new atmosphere of mutual trust between the Commission and NCAs. We can only hope that in the future the same will happen in Spain, but I am not very optimistic about this. I have to say that, while at the European level the move is towards further integration, for the time being in Spain the debate is about further decentralisation, or possibly disintegration. FRANCES BARR — I just wanted to pick up on the points made in relation to sectoral regulators, and in particular on Peter Cameron’s concluding points. To give a UK domestic example, the Office of Fair Trading and the UK sectoral regulators are actually a network. In the UK all sectoral regulators have concurrent powers to apply national competition rules. There are actually formal rules on case allocation, but more importantly, we have informal meetings every month where we discuss problematic issues and cases. In addition, the Director General of the OFT is under an obligation to produce guidance on how the competition prohibitions are to be applied or enforced including in the regulated sectors, and so far we have produced a large number of guidelines acts whose content has been agreed upon in meetings with the regulators.



JACQUES BOURGEOIS — Professor Monti has said that the case allocation decision is not going to be challengeable. I think that he can expect quite a bit of resistance on this, and subsequently quite some fight about that as a result of the consequences of a case being allocated to this or that NCA. I am not sure that the EC courts will follow the Commission, and even the Council, if the regulation stipulates that this is not a challengeable act. I also had a question for Dr Böge: what proportion of the total number of cases in the application of German antitrust law is dealt with by the Länder competition authorities?



Panel Discussions

29

MARIO MONTI — I will try to respond to this specific point through an example. Imagine a cartel operating in the Benelux area. The case could be prosecuted by the European Commission, the Belgian, Dutch or Luxembourg authorities. Suppose that it is decided, so to speak, to allocate the case to the Dutch competition authority. This means that the other competent authorities decide not to prosecute. In my view, this decision does not harm the interests of the companies concerned, and therefore it should not be open to appeal before the Court of First Instance. Of course, the interested companies will be able to challenge in court the final decision of the Dutch competition authority. This is how I would see, as a first approximation, the problem. Perhaps Mr Wils would like to add a few words on this.



䉴 WOUTER WILS — I think that the example given by Commissioner Monti is perfect. The point is that in such a case there are parallel enforcement competencies, so the undertakings in question are under threat, so to speak, from the European Commission and the Belgian, Dutch and Luxembourg authorities. What changes in the legal situation of the undertakings when the network of competition authorities adopts a so-called ‘decision’ in the sense that the case will be allocated to the Dutch competition authority? The change is that the undertakings are liberated from the pressure of being targeted for prosecution by several competition authorities. The only thing unfavourable to the undertakings will be an eventual sanction decision by the very effective and tough Dutch competition authority — which of course can then be challenged before the very competent and efficient Dutch courts. 䉴 JAMES RILL — I find it astonishing that members of the Bar would think that there should be a legal basis for challenging the decisions on allocation of cases within the network — decisions that are purely discretionary. I can conceive of only one area where there could be problems, and that is the sharing of information between jurisdictions where there are only civil sanctions for antitrust infringements, and jurisdictions where criminal sanctions are also available. But it seems to me to be very hard to find a legal basis for challenging the discretionary exercise of authority.

ULF BÖGE — In reply to Jacques Bourgeois’ question, it is difficult to tell exactly the proportion taken by the cases dealt with by the state competition authorities. But consider that merger control is the exclusive responsibility of the Bundeskartellamt, and my impression is that the share of the Bundeskartellamt in the enforcement of German antitrust rules is higher than that of the Länder authorities. For example, the number of cases concerning abuse of a dominant position dealt with by the Bundeskartellamt has increased since last August, when we created a special unit dealing with such cases in the electricity sector. We had about 20 cases in this sector, on top of the around 200 cases abuse cases related to other sectors, which were dealt with by the federal and the state authorities.



30

The Network Concept, Competition Authority Networks

The state authorities have the same investigative powers as the Bundeskartellamt in cartel cases, but of course the most important cartel cases were dealt with by the Bundeskartellamt, because the effects of the infringement were felt in several Länder. CALVIN GOLDMAN — I would like to make two brief comments, from a Canadian perspective, on the presentations made by Barry Hawk and James Rill. More specifically, concerning the NAAG, from what I have seen there seems to be a critical gap in the ability of the State Attorney Generals to work effectively together in one context: they cannot share confidentially-protected information. They do not have the legal authority to take information that they may have subpoenaed and turn it over to other members of the network without the consent of the party concerned. And — as I understand it, having worked in the Canadian and US antitrust enforcement both on the government and the bar sides — the State Attorney Generals cannot share documents obtained from the federal authorities. For example, if the Department of Justice compels documents or testimony by grand jury process to investigate a cartel, those grand jury documents and transcripts, to the best of my knowledge, cannot be shared with the State Attorney Generals. However, they can be sent to Ottawa, and in fact be shared with the Canadian Competition Bureau. This is quite an interesting dichotomy, or anomaly: the proof cannot be used by the State Attorney Generals, but can be used by the Canadian authorities. The US–Canada Mutual Legal Assistance Treaty makes the latter possible. If I understood well, according to Article 21 of the draft regulation, the members of the EC network can exchange information without the consent of the companies concerned. This is essential for the ability of any network to operate effectively together, but it raises other questions, such as that of legal guarantees or safeguards to be provided to the concerned companies. My second comment relates to James Rill’s remarks about the difficulties that a competition authority is facing in undertaking concurrent review together with the sectoral regulators. We had the same experience, of course, in Canada. And I can tell you from my personal experience that we made extensive use of our intervention power — i.e., the statutory right of the competition authority to appear and make submissions in the hearings of any other sectoral authority. That is one way of dealing with the problem. Another way that, in my experience, is far more effective in most contexts, is to follow what Dr Böge and James Rill refer to as critical, namely the way of informal dialogue between the agencies. Dialogue should occur not only at the level of heads of agencies, but also at the staff, operational level. Sometimes by following the informal channel we have accomplished a lot more than by formal, statutory interventions.



BARRY HAWK — It is true that the members of the NAAG cannot share subpoenaed information unless the target companies consent, but that is not as big an obstacle as it might first appear to be. If you are investigating a cartel among several companies, you can subpoena individuals. As a matter of fact, this is one



Panel Discussions

31

of the glaring weaknesses in the EC antitrust enforcement. Sometimes one wonders how they ever get to prove a cartel case in Brussels without the power to subpoena, i.e., to compel testimony from individuals. And you do not need that many people to volunteer to cooperate in the investigation if you can sanction individuals. 䉴 JAMES RILL — Barry Hawk referred earlier to the 1984 Vertical Restraint Guidelines issued by the Department of Justice which were not followed even by the Department itself. The States felt compelled to put out their own Guidelines, which were quite different. But let me give you an example of properly designed Guidelines, such as the Horizontal Merger Guidelines drafted by Professor Baxter in 1982. In 1992 these Guidelines were revised, and I think that the biggest improvement was that we got the Federal Trade Commission to discuss about regulatory controversy and sign those Guidelines. The States de facto endorsed those Guidelines for the most part. I think that the use of such Guidelines serves not only the purpose of transparency, but, if properly designed, the Guidelines can make a major contribution towards both law enforcement convergence and compliance in the private sector.

MARIO MONTI — One comment addressed to Professor Thatcher: you have shown clearly in your intervention how ‘decentralised’ telecoms regulation undermines the level playing field in the Internal Market, and you implied at the end that the same risk might arise with the modernisation of EC competition policy. Yet I wonder whether we should not draw a distinction between areas where the EC regulatory system was set up from the beginning as largely decentralised and areas where it has been set up and it remained centralised for decades, as in the case of EC competition law? And one question for James Rill: the European press recently reported a proposal coming from the US Department of Defence concerning a possibly enhanced role of the Committee on Foreign Investment in reviewing mergers. Where would you place this proposal within the framework of several US agencies having concurrent jurisdiction to review mergers?



䉴 MARK THATCHER — That is a very interesting question. I suspect that in both cases the mechanisms at work are quite similar. I think this is so because even in areas where the European regulation is quite detailed it is still difficult to have a sufficiently encompassing set of rules. Moreover, there are cases where the European rules on the substance are quite detailed, but in practice there are also issues of procedure and resources, and I doubt that the rules concerning the latter are sufficiently developed. To use a simple example: the resources of regulators vary from one Member State to another, so the time necessary to deal with a case would also vary from one Member State to another. As for accumulation of trust, there may well have been an accumulation of trust between certain regulators, but regulation is also about conflict. I would have thought there is even more potential for conflict in competition law enforcement

32

The Network Concept, Competition Authority Networks

than there is in telecoms regulation. Here it is not just potential of conflict between regulators, but also, private parties have directly conflicting interests. One party may wish to have its case heard in one particular Member State where the procedures are slower, and the counterpart will want the opposite. So, I think that in this area of law enforcement there are fundamental mechanisms opposing towards conflict, and then towards conflict resolution. JAMES RILL — Your question, Professor Monti, raises the issue of the authority the US Committee on Foreign Investment. I actually sat as a member of it for a while. It is not particularly startling, nor is it news, that national security issues prevail over competition considerations, particularly when there are genuine national security issues at stake. This order of priority is recognised in the US–EC cooperation agreement of 1991, in the antitrust agreement between the US and Australia, and I think it is a generally recognised principle of international law. When there is a particularly sensitive merger, let us say one involving highly sophisticated electronic system used in missile guidance, it seems appropriate that the US Department of Defence would have some input into the review, on aspects such as whether or not that particular system should become available for use by another country. The danger, of course, is that interests other than national security, or superficial national security interests, intrude into that kind of decisionmaking. Yet, while the US Department of Defence understandably has a major role to play on this Committee, I was always somewhat bemused by the US Department of Commerce also having a seat on the same Committee. I found that some of the views expressed by the US Department of Commerce had nothing to do with the national security of the United States, but rather with the health and welfare of a rather limited sector of the American business community. The problem that arises is how much transparency you can get in that kind of decisionmaking process. And, quite frankly, if the decision rests with the governor or the chief magistrate of any State jurisdiction, that decision is not going to represent national security interests. Perhaps the most that one can ask for is that the decision does not rest with the vote of elected politicians or magistrates. Maybe the only guarantee is to invest responsibility for decisions that set aside competition for national security considerations in the President, in other words, to place responsibility for this kind of decisions at the highest level competent to make such evaluations, and at least make the decision-making process transparent.



䉴 MARIO SIRAGUSA — I would like to make a brief comment on the presentations of this morning. First of all, I think that the EU system has always been a network system. The genius of the Treaty is that, through the principles of supremacy of Community law and direct applicability, there has always been a role for national authorities and courts in the application of EC law. This enforcement system is by definition a network system. I think that what we are discussing about here is the change of role for the various members of this network.

Panel Discussions

33

Under the old enforcement system, the Commission was the promoter of EC competition policy and had the power to enforce EC competition rules so as to avoid differentiation in their application at the local level — an effect that decentralisation necessarily entails. For me, the key question is whether the new enforcement system will allow the Commission to continue to play this role. Will the Commission have the political strength to continue doing what it is doing now? I have my doubts about that. This has always been my concern about the reform proposal. I think that the Commission’s proposal has already gone as far as it could in terms of decentralisation. And I absolutely do not understand why there are efforts to take away even more powers from the Commission, unless we believe that Europe does not need any more a policy promoter in the competition area. If you really believe that we are already in Europe in a situation in which national authorities themselves can carry out that role…. So I think that we should seriously reflect on the implications of these requests. To give just one example, could the Commission under the new enforcement system issue a new block exemption regulation — like the one for the motor vehicles sector, that is now discussed — in spite of the opposition of some Member States and a large part of the industry? 䉴 KAREL VAN MIERT — If I may say one thing: if I thought that the Commission would no longer be able to continue to play this role of policy promoter, I would have never taken the responsibility for the Modernisation White Paper. But I was convinced that we could change the system while at the same preserving this role for the Commission. 䉴 MARIO MONTI — I liked very much what Mario Siragusa said, and the example of the block exemption for motor vehicles distribution is very well chosen. I have one remark and one question for Mario Siragusa. The remark is that there are of course political difficulties with the adoption of this block exemption regulation, but nevertheless we are confident that it will be adopted before the summer,12 obviously having taken into account the outcomes of the consultation with the Member States and representatives of the industry. The question is the following: what exactly are the aspects of the modernisation project that you see as weakening the power of the Commission in relation to decisions like this one? 䉴 MARIO SIRAGUSA — For me, the main problem is the elimination of the Commission’s monopoly to apply the exemption provision under Article 81(3) EC. I think that the Commission’s ‘power’ to deal with problems such as those arising with respect to motor vehicles distribution is based on the monopoly that the

12 See

Commission Regulation (EC) No 1400/2002 of 31 July 2002 on the application of Article 81(3) of the Treaty to categories of vertical agreements and concerted practices in the motor vehicle sector, OJ L 203 of 1.08.2002, pp. 30–41.

34

The Network Concept, Competition Authority Networks

Commission has had so far to apply the exemption provision under Article 81(3) EC. Since the main feature of the reform proposal is the elimination of that exclusive competence for the Commission, I wonder whether this would not have political implications. From the standpoint of legal practitioners it does have some implications, as there are aspects that still remain unclear; for example, what will be the role of group exemption regulations in a legal exception system? 䉴 ULF BÖGE — I think that Mr Siragusa asked the wrong question. The correct question is, why does the Commission want to decentralise the application of Article 81(3) EC? The objective of the reform is to improve the implementation of EC competition rules. Suppose that in Germany we would discuss now re-centralising the application of national competition rules. Centralisation would weaken our enforcement system, because the Länder authorities are closer to the market and know exactly what is to be done. And, as we have this network between the Bundeskaterllamt and the Länder authorities, we are much stronger than we would be if each of the members of the network were to act on its own. So, I think you should see this question of decentralisation in combination with the idea of the network. At the level of the EU, we will be more successful if we combine decentralisation with the set-up of a network of competition authorities, and of course the Commission has to play a central role in this network.

JAMES VENIT — I agree with what Dr. Böge just said. The modernisation exercise started from the need to accelerate and reinforce the Commission’s enforcement agenda, and was based, I believe, on the perception that the prior authorisation system was inefficient. At that point, the possibilities were either to cut back on the application of Article 81 EC so as to reduce the burden of the prior authorisation system, or to expand the possibility to grant exemptions under Article 81(3) EC to the courts, thereby taking pressure off the prior authorisation system. What frankly surprised me is that the Commission opted for totally abandoning its exemption monopoly. I think that this is what generated the fears that the Commission will lose its primordial role in the system. My question is, why did the Commission opt for total decentralisation, when it could have opted for a ‘middle way’, such as allowing the courts to apply Article 81(3) EC?



䉴 ALEXANDER SCHAUB — This was, indeed, one of the central issues during the more than two years of Commission internal debate on the reform project. One could have imagined intermediate solutions. Why did we choose this one? We chose it because we were convinced that it would be possible to develop cooperation with the competition authorities of the EC Member States, which is essential for making the new enforcement system work. I do not think that we were wrong, but I must confess that I am a bit worried to see that some see this reform largely in terms of a power game. We are trying instead to develop an enforcement system that is appropriate for the current state of European integration. The present enforcement system was designed at a time when the Community had only six

Panel Discussions

35

Member States with no competition law enforcement experience. Today all fifteen Member States have highly respected and competent competition authorities, with which we succeeded to develop a remarkable degree of cooperation and mutual confidence. Against this background, our conclusion was that while maintaining the prior authorisation system we could not achieve efficient enforcement. So the reform had to go deeper. We were convinced that the Member States’ competition authorities would be cooperative and reasonable enough to accept that even in a totally decentralised enforcement system there is a need for the Commission to play a primary role. This is the real issue we are discussing here. Now, coming back to Mario Siragusa’s question, the Commission has, and will continue to have, the determination to adopt group exemption regulations in spite of political pressures. I do not see why that should change. Another issue is whether the adoption of group exemption regulations will be possible in the new enforcement system. That is one of the points of disagreement still pending. The Commission suggested that, in the future, it should get a generalised authority for the introduction of new group exemptions. This has provoked uproar in some Member States — understandably so, if one looks at the issue only in terms of power distribution. However, one should look at the real important question, which is, whether the new enforcement system can function without group exemptions. The Commission believes that in the new enforcement system group exemptions will become even more important, because they are an instrument to obtain rapid clarification of highly complex and controversial legal questions. We thought it would be easier for the members of the network to have group exemptions to resort to, this would offer them better protection from the national governments. One thing is clear, the political pressure on national competition authorities will not decrease in the new enforcement system. National competition authorities have become a factor to consider in the economic policy-making at the Community level and in each Member State. The addition of new enforcement competencies will raise their profile even further. If you look at developments over the last few years, there cannot be the slightest doubt that the national competition authorities have increased their institutional weight, and at the same time the Commission has certainly increased its own institutional weight. I would contest anybody pretending that in the new enforcement system the national competition authorities will be losers rather than winners. Therefore I cannot understand the theory that the Commission seeks to grab more power through this reform. KAREL VAN MIERT — Apparently, the way that people are looking upon the reform is coloured by subjective assessments, at least for the time being. Some find that the Commission goes too far in totally giving up its exemption privileges, others consider that the Commission will still be in too strong a position vis-à-vis the national competition authorities 䉴

36

The Network Concept, Competition Authority Networks

䉴 ROLAND STURM — While listening to this discussion it came to my mind that the principle of subsidiarity was not mentioned even once. And I wonder whether this debate is not too much concentrated on ‘insider’ competition law enforcement aspects, whereas it should perhaps be cast against the more general debate over the division of competencies within the EU. This would perhaps also give more legitimacy to the reform. But it would also raise questions, such as, should we start from the idea that with the reform the Commission gives up some of its present competencies, or from the more general subsidiarity perspective, and then discuss how competencies should be divided? If one takes the latter as a yardstick, one may come to different conclusions as to how far the reform proposal has gone. Next, I would have one question for Alexander Schaub. He said before that in the new enforcement system the Commission should be empowered to adopt group exemption regulations. My question is, does the Commission intend to involve the members of the network in the decision-making, or does it intend to do this on its own? I think that if there will be a network, then its members should have the possibility to discuss the proposals of group exemption regulations, and have a real influence in the decision-making process. 䉴 ALEXANDER SCHAUB — You should not believe for a second that the Commission envisages to adopt group exemption regulations without having first undertaken the most intensive debate with all those concerned. In the case of the regulation concerning motor vehicles distribution, about to be adopted this summer, we are already at the stage of a third version subject to public debate. I would like to see one legal project in Germany that has been exposed to the same intensity of public debate…. My impression is that the nervousness among some on the German side concerning this issue is due to an incomplete knowledge of the practical reality. 䉴 FRANCES BARR — I just wanted to make a practical remark in relation to the block exemption regulations and how they are going to be adopted in the future. The Commission has always had the monopoly in adopting them, motivated by its experience in dealing with EC competition cases. In other words, the Commission would see when the adoption of a block exemption regulation was necessary for a specific sector/market, because it had knowledge of that area via the cases it had dealt with. In the future, however, once the reform is in place, it will be the competition authorities of the Member States that have more practical knowledge and case experience. Is it not logical, then, that the Member States’ competition authorities become even more involved in adopting block exemption regulations?

ALEXANDER SCHAUB — More involvement of the national competition authorities will of course be highly welcome. But we were talking here about the involvement of national governments, and these, to my knowledge, not only do 䉴

Panel Discussions

37

not have a lot of experience with competition law enforcement, but also have shown in the past not to be very interested in competition considerations, and are often guided by different considerations. If the national competition authorities were to decide together with the Commission, I would be much more relaxed than with the present system, whereby the Commission discusses with the national governments. To be clear, mine was not a declaration of mistrust against the national competition authorities, but rather concern about the behaviour of national governments. MARIO MONTI — I just wanted to mark a sharp disagreement with Alexander Schaub on one point, where he criticises the intervention of heads of governments concerning our block exemption regulation on motor vehicle distribution. I find this perfectly legitimate, especially in an area where Member States are consulted only on an advisory basis. Of course, it would be very worrisome if the Commission were not to face its responsibilities, and both Alexander Schaub and I know that the Commission will not do that. For the rest, if the opposition comes from the head of government of a very large Member State, we at the Commission are particularly at comfort with our conscience, and the consumers from that Member State have suffered more than proportionally so far from the restrictive practices of the automobile industry.



CLAUS-DIETER EHLERMANN — I wanted to thank you all, and especially the political scientists who intervened during this session. I think we opened here a huge agenda for political science research. Actually, one of the more interesting questions that was raised is the relationship between national governments and national competition authorities. I think this is an absolutely fascinating subject. Also, during your interventions I sat here trying to visualise how this network is going to operate, and I conceive it as a social unit, rather than a sort of legallyprescribed unit. The question for me is: what are the incentives for the national competition authorities to cooperate actively, positively, within the network? Within the new enforcement system they can expand their investigative powers, and that is an asset. There will be exchange of information and exchange of best practices, and that is also an asset. There will also be an educational role of the network, particularly for the competition authorities of the new Member States, which is tremendously important. And then, as Alexander Schaub mentioned before, the national competition authorities will strengthen their own positions within their national institutional environment. Now, all of these are benefits for the members of the network, but when one normally gets a network going there is a sort of day-by-day necessity to cooperate. I think that so far the network is largely a legally driven concept, and I do not see quite yet how those day-by-day active incentives are going to work.



䉴 DAVID GERBER — I wonder if Mario Siragusa’s question, which got us started on this debate, might not depend to some extent on this notion of ‘competition

38

The Network Concept, Competition Authority Networks

culture’ that has often been talked about. Some people think it means one thing, and other people think it means another. If we apply it to the interactions between the members of the future network on a horizontal basis — i.e. not including the Commission — then we see that they agree on many things in many ways: on the interests shared, the legal texts that are used, the interpretative methods, etc. One of the extraordinary things in the history of European competition law is the extent to which the system has developed, starting from a situation where there was almost nothing to work with. So, there is not much doubt in my mind that the common cultural factors will be a solid basis for the future network. There will still be some problems, but the common foundations are there. Then again, one of the issues that we have not touched upon so far, and which, I think, plays an important role, is the eastern enlargement. As I said before, there is by now fairly good agreement on so many aspects within the existing group of competition authorities, but undoubtedly developing that competition culture within [the] new entrants’ group will be more problematic. I think that this is something to be considered in the design of the future network. 䉴 IAN FORRESTER — If this debate were conducted within a single country, or within a more homogeneous continent, I think it would be a great deal easier. One of the problems is that there are large discrepancies among the EC Member States in terms of civil service tradition, resources, acceptance of the merit of competition law enforcement, and the discrepancies will increase following enlargement. I wonder to what extent such issues are perceived as problematic within the NAAG, for example. I wonder if the US State Attorneys General share the same values, resources, civil service traditions and so on.

JOHN FINGLETON — I think that there are two questions that arise when you get rid of a monopoly, like the Commission does in this regard. The first is: who does what under the new system? The second is: how will they do it? I do not think that what the political scientists say about power is so much important as the dynamic of the system we are about to create. One of the optimistic points is that in time there will be convergence within the network on how we do things. If that happens, then the question of, for example, case allocation, and who does things, is simply going to be solved on the basis of efficiency criteria, and not on the basis of the possible manipulation of the outcome. Also, it is true that the national competition authorities are going to get a lot more power, but once the new regulation comes into force, that would be a fait accompli. Yet there will still be a power play within the national institutional systems, and I do not think we should ignore this in our debate. From an Irish point of view, much of the disquiet with respect to the reform proposal is related to our national procedures. At present the Irish competition authority has a similar role to the US Attorney General: we prosecute and go to court, but we cannot impose sanctions on our own (even fines). My vision is that Ireland is going to have to harmonise at some



Panel Discussions

39

level with the rest of Europe, so we should have a more collective vision of where we are going to with this reform. ALEXANDER SCHAUB — I just wanted to reassure you about the situation in the candidate countries — I am talking here specifically about antitrust enforcement; State aid rules enforcement is a different matter, but that is not an area where national competition authorities are involved. The way in which the competition authorities of the candidate countries have progressed over the last six to seven years is really remarkable, and this is also thanks to the efforts of the not-yet-formally-existing network. We now have every year a conference with them, and this is not just a diplomatic meeting, but a real working meeting, where the top officials and the working level are coming together, and where there is a policy debate. One can see how year after year this dialogue becomes ever more professional, intensive and mutually valuable. The competition authorities of the current Member States have been directly involved in this process through the so-called ‘twinning exercises’, which practically involved the dispatching of national experts from the competition authorities of the current Member States to those of the candidate countries. I believe that the candidate countries will be much better prepared in this respect than most of the current Member States at the moment of their entry in the Community. This fact is also illustrated in the accession negotiations: in the past, the accession negotiations on the so-called ‘Competition Chapter’ were mostly about whether the candidate countries had legislation which could be considered as more or less compatible with the EC competition rules. That was, of course, not very satisfactory. Nowadays we examine for every candidate country not only the existence of national legislation in compliance with EC competition rules, but also whether they have enforcement authorities, whether these are they sufficiently well equipped, professional and independent, and third — and most importantly — what is their enforcement record. Only when these three criteria are fulfilled are we prepared to close the Competition Chapter in the accession negotiations. In concreto, we examine the antitrust decisions taken. It is not by chance that, until now, we closed the competition chapter of the negotiations only with four of the candidate countries. In earlier accession negotiations the Competition Chapter was totally irrelevant, and was closed almost automatically. This is the first time that competition policy is given such importance in the accession negotiations. This also reflects the increasing importance and recognition of this policy area in the broader political agenda of the European Union. Coming back to the issue of incentives to cooperate within the network, a similar question has arisen in relation to our bilateral cooperation agreement with the US antitrust authorities. In this case one could have also asked, what are the incentives for the US, who are the first world power, to cooperate with us? Or, what is our interest to cooperate with them? But, as we have discovered, in reality there is a strong incentive to cooperate.



40

The Network Concept, Competition Authority Networks

BARRY HAWK — Regarding Ian Forrester’s question concerning the NAAG, there is certainly far more commonality among the NAAG members in terms of the so-called ‘common competition culture’ than among the members of the future EU network. In the US the State and federal antitrust laws are by and large the same. What differs radically is the resources: some States may have one person in the antitrust division, and others are better staffed, but in practice this does not make much of an impact. And last but not least, there are certainly very different egos at stake. Remember that the position of State Attorney General is one for aspiring governors, so this is all very political.



䉴 JAMES RILL — I just wanted to add another gloss on Barry Hawk’s answer, because I think that the similarity of outcome between the NAAG and the EU network is somewhat over-emphasised. One thing that is not going to be readily transposable to the European experience, fortunately, is the fact that in the US the antitrust officials of several States are elected, a fact that on occasion produces a rather curious effect on antitrust enforcement. To give just two examples: the Westpoint/Peppel merger was approved in North Carolina on condition that the factory be kept open; the Russell Stover Chocolates merger was approved in Pennsylvania on condition that a given quantity of the raw material be purchased from local suppliers. One would wonder about the overall consumer welfare considerations behind such decisions.

I David J Gerber* The Evolution of a European Competition Law Network

Networks necessarily have a time dimension. They consist of relationships that take shape and function over time. The European Commission’s proposals to ‘modernise’ the EU’s competition law system call for the establishment of a ‘European competition law network’ that will perform key functions within the system.1 In assessing these proposals, the time dimension thus takes centre stage. This paper examines that dimension by locating the Commission’s network proposals within the development of European competition law. A European competition law network should not be understood as something new. Although the term ‘network’ has only recently come into common use, relationships among competition law officials have shaped competition law decisions since the inception of the system, and this is the core idea of ‘network’. Europe has thus had experience with competition law networks, and this will condition the operations of the proposed network (‘PNW’) in important ways. It also provides a basis for thinking about how to structure and operate the PNW most effectively. The proposals carry much potential, but they also entail risks that deserve careful analysis. This paper has three main objectives. One is to identify the changes that will be brought about by the proposals. Effective assessment requires analysis of the ways in which the PNW differs from past and current competition law experience. A second goal is to explore some possible implications of these changes for the proposals. How, if at all, they should be modified? What strategies should aid in achieving their goals? And what we can expect from them? Finally, I consider how we talk about networks. The network concept has become fashionable in discussions about Europe’s future, and much potential value is ascribed to it. However, the idea is often so vague that it carries little useful content. If networks are to function as many hope they will, we need to develop a language for talking about what they are, what they do, and how they do it. I look first at the concept of ‘network’ used in the proposals and explore ways of analysing it. I then use these analytical tools in looking at three models of network: the foundational model, the current model and the model proposed by the Commission. Finally, I suggest some implications of this analysis for shaping and implementing the PNW.

* Professor, Chicago-Kent College of Law, USA. 1 European Commission, Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, COM (2000) 582 (Sept. 27, 2000).

44

Working Paper I

I.

The Network Concept: Framing the Subject

1.

The Concept in the Modernisation Proposals

My assignment here is to explore the ‘network’ concept as it is used in the proposals.2 That concept has three main features. First, it is defined formally by reference to employment status. A competition law official in the Commission or in a Member State competition authority (national competition authority: ‘NCA’) is automatically a member; others are excluded. Second, it refers to relationships among these officials. Third, it is unitary: all EU and Member State competition law officials are conceived of as being part of the same set of relationships.3 The proposals ascribe several functions to the proposed network. One is to reduce the Commission’s burdens by eliminating or transferring from it tasks that it currently performs. The most prominent of these burdens is the review of applications for exemption under Article 81(3) EC. This is an efficiency claim about how the EU should use its resources, but it is based on claims about the capacity of the PNW to achieve particular results. Three such claims are central. The first is that the PNW will improve the enforcement of competition law. Local competition officials are said to be closer to markets than the Commission can be, and thus the enforcement of the law should improve if they assume greater enforcement roles.4 A second claim is that the PNW can achieve better voluntary compliance with the competition laws, primarily by developing and promoting what the proposal calls a ‘competition culture’.5 Third, the PNW is expected to improve the quality of individual competition law decisions. The rationale is that there will be more of them; they will be made in more varied contexts; and they will thus create a richer and more valuable body of case-law.6

2.

Analysing Networks

The analytical value of this network concept is limited by its static and formal definition. It does not relate directly to the operations of the network, and it does 2 Another

use of the network concept is more inclusive, but also more focused. It includes in a NW all who regularly interact in interpreting and applying a particular set of laws. This would permit, for example, the inclusion of judges and academics. 3 For detailed discussion of the proposals, see Ehlermann C.-D. and Atanasiu I., eds. (2001): European Competition Law Annual 2000: The Modernization of EC Antitrust Policy, Hart Publishing, Oxford and Portland, Oregon. 4 European Commission, Explanatory Memorandum accompanying the Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, p.6 (hereinafter ‘Explanatory Memorandum’). 5 Id. at 9. 6 See Wils W. P.J.(2001): ‘The modernisation of the enforcement of Articles 81 and 82 EC: a legal and economic analysis of the Commission’s proposal for a new Council Regulation replacing

The Network Concept, Competition Authority Networks

45

not, therefore, provide a basis for analysing the decisions that comprise those operations. In turn, this obscures the causal relationship between the network and the benefits that it is intended to provide. I here employ the network concept in a more precise way that refers directly to its operations. It contains three basic elements. First, for our purposes the network consists of relationships; the term refers to how individuals and institutions relate to each other. Secondly, the analysis looks specifically at only one aspect of those relationships: how they influence decisions of those within the network.7 Relationships have many aspects, but we are concerned with them only insofar as they relate to decisional outcomes. And thirdly, we are concerned only with particular categories of decisions: those that make, interpret and enforce competition laws. Therefore, our analysis focuses on how network relationships influence the decisions of network members. This focus on decisional influences allows us to ask more precise questions about the network and its operations. We ask two sets of questions. One relates to the structure of the network. We need to know how it is organised: who influences which decisions? This leads to a basic distinction. Members of the network make and influence decisions with regard to two separate sets of competition laws: those of the EU and those of its Member States. These represent what we can call ‘spheres of operation’.8 The ‘EU sphere’ refers to network relationships as they relate to EU law, while the Member State sphere includes those relationships insofar as they deal with Member State law. The distinction is essential to an effective analysis of the network. Each sphere has its own internal structure. We need to know what kinds of relationships exist among participants within it. We will ask questions such as: ‘What kinds of subgroups are there?’ and ‘What are the lines of authority and the positions of status within it?’ Some structuring factors are formal and institutional; others are personal and social. The other set of questions relates to network operations. Two functions are of primary concern: the flow of information through the network, and the normative authority that network relationships exert on decisions within it. Often these two factors are not clearly distinguished from each other, but the distinction is important for analysing the proposals. Claims about the transmission of information are central to the proposals, and we ask several questions about that process. First, what does the network Regulation No. 17’, in Hawk B, ed., 2000 Fordham Corporate Law Institute, Huntington, New York. 7 This

form of decisional analysis is discussed further in Gerber D. J. (1998): ‘System dynamics: toward a language of comparative law?’, 46 American Journal of Comparative Law 719. 8 The term is important, because it emphasises the decisional focus of the analysis. The issue here is not who is part of the series of relationships that constitute the network, but to what extent members of the network influence decisions relating to a particular set of laws — either national or EU. An official may participate in both spheres of operation or merely in one. The term ‘sub-network’ could also be used, but it would focus attention on formal membership, and it would fail to capture the decisional element of the analysis.

46

Working Paper I

transmit? What are the characteristics of the information? It might include specific information (such as full reports of cases, legislative rules, industry data and policy guidelines), but it might also include general principles, values, expectations and interpretations of experience. Second, how much does it transmit: what quantity of information? Third, in which directions does the information flow: who sends and who receives? Fourth, how is information transmitted: what are the methods of transmission? Finally, what factors affect the reception and use of the information by the intended recipients? Equally important is the flow of normative authority through the network. By ‘normative authority’, I simply mean the capacity of the network to shape competition law decisions and influence decisional outcomes. The transmission of information is intended to affect decisions, and it is thus necessary to understand the mechanisms by which the network expects to achieve that result.

3.

Modelling the Network

The concept of ‘model’ is a valuable tool for capturing the key elements of the story. Its use necessarily oversimplifies complex situations, but it has the benefit of clarifying lines of development and changes over time. I do not use ‘model’ as a theoretical construct. I have not begun with an abstract explanatory model and applied it to the evidence. I instead use it as an empirical construct to capture patterns, and pattern changes over time. The network has operated according to three basic models. A foundational model evolved in the early years of European integration. It was characterised by virtually autonomous spheres of operation. There were minimal institutional links, and the exchange of information and ideas on the basis of personal contacts and informal authority was its most salient feature. As European integration proceeded, it precipitated a gradual change in the operation of the network. This produced what can be called a ‘solar model’, in which the spheres are partially integrated but remain largely distinct and separate. The modernisation proposals call for changes that will produce yet a third model of network operation, which is characterised by a high degree of integration between the spheres.

II.

The Foundational Model: Separate Spheres

The foundational model was conditioned by the political and institutional circumstances of the first decades of EU integration. It remained more or less in place until about the mid-1980s. As a result, it set the basic patterns of operation for the network, and it is the form of network that has operated the

The Network Concept, Competition Authority Networks

47

longest and with which many competition law decision-makers have the most experience.9 In this model, the officials of each competition law system operated almost exclusively in their own spheres (EU or Member State). Virtually all decisionmakers were concerned almost exclusively with the development and application of their own legal regimes. They sometimes transmitted information to officials in other systems, but such contacts were limited and sporadic, and they were often based on personal relationships. They were rarely formalised or institutionalised. This network structure was consistent with the political and economic circumstances in which the network operated. In the early stages of integration, European economies were largely national in structure, and national governments operated largely autonomously. A network such as the one currently proposed would have been almost unthinkable.

1.

The EU Sphere

The EU sphere initially operated in a highly uncertain context. The process of integration was halting and uncertain, and many doubted its future. At times it moved forward, while at other times it stagnated. As late as the early 1980s, the references to ‘Eurosclerosis’ — the unpromising condition of a bureaucratically hampered community with limited forward momentum and little internal dynamism — were common. This conditioned the development of competition law in many ways. Perhaps most important was the urgency it imparted to the enterprise of competition law, which often served as a principle tool for moving the process of integration forward.10 The European Court of Justice repeatedly established broad principles of competition law that reduced the capacity of Member States or private firms to segment markets along national lines, thereby also maintaining the forward momentum of the integration process. The EU sphere consisted primarily of employees of DG IV (as DG Competition was then called). In general, only they made or influenced EU decisions; Member State officials had very limited influence on those decisions. However, there were two exceptions.

9 For perspectives on this period, see, e.g., Gerber D. J. (1998): Law and Competition in Twentieth Century Europe: Protecting Prometheus, Clarendon Press, Oxford, at pp. 346–358; Hawk B. E. (1972): ‘Antitrust in the EEC — The first decade’, 41 Fordham Law Review, pp. 229–292; Goyder D. G. (1998): EC Competition Law, 3rd edition, Clarendon Press, Oxford, at pp. 34–74 and 561–586; and Holley D. L. (1993): ‘E.E.C. competition practice: A thirty-year retrospective’, in Hawk, B. E., ed., 1992 Fordham Corporate Law Institute, Huntington, New York, pp. 669. 10 For discussion, see Gerber D. J. (1998/2001), supra note no. 9, at pp. 351–64; and Gerber D. J. (1994): ‘The transformation of European Community competition law’, 35 Harvard International Law Journal 1, pp. 97–148.

48

Working Paper I

One was the Advisory Committee on Restrictive Practices and Dominant Positions, which meets to review important decisions of the Commission before they are finalised. Although this group has undoubtedly given valuable comments to the Commission in some situations, its influence has generally been marginal. A more important source of influence was informal. The Commission was developing EU competition law, and it sometimes turned to Member State officials for advice or information. As a result, some officials acquired significant influence on decision-making and became significant parts of the EU sphere, but their membership was primarily personal rather than institutional. German officials were the most prominent, particularly in the early years. Germany had introduced a competition law in 1957 after almost a decade of intense debate and analysis, and the German Federal Cartel Office was assiduously developing competition law in Germany during the same period that the Commission was developing EU competition law.11 Moreover, no other Member State had comparable experience with competition law during this period. Naturally, therefore, German officials were accorded roles relating to EU law that were seldom available to officials from other states. The EU sphere was primarily structured by the official status and authority relationships within DG IV. To the extent that it included members from outside the Commission, their status was based largely on reputation for knowledge or expertise. The amount of information transmitted among the members of this group was limited, primarily because there was relatively little information to transmit, particularly in the early years. In the competition law area, the acquis communautaire developed slowly during its first decade. There was little case-law, and there were relatively few well-developed rules and policies. For example, the block exemptions are an important feature of EU competition law today, but they became a significant factor only late in the period. When information was transmitted beyond the boundaries of DG IV, it generally flowed through informal channels, and it was usually directed to specific needs of the Commission. The Commission generally did not provide regular information regarding its activities to the Member States other than in the limited, formal context of the Advisory Committee. When EU decision-makers sought information and advice, it was typically with a specific purpose. This meant that the information and its normative influence were often inseparable. Member State officials were included in the network precisely because of the expected value of their knowledge or skills. As a result, they were expected to influence outcomes, and the information they transmitted automatically had status and often a degree of urgency. The NCAs transmitted little information to the Commission or to other NCAs.

11 For

discussion, see Gerber D. J. (1994): ‘Constitutionalizing the economy: German neo-liberalism, competition Law and the “New Europe” ’, 42 American Journal of Comparative Law, pp. 25 et seq.

The Network Concept, Competition Authority Networks

2.

49

Member State Spheres

Each national sphere had its own statutes, procedures and institutions. Although there were broad similarities among them, they often differed in important ways from each other and from EU law and procedure.12 Each identified its own policy objectives in the light of national political traditions and economic circumstances, and each operated within a unique juridical space. In the early years of competition law development, some Member States either had no competition authority or one that lacked any significant role in legal and economic life. Except in Germany, these authorities were initially small, under financed and with little political or popular support. However, this was changing rapidly by the 1970s. Member State competition law systems were becoming increasingly important, increasingly juridical and increasingly independent of domestic political control.

3.

Relationships between the Spheres

EU and Member State spheres were officially related in three basic ways. One was on the above-mentioned Advisory Committee, where the representatives of the NCAs had to become familiar with EU law and with the objectives and decisions of the Commission, and where they were expected to take positions regarding them. They were also related jurisdictionally. When the Commission took action to enforce EU law, NCAs were required not to interfere with those actions.13 This meant that NCAs had to take into account Commission policies and decisions in making their own policy and enforcement decisions. Finally, they were related by the didactic, educational role of the Commission. Particularly during the 1960s, the Commission considered the education of Member State governments regarding competition law matters to be an important part of its responsibilities.14

III.

The Current Network: A Solar Model?

The current competition law model took its basic shape during the 1980s. In it, the EU and Member State spheres have moved closer together. They still

12 See Gerber D. J. (1998), supra note no. 9, at ch. 13 The relationship between the Commission and

6. Member State laws was given lasting structure in Case 14/68 Walt Wilhelm v. Bundeskartellamt [1969] ECR 1. 14 See, e.g., Graupner F. (1973): ‘Commission decision-making on competition questions’, 10 Common Market. Law Review 291.

50

Working Paper I

generally operate separately, but they have become more closely related. Their relationships can be represented by a solar model. The Commission is the centre of the system. It makes decisions that affect all NCAs, but it gets little information from them and generally pays little attention to them. It sets the constraints within which they must operate, but it does not interfere directly with their operations. The NCAs operate independently, and they have relatively fixed relations with each other and with the Commission. They are firmly tied to the central source of power and influence, but their role is largely passive with respect to it. They are expected to heed information from the Commission, and to receive and act on its directions. What the Commission does is often important to NCAs, particularly if their national legislation tracks EU competition law.15

1.

The Context

This model began to evolve in the mid-1980s as the pace of European integration increased. During this period the process of European integration was re-energised under the Commission presidency of Jacques Delors. The Single European Act in 1986 signalled new levels of both political and economic integration, and this changed expectations regarding the role of the EU and its future. In its wake, the idea that competition law decision-makers would operate in isolated spheres seemed increasingly untenable. National competition laws have gradually converged in response to this political and economic integration. The new states that entered the Union in the 1990s generally copied the competition law legislation of the EU, and existing Member State legislation has also been frequently revised to reflect EU law. In 1990, for example, Italy introduced its first competition legislation, which it modelled to a large extent on Community legislation.16 The introduction of an EU merger control regime in 1989 was a further step toward convergence and centralisation.17 It seemed to many to herald a future in which competition law would be centralised in Brussels, as much of merger control now was. Some even anticipated that Member State competition laws and Member State competition authorities would disappear. But the direction of development began to change in the early 1990s. Two factors were primarily responsible. One was the growing recognition that the 15 In Sweden, for example, current legislation is almost identical to EU law, and it specifically requires

that the text be interpreted on the basis of the same sources used in applying EU law. See Bernitz U. (1993): Den Nya Konkurrenslagan, 2nd edition, Stockholm. 16 For discussion, see Donativi V. (1990): Introduzione della Disciplina Antitrust nel Sistema Legislativo Italiano, Giuffrè, Milano. 17 Commission Regulation 4064/89, OJ L 257/14 [1990] (corrected version of OJ L 395 [1989])]. For discussion, see Mestmäcker E.-J. (1989): ‘Merger control in the Common Market: Between competition policy and industrial policy’, in Hawk B., ed., 1989 Fordham Corporate Law Institute, Huntington, New York, ch 20.

The Network Concept, Competition Authority Networks

51

demise of the Soviet Union was likely to lead to inclusion of former Warsaw pact members as members of the Union. The other was growing popular dissatisfaction with bureaucratic centralisation and resentment toward too much interference from Brussels. This led to the introduction of subsidiarity as a fundamental structural principle of European integration. In recognition of these two factors, the Commission began to reverse the centralisation process in the competition law area by increasing the role of NCAs in the application of EU law. In 1994, DG IV issued a call for increased participation by Member States,18 and in 1997 it issued more comprehensive guidelines defining the respective spheres of competence of the NCAs and the Commission.19 The primary focus was on inducing more of those aggrieved by arguably anticompetitive conduct to go to the national courts for relief rather than to Brussels. These calls did not yield a significant increase in the use of national courts, but they did begin a process of increased interaction between the Commission and NCAs.

2.

The EU Sphere: Integration and Authority

The structure and operations of the network reflect these changes in the legal and political context of competition law. We look first at that portion of network operations that relates to EU law. It now includes more participants, and it carries more information of more kinds that is used for more purposes. It is also performing functions that had not been conceived before the 1990s.

2.1.

Participants and Structure

The size of the network has grown significantly. Expansion of the EU to 15 members during this period has significantly increased the number of officials that are part of the network. In addition, existing authorities have generally grown substantially during the period.20 The formal structure of the EU sphere now includes additional authority for NCAs to apply EU law, even though this authority is seldom used and some states have not yet authorised their authorities to apply EU law.21 However, the authority structure is little changed in other respects, with the Advisory Committee continuing to provide the most direct NCA input into EU decision-making. 18 For

discussion of these efforts, see Ehlermann C.-D. (1986): ‘Implementation of EC competition law by national antitrust authorities’, 17 European Competition Law Review 88. 19 Commission Notice on Cooperation Between National Competition Authorities and the Commission in Handling Cases Falling Within the Scope of Articles [81] or [82] of the EC Treaty, OJ C 313/3 [1997]. 20 For discussion of this process, see Gerber D. J. (1998), supra note no. 9, at ch 10. 21 At this writing those countries are: Austria, Finland, Ireland, Luxembourg and the UK.

52

Working Paper I

The informal structures based on experience, expertise and personal relationships that were often important during the foundational period have become generally less so, basically because the Commission has less need for specific individual expertise from NCA administrators. The introduction of merger controls temporarily increased the need for specific expertise in that area of competition law for the first few years after its introduction, when the merger task force included many officials on loan from members states (again, the knowledge and experience of German officials made German influence predominant).22 However, that need has also diminished.

2.2.

Process

There is also significantly more information transmitted through this expanded set of relationships. The Commission more frequently informs NCAs of its activities and policy views, primarily in the context of specific factual situations but sometimes also on a more general level. The most dramatic increase in information flow has occurred in the process of evaluating and discussing the modernisation proposals, which is a process that has now lasted for several years.23 The normative force of network relationships remains little changed. The Commission basically directs developments involving EU law. It expects the Member State participants to listen and adhere to its instructions and guidelines, but they are not expected to wield influence. Again, the modernisation negotiations represent a special situation in which the Commission needs political support from Member States in campaigning for the changes it wants, and it must therefore consult extensively with Member State representatives about the specifics of the proposals.

2.3.

Member State Spheres

As noted above, NCAs have generally become significantly larger and more influential than they were during the foundational period. The last of the original Member States to create a competition authority was Italy, but within a short period of time the Italian competition authority became a formidable and highly respected institution.24 Moreover, more officials from each authority are now involved with EU law in one-way or another. 22 For

discussion, see Krause H. (1995): ‘EC merger control: An outside view from inside the Merger Task Force’, 39 Journal of Business Law 627. 23 For extensive discussion of the modernisation programme, see Ehlermann C.-D. and Atanasiu I., eds. (2001): European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, Oxford and Portland, Oregon, and Ehlermann C.-D. (2000)” ‘The modernization of EC antitrust policy: A legal and cultural revolution’,37 Common Market Law Review 3, 537 et seq. 24 For discussion, see Amato G. (1998): Il Gusto della Libertà: Italia e l’Antitrust, Laterza, Roma-Bari.

The Network Concept, Competition Authority Networks

53

The NCAs continue to operate primarily as separate spheres of decision-making. Neither the Commission nor other Member States have any significant role in or influence over the internal decisions of a NCA, except insofar as the latter is aware that the Commission has the ultimate authority to remove a case from its jurisdiction by commencing proceedings of its own. NCAs do not transmit significant amounts of information to either the Commission or other NCAs.

2.4.

Toward an Integrated Model

Tentative first steps have been taken toward integrating the EU and Member State spheres of operation. The members of the network have generally grown increasingly interdependent. Moreover, the trend during the 1990s for Member States to adopt EU substantive law has created a situation in which most or all decision makers share basic principles and ways of thinking about competition law issues, and they often share the same texts. However, European political and economic integration has yet to be given full expression in the relationships among those who make and enforce competition laws within it. There is currently a very limited flow of information between them, and there is limited coordination of activities. This sets the stage for the modernisation proposals.

IV.

Modernisation: A Centralised Interactive Model

The network envisioned by the Commission is similar to these earlier models in important ways. It evolves from them and would be inconceivable without them. Yet it also represents fundamental change. This section looks at the relationship between the PNW and its predecessors. The PNW represents what can be called a centralised interactive model.25 It is centralised in the sense that it is, for most purposes, a unitary system that is directed by the Commission. Except in the limited situations in which NCAs apply their own laws, all members of the network are primarily concerned with the interpretation and application of EU law. This means that the Commission directs decision-making throughout virtually the entire network rather than merely within a limited EU sphere. The PNW is interactive in the sense that all 25 I am tempted to call this a ‘ship model’. A ship is a single unit in which there are many separate operating departments, but each one is ultimately subject to the direction of the ship’s commander, and they must function together with respect to a common plan and set of procedures. This is what the Commission envisions for the network. For insightful and here remarkably relevant analysis of the ways in which the members of a ship’s crew relate to their common objectives and rules and to the directions of the ship’s officers, see Hutchins E. (1995): Cognition in the Wild, MIT Press, Cambridge, Massachusets.

54

Working Paper I

members are to react and respond to each other. Information and influence are intended to flow multidirectionally through the network.

1.

The Modernisation Process

The PNW will operate in a context that differs — in some ways dramatically — from earlier contexts. Political and economic integration has significantly increased in the late 1990s, with the successful introduction of the Euro and the recently convened ‘constitutional’ convention as symbols of this integration. The Member States have come to accept and operate within a system of highly interdependent relationships, and this interdependence is becoming more widely recognised and fully accepted. However, the most dramatic change of circumstances will come from the Union’s planned expansion. Membership could increase by as many as 10 or more over the next few years, and much of the impetus in the modernisation proposals comes from the claim that the current system cannot function in such a large Union. This expansion also differs from previous expansions in two important ways. First, the expansion members generally differ markedly from current members with regard to economic structure, political and legal traditions, and experience with competition law. In contrast, previous entrants tended to be largely similar in these ways to existing members. In particular, they already had competition law systems, some of which were — as in Sweden — well developed. The second major difference between this expansion and previous ones is that a large number of members are likely to enter the Union in a very short space of time.

2.

Network Membership and Structure

In the PNW the distinction between EU and Member State spheres will continue to exist, but it will lose much of its importance because most network members will deal primarily or exclusively with EU law. In this sense, the competition authorities of the Member States and the Commission will become far more highly integrated. There will be some residual role for Member State competition laws, but under the current proposals it will be quite small. The network will become integrated at a time at which membership is likely also to increase significantly. Each new Member State will bring with it a new competition authority whose members will become part of the network. Moreover, the increased obligations of Member States under the proposals are likely to lead to an increase in the number of officials they employ. This rapid increase in the number of members will put additional strains on network relationships.

The Network Concept, Competition Authority Networks

55

The structure of the PNW will also be significantly more complex. For example, the new members will initially form an identifiably separate group. Formally they will have the same status as existing members, but they will have less informal or operational status because they will have less experience and knowledge that is relevant to competition law and the operation of the network. They will also be distinguished by differences in their economies and their general political and legal experience. The economies in which they operate will generally be less developed than those of existing members, and the political traditions and experience of most will continue for some time to be influenced by their recent history under Soviet dominance. In addition, their experience of market economy, Western European legal systems, international competition, and particularly EU institutions and operations, will differ markedly from the experience of current members.26 The separateness of the group will be reinforced by the fact that its members are likely to enter as a block rather than one or two at a time as has been the pattern in previous accessions. Integration of the EU and Member State spheres also means that each NCA will become a structural element within the EU sphere. Each has its own interests and its own economic and political realities, and these will become part of the decision-making process of EU law for the first time. Issues such as funds for performing network obligations now become directly relevant to the process of applying and developing EU law, as does the relative political and economic power of the Member States themselves. These factors have little relevance for EU law while the NCAs operate in their own separate spheres, but they will take on new significance when the network is integrated. Expansion of the network is likely also to generate informal sub-networks in which information will flow more freely, communication will be more frequent, and mutual attention will be higher than with respect to the rest of the network. Language will play a role in fashioning these sub-networks, as will regional and local traditions and interests.27 Formal authority will still be an important structuring factor. It will continue to be concentrated in the hands of the Commission, and it will now reach into far more activities and decisions by far more network participants. NCAs will have authority to interpret and apply EU law within their areas of competence, but the Commission will have the authority to override their decisions.

3.

Transmitting Information

The most prominent change in the operation of the PNW will be a vast increase in the amount of information flowing through it. The proposals make information 26 For

discussion, see Fingleton J., Fox E., Neven D. and Seabright P. (1996): Competition Policy and the Transformation of Central Europe, CEPR, London. 27 This can already be seen, for example, in intensified cooperation among the Nordic Member States in competition law matters.

56

Working Paper I

flows a key part of the network’s operations, and they create numerous new obligations to transmit it. These obligations fall primarily on the NCAs, who will be required to inform the Commission of many of their activities, especially if they intend to make particular types of decisions. NCAs will also be required to transmit extensive information about their activities to each other. Information will also flow in many directions, thereby increasing the complexity of the transmission process. In the current model, information flows primarily from the Commission to the NCAs, but the latter have few general obligations to send information to the Commission and they have no obligations to send information to each other. The proposals envision a network in which extensive information will also flow from NCAs to the Commission and horizontally among NCAs. Information will flow through the network under different conditions. Information has hitherto generally been transmitted in response to particular needs, and often on an ad hoc and informal basis. Under the proposals, information will be transmitted primarily pursuant to formal obligations and in response to the exercise of formal rights to information. It may also be transmitted in specific procedural contexts that will require close cooperation between the NCAs and the Commission (for example, when the Commission wishes to present its case in a Member State court).28 The information that is transmitted will also often have different characteristics: frequently more procedural, more technical, and more specific. A large portion of it will be highly specific information about particular cases. The proposals require the NCAs to submit to the Commission, for example, detailed information about each case in which they plan to prohibit conduct, vary the applicability of a block exemption, or accept commitments.29 NCAs may also be expected or required to transfer third party information, that is, information that it acquires from private firms pursuant to its own procedures. In addition, the Commission has promised a new set of detailed regulations and guidelines designed to guide Member State decision-makers.30 Finally, the information will be received in a different context. Information is now generally transmitted only if its potential relevance for a particular purpose is established and there is thus a recipient interested in receiving and using it. Under the proposals, in contrast, much of the information that the PNW carries will not be targeted in that sense. It will be transmitted because it falls within a particular category described by general rules, and it will be received only to be sorted and processed. Officials may review it in order to determine its relationship to policy positions or pending decisions, but it may just be put in a database. This changes the role of the information and its transmission in ways that have yet to be analysed carefully. 28 European Commission, supra note 29 Id., Article 11(4). 30 Explanatory Memorandum, p.8.

no. 1, Article 15.

The Network Concept, Competition Authority Networks

4.

57

Normative Influence

The capacity of the network to exert normative force is in many ways the key to the proposals. As the Commission is aware, it is relatively easy to require the transmission of information and to monitor the extent to which participants have fulfilled their obligations in transmitting it. It is more difficult to ensure that the information has the desired effect on decisional outcomes. Yet information must affect decisions if the network is to achieve the desired results. The proposals envision the network having normative force, both formally and informally and in both vertical and horizontal relationships.31 The vertical component consists primarily of the Commission’s direction and control of the decision-making of NCAs. The Commission intends to exercise this formal authority in specific cases, as well as by issuing extensive and detailed regulations whose purpose is to tell the NCAs what they must do. It also intends to exercise informal authority in the context of increased personal interaction with Member State officials, but the proposals are vague as to how this is to occur. NCAs will have no formal authority vis-à-vis the Commission, but the Commission suggests that it will take into account the views of Member State officials in making its decisions. The proposals are vague about how this will be done, and there is concern among Member State officials that their influence on Commission decisions will be marginal. The Commission acknowledges that vertical influence will be inadequate to achieve the desired results. It intends that horizontal relationships among NCAs will add the necessary constraints and influence.32 The proposals do not include formal mechanisms for this horizontal influence. Instead, the Commission seems to view it as an informal process that will result from participation in the network itself. It anticipates that the increased interactions of Member State officials among themselves and with Commission members will shape decisional outcomes. It is in this context that it refers to the idea of a competition culture that will naturally tend to produce convergence in decision-making, but it provides little indication of how it expects this to work.

5.

Member State Authorities — Residual Roles

Even under the proposals there will be residual individual spheres of operation for the application of Member State competition laws. Those residual spheres

31 See

gen. Monti M. (2001): ‘The modernisation of EC antitrust policy’, in Ehlermann C.-D. and Atanasiu I, eds., European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, Oxford and Portland, Oregon, pp. 3–13. 32 Schaub A. (2000): ‘Modernization of EC competition law: reform of Regulation 17’, 23 Fordham International Law Journal 752, at p. 764.

58

Working Paper I

will be particularly limited if Article 3 of the draft Regulation, which calls for exclusive application of EU law when conduct affects trade between the Member States, is enacted.33

6.

Continuity and Change

This brief review reveals the extent of both continuity and change in the network proposals. The PNW shares important features with previous and current versions of the network. It will continue to operate as part of a largely administrative system in which private law mechanisms play a limited role. Although the proposals encourage private suits in national courts, it remains to be seen whether such suits will increase, much less become common. The Commission will continue to be the dominant institution within the network. It will have the authority to control the flow of information and to determine who handles cases and what the substantive outcomes should be. The proposals increase the Commission’s dominance in the sense that it will control a largely integrated network that includes all European competition law officialdom rather than merely one of many decision-making spheres within that network. The network will still consist of administrative agencies that are not organisationally integrated. The Commission and the NCAs will remain institutionally separate. The proposals call for changes in function and relationships, not changes in formal organisational structure. The PNW will continue to transmit information and exercise influence across the boundaries of these institutions. However, the changes are also extensive. Perhaps the most fundamental is that the network will become a formal legal concept. It is part of the modernisation plan and, as such, it will have official functions and be understood and evaluated in terms of its performance of those functions. In short, it will be institutionalised, and will therefore be encumbered with rights, duties, expectations and power positions. In part as a result of this institutionalisation, pressure on the network’s structures and operations will increase significantly. It will be asked to perform far more than it has performed in the past. Vastly more information will pass through it, and it will move in more varied and complex directions. In addition, it is expected to exert normative influence in new and more extensive ways, many of which are untried and unclear. This means, in turn, that businesses and political entities will attach more value to network relationships and to the many decisions that will constitute its operations, and they can be expected to expend correspondingly more resources to influence them. 33 At

the time of writing there were reports of some doubts as to whether that Article would be enacted.

The Network Concept, Competition Authority Networks

V.

59

Some Implications of the Proposed Changes

The modernisation proposals represent a bold step in the development of competition law in Europe, and their success will depend to a significant extent on how the proposed network of competition law authorities functions. As we have seen, a network has operated throughout the development of European competition law, and experience with it provides a basis for assessing the proposals. That experience will condition the network’s operations, particularly in its early years, and it yields insights into what can be expected from the network.

1.

Conditioning Factors

One conditioning factor is the trajectory of competition law development. Competition law in Europe has become steadily more important in business and legal life throughout the Union, and this has increased the status and influence of competition law authorities and the officials in them. Moreover, the relationships among these officials have become closer, more extensive, more effective and more important. This experience colours both the proposals and the reasoning behind them, imbuing them with an unmistakable sense of optimism. The tone is: if the network can develop this far this fast, we can expect to develop it further along the same lines. The justifications for the network proposals reflect confidence that it can bear much increased demands and perform new functions, and that the Commission will be able to direct the flow of information and influence within the network. Yet the impending changes impose significant costs on decision-makers and their institutions. All will have to operate as part of a dense web of relationships that places extensive new demands on their financial and human resources. Institutions that have operated more or less independently will be required to change their operations to respond to these new demands, and the costs of these changes will be a critical factor in the success of the PNW. Perceptions shaped by previous experience will influence how network members assess and respond to these added costs and burdens, particularly during the early stages of the PNW’s development. For example, Member State officials often believe that the Commission attaches little value to their views (except perhaps in the context of the modernisation negotiations) and that Commission officials view Member State officials not as equals but merely as enforcers of Commission decisions. The network proposals are built on the assumption that this perception can be changed, but changing it is likely to require sustained effort on the part of the Commission. One way of moving in this direction may be, for example, to increase the participatory rights of Member State officials in their dealings with the Commission, thereby anchoring their influence.

60

Working Paper I

Past experience also creates expectations that entrain their own incentives and disincentives, depending on the extent to which such expectations are met and how disappointments are perceived and handled. For example, German competition law officials have long played a particularly influential role in the network. This role and the status that accompanies it are of much value, but that value is likely to decline under the new system, particularly when coupled with expansion of the Union. Similarly, current members of the network might expect their opinions to be valued more highly than those of expansion members, but officials in the latter group might not always share this expectation. The resulting disappointments could generate conflicts and thereby burden network operations.

2.

Lessons?

European competition law experiences, in general, and the evolution of the network, in particular, provide some insights into the factors that are likely to influence the PNW. Drawing ‘lessons’ from any historical development is a matter of conjecture and interpretation, but it might nonetheless be of value.

2.1.

Responding to External Pressures

One such lesson is that the effectiveness of the network is likely to depend to a significant extent on its capacity to withstand economic pressures. Officials who apply and enforce competition law are everywhere targets of pressure from business firms because of the often considerable economic consequences of their decisions. The battle to resist such pressures has been a constant theme of European competition law development at both Member State and EU levels. The Commission and some Member State authorities have proven to be generally effective in dealing with such pressures, but this has not always been the case with some authorities. The increased size and enhanced importance of the PNW will create new targets for such pressure. Each of the many decisions that will have to be made about the transmission of information (what to send, how to send it, how to receive and store the information, and so on) and about the exercise of influence represents a point of vulnerability. Moreover, the decisions that shape substantive outcomes will become more numerous, more complex and often less transparent. This calls for designing network operations to protect those decisions from outside influence. The goal of creating a ‘competition culture’ should include creating a culture of resistance to such influence, and European experience has much to offer in that context. Political influence over network decision-making creates similar issues. The network will function effectively only to the extent that (1) officials are permitted to pursue the shared public interest goal of developing and enforcing competition

The Network Concept, Competition Authority Networks

61

law, and (2) they are widely believed to have that independence. The more independent they are and are perceived to be, the more likely it is that they will identify with common goals set for the network, and will seek to conform not only to their letter but also to their spirit. Transmitting information and exercising influence within the network involves making decisions that will often be voluntary and non-transparent, making outside influence difficult to detect. Therefore, independence from political influence cannot be limited to formal independence but should also include operational independence. This may include, for example, protection for officials who reveal attempts to exert influence and obligations of mutual assistance from other network members.

2.2.

Commitment

A key factor in the development of competition law in Europe has been the high level of personal commitment and engagement of administrative officials. The crucial early successes of the German Federal Cartel Office and other Member State authorities were due in large measure to the intense personal engagement of officials who believed — often passionately — in the importance of their project.34 At the EU level, competition law development has often been driven forward by officials committed to a particular vision of the role of the market in Europe and willing to invest energy and enthusiasm to realise that vision. This type of commitment is likely to be necessary to create and operate the kind of network envisioned in the proposals. Otherwise the pressures will be too strong and the costs of operation will be too high for it to succeed. An uninspired, routinised conception of competition law is not likely to generate high levels of commitment. The proposals will create extensive routine obligations to transmit, store and review information, and there is a risk that the weight of bureaucratic routine will overshadow and even blunt the necessary commitment. This suggests that network obligations should be calibrated to avoid excessive routinisation and to maximise opportunities and incentives for active commitment by all network members.

2.3.

Aligning Interests

Resisting external economic and political influences and developing high levels of commitment will depend, in turn, on aligning the perceived interests of network participants. To the extent that network participants perceive the same set of interests as guides to their decisions, they are likely to use the decisions of others as guides to their own decision-making and to reach generally similar

34 See,

e.g., Gerber D. J. (1998), supra note no. 9, at pp. 280–87.

62

Working Paper I

conclusions, and this coherence is critical to the success of the network. The greater the alignment of perceived interests, the greater also is the likelihood that network members will reinforce each other’s decisions and interpret intra-network messages in a manner consistent with the intent of the sender. This kind of alignment has been critical to the development of the network in the past, as leading decision makers have generally identified with a particular conception of competition law and its role in European integration. Increased pressures on the network could undermine this commonality of interests. There will be many more decisions to be made and many more possibilities for conflicts of interest, and such conflicts can reduce network effectiveness. One way of counteracting this tendency is to increase the participatory role of the NCAs. The more such officials are actively engaged in policy formulation and implementation, the more likely they will be to identify their own interests with the interests of other network members.

2.4.

The Structure of Information

European competition law experience also suggests that the structure of information that moves through the network will play a role in network operations. Where it is organised in ways that make it easily accessible and understandable, it will be more readily assimilated by those who receive it and more likely to influence their decisions. As the network transmits increased amounts of more detailed information, this structuring will become correspondingly important. Structuring the information — for example, by clearly identifying the categories into which it falls and the goals and principles to which it relates — may be more important than the quantity of the information transmitted in the operation of the network.

2.5.

External Support

Finally, external support for competition law has been critical in the development of competition law in Europe, and it is likely to be important in the success of the PNW. Political support is key because network relationships are influenced by the relationships around them. Officials within the network will be employees of separate institutions, whose policies and interests can influence their actions. Information will flow effectively through the network only if politicians provide the necessary resources and intra-institutional incentives for transmitting, receiving and using it. These institutions can also control the extent to which network members can put the goals and methods of the network above those of institutional and national interests. In general, network relationships will exert the expected normative influence only insofar as they operate within an atmosphere of cooperation and support from surrounding institutions.

The Network Concept, Competition Authority Networks

63

Similarly, relationships with intellectual communities have often been important in the development of the network. As competition law has developed in each Member State and within the EU, and as information and influence have passed through the network, experience and ideas developed within academic and practitioner circles have played key roles. Influence within the envisioned network will continue to depend to a significant degree on the perceived strength of arguments and the capacity of network members to provide convincing analyses, and external intellectual input will continue to be an important factor in fashioning both.

3.

Agenda

This review of the development of the European competition law network reveals areas that deserve further study. Most basic is the need for a more developed language for talking about networks and their operations. The network concept is often used in vague, imprecise ways that obscure more than they reveal. If the network is to play the central roles envisioned for it, those involved in it will need more sophisticated and widely accepted ways of talking about it. Social scientists have begun developing a language for analysing networks, and this work deserves further attention and application to the competition law context.35 The modernisation proposals are based on several sets of assumptions about networks that call for further analysis. One relates to the flow of information. The proposals assume that this process will yield a common competition law ‘voice’. They assume that the more information flows through the network, the clearer and more consistent the law will be, and the more effective enforcement will be.36 However, this assumption is far from well grounded. It requires additional assumptions about how information will be sent, received and used that have been little studied. Our developmental analysis provides some points for evaluating the assumption, but more is needed. The methods by which information is transmitted also call for more study. The proposals seem to assume that vast amounts of information can be transmitted through the network at minimal costs and in ways that tie the information directly to the decision-makers that are supposed to use it. That assumption is based, in turn, on assumptions about the capacity of information technology to create that result. Experience with information technology in comparable situations leaves room for doubt about such assumptions. At a minimum, they deserve more careful analysis. 35 For

valuable examples of political science analysis in the competition law area, see, e.g., Doern B. G. and Wilks S. (1996): ‘Conclusions: international convergence and national contrasts’, in Doern B. G. and Wilks S., eds., Comparative Competition Policy: National Institutions in a Global Market, Clarendon Press, Oxford, at p. 327, and Sturm R. (1996): ‘The German cartel office in a hostile environment’, in the same volume, at p. 185. 36 ‘More decision-makers also mean more case-law and administrative decisions, which will further clarify the scope of the competition rules’. Explanatory Memorandum, p. 9.

64

Working Paper I

A second set of assumptions refers to the normative force of the network. The proposals recognise that the Commission cannot effectively control or police all decision-makers within the network and that network relationships must therefore exert the requisite influences on network decision-making themselves. Information transfer is not enough to ensure that laws will be interpreted, applied or enforced in the ways that the Commission envisions. Decision-makers will often have interests and perceptual biases that could lead them in divergent directions, and the network is expected to have the normative force to counteract these obstacles to convergence and cooperation. This normative task might be more difficult than the Commission sometimes seems to expect, and, in any event, the normative capacity of network relationships deserves more study.

4.

The Network and European Integration

Competition law has always had a special status and played special roles in the process of European integration. It has often been used as a motor of integration, pushing the process forward when its impetus has stalled. A central question in evaluating the current modernisation proposals is whether they will enable competition law to continue to play this role. If successful in the context of competition law, the network concept is likely to be more widely applied in other areas of European integration. However, to the extent that it is seen as a failure, this experiment will cast doubt on many current plans and could have serious repercussions for the future of integration.

II Stephen Wilks* Understanding Competition Policy Networks in Europe: A Political Science Perspective

I.

The Interest in Policy Networks

In 1974 two Americans published a book about the British Treasury that came as a revelation to those political scientists who study public policy. The book was wonderfully revealing about the notoriously secretive world of the Treasury and its control of public expenditure, but it was the theoretical underpinning that generated shock waves. Heclo and Wildavsky presented the world of Whitehall administration as a ‘village’ in which similar people with similar world views and values functioned within similar incentive frameworks in annual cycles of work.1 They developed shared understandings, methods of cooperation and frameworks of compromise. The analysis presented a plausible picture of a ‘community’ or ‘network’, closed to outsiders, difficult to penetrate or to understand, and confident in its unchallenged control of one of the most powerful policy instruments in the British political system. The ‘village’ analogy owed more to insights from sociology and anthropology than to theories of democratic process or legal accountability. It offered some fairly basic propositions but spawned a fertile process of theoretical development, which included an influential article by Heclo and a series of books and articles.2 By the end of the 1980s the ‘policy network’ approach had come to dominate policy studies and was the preferred framework within which political scientists and their research students studied policy-making. Networks have always existed. Indeed, if networks are about co-operation then they underpin human civilisation. The cynical formula of ‘it’s not what you know, it’s who you know’ has a more benign face in the recent sociological emphasis on ‘social capital’ as a basis for successful societies. In the world of competition law and policy the recent surge of interest in the concept of ‘network’ has of course

* The author is Professor of Politics and Deputy Vice Chancellor at the University of Exeter. He is a member of the UK Competition Commission but writes in an academic capacity. None of the views expressed here should be attributed to the Competition Commission. 1 Heclo H. and Wildavsky A. (1974): The Private Government of Public Money, London, Macmillan. 2 Heclo H. (1978): ‘Issue networks and the executive establishment’, in King A., ed., The New American Political System, Washington DC, The American Enterprise Institute, pp. 87–124; Richardson J. and Jordan G. (1979): Governing Under Pressure, Oxford, Martin Robertson.

66

Working Paper II

been prompted by the Commission’s new found commitment to decentralisation of competition law implementation. This has been given revolutionary expression in the White Paper and draft proposals to ‘modernise’ competition policy by entrusting enforcement to the national competition authorities (NACs) and courts. Ehlermann asserts that the White Paper: is the most important policy paper the Commission has ever published in more than 40 years of EC competition policy. It suggests a legal and cultural revolution in proposing the fundamental reorganization of the existing responsibilities between the Commission, national anti-trust authorities and national courts.3

In giving up its monopoly of implementation the Commission is clearly opening the floodgates to the possibility of substantial divergence in the application and evolution of law across the Union. It has proposed various legal, administrative and judicial safeguards to avoid the ‘renationalisation’ of competition policy. It has also suggested organisational and sociological solutions including the ‘proper functioning of the network between the Commission and the members states’. As Ehlermann points out, the White Paper says less about the possibility of horizontal co-operation between NCAs but, as he also points out, this is a legitimate concern of the Commission which could look to the models of interstate cooperation in the United states. Hence the interest in networks is fed by the potential they offer for supplementing legal provisions and facilitating cooperation, coordination and consistency between the varied multiplicity of NCAs across Europe. It is doubtful, however, that our interest in policy networks should be restricted to their technical or functional advantages. As the political science literature emphasises, policy networks are about power. The concept of policy network purports to explain who participates in policy-making, how decisions about new policies are made, and how policy is adapted and interpreted as it is implemented. Members of policy networks are influential, it is important to be an ‘insider’ and actors co-operate within policy networks largely to ensure that they will achieve their personal or organisational goals. Competition policy networks are no different. Interested parties — the NCAs, lawyers, ministers, officials, courts, economists — need to understand networks in order to understand who has influence. They need to have a hand in constructing networks to ensure that they are insiders, that their voice is heard and their interests safeguarded. This conference and subject area is energised by the imperative to ‘construct’ helpful networks but the first insight from political science would be that this is not a disinterested process.

3 Ehlermann

C.-D. (2000): ‘The modernization of EC antitrust policy: a legal and cultural revolution’, 37 Common Market Law Review 537–90, at p. 537.

The Network Concept, Competition Authority Networks

II.

67

The Theoretical Shop Window

The purpose of this paper is to review the theoretical material on policy networks in the context of the concern with the ‘competition policy and law network’. The literature is extensive and the discussion will inevitably be selective. No attempt is made to present empirical findings on the presence, absence or form of the competition policy network in Europe. On the other hand, the discussion is influenced by previous work on competition policy.4 The following treatment is hence eclectic and deals with four main areas. First, it explores the concept of a ‘network’ and reviews some political science definitions. Secondly, it looks at the dynamics of networks and reviews how they are constituted, maintained and institutionalised. Thirdly, it looks at particular types of network, focused on the concepts of policy sectors and epistemic communities. Fourthly, it explores the political biases attached to network modes of policy making and presents some comments on network construction.

III.

What is a Network?

Our concern here is with the idea of the ‘policy network’ and hence with a mode of governance of some areas of public management. A policy network is a method of co-ordinating activity and may arise either spontaneously or through deliberate creation. Networks are conventionally contrasted to hierarchies and markets. A network is unlike a hierarchy in that it does not have central direction or an established set of authority relations; it is unlike a market in that relations are personalised and based on informal exchange rather than formal contractual exchange. One recent study declares that: we define policy networks as (more or less) stable patterns of social relations between interdependent actors, which take shape around policy problems and/or policy programmes.5

This serves to emphasise the idea that policy networks grow out of policy sectors or sub-sectors. Policy networks will exist for health, childcare, defence policy, education and so on. This definition talks of ‘actors’, earlier definitions tended to 4 In

particular, Doern B. and Wilks B., eds. (1996): Comparative Competition Policy: National Institutions in a Global Market, Oxford, Clarendon Press; Wilks S. (2001): In the Public Interest: Competition Policy and the Monopolies and Mergers Commission, Manchester University Press; Drahos M. (2002): Convergence of Competition Laws and Policies in the European Community, London, Kluwer; Wilks S. and Bartle I. (2002): ‘The unanticipated consequences of creating independent competition agencies’, 25 West European Politics 1, 148–72. 5 Kickert W. J. M. et al. (1999): ‘Introduction: a management perspective on policy networks’, in Kickert W. J. M., Klijn E.-H. and Koppenjan J. F. M., eds., Managing Complex Networks: Strategies for the Public Sector, London, Sage, at p. 6.

68

Working Paper II

identify the actors as organisations so that Benson’s influential work treated policy networks as a complex of organisations connected to each other by resource dependencies and distinguished from each other by breaks in the structure of resource dependencies.6 The key elements of these and like definitions include: • dependency — networks exist due to interdependencies between actors; • actor diversity — there are a range of actors with no ‘centre’ or dominant actor; • goal diversity — the actors have a variety of goals which are not necessarily shared by other actors; • continuity — networks have by definition a degree of stability which involves repeated interactions; • shared norms — networks embody agreement on norms expressed as shared understandings about substance and process (norms, which are tacit and specific, should be distinguished from values, which are explicit and generalised, e.g. ‘do not reveal confidential information’ is a norm; ‘competition is beneficial’ is a value). Networks display an infinite variety. In some policy sectors they may be long-lasting and dominant, in others they may be transitory and fragmented. In a European setting each country displays different patterning of networks. In the UK networks are ubiquitous, hidden and informal; in Germany they tend to be more formalised and government-sponsored (in a pattern sometimes labelled ‘neo-corporatist’); in Italy networks are more likely to take on ‘client-patron’ characteristics. Analysis of these patterns is part of the stock-in-trade of comparative political science. Here networks are seen as part of the ‘governance’ of the policy sector and will almost always reach out to include non-governmental actors. The cement of the policy network is the functional necessity for actors to co-operate to achieve their goals. Thus, in the area of health care officials in the Ministry of Health need to co-operate with colleagues in other ministries, with the professionals (doctors, nurses, pharmacists and their unions), with insurance companies, with pharmaceutical companies, with universities, charities and the decentralised agencies operating hospitals. In this process any one actor will be dependent on others for certain ‘resources’. The resources are most obviously money and manpower but they also include information, advice, legitimacy and authority. In some respects research into networks is relatively easy. Formal organisational arrangements provide a starting point and it can be seen who sits on key committees, which organisations are on advisory committees, who is consulted routinely and regularly, which people attend conferences and seminars, who receives resources. Indeed, there have been some ambitious attempts to ‘map’ 6 Benson

J. K. (1975): ‘The inter-organizational network as a political economy’, 20 Administrative Science Quarterly, 229–49.

The Network Concept, Competition Authority Networks

69

networks in formal ways. To use concepts such as authority, frequency of contact and number of contacts with other network actors to construct mathematical maps. On the whole, however, such descriptive research is one-dimensional and unsatisfactory. It is equally important to study the hidden aspects of the network — the compromises, deals and pressures. These are far harder to research and require sensitive, informed programmes of interviews, It is also important to reach conclusions about the ‘norms’ of the network. In coherent and stable networks these shared norms are likely to be well developed. They constitute a ‘world view’, a set of understandings and shared assumptions which constitute what has been called a ‘logic of appropriateness’.7 Actors within the network will share assumptions about how the world operates, their cognitive frameworks will coincide and certain courses of action will appear obvious or normal. To take a competition policy example: in assessing monopoly or merger cases the importance of market definition and market share would be accepted be all network participants as a key, if not the dominant, variable. Imagine an actor who joined the discussion and argued that market share was unimportant and that instead the decision should revolve around the employment practices of the companies concerned, specifically about the companies’ willingness to employ ethnic minorities. The network members would regard such arguments as nonsense, outside their cognitive framework, and would reject both the argument and the actor. This normative dimension is immensely important and very difficult to capture and to research. Thus we can visualise with some precision how a European competition policy network might be defined. I would involve a clear set of actors including leading officials in: CORE: • competition authorities, Brussels and NCAs • competition officials in Industry and Finance Ministries • prominent specialists, lawyers and economists • leading figures in the major law firms • courts and appeal agencies PERIPHERY: • • • •

leading judges and advocates economic consultancies lobbyists and lawyers in major companies and trade associations leading academics

But what do these potential network participants agree upon? They probably agree on the virtues of competition, but workable competition rather than pure competition. They may agree on legal certainty for business, on independence 7 Marsh J. G. and Olson J. P. (1998): ‘The institutional dynamics of international political orders’, 52 International Organization 4, 943–69, at pp. 951–52.

70

Working Paper II

from political intervention, and on exchange of case data. Research into the normative cement of a presumptive European competition network would be invaluable.

IV. The Dynamics of Networks — and is there a European Competition Policy Network? Our primary concern is with inter-organisational networks which are created through self-interest and sustained through a combination of self-interest and trust. The self-interest comes from the recognition that modern society is so complex that policy making and implementation requires the cooperation of a series of agencies and actors. Typically no one organisation can secure its objectives alone and is therefore ‘dependent’ on others. This idea of ‘resource dependency’ is at the root of network theory. Emphasis has been put on the resources of money and authority8 but resources would also include information, access, personnel, legitimacy or inaction. Some organisations within a network may be more central and therefore powerful, indeed, networks may be officially sanctioned through devices such as consultative committees or trade associations. Nonetheless, the network rests on the voluntary co-operation of the organisations concerned. Networks cross the public: private divide. Indeed, in the increasingly liberalised and privatised economies of Europe, private sector involvement must be standard. Similarly networks are frequently trans-national. In the globalised, or Europeanised social and economic space cooperation across borders in unavoidable and the EU itself has been likened to governance ‘by network’. The whole EU implementation apparatus of ‘comitology’ is an exercise in the science of network creation and operation. Networks are maintained by realisation of self-interest, by a reduction in transaction costs, by familiarity and by closure. Realisation of self-interest could include the adoption of preferred policies, the passage of desired legislation, the allocation of money or simply the achievement of an organisation’s goals. Organisations will have to compromise. There is no room in this world for fundamentalism, but compromises will be justified by the expectation of future incremental gains in resources. The cooperation resulting from an effective network will also reduce transaction costs. Information will be volunteered, programmes will be adapted, timetables will be synchronised. This is largely informal. Participants in a network do not resort to formal modes of direction or arbitration. They do not resort to ultimatums, to legal actions or to political interventions. As networks operate they build up familiar routines, standard operating procedures and a repertoire of successful approaches to problem solving. Familiarity, 8 See

supra note no. 8, at pp. 229–30.

The Network Concept, Competition Authority Networks

71

routine and success breed trust. This includes trust in individuals, in organisations and in joint approaches to problem solving. Thus the network becomes ‘normalised’ and builds up a collective cognitive mindset of acceptance which in turn generates a ‘logic of appropriateness’. Networks may become so established as to become ‘institutionalised’. Van Waarden suggests that ‘institutionalisation will tend to be greater in closed networks, with compulsory membership, ordered linkages, high intensity, multiplexity and symmetry of relationships, overlapping memberships and interlocking leaderships, and with a central policy unit’.9 Networks that become institutionalised may well have a regular cycle of intensive business involving expenditure programmes. The agricultural support policy network would be a standard example. Fourthly, we come to closure. Networks are defined by their boundaries and the rigour of inclusion or exclusion. A network may have a core, a periphery, but also interested parties who are excluded, for whom the network is closed. These are people and organisations who are not consulted, not copied in to reports, not involved in decisions. They may have perfectly legitimate interests and may be major bodies such as labour unions or regional governments, or critics such as environmental activists or other pressure groups. As Schaap and van Twist point out, closure also has a cognitive dimension. Actors in the network may be unwilling or unable to think about issues other than in the language and the framework of the network. ‘Network closure causes this form of closedness due to the fact that each discourse has its own symbols, its own rules for correct use of language. Communication with and in the network is only possible if one can speak the network language’.10 Thus, if actors use the wrong language or conduct themselves inappropriately the network closes itself off from them. This points up one major weakness of networks, their inability to think out of the frame, to be imaginative and to innovate. Network maintenance may clash fatally with network adaptation. Is there then already a competition policy network in existence in Europe? I am not going to offer an empirical analysis, simply some preliminary observations, some ‘sighting shots’. Taking the four elements of network maintenance reviewed in this section we can briefly map the degree to which competition policy would embody elements that would be conducive to network maintenance. The ‘Yes’ categories are areas where competition policy would lend itself to network maintenance, the ‘No categories the reverse: (i)

Self-interest Yes — need for authority, mainly from DG IV — need to collaborate over cases, both allocation and decision — exchange of information

9 F. van Waarden F. (1992): ‘Dimensions and types of policy networks’, 21 European Journal of Political Research, 29–52, at pp. 35–36. 10 Schaap L. and van Twist M. J. W. (1999): ‘The dynamics of closedness in networks’, in Kickert et al. (1999), as cited in supra note no. 5 at p. 72.

72

Working Paper II

No

— — — — —

formal co-operation through Advisory Committees inability to exchange confidential information no need for financial assistance separate legal systems no ‘routine’ contact, cases episodic

(ii)

Reduction in transaction costs Yes — adoption of common legal terminology and processes — processes of consultation in developing regulations and block exemptions — sharing and repatriation of cases — collective benefit if enterprises become competition compliant across the EU No — each case specific and unique — need to meet legal requirements of network partners may increase costs — companies and governments may insist upon opposition, not co-operation

(iii)

Familiarity Yes — organisations know each other, common meeting places and conferences — some transfer of staff between agencies — in some cases long history of co-operation No — no spending programmes or cycles of meetings — each legal and administrative system different — different national attitudes to mergers, restrictive practices etc.

(iv)

Closure Yes — — — No — — —

language and discourse legal and exclusive inclusion by professional status, e.g. law firms strong conventions of secrecy and official confidentiality economic and case-specific expertise need to be brought in legal process allows for appeal and review powerful corporate interests can insist upon political intervention

From this preliminary sighting it could be suggested that the European competition law network does have some tentative existence but is partial and fragile. Partly this is due to the difficulty of creating a trans-national network (national networks are easier to establish), but it also reflects the legal basis of the network which is grounded in national law and energised by individual and unique cases. Competition policy does not rest on the annual cycle of activity

The Network Concept, Competition Authority Networks

73

or the negotiation of massive spending programmes which have fostered networks in other policy areas.

V.

Types of Network: Epistemic Communities

There have been various attempts to construct a typology of policy networks, with rather limited success.11 Political scientists have tended to focus on relationships, especially exchange relationships, and on questions of who has power and influence within the network. The focus of the Commission and the proposed Council regulation is rather on the use of the network as a method of implementation and co-ordination, rather than policy-making. Both approaches tend to underplay one fascinating aspect of the embryonic competition policy network which is its role in producing and embedding ideas (although the Commission does talk of the importance of ‘competition culture’). The importance of ideas and control over the policy agenda is central to the concept of ‘epistemic communities’ but before turning to that possible avenue let us rehearse the actual and potential basis of the ‘competition network’. First, we should assume that the Commission either does not believe that such a network exists, or has not formed a view of its nature. The draft regulation observes in the explanatory memorandum that ‘it is a core element of the Commission’s proposal that the Commission and the national competition authorities should form a network’.12 Ehlermann points out the need to construct an horizontal network between NACs, as well as a vertical network between them and the Commission. Clearly there is already a level of trust and amicable relations at least at the vertical level. What are the ingredients for deepening and extending existing arrangements into an effective network? As we have seen, the ‘need’ to cooperate is, at the moment, rather slight. Each NAC and the Commission, treats each case separately working within specific legislative and procedural requirements. The work in compartmentalised by legal stipulations, by the need to retain legal independence, and by the constant reaffirmation of political independence. In this respect the competition policy and law provides infertile soil for network growth. Such network arrangements as exist tend to be formally constituted through advisory committees and arrangements to share information. The network involvement of the private sector, both companies and professional firms, comes through formal case meetings and hearings. Legal process and propriety, and the need to establish the authenticity and provenance of information which may be used in Court, means that informal contact has to be undertaken with caution.

11 Rhodes R. A. W. and Marsh D. (1992): ‘New directions in the study of policy networks’, 21 European Journal of Political Research 1/2, 181–205. 12 CEC Proposal for a Council Regulation, Brussels, COM(2000) 582 final, 27.9.2000, at p. 6.

74

Working Paper II

The committees, meetings, lunches, joint activities which permeate so many administrative activities can become problematic when legal challenge can be based on conflict of interests, bias and unequal access to the law. There is, of course, another face to this portrait of a formal and professionalised set of relationships. The competition policy world is populated by experts who may be specialised officials, lawyers, economists or trained business people. They share a familiarity with economics, with law and with legalised procedures. These tend to be people who share the same background. They tend to have in common University degrees, legal training, years of work within competition agencies. They will know the same people, go to the same conferences, read the same journals, admire the same ‘heroes’ and share the same professional standards. In other words a professionalised network already exists despite the formalism which due procedures and legal objectivity require. Moreover, whilst policy implementation, and the processing of cases, must be attended by inhibiting protocol, the same is not true of policy making. In discussion over new legislation — whether it is an EU regulation, reform of national laws, or (regularly) the transfer of the EU competition model into a new agency in an aspirant new member state, then contact can be free, informal and intense. The picture of a professionalised policy network raises crucial questions about the normative cement of relationships. It is likely that the network membership, both individuals and organisations, will share procedural norms. Do they also share substantive norms and values? Can we hypothesise that most if not all leading members of the policy-making network will share the following sorts of values? — — — — — — — —

competition is beneficial market integration should always be pursued competition policy should take precedence over trade policy competition promotes social justice competition promotes economic growth and welfare administration of competition policy should be free from political intervention economic conflicts are justiciable competition decisions should be driven by economic, not social, criteria.

In turn these values will drive certain sets of norms, for example, the need to employ economists, the need for economic approaches to market definition, the antagonism to market segmentation, the need to treat all complains equally, the need to keep at arms-length from politicians, the pre-occupation with ‘hard-core’ cartels. What is being sketched out here is a cognitive framework, a world view, which provides a concrete foundation for the relationships and the trust which will facilitate network maintenance. The sociology of this cognitive framework would suggest that it is largely people with pro-competition views who join competition agencies or who succeed to be promoted within them.

The Network Concept, Competition Authority Networks

75

The final ingredient in network creation is size. In fact the potential competition policy network across Europe is relatively small. It is concentrated in government ministries and agencies, in the courts, in legal practices, in economic consultancies and University departments, and in the legal sections of large companies. Compared with the profusion of actors concerned with policy areas like education or health care competition policy is a positively intimate sector. This opens up the possibility of an inter-organisational network also developing as a ‘social network’ of a number of prominent individuals who are in regular contact. Such groups, of judges, advocates and prominent officials in the Commission and the NCAs certainly seem to be characteristic of the European scene. Indeed, Professor Ehlermann’s annual conferences provide a Who’s Who of such people. A recent paper expresses the implications of a social network in the following terms: In-groups and dense ties create the conditions — trust, familiarity, mutual understanding and common language, shared preoccupations and values — that facilitate and speed up the circulation of an idea or practice within a particular group leading in time to cohesion and relative uniformity. Dense ties are therefore conducive to patterns of rapid and contagious adoption, to the ‘exploitation’ in March’s (1991) words, of a particular practice and idea that is already available within the group.13

If the points made about the ingredients of the competition policy-making network have validity, then they map rather accurately against the concept of ‘epistemic communities’ which has been developed in the field of international relations. The concept of the epistemic community goes beyond the idea of a professionalised network to try to capture the notion of a small, cohesive, international group of policy experts who have effectively moulded policy. The term derives of course, from ‘epistemology’ — the study of a common knowledge base. Policy experts become influential in periods of transformation when uncertainty about policy options obliges policy makers to turn to, or to create, networks of experts. Such networks have been identified as influential in, for instance, the creation of policy towards pollution control to protect the ozone layer and in the creation of free trade in services. Such networks have also been identified in the early creation of competition policy in the shape of the Franco-American group of specialists assembled around Jean Monnet.14 Networks of this sort are relatively small (as small as 35 people), transnational, and composed of individuals rather than organisations. One could hypothesise that such networks were influential in the creation of Regulation 17, in the re-invigoration of European competition policy in the 1980s (centred on Lean Brittan) and in the current re-definition of policy (centred on Karel van Miert).

13 Paper presented at the SASE meeting in Amsterdam (mimeo). 14 Djelic M. (1998): Exporting the American Model, Oxford University

Press, Oxford.

76

Working Paper II

A more formal definition of epistemic community is offered by Haas, with apologies for the length: An epistemic community is a network of professionals with recognised expertise and competence in a particular domain and an authoritative claim to policy-relevant knowledge within that domain or issue-area. Although an epistenic community may consist of professionals from a variety of disciplines and backgrounds, they have (1) a shared set of normative and principled beliefs, which provide a value-based rationale for the social action of community members; (2) shared causal beliefs, which are derived from their analysis of practices leading or contributing to a central set of problems in their domain (3) shared notions of validity — that is, intersubjective, internally defined criteria for weighing and validating knowledge in the domain of their expertise; and (4) a common policy enterprise — that is, a set of common practices associated with a set of problems to which their professional competence is directed, presumably out of the conviction that human welfare will be enhanced as a consequence.15

This definition offers a powerful combination of attributes — shared norms, shared causal beliefs, shared notions of validity, a common policy enterprise. This captures the idea of a rather evangelical group of senior officials, lawyers and perhaps economists who shared a common aversion to dirigisme, to industrial intervention, subsidies, industrial champions, indeed, to ‘planning’. The alternative concept of the market economy based on sound economic principles and implemented through a robust legal framework offered a promising way forward for liberal politicians uncertain about a ‘post-planning’ vision for the management of the economy. To sum up the argument: whilst there are clearly formal arrangements for co-operation in competition policy across Europe, there is no ‘policy implementation’ network of any degree of coherence. There is, however, a ‘policy-making’ network that is small, elite, and with the characteristics of an ‘epistemic community’. The epistemic community is not driven by the conventional network pressures of functional goal achievement, familiarity and resource dependency, instead it is driven by its ‘common causal beliefs’ and its ‘common policy enterprise’. The epistemic community is not therefore a good source of routine coordination but it is a source of policy innovation. It might thus be argued that it is this community which is responsible for the radical and momentous proposals which are the main concern of this workshop. The challenge is therefore threefold. First, to extend an epistemic community into a ‘policy network’ capable of co-ordinating policy implementation. Secondly, to deepen the network so that more junior individuals and organisations, including national courts, become familiar with the technical issues and normative assumptions (what the Commission calls a ‘competition culture’). Thirdly, to ensure that 15 Haas P. (1992): ‘Introduction: epistemic communities and international policy coordination’, 46 International Organization 1, at p. 3. (Special issue on Knowledge, Power and International Policy Coordination, edited by Haas P.).

The Network Concept, Competition Authority Networks

77

the network is sufficiently flexible and open to accommodate new ideas and critical perspectives. Competition policy has come of age and will have to escape some of the elite certainties of its progenitors.

VI.

Constructing Implementation Networks

This workshop is centred on the construction of an effective competition network. The following comments are simply a small contribution to that larger discussion. They bring together some perspectives from political science and follow on from the sections above. There are three areas of comment: design principles, the dialectic of network closure, and the danger of professional dominance.

1.

Design Principles

The proposed Council regulation represents a moment of enactment of new institutional arrangements. There is a well developed literature on institutional design and widespread agreement that the moment of enactment is critical to the development of policy with ramifications lasting for decades.16 There is a debate about whether institutions can be designed from scratch (suggested by rational choice theorists), or whether they are largely pre-determined by the existing endowment of arrangements (path dependency theorists). The rational choice theorists assert that institutions can be designed to favour certain coalitions of interest. They assert that they can be ‘hard-wired’ to privilege certain interests through manipulation of the level of independence, resources, appointments, access, administrative structure and so on.17 The creation of a network is not quite the same as the creation of an institution but some of the same insights apply. This repeats the point that a new network will embed existing power relationships, it is not ‘just’ about efficiency. On the question of efficiency the discussion above indicates that network participants must be given incentives to cooperate. They should be made ‘dependent’ on other participants in the network. Such dependency will require an acceptance of informal negotiation, agreement and compromise. These processes will be lubricated by the build up of trust and mutual respect but this must be at the level of the organisation rather than the individual. Implementation networks should be built on a stable basis and not be vulnerable to staffing changes. 16 See Wilks S. and Bartle I. (2002), as cited supra note no. 4. 17 Macey J. (1992): ‘Organizational design and political control

of administrative agencies’, 8 Journal of Law, Economics and Organization 1, 93–125; see also Horn J. (1995): The Political Economy of Public Administration: Institutional Choice in the Public Sector, Cambridge University Press.

78

2.

Working Paper II

The Dialectic of Network Closure

The network will not be coherent and effective unless it develops a set of normative understandings. The analysis above referred to shared understandings and a common language. These shared understandings clearly exist. A competition policy practitioner can walk into a competition agency anywhere in the world and within five minutes be engaged in familiar discussions about market dominance, industrial structure, types of restrictive practices and the irritating reluctance of industrial and trade policy-makers to understand that free competition ultimately strengthens the national economy. The European network will be based on this ‘world view’ and on the more technical concerns with legal principles, legal processes and analysis of the case-law emerging from the ECJ and national courts. Network creation should clarify and reinforce these understandings as it creates a ‘competition culture’ as a foundation for the competition network. But, and here the dialectic, the more coherent and embedded the competition culture becomes, the more it generates its own downfall by alienating or exasperating outsiders. The competition network must gain wide acceptance, in government, in industry and amongst the wider public. Martin Bangeman’s famous attack on the ‘competition Ayatollahs’ provides a pertinent warning that there are alternative and perfectly respectable world views which lead to different conclusions about economic efficiency, welfare and the public interest. It will be important for the network to develop a culture which can relate the benefits of competition to a wider set of public policies and public preoccupations. The competition culture will need to be true to its principles but also have a strong pragmatic element along the lines established by Karel van Miert.

3.

The Dangers of Professional Dominance

An easy option and a possible outcome is that the network will be based upon, and dominated by, lawyers. In the UK economists have become increasingly prominent and influential in competition policy but in the rest of Europe the norm is for lawyers to take the leading positions. This may a peculiarly British preoccupation but the ‘legalisation’ of competition policy carries dangers. There has been a long-standing concern about the ‘justiciability’ of complex economic cases and it is far from certain that the application of legal principles is always conducive to economic efficiency or social justice. In more practical terms the network will have to embrace administrative and industrial interests who will fine legal language largely opaque. If the key to a successful competition policy is compliance, then the policy, through the network, must be made transparent and comprehensible. Network participants may be entirely persuaded of the virtues of competition, the network itself must be created in such a way that it can convey those virtues to policy makers and to a sometimes sceptical public.

The Network Concept, Competition Authority Networks

VII.

79

Conclusion

This paper has aimed to contribute towards the theme of the Workshop by outlining the approaches to the analysis of policy networks which have been developed within political science. The analytical approach and the conceptual frameworks may be helpful in evaluating the potential for the creation of the competition network within Europe. Work within political science has tended to concentrate on the dynamics of networks and the way in which they structure power and influence. The focus has been on understanding policy rather than designing networks. More recently the policy network approach has been overtaken by a wider concern with the institutional configuration of given policy areas so that research has stressed the biases embedded in institutions and has sought to understand them by analysing their historical evolution. In other words the preoccupation with policy networks that dominated policy studies in the early 1990s has waned and this approach has tended to recede in importance. Nonetheless, the policy network approach does allow for more conceptual clarity and does allow us to define the requirements of an effective policy implementation network. The paper has argued that a ‘policymaking’ network, which could also be termed an ‘epistemic community’, already exists; it is far less certain that a coherent policy implementation network exists, either vertically between the Commission and the NCAs, or horizontally between the NCAs. This paper has therefore been relatively ‘apolitical’. It has not drawn conclusions about the potential effects of the reconfiguration of competition policy responsibilities anticipated by the proposed Council regulation. Others have argued that the reforms would weaken EC anti-cartel policy, or that the scope for an autonomous national competition policy will be substantially reduced.18 This paper would certainly share the assumption that the way in which the legislation is formulated and the network is created will have profound impacts on the power of competition policy within Europe and the ability of interested parties to deploy and to benefit from that power . But at this point the paper simply makes the point that would be expected from a political scientist, that the process of network creation is in itself an acutely political process in which arguments about efficiency, effectiveness, value for money, equity and legal certainty are almost certainly masking deeper seated concerns about protecting and extending the interests of the protagonists.

18 See

Ehlermann C.-D. (2000), as cited supra note no. 3, at p. 560; Bergeron J. H. (2001): ‘Antitrust federalism in the European Union after the modernization initiative’, Antitrust Bulletin 513–35, at p. 533.

III James F Rill* Why Sectoral Regulators Should Not Compete in US Merger Analysis

I.

Introduction

The allocation of responsibilities among governmental bodies is often imperfect, placing responsibilities in the hands of agencies that are not adept at executing that authority. One area of particular concern, and potentially severe economic impact, is the grant of authority to sectoral regulators to assess antitrust issues that should be left to competition experts. In the United States, two federal agencies — the US Department of Justice Antitrust Division (DOJ) and the Federal Trade Commission (FTC) — share generalised responsibility for merger review and competition policy formulation across the economic spectrum. In addition to these two agencies, certain sectoral government agencies have been allocated specialised authority for competition assessment and enforcement in designated regulated or semi-regulated industries. The role of the sectoral government agency varies from concurrent jurisdiction with the DOJ and the FTC to preemptive jurisdiction over mergers in a specific industry. The decisions made by these specialised agencies are sometimes at odds with the recommendations and decisions of the antitrust enforcement agencies. The problem of overlapping sectoral and generalised antitrust review has global significance. The frictions and complexities presented by the possibility of merger inquiries in 60 to 70 jurisdictions presents in itself an extraordinarily daunting challenge to expeditious and consistent enforcement. Concurrent or subsequent sectoral antitrust determinations elevate further the spectrum of delay, cost and inconsistent result to impede the progress of what may well be pro-competitive, efficient transactions or, conversely, in the case of sectoral preemption authorise transactions which would merit antitrust challenge. This problem was directly addressed by the US Department of Justice when it formed the International Competition Policy Advisory Committee (ICPAC) in November 1997 to study and present recommendations to the Justice Department on the future of international antitrust policy.1 A majority of the ICPAC members * James

Rill is a partner and co-chair of the antitrust section of Howrey Simon Arnold & White, LLP in Washington, DC. The author gratefully acknowledges the assistance of Jennifer Vasta in the preparation of this paper. 1 The Committee was co-chaired by James Rill and Paula Stern. Committee members included the following leading representatives from the worlds of business, law, and academia: Zoe Baird, Thomas E. Donilon, John T. Dunlop, Eleanor M. Fox, Raymod V. Gilmartin, Vernon E. Jordan, Jr., Steven Rattner, Richard P. Simmons, G. Richard Thoman, and David B. Yoffie.

82

Working Paper II

recommended removal of the oversight authority for the competition aspects of merger review from the sectoral agencies.2 In offering this recommendation, the ICPAC majority explained that overlapping sectoral and generalised agency authority threatens: (1) efficient review; (2) substantive international convergence; (3) case-by-case cooperation; and (4) consistency and transparency.3 This paper examines the effect of overlapping responsibility for competition review and enforcement in four regulated industries — railroads, airlines, telecommunications, and agriculture. It discusses the roles and mandates of the various agencies and identifies the consequences of multiple agency oversight. It also discusses the effect of decentralised US enforcement on global cooperation efforts. A review of this jurisprudence shows that the role of these sectoral regulators in merger review, if any, should be limited to industry specific, non-competition based analysis, and that sectoral regulators should cede all competition-based review to the FTC and the DOJ.

II. The Jurisdiction of the Department of Transportation over Certain Airline Mergers 1.

The Source of the DOT Jurisdiction and the Standard of Review

Agreements between US airlines and foreign carriers are subject to the sole jurisdiction of the US Department of Transportation (DOT). Prior to 1989, the DOT had authority over domestic and international airline mergers. While the DOJ now has jurisdiction over domestic airline mergers, the DOT retains authority over the approval of proposed international airline alliances and the grant of antitrust immunity for such agreements.4 The DOJ continues to conduct its own merger analyses regarding international alliances and files comments in DOT proceedings concerning such matters in an advisory capacity. Federal law requires the DOT to approve those agreements that are not adverse to the public interest provided that they do not substantially reduce or eliminate competition. An exception is provided for agreements deemed necessary to meet a serious transportation need or to achieve important public benefits — including international comity and foreign policy considerations — that cannot be met by reasonably available alternatives that are materially less anticompetitive. Approval 2 See

Final Report, Department of Justice’s International Competition Policy Advisory Committee to the Attorney General and Assistant Attorney General for Antitrust (hereinafter ‘ICPAC Report’), at p. 154. The Advisory Committee was assisted in large part by a paper prepared for the Committee by William E. Kovacic, ‘The Impact of Domestic Institutional Complexity on the Development of International Competition Policy Standards’, March 15, 1999 (hereinafter ‘Kovacic Submission’). Note that entrusting competition policy exclusively to the federal agencies requires Congressional action. 3 ICPAC Report, at pp. 145–47. 4 The DOT has authority over international alliances and the decision to grant immunity for such alliances under 49 U.S.C. §§ 41309 and 41308, respectively.

The Network Concept, Competition Authority Networks

83

of an agreement by the DOT specifically exempts the parties affected by the order from the antitrust laws to the extent necessary to proceed with the transaction.5

2.

The Failed Alliance between American Airlines and British Airways

The most recent example of the divergent results of a DOJ and DOT merger review in this area is the failed effort by American Airlines (AA) and British Airways (BA) to secure DOT approval for its proposed international alliance. The DOT originally conducted a conventional merger analysis of the alliance and concluded that it would not substantially harm competition in any market if seven of the alliance’s slots at London Heathrow airport were transferred to Continental and Delta for service from New York and Boston. The DOJ conducted its own merger analysis and arrived at essentially the same conclusion.6 The transfer of slots and facilities for those seven daily services was an aggressive remedy that would have eliminated any competitive concern according to the DOJ. However, the DOT went beyond its original analysis and announced that tentative approval of the alliance was subject to the divestiture of the slots and facilities equal to American’s sixteen daily round-trip services to Heathrow.7 Rather than permitting economic or commercial conditions to determine the outcome, the agency also determined exactly how the slots would be allocated between the other major US airlines. The DOT stated its intention to distribute American’s slots, free of charge, to US airlines that actively opposed the alliance.8 Notably, the underlying statute does not require a finding that an agreement would have public benefits. Rather, it simply requires a determination that the agreement is not ‘adverse to the public interest’. By extending its reach beyond its statutory mandate, the DOT engaged in a form of government planning that is completely at odds with market principles. It went beyond its requirement to address potential anticompetitive effects of the proposed alliance and determined that it would use the proposed alliance between AA and BA to attempt a complete restructuring of the market for airline service between the US and the UK. Clearly this dirigiste formula would benefit the recipient airlines — Continental, Delta, Northwest, and US Airways — but not consumers. Confiscation 5 49 U.S.C. §41309. 6 Specifically, DOJ recommended divestiture of enough slots for new entrants to offer nine daily round

trips from New York and Boston as well as substantial new air service from other US cities. See Department of Justice Press Release, ‘Justice Department Urges DOT to Impose Conditions on American Airlines/British Airways Alliance’, December 17, 2001. 7 Department of Transportation Press Release, ‘Department of Transportation Tentatively Approves Antitrust Immunity Covering the Alliances of American Airlines and British Airways, and United Airlines and BMI’, January 25, 2001. The DOT announcement stated that approval conditions included the divestiture by American and British Airways of 17 daily round-trip slots for flights by new competitors between US cities and Heathrow. 8 Id. The DOT’s order allocated the slots as follows: Continental (5), Delta (6), Northwest (3), and US Airways (2).

84

Working Paper III

of assets and summary reallocation preempts the market from determining whether the airlines are able to acquire slots through normal competitive efforts. Final approval of the alliance also was contingent on the finalisation of an ‘Open Skies’ agreement between the US and the UK. The irony of this condition is that the asset reallocation plan that DOT set as a condition to approval of the proposed AA/BA alliance jeopardised the ability to put an Open Skies agreement in place. The DOT and the DOJ both support the implementation of an open skies arrangement to replace the Bermuda II bilateral agreement. Bermuda II is a highly restrictive agreement that severely limits new entry between the US and London’s two major airports. It allows only four carriers (AA, BA, United and Virgin Atlantic Airways) to provide any service between the US and Heathrow. On the US side, the carriers may serve only ten US cities. The UK government recognised that Open Skies must include access for its national carriers to the entire US Therefore, it made clear that obtaining an Open Skies agreement was contingent upon the approval by DOT of the AA/BA alliance. By rendering the alliance unworkable through its severe conditions for approval, the DOT squandered an important opportunity to eliminate the Bermuda II restrictions and enhance competition from the US to London. The US and the UK. governments are no longer in active negotiations on an Open Skies agreement.

III. 1.

Mergers in the Telecommunications Industry Concurrent Jurisdiction and the Role of the Federal Communications Commission

Since passage of the Telecommunications Act of 1996,9 the DOJ and the Federal Communications Commission (FCC) have had concurrent jurisdiction over mergers and acquisitions in the telecommunications industry. The FCC’s jurisdiction to review such mergers is derived from both the Clayton Act10 and the Communications Act of 1934.11 The Clayton Act grants the FCC the authority to bar acquisitions of ‘common carriers’ engaged in wire or radio communications where the effect of such acquisition may be substantially to lessen competition, or tend to create a monopoly. This is the same mandate given to the DOJ under the Clayton Act to review mergers generally.12 The Communications Act requires

9 Telecommunications Act of 1996, Pub.L. No. 104–104, 110 Stat. 56 (1996) (codified at 47 USC §§151 et seq.). 10 15 U.S.C. §§21(a), 18 (1994). 11 Communications Act of 1934, Pub.L. No. 73–416, 48 Stat. 1064 (codified as amended in scattered sections of 47 U.S.C. 12 The Clayton Act grants the DOJ statutory authority to review mergers, and prohibits the acquisition of stock or assets by any ‘person’ where ‘the effect of such acquisition may be substantially to lessen competition, or tend to create a monopoly’.

The Network Concept, Competition Authority Networks

85

the FCC to review mergers between telecommunications common carriers under a ‘public interest’ standard. While both agencies undertake a competition analysis, the principles guiding each agency’s review differ significantly. The DOJ review focuses on whether possible anticompetitive effects of the merger violate antitrust laws — the mandate of the Clayton Act. While the FCC has authority to challenge proposed transactions under the Clayton Act, it rarely, if ever, does so. Instead, the FCC uses its broader powers under the Communications Act to apply the public interest standard, which includes an assessment of the competitive effects of the merger. Thus, from the merging parties’ perspective, it gives the government a ‘second bite at the apple’ to reject the proposed merger on competition grounds. The FCC’s scope of review under this standard is not clear, creating confusion and uncertainty for the merging parties. Former FCC Commissioner Furchgott-Roth addressed this issue in Congressional testimony, stating: I also have grave concerns about the process and practices employed in FCC merger review. The current system — or rather, the lack of a clearly delineated one — puts merging entities in an inequitable and difficult situation … . Unfortunately, there is no established Commission standard for distinguishing between the license transfers that trigger extensive analysis by the full Commission and those that do not. Nor do any of the Commissions orders in ‘merger’ reviews elucidate the standard.13

While the scope of review under the FCC’s public interest standard is not well defined, it clearly extends well beyond the parameters of review under the antitrust laws and encompasses the policy objectives of the Telecommunications Act. Such objectives include providing services to rural areas (i.e. universal service) and sustaining high levels of residential service quality. In short, while a DOJ merger analysis ensures that a proposed merger is not harmful to competition, the FCC is required to conclude that a merger will enhance competition and benefit consumers.14

2.

The Bell Atlantic–NYNEX Merger Review

With such differing review standards, it is not surprising that the DOJ and the FCC have come to different conclusions for the same merger. In the case of the merger between Bell Atlantic and NYNEX, the DOJ did not oppose the merger while the FCC imposed several conditions.15 The FCC’s conditions included (1) performance monitoring reports, negotiated performance standards, and enforcement mechanisms; (2) carrier-to-carrier testing of uniform operations 13 Testimony

of Federal Communications Commissioner Harold W. Furchtgott-Roth before the House Committee on Commerce, Subcommittee on Telecommunications, Trade, and Consumers, March 14, 2000, available at www.fcc.gov/Speeches/Furchtgott_Roth/2000/sphfr004.html. 14 See Statement of Commissioner Susan Ness, Federal Communications Commission, Before the Committee on the Judiciary, US House of Representatives, June 24, 1998. 15 ‘Antitrust Division Statement Regarding Bell Atlantic/NYNEX Merger’, US Department of Justice Press Release, April 24, 1997; ‘FCC Approves Bell Atlantic/NYNEX Merger Subject to MarketOpening Conditions’, Federal Communications Commission Press Release, August 14, 1997.

86

Working Paper III

support systems; (3) prices based on forward-looking economic costs; (4) shared transport facilities on a minutes-of-use basis; and (5) easy payment plans for non-recurring charges. These conditions are clearly directed toward meeting the policy objectives of the Telecommunications Act rather than addressing potential antitrust harm. This reflects the FCC’s belief that ‘the proper role of competitive forces in an industry must… be based not exclusively on the letter of the antitrust laws’, but on the ‘special circumstances of the industry’.16 However, the FCC decision in the Bell Atlantic/NYNEX case implies that the Commission reached a different conclusion than the DOJ on the competitive effects of the transaction.17 The DOJ is not required to consider whether a proposed merger’s effect on an industry is consistent with a particular policy objective or congressional mandate. In fact, the DOJ has no authority to impose conditions on a merger that it finds either ‘competitively beneficial or neutral’.18 The FCC’s public inquiry focuses precisely on such factors. Not all FCC Commissioners are comfortable with their role in merger review. Commissioners Furchgott-Ross and Powell have been vocal in their concerns regarding concurrent DOJ and FCC authority for telecommunications merger review. Commissioner Harold Furchgott-Roth’s testimony before the House Committee on Commerce, Subcommittee on Telecommunications, Trade, and Consumer highlighted the problems associated with shared jurisdiction in this area. Noting that the FCC has authority under the Clayton Act but rarely invokes such power, Commissioner Furchgott-Roth stated: If the Commission intends to exercise authority over mergers and acquisitions as such, it ought to do so pursuant to the Clayton Act, not the licensing provisions of the Communications Act … . The Commissions focus on mergers rather than on license and authorization transfers creates another problem: our work often duplicates that of the Department of Justice’s Antitrust Division and the Federal Trade Commission. Merging companies should not have to jump through excessive federal antitrust hoops, and those hoops should be held out by the institutions with the express statutory authority and expertise to do so. Those agencies are the Department of Justice and the FTC. When the FCC gets into the game as well, it increases the costs of the merging parties and expends taxpayer funds, while adding little from an antitrust perspective.19

Chairmen, then-Commissioner Powell has also expressed concern about the differing standards and duplicative functions of the FCC and the federal antitrust authorities: The antitrust authorities must prove their case by a preponderance of the evidence to block a merger, whereas the FCC places the burdens on the applicants to affirmatively 16 United States v Federal Communications Commission, 652 F.2d 72, 88 (DC Cir. 1980). 17 See Memorandum Opinion and Order in the Application of NYNEX Corp. and Bell Atlantic

Corp. for Consent to Transfer Control of NYNEX Corp. and Its Subsidiaries, No. NSD-L-96-10, at §37, August 14, 1997. 18 See US Dept. of Justice and Federal Trade Comm’n, Horizontal Merger Guidelines (1992), reprinted in 4 Trade Reg. Rep. (CCH) 13, §194. The 1992 Guidelines stipulate that while the DOJ will challenge ‘competitively harmful mergers, the Agency seeks to avoid unnecessary interference with the larger universe of mergers that are either competitively beneficial or neutral’. 19 See supra note no. 13.

The Network Concept, Competition Authority Networks

87

prove the transaction is ‘pro-competitive’, a fairly recent pronouncement [referring to the Bell Atlantic-NYNEX case] … The Commission should be constrained to consider only such issues as whether the merger would violate an express provision of the Communications Act or the Commission’s rules. In addition, it is appropriate for it to consider the merger’s impact on other communications policies such as media diversity and universal service that are not appropriately considered by antitrust authorities. However, I believe that the Commission should be required to defer to the antitrust authorities’ competitive analysis and leave it up to them (and the courts) to address the specific competitive harms that they identify … . The Commission could file comments with the appropriate antitrust authority reviewing a merger or issue an advisory opinion on a given merger.20

If Congress will not clarify the appropriate role for the FCC with respect to competitive analysis of mergers, then the FCC should exercise restraint in exercising its authority. The FCC merger review increases transactional costs of the parties and introduces a significant degree of uncertainty into the process. The standard for review is unclear and the time-sensitive nature of mergers is not compatible with the FCC’s unlimited review period.21 The competitive analysis should be ceded to the FTC or the DOJ.22 Worth noting is a mirror image issue arising from legislative assignment of dual agency review. Take, for example, the responsibility of the DOJ for advising the FCC in connection with the authorisation of incumbent local exchange carriers’ entry into the inter-exchange voice telecommunications markets.23 Under the Telecommunications Act, the FCC is given power to lift the statutory restriction against such entry and is to receive advice from DOJ as to matters including whether the applicant complies with a 14 point checklist set forth in the Act. DOJ’s assessment of compliance with the checklist includes a determination as to, among other issues, (1) non-discriminatory access to telephone numbering, signaling systems, and 911 services, (2) unbundling of local loops, transports, and switching, and (3) white page directory listings. Forcing DOJ to stray from traditional competitive concerns into such ‘public interest’ fields produces an 20 Testimony

of Federal Communications Commissioner Michael K. Powell before the House Committee on Commerce, Subcommittee on Telecommunications, Trade, and Consumer Protection, on The Telecommunications Merger Act of 2000, March 14, 2000. See www.fcc.gov/Speeches/ Powell/Statements/2000/stmkp005.doc. 21 The FCC recently adopted rules in an attempt to streamline review of Section 214 applications. The new rules provide that certain applications, such as small incumbent local exchange carrier transactions or transactions in which the acquiring party is not a telecommunications provider, will automatically be granted 30 days after public notice unless otherwise notified by the Commission. See FCC Press Release, ‘Federal Communications Commission Updates Merger Review Process’, March 14, 2002. 22 Under the recent agreement between the DOJ and the FTC regarding clearance procedures, media transactions were to be allocated to the DOJ. ‘Memorandum of Agreement Between the Federal Trade Commission and the Antitrust Division of the United States Department of Justice Concerning Clearance Procedures for Investigations’, March 5, 2002. However, due to Congressional opposition to the agreement, the DOJ ceased adhering to the agreement. See US Department of just a Press Release, Statement by Charles Jawes regarding DOJ/FTC Clearance Agreement, May 20, 2002. 23 Telecommunications Act of 1996 at §271.

88

Working Paper III

unwonted drift from consumer welfare objectives into the realm of industrial policy. The result does not seem salutary.

IV. The Jurisdiction of the Surface Transportation Board in the Railroad Industry 1.

The Source of the STB Jurisdiction and the Standard of Review

Jurisdiction over mergers and acquisitions in the railroad industry resides solely with the Surface Transportation Board (STB). The DOJ provides non-binding advice that the STB must consider but need not follow.24 In this regard, DOJ is treated like any other party that may submit comments under the Interstate Commerce Act. The DOJ, or any other party, can only challenge the decision on the grounds that (1) the original proceeding involved material error; (2) there is new evidence; or (3) circumstances have substantially changed.25 The DOJ cannot challenge a decision on antitrust grounds. By statute, approval by the STB of a railroad merger exempts the carriers involved from the antitrust laws for all actions necessary to effectuate the transaction.26 The STB was established within the US Department of Transportation on January 1, 1996 by the ICC Termination Act of 1995 (ICCTA) as a successor to the Interstate Commerce Commission. The STB has broad jurisdiction over all matters concerning railroads, including the review of mergers.27 By statutory mandate, the STB applies a broad ‘public interest’ standard in its evaluation of mergers and acquisitions.28 Factors that must be considered by the STB in the merger of two Class I railroads include, at a minimum, adequacy of transportation to the public; other rail carriers in the area involved; impact on employees; financial impact on the involved carriers; and competitive effects.29 In addition, statutory rail transportation policy directs the STB to promote safety, efficiency, good working conditions, an economically sound and competitive rail transportation system, and the needs of the public and national defense.30 24 Specifically, 49 USC §11324 requires the STB to ‘accord substantial weight to any recommendations of the Attorney General’. 25 49 C.F.R. §1115.4. 26 49 U.S.C. §11321(a). 27 See 49 U.S.C. §10501. 28 49 U.S.C. §11324. 29 Id. Class I railroads are defined as railroads with annual revenues of at least $250 million. Interestingly, a merger that does not involve the merger of at least two Class I railroads is subject to a different standard. The standard for non-Class I railroad mergers focuses on the competitive effects of the transaction. Under §11324(d), the STB shall approve such mergers unless it finds that the transaction will (1) result in a substantial lessening of competition, creation of a monopoly, or restraint of trade in freight surface transportation in any region of the United States; and (2) the anticompetitive effects of the transaction outweigh the public interest in meeting significant transportation needs. 30 49 U.S.C. §10101.

The Network Concept, Competition Authority Networks

2.

89

STB’s Statutory Mandates Corrupt Competition Analysis

In 1996, the Surface Transportation Board analyzed the proposed merger of the Union Pacific (UP) and Southern Pacific (SP) railroads. The DOJ strongly opposed the UP/SP merger, concluding in testimony before the STB that it was the ‘most anticompetitive rail merger ever proposed’.31 The STB approved the merger despite DOJ’s vocal concerns and instituted two conditions. First, UPSP was required to grant certain trackage rights to the Burlington Northern and Santa Fe Railway Company (BNSF). Second, the STB subjected UPSP to oversight for five years during which time the STB would determine whether the conditions effectively addressed the competitive issues they were designed to remedy.32 At the time of the approval, Assistant Attorney General Anne Bingaman issued a strongly worded press release criticising the STB decision, stating that: The Surface Transportation Board, a regulatory agency created by Congress to succeed the Interstate Commerce Commission, decided today to approve the $5.4 billion Union Pacific and Southern Pacific railroad merger. This merger will create a monopoly in many markets and leave shippers throughout the West with just two major rail competitors. As a condition of approval, the Surface Transportation Board has given itself a significant regulatory oversight role for years to come in an effort to prevent this merger from being anticompetitive. We continue to believe that a competitive market structure — and not more regulation — is the best way to keep prices low in the railroad industry, as in every industry. We are concerned about the effect of today’s decision on shippers and consumers and will review carefully the Surface Transportation Board’s written decision when it becomes available.33

The UP/SP merger review and decision reflect the fundamentally different mandates of the DOJ and the STB. The STB’s broad regulatory authority, granted by statute, allows the STB to review potential acquisitions with limited attention paid to their competitive effects. The STB took the position that a merger reducing the number of competitors from three to two was not problematic and that the effect of a reduction in competitive service from two to one could be mitigated simply by the remaining firm’s grant of ‘trackage rights’ to another railroad. As a result, a merger that the DOJ deemed highly anticompetitive was approved and thereby exempted from the antitrust laws. Once the merger was approved, any anticompetitive harm that ensued could be addressed only by the STB as part of its five-year oversight condition and not by the DOJ.

31 Testimony of Anne K. Bingaman, Assistant Attorney General, Antitrust Division, US Department of Justice before the Surface Transportation Board, July 1, 1996. 32 The STB ended its formal oversight of the UP/SP merger in December 2001. See STB Press Release, ‘Surface Transportation Board Ends Formal Oversight of ‘Union Pacific-Southern Pacific’ Railroad Merger; Resolves Various Disputes; Will Continue Enforcement of Agency-Imposed Merger Conditions’, December 20, 2001. 33 Anne K. Bingaman, Assistant Attorney General, Antitrust Division, Department of Justice Press Release, ‘Statement on the Surface Transportation Board’s approval of the Union Pacific and Southern Pacific merger’.

90

Working Paper III

After the UPSP approval, a group of rail customers and railroad competitors sought to reopen the UPSP case through the oversight process, citing stifled competition and massive service failures. The problems were particularly severe in the Houston, Texas area. The STB issued a series of unprecedented emergency orders based on their finding that a ‘failure of traffic movement exists which creates an emergency situation of such magnitude as to have substantial adverse effects on shippers, or on rail service in a region of the United States’.34 However, the STB refused to reopen the UPSP merger and official oversight of the transaction ended in December 2001. At the time some major rail customers concluded that they would have to look to Congress to force the board to boost rail-to-rail competition.35 Prior to the UP/SP merger, the DOJ participated in hearings before the STB’s predecessor, the ICC, regarding the proposed merger of Burlington Northern (BN) and Santa Fe (SF). The DOJ raised numerous competitive issues through both written filings and oral arguments. Over DOJ’s objection, the ICC approved the BN/SF merger.36 The DOJ supported efforts to reopen the ICC’s decision in the BN/SF case, stating that the conditions imposed on the merger did not effectively remedy the resulting loss of competition.37 The STB rejected that request and efforts to appeal the decision were unsuccessful.38 Six utilities solely served by Santa Fe also raised concerns regarding the BN/SF transaction, arguing that the vertical effects of the merger would harm their interests. However, the ICC concluded in the BN/SF decision that ‘the market power faced by a utility was neither created nor increased by consolidation of a monopoly destination carrier with an origin carrier’.39 The Commission further concluded that while it was not impossible that the benefits of origin competition might flow through to a utility, it was presumed that they would not.40 The only way to overcome the presumption of legality of such vertical arrangements was through an affirmative showing that a specific utility was able to obtain real benefits from origin competition. While the utilities introduced extensive evidence on the benefits of origin competition, the ICC rejected their data and denied all requests for conditions based on the argument that the merger would have anticompetitive vertical effects.41 34 See

‘Surface Transportation Board News’ release No. 97–92, issued October 31, 1997; ‘Surface Transportation Board News’, release No. 97–102, issued on December 5, 1997. Daniel Machalaba and Anna Wilde Mathews, ‘US Regulators Reject an Effort to Revisit Union Pacific Merger’, The Wall Street Journal, December 22, 1998. 36 10 I.C.C.2d 661 (August 16,1995). 37 See Response of the United States Department of Justice to Petition of Roquette America, Inc. to Reopen Finance Docket No. 32549, March 6, 2000, available at www.usdoj.gov/atr/public/comments/4282.htm. 38 FCC decisions are reviewed by the US Courts of Appeal. The Appeals Courts apply a deferential standard of review to the Commission’s decisions. 39 Id. at 181. 40 This is sometimes referred to as the ‘one lump’ theory, meaning that a firm with market power in a vertical chain can extract only ‘one lump’ of profit even though it operates at two or more levels in the chain. Id. at 182. 41 In addition to the utilities and rail labour, 21 entities requested that the ICC impose various conditions on the BN/SF merger. The ICC denied all requests for conditions from these entities except for two limited conditions respecting one entity. 35 See

The Network Concept, Competition Authority Networks

91

In response to pressure from major industries and members of Congress, the STB held a series of hearings on March 7–10, 2000 regarding major rail mergers and the present and future structure of the railroad industry. The hearings were prompted by the application of BNSF to acquire Canadian National and the fear that another round of railroad consolidation was imminent. The DOJ chose not to participate in these hearings largely, one could observe, as a matter of protest over the STB’s disregard of prior DOJ commentary. Pending the issuance of new merger guidelines, the STB ordered a 15-month moratorium on all railroad mergers on March 17, 2000.42 On April 5, 2001, the STB held a hearing concerning its proposed new major railroad merger regulations. While a number of railroads and rail labour organisation presented their views at the hearings, the DOJ did not participate. Given the STB’s record of disregarding the DOJ’s comments and findings, it is not surprising that the Department decided not to take part in the proceedings. As a result, while the STB did exhibit some concern for traditional competition analysis, the final regulation does reflect a not insubstantial tilt toward competition regulation based on factors other than consumer welfare. The ICC and the STB have significantly structured railway competition through their ‘public interest’ reviews. These reviews have largely disregarded competitive effects and relied, with little success, on the STB’s ability to enforce behavioral conditions. Moreover, the STB’s moratorium was instituted ad hoc without consideration of potential case-sensitive competitive effects, showing further disregard for fundamental principles of competition analysis. Even if a moratorium ultimately may have been the right result, it was undoubtedly reached by the wrong means. The railroad industry has consolidated rapidly under the watch of the ICC and the STB. Since 1980, the number of major railroads has been reduced from 40 to only four with more acute effects on a localised level. The analysis by the STB that has sanctioned this result is both a departure from established antitrust principles and somewhat incoherent. Legislation vesting authority for the review of antitrust issues as to railroad mergers and joint ventures in the antitrust agencies should be accorded serious consideration.

V.

Merger Review in the Agricultural Industry

In 2000, several bills were introduced in the US Congress to authorise the US Department of Agriculture (USDA) to challenge agribusiness mergers. According to their sponsors, these bills were a response to a wave of mergers in the agriculture sector. The proposed legislation would have empowered the USDA to 42 STB

Ex Parte No. 582, ‘Public Views on Major Rail Considerations’, March 17, 2000. See also STB Press Release, ‘Surface Transportation Board Directs Railroads to Take a Breather from Mergers While New Merger Rules are Developed’, March 17, 2000; Don Phillips, ‘Board Puts Hold on Rail Merger Plans’, The Washington Post, March 18, 2000 at E1.

92

Working Paper III

challenge mergers based on a broad policy mandate. S.2252 would have authorised the USDA to challenge an agribusiness merger if the merger would cause ‘substantial harm to the ability of independent producers and family farmers to compete in the marketplace’.43 S.2411 would have empowered the Secretary of Agriculture to declare unlawful an agricultural merger that ‘could be significantly detrimental to the present or future viability of family farms or ranches or rural communities in the areas affected by the merger or acquisition, pursuant to standards established by the Secretary’.44 At Senate hearings on S.2252, USDA Secretary Daniel Glickman testified that: Current antitrust review does not include socioeconomic issues and is limited to more traditional legal and economic criteria. I do not know the best way to address these nonspecific legal and economic criteria in addressing the other impacts of agricultural mergers or who is best equipped to perform this analysis, whether it is the DOJ or the Department of Agriculture or somebody else. But it is my personal judgment that these issues should be examined … .45

But Secretary Glickman acknowledged that the DOJ and USDA work closely on merger reviews involving the agricultural sector and that the DOJ has taken steps to protect American farmers from the potential adverse effects of acquisitions.46 In Congressional testimony in April and September 2000, the DOJ emphasised their strong record of enforcement in the agribusiness sector.47 In the past three years, the DOJ challenged four mergers in the agricultural markets: Monsanto/DeKalb Genetics Corporation; Cargill/Continental; New Holland/Case; and Monsanto/Delta & Pine Land. The DOJ conducted 43 Under S 2252, factors the USDA would have been required to consider included (1) the effect of the transaction on prices paid to producers; (2) the likelihood that the transaction would significantly increase market power for the surviving entity; (3) the likelihood that the transaction would increase the potential for anticompetitive conduct by the new surviving entity; and (4) whether the transaction would adversely affect producers in a particular region, including an area as small as a state. 44 Under S 2411, the USDA could also challenge a merger if it violated a long list of vague unfair practices, including (1) using unfair, unreasonable, unjustly discriminatory, or deceptive practice or device in the marketing, receiving, purchasing, sale, or contracting for the production of any agricultural commodity; (2) giving unreasonable preference to a particular person or locality; (3) making false or misleading statements; (4) failing to perform any specification or duty, express or implied, arising out of any undertaking in connection with any transaction involving any agricultural commodity purchased or received in interstate commerce; (5) retaliating or disadvantaging someone because of statements he/she make to any person regarding alleged improper actions or violations of law; and (6) price discriminating for agricultural commodities of like grade and quality. 45 See Transcript, Hearing before the Subcommittee on Antitrust, Business Rights, and Competition of the Committee on the Judiciary, United States Senate, on S 2252 to Examine Issues Related to Competition and Mergers in the Agricultural Industry, and Related Proposals, September 28, 2000, at 5. 46 Id. at 8. 47 See Statement of John M. Nannes, Deputy Assistant Attorney General, Antitrust Division, US Department of Justice, Hearing on Antitrust Enforcement in Agriculture, Senate Committee on Agriculture, Nutrition, and Forestry, April 27, 2000; Statement of John M. Nannes, Deputy Assistant Attorney General, Antitrust Division, US Department of Justice, Before the Subcommittee on Antitrust, Business Rights, and Competition, Senate Committee on the Judiciary, Concerning Agriculture Concentration, September 28, 2000.

The Network Concept, Competition Authority Networks

93

exhaustive investigations and effectively identified and addressed the anticompetitive effects of the proposed combinations. Additionally, in 1999, the DOJ created the position of Special Counsel for Agriculture to address the competitive concerns in agricultural markets and advise the Assistant Attorney General of specific industry implications affecting the antitrust review of mergers and other transactions. The working relationship of the USDA with the DOJ and FTC is an example of how a sectoral agency can work with the DOJ to enhance the merger review process. The DOJ calls upon the expertise of USDA during merger investigations and taps into the USDA’s extensive information about agricultural markets.48 In 1999, the DOJ, FTC, and the USDA entered into a memorandum of understanding with respect to monitoring competitive conditions in the agricultural sector.49 The memorandum formalised the agencies’ cooperative working relationship, including the exchange of legal, economic, and technical assistance that would help the agencies’ carry out their enforcement responsibilities. The relationship between the USDA and the antitrust agencies ensures that comprehensive industry specific information is provided as part of the merger review process but that the underlying antitrust analysis is carried out by the agency with the proper expertise, namely the DOJ. Concurrent, to say nothing of preemptive, sectoral authority for merger review would undercut the goal of ensuring consistent and coherent competition policy and procedures. The DOJ and the FTC aggressively enforce the antitrust laws in the agriculture sector. Extending jurisdiction to the USDA would duplicate efforts and create inefficiency. Transaction costs for agribusiness community would increase, as parties would be forced to respond to investigations from multiple agencies. Additionally, the legislation that was proposed would have imposed legal standards for agribusiness merger review that are different from, and vague compared to, the well-defined antitrust standards applied by the antitrust agencies. Fortunately, the legislation failed for now and federal merger review in the agricultural sector remains with the DOJ.

VI. Merger Review by the Federal Energy Regulatory Commission Transactions involving energy companies are subject to potential review by five separate entities: (1) one of the federal antitrust agencies, (2) the Federal Energy Regulatory Commission (FERC); (3) the Securities and Exchange Commission; 48 See

Douglas Ross, Special Counsel for Agriculture, Antitrust Division, Department of Justice, Remarks before the Organization for Competitive Markets, July 20, 2001, at 9. See www.usdoj/atr/ public/speeches/8661.htm. 49 Memorandum of Understanding Between The Antitrust Division, Department of Justice and The Federal Trade Commission and The Department of Agriculture Relative to Cooperation with Respect to Monitoring Competitive Conditions in the Agricultural Marketplace, August 31, 1999. See www.usdoj.gov/atr/public/guidelines/3675.htm.

94

Working Paper III

(4) the public service commission in each state in which the parties do business; and (5) the attorney general of each state in which the parties do business. Approval by one of these entities does not preclude a challenge by another. FERC’s authority to review public utility mergers derives from the Federal Power Act.50 The Act requires that the Commission approve those mergers or acquisitions that are consistent with the public interest. FERC has issued its own guidelines for merger review and in December 1996, the agency issued a Merger Policy Statement with the design of streamlining its merger review process.51 Under the policy, mergers must be consistent with FERC’s competitive goals and the public interest and must account for changes in market structures. The Order simplified FERC’s review by focusion on three factors: (1) effect on competition; (2) effect on rates; and (3) effect on regulation. The Commission also adopted the Horizontal Merger Guidelines promulgated by the DOJ and the FTC. FERC’s merger review, however, is much different in practice than that of the antitrust agencies. FERC’s broad public policy review has invoked criticism from merging parties and a FERC commissioner. In a strongly worded speech, FERC Commissioner Hébert asserted that ‘Congress shuold remove us from the merger business.’52 Commissioner Hébert’s speech was in response to FERC’s decision in the American Electric Power Company/Central and Southwest Corporation case, where the Commission reversed the findings of fact of the Administrative Law Judge and disregarded expect testimony in imposing additional conditions on a merger that did not pose anticompetitive harm. In response to criticism regarding FERC’s role in merger review, the House of Representatives recently passed H.R. 4. The bill repeals FERC’s authority under Section 203 of the Federal Power Act to review proposed mergers and dispositions of public utilities. Several FERC Commissioners defended their involvement in merger review and specifically argued that since FERC’s ‘duty to protect the pubilc interest is broader than the focus of the antitrust agencies’, the Commission is in a better position to participate in the restructuring of the public utilities industry.53 However, as noted by Commissioner Hébert in prior Congressional testimony, FERC merger review ‘brakes utility efforts to restructure themselves as they deem best to respond to and take advantage of competitive opportunities and challenges. More troubling, FERC uses mergers to further policy goals that it has no authority to order directly.’54 50 Federal Power Act of 51 Inquiry Concerning

1935, 16 U.S.C. §824b. the Commission’s Merger Policy Under the Federal Power Act: Policy Statement, FERC Order No. 592, December 18, 1996. In November 2000, FERC issued revised merger filing requirement through Order No. 642. The Order generally reaffirmed the December 1996 policy statement. 52 Remarks by Commissioner Curt Hébert on FERC’s Role in Merger Review, Washington, D.C., March 15, 2000. 53 Testimony of Commissioner William L. Massey, Federal Energy Regulatory Commission, Before the Subcommittee on Energy and Air Quality, United States House of Representatives, December 12, 2001. 54 Testimony of Commissioner Curt L. Hébert, Jr., Federal Energy Regulatory Commission, Before the Committee on Commerce, Subcommittee on Energy and Power, United States House of Representatives, September 11, 2000.

The Network Concept, Competition Authority Networks

95

While the Senate rejected an amendment to broaden FERC merger review, it did not repeal FERC’s merger review authority in S. 517, its version of H.R. 4. In fact, the Senate bill added a requirement that FERC approve generator transfers in addition to mergers. As of this writing, the bills are in Conference, where the provisions may be reconciled. Congress should take advantage of this opportunity to eliminate duplicative efforts and introduce consistency to public utility merger review by review by removing FERC from the process.

VII.

Conclusion

As the examples above illustrate, placing antitrust jurisdiction outside the federal antitrust agencies results in inconsistent investigations of potential mergers. In some cases, the specialised regulatory agency approves mergers that the DOJ or FTC finds clearly anticompetitive, as in the case of the STB’s approval of the UPSP and BNSF mergers. In other cases, the agency places additional conditions in the name of remedying anticompetitive effects when the real motivation is policy or political concerns, as evidenced by the American Airlines/British Airways and Bell Atlantic/NYNEX transactions. The result is a system plagued by uncertainty and, ultimately, irrational economic consequences. The DOJ and the FTC are more skilled in antitrust matters and should be solely responsible for the review of the competition effects of mergers and acquisition. These agencies are accustomed to learning about, assessing, and making enforcement decisions about a wide variety of complex industrial organizations and industries. In those instances where additional industry expertise is required, the agencies can and do cooperate with sectoral regulators but base their decisions, ultimately, on economic results, as proper antitrust analysis dictates. Sectoral agencies may still carry out the express policy mandates prescribed by Congress for their particular industries. As FCC Commissioner Powell stated, the regulatory agency should be constrained to consider only whether a transaction would violate an express provision of the statute they are required to enforce. Placing authority over competition solely with the DOJ and FTC will ensure that the sectoral agencies cannot effectuate other social and economic policies under the guise of a competition analysis. Taking the competition analysis from the agencies’ purview would make their policy and political motivations more transparent. Transparent decisionmaking is an important constraint on the exercise of discretion by public officials.55 In addition, empowering specialised government agencies to challenge mergers and acquisitions in certain regulated sectors is an inefficient use of resources. 55 See

Kovacic W. E. (2000): ‘Lessons of competition policy reform in transition economies for US antitrust policy’, 74 St John’s Law Review 361, at pp. 383–384.

96

Working Paper III

Even when the DOJ does not have specific authority over a merger in a regulated industry, as in the area of railroads and international airline alliances, it still conducts a merger analysis in order to provide advisory comments to the deciding agency. Often DOJ’s analysis and recommendations are summarily ignored. Some regulatory agencies recognise the problems associated with concurrent jurisdiction and believe that they should not have authority for reviewing the competitive effects of mergers and acquisitions.56 Consistency in antitrust enforcement is an important component of cooperation efforts. It is necessary to allow national governments to establish common policies and procedures with foreign counterparts and to work together on assessing individual transactions. There is no mechanism by which foreign competition authorities can share information with sectoral regulators as they can with the FTC and DOJ. 57 It is difficult for the US to encourage foreign governments to cure imperfections in their competition policies and procedures when the US system is more decentralised with different substantive standards implemented by different government bodies. The ICPAC Report raised this issue noting: The United States by virtue of its experience and developed practices can and should play a leading role in the effort to implement reforms in the international arena. Perhaps one of the most effective ways in which the United States can stimulate global reform is through leading by example. It is therefore important that the United States continue to examine and perfect its own merger review process. After addressing problems within its own borders, the United States is well positioned to advocate that other jurisdictions make modifications in their merger review systems.58

The FTC and the DOJ continue to work together to improve the merger review process. Many of the recent initiatives were adopted from the ICPAC recommendations, including the guidelines issued last October regarding the merger investigation process and the recently released agreement between the two agencies concerning clearance procedures for merger investigations. 59 However, improved coordination between the DOJ and FTC is not enough. The increasingly global nature of transactions requires consolidation of authority for competition merger review and policy formulation solely with the antitrust agencies.

56 See supra notes no. 13 and 20. 57 Kovacic Submission at p. 21. 58 ICPAC Report at pp. 87–88. 59 Memorandum of Agreement Between

the Federal Trade Commission and the Antitrust Division of the United States Department of Justice Concerning Clearance Procedures for Investigations, March 5, 2002.

IV Barry Hawk and Jeffrey Beyer* Lessons to be Drawn from the Infra-National Network of Competition Authorities in the US: The National Association of Attorneys General (NAAG) as a Case Study

I.

Background on the NAAG

1.

What is the NAAG?

The National Association of Attorneys General (NAAG) is a trade association of attorneys general in the United States of America. It was established in 1907, with the primary mission ‘to promote communication among the attorneys general, the chief law enforcement officials in their state, and to assist in the delivery of high quality legal services to the states’. The NAAG membership consists of the Attorneys General and chief legal officers of the 50 states, the US territories, the District of Columbia and the Commonwealth of Puerto Rico. It must be emphasised that NAAG is an association of law enforcement bodies, and not an association of national competition authorities, as are the NCAs in the EC. In other words, the state advocates general prosecute actions before courts; they do not decide cases, unlike most NCAs in Europe. This difference cautions against an overly enthusiastic drawing of lessons from NAAG by the ECA, or their application to the proposed network of competition authorities that is to be established in Europe. Notwithstanding the importance of collective action fostered by the NAAG, each attorney general retains his/her sovereignty regarding enforcement decisions; the NAAG itself has no enforcement authority. Enforcement activities are actually ad hoc actions taken by particular attorneys general, either individually or collectively. The NAAG has several permanent committees, including the Antitrust Committee, Civil Rights Committee, Consumer Protection Committee and Criminal Law Committee.

* Skadden, Arps, Slate, Meagher & Flom LLP, Brussels, Belgium.

98

2. 2.1.

Working Paper IV

Specific Focus on Antitrust Enforcement State Involvement in Antitrust Enforcement, Generally

The office of the Attorney General of each of the 50 states has the power to enforce both federal and state antitrust laws. All states are free to participate in the NAAG. In that sense, NAAG is open and lacks ‘closure’, to use the political science jargon. The end of the nineteenth century saw the adoption of aggressive state antitrust enforcement statutes in the US. At the time when the Sherman Act was enacted, 13 states had antitrust laws on the books and another 14 state constitutions included antitrust provisions. The state antitrust measures took many forms and often established extremely detailed prohibitions. Several state laws variously targeted monopolies, restraint of trade, restraint of competition, pooling, price fixing, output limitations, territorial divisions, resale restraints, refusals to deal and predatory pricing, among others. By 1915, statutes in as many as 35 states authorised fines, imprisonment, or both, and often provided penalties that were much more severe than the one year, $5,000 limits established in the Sherman Act as passed in 1890. Statutes in North Carolina, South Carolina, Oklahoma, Tennessee and Texas authorised maximum prison terms of up to ten years for antitrust violations. Nor were the state enforcement activities confined to purely local concerns; combinations of multistate scope were often targeted. For example, ten states and the Oklahoma Territory brought 24 cases against members of the Standard Oil Trust between 1890 and 1906. Other state suits challenged the activities of the sugar, beef and tobacco cooperatives. The aggressive enforcement efforts undertaken by states in the late nineteenth and early twentieth centuries can be shown through a couple of examples. In the years between 1912–1918 and 1922–1924, Texas courts imposed $1.5 million in total fines in state antitrust cases. By the end of 1915, Missouri courts had levied unsuspended fines of $678,000 against defendants in five actions charging violations of state antitrust standards. With the passage of the Sherman Act (1890), Clayton Act (1914) and Federal Trade Commission Act (1914), federal regulators began active antitrust enforcement at the federal level, and state antitrust enforcement declined significantly. One study found no state merger actions during the 1930s, 1940s and 1950s; two during the 1960s; and two during the 1970s. However, federal enforcement under the Reagan Administration was greatly reduced in the 1980s, and it was limited to prohibiting cartels and large horizontal mergers. In response, state antitrust enforcement was revitalised, and its focus expanded to cover multistate and national matters as well as local ones. Twenty-nine cases and 16 amicus briefs or comments were filed during the 1980s. During 1988, the number of merger cases filed by the states outnumbered those filed by the DOJ Antitrust Division. The late 1980s and 1990s again saw increased enforcement at the federal level.

The Network Concept, Competition Authority Networks

99

But rather than recede, states have generally remained active, with many states increasing their antitrust enforcement in recent years.

2.2.

The NAAG and Antitrust

The NAAG Antitrust Committee is composed of nine attorneys general who study substantive antitrust matters and recommend policy positions to the attorneys general. These positions are often the basis for testimony by attorneys general at congressional hearings or in letters to Congress. During the ‘renaissance’ of state antitrust enforcement in the 1980s, the NAAG created the Multistate Antitrust Task Force. Formed in 1983, the Task Force was intended to coordinate the exercise of powers of the individual state attorneys general in antitrust matters. It is a permanent subcommittee of the NAAG’s Antitrust Committee. The Task Force comprises the chief antitrust attorneys for each of the 50 state attorneys general. The Multistate Antitrust Task Force and the Antitrust Committee attempt not only to enforce state antitrust laws, but also promote state enforcement of federal antitrust laws. The concept underlying the Multistate Task Force surfaced in the mid-1970s when the NAAG assigned an assistant attorney general to report semi-annually to the attorneys general on the progress of multistate cases in which the states were parties or had some other interest. The Task Force was originally staffed by six assistant state attorneys general; but more and more states wished to participate as it became more active, and the Task Force eventually became a de facto ‘committee of the whole’. All 50 states are now members of the Task Force, and formal Task Force meetings are held several times a year; representatives of more than 30 states regularly attend the meetings. The activities of the Task Force include conducting multistate investigations and filing multistate actions, writing amicus curiae briefs, drafting and lobbying for proposed legislation, and preparing antitrust policy guidelines. A Chair, selected by the Chair of the NAAG Antitrust Committee, leads the task force. There are also seven national vice-chairs; two for investigations, three for amicus matters, one for policy and legislation and one for education and training. The regional vice-chairs coordinate regional matters that are not appropriate for Task Force activity, and they also facilitate communication with the chiefs of the regional offices of the FTC and Antitrust Division.

II. The Relationship Between and Among Members of the NAAG The members of the NAAG cooperate with one another with respect to legal and law enforcement issues, policy research, and analysis. They jointly

100

Working Paper IV

conduct investigations and litigation. They have promulgated enforcement guidelines as well as a voluntary pre-merger disclosure compact, and they also work together to draft amicus briefs in important antitrust cases, and propose new legislation.

1.

Coordinated Investigations and Litigation

The NAAG Multistate Antitrust Task Force coordinates multistate investigations and litigation. Virtually all multistate investigations and litigation are conducted through the Task Force. In these cases, the Task Force functions like an ad hoc antitrust enforcement unit. When the Task Force decides to pursue an investigation, a small group of states typically takes the lead. One state will issue subpoenas or civil investigative demands authorised by some state statutes, and it will inform the respondents that their responses will be shared with other interested states. This approach serves several purposes: it relieves respondents of having to respond to multiple subpoenas; it facilitates coordination among the states; and it allows the states to use their resources more effectively. The lead states report to the Task Force, and eventually recommend whether a suit should be filed. States often arrange to share costs, and they regularly deputise staff attorneys from one state to act as assistant attorneys general in other states for investigation and litigation purposes. The states coordinate their enforcement efforts in several distinct areas, as discussed below.

1.1.

Price Fixing

Recent coordinated enforcement actions and activities in this area included: CDs (coordinated action by 33 states challenging alleged price fixing in the sale of audio compact discs); sanitary paper litigation (allegation that several paper manufacturers conspired to fix prices in the commercial sanitary paper market); oil (Arizona, California, Oregon and Washington challenged alleged horizontal price fixing by major oil companies, resulting in 17 years of litigation culminating in a settlement worth over $134.5 million in cash and $11 million in fuel vouchers); and shoes (50 states and the District of Columbia reached a $7.2 million settlement with Keds Corporation in a case involving allegations that Keds conspired to fix the resale price of women’s athletic shoes).

1.2.

Tying Arrangements

States have also jointly undertaken enforcement action in the area of tying arrangements. For example, in the mid-1980s, Maryland, Delaware, West

The Network Concept, Competition Authority Networks

101

Virginia, Pennsylvania, Virginia and the District of Columbia jointly challenged an agreement between a Toyota distributor and its dealers to offer Toyotas only with a $500 rustproof package, which resulted in refunds for 33,000 consumers. In September 1993, 34 attorneys general settled a suit charging Sandoz Pharmaceuticals Corp. with unlawfully tying the sale of its schizophrenia drug, Clozapine, to a blood monitoring and laboratory testing system operated by Caremark, Inc. The attorneys general recovered $20 million for patients and state mental health programs that had been forced to buy the lab tests.

1.3.

Mergers

Although mergers are commonly thought of as generally being challenged by the Justice Department and/or the FTC, the states have also brought merger cases. For example, in 1988, Massachusetts, Maine and New Hampshire challenged the acquisition of Federated Department Stores by the Campeau Corp., a department store chain. The states’ challenge resulted in a consent decree, whereby Campeau agreed to divest its Filene’s Basement division if the take-over were to succeed.

1.4.

Joint Ventures

Joint ventures have been the subject of coordinated state enforcement efforts as well. For example, in 1989, Visa and Mastercard formed a joint venture to develop and market a point of sale debit card called ‘Entree’. Twelve states alleged the joint venture stifled competition because Visa’s and Mastercard’s members included almost every major bank in the US, and the joint venture in the emerging debit card market therefore constituted an attempt to create a monopoly in a new payment system industry. A consent decree was ultimately entered, requiring Mastercard and Visa to terminate the Entree card.

1.5.

Monopolisation

In one prominent recent monopolisation case, the generic drug manufacturer Mylan Laboratories reached a settlement with 33 states in early 2001. It paid $100 million in a case in which Mylan was charged with unlawfully attempting to corner the market in two prescription drugs commonly used to treat Alzheimer’s disease. And, in the mostly widely known monopolisation case of recent years, 19 states and the District of Columbia joined the Justice Department in suing Microsoft for alleged monopolisation of the computer software market. Although nine states and the Justice Department settled with Microsoft, nine other states and the District of Columbia elected not to settle. The Microsoft case will be explored more fully below, in the context of the discussion on the relationship between the states and federal antitrust authorities.

102

1.6.

Working Paper IV

Other Areas of State Enforcement

The states have also actively enforced antitrust law in other areas, including health care and boycotts. Furthermore, the states were intimately involved in the tobacco litigation that resulted in 46 states entering into a $206 billion settlement with the tobacco industry in 1998 to resolve a variety of claims, including antitrust and consumer claims. The above are just a few examples of the successful joint litigation and investigations undertaken by the states in the spirit of cooperation. Coordinated state enforcement has clearly proven to be extremely effective. Between 1995 and 1997, attorneys general reached settlements in multistate cases with America Online, American Cyanamid, Bausch & Lomb, General Motors, Louisiana Pacific, Mazda, Packard Bell and Sears Roebuck.

2.

Enforcement Guidelines

In addition to conducting joint investigations and litigation, the states have also used the NAAG and the Multistate Antitrust Task Force as a vehicle to develop and adopt policy positions. Work by the Task Force led to the adoption of the NAAG Vertical Restraints Guidelines and the NAAG Horizontal Merger Guidelines. Both of these guidelines reflect the antitrust philosophy of the states and ‘embody the general enforcement policy of the NAAG and its members’.

2.1.

The NAAG Vertical Restraint Guidelines

The Guidelines were unanimously adopted by NAAG in December 1985 and, as such, represent the first collective venture by the states into the field of policy formulation. The Guidelines were amended in December 1988 and again in 1995. The Vertical Guidelines purport to state the general enforcement policy of the 50 states under both federal and state antitrust law, with respect to resale price maintenance and non-price vertical restraints. However, the Guidelines acknowledge that differences in state antitrust laws and in prosecutorial decisions made by attorneys general may result in some deviations from the stated enforcement policy. The Vertical Guidelines have two stated primary purposes: • to ‘mark the boundaries’ between per se unlawful agreements and those to be judged under the rule of reason; and • to identify the factors that the attorneys general deem relevant in a rule of reason analysis.

The Network Concept, Competition Authority Networks

103

Generally speaking, the NAAG Vertical Guidelines reflect a clear intention to challenge vertical price-fixing agreements, a scepticism regarding pro-competitive inter-brand effects as a justification for intra-brand restraints, and a 10 per cent market screen to weed out less important non-price cases. In addition, the Guidelines unequivocally condemn resale price maintenance. With respect to tying arrangements, the key inquiry is whether the firm imposing the tying arrangement has ‘economic power’ in the market for the tying product. The Guidelines have prompted coordinated challenges by NAAG members to various national distribution schemes in industries including shoes and crop protection chemicals.

2.2.

The NAAG Horizontal Merger Guidelines

The Horizontal Merger Guidelines were approved by the NAAG in 1987 and revised in 1993. For the most part, these Guidelines are in harmony with their federal counterparts, the DOJ-FTC Horizontal Merger Guidelines of 1992. However, there are some differences between the two. For example, the NAAG approach to market definition in section 3 relies on actual market experience and data. Merging parties may present an analysis under the federal market definition methodology, which is recognised as an alternative method of defining markets to investigating states. Mergers involving a leading firm in the market or an innovative firm also receive special scrutiny under the NAAG Merger Guidelines. Additionally, the failing division defence permitted under the federal guidelines is not recognised by the NAAG Horizontal Merger Guidelines. The NAAG Guidelines were developed largely to promote uniformity among the states. The Guidelines provide a practical alternative to the federal merger guidelines. The NAAG drafters concluded that the federal guidelines represented an overly theoretical method of analysis and lacked predictability. However, it should also be noted that NAAG recognised that uniform results might not always be achieved within the Horizontal Guidelines. Though the Merger Guidelines were intended to create predictability in merger enforcement, the Guidelines provide that state officials ‘may vary or supplement this general antitrust enforcement methodology in the exercise of their individual prosecutorial discretion or to account for differences in state antitrust laws and variations in precedents among the federal circuits’.

3.

Voluntary Pre-Merger Disclosure Compact

NAAG published a Voluntary Pre-Merger Disclosure Compact in 1987, which was signed by 45 states. The Compact was subsequently revised in 1994. The Compact allows parties to voluntarily file with a designated ‘liaison state’ a copy

104

Working Paper IV

of their initial Hart-Scott-Rodino filing, copies of any subsequent requests for information by the federal enforcement agencies, and — upon the request of any state — the additional materials provided in response to these requests. The initial filing must be simultaneous with those to the DOJ and the FTC. In return for this voluntary filing, signatories of the NAAG Compact agree not to serve the parties with any additional requests for information or to make any other compulsory pre-complaint investigative demands on the parties during the Hart-Scott-Rodino waiting period, including any extensions of time. However, the NAAG Compact signatories do not agree to be bound or constrained by the decisions of the federal enforcement agencies. Under the amended Compact, the signatory states agree not to use compulsory process to investigate a merger that is undergoing HSR merger analysis by one of the federal agencies unless the merging parties have declined to provide supplementary materials that have been voluntarily requested by a state in a timely fashion. Additionally, the Compact addresses confidentiality concerns by pledging to maintain the confidentiality of information contained in the pre-merger notification forms.

4.

Amicus Briefs

NAAG members also coordinate the states’ participation as amicus curiae in major antitrust cases, through the amicus curiae program. States have filed amicus curiae briefs before the US Supreme Court, in order to inform the Court of the states’ positions in many antitrust cases, including Eastman Kodak Co v Image Technical Services, 504 US 451 (1992) (a tying and monopolisation case), and Matsushita Electric Industrial Co v Zenith Radio Corp 471 US 1002 (1985) (a case concerning an alleged conspiracy to fix prices and a predatory pricing scheme). The states have also submitted amicus briefs to federal courts in the merger context. For example, after an independent investigation, several states submitted an amicus brief and an expert affidavit in support of the FTC’s challenge of the proposed Staples/Office Depot merger, but they were ultimately denied the right to intervene in the proceedings.

5.

Legislative Positions

The Multistate Task Force also proposes legislation that attempts to harmonise and improve antitrust enforcement at both the national and state levels. For example, the Task Force drafted both the State Attorneys General Antitrust Improvements Act of 1987 and the Model State Statute Governing Pre-Merger Notification. The State Attorneys General Antitrust Improvements Act consisted of a package of federal legislative reforms, including (among other items) ending the antitrust

The Network Concept, Competition Authority Networks

105

immunity enjoyed by the insurance industry under the McCarran-Ferguson Act. The bill did not pass. The NAAG Model State Statute provides for mandatory pre-merger notification to state antitrust authorities under certain circumstances. A bill based on the NAAG Model Statute was introduced in the Connecticut Legislature but was later defeated.

III. The Relationship Between Members of the NAAG and the Federal Antitrust Authorities (FTC and DOJ) 1.

How the States and the Federal Agencies Cooperate

In addition to serving as a vehicle to encourage the states to work together in antitrust enforcement, the NAAG has also played an important role in improving cooperation and coordination between federal and state antitrust enforcers. The enforcement policies of the states generally and the federal antitrust agencies have recently been going through a process of close coordination. The relationship between the states and federal antitrust regulators has been described as evolving ‘from outright hostility in the early 1980s to grudging recognition in the late 1980s and joint prosecutions in the early 1990s’. In any event, it is clear that the relationship between the states on the one hand and the FTC and DOJ on the other, has undergone a huge transformation. This is largely due to the cooperative work of the NAAG Multistate Task Force and the federal agencies over the past several years. The federal-state coordination is aimed at promoting efficient use of the antitrust law resources and improving the consistency of the federal and various state antitrust laws. The cooperation between the NAAG members and the federal agencies has taken many forms, including: joint federal-state investigations and litigation; information sharing between state and federal agencies; formation of new institutional structures to promote federal-state coordination (such as the Executive Working Group for Antitrust); and direct communication between the states and the federal government on antitrust matters.

1.1. a.

Examples of Joint Federal–State Investigations and Litigation Major Litigation

Just as the states have coordinated multistate litigation among themselves, there have been many recent examples of major cases that have been brought jointly by both federal and state authorities.

106

Working Paper IV

In 1994, the FTC and all 50 states jointly settled one case against Reebok that involved price fixing in the athletic footwear market in violation of federal and state law, and a second case concerning a resale price maintenance scheme involving Keds in the women’s casual footwear market. In the course of the same year (1994), the Antitrust Division of the Justice Department and the state of Utah coordinated their investigations regarding price fixing by hospitals of their compensation levels for nurses. Eventually, the District Court of Utah entered three consent decrees under which each of the defendants — eight Utah hospitals, the Utah Hospital Association and the Society for Healthcare Human Resources Administration — were prohibited from, among other things, ‘conducting or facilitating’ any exchange of information concerning ‘the current or prospective compensation paid to nurses’. In the matter of Northwest Crab Fishermen, a joint action was undertaken by the DOJ, California, Oregon and Washington to prohibit price coordination by crab fishermen. b.

Microsoft

The most significant litigation of recent years in which the federal antitrust regulators and the states cooperated was the lawsuit brought jointly against Microsoft by the Justice Department and several state attorneys general. Nineteen states joined the Justice Department in suing Microsoft in 1998, accusing the company of using its dominance in computer software to drive competitors out of business. After a trial before a federal court, Microsoft was found in 2000 to have engaged in various anti-competitive practices, and the court ruled that the company was to be broken up as a remedy. However, in 2001 an appeals court overturned the order that Microsoft be broken up, though it agreed that the company had violated the antitrust law by illegally maintaining a monopoly. In late 2001, nine states and the Justice Department reached a settlement agreement with Microsoft. But another nine states and the District of Columbia rejected the proposed settlement, concluding that the deal did not go far enough to prevent, in the words of Connecticut Attorney General Richard Blumenthal, a ‘recurrence of violations’ of the law. In November 2002, a federal judge approved the settlement, with some modifications as recommended by the ‘main settling’ states. Most of the non-settling states thereafter dropped their objections to the settlement. Massachusetts, however continued to object, and, in November 2003 as the lone holdout state, they argued to a Washington A.C. appeals court that the settlement should not be accepted. As of June 2004 the Court had not yet ruled on the Massachusetts appeal. During the progress of the case, the NAAG established the Microsoft Working Group of Nineteen States and the District of Columbia, which coordinated the efforts of the states. Thus, when the states commented on the case, or made submissions to the court (for example, submitting a set of proposed remedies in May 2000), they spoke with one voice.

The Network Concept, Competition Authority Networks

107

Microsoft is an interesting example of how the states and federal authorities work together to enforce the antitrust laws, while still retaining independent authority to seek the solutions that they deem most appropriate. The Justice Department and the states worked jointly for many years, yet half of the states initially elected to join the DOJ and settle the case, while the other half felt the better course of action was to seek tougher penalties. c.

Mergers

The states and federal agencies have also joined forces to conduct merger investigations and, when necessary, to challenge transactions as anticompetitive. In the Exxon/Mobil merger, the FTC, 13 states and the District of Columbia entered into settlement agreements with the merging companies after lengthy negotiations. The settlement included a substantial divestiture of more than 2,400 Exxon and Mobil gas stations nation-wide. In addition to the consent agreement accepted by the FTC, the merging parties entered into separate settlement agreements with four states that addressed the local impact of the merger on consumers. In 2001, 12 states — including Iowa and Pennsylvania — joined the Justice Department in deciding to sue to block United Airlines from acquiring US Airways in a proposed deal that was valued at $4.3 billion. In USA Waste Services, 19 states and the Justice Department investigated the merger of waste disposal companies and obtained divestitures in 20 local markets. In the Sony Corp. matter, New York and Illinois joined the DOJ in investigating and obtaining a consent judgment divesting 14 movie theatres in Manhattan and Chicago. In Shell/Texaco, California, Hawaii, Oregon and Washington joined the FTC to investigate and settle competitive concerns arising from the transaction. In US v Morton Plant Health System, the DOJ Antitrust Division and the Florida Attorney General filed a joint lawsuit against the merger of two hospitals in the Tampa Bay, Florida area.

2.

Information Sharing Between the States and the Federal Agencies

Significant efforts have been made to increase information sharing between the federal agencies and state antitrust authorities. In the area of pre-merger disclosure (discussed above), 45 states executed the Voluntary Pre-Merger Disclosure Compact, which is designed to provide an incentive for merging parties to provide voluntarily the states with copies of their HSR filings and other materials that they provide to the federal agencies. Shortly thereafter, the states formulated an Information Sharing Protocol under which they could obtain access to HSR filings with the consent of the merging parties. The Protocol was adopted by the federal agencies, and the FTC later ruled that the

108

Working Paper IV

Protocol would provide the exclusive means by which states could obtain HSR material from the FTC. In 1995, the FTC and DOJ announced a new policy to increase the sharing of information regarding mergers. The policy provides that states may receive information obtained from third parties, without revealing their identity and staff analytic memoranda, after the FTC has decided whether it will challenge the merger. Federal-state information sharing as embodied by the Compact and the Protocol is intended to reduce the risk of inconsistent enforcement by allowing coordination and communication between state and federal antitrust enforcers, starting at the earliest stages of the investigation.

3.

Institutional Structures to Promote Federal-State Coordination

The Executive Working Group for Antitrust comprises the heads of the two federal enforcement agencies and the chairman of the Antitrust Committee of NAAG. The group’s purpose is to discuss enforcement policy issues of common concern to the three agencies, and to coordinate overlapping federal and state enforcement activities, thus preventing the duplication of efforts. The Working Group has made significant contributions to cooperation among the three enforcement entities, with senior NAAG officials stating that the Executive Working Group has been an important vehicle for increased state-federal cooperation on issues of mutual interest. In order to ensure regular communication between the federal and state authorities, in the mid-1990s the Justice Department appointed a Senior Counsel to the Assistant Attorney General for Antitrust. The primary responsibility of this individual is to serve as the Antitrust Division’s liaison to the state attorneys general.

4.

Direct Communication from the States to Federal Authorities

Aside from coordinating efforts directly with federal enforcers, states have taken the initiative to express their views directly to Congress and the agencies regarding federal antitrust legislation initiatives. For example, in 1995, 24 attorneys general submitted a letter to Congress regarding legislation seeking to deregulate the telecommunications industry. In 1985, when the DOJ released its vertical restraint guidelines, the NAAG criticised the guidelines for being too lenient on vertical restraints. The NAAG publicised its opposition to the federal guidelines and, later in 1985, it released its own set of guidelines in an attempt to state the law as the NAAG understood it at the time. The NAAG also lobbied — successfully — the DOJ to have the federal guidelines withdrawn. In addition,

The Network Concept, Competition Authority Networks

109

the NAAG submitted comments regarding the DOJ/FTC Joint Horizontal Merger Guidelines, some of which were accepted regarding merger analysis.

5. 5.1.

Advantages and Disadvantages of Federal-State Cooperation Advantages

Because all antitrust enforcement is conducted under a common body of statutory and case-law, enforcers at the state and federal levels have a common interest in promoting a competitive business environment. Rather than being a hindrance, the differing perspectives and incentives between the states and the federal agencies can actually promote better enforcement. Another level of independent antitrust enforcement means that overall enforcement will be less affected by policy shifts at the national level. Given the variety of incentives and concerns at both the state and federal level, antitrust enforcement does not depend on the policy agenda of one primary enforcer. This was illustrated in the 1980s when the states increased their enforcement of federal antitrust law as a result of the less aggressive enforcement approach of the federal agencies.

5.2.

Disadvantages

Differences in substantive law governing inter-state transactions can increase legal uncertainty, although this does not appear to be a significant problem or issue in the US When transactions have disproportionate effects in certain geographical areas, states may successfully try to block these deals, even if they have efficiencies on a wider scale. In addition, the presence of another layer of antitrust enforcers necessarily means additional bureaucracy, and this can naturally lead to increased transaction costs. Furthermore, the presence of more enforcers can lead to rivalries between agencies that are anxious to block or otherwise affect the efforts of others. At a theoretical level, many of these problems could be reduced or eliminated by giving exclusive enforcement authority to a central antitrust enforcer, which would not be affected by local interests when enforcing the law. A central enforcement agency would also eliminate extra layers of bureaucracy, as well as the potential for different enforcement guidelines and policies among the multiple enforcers. However, a central enforcement agency lacks the sensitivity of state enforcers to local concerns. Many effects that appear to a central enforcer to be minimal, might be more significant to local enforcers because the markets affected might be primarily local ones. Furthermore, the entire business community can be affected when the central enforcers undergo a policy shift. The effects of such a shift could not be tempered without a coordinated group of local

110

Working Paper IV

enforcers that work with each other as well as with the central enforcement authority.

IV.

Does the NAAG Offer any Lessons for the European System?

There are several significant differences between the NAAG (and the US antitrust regime) and the EC network (and the European antitrust system). These differences make it difficult to draw lessons from the NAAG, and certainly caution against a clumsy application of the NAAG experience in Europe. These differences include: • The state Attorneys General are not competition authorities and do not decide cases, unlike the competition authorities of the EC Member States. • There are no courts or judicial bodies in the US ‘network’. • Private actions are available and play an important role in US antitrust enforcement, unlike in the EC; issues involving case allocation are therefore absent or unimportant in the US. • There is nothing similar in the US to the European Walt Wilhelm doctrine, and federal antitrust law does not pre-empt state antitrust law; again, this radically changes the relationships between the state and federal authorities. • There has been little, if any, use of the NAAG for the purposes of harmonising antitrust law in the US. Given these differences, the lessons for the EC are modest. First, the importance of the human factor cannot be sufficiently emphasised. Respect, trust and openness among antitrust enforcers and authorities are probably the most important factors in a successful cohabitation of state and federal powers. Secondly, the successes of the NAAG — notably in the cartel area — result from the pragmatic self-interest of each independent enforcer. This suggests that increased cooperation among NCAs and the EC Commission in the network might be expected to lead to increased enforcement activity.

V Ulf Böge* The Bundeskartellamt and the Competition Authorities of the German Länder

I.

The Network of Competition Authorities in Germany

National competition authorities will be more involved in the application of EC competition law after the revision of Regulation 17/62. The further decentralisation of the application of EC competition law is based on the insight that the effective enforcement of competition rules requires geographical proximity between a competition authority and the relevant facts of a case. At the same time, such decentralisation places new demands on the basis of cooperation between those agencies that are responsible for the application of EC competition rules. Accordingly, the Commission’s proposal to establish a network of European competition authorities proceeds from the same idea. Intensive discussions are taking place between the Commission and the EC Member States as to whether this network should take a rather hierarchical or cooperative form. However, some experience with a ‘network of competition authorities’ has already been accumulated in Europe, if only at Member State level. Following the entry into force of the Act against Restraints of Competition (ARC)1 in 1958, German national competition law has been jointly enforced by the federal and Länder competition authorities. In creating such a network of competition authorities, an element of conflict inevitably arises between two objectives that are equally worth striving for. On the one hand, it is essential to maintain the autonomy of the individual network members so as to enable them to rigorously enforce competition law. On the other hand, consistency in the application of the law within the network as a whole is an important objective. The autonomy of the authorities in the network is primarily connected with the degree to which the law guarantees the independence of the individual authorities in deciding upon individual cases. Thus, the members of the network have a particularly high level of autonomy if the individual authorities have legally defined competencies. However, the absence of instructions from certain network members, or the lack of other ways with which to influence * President of the Bundeskartellamt. 1 The ARC came into force on 1

January 1958; the text of the ARC is available on the Bundeskartellamt’s website (at www.bundeskartellamt.de, ‘English Pages’, ‘Competition Act’).

112

Working Paper V

decisions, are also important parameters for determining the degree of independence. Uniform law can be of great significance for establishing homogeneous competitive conditions. At the same time, close cooperation between the network members is elementary in order to achieve a joint interpretation of the law. A uniform application of the law can ultimately be ensured by a common court of final appeal. The purpose of this paper is to show how these instruments are put to use within the network of German competition authorities so as to reach a balance between the poles of autonomy and consistency.

II. Institutional Framework of the Network of German Competition Authorities The network of German competition authorities comprises the Bundeskartellamt, the Federal Ministry of Economics and Technology, and the respective Land competition authorities of the 16 federal Länder (‘the State competition authorities’). This paper focuses on the cooperation between the Bundeskartellamt and the State competition authorities, as the Ministry of Economics and Technology is only involved to a limited extent in anti-cartel proceedings.2 The Bundeskartellamt is an independent authority that is responsible to the Ministry of Economics and Technology; the State competition authorities are working units within the economics ministries of the Länder. The way in which the German network functions is first conditioned by a legal framework that regulates aspects such as the applicable law, the competencies of the individual network participants, and successive stages of appeal. It also sets minimum standards for the reciprocal exchange of information. All network members apply the same substantive competition rules. This, as well as the main body of cartel procedural law, is laid down in the Act against Restraints of Competition (ARC), which is an integral part of federal law. The Länder do not have their own competition statutes. When there are no special regulations for individual procedural questions in the ARC, the respective general federal or State provisions (such as those governing administrative proceedings) apply. These are largely identical within the Federal Republic, with the result that the competition authorities at federal and State level operate within (virtually) uniform legal frameworks. The allocation of competence between the Bundeskartellamt and State competition authorities, as provided for in the ARC, is generally determined by the 2 The

Federal Minister for Economics and Technology is responsible for granting the ‘ministerial authorisation’ under section 8 of the ARC (special authorisation for cartels) and under section 42 of the ARC (special authorisation for concentrations).

The Network Concept, Competition Authority Networks

113

geographical effects of the restraints of competition under examination.3 The Bundeskartellamt is the competent authority if the restraint in question has effects in more than one State. If this is not the case, competence lies with the relevant State competition authority. In addition, the Bundeskartellamt is vested with some areas of exclusive competence, particularly with respect to merger control and the application of European competition law. Each network member decides within its area of competence, independently of the other competition authorities. Above all, the Bundeskartellamt has no authority to give instructions to or supervise the State competition authorities, and even less to withdraw individual cases from them. The availability of judicial review of the competence of a competition authority is limited. If the parties involved in a case wish to claim that an authority lacks the jurisdiction to adopt a prohibition decision, they must do so during the proceedings that take place before the respective competition authority; this is particularly the case if they later intend to invoke this argument in court. A competition authority’s lack of jurisdiction to impose fines only renders the decision ineffective if the authority evidently had no jurisdiction. Irrespective of this, every competition authority is obliged to examine its own competence. If it becomes apparent — even within the course of a proceeding — that an authority lacks jurisdiction, it would have to transfer the case to the competent authority. In practice, the existing rules avoid conflicts over jurisdiction between the network members. There have been few cases in which companies have tried to appeal the decisions of network members on the grounds of (alleged) lack of jurisdiction. As far as the stages of appeal are concerned, the Federal Supreme Court is the judicial authority that is competent to hear all cartel proceedings, and it examines only the legal aspects of a case. The first and only court of lower instance is the respective higher regional court of the State in which the deciding competition authority is located. This court also examines the findings of fact on which the authority’s decisions are based. Special cartel divisions have been established within the Federal Supreme Court and higher regional courts. As a result, all essential legal questions are ultimately decided by one and the same court and division (after the earlier appeal stages have been exhausted). As part of their information obligations (which are regulated by the the ARC), the State competition authorities are obliged to inform the Bundeskartellamt when they institute proceedings or carry out investigations. Conversely, the same obligation applies to the Bundeskartellamt, which is to inform the relevant State competition authorities that cover the region in which the company concerned is located. There is no formal obligation for the State competition authorities to

3 The provision under section 130 (2) of the ARC, which defines the area of application of the ARC, is

based on the same principle, that is, requiring effects on competition in the domestic market.

114

Working Paper V

consult the Bundeskartellamt before a decision is made. In merger control proceedings, the competition authorities of the Länder in which the relevant companies are located can submit comments if the Bundeskartellamt plans to prohibit a merger. In addition to this statutory obligation to cooperate, the forums for the exchange of experience and cooperation on conceptual issues have proved particularly important features of the German network. For example, representatives of the Bundeskartellamt, the State competition authorities, and representatives of the Ministry of Economics and Technology meet twice a year. The location of these meetings is rotated among the Bundeskartellamt and the State competition authorities. All network members propose topics for the meetings. These can be either concrete cases and experience reports, or more general legal or factual questions that have evolved during enforcement. The primary objective of these meetings is the exchange of experiences and views. Occasionally, common viewpoints of the competition authorities are also adopted.4 In addition to these meetings, the Working Committee on the Supply Industry (electricity, gas and water) also meets twice a year. The current focus of such meetings is on coordination in the area of abuse control in the electricity sector.

III.

Dealing with Potential Network Problems in Practice

In practice, the following questions regarding the joint enforcement of certain substantive rules by a network of competition authorities would have to be considered, irrespective of the specific features of the legal framework that the network is required to apply: • How do the member competition authorities manage to jointly ensure the reasonably consistent enforcement of the same competition rules despite their separate responsibilities? • How can the member competition authorities effectively prevent being played off against each other by the market participants? • How do the member competition authorities manage to develop an effective strategy by which to enforce competition law if the basic conditions of a particular sector change drastically? Thus far, the network of German competition authorities has always been able to find joint solutions to these problems.

4 The latest example of this is the joint position paper on the assessment under competition law of bid-

ding syndicates in tenders for local transport services (see www.bundeskartellamt.de ‘Merkblätter’, ‘Kartellrecht’ (unfortunately available only in German).

The Network Concept, Competition Authority Networks

115

Although the German competition authorities all apply the same substantive law, a question arises as to whether these provisions are interpreted consistently, as in Germany no competition authority may bind the other competition authorities to its own interpretation of a provision or push through its own interpretation by making use of its supervisory powers or by withdrawing proceedings. This may indeed result in differences in the application of the law; incidentally, this situation also occurs in other areas (that is, beyond competition law enforcement). These principles continue to apply at the level of the courts of first instance. It is only before the Federal Supreme Court (the final appellate stage) that the application of the law becomes subject to uniform control. In Germany, the level of consistency in the application of the law is ultimately achieved through this graduated system, and these standardising mechanisms are considered to be sufficient. Hence, throughout the many years of joint application practice, and despite the relative independence of the individual network members, neither the business community nor academics or politicians ever called for more ‘standardisation’ in the application of the law (such as by introducing consultation mechanisms or through even more far-reaching coordination instruments). The practice has shown that the independence of the individual network members does not necessarily produce diverging results; it is instead a useful enhancement of the system, due to the different approaches involved. For example, the competition authority of Bavaria published its own de minimis notice on anticompetitive agreements.5 Although this does not directly contradict the corresponding notice of the Bundeskartellamt, it does set distinct priorities (for instance, by dealing more thoroughly with cooperatives on the demand side). Although the ARC definition of ‘competence’ rarely gives grounds for legal disputes, companies may be tempted to submit a case to a competition authority that is not actually within its jurisdiction. Such ‘forum shopping’ can be successfully addressed by means of a constant and effective exchange of information between the competition authorities, as the following case exemplifies. An association in the building sector notified the Bundeskartellamt of a nationwide market information system. The project would have de facto re-established the organisational framework of a quota cartel that previously existed in various Länder. The project was eventually abandoned, after the Bundeskartellamt expressed concerns and threatened to prohibit the agreement. However, various State competition authorities received notifications of similar, although geographically more narrowly defined, reporting systems shortly afterwards. The network became aware of this conduct through the information exchange between the competition authorities. Although the project was now divided up geographically, there was again a question as to whether the Bundeskartellamt

5 See

www.bayerische-landeskartellbehoerde.de, ‘Informationen’ (available only in German).

116

Working Paper V

was competent to investigate some of the reporting systems which had been notified to the states; this is because they would have had competitive effects in more than one Land. Accordingly, one of the State competition authorities concerned referred its proceedings to the Bundeskartellamt as the competent authority, and the Bundeskartellamt formally prohibited the reporting system. In this way, a circumvention of competence and an attempt to play the network members off against each other was prevented effectively. In the course of the liberalisation of the network-based energy sector and the introduction of competition in the areas of electricity and gas, the German competition authorities have faced some entirely new enforcement problems. These resulted above all from the fact that electricity and gas are network-based sectors, and network operators continue to hold monopoly positions in the individual regions. Therefore, the control of abuse of dominant positions has gained particular significance. On account of the network-related regional market definition, some abuse proceedings against network operators fall within the Bundeskartellamt’s competence, and some fall within that of the State competition authorities. Therefore, nation-wide uniform control can only be achieved if the competition authorities act in a coordinated fashion. Thus, at the initiative of a State competition authority, in Autumn 2000 the Working Committee on the Supply Industry established a working group on electricity network utilisation. This group was supposed to prepare a report concerning the issues of abusive fees for network use and abusive practices relating to network access by competitors. The working group consisted of members of the Bundeskartellamt and several State competition authorities. The report was completed in Spring 2001 and jointly adopted by the member competition authorities. The report6 formulated fundamental guidelines for the treatment and evaluation of abusive practices by network operators. It contains statements on the key areas of the control of abuse of dominance in this sector, and in particular on the application of the comparable market concept and cost control. This ensured that action taken by the network members against abusive conduct by network operators is now similar in its principal features. However, the report does not determine the network members’ future action in all its details. This leaves the individual competition authorities sufficient scope to react flexibly to special conditions within their particular territory and to be able to react adequately to new developments. Although the report is not legally binding, it is of considerable significance for the practical application of the law. As the State competition authorities cooperated in drafting the report, and they finally endorsed it, they feel bound by its statements. This is demonstrated by the fact that both the Bundeskartellamt and the State competition authorities regularly refer to the report in their proceedings. 6 Report

dated 19 April 2001, excerpts available at www.bundeskartellamt.de, ‘English Pages’, ‘Competition Act’.

The Network Concept, Competition Authority Networks

IV.

117

Conclusions

To what extent can the experience gained by the German network of competition authorities be of relevance in a European context? The most important underlying ideas of the German network are similar to those of the European network that is about to be established. Both networks are based on the notion that the effective enforcement of competition law can best be ensured if decisions are made by the authorities that are the most competent to deal with the respective cases. The differences between the two are to be found essentially in terms of administrative law and rules on sanctions that the respective networks follow or will follow in future. Equally different are the concrete methods of cooperation within the two networks. For various reasons, the precise system of allocating competencies between the competition authorities within the German network cannot be transferred to the European network. The fact that the Commission and the national competition authorities have generally parallel competencies inevitably leads to the EC competition authority playing an outstanding role. This also applies to a limited extent to the Bundeskartellamt’s role within the German network. However, it has become clear that the precondition for an effective enforcement of competition law by various authorities is a trustful cooperation of all parties concerned based on equal rights, that is, participation of all authorities in the development of cooperation practices and the interpretation of substantive law. Without such an orientation scenario, members of the German network might conduct proceedings without any coordination. In such a case, the control of abuses of dominant position in the German energy sector would surely have been less effective. Even if the Commission’s outstanding role must be taken into account to an even greater extent in the European network, one thing is clear: if the Commission wants to have the last word in all proceedings and in the development of concepts, affecting everything, up to the setting of conditions for cooperation practices within the network, then there is a danger that the national competition authorities will feel less obliged to take action and will be less committed to their tasks. This would have negative consequences on the effective enforcement of EC competition law. Although it is true that certain corrective mechanisms are required in a system of parallel competencies, these mechanisms must be reasonably proportionate to the expected need for corrections. This need for corrections can be largely reduced, particularly by means of a form of cooperation that is based on genuine partnership within the network.

VI Mark Thatcher* The Causes and Consequences of Regulation by Networks: Telecommunications in Europe

I.

Introduction

In both European general competition policy and telecommunications, regulation is made at two levels: the EC/EU and the national level. Although organisationally and legally distinct, in practice they are intimately linked. Moreover, power is diffused at each level among several actors. This article examines telecommunications in Europe: not only is the sector important in its own right, but it also offers an example of EC regulation that is implemented at the national level, which is very relevant to other domains in which the EC/EU relies on national bodies to ensure that its regulation is enforced. More generally, it offers a detailed case in which there are networks of regulators at EC and national levels, and hence can be used for comparison and lesson-drawing for general competition policy. The various actors in the European telecommunications ‘regulatory space’ (cf Hancher and Moran 1989) form ‘policy networks’. Organisationally distinct actors who are partly inter-dependent but also partly autonomous, become linked without being part of a single organisational hierarchy; their relations are based on exchange, thereby producing policy networks, and combine elements of conflict with those of cooperation (Marin and Mayntz, 1991b, pp. 15–16; Kenis and Schneider, 1991, pp. 41–42). Policy is seen as the product of complex interactions amongst actors drawn from both ‘public’ and ‘private’ sectors (Knoke et al., 1996, p. 6). This article therefore examines the various linkages among key actors. It begins by analysing the shape of EC regulation in telecommunications. Then it looks at the national level, drawing on the experiences of six countries — Britain, France, Germany, Italy, the Netherlands and Spain. Thereafter, it considers the feed back effects of regulation by network on the EC level.

* London School of Economics, Centre for Analysis of Risk and Regulation, and Department of Government (e-mail: [email protected]).

120

Working Paper VI

II. The Shape of European Community Telecommunications Regulation1 EC regulation consists of a mixture of sector-specific legislation and the application of general competition law to the sector. A broad framework has gradually developed over the period from the late 1980s onwards (Thatcher 2001, Schmidt 1997, 1998). It prohibits Member States from continuing legal monopolies (‘special and exclusive rights’) and establishes ‘re-regulatory’ rules governing competition- from interconnection rights to cost-based tariffs.2 The regulatory framework was mostly created through cooperation between national governments and the EC Commission (Thatcher 2001); latterly the European Parliament has also played a significant role. Although there were disputes over the legal basis of Commission action that reached the European Court of Justice (Schmidt 1996, 1998), the Commission acted in partnership with national governments, obtaining agreement on the substance of legislation before proceeding with it. The most recent example concerns local loop unbundling, where a Regulation was rushed through Council and European Parliament in 2001. Although wide-ranging, the EC regulatory framework does not mean the end of national regulation. On the contrary, no central ‘Euro-telecom regulator’ has been established. Instead, EC regulation must be implemented by national regulatory authorities — NRAs. The European Commission monitors implementation and can take action against Member States, including legal cases before the European Court of Justice. The EC regulatory regime permits scope for differences among Member States because it is incomplete, allows different methods of compliance with its requirements and is often very broad. EC legislation is neutral with respect to ownership,3 and hence Member States can decide whether and how to have public ownership of PTOs (public telecom operators). Even following the 2002 Regulatory Package, it does not require national regulatory authorities to be ‘independent’ of elected politicians, but merely that there regulators be ‘legally distinct and functionally independent’ from suppliers of telecommunications services and equipment.4 Indeed, government departments retaining regulatory powers can be counted as NRAs, together with specialist sectoral regulators. EC legislation does not cover the appointment, dismissal or terms of regulators, or

1 References are to the EC as telecommunications regulation took place under the EC pillar of the EU. 2 The key directives have been: Commission 1988, 1990, 1996; Council 1990, 1995, Council and

European Parliament 1997a, 1997b; these are now being replaced by Directives within the 2002 Regulatory Package — discussed in part IV. 3 Article 295 [previously 222] of the Treaty of Rome. 4 European Parliament and Council (2002a).

The Network Concept, Competition Authority Networks

121

their resources. Most EC regulation is in the form of directives; these are binding on Member States as to their aims, but not the means to achieve these. NRAs enjoy considerable discretion under the legislation in key areas for competition such as licensing, universal service and interconnection. Finally, EC rules leave NRAs much scope to determine their internal decision-making procedures: they must respect general principles such as acting in a manner that is ‘objective, proportional and non-discriminatory’ and ‘transparent’, but these principles are very general. There are several reasons why EC regulation provides such scope for national regulation. First, attempts to create a centralised Euro Telecom regulator have been rejected by Member States. Thus Ministers rejected a powerful EC ‘licensing committee’ to police the award of licences (Commission 1992); the Commission did not pursue the idea. During the late 1990s the Commission expressed concerns that NRAs had insufficient powers, resources and independence from incumbent PTOs and that Member States were failing to transpose correctly EC legislation. There was support within it, notably by Martin Bangemann (DG XIII and next DG III Commissioner), for establishing a European-level agency to ensure even and effective implementation of EC regulation.5 The European Parliament repeatedly called for an Euro telecom authority or Committee to prevent fifteen differing regulatory areas developing.6 Despite these pressures no EC-level regulator was created and the Commission pulled back from seeking to establish one (cf Commission 1999a). The main reason was opposition from Member States, who were not ready to accept such a powerful authority (Agence Europe 25.2.97, Financial Times 19.12.97, personal interviews). In the June 2000 proposals, the Commission did not seek to revive ideas of an Euro telecom regulator (Commission 2000). A second reason is that the EC’s regulatory framework was based on compromise and agreement between the Commission and national governments. The latter accepted liberalisation, but wished to keep a degree of national discretion over ‘re- regulation’ in order to meet their industrial and social policy objectives. Thus for instance, they wished to control the licensing of mobile services or many aspects of universal service. In particular, countries such as France sought to reconcile EC liberalisation with ‘service public’ traditions. A third factor is that Commission does not have the resources or expertise to implement EC regulation across all Member States- it would require a large increase in its staffing and budget.

III.

The National Level: Networks of Actors

At the national level, telecommunications regulation also consists of networks of interdependent actors. The key ones are: independent or semi-independent 5 See Agence Europe 24.5.96, 25.2.97, Financial Times 3.7.96, 6 Agence Europe 11.4.95, 20.2.96, 24.5.96, 21.12.96, 24.2.97.

19.12.97.

122

Working Paper VI

regulatory agencies; governments; public telecommunications operators, notably former incumbents; general competition authorities and legislatures; courts. This section analyses the positions of the key actors and shows how they remain dependent on other actors. Cross-national comparison reveals two sets of features. First, all countries have followed a similar path towards privatisation of incumbent operators such as BT, Frauer Télécom, Deutsetre Telecom and Telecom Italia, towards the creation of IRAs and towards powers for both general competition authorities and sectoral IRAs. Secondly, the institutional frameworks differ somewhat among the four countries.

1.

Independent/Semi-Independent Sectoral Regulators

All four countries analysed in this study have established sectoral regulators. However, they did so at different times (see table 1). Oftel is by far the oldest national telecommunications regulator; the others are much more recent, dating from 1996–1998. Thus in most EC countries, the sectoral regulators are very young. Table 1:

National Independent Sectoral Regulators

Country Britain France Germany Italy Netherlands Spain

Sectoral regulator

Date of creation Telecommunications)*

Oftel (The Office of ART (Autorité de Regulation des Télécommunications) Regulierungsbehörde für Telekommunikation und Post (RegTP) AGCOM (Autorità per le Garanzie nelle Communicazioni) OPTA (Onafhankelijke Post en Telecommuniatie Autoriteit) CMT (Comisión del Mercado de Telecomunicaciones)

1984 1997 1998 1998 1997 19977

The organisational status of the sectoral regulators varies. Some are legally independent or semi-independent bodies from governments — for instance, the ART, AGCOM, CMT and OPTA. In contrast, technically Oftel is a non-ministerial government department, whilst the RegTP falls within the scope of the Federal Ministry of Economic Affairs and hence, whilst independent, is also accountable to it. Three sectoral regulators have a broader remit than telecommunications: the German RegTP cover both postal and telecommunications services; the Italian Autorità per le Garanzie nelle Communicazioni (AGCOM), and since 2003 Ofcom in Britian, are responsible for both telecommunications and broadcasting. * Replaced in 2003 by Ofcom (Office for Communications). 7 Created by Decree in 1996, but functions established in 1997

law.

The Network Concept, Competition Authority Networks

123

The numbers of regulators range from three in Germany and the Netherlands, five in France, to nine in Italy and Spain. Britain was an exception, as powers were vested in one individual regulator, the Director General of Telecommunications; however, a move to a nine-person Commission took place with the 2003 Communications Act. Although sectoral regulators are independent from governments, elected politicians play a leading role in the selection of members. In Britain, the Secretary of State for Trade and Industry appointed the Director General of Telecommunications and had great discretion over the choice of person. In other countries, there are more complex and sophisticated systems of appointment. Powers over membership are less concentrated in an individual government minister, as several bodies nominate or elect members, including the government and the legislature or heads of its chambers. This spreads the likely range of political parties which can nominate members; in Italy, the rule of electing several members was devised so that the two largest blocs in each house of the legislature would each get one member of the telecommunications commission and one of the broadcasting commission. Regulators are appointed for fixed terms and cannot be dismissed except under exceptional circumstances. Lengths of appointment vary, the longest being seven years for the CMT. When there are multi-person regulatory bodies, terms of members may expire at different times (for instance, in France). The pay and other conditions of regulators also vary, together with who has the power to Table 2:

Membership of Sectoral Regulators

Country

Number of regulators

Powers to nominate

Britain

One — The Director General of Telecommunications (after 2003, nine members of ofcom) 5 members of ART

Secretary of State for Trade and Industry

France

3 nominated by government, including its President; 2 nominated by heads of National Assembly and Sénat respectively Germany President and 2 Vice-Presidents, Nominated by the federal government but decisions taken by on proposal of the Advisory Council, chambers set up by the composed of members of two houses Ministry of Economic Affairs of Bundestag Italy 9 members of AGCOM President nominated by the President on proposal of prime minister; 4 elected by Chamber of Deputies and 4 by Senate (2 for Telecommunications Commission and 2 for Broadcasting Commission) Netherlands Chairman and 2 Vice-Chairs Appointed by the Minister Spain 9 — President, Vice-President Appointed by government, but legislature’s and 7 Counsellors Telecommunications Commission can approve/reject

124

Working Paper VI

determine those these. In Britain, the Secretary of State decided the Director General of Telecommunication’s remuneration and other conditions and could reappoint him/her. In other countries, more of these matters are defined in legislation, reducing the legal powers of ministers; thus, for instance, in France, legislation stipulates that members of the ART have non-renewable six year terms, and enjoy the employment conditions of senior civil servants. Table 3:

Length of Term of Office

Regulatory body

Length of term of members

Oftel ART RegTP AGCOM OPTA CMT

Maximum 5 years renewable 6 years non-renewable 5 years 7 years non-renewable 4 years renewable 7 years

There are very few criteria for selection in most countries; even if specified, they remain very general (for instance, in Spain, the law merely states that CMT members will have ‘relevant acknowledged competence’). Similarly, most procedures merely concern who has the power to nominate regulators. There are almost no rules to choose candidates on merit, such as opening the posts to public competition, holding interviews or forming panels to assess rival candidates. National legislation gives the sectoral regulators duties and functions. These include ensuring competition, but also other duties such as aiding consumer interests, ensuring provision of affordable services, safeguarding various groups (for instance the disabled or rural dwellers) and safeguarding ‘public interests’ such as national security. They are often set out in very general terms. Hence sectoral regulators face broad objectives that may conflict with each other. This allows them to enjoy considerable discretion in interpreting their aims. Moreover, they have decision-making powers and advisory functions, especially towards governments; thus, they are both independent decision-makers and government counsellors. In general, governments (often specifically Finance Ministries) and legislatures set the maximum spending allowed by sectoral regulators. Levels of resources vary greatly. thus for example, Oftel spent 19.6m Euro in 1998–1999, whereas the RegTP spent 63.8m Euro (against a maximum possible of 150.5m Euro). In terms of numbers of staff, several of the sectoral regulators are relatively small — for instance, the ART has only 145 staff and AGCOM is allowed a maximum of 320 to cover both telecommunications and broadcasting, whereas the RegTP has 2600 employees at its headquarters (albeit that it covers both telecommunications and postal services that that 12% of staff are part-time) and 1600 in regional offices. Staff pay and conditions are often those of the public sector, although some regulators (such as AGCOM) have scope to vary terms and salary levels from the general civil service. The majority of staff are current or former civil servants,

The Network Concept, Competition Authority Networks

125

especially in newly created regulators in Italy, Spain and Germany, where most of the sectoral regulators’ employees merely moved from PTT Ministries. In addition, a significant number of regulatory staff are civil servants on secondment from government departments. Even in a long-established body such as Oftel, more than half of employees (54%) were civil servants. Thus links with the public sector remain strong, at least in terms of personnel. Table 4:

The Resources of Sectoral Regulators (1999/2000)

Regulator

Number of Staff

Budget (Million Euro)

Oftel ART RegTP AGCOM OPTA CMT

190 145 4200 178 101 113

22.4 14.1 63.8 26.8 9.9 7.6

2.

Elected Politicians: Governments and Legislatures

Although sectoral regulators have been established, governments and legislatures retain important powers. They nominate some or all of the members of the regulatory body. They control key resources available to the NRAs, especially their finance. They often have high-level functions of guidance and accountability. For instance, in Britain, the Secretary of State could issue general guidelines to Oftel; in Germany, the RegTP is within the scope of business of the Ministry of Economic Affairs and hence accountable to it, while the Ministry also organises the RegTP’s ‘ruling chambers’. Governments can legislate to alter the position of semi-independent regulators. Moreover, in Germany, France, Italy, the Netherlands and Spain, areas such as the definition of universal service or the allocation of responsibilities for universal service provision, are covered by legislation, frequently in decrees issued by governments. In addition, governments continue to hold key powers over licensing (especially for mobile networks and services/networks for which individual licences are required). Thus governments are both themselves regulators of the market, due to their continuing powers over the market, and regulators/supervisors of sectoral regulators. Continuing public ownership of PTOs further complicates the distribution of functions. Governments enjoy powers over publicly owned incumbents thanks to their shareholdings. In France, these powers are made explicit in legislation, which specifies, for example, that France Télécom’s tariff changes must be approved by the government. Public ownership means that governments are both owners of suppliers in a competitive market and regulators. Governments enjoy great regulatory discretion over the use of their powers. They face few specific duties; even in Britain, these are very broadly defined. Moreover, there are few procedural guidelines specific to telecommunications.

126

Working Paper VI

As a result, control of the government’s regulatory roles depends heavily on the political process and EC and administrative law. Although elected politicians retain significant powers, they too are dependent on other actors, notably the independent sectoral regulators. The latter have expertise and information concerning the sector that government departments often lack. Moreover, independent regulatory agencies can have legitimacy, as a result of being seen to be independent of political pressures and of transparent decision-making procedures. Finally, governments may need agencies to be strong and independent in order to enjoy the benefits of delegation, such as enhanced credibility to attract investment, blame-shifting and dealing with information asymmetries (cf Thatcher 2002, Levy and Spiller 1996)

3.

PTOs (Public Telecom Operators)

Incumbent PTOs continue to dominate most national markets, with shares that vary from 60%–90%. Moreover, many entrants are consortia with well-established national firms — for instance, Vivendi or Scottish Power. Only in mobile markets are Europe-wide PTOs now emerging, notably Vodafone and Orange (owned by France Télécom). There continue to be significant differences in the ownership and status of the incumbent PTOs. At one extreme, there is no state shareholding or golden share in BT, and only a small state shareholding in Telecom Italia, together with a golden share (which the government has pledged not to use and whose legality has been attacked under European law). However, at the other extreme, France Télécom still has a substantial minority state stake (92%). Moreover, over 85% of France Télécom’s employees are still civil servants. In Germany, Deutsche Telekom remains majority state-owned, although there are plans to sell further stakes. Public ownership of PTOs means that governments are regulators and owners of a supplier in a competitive market. It provides incentives for government action to protect the value of its asset. Hence the industrial policy and financial pressures on governments differ considerably across nations. Nevertheless, even privately owned firms such as BT or Telecom Italia remain symbols of national industrial policy and have political influence.

4.

General Competition Authorities

The position of general competition authorities is crucial in regulating a liberalised telecommunications market: they offer an alternative regulator to sectoral bodies; often they have a longer history; they provide a different model to sectoral regulation, notably ex post general rules that apply to all sectors and have been developed from broad experience; frequently their position has been strengthened in recent years (Wilks 2002, Doern and Wilks 1996).

The Network Concept, Competition Authority Networks

127

National legislation contains a number of provisions concerning the potential overlap in jurisdictions between sectoral and general competition regulators. In most of the six countries, it requires the two regulators to cooperate and share information. Thus, for instance, in France, the President of the ART must inform the Conseil de la Concurrence of any apparent abuse of a dominant position and anti-competitive behaviour, whilst the latter must ask the ART’s advice on cases in telecommunication and then communicate information about cases. Similarly, in Germany, the RegTP and the Federal Cartel Office (Bundeskartellamt) must exchange information on matters that touch on both their jurisdictions and must reach agreement on whether a form has market dominance. Thereafter, however, lie difficult issues of delineating the scope of the two. The division of activities is generally defined only at a high level of generality. In Britain, the Director General of Telecommunications and the Director General of Fair Trading (DGFT) have concurrent jurisdiction to provide guidance or decisions; the latter can ask the former to exercise his powers and under the Competition Act 1998, matters relating to the telecommunications sector will normally be passed to Oftel. In France, telecommunications remain subject to general competition law and the ART is obliged to notify matters to the Conseil de la Concurrence which fall within the latter’s competencies; however, the circumstances under which this would be so are not clear cut. In Germany, the 1996 Telecommunications Act is intended to complement competition law, and distinguishes between areas requiring sector-specific regulation in order to ensure universal service and control market power, and those where general competition law can apply. In Italy, AGCOM has powers over concentrations and investigative powers, but the general competition authority (the Autorità Garante della Concorrenza e del Mercato) has final jurisdiction over anti-trust matters, notably abuses of dominant positions.

5.

The Courts

Institutionally, the courts stand in the background of telecommunications regulation. Nevertheless, they can be key players. The implicit threat of legal action exists in many relationships: regulators’ decisions can be open to forms of administrative or judicial review; if legal rights are granted, they are enforceable through the courts. In some countries, notably Germany, actors have had frequent recourse to the courts. In others, such as Britain, court actions have been rare.

IV.

Key Regulatory Decisions in Practice

How have national regulatory authorities used their discretion in practice? This section looks at four key areas.

128

1.

Working Paper VI

Licensing

Individual licences (‘authorisations’ under the new Regulatory package) are no longer needed for many telecommunications services under EC and national law.8 Nevertheless, EC Member States can insist on individual licences for mobile networks (due to limits on frequencies available) and were also able to do so for public voice telephony services and networks. Moreover, whilst EC law restricts the circumstances under which an individual licence can be refused and the licence conditions that can be imposed, Member States retain a degree of discretion and considerable differences exist from one Member State to another.9 Most countries have required individual licences for public networks and voice telephony, as well as mobile networks; the Netherlands is an exception, being one of the most liberal regimes in the EU by restricting individual licences to mobile networks. In several Member States, including Britain and France, the power to issue individual licences, and hence determine licence conditions (within EC law) lies with telecommunications ministers, whereas in others, such as Germany, the responsibility lies with the sectoral regulator. In France, the minister can insist on conditions (a cahier des charges) covering matters such as confidentiality, interconnection, interoperability of services, protection of defence and public security and funding of public service. The power to issue mobile licences has become particularly significant in the last ten years, as mobile networks have provided the main source of competition, but spectrum capacity limits the number of networks that can operate. In most EU countries, the method of allocating licences, notably the choice between auctions and beauty contests, fees and the criteria applied, have been decided by governments. Strikingly different choices have been made. Thus for instance, UMTS licences have been allocated according to very different systems and costs. Spain made its decisions early and chose a low-fee beauty contest that raised 0.5b Euro, but then sought indirect means of raising further revenue. In contrast, Britain and Germany used auctions, and raised massive sums (37.4b Euro and 50.5b Euro). Italy developed a two-stage method, whereby candidates qualified and then bid, raising 14 Euro. After much political debate, in which the ART opposed an auction, an option supported by the Finance Ministry, France has adopted a high fee beauty contest (20 billion Euro over 15 years), but then reduced the real costs following the ‘telecom crash’ and a failure to allocate all licences. Equally, the number of licences varies from four to six across the countries. Powers over the modification of individual licences offers another potentially significant regulatory matter, especially as the telecommunications sector is

8 Many

services either do not require a licence or are covered by a class licence and hence are merely registered. 9 Cf also European Commission 1999b.

The Network Concept, Competition Authority Networks

129

evolving rapidly. There is a significant contrast between continental Europe and Britain. In the latter, under the 1984 Telecommunications Act the Director General of Telecommunications can agree a licence change with an operator, subject to a veto by the Secretary of State for Trade and Industry; if agreement cannot be reached, or if the Secretary of State vetoes it, the Director General of Telecommunications can refer the matter to the Competition Commission, setting out why it could operate or be expected to operate ‘against the public interest’. If the Competition Commission issues a report favourable to the Director General’s reference, the latter can make such modifications of licence conditions ‘as appear to him that are requisite for the purpose of remedying or preventing the adverse effects specified in the [Competition Commission] report’, without the agreement of either the licensee or the Secretary of State. Thus the Director General has considerable discretion over the terms of any reference and then, if the report is favourable, what licence changes to make. Oftel has used this power, or more often the threat of using it, in many negotiations with operators, especially BT. In contrast, NRAs in other EU Member States lack such powers; licence changes usually involve action by governments, as licence issuer and/or legislation. In Spain and Italy, the position of the sectoral regulator is further weakened by the fact that Telecom Italia and Telefonica both operate under concessions issued by the government and not under licences.

2.

Universal Service

EC law states that a limited number of services, notably low-speed fixed public telephone lines and emergency services, are to be available to all users ‘at an affordable cost’.10 Nevertheless, national regulatory authorities have powers over the financing and provision of universal service. NRAs are allowed under EC law to finance losses due to the provision of ‘affordable’ universal services either through a special fund or through interconnection charges. Only France has chosen to levy a special interconnection fee; other Member States have a special fund, into which PTOs pay according to their share of the market and which is paid to operators who provide universal service. In several Member States, rules limit that PTOs is able to supply universal service. In France, the 1996 law states that universal service operators must be able to meet obligations throughout France and also that France Télécom is entrusted with universal service.11 A similar position holds in the Netherlands. In Spain, the CMT has entrusted universal service to Telefonica. In contrast, in Italy, suppliers can offer to provide universal service over all or part of the country and hence draw on the universal service fund. 10 Other services are to be available to all users, but without a requirement of affordability. 11 In 2000, provision of the new ‘social tariff’ for low-income users has been opened to other operators.

130

Working Paper VI

The cost of universal service is determined by the sectoral regulators. They have followed different procedures and reached very different figures. Whilst great reliance has been placed on figures provided by the incumbent, in Britain, Oftel has declared that it was unconvinced by BT’s initial figures, especially given the long-term benefits of universal service provision; it has therefore refused to accept that there is a ‘undue’ universal service cost to BT, until and unless more reliable figures are provided. In contrast, the universal service fund in Italy is 62m Euro, the figure having been set after consultations, whilst in France, it is a maximum of 0.8% of fixed telephony revenue, this being 183m Euro in 2000. No ‘undue’ test of the cost to operators of providing universal service is set in these countries. At present, universal service offers a limited, albeit significant, cost relative to the overall telecommunications market. However, it has symbolic importance, especially for ‘public service’. When ‘social exclusion’ becomes politically significant, new programmes are introduced; thus, for example, in France, a reduced tariff scheme has recently been put into place for low-income users. Moreover, the scope of universal service may be expanded. Finding appropriate mechanisms to determine the scope of universal service may therefore become important both financially and for reconciling competition and ‘the public interest’.

3.

Interconnection, Unbundling and Information

Interconnection is a key feature of competition, and hence features prominently in EC law. Operators of public services and networks must be able to interconnect with each other and PTOs with ‘significant market power’ had to produce a Reference Interconnection Offer (RIO), which (like all prices) should be cost-based, and must meet all ‘reasonable’ requests for interconnection (under the new Regulatory Package, they must produce Reference Offers for unbundled access to twisted pair wire used for the local loop). There are EC Recommendations and Notices for interconnection and for accounting separation and cost accounting. Although EC law was relatively well developed in its coverage of interconnection and related issues in the 1990s, NRAs continued to have significant discretion, which they have used, resulting in important differences across Member States.12 One issue concerned how and when the RIO is drawn up. In Britain, interconnection is negotiated between operators, disputes go to Oftel and BT’s RIO is adjusted throughout the year, frequently following disputes. In other countries, sectoral regulators play a more direct role and adjustments can be slower- for instance, in Spain, Telefonica’s RIO is negotiated between it and the CMT and is fixed for the year. Another area of variation concerns information. 12 Cf

European Commission 1999b.

The Network Concept, Competition Authority Networks

131

NRAs are empowered to ask to for cost information to verify whether the tariffs, including interconnection rates, of operators with significant market power, are sufficiently cost-based; they therefore enjoy discretion over what information to demand — for instance, what level of disaggregation between different services. The extent and nature of accounting separation offers another important area of discretion: the extent to which it is taken varies, with Oftel having developed particularly sophisticated systems. Effective accounting separation is important to allow challenges on matters such as excessive interconnection rates and crosssubsidies. Charges must be sufficiently unbundled so that applicants for interconnection do not have to pay for unnecessary services. This again allows for interpretation by PTOs and discretion for NRAs. Local loop unbundling has offered another route for national variation. Britain has consciously opted for competition among infrastructures and not accelerated local loop unbundling, while France Télécom also sought to postpone it; in contrast, certain states, led by Germany, have pressed forward with it. Despite these variations, interconnection offers an example of relatively successful EC regulation. All Member States have ensured that RIOs are available and that interconnection occurs in practice. Remaining issues therefore concern detailed matters such as delays and more detailed cost accounting. In contrast, local loop unbundling shows the difficulties of regulating a rapidly changing market solely at the EC level.

4.

Retail Prices

EC legislation states that tariffs must be cost-based. Yet it contains few specific details on forms of retail price control; such matters are largely left to NRAs, although they must follow general principles such as non-discrimination and transparency. The result has been that although incumbents face retail price controls in all Member States, significant variations have arisen- for example, which services have price controls and the use and extent of ‘baskets’ of services. In France and Spain, the Minister has to approve France Télécom and Telefonica’s tariffs; in France; the ART must be consulted, but the government can ignore its opinion, and has done so on several occasions. In Italy, some of Telecom Italia’s tariffs are covered by a price cap and AGCOM has the power to challenge tariffs of all operators (but its prior approval is not required). In Britain, a retail price cap is a condition in BT’s licence, and hence modification involves negotiation between BT and Oftel, subject to a veto by the Secretary of State, or a reference to the Competition Commission (see above). Similarly, the degree of cost information and its form differ across countries. The result is that it is not easy to compare across nations, whilst if NRAs do not demand sufficiently detailed information, it becomes difficult to mount challenges based on cost-orientation of tariffs.

132

V.

Working Paper VI

Conclusion: Feedback Effects at the EC Level

Telecommunications regulation in Europe is characterised by the existence of networks of inter-dependent but also partially autonomous actors. The development of the EC’s regulatory framework has involved several actors, notably national governments and the Commission. The EC relies on national authorities for implementation of its regulation. At the national level, there are many actors with powers over regulation. Moreover the combination of implementation of EC regulation at the national level and gaps in the EC’s regulatory regime has allowed cross-national diversity in institutional arrangements and substantive regulatory decisions. Several strategies can be envisaged at the EC level to respond to such diversity: accepting cross-national differences; benchmarking and cross-national comparison; the creation of new networks to aid coordination of existing actors; strengthening of the EC’s legal framework. No one strategy has been decisively chosen, although the 2002 Regulatory Package has many characteristics of the last strategy. Rather, elements of each have existed. They have represented ‘feedback effects’ as the effects of earlier phases of regulation have fed back into new EC action. The first strategy of accepting cross-national differences in certain aspects of regulation was the basis of the original partnership between the Commission and national governments in developing EC regulation. Certain aspects of the Regulatory Package continues the tradition- notably those which are ‘permissive’ for NRAs, notably over universal service (European Parliament and Council 2002d). Diversity may allow regulators to be sensitive to local conditions and to experiment. If ‘regulatory competition’ operates beneficially, better regulation may drive out worse regulation, through mechanisms such as the relocation of firms. However, regulatory diversity poses problems in a single market, notably that of ‘fair competition’. It raises the transaction costs of cross-border activity by companies, who find that organisational arrangements and regulatory decisions differ. Finally, regulatory competition may not operate strongly or may lead to undesired ‘races to the bottom’. Cross-national comparison and learning has provided a second strategy and feedback effect. The Commission monitors and compares ‘performance’ across the EU and publishes indices (notably in its Annual Reports on the Implementation of the Telecommunications Regulatory Package). Such comparison rapidly slides into ‘benchmarking’ and represents an informal pressure on Member States. Much less formal learning and copying has taken place among regulators, especially IRAs, whom now regularly exchange information and examining each other’s practices. A third feedback effect has been the establishment of Europe-wide formal and informal groups of national regulators. Thus for instance, the Commission established the High Level Regulators Group, which under the 2002 Directives will be

The Network Concept, Competition Authority Networks

133

replaced with the more formal European Regulators Group for Electronic Communications. The Commission is also dependent on NRAs — for implementation, information and often support in developing the EC regulatory framework. Meanwhile, the independent regulators have formed their own Independent Regulators’ Group. Its functions are to allow them to learn form each other but also to coordinate their positions (for instance, over new EC legislation). The final strategy has been the traditional one driving European integration: new legislation, ‘soft law’ and occasional legal action by the EC Commission to deal with gaps in EC regulation to attempt to prevent ‘excessive’ diversity. Thus for instance, interconnection has seen Recommendations and Notices on matters such as the methods to be used by NRAs to calculate costs for interconnection. When large-scale discrepancies arise, the Commission is quick to point out the need for coordinated action. Thus for instance, the recent UMTS licensing round has led to discussion of new EC rules for mobile licensing, a matter traditionally largely left to Member States. Legal action has been rare, but has been used on occasion.13 Then in 2002, the Regulatory Package was passed, comprising six directives. In several areas, notably the behaviour of NRAs, access and authorisation, it laid down a series of new rules. Thus for example, Member States are to publish the tasks to be undertaken by NRAs, provide rights of appeal from the decisions of NRAs to a body independent of that NRA, ensure transparency, notify the Commission and other NRAs over measures that may affect trade among Member States, with the Commission being empowered to order the NRA to withdraw measures. (European Parliament and Council 2002a). There are detailed rules on matters such as numbering, collocation and defining significant market power/dominance. Similarly, for access and interconnection, there is a host of rules concerning the obligations of undertakings, transparency and the duties of NRAs (European Parliament and Council 2002b). The Directive on authorisation of electronic communication systems (European Parliament and Council 2002c) greatly limits the scope for individual licences and has rules on matters such as fees that could be used to hinder competition. The effects of the Package are yet to be seen- NRAs remain and retain powers and discretion, but it does appear to provide many legal opportunities and rights for the Commission and private actors. These feedback effects are in their early days. Nevertheless, regulation by networks can be seen to give rise to processes whereby new pressures and opportunities for EC action arise, resulting in the further development of European regulation.

13 For instance, there is currently a case against France concerning its calculations for the costs of universal service.

VII

Ex Cursus

Burkard Eberlein* Policy Coordination without Centralisation? Informal Network Governance in EU Single Market Regulation

I. Introduction: Informal Governance and the Challenge of Policy Coordination This paper takes the following approach to the theme of informal governance. Based on empirical evidence from single market regulation (utility regulation in particular), it investigates the potential of informal network governance to address the challenge of supranational policy coordination. Can informal, network governance offer an ‘escape route’ (Héritier 1999) to effective policy coordination, in a context of diversity and fragmentation of political authority, where centralisation and hierarchical control are not available? This paper deliberately takes a policymaker’s perspective that gives more attention to ‘system effectiveness’ rather than ‘citizen participation’ (Dahl 1994). I should stress that this is not to downplay the issue of democratic legitimacy, which is in the centre of much current analysis of informal governance (e.g. Rhinard 2002). On a very general level, the challenge of policy coordination in ‘post-national’ settings can be construed as a ‘dilemma of governance’ (Keohane/ Nye 2001: 1): on the one hand, the growth of functional interdependence (be it economic or environmental) generates a greater need for policy coordination, due to the increase of externalities between different political units. On the other hand, political authority is fragmented, both horizontally and vertically. The nation-state has lost its role as the classical container and ‘centraliser’ of political authority: policy responsibilities and powers are increasingly shared with private, for-profit and non-profit organisations; and substantial policymaking powers have been either devolved to sub-national levels of government, or flown upwards to international organisations (e.g. Kettl 2000). Thus, centralisation and hierarchical control are not easily available options to ensure policy coordination.

* Department of Political Sciences, TU München. The author can be reached by email at [email protected]. This paper was written in the context of a larger research project on the regulation of public utilities in the EU (Project Director Edgar Grande, TU München) and partly draws on Eberlein/Grande 2002. For helpful comments on earlier versions I am indebted to Nicolai Dose, Philipp Genschel, Virginie Guiraudon, Dieter Kerwer, Dieter Wolf, and the participants of the Research Seminar, Joint Center for International Studies (CIS), University and ETH Zürich.

136

Working Paper VII

While the EU is an exceptionally powerful and intrusive international organisation, it is, I will argue, nevertheless faced with the same kind of dilemma when we look at market regulation. After briefly explaining why the EU qualifies for the governance dilemma, I will proceed as follows. In a first step, I will discuss two scenarios of how regulation is organised and coordinated in the EU. Secondly, I will confront the two scenarios with empirical evidence. This will lead me, thirdly, to specify the regulatory dilemma faced by EU policy-makers. I will then, fourthly, present transnational regulatory networks that have emerged in response to this dilemma. I shall specify the logic of network governance (decentralised coordination) and show how regulatory networks may provide for ‘soft coordination’. Fifthly, and finally, I will investigate the conditions and limits of network effectiveness. In a nutshell, my argument is the following: transnational regulatory networks can offer important escape routes to EU policy coordination without formal centralisation. Networks preserve important benefits of decentralisation while providing for mechanisms designed to achieve ‘soft coordination’. However, I also argue that network governance should not be viewed as a panacea but rather as one mode of governance that is endowed with comparative (not absolute) advantages.

II.

The EU and the Governance Dilemma of Policy Coordination

Why does the European Union qualify as a case study for the governance dilemma? On one side of the equation, the answer is pretty straightforward. With the completion of the Single Market and the transition to a single currency, the EU has reached an exceptionally high degree of economic interdependence between Member States, which requires extensive policy coordination. But what about fragmentation of political authority? Compared to other international organisations, the EU is, no doubt, the most significant and powerful of its kind. It has strong supranational institutions such as the European Commission and the European Court of Justice, and there is a sophisticated set of common rules governing multiple policy sectors. And this should be particularly true for Single Market issues, with the strong legal footing of ‘negative integration’. However, even in most integrated Single Market areas, the EU polity has not matured to the point of resembling a federal state with a clear political centre. It is not necessary here to engage in the traditional debate between inter-governmentalists who view the EU as a creature of Member States (e.g. Moravcsik 1998) and neo-functionalists or supra-nationalists who stress the independent role and life acquired by supranational ‘agents’ (e.g. Sandholtz/ Stone Sweet 1997). If we view the EU as political authority structure, it seems fair to state that divided authority and power sharing across multiple layers of governments are its constitutive features. It is a polity that lacks a ‘centre of

The Network Concept, Competition Authority Networks

137

accumulated authority’ (Hooghe 1996: 18), that is ‘highly decentralised and atomised’ (Hix 1999: 5) and that is, to a large extent, governed by consensus and negotiation (Kohler-Koch 1999).

III. Market Regulation in the EU Multi-Level Setting: The Functionalist and the Realist Scenarios Regulation, defined as specific form of rule-based government intervention in the economy (e.g. Noll 1985), is a growth industry in Europe. Two sources of regulatory growth are of particular importance. First, in the aftermath of (partly EU-driven) privatisation and liberalisation that ended state ownership and monopolies across a host of economic sectors, regulatory bodies have been created or strengthened to both foster and monitor market competition. They received powers mainly to protect competition but also to promote other social and political goals typically not served by competitive markets. Privatisation and liberalisation did, thus, not result in a simple retreat of the state but rather in a redefinition of its role (Müller/ Wright 1994): ‘from a producer of goods and services to that of a regulator’ (Majone 1991: 84). Quite paradoxically, ‘freer markets’ are monitored by ‘more rules’ (Vogel 1996; also Majone 1994a). Secondly, regulatory growth has also been fuelled by the increasing political salience of risk-related policy fields such as food safety and environmental protection. Increased public awareness and demands, often reinforced by scandals (e.g. the BSE food scandal), and technically very complex policy features, combined to create strong pressures for more expert regulation. Furthermore, other public interest goals, such as gender equality, mobilised sufficient support to get onto the regulatory agenda in Europe. Thus, the Single Market is not simply about de-regulation but involves considerable re-regulation, both to make and to correct markets. Regulation in Europe has grown more important — but how is regulation organised in a multi-layered polity, and what does this mean for policy coordination? In the literature, we find two contrasting scenarios on the organisation of regulation in Europe: a functionalist and a realist account. The first scenario is grounded in a transaction-cost politics approach to delegation. It emphasises the functional pressures for regulatory functions to be delegated to the EU level in the context of the Single Market (Majone 1994b, 1996). There are a number of reasons for a centralisation of regulatory tasks and powers on the EU level. First, on the demand side, centralisation serves to correct policy externalities and to reduce transaction costs as compared to decentralised rule making and application in lower-level units. But not only multi-national companies operating in the European market (or environmental pressure groups) should prefer uniform rules. Also the Member States collectively stand to benefit from a self-commitment to a delegation or centralisation of regulatory powers. This is because centralisation

138

Working Paper VII

prevents national regulatory opportunism that would reduce aggregate EU welfare. Secondly, on the supply side, regulation is the privileged avenue of supranational institutions, and notably of the European Commission, to advance EU powers and policies. This is because regulation is not subject to fiscal restrictions that limit EU growth as a distributive welfare state. The EU can thus most easily grow as a regulatory state and, therefore, suggests itself as the most appropriate level of regulation. But does form follow function (functional pressure)? This is where the alternative scenario comes in with a much more sceptical view on the chances for regulation to be centralised and managed by an EU regulatory state. This realist account of European regulation starts from the asymmetrical nature of the European integration process (Scharpf 1997, 1999). Regulatory measures of ‘negative integration’ (market-creation or opening) can build on a strong legal footing in the supranational dynamic of integration, and this quite independent of member state control (the European Treaties and the pro-market rulings of the European Court of Justice). ‘Positive integration’, that is re-regulation of markets on the EU level, by contrast, is very difficult to achieve since it depends, in spite of qualified majority voting, on explicit political consensus among Member States that often have diverging regulatory interests rooted in stark economic differences (e.g. low-wage versus high-wage economies). The realist scenario concedes that these conflicts of interests do not rule out EU level regulation altogether. European regulation is expected to be more successful in the realm of social regulation (environment, health, and workplace safety). Here, regulation is mostly about product-related rules (enhancing the value of products), that complement the market-creating logic of negative integration (Scharpf 1996: 119). Economic regulation, such as the regulation of energy or telecom markets, is a different matter. It makes rules about the conditions of production (increasing costs without enhancing product value), thus bringing economic differences and diverging member states’ interest into play and conflict.

IV. Testing the Two Scenarios: National Regime Divergence in the EU Framework How do the competing scenarios compare to the empirical picture of regulation in Europe? If we take the distribution of regulatory powers and institutions as measuring rod, the realist scenario seems to fare much better. Sure, expressed in numbers of legal directives and regulations, there has been an impressive growth of EU-level regulation, and the policy scope of EU regulation has considerably widened as well. Also, new independent agencies have been created on the EU level, for example for the environment and for safety and health at work (Kreher/Meny 1997).

The Network Concept, Competition Authority Networks

139

However, a closer look at these 12 bodies reveals that, if they perform regulatory functions at all, they do not have the kind of regulatory powers (rulemaking, application, and litigation) that are usually given to independent regulatory agencies as we find them in the US. Take the European Environment Agency (EEA), for example, which, by statute, is limited to information gathering and dissemination, and compare to the US counterpart EPA. And even the European Agency for the Evaluation of Medicinal Products (EMEA), that comes closest to a fully-fledged regulatory authority, formally can only make recommendations, while decision-making lies with committees controlled by Member States. In short, EU agencies are not designed to take over from national regulators or even less to constitute a regulatory ‘fourth branch of government’. Furthermore, the fact that all European agencies operate in the realm of social regulation points to a broader asymmetry of European regulatory activity, roughly consistent with the realist account. The focus of European regulation is quite clearly on social regulation, and here, as evidence from environmental or occupational safety regulation shows, EU regulation is significant and rather successful in terms of policy coordination by imposing common rules (e.g. Eichener 1997). By contrast, in the field of economic regulation of markets, we do not find EU-level regulatory agencies at all. Attempts to establish an independent EU regulator for telecommunications, for example, failed. Rather, regulatory institutions to oversee market competition in the aftermath of privatisation and liberalisation were created or strengthened on the national level (e.g. Thatcher 2002). Sure, these bodies operate in a larger ‘regulatory framework’ defined on the EU level (Coen/Doyle 2000; Eberlein/Grande 2000). This framework typically consists of two elements. First, a sector-specific directive (that opens markets to competition) sets out minimum regulatory requirements and rules for Member States to implement. The implementation process is monitored by the sectorally responsible Directorate of the European Commission, and other Directorates may pursue other polices affecting the same sector as well (e.g. environmental policy in the regulation of energy markets). For example, the (1996) Electricity Directive, that prescribes a progressive and partial opening of electricity markets, requires Member States to ensure management unbundling of their transmission system (grid) operators and accounting separation of transmission from supply, so as to prevent discrimination against new competitors. But the Member States are free to choose between different forms of unbundling (creation or a separate legal entity or independence in management and accounting terms only). These requirements are often rather moderate or even vague because they result from a difficult political compromise (among the Member States) necessary to secure the passage of the EU directive in the first place. Secondly, national implementation is monitored by general EU competition law and policy. It is true that the Directorate-General Competition in the European Commission is endowed with exceptionally large powers to police anticompetitive practices (e.g. abuse of a dominant position, cartel agreements). These powers include merger control and also cover the behaviour of the Member

140

Working Paper VII

State governments or government-controlled businesses (e.g. state aids). Also, EC competition law offers a valuable avenue of redress for parties affected by national infractions of the European competitive order. However, general competition law is, as a rule, not well suited for sector-specific regulatory fine-tuning. Even if, in some instances, merger approvals were tied to specific domestic regulatory requirements (e.g. the German E.ON utility merger), competition law is a negative control instrument but not a positive tool to achieve harmonisation of national practices. Also, the Competition Directorate is not immune to political member state pressures (and subject to intra-Commission bargaining), which might also hamper a consistent application of EU rules, and thus harmonisation (e.g. McGowan/Cini 1999). The upshot of this is: the EU regulatory framework leaves to the Member States a large margin of discretion and choice in the regulation of markets. This concerns both the way they design regulatory institutions and the regulatory policy choices they make in the face of competing concerns (the following draws on Eberlein 2000a). The national regulatory institutions put into place after market opening exhibit important differences. In electricity and telecom, for example, national systems differ as to how they distribute regulatory powers between ministries, independent regulatory agencies, competition authorities or legislators. In the case of Germany, for instance, non-discriminatory access to the monopoly grid is governed by a ‘self-regulatory’ industry agreement, under the general competition oversight by the Bundeskartellamt (Eberlein 2000b). We also find important national variations as far as the scope and the goals of regulation are concerned. The EU Electricity Directive, for example, enables the Member States individually to define certain public service obligations ‘in the general economic interest’, complementing the overall objective of market-driven efficiency. Thus, the Member States have developed different sets of specific provisions to achieve public service goals such as universal service. In the field of environmental protection, different kinds of constraints are imposed on electricity suppliers to promote renewables in electricity generation. Variations in national (and sectoral) regulatory choices and regimes can be attributed to the weight of different national institutional settings. Sectoral regulatory design and goal selection are mediated by macro-level institutional variables. As neo-institutional writings have shown (e.g. Hall 1986; Immergut 1992), these factors include in particular: structures of political organisation, the nature of state-interest group relationships, and overarching sets of policy ideas and frames. If sectoral arrangements are embedded in national macro-level institutions, they are particularly resilient to change and likely to persist, even under functional pressures for convergence (Eberlein 2000b, 2001). It follows that different national regulatory regimes that have grown under the EU framework roof should not be expected to easily converge on a common European model of regulation.

The Network Concept, Competition Authority Networks

141

In sum, while there is some degree of delegation of the regulatory functions to the EU level or ‘framework’, this does not correspond to the expectations of the functionalist model. Rather, regulation remains largely decentralised and anchored in distinct and divergent national regulatory regimes, as predicted by the realist account.

V. The Regulatory Dilemma: Negative Externalities Without Centralisation The EU system of framework regulation grants broad discretion to the Member States to maintain different regulatory institutions, dealing directly with firms and consumers. This discretion is crucial in securing the consent of the Member States to initiate and pass EU rules in the first place. However, as part and parcel of this set-up, EU policymakers face what might be termed as a ‘decentralisation problem’. The diversity of national rules and regulatory bodies results, as set out above, in patterns of divergent implementation of the EU framework. National patterns of regulation differ, notably, with respect to the power the give to different market players: national regulatory choice and design will, for instance, affect the relative position of incumbent (domestic) operators versus new (foreign) market entrants. Thus, divergent national implementation may run counter to the overall goal of EU framework, i.e. ensure the equal treatment of market competitors across the Single Market. In this case, competition on an equal footing cannot be effective without a uniform set of rules and standards applied to all market participants (level playing field). National regulatory regimes do not only differ in design and policy choice and tend to produce different effects in their national markets. If they are joint in a single, larger market — and this is where functional interdependence comes in — they generate significant negative externalities for each other, that is regulatory regime difference harms another regime and no compensation is paid, by the beneficiary, to the adversely affected party. This includes cases in which regimes differences are opportunistically exploited by the beneficiary. Take the example of an incumbent operator in the electricity business. A lenient domestic regime helps to protect the home market from new competitors. Almost unassailable at home, the operator can more easily expand into foreign markets that are more rigorously regulated and are thus more open to new entrants. The current expansion of French electricity giant EDF from a rather secure home base into other European markets illustrates this point quite clearly. A good example of a high degree of functional interdependence is cross-border trade of electricity. Different rules for the use of networks and different rates of network charges affect the relative competitiveness of transit paths across Europe and may even obstruct certain trade flows altogether. This is, for example the case, when in some countries, (high-voltage transmission) network costs are not

142

Working Paper VII

socialised to all consumers but charged to exporting generators only. These generators will be put at a clear competitive disadvantage compared to their competitors in other countries. Between the Member States, the regulatory dilemma is as follows: trading countries depend on the use of transmission networks and interconnection capacity of transit countries. Trading countries have an incentive to free ride on the network capacity of transit countries, while the latter have an incentive to overprice network costs, so as to limit market penetration by foreign competitors. In short, the idea is that regulatory opportunism will prevail unless parties are constrained by higher level, common rules on how to charge for network use. As a rule, the winners of regime difference and interaction have no incentive to give up unilateralism and seek coordination or even less to agree to centralisation on the EU-level. As seen above, crucial regime differences may also be institutionally ‘locked-in’ by national macro-level constraints. This might also constrain the reaction of the losers. If unconstrained, they have an incentive to adjust regulatory standards in a ‘race to the bottom’. However, regulatory competition between national regimes, for instance by granting competitive advantages to national operators and suppliers through laxity in domestic network access regulation (‘national champion approach’), defeats the very idea of fostering competition and undermines the establishment of a common and competitive European market. The aggregate welfare of a true single market (and of unfettered competition in the home market) will not be realised. Unlike in other international contexts, there is no hegemonic player (Member State) able to unilaterally impose its regulatory model upon other national jurisdictions. And market pressures working towards a ‘trading up’ of regulatory standards (rigorous, pro-competitive regulation) only operate if the aim is to attract foreign business to the home market. If the aim is to fend off market penetration by foreign competitors, laxity towards domestic operators will be rewarded. This leaves us with the following regulatory dilemma: on the one hand, national regime divergence and interaction produce considerable negative externalities that require EU-level policy coordination. On the other hand, the Member States, due to conflicting interests caused by regime interaction, are collectively not willing to grant sufficient powers to the EU to erect and enforce uniform rules and standards. The broad EU regulatory framework that gives certain powers to the Commission (e.g. EC competition law) does not sufficiently address the need for coordination. Speaking in more abstract terms, we can rephrase the governance dilemma of policy coordination as follows: hierarchy, as one mode of governance, is not available to ensure EU policy coordination, while the alternative mode of governance, i.e. ‘market coordination’ (national regime competition), produces undesirable results. But the story does not end here. I argue that we can identify patterns of informal coordination without centralisation, patterns that have emerged, beyond the legal status quo, to address the regulatory dilemma. These patterns form the basis

The Network Concept, Competition Authority Networks

143

of a third, alternative account of market regulation in the multi-layered EU-polity, grounded in a third mode of governance.

VI. Escape Route from the Governance Dilemma? Transnational Regulatory Networks as Alternative Scenario Centralised EU coordination of national regulatory regimes is not a political option to satisfy the need for regulatory policy coordination. This is because the EU level, and notably the Commission, lacks both the formal powers and the legitimacy to impose a uniform implementation of regulatory rules and standards. In this context, transnational regulatory networks have grown to become an important avenue to achieve informal or soft coordination of national regulatory practices. What do these transnational regulatory networks look like and how do they work? A more general remark is in order here: when I speak of networks in this context, I do not subscribe to the sociometric network concept that aims for social structure analysis, and that is often based on quantitative methods (e.g. Laumann and Knoke 1987, Knoke 1990). Furthermore, within the broader analytical perspective of networks as patterns of resource interdependence, exchange and mobilisation, I do not see networks simply as patterns of interest intermediation between the state, or public actors, and private actors. Rather, networks are viewed as a specific form of governance (as opposed to market and hierarchies) in functionally differentiated settings where resources are widely dispersed.1 Finally, while networks are usually about the interaction of public and private actors, networks may also primarily link up different public actors drawn from different jurisdictions. This is obviously relevant in the case of EU governance, and brings in the ‘transnational’ dimension, to be discussed further below. Regulatory networks are essentially composed of representatives of national regulatory bodies, but with the input of ‘stakeholders’, i.e. affected industry and consumer groups. On an informal basis, they develop ‘best-practice’ rules and procedures for regulation in their sector. Best practice is grounded in dominant professional standards. In this respect, regulatory networks are close to professional networks (Burley/Mattli 1993) and epistemic communities (Haas 1992), in which technocratic expertise and cognitive resources play a central role (for EU examples see Radaelli 1997, Verdun 1999). 1 ‘In

this view, policy networks are best understood as “webs of relatively stable and ongoing relationships which mobilise and pool dispersed resources so that collective (or parallel) action can be orchestrated toward the solution of a common policy” ’ (Kenis and Schneider 1991, p. 36) A policy network includes all actors involved in the formulation and implementation of a policy in a policy sector. They are characterised by predominantly informal interactions between public and private actors with distinctive, but interdependent interests, who strive to solve problems of collective action on a central, non-hierarchical level (Börzel 1997: 4).

144

Working Paper VII

These networks perform effective coordination if they manage to set the agenda for formal endorsement of best-practice rules by the competent authorities, thus paving the way for some de facto harmonisation. I define effectiveness as the capacity to produce (including to make other actors produce) collectively binding decisions on the supranational level. For the European Commission, regulatory networks are an important tool to promote the emergence of common regulatory approaches. Thus, the Commission is the main sponsor of transnational networks. Yet, it does not simply control network operation. How do these networks relate to the EU committee system? Obviously, they do not have the legal status of Council working groups, but they may work en amont of Council decisions (paving the way). Nor do they belong to the comitology committees. Following the classification by Rhinard (2002: 192–3), they come closest to the third category of what he calls ‘Commission advisory groups’. Christiansen and Kirchner (2000: 7) interestingly use the term ‘coordination committees’. While these networks also operate in EU agenda-setting and policy formation, they are primarily designed to coordinate national regulatory policies (implementation) in the context of EU framework rules. A good example is the so-called European Electricity Regulation Forum. The Forum was established in 1998 to address the regulatory needs arising from the national implementation of the Electricity Directive. Its objective was to provide a platform for ‘informal discussion and the open exchange of experience’, complementing the work of bilateral follow-up groups between the Commission and national authorities. The main challenge identified at the outset was to ensure ‘that emerging differences in domestic approaches do not create barriers to the establishment of an EU Internal Market in electricity’. Organised by the Commission’s Directorate General responsible for energy markets (now DG for Energy and Transport), the forum convenes twice a year, bringing together national regulatory and competition authorities, transmission system operators, as well as industry and consumer representatives, and, occasionally, outside, commercial or academic experts. Additional meetings of smaller, specialised Working Groups are held in between Forum sessions. Progressively, two key issues for an integrated EU electricity market came to dominate the agenda: one, to develop a common system of transmission tarification (for the use of networks across Europe) and to establish common rules for the allocation and management of scarce interconnection capacity between national grid systems. In a series of meeting, partly with the input of outside expertise, the ‘Florence process’,2 as it began to be called in Euro-jargon, managed to develop a common understanding of regulatory needs, concepts and tools. For example, a consensus was reached on a set of principles that should guide the calculation of transmission tarification. These often quite technical policy solutions proposed by the Forum directly feed into and set the agenda of the formally

2 The

Forum meetings take place at the European University Institute in Florence, hence the label.

The Network Concept, Competition Authority Networks

145

competent body on the EU level, i.e. the Energy Council of Member States. The same kind of Forum was set up for natural gas regulation, the so-called ‘Madrid process’. An approach similar to the Forum method is used in the telecommunication sector. The Competition Directorate and the Directorate with sectoral responsibility for telecommunications (Information Society) established groups such as the Open Network Provision (ONP) Committee, composed of national regulators and ministries, Member State representatives in Brussels, and the Commission. The ONP Committee meets every two months, discussing issues such as licensing and implementation, and has progressively acquired an important regulatory role. The creation of two new committees, designed to coordinate and regulate national regulators, has been proposed by the Commission. Since the core of these EU regulatory networks consists of government officials, the networks have a strong ‘trans-governmental’ flavour, as opposed to intergovernmental and transnational,3 following the useful distinction grounded in IR theory and recently revived by Pollack/ Shaffer (2001). In trans-governmental governance, ‘lower-level government officials interact directly’ (ibid: 20). And they do so pursuing their own (sectoral and professional-driven) agenda, independent of national decisions (Risse-Kappen 1995: 9). Trans-governmental networks are potentially global in reach. Slaughter (1997: 184) observes that ‘government officials are networking with their counterparts abroad, creating a dense web of relations that constitutes a new, trans-governmental order’. And she goes on to claim that ‘trans-governmentalism is rapidly becoming the most widespread and effective mode of international governance’ (185), constituting what she calls ‘The Real New World Order’. Whether her broad claims are valid or not, Slaughter shows that the regulatory (and the judicial) realm is a particularly propitious environment for the development of trans-governmental networks that achieve coordination (‘order’) not through formal agreements based on government authority but rather by way of informal instruments such as memoranda of understanding concerning regulatory practice. We clearly see the trans-governmental component at work in our EU cases. In European electricity and gas regulation, the work of the Regulation Fora (the socalled Florence and Madrid processes) described above has progressively given an important role to the ‘Council of European Energy Regulators (CEER)’. This informal group of national regulators was established in March 2000 through a memorandum of understanding (Vasconcelos 2001). It includes regulators from most European countries (some not EU members), and maintains working contacts with regulatory counterparts in North America and EU candidate countries. It is organised in several working groups that develop best-practice rules 3 EU regulatory networks are much less of the ‘transnational’ kind, in which ‘nongovernmental actors participate directly in the process of global governance’, and not intergovernmental, in the sense that states are represented by their chiefs of government when bargaining and cooperating within international regimes.

146

Working Paper VII

and standards. It views itself as a ‘focal point for contacts between regulators and EU institutions’, and its creation has been actively promoted by the Commission. The CEER now plays the crucial role of agenda setting: it drafts guidelines and policy proposals that are submitted for discussion to the larger Forum Group. This is not to say that (private) stakeholders (diverse industry and consumer interests) have no say or importance. As a matter of fact, the Florence process in electricity not only brought together the established industry actors but also worked as an important platform for new sectoral interests to voice and press their concerns. The European Federation of Energy Traders (EFET), for example, played an active role in pressing for a user-friendly system of cross-border transmission tarification. Over time, the Forum opened itself to more and more participants (e.g. distribution companies), beyond the core of the transmission system operators and the traditional industry (incumbents) representatives. Therefore, while acknowledging power asymmetries, it would be wrong to speak of a ‘corporatist cartel’, or of ‘regulatory capture’ by industry interests. How do transnational regulatory networks tick? Which are the underlying mechanisms that allow them to perform decentralised coordination? I will look first at the first component: decentralised. Regulatory networks respect the discretion and autonomy of national regulatory regimes. This is not only essential in securing the consent of the Member States to EU-level activities in the first place. More importantly, by respecting, in the spirit of subsidiarity, a certain degree of Member State autonomy, this arrangement allows to reap the benefits associated with diversity and decentralisation. ‘Street-level’ national officials have specific ‘local knowledge’ necessary to properly address regulatory problems on the ground. Decentralisation encourages flexible rule adjustment tailored to specific local conditions that may vary substantially within a heterogeneous polity such as the EU. Conversely, centralisation implies a certain loss of information and flexibility. Centralised decision-making suffers from either a lack of information or from information overload in a ‘boundedly rational’ decision world. In this sense, one may think of decentralisation as a source of positive (instead of negative) externalities: national regime difference provides a greater variety of alternative regulatory models to compare with one another and of experiences to adopt or adapt from. The dissemination and comparison of diverse regulatory practices make ‘tacit knowledge’ available to other policy-makers beyond national borders. More broadly, diversity generates more opportunities for mutual learning, to have a richer stock of experiences to draw on when devising new policies. But how can these virtues of decentralisation be married to the need for policy coordination? How do national officials come to accept or converge on regulatory practices that may diverge from their national regime pattern? I suggest to distinguish between two elements. The first element builds on the power of cognitive resources and the opportunities for policy learning and transfer. The crucial resource of informal coordination

The Network Concept, Competition Authority Networks

147

through networks is not formal authority but information. In policy domains in which regulatory decision-making depends on science-based, technically complex knowledge, control over credible information, supported by professional standards, becomes an important tool of ‘soft steering’. If a consensus on expert level can be achieved (‘best-practice’), and be successfully injected, by national experts, into the different domestic arenas, then this will help to produce some de facto level of harmonisation of policy practice. From the Commission’s policymaking perspective, ‘the dissemination of best practices and models is an efficient strategy for building technocratic legitimacy’ (Radaelli 2000: 38). The broader argument is that density of interaction will facilitate ‘policy transfer’ and ‘lesson-drawing’ across different institutional settings (for an overview see Dolowitz/ Marsh 2000). Frequent interactions in a dense ‘organisational field’ such as the EU can catalyse processes of ‘institutional isomorphism’ (DiMaggio/ Powell 1991) that can account for increasing similarities of regulatory practices (Radaelli 2000). In our case of expertise-based regulatory policy-making, ‘normative isomorphism’ is the most important source of the tendency to become alike. More specifically, professionalisation creates a strong, shared frame of reference that facilitates convergence and homogenisation. National officials are driven by a ‘reputation game’ with their national counterparts. They will seek to comply with ‘best-practice’, regulatory standards to maintain their good standing in the professional community (Majone 2000: 6) This is precisely the real strength of EU agencies. As seen above, they only have very weak formal powers but are they are centres of expertise. They make a virtue of necessity and, instead of working by ‘command-and-control’, they operate through ‘regulation by information’ (Majone 1997). They are not engaged in formal decision-making, rather they are information ‘brokers’ who promote the creation of networks between the different national regulatory agencies (Dehousse 1997: 257). One strategy is to organise the exchange of staff. The idea is that by integrating national representatives or experts into European-level networks, they will be socialised to have a ‘cosmopolitan’ instead of a ‘local’ outlook on regulatory issues. Regular interaction will promote a European, instead of national perspective. But there is more to coordination than isomorphic convergence by professionalised networks. A second element, crucial to the production of coordination, is to tie participants (and their political principals) to a process of monitoring of national performances. Networks do not only disseminate the richness of national experiences and practices. They also subject them to a joint process of peer evaluation using the business technique of benchmarking. National performances are exposed to scrutiny. To be sure, regulatory networks do not have the formal powers to punish or reward national performances, but informal strategies such as naming and shaming can act as soft constraints on opportunistic behaviour. Take the example of transmission tarification in cross-border trade of electricity, and the consensus, reached by the European Electricity Regulation Forum, on best-practice principles and methods of cost calculation for network use.

148

Working Paper VII

In the course of the process, Germany as a transit country was drawn into a postage-stamp and marginal cost-pricing tariff model that excluded stranded costs of network investment. This in line with professional best-practice, but diverges from national distributional interests of a transit country in the centre of Europe. Why not defect? Well, any interest-based attempt to challenge the tarification principle, or to divert from it, can now be denounced as unprofessional and as uprooting an earlier consensus. It is politically very difficult to roll back the process. Put differently, the Forum was successful in increasing the costs of exit. The successful framing of the policy issue as ‘how do we design a best-practice cost-tariff model for an integrated electricity market’ greatly reduces the legitimate room for regulatory opportunism. The Forum managed to de-emphasise the distributional implications and to magnify the common concern: ‘we need rules for an integrated Single Market in electricity’. Thus, an argument for network effectiveness cannot only be made from a constructivist perspective that stresses the role of professional norms and socialisation. A rational choice approach to cooperation that focuses on changes in the perceptions of costs and benefits, can also provide clues to network effectiveness. But how effective is network governance?

VII.

Scope, Conditions and Limits of Network Effectiveness

Network governance is not a panacea for the resolution of the governance dilemma presented at the outset. Under which conditions can we expect transnational network governance to perform effective regulatory coordination, which are its limits? And what about the price attached to effectiveness? It is useful to start by distinguishing between internal and external conditions. Internally, the main condition for network effectiveness is a certain level of mutual trust between network participants (national officials), so that information will be disclosed and disseminated. In regulatory, trans-governmental networks this internal cohesion is greatly encouraged by the fact that members share a common professional background and often have strong personal contacts that can develop in small groups. Externally, regulatory networks can only perform effective coordination if their participants have some degree of autonomy vis-à-vis their domestic governments, as the definition of trans-governmental relations rightly points out (RisseKappen 1995: 9). Are national officials endowed with sufficient room for manoeuvre to engage in mutual learning processes and to possibly adjust national avenues to best-practice standards? This raises the central question of how regulatory networks deal with conflicting national interests and regulatory opportunism, as stressed by the realist account of EU market regulation. Will ‘learning through monitoring’ and soft constraints such as ‘naming and shaming’ be sufficient to discipline national discretion?

The Network Concept, Competition Authority Networks

149

As a rule, national governments will only agree to delegate de facto regulatory powers to informal networks if the perceived benefits of cooperation are greater than the costs. If conflicts over distribution between Member States begin to outweigh the common concern for coordination, then informal coordination will be ineffective or break down altogether. Under conditions of uncertainty, the potential benefits from this kind of cooperation are high. Governments who face a new, technical policy problem on which there is little national experience available, will be willing to delegate regulatory powers in exchange for policy solutions. However, once ‘lessons are learned’ and distributional implications and concerns move to take centre stage, the risk of deadlock or breakdown is high. Therefore, networks will have to make sure to frame policy issues as common and technical concerns, and downplay the distributive and ‘political’ elements or effects. Also, they may seek to confront governments with a technically ‘sound’ fait accompli, so that after-the-fact government intervention can be denounced as improper ‘politicisation’ of technical issues. But, sure, this might not be enough. As the regulatory literature tells us, political principals are more likely to intervene into the discretion of regulatory agents under conditions of low technical complexity and high issue salience, as opposed to high complexity and low salience (e.g. Eisner et al 2000: 29). Political salience is crucial: if a regulatory issue receives broader politico-electoral attention (salience, e.g. food scandal), then informal networks are likely to be disempowered. Our example of EU electricity regulation helps to illustrate this interplay of mechanisms. The Forum network was successful in the production of technical policy concepts necessary to respond to a new challenge: how to organise transmission tarification for cross-border trade in an integrated European market. Once, however, the distributional implications for transit versus trading countries began to surface and received more attention, the engine of the Forum process started to splutter. Attempts were made to denounce the blockade of transit countries as non-technical politicisation and as undermining an earlier consensus on a ‘best-practice’ solution. But nevertheless, the effectiveness of the network process was seriously called into question. Does this mean that regulatory networks break down exactly when we need them most? I argue that this is not so, because this view of delegation misses what is most important about networks, i.e. the process and dynamic character of cooperation. The point is that governments cannot simply withdraw delegation at any given time. For one, in the regulatory realm, national agents enjoy some institutionalised protection from direct government intervention. Most important, however, there are distinct costs associated with exit and defection from networks: loss of information and of future cooperative gains, but also remember that defection in one case can adversely affect cooperation in other sectors and for the future. The point is that the progressive increase and shadow of long-term gains can discipline short-term opportunism. Networks tie participants to a process that

150

Working Paper VII

modifies the cost-benefit-ratio over time. The costs of exit tend to increase, and these costs are especially high if networks have managed to deliver an accomplished fact that is difficult to unwind. Take again our example of EU electricity regulation and the case of transmission tarification. Germany did not exit the Regulation Forum even though it was unhappy with the distributional implication of the tarification model. To be sure, it complained and activated alternatives decision-making routes, but it could not roll back the Forum process. This brings us to a more general point on network effectiveness. To be sure, it would be foolish to believe that network governance can be just as effective as hierarchy in sanctioning opportunistic behaviour and that network governance can address all problems posed by cooperation equally well. No doubt, it does a better job at producing a common good, on the basis of information sharing and learning, than at settling distributional conflicts. However, networks can at least mitigate and de-emphasise conflicts by framing issues in terms that reduce legitimate room for opportunism and that are more compatible with a common solution. Also, if conflicts ride high, they provide a permanent forum for exchange, so that communication does not break down (keep the process going). It would be unfair to evaluate network effectiveness with reference to the nirvana of ideal-type hierarchies. If we take as reference point the status quo of decentralised non-coordination (or undesirable market coordination) and the failure of centralisation, networks offer considerable coordination benefits. In the same vein, we should, more generally, not think of network governance as simply replacing other modes of governance. No doubt, the regulatory game is played simultaneously in different arenas governed by different mechanism. But networks a) offer comparative advantages, and b) they generate added value in combination with other modes of governance. Take the example of networks operating in the ‘shadow of hierarchy’. This does not eliminate the benefits associated with decentralisation, but it can increase effectiveness. Thus, we should view network governance as enriching the toolbox of governance mechanisms. Finally, while the focus of this paper has been on effectiveness, and not on democratic legitimacy of informal governance, I will at lest recall the high price attached to network effectiveness. In transnational regulatory networks, nonelected officials (and private actors) exercise substantial, de facto regulatory powers. Trans-governmental networks are effective insofar and because they are informal, secretive and manage to isolate themselves from political pressures. While this helps to increase ‘output legitimacy’, it poses a serious problem for ‘input legitimacy’ and accountability. Sure, there are non-electoral forms of accountability such as transparency and reason-giving requirements. But transgovernmental networks are not only non-elected. They are also typically elitist and non-transparent in their deliberations. Other forms of informal, ‘new governance’ that aim for a broader participation of stakeholders are viewed to have much more potential for input legitimacy.

The Network Concept, Competition Authority Networks

151

They draw on the innovative features of new governance (flexible, heterarchical, procedural) much more in tune with modern ‘networked polities’ such as the EU (Ansell 2000). What is needed here is to tie the promise (and limits) of ‘new governance’ to a broader theoretical approach that manages to link the virtues of decentralisation and participation with effective coordination. To explore if the theory of ‘democratic experimentalism’ (Dorf/ Sabel 1998; Sabel 2000) that tries to deliver just that, offers a plausible account of decentralised coordination, would go well beyond the scope of this paper (but see Eberlein/ Kerwer 2001).

VIII.

Conclusions

The example of utility regulation shows that EU regulation is faced with a governance dilemma. On the one hand, national regime diversity and interaction in a Single Market produce significant negative externalities that require EU-level policy coordination. On the other hand, the Member States are not willing to delegate sufficient regulatory powers to the EU to erect and enforce uniform rules and standards. I argue that transnational regulatory networks can offer an effective escape route to EU policy coordination without centralisation. Network governance respects Member State diversity and reaps the benefits associated with decentralisation (information, flexibility), while providing mechanisms of ‘soft coordination’. They are based on a) professionalisation (expertise-based convergence) and b) monitoring techniques (benchmarking) set in a dynamic process that reduces the room for national opportunism. However, the effectiveness of transnational regulatory network hinges upon certain conditions: they perform best under conditions of uncertainty, high technical complexity, and low political issue salience. More generally, network governance is not a panacea, it is simply endowed with comparative advantages. It enriches the toolbox of governance mechanisms by offering added value, not least in combinations with other governance modes, hierarchy and markets. Finally, if regulatory networks are effective, they create new problems of democratic legitimacy: non-elected officials (and private actors) exercise substantial regulatory powers, with no electoral and little non-electoral accountability.

Technical Problem Areas

285

Panel Three: Technical Problem Areas

JOHN FINGLETON — When talking about a large European project, it seems appropriate to quote Bismarck. Asked how to solve the Irish problem, Bismarck said that the solution would be to swap the populations of Ireland and Holland: the Dutch would make Ireland into an industrious and prosperous nation that would feed the rest of the world, and the Irish would be careless and fail to maintain the dykes, and would drown. I think we have all moved on a little since then, but a question that remains valid is, how is the network going to function in light of the cultural diversity in Europe? I personally think that Europeans work very well together nowadays. Returning to the topic of my presentation: since I am not a lawyer, but an economist, I have a lot of questions about how this network is going to operate. However, that is not to say that I am not enthusiastic about this project. My first question is about the legal status of the decisions concerning case allocation. As I said yesterday, there are two issues that I see coming up here: who does what? And how do they do it? If the outcome of a case can vary depending on who deals with it, then the rights of the undertakings concerned may also be affected. We need to clarify this. It is not apparent to me whether the parties will be able to challenge case allocation decisions. Yet whether they can or cannot do so will affect how the network itself operates. Unless one authority knows something more than the complainant does, a case ultimately has to be allocated on the basis of what the complainant argues in his petition. I think that we need greater clarity, particularly with respect to the complainants’ rights, and we should even see whether a harmonisation of the complainants’ rights across the national jurisdictions within Europe is necessary. Ultimately, this is a question about procedure, and about the harmonisation or non-harmonisation of procedures across the Member States. With regard to the rights of the defendants, I think that the issue of sanctions is probably the most important. In other words, if the main issue for complainants is one of procedure, for defendants the main issue is one of sanctions. Returning to case allocation, in practice a case may have be allocated either between the Commission and the competition authority of a single Member State, or between the competition authorities of several Member States, or between a group of national competition authorities and the Commission. With respect to the first type of case allocation, between the Commission and the competition authority of one Member State, the Commission has had the power to take up a case from the national competition authorities since 1962 (as Alexander Schaub mentioned yesterday). This is an issue that may be of greater concern to larger Member States than to smaller ones: if an anticompetitive agreement or behaviour has an effect in Germany, it might be of more interest to the Community than if it had a similar effect in Ireland. Clearly, its effect on intra-community 䉴

286

Panel Discussions

trade is going to be the determining issue. The cross-border effects doctrine is not relevant here, because the effect is felt in one Member State. The effects doctrine is relevant to the allocation of cases among the EC Member States, not between a Member State and the Commission. An issue about which I have a question mark is whether the Irish competition authority, for example, could actually ask the Commission to take up a case simply because it cannot itself impose sanctions. If you take the Irish Sugar case as an example:1 after the implementation of the reform, the Irish competition authority would be able to look at the case but not to impose a fine. (The fine imposed by the Commission in Irish Sugar under Article 82 EC — then Article 86 EEC — was 8,8 million Euro). So my question is: could we investigate the case and then pass the file on to the Commission, asking them to make a decision? Is that an option for national competition authorities wanting to deal with the lack of effective sanctions in their national legal systems? The Irish competition authority is often asking the Commission to take up cases, and not because it does not find them interesting, but because of lack of effective remedies available in its own jurisdiction. It takes a long time to get a case through court in Ireland, and we do not have the ability to impose fines or other remedies comparable to those available to the Commission. Similar situations might occur under Article 86 EC: we might run into a state restriction that the Commission would be in a much stronger position to challenge. As to the allocation between the competition authorities of the Member States: here the effects doctrine applies. I think this will be the most exciting area of enforcement in the future, when a group of national competition authorities will have the opportunity to work together on a case. I think that there are already subgroups of Member States whose markets are already much more closely integrated than in the rest of the EU; for example, Ireland and the UK, or the Czech and Slovak Republics. In some of those groupings the procedures and sanctions may also be very similar. Consequently, when we talk about groupings or pairings of competition authorities, it is quite likely that these will not happen in a random way, but rather within some fairly coherent group of competition authorities. And I think that this is probably how we are going to start cooperating within the network. For example, the Irish and UK authorities will start working more closely, or the Belgian, Dutch and Luxembourg authorities will do the same. In my written contribution I also made a note on the four Nordic States, questioning whether in their case we should not allow for greater flexibility. I also have some questions about how these joint teams will operate; these are technical questions that we need to clarify. I think that the Commission and the Member States should agree on a new notice on market definition. Market definition is going to be crucial in the new enforcement system, and if the national competition authorities do not share 1 Commission Decision of 14 May 1997, IV/34.621, 35.059/F–3 Irish Sugar plc, OJ L 258, 22/09/1997, p.1 et seq.

Technical Problem Areas

287

the same fundamental principles about market definition, the issue of case allocation will become more complicated as well. We do not want to see situations in which one national competition authority says it does not take into account supply substitutability — that would be an invitation for undertakings for whom supply substitutability is essential to seek to have their cases heard there rather than somewhere else. I would very much hope that the first sign of our ability to deliver within the network would be the adoption of a joint notice on market definition. I do not see big problems with the existing Commission notice on market definition, but we might need to change a few things here and there in order to take our collective experience into account. As to the third type of case allocation, between the Commission and a subgroup of Member States, I think that decisions will depend on how well the subgroup works together. This is a type of decision-making that will become smoother over time, and it is not going to be a big issue. In my written contribution I raised the issue of conflicting decisions. I wonder whether that is going to be interpreted conservatively, meaning that the same mind could not possibly have reached the two conflicting decisions. Or will it be interpreted broadly, meaning that incompatible decisions cannot be imposed upon the same undertaking(s). I prefer the broader interpretation, as it leads to more experimentation and learning, and I think it allows the national competition authorities to reason independently from the Commission, perhaps even challenge the Commission’s reasoning, find innovative remedies, and so on. Another issue that I touched upon in my written contribution is that of resources. Resources will be required for doing a whole range of things in the new enforcement system. The first of these is the actual investigations. This touches upon a point that I made yesterday, about whether the whole purpose of the reform is to do what we are currently doing at a lower cost — which I do not think is the case — or whether it is to improve the quality and quantity of what we are doing. We need resources in order to improve the quantity of the decisions. The national competition authorities are going to have more work to do in the new enforcement system, and not all Member States seem to be aware of this. Resources will also be needed for the operation of the network itself. For example, who is going to read all the information circulated within the network? For example, when information is sent to the Commission, will it just be filed or will somebody actually read it? It would be good if the information were actually to be read. Another additional cost would result from translations, albeit into a single language. Another issue is that of staff exchanges: I would be very surprised if any competition authority had staff to spare for exchanges. To sum up, these are some of the technical issues that I mentioned in my written contribution, and which I hope will be discussed within the Council Working Group. I would conclude by saying that, ultimately, there is a self-equilibrating mechanism in the functioning of the network. I also believe that the system is going to converge towards efficiency, but that may require some harmonisation of sanctions, and possibly of procedures. As an economist, I would put emphasis

288

Panel Discussions

on the harmonisation of sanctions because they have an important overall economic effect on the functioning of the system. The more we can do up front in sorting out these problems, the faster the convergence will occur. As I said before, convergence will occur anyway, but the solving of these problems would accelerate the process. I believe that, even in the absence of the modernisation project, we would have assisted with the development of competition law enforcement at the Member State level. IAN FORRESTER — I must first confess that I had some difficulty in composing my intervention for this workshop, and I do not offer my thoughts with great confidence. Subject to that warning, I would observe, first of all, that the old rules functioned adequately and did their task in their time. They were not logical, they were not coherent, but they were made to work by talented officials and energetic lawyers. The new regime will certainly have imperfections, and I assume that at least twenty percent of what has been proposed in the draft regulation is inadequate or wrong, and will have to be changed. The same happened with the merger regulation adopted twelve years ago: some of it was good, and a little bit of it had to be altered. However, I am not troubled by the idea that you won’t get it right the first time. I do think that building flexibility into the new system is highly desirable. The biggest news is that the reform is under way. Some people around this table, such as Prof. Ehlermann, Karel van Miert, Alexander Schaub, and others, deserve the credit for a process that has led to a bloodless reformation without the Inquisition. This is to make an analogy with one of my old remarks about the laicisation of Community competition law, with DG IV as the Vatican pushing out onto the laity the enforcement of the EC competition rules. Yesterday we heard two very diverging points of view: on the one hand, Mario Siragusa said that the Commission is losing its control and we need order and predictability. On the other hand, we heard Frédéric Jenny and Ulf Böge say that the national competition authorities will be placed in a subordinate position by the Commission. I hope both these views are wrong. I presume that reality lies somewhere in the middle. I also presume that the Commission will use the power to take up cases from the national competition authorities, and will exercise its powers with respect to case allocation in an intelligent way. At the same time, I do not worry too much about legal certainty in the new enforcement system. Competition lawyers have done fine for forty years without legal certainty. I do worry about quality control and homogeneity, for reasons that I have set forth in my written contribution. We can already see instances of naked use of political power and problems of quality control in the current Member States, and those will be amplified after enlargement. The competition authorities in the Eastern European countries apply a formalistic, mechanical competition law that does not have much to do with what is going on inside the EU. They need the assistance of the competition authorities in the EU in resisting political pressure. But sometimes not even the competition authorities inside the EU succeed in resisting political pressure. 䉴

Technical Problem Areas

289

So, let us get the network functioning as quickly as possible, so that the competition authorities of the new Member States will have the support of its members. It is easy to conceive clever legal points about inconsistency and so on. The new rules are not perfectly clear, but I don’t think they could be, given the current state of affairs. It seems to me that, when confronted with difficult decisions, the Commission is telling the national competition authorities: ‘Do your best. Work on it. We are not going to intervene quite yet.’ I think this is an appropriate and understandable approach. Lawyers will take every chance they can get to make procedural trouble. So do not assume that the lawyers will give you an easy ride; they will not. But that is fine. Controversy will make decisions, and out of the decisions will come adequate results. I did a little bit of anecdotal research during the course of preparing my written contribution, and I spoke to officials of about ten national competition authorities. They all said they meet with each other as if they were members of a family, and so on, but when asked if they ever cooperated in allocating a case, not one said ‘yes’. I may be wrong, and I am sure there are instances where the national competition authorities have agreed on an efficient allocation of cases, but it is not enough to just meet. It is necessary to do things together. Family weddings are nice, but families don’t (always) run a business. Lawyers also meet and discuss things — for example, in the context of practising law, when a conflict arises between clients of a law firm — and the issue is then discussed by looking at the relative advantages and making a rational decision. This model can also work, I believe, for competition agencies. There is lots of ground for hope. Practitioners are appearing regularly before the national competition authorities, and there is a desire to make the reform succeed. In my personal experience, the national administrations also have great respect for the decisions taken by other national administrations. Mario Siragusa, Frédéric Jenny and Ulf Böge raised beautiful questions to which I do not know the answers. But I believe that, on the basis of common sense and Europe’s capacity to deal with the illogical, we will have a successful reform. 䉴 ANNE WILLEM KIST — I empathise with Ian Forrester’s views on the subject, so I would like to start with an anecdote: there was a legendary figure in my university, who had the reputation of being what we call in Holland an ‘eternal student’. He was to become a medical professional. It took him years and years to graduate, and in the end, he started his own practice. He chose to locate his practice next door to a very highly reputed general practitioner. On the door of his own practice he put a plate: ‘For difficult cases, go next door.’ This is exactly how I feel when dealing with this subject: for the difficult technical issues, go to others. Now allow me to make a few observations about the issues discussed in this session. First of all, I think that we should consider in more detail what the exact scope and effects of Article 12 in the draft regulation will be, and particularly its likely effects on national procedures. This is, of course, an issue that will develop through case-law, but I think that we should nevertheless give it some more consideration before the adoption of the new regulation.

290

Panel Discussions

Secondly, it is clear to me that we need to have as much clarity as possible when the new enforcement system is introduced. It would be impossible to arrive at the necessary degree of clarity if we want to implement the reform soon. Therefore we should find a balance between the objective of having as much clarity as possible as to the terms of cooperation within the network, on the one hand, and relying on emerging case-law and best practice procedures, on the other. Maybe practitioners will not be very happy with such a compromise but, as Ian Forrester said before, did we ever have it any better? Thirdly, what will be the position of undertakings in the event of an information exchange within the network? One possibility would be to make an administrative decision as a preparatory phase: if one does not agree with it, one can always challenge it in court. Another would be to adopt a solution similar to what we have in the Dutch legal system with respect to the application of both EC and national competition rules. The Dutch system allows the interested parties to go to a specialized competition court in summary proceedings. Furthermore, the Dutch competition authority is under a legal duty to announce to the parties involved that it is about to exchange a certain kind of information, this gives them the opportunity to go to court. I am not sure whether this is the best solution, but a solution must nevertheless be adopted. Finally, a very complicated and thorny issue arises from the fact that nowadays we see various European countries moving in the direction of criminalising certain aspects of antitrust law. Other countries, such as the Netherlands, are determined to maintain antitrust in the administrative sphere. This oncoming divergence may very well raise thorny and difficult issues with respect to the exchange and use of information within the network. For example, I am not sure that a Dutch court would accept evidence obtained under criminal law proceedings in another EC country, or where the other competition authority has powers that the Dutch authority does not have. Some may argue that any national legal obstacles to the use of evidence obtained through exchange within the network should be set aside in order to give full effect to Article 12 in the draft regulation. But this is an arguable position. 䉴 CLAUS-DIETER EHLERMANN — This is indeed a technical issue of great practical relevance: during my six years at the WTO Appellate Body, I came to appreciate that it is essential to have rules concerning the exchange of confidential information, otherwise the WTO dispute settlement system can be brought to a halt. However, there are no such rules at the WTO. 䉴 MARC VAN DER WOUDE — I would like to start with some general observations. First, under the present rules, as defined in the Spanish Banks case,2 the

2 Case

C–67/91 Dirección General de Defensa de la Competencia (DGDC) v. Asociación Espanola de Banca Privada (AEB) and others (‘Spanish Banks’) [1992] ECR I–4785.

Technical Problem Areas

291

information gathered in the context of one proceeding may not be used as evidence in another one. However, such information may be used as a reason to initiate new proceedings under either EC or national law. This has important consequences in terms of designating the authority competent to deal with a case: one should be careful not to send information to those authorities that may, in the future, be put in a conflict of interest situation. Furthermore, problems may arise if a competition authority uses the information received for purposes other than the enforcement of competition law (for example, for commercial negotiations). My second observation also concerns conflicts of interests. In antitrust enforcement we have, on the one side, the interests of the competition authorities, which want to obtain the maximum amount of relevant information for solving a case, and who do not want to be disturbed by anybody — and certainly not by lawyers — in the gathering of such information. On the other side we have the interests of the companies concerned: those from which the evidence originates, and whose interests are obviously different from those of the companies against which these documents are going to be used. Somewhere a balance has to be reached between these conflicting interests. Thirdly, I would like to say a few words about differences between the present and the new enforcement system in terms of exchange of information. In particular, the new enforcement system will be characterised by horizontal flows of information. This contradicts the current rules: in Spanish Banks, the ECJ specifically ruled out the exchange of information among competition authorities. The draft regulation not only envisages such an exchange of information as essential for the functioning of the new system, but also requires that any rule of national law that would prevent such an exchange be set aside. This sets aside the Spanish Banks case-law. However, the draft regulation speaks about a right to exchange information, and not about a duty to do so. The limits to the exchange of information are set out in Article 20 of the draft regulation. They re-confirm the existing case-law on the subject; see Dow Benelux.3 Curiously enough, the current limits will continue to apply with respect to the Commission services, but not in relations between the national competition authorities that are part of the network. Of course, for the latter there is another limit, insofar as the information obtained may be used only for the purposes of applying EC competition law. But what happens if the information is sent to another authority that intends to apply national competition law? May the latter authority use the information obtained in order to initiate proceedings under national competition law? Here Article 3 of the draft regulation also comes into the debate. What is its interface with Article 12 in the draft regulation? Finally, the draft regulation establishes that the information obtained through the network may be used only for applying financial sanctions.

3 Case

85/87 Dow Benelux v. Commission [1989] 3150.

292

Panel Discussions

And now to the more provocative part, which concerns technical problem areas as seen from the perspective of a practicing lawyer. The first problem that I identified is as follows: is Article 12 of the draft regulation a valid provision? In both Spanish Banks and Post Bank4 — the latter being a case that concerns the exchange of information with national courts outside the scope of Regulation 17/69 — the European courts do not make reference to Regulation 17/69, but to primary Community principles: the rights of defence and confidentiality obligations. One could easily say that Article 12 in the draft regulation is not valid, but that would go too far. There are means to reconcile the need to protect the rights of defence with the objective of promoting a free exchange of information. I would say that the interested companies have three rights: first, they have a right to the highest standard of defence. This includes, as Mr Kist has already suggested, the principle that a competition authority may not rely upon evidence that it would not have been authorised to gather in its system. Secondly, the company from which the documents originate should have the highest level of protection. This is not only an issue of confidentiality, but also one of access to public documents. In some countries, such as the Netherlands, the system of access to public documents is very generous. Does this mean if I were to send some documents to the Dutch competition authority under the Dutch law, a third party could get access to the documents, which are perhaps are not confidential in the Community’s legal conception, but which nevertheless I would not like to be made public? Here again, I would support establishing a right to the highest level of protection. (I feel similarly with regard to the issue of legal professional privilege, but I think that Mr Santiago Martínez Lage will speak about that later on). Also important is the right to having the most favourable criminal law applied, especially when we speak about leniency programs. For the moment there is no harmonisation of these programmes, and I do not think it will be possible to harmonise them in the future either. In summary, if someone sends information to one national competition authority, s/he should be entitled to the same conditions under which the information is granted in the country of destination. One could even push it as far as to say that, if the level of protection in the country of destination is higher than in the country of origin, the information provider should benefit from the higher protection available in the country of destination. Finally, I would support the introduction of a right to the complete transfer of the file. In complex cartel cases, particularly when the prosecution relies on circumstantial evidence, exculpatory documents can be very important. So, if a file is sent from one competition authority to another, it should include both incriminating and exculpatory documents. To conclude, if these rights — the right to the highest standard of protection, the right to the most favourable conditions, and the right to the complete transfer

4 Case

T–353/94 Post Bank v. Commission [1996] II–926.

Technical Problem Areas

293

of the file — are respected, Article 12 of the draft regulation would be compatible with the ECHR on the rights of defence. Then the question arises: how can we safeguard these rights? I think that Mr Kist has already given the answer: before documents are sent, the company from which the documents originate should be informed about the fact that this information will be sent and given the opportunity to oppose the transmission of the documents. This is nothing new; it is a procedure that already exists (we saw it applied in Post Bank). I do not see any reason why this should be changed, and I think it can be made compatible with a smooth exchange of information by imposing very strict deadlines on the companies involved. I think that in many cases the rights of defence will be protected in the country of destination, but the interests of the company from which the documents originate will not necessarily be protected in the country of destination. FRANCES BARR — The written contributions prepared for this workshop and the discussions we have had so far show that this is an area raising a great many issues. The ones that I would like to cover in my presentation include:



• the type of information covered by the provision in the draft regulation on the exchange of information within the network; • possible practical problems in the application of Article 21 of the draft regulation; • possible implications of the exchange of information for the rights of defence; • the impact of the exchange of information within the network on the effectiveness of leniency programmes; and • possible consequences of the existing differences between national procedures and sanctions in the new enforcement system. This is long list, and obviously I shall only set out some preliminary and personal thoughts that are intended to highlight some possible practical problems and consequences of the exchange of information within the network. These issues are covered in more detail in the paper prepared by my colleague Margaret Bloom for this workshop. As noted, these views are all personal. As to the type of information covered by the exchange of information within the network, Article 12 of the draft regulation refers to the Commission and the Member States having the power to provide one another with information, including confidential information, and exchange evidence on any matter of fact or law. Article 21 in the draft regulation refers to investigations, and in this context the reference is simply to ‘information’. The question is therefore: what type of information is actually covered by these provisions? And, what is actually protected? The draft regulation uses the concept of confidential information and makes some qualifications to this concept: first, Article 27 sets out protection for

294

Panel Discussions

information of the kind covered by the obligation of professional secrecy. In addition, the protection of business secrets in the process of access to the file is recognised in Article 26. The obligation to maintain professional secrets is established by the EC Treaty itself. The idea that particular types of information may require special treatment in administrative proceedings has also been developed in Community case-law. Similar concepts are also dealt with in national laws, and these may provide a different type of coverage and protection. In the UK, for example, the Competition Act establishes a prohibition system modelled on Articles 81 and 82 EC, and the regime defines confidential information as commercial information, the disclosure of which would, or might, significantly harm the legitimate business interests of the undertaking to which it relates. This is very different from the concept of professional secrecy in the EC Treaty, which refers to information about undertakings, their business relations, or their cost components. The differences between national laws in this regard could have an impact on the actual nature of the information that can be collected and exchanged within the network. In addition, there are also differences in national laws with respect to the powers of investigation of the competition authorities. These include, for example, differences as to the power to gather different types of information, the ability to ask those under investigation for information, the scope of the protection provided in areas such as legal professional privilege, the type of premises that can be inspected, and so on. In the UK, for example, the Office of Fair Trading can carry out on-site investigations on domestic premises provided that these are used in connection with the business of an undertaking, or documents relating to the business are actually kept on the domestic premises. Article 21 in the draft regulation requires that the competition authorities of the Member States assist each other as far as national procedures allow. This may mean, for example, that if an investigation is carried out in the UK on behalf of another Member State, our investigators would not take documents that had been prepared by an in-house lawyer. These would be protected by legal-professional privilege under UK law. I think that this may well have an impact on the actual effectiveness of the mutual assistance envisaged by Article 21 in the draft regulation. Information obtained under Article 21 can be exchanged and used in accordance with Article 12 of the draft regulation, which allows the receiving authority to use the information even in situations in which it could not have been obtained by its own on-site investigation. In this respect, the only limitation appears to be in terms of the sanctions that can be imposed. Taking the example of a document produced by an in-house lawyer: this would mean that this document could be used by the Office of Fair Trading if it was information received under Article 12 of the draft regulation following an investigation carried out by the competition authority of another Member State at the request of the Office of Fair Trading based on Article 21. The same document could not be used if the Office of Fair Trading’s investigators had themselves come across it during an investigation in the UK.

Technical Problem Areas

295

This is my interpretation of how Articles 12 and 21 interact. However, looking at the written contributions prepared for this session, and taking into account the discussion we have had so far, it seems to me that others have a contrary view. So I think that this is an issue we could examine in more detail. If the issue is not settled in the new regulation, I fear that lawyers will challenge our (the Office of Fair Trading’s) acts relying on these provisions. Furthermore, lack of clarity in this respect this may also have implications for the rights of defence. For example, will undertakings have a right to make representations about the use of such information during national legal proceedings? If so, when would those rights actually arise? Would this be when the investigation is carried out on behalf of the requesting Member State in accordance with Article 21, or when the information is being used by that requesting Member State to pursue the case? These are all issues that arise when the information obtained through the exchange is being used for the actual purpose for which it was acquired under Article 21. I have not considered additional concerns, that may arise when information exchanged under Article 12 is used for another investigation, albeit in relation to the application of competition law. However, the issue has been tackled in other written contributions for this session. I have also not considered the nature of the rights of defence, which I believe to be a very complicated area. I would like to turn instead to an area that is slightly more straightforward, namely whether the exchange of information within the network will have an impact on leniency programmes. Leniency programmes operate on the basis of confidentiality of the information that an undertaking has provided in return for a grant of immunity from sanction. It is generally accepted that this confidentiality can only be respected for a limited period during the investigation. In the UK we have a new leniency system that has been operating for a couple of years. It is extremely effective but, at every single stage of the proceedings, companies argue about the confidentiality of every single aspect of their application. I think that the ability of the competition authorities to exchange information within the network might mean that an undertaking would apply for leniency in every Member State that has a leniency program. This will happen because undertakings know that the information provided as part of one leniency application could be exchanged with other Member States in accordance with Article 12, even though no immunity would arise in any other jurisdiction in the absence of an application in that particular Member State. If a case has already been allocated by the network, the undertaking will know who will be handling its case, and therefore the choice of where to make any application for leniency will be fairly straightforward. But if a case has not been allocated, the undertaking will need to make a choice of where to apply for leniency, taking into account the considerations I have just outlined. For these reasons it may actually decide to apply first to the Commission. I think it will be very interesting to see what impact, if any, this choice — which is prompted by potential concerns about information exchange — may have on the operation and effectiveness of national leniency programmes.

296

Panel Discussions

Considering the issues that arise in relation to information exchanged within the network, one might conclude that significant differences between the laws of Member States cannot actually be sustained in the new EC competition law enforcement system. Most significant in this respect are differences between civil and criminal law procedures, and those concerning the scope of sanctions. I do not think that these differences are necessarily a problem right now, so I agree with Céline Gauer and I share Professor Ian Forrester’s optimism about the new system. But it may be likely that pressure to harmonise national procedures will increase as organisational reforms take effect. And this could occur by ‘soft harmonisation’, for example, by adopting practices and procedures developed as the experience of the network grows, and through continued discussions such as the one here, and other discussions within the ‘family’ of the national competition authorities. But it may be that only formal changes to national laws will produce a practical and workable system, or that there will be a need for further reform of EC competition law. 䉴 SANTIAGO MARTÍNEZ LAGE — My intervention will focus on four points related to the exchange of information. First, my colleague Helmut Brokelmann and I think that there is a need for a common notion of confidentiality and business secrecy. Under Article 26(2) of the draft regulation, access to the Commission file is limited by business secrecy principles and excludes confidential documents and internal documents. This provision refers only to procedures before the European Commission, so it does not apply to procedures before national competition authorities. We think that it would be necessary to harmonise this concept in national laws in order to avoid the loss of information’s confidential character as a result of the transfer of documents from one national competition authority to another. Furthermore, even if the notion of confidentiality is harmonised, a rule will still be needed to maintain the classified or confidential character of a document. We suggest that something similar to a mutual recognition principle should be applied in this case. That is, the assessment of the document by the authority of the country of origin should be respected by the authority of the country of destination. Probably one exception is needed from this rule: the country of destination’s authority should be allowed to make such a decision only if the declassification is based on a new fact, such as voluntary disclosure by the party that claimed confidentiality, or if the document has been made public by the other party. The second point concerns limits on the exchange of information within the network. Here we share the views already expressed by Marc van der Woude. Under Article 27 of the draft regulation, information obtained through an exchange within the network is to be used only for the purpose for which it has been acquired. However, Article 27 also applies ‘without prejudice to Article 12’, which stipulates that the information obtained can be used for the application of EC competition law. In the latest version of the regulation (presented as ‘Progress Report to the Council’) this provision is completed by ‘and in parallel, national

Technical Problem Areas

297

competition law’. In our opinion, the information obtained through an exchange should not be used for the application of national competition law alone and, even when used for the application of national competition law together with Community law or for the application of Community competition law only, it should be used with the limits established by the Dow Benelux and Spanish Banks case law.5 The Dow Benelux decision says very clearly that ‘the right of the defence of the undertaking concerned would be seriously endangered if the Commission could rely on evidence against undertakings, which was obtained during an investigation but was not related to the subject matter of the purpose thereof.’ In Spanish Banks, the Court established that ‘the right of the defence requires that the information thus obtained should not subsequently be used outside the legal context in which the request was made’. The third point has already been touched on by Willem Kist and Marc van der Woude; it concerns the introduction of a procedure whereby the undertakings concerned about an exchange of information should be heard before the competition authority of the country of origin makes a decision to transfer the information. In particular, the undertaking should be heard on two points: whether there are circumstances that justify the transfer of information, and the limits to this transfer (that is, what documents, or parts of documents, should be considered confidential or business secrets?). In our opinion, the competition authority of the country of origin should issue a reasoned decision on this matter, and there should be an opportunity to review it in court. The fourth point concerns legal privilege. It is a bit disappointing that the draft regulation does not make any reference to the existing case-law on legal privilege, in particular to the AM&S and Hilti doctrine.6 Nevertheless, it is obvious that this case-law should always apply to procedures before the European Commission, so that the Commission is prevented from obtaining — either through a request for information or by confiscation — any document that is the object of an exchange of correspondence between an undertaking and external lawyer in the context of an actual, or a potential, procedure before the European Commission. This is the case law, but we think that we must go even further if we want the legal privilege to be effectively protected; here I disagree with Frances Barr’s position. I think that here we have two different situations. In the first, national competition authorities are obliged to respect the legal privilege attached to documents generated in proceedings before the European Commission. In other words, if a company receives legal advice from external lawyers in an actual or potential proceeding before the European Commission, the national competition authorities cannot try to obtain this document, as this would be contrary to Article 10 EC. Even if legal privilege is not regulated under national law, the national competition authorities cannot obtain the documents 5 See supra notes no. 2 and 3. 6 Case 155/79 AM&S Europe v.

Commission [1982] ECR 1575; Order of the Court of First Instance, Case T–30/89 Hilti v. Commission [1990] ECR II–163.

298

Panel Discussions

generated — I insist — in procedures before the European Commission. The same holds true if the national competition authorities apply Articles 81 or 82 EC: when applying Community law, the Member States are bound by the general principles of Community law. In relation to the examples made by Frances Barr: it is difficult for me to imagine that the OFT is authorised under UK law to obtain from a foreign country information that it could not itself have obtained in its own jurisdiction, thereby taking advantage of the weaknesses of less developed foreign legislation that does not protect legal privilege. At the same time, I share Marc van der Woude’s suggestion that the highest standard of protection should be applied in both the country of origin and that of destination. Probably the best way to deal with these matters is through a directive. 䉴 MONICA WIDEGREN — I will turn now to a different subject — consultations within the network — which, along with the exchange of information, is one of two pillars necessary for a robust functioning of the network. In the legal exception system, the network will have to deal with more cases than at present (leaving aside, of course, applications for individual exemptions or negative clearance). We shall also have to deal with the newcomers in the network — the competition authorities of the new Member States — which have less experience in cooperating with the Commission or between themselves. These factors taken together will have an impact on the consultation process. The concept of consultation is not defined in either the existing Regulation 17/69 or the draft regulation. I suppose it could be defined as the initiative of a national competition authority to exchange information in order to obtain a reaction from another national competition authority and, in turn, take further action in a given case. We are, of course, talking about formal consultations among the members of the network. On the basis of the draft regulation, one can identify three types of consultation process: (1) consultation regarding the allocation of cases; (2) consultation prior to making decisions; and (3) assistance with the information gathering process. Following the implementation of the reform, the Commission will retain more or less the same powers with respect to consultation as it has in the current system. The Commission’s powers to provide information to the national competition authorities about new cases before it, and to consult within the Advisory Committee prior to adopting decisions, will essentially remain the same. With regard to Article 11(6) in the draft regulation, I was pleased to hear Alexander Schaub say that the Member States will be consulted before such a measure is adopted. The Member States were mentioned in plural, so I take it that Mr Schaub was referring to such consultations taking place within the network. I think that the consultation procedures should be explicitly addressed in the text of the regulation itself in order to underline the horizontal nature of cooperation within the network. The national competition authorities are required to submit information about new cases to the network. This provision was discussed in the Council Working

Technical Problem Areas

299

Group, and we have now specified that such information should be submitted after the first formal investigative measure is taken, whatever its nature. As I understand it, this means that the members of the network must ensure, for example, that a complaint has some substance before it is disclosed within the network. In order to facilitate consultation, we would also have to elaborate a standard format for this kind of information-sharing. As to consultations prior to the adoption of a decision, we agreed at the Council Working Group that every Member State needs to clarify for itself what a draft decision would mean and by whom it should be submitted to the network. For instance, would the national court or the national competition authority submit it? The answer to these questions depends on the particularities of each national decision-making system and the kind of information involved. The Commission says that a statement of objection would be fine, but of course it depends on whether all the Member States’ national legal systems allow for procedures similar to a statement of objection. In my view, we are on safer ground if referring to draft decisions. Finally, with respect to consultations following a request for assistance to obtain information: I prefer to see this procedure as a consultation process. When a member of the network requires assistance to gather information, that the members of the network must do their best to cooperate, or the network will never function. I am now referring to what Frédéric Jenny said yesterday: perhaps the new regulation should establish an obligation for the national competition authorities to cooperate. Finally, the functioning of the network will require a lot of informal consultations, which of course we already have. The members of the network will need to discuss competition policy problems: actual cases and also how to investigate distinct competition problems. A last point on language: the draft regulation makes no mention about this, but I think it is quite obvious that we will need a single working language. We could not possibly continue the current system, whereby we work in eleven languages; the upcoming enlargement makes the need for a single language more important. A network intranet will also have to be set up. I do not think that will be very difficult; we already use an internal system of transmitting encrypted information from the Commission to the national competition authorities. What we have to do instead — and here I agree with John Fingleton — is to design and set up a workable internal organisation within the national competition authorities with a view to their new tasks ahead. I do not see this as a problem of resources, but perhaps we could discuss that later on. It is all a matter of priorities, and I foresee that most of the cases before the Swedish competition authority will be Article 81 and 82 EC cases. 䉴 JACQUES BOURGEOIS — I have two short introductory remarks to make, and four brief questions to ask. First, I think there are two fundamental underlying policy choices to make when we talk about consultation within the network. The first is to say: ‘Well, let’s let the chips fall where they may’. In other words, let’s live with the

300

Panel Discussions

inconsistencies or lack of uniformity of application, because it will all sort itself out over time. If I may say so, this is somewhat like the approach in the US. The other choice is to say: ‘No, uniformity and consistent application is of paramount importance’, and the only way to achieve them is to allow the Commission to step in whenever there is a risk of inconsistent application and non-uniform application of the EC rules. This is, in fact, the option advocated by a good number of practicing lawyers. Could the solution be somewhere in between? This mid-way would be a form of cooperative joint management of the cases within the network, to be achieved through reinforced consultation and allowing everybody to have a say about things. In the end, if there is no consensus and a decision is to be made, that decision could be made by qualified majority. If we were to opt for such a system, a certain number of consequences flow in terms of consultation; more precisely, with regard to the scope and consequences of consultation. My second remark is that I find myself in agreement with many of the views put forward by Monica Widegren in her written contribution. She defined consultation as the initiative taken by one or several competition authorities to share information within the network in order to get reaction from other network members that will lead to further action. In her paper she actually extends the concept of consultation to the request for assistance and other matters. She insists very much on the horizontal nature of consultation, and I think that is an important point. And now the first question: ‘Consultation about what?’ In other words, should consultation occur with regard to competition policy issues that arise in general, or about cases? In my view, consultation about cases would, of course, involve considerable costs. How could we handle this matter? I think that, to start with, the information to be brought before the network should be limited to the essential elements of each case. Then, if the national competition authority that submitted this information also wants to request assistance from other members in relation to the case, a consultation process would be triggered. Or vice versa, other members of the network might request consultation about the case. My second question is related to Monica Widegren’s proposal that information about cases should be submitted to the network after the first formal investigative measure. Isn’t that a bit late? If we look at the Commission’s enforcement practice, we see that quite a few things would have happened before the first formal investigative measure is taken. I do not think that this is a good solution. It would be more sensible and efficient if the other members of the network were consulted before the first formal steps of investigation were taken. What should be the object of such consultations? The answer depends again on what one wants to achieve, but certainly this should include advice from the Commission, interpretation of the law, mutual practical assistance and, finally, the draft decision or statement of objection. My last question concerns the legal effect of the consultation process. On this issue too, the views appear to be very mixed. On one view, the outcome of the

Technical Problem Areas

301

consultation should not necessarily be given legal effect as the new regulation will give the Commission enough power to intervene to avoid major inconsistencies in the application of EC law. On another view, the results of the consultation process should be legally-binding on the national competition authority that is handling the case, as this would be a way of ensuring consistency and uniformity in the application of the EC competition rules. Finally, some maintain that there is no need to take a formal position on this issue because the outcomes of the consultation are a form of ‘soft law’. In other words, it would be difficult for the national competition authorities to depart from the line of action suggested as a result of the consultation unless there were very serious reasons for doing so. Finally, I think we should also take into account how these consultations are going to affect private parties. I am not going to come back to everything that has already been said on this. One extreme view is that the parties should have a say during the consultation process on how a particular rule should be applied to a specific case. I think that this would probably be very difficult to manage in practice. On the other hand, it is not very satisfactory to have the national competition authority dealing with the case imposing the result of the consultation process on the parties as a decision that cannot be challenged. 䉴 WOUTER WILS — I would like to start with a cautious disclaimer: I will not speak for the Commission or the national competition authorities; I will just express my personal views on this subject. Also, I will base my remarks on the Commission’s original proposal for a new regulation (other texts that have been circulated in the Council have never been approved by the Commission, and have been even less approved by the Council). My assignment for this workshop was to talk about the network from the perspective of the European Convention of Human Rights (ECHR). When I received the task I did not really see the link between the two, but in the meantime I have discovered that this is an extremely interesting subject. In my written contribution I have actually broadened it, in the sense that I also took into account the Charter of Fundamental Rights of the European Union (CFR), a document that has been ‘solemnly declared’ by the European Parliament, the Council and the Commission on 7 December 2000 in Nice (whatever that means in terms of its status in the architecture of the EC legal system). In any event, it would be unthinkable for the new system not to fully respect the provisions of both the ECHR and the CFR. The Commission’s proposal for a new regulation envisages three functions for the network: the efficient allocation of cases between its members, assistance between the members in terms of fact-finding, and coordination of action in order to ensure the consistent application of EC competition rules. Within the latter I would distinguish between exchange of information, consultations, and then the mechanism of control by the Commission over the national competition authorities on the basis of Article 11(6).

302

Panel Discussions

There are two sets of provisions in the ECHR and the CFR that are relevant in this context. On the one hand, there is the right to fair trial and other guarantees for defendants in criminal proceedings. You may ask yourselves: ‘What does this have to do with the network? EC competition law is not criminal law.’ The concept of ‘criminal’ has its own autonomous meaning in the ECHR. Maybe a more appropriate term would be ‘penal’ because it has a wider sense than ‘criminal’ and I think it is impossible to deny that EC competition rules are of criminal law nature within the meaning of the ECHR. I think that the meaning of the concept of ‘criminal law’ is the same in the CFR, because if you look at its preamble and the last but one article, they take over — possibly even extending — the defence rights established by the ECHR. In the second half of my written contribution I went through all the different elements of the network looking for possible problems of compatibility with the ECHR and CFR. After discussing them, I concluded that most of them are not real problems. I would like to mention now two of the seven problem areas that I identified in my written contribution. The first is the problem of what American legal jargon would call ‘double jeopardy’: the right not to be prosecuted and tried for the same offence twice. This right is laid down in Article 4 of Protocol VII ECHR, respectively Article 50 CFR. Back in 1969, in Walt Wilhelm, the ECJ held that there was no problem of double jeopardy if a company was prosecuted both by the Bundeskartellamt under German competition law and by the European Commission under EC competition rules, because these laws were not the same.7 The difference laid in the EC competition law’s orientation concerning effects on trade between the Member States. If I compare this case-law with a recent case from the Strasbourg Court, I am not sure that this idea that the two sets of law are different still holds true. The principle that seems to have emerged in Strasbourg is that two prosecutions are for the same offence when their essential elements overlap;8 vice versa, two offences will be separate when each contains an essential element that is not found in the other. The same principle is applied by the US Supreme Court in the interpretation of the 5th Amendment to the US Constitution. To cite the US Supreme Court: ‘Each provision should require proof of an additional fact, which the other does not, for it to be separate offences.’9 So, we may potentially have a problem with the Walt Wilhelm doctrine. Article 4 Protocol 7 ECHR, prohibits prosecution of the same offence twice within the same jurisdiction. One could say that two prosecutions, taking place in two distinct Member States, do not infringe the ECHR. The situation is distinct if the same case is prosecuted by the Commission and one Member State. Community law is part of the national law, and therefore not a different jurisdiction. At any 7 Case 14/68 Walt Wilhelm v. Bundeskartellamt [1969] ECR 3. 8 Judgment of the European Court of Human Rights of 29 August

37950/97. 9 Blockburger v. US, 284 US 299 (1932).

2001, Fisher v. Austria, case

Technical Problem Areas

303

rate, Article 3 of the draft regulation establishes the primacy of Community competition rules. There is no doubt that in such cases it will be the same offence for which there is a risk of double prosecution. While the ECHR prohibits double prosecution of the same offence within the same jurisdiction, the CFR prohibits double prosecution of the same offence within the European Union. I think the risk of double jeopardy is augmented in this situation. However, in recital 16 of the draft regulation, the Commission proposes that there should always be one single competition authority to deal with each case, and that should be the authority best placed to handle the matter. We do not risk double jeopardy if this proposal is embraced. The second problem area is that of who is going to be a member of the network. The problem is that in some Member States the courts have a role to play with respect to the decision to start prosecuting a case. What is relevant here is Article 6 ECHR and the case-law of the Strasbourg Court on the conditions of independence, impartiality and full jurisdiction that courts have to satisfy. I think that there is an obvious tension between independence, impartiality and full jurisdiction, on the one hand, and having to cooperate with the Commission on the other; this tension involves consultation, the transmission of documents in response to a request from the Commission, and possibly even the case being removed by the Commission under Article 11(6) of the draft regulation. For most Member States there is no problem in this respect because their antitrust prosecution systems do not involve the courts, and the decisions of the competition authorities can be appealed before courts. There are three Member States for which I do see a problem: Sweden, Ireland, and Finland. So, I believe that this issue should be clarified in the new regulation. The draft regulation is not clear on this point, as it does not define what constitutes a ‘competition authority’. The Commission has not taken a position on this. I think it should be made clear that the national bodies, which we sometimes call competition authorities but that qualify as independent and impartial courts within the meaning of Article 6 ECHR and Article 47 CFR, should not be brought under the scope of Articles 11 and 14 of the draft regulation. At the same time, and in order to achieve consistency, they should be brought under the scope of Article 15; that is, they should be brought under the same rules covering courts that apply Community law in litigation between private parties. And, of course, because these entities are courts, they should also be able to ask for preliminary rulings from the ECJ. I fully agree with Dr. Böge, who said yesterday that the Community body mainly responsible for ensuring coherence in the application of EC competition rules should be the ECJ. To my mind, a possible counter-argument — which was made yesterday by Céline Gauer — is that, just as there is no problem if the Bundeskartellamt applies EC competition rules, because its decisions can be appealed in court, there should be no problem if the Irish courts were to cease to be independent, because there would still be an appellate process and the Irish appeal courts would still be independent. I do not think that my definition would go against De Cubber v. Belgium and Findlay v. United

304

Panel Discussions

Kingdom case law, 10 where the Strasbourg Court said that its acceptance of ‘decisions taken in first instance by administrative authorities open for further appeal’ does not apply to traditional criminal law. I think that this definition certainly applies to Ireland, and could also be transposed to the Swedish and Finnish jurisdictions. In my view, the essence of the Strasbourg case law is to accept the historical situation of administrative authorities whose decisions are open to appeal. But I do not think that the Strasbourg Court would accept a situation in which a court that was once independent became dependent. DENNIS WAELBROECK — Yesterday we spent a lot of time trying to figure out whether the Commission was the sun or the moon in the new system and, more generally, where is the centre of the new universe? Since we are in the city of Galileo Galilei, I wonder whether we have put the centre of the system in the right place. If you read the Bible, at the centre of the creation are Adam and Eve, and it is for them that the rules are made. So, to return to our world, I think that the new rules are not for the competition authorities but for the business community. After this short introduction, I would like to make two preliminary remarks. The first is that competition law is not only about cartels. As I see it, competition law is about trying to strike the right balance between economic efficiency on the one hand, and workable competition on the other. It is there to define the sound limits within which companies should conduct their daily business. I emphasise this because I have the feeling that too often when we discuss the modernisation project we have a very repressive vision of the purpose of competition rules. This distorts the perspective because, if the law is only about cartels, there is almost no limit to what enforcers should do in order to fight them. I would like to think that not all companies are wrongdoers, and that they try to know what the law is about and observe that law. This area of law is extremely complex and difficult to apply, and what I would therefore expect from any reform project is clarity and simplification. My second preliminary remark is that the network is not some sort of uncertain ghost in the future; it is a daily reality we already live, with or without the modernisation project. What we are seeing now in our daily practice is that, instead of having only one interlocutor — as, in the good past, the Commission — cases emerge everywhere. As Giuseppe Tesauro said yesterday, this state of affairs is probably very lucrative for the lawyers, so we should not complain. But I am not sure this is necessarily the best way to operate. Just to give you a few examples: in the petrol station business there is an interesting question that emerges now, after the adoption of the new regulation on vertical restraints, which is whether a system of guaranteed margins for the retailers is legal or not. In Italy, the position 䉴

10 De Cubber v. Belgium, judgment of 26 October 1984, Series A no. 86, paras 31–32; Findlay v. United Kingdom, judgment of 25 February 1997, Reports 1997–I, para 79.

Technical Problem Areas

305

appears to be that they are not; in Spain, the position appears to be the contrary. The problem is, which rule would be applied within the network? And finally, is having all these procedures operating in parallel the most efficient way of handling cases? Returning to the issue of the network and the ECHR, I would like to make three points. I may be a bit provocative, but this is why we are here, so this should not be seen as a criticism of the modernisation project. The first point is that, as Wouter Wils said before, in this area of law we are speaking of imposing sanctions that may be of a criminal nature. Article 7 ECHR establishes the principle of the legality of the sanction applied. In other words, the law and the sanction should be predictable. There is a long line of case-law from the Strasbourg Court on the predictability of the law. In simple terms: it establishes that one should be able to know in advance what the law asks of him/her, and what sanctions will be applied if s/he infringes it. I agree, of course, with Dr Böge and Ian Forrester that there cannot be absolute legal certainty, particularly in this area of the law. This is not the point; I just wonder how much legal uncertainty is acceptable, or desirable, in a system where about 26 competition authorities will very soon apply the same law but may often go in different directions. Just to give a few examples: in Michelin, the European Commission imposed a fine of 20 million Euro for a rebate system in place in France, even though that system had been reviewed every year and approved by the French competition authority.11 So, one national competition authority says your behaviour is legal, and then the Commission comes in and imposes a fine of 20 million Euro for the same behaviour. Is this compatible with Article 7 ECHR? Another recent example is Visa International:12 following the Commission’s decision, that had found no restriction of competition in the non-discrimination rule, Sweden and the Netherlands prohibited the non-discrimination rule on the basis of national competition rules. And Belgium is now starting a procedure against the same company, threatening to impose a major fine for the same behaviour. I could go on and on with such examples, because lately I feel like all my practice is going in that direction. I recently notified the Commission of a selective distribution system for a major software company: while the Commission issued a comfort letter, the French competition authority imposed a fine for the same system. This is the reality of our practice, and I am not sure that the modernisation project fully addresses these issues. What I see instead is that we are going to lose the benefit of one-stop shopping in the current system — that is, application of Article 81 (3) EC by the Commission — and, frankly, I do not see any compensation for the companies. Once again, is it acceptable to have no rule on the allocation of cases and mutual recognition? Is this compatible with Article 7 ECHR? 11 COMP/36.041 – PO/Michelin, Commission Decision of 20.06.2001. 12 Case COMP/29.373 Visa International — Multilateral Interchange Fee,

24 July 2002, OJ L 318 of 22.11.2002, pp. 17 et seq.

Commission Decision of

306

Panel Discussions

If I go back to my starting hypothesis, when Adam and Eve breach the rule that forbids eating the apple, as far as I know, there is only one authority to sanction them, and not 26. My next question is: was it impossible to introduce another system? If I look at the Community legal system, I see everywhere mutual recognition where regulation is done by networks. For example, in the medical, pharmaceutical, veterinary, pesticide sectors there are networks of national agencies working together; they appoint one competent authority to deal with a case, and the others are consulted. There is a procedure of mutual recognition, and everybody accepts the ruling of the designated authority. We also heard yesterday about the NAAG in the US, whereby various authorities bring joint actions before one single court. There is also the alternative of having a system of joint management of cases, as proposed by Jacques Bourgeois. I find that the chosen system is minimalist, and it creates problems under Article 7 ECHR insofar as there is uncertainty about the forum in which one’s case will be decided. Very briefly, with respect to the network and the CFR: the reform does not envisage any harmonisation of the national procedural rules whatsoever, and I think that this is really a mine-field of problems. Finally, the EU is still not a signatory to the ECHR. So, if the Commission will apply Article 11(6) of the draft regulation, I wonder if this in itself is not a violation of the ECHR, because it denies companies the defence rights they would enjoy under the ECHR if prosecuted by a national competition authority. Having said this, I repeat that I support the modernisation project generally; I have just tried to make some constructive suggestions for the future. 䉴 CLAUS-DIETER EHLERMANN — Your remarks remind me of the fact that monopolies are traditionally easier to run than deregulated industries, for which you need rules, and here we are discussing the new rules for the post-monopoly period. My second point is that mutual recognition requires mutual trust, and if there is no mutual trust there is no mutual recognition. Having said this, the effect of further integration in the internal market should normally be at least a minimum of harmonisation.

CÉLINE GAUER — A lot of interesting critiques and remarks were made this morning, and I would like to make a few comments on two of the aspects touched upon in our discussion. On the exchange of confidential information under Article 12 of the draft regulation, I would like to give some answers to the points and criticisms raised by Marc van der Woode and Santiago Martínez Lage. First of all, Article 12 applies notwithstanding any national provision to the contrary. This means that no national rule can prevent the national competition authorities from exchanging information or using the information received. Even if the information has been collected in the course of criminal law proceedings, this would not prevent a national competition authority — which is traditionally an administrative authority — from



Technical Problem Areas

307

using the information. This is because Article 12 empowers national competition authorities to use the information received. Article 12 ultimately creates a new legal ground on the basis of which information may be obtained. Some of the speakers this morning called for the introduction of a procedure whereby we should have a legal decision, and the possibility of challenging it, before information is exchanged. If this were to happen, we might as well get rid of Article 12; the time possibly needed to have such a decision reviewed in court, with a risk of appeal, would be prohibitive. One may have to wait for two years before the court decides that, for instance, the Commission could take up a case and require the necessary information from the national competition authority. By then the case would be dead anyway. Furthermore, it could be that we need to exchange information even before a company becomes aware that the receiving authority will be carrying out an investigation. If the company became aware at this stage, it would probably be delighted to destroy all evidence that the receiving authority could later seek to find in its own jurisdiction. This is the reason why I do not think it is possible or desirable to have such a procedure. Furthermore, I do not think it is necessary; the level of confidentiality, which was mentioned several times this morning, will be the same. Here again (I mentioned this yesterday), I think that Article 27 of the draft regulation harmonises the level of confidentiality within the Community for all information exchanged pursuant to Article 12. It is not really important to establish who is going to decide the status of the information transmitted — be it the transmitting or the receiving authority — because the rules and standards to be applied will be the same anyway. Moreover, the receiving authority will generally be in a better position to assess the status to be given to the received information. At the beginning of an investigation, companies would normally claim that everything is confidential. Then, when we prepare the file for access at the Commission, we have more a precise discussion with the companies about each and every document that is contested, and we find a solution. So, for me, it is more for the authority that is going to make the final decision on the case (rather than the transmitting authority) to have this kind of constructive discussion with the companies. The second argument I wanted to comment upon is related to defence rights, and I refer in particular to the legal privilege for in-house lawyers. I do not agree with Marc van der Woude that the highest standard of protection should always prevail. I think we have a common standard of protection within the EU, which is already high, and that derives from the ECHR. The Commission is not a signatory to the ECHR, but it nevertheless respects it because of the case law of the European courts. If some Member States give legal privilege to in-house lawyers, I do not believe this is an expression of a fundamental right; it is simply a political choice that was made in a particular country. Companies carrying out their activities in Europe know that they are exposed to the possibility of having advice given by their in-house lawyers used in competition proceedings. I fully share the opinion expressed by Frances Barr this morning that the OFT could, for example, use an advice given by an in-house lawyer that it receives via the French

308

Panel Discussions

competition authority (which does not recognise the privilege). Again, this does not impinge on a fundamental right. For these reasons I do not think that we really need the sort of procedure proposed by Marc van der Woude. Article 12 of the draft regulation should be left to the competition authorities. It is for the deciding authority to make all the necessary assessments, and for the company to contest the use of this sort of evidence before the court. Of course, if the information were to be illegally collected by a national competition authority, then companies would have the normal recourse to the courts of the country where the collecting authority is located. But this is a completely different issue. Thirdly, I would like to comment very briefly on Wouter Wils’ intervention. He made it clear that he was expressing personal views, and not the official position of the Commission. I would like to contradict him on two aspects, also from a personal standpoint. If Wouter Wils is right on the issue of double jeopardy, then everything we have been talking about this morning with respect to teamwork within the network is worth nothing. The members of the network will not be able to work together if we can have only one decision-making body. Furthermore, if we allow only one decision to made, the result is that we will under-punish infringers. By contrast, I believe we should allow each national competition authority to take into account the effects of the infringement in their jurisdiction and to impose sanctions. The same infringement does not necessarily have the same effects in different jurisdictions. For example, with respect to an infringement producing effects in Italy and France, the Italian competition authority will take into account mainly the effects on Italian consumers, and the French competition authority will look mainly at the damage it causes to French consumers. I believe that — at least from the horizontal perspective — parallel action is perfectly sustainable and does not violate the principle of double jeopardy. Having said this, the issue is of course more complicated from the vertical perspective. If a national competition authority has adopted a decision concerning a certain infringement and the Commission then wants to make a more restrictive decision, this would probably create problems from the point of view of the principle of double jeopardy. Finally, Wouter Wils gave an answer to my comment yesterday about the possibility of making Article 11(6) applicable to courts. I take note of his answers, but I do not think it really convinces me. I do not think that competition rules, when applied to companies, fall within the scope of what the Strasbourg Court defines as ‘traditional criminal law matters’. Furthermore, I am not sure of the fact that, if the Commission could use Article 11(6) up to the first level of decision-making, this would disqualify the Irish court (for example) as a fair and impartial tribunal. However, that issue will probably have to be clarified by the court. ALEXANDER SCHAUB — First, I hope that everybody has appreciated the rich diversity and the liberty of opinion within the European Commission, to which I



Technical Problem Areas

309

will not immediately add further examples. Then, while listening to Dennis Waelbroeck’s critical remarks and the cases of conflicting decisions under the existing enforcement system that he mentioned, I was thinking that one cannot find better arguments for the reform. In response to the suggestion that we should introduce the principle of mutual recognition: this is basically what we are doing. The new system is based on the application of Community rules in all Member States, and the replacement of the Commission’s exemption monopoly by a system of parallel enforcement competencies. De facto, this means that there is the confidence among the members of the network that cases can be decided by their colleagues. Technically speaking this might not be ‘mutual recognition’, but I would say it goes even further that that. By the way, Dennis Waelbroeck’s optimistic description of regulation in the pharmaceutical sector is certainly well intentioned, but things are much more complicated in the real world. It so happens that I have personally been involved in the launch of the European pharmaceutical agency — and I continue to be very proud of this — but the realities of this sector are exactly the opposite of what you were describing. This is a case where you see very clearly the limits of the mutual recognition principle. What we are proposing for the competition area actually goes much further than what is currently practiced in the pharmaceutical sector. On a different note, I fully agree with John Fingleton on the fundamental importance of adopting pragmatic attitudes to the launch and application of the new rules, and avoiding by all means any obsession with perfectionism. I also agree with those who said areas of the reform still need to be clarified and there are rules that need further improvement. We have already seen this in the merger control area. At the same time, this reform has been prepared and thought over for years now, and I believe there are few other European projects to have been so carefully prepared. Of course, there will be room for further improvement, but those who expect everything to be absolutely clear and perfect from day one are not the right persons to make the new system a success. Another point that emerged clearly from this discussion is that the new system will trigger the need for convergence (which is the term that I prefer to ‘harmonisation’). I think we should avoid falling again into the harmonisation obsession of the past. As Commissioner Monti said yesterday, the mere fact that there will be such close cooperation between the competition authorities should serve as an eye-opener. The areas that have been identified here as problematic — and that were sometimes depicted in frightening colours — already exist under the current system, but perhaps there is less awareness of their existence. An example is the need for extra resources and staffing. The first question is: do we use our staff in the most efficient way? I am sure that close cooperation between the competition authorities has already improved our working and training methods and our enforcement priorities. For instance, the next ECA meeting will discuss common training programs, through which best practices are transferred much more easily, and which encourage peer pressure and peer reviews. Those who have

310

Panel Discussions

followed the developments over the last three years will remember quite a few examples of changes already induced by the prospective reform. A last remark concerns timing: at the last Council meeting (in Barcelona), the European heads of state and government expressed their hope that the reform will be approved by the Council of Ministers before the end of 2002.13 If this timing is respected, then certainly another full year will be reserved for the preparation of the effective start of the new system. CLAUS-DIETER EHLERMANN — I fully agree with you with respect to the pharmaceutical sector: this is an example where mutual recognition did not work, and where it was therefore necessary to establish a central agency.



ALEXANDER SCHAUB — The initial effort to apply the principle of mutual recognition was a total failure. Then the concept of a European agency came into being, and it is not a great success either. How does it work? All the cases that are ‘controversial’ by nature (for instance, those concerning biotech products) go directly to the European agency, and so do other cases in which some Member States are hesitant to recognise decisions previously made in other Member States. This mechanism is very close to that established by Article 11(6) of the draft regulation, although the Commission has never used its power to take up a case from a national competition authority against the will of the Member State concerned, and it does not intend to do so in the future. However, this difference of approach is fully justified in the pharmaceutical sector, because issues of health are at stake.



CLAUS-DIETER EHLERMANN — I totally agree. If the principle of mutual recognition is not workable and there is a need for a certain degree of uniformity in the application of certain decisions, then a mechanism of central control is indispensable. By the way, following the discussion we had yesterday, it is still not very clear to me what the current situation is in the telecom and energy sectors (that is, now the principle of mutual recognition no longer applies, and the Commission is no longer the decision-making centre). Personally, I do not believe in the sustainability of the telecom regulators working together without a very strong element at the centre, and that is absent the moment.



JAMES RILL — I think that Dennis Waelbroeck’s remarks focused on a very important split in the decentralisation process. It is something that people should really understand because it frames the whole discussion. Basically everybody accepts that a cartel is bad thing and should be punished, so there is not a whole lot of legal uncertainty in this area of antitrust law; there are some procedural issues to be dealt with, but these are not subtleties of the law. One of the objectives of the



13 The

Regulation was adopted 16 December 2002, and will become applicable from the 1 May 2004. (See Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ L 1 of 4.1.2003, pp. 1–25.)

Technical Problem Areas

311

modernisation project is to render more effective this area of law enforcement. So long as the procedural rules are respected, I think it is largely their own business how the Commission and the competition authorities of the Member States work this out among themselves. The issues are less clear-cut with respect to an abuse of dominant position. The non-discrimination rule example invoked by Dennis Waelbroek is very telling. In such cases the application of the law involves economic assessments that are very different from and more complex than those to be made in cartel cases. Prima facie it may be concerning to see different rules about the same kind of anticompetitive behaviour in different EC Member States, but instead we should see this as a laboratory to test the impact of the rules on the consumer. The rules can be harmonised afterwards. Then the question is: should that be done by the Commission and the Member State authorities? Or should it be left to the courts? Moreover, should the court system eventually evolve, or does it already have the ability to bring about a solution to the apparent irrationality of different rules in different countries? It seems to me that, in a decentralised enforcement system, the Commission and the competition authorities of the Member States should discuss together certain important enforcement and policy issues. For example, is resale price maintenance really a bad thing or not? Should it be prohibited with a per se rule or not? Other issues may be better left to the courts (such as whether there is a need for a non-discrimination rule or not), if only because they would otherwise overload the public enforcement system. To conclude, I see first an institutional issue about who does what. As to the more subtle questions of competition law, I believe that the courts should be given a bigger role to play. One should not lose sight of the main objective of the reform proposal, which is to liberate the Commission from the tedium of routine enforcement issues in order to allow it to focus on the big enforcement issues. In a certain sense this is also an answer to Mario Siragusa’s criticism about the fragmentation of the EC competition law system, and to doubts about whether the Community’s judicial system is robust enough to ensure coherence in the application of EC competition rules. As a practicing lawyer, I have greater confidence in the courts than in the administrative authorities. They may provide a more impartial resolution mechanism for some legal issues and, due to the adversarial process, they certainly have the ability to arrive at informed decisions. 䉴 CALVIN GOLDMAN — Listening to the remarks made this morning concerning the very interesting and challenging subject of the exchange of confidential information, I thought this is a discussion that is quite parallel to one that a number of us in this room were engaged in very recently at the OECD. And, of course, the issues addressed today are parallel to those that we wrestle with in the area of international cooperation between competition authorities. Most recently, I was asked — along with James Rill and a few others — to put together a response to the OECD’s very long list of questions in the area of exchange of international

312

Panel Discussions

confidential information in the context of trans-border cartel investigations. We attempted to address the two topics of safeguards and process. The issues that emerged are almost identical with those arising within the EU, and with those that arise in the context of bilateral cooperation between the EU and the US, Canada, Australia, and so on. To summarise them very briefly: on the safeguards side, the thorniest issues are those of legal privilege and immunity, and testimony protections. Then there is the question of giving notice about the exchange of information. Today I heard some very interesting remarks from the head of the Dutch competition authority, in the sense that in the Dutch legal system the parties have access to a special procedure that allows them to go to a special court for a right to be heard in advance of the exchange of information. I am not sure that a similar procedure exists anywhere else. On the contrary, the US Department of Justice very strongly opposes any notice about the exchange of information whatsoever. In Europe, the business community has asked for one. In fact, in our work for the OECD we suggested a compromise solution that tries to bridge the gap between these two positions in the sense that there ought to be notice only when it will no longer prejudice the trans-border investigation. So the authorities themselves can decide when to give notice, but they need not give notice if there is a risk of prejudicing a dawn-raid or any other instruments of investigation. So far this proposal has not been sufficient to close the gap. Nevertheless, the issue must be addressed, because we will otherwise not see effective cooperation in the pursuit of cartels. This is an issue of balance and of fair trial. If we can reach some consensus on this, the competition authorities will have a set of safeguards through which cooperation in enforcement becomes more transparent. JAMES RILL — On the same point of the issuing of notice, I think there is a difference between what might be the unofficial position of some in the management of the Department of Justice’s Antitrust Division, and the official position of the same. I would not want to contradict Calvin Goldman on the informal position, but if one wants to understand the official position of the Department of Justice’s Antitrust Division, one first needs to understand the circumstances under which information may be shared. Apart from the court procedures concerning the exchange of information — the delivery of information and aid of formal proceedings under the Grand Jury rules — there are only two legal instruments in the US that cover the exchange of confidential information. One is the Mutual Legal Assistance Treaty with Canada, which expressly refers to the exchange of information in antitrust proceedings. The other is the US–Australian Agreement, that was entered into under the International Antitrust Enforcement Agreement (commonly known as ‘the IAEAA’), which authorises the exchange of confidential information. With respect to notice prior to the exchange of that information, it was a lively topic of debate during the legislative proceedings in 1994, leading up to the enactment of the IAEAA. In its Committee Report, the Committee on Judiciary of the House



Technical Problem Areas

313

of Representatives provided that notice should be given prior to the exchange of confidential information with foreign agencies operating under an IAEAA agreement, except where that notice might prejudice the progress of investigation. That particular provision was endorsed and actually put into the report at the urging of the US Council for International Business (USCIB), and was subsequently endorsed by the then Assistant Attorney General in a formal response to a request for information from the USCIB. This is the formal position of the Department of Justice’s Antitrust Division. Again, I would not like to contradict Calvin Goldman’s perception of the informal position as possibly being something along the following lines: ‘No way we are going to give advance notice.’ Regrettably, the prejudice to the course of an investigation is somewhat in the eye of the beholder. What qualifies as confidential information is not always so easy to determine, and what will prejudice the process of an investigation can be very subjective, very nebulous. My guess is that, in any kind of exchange of information taking place in relation to a cartel investigation, the enforcement agency tends to believe that prior notice will jeopardise the investigation in almost every circumstance. Whose investigation is jeopardised — the one of the authority requesting the information, or the one of the authority providing it — is another question. They will probably both take the same position. So the likelihood of the prior notice being given does not provide an enormous hope for a safeguard (unless there is some negotiation prior to the information being given to the authority in the first place, which is something that has happened). The question is, what should be done about it? I think that there is a strong argument in favour of the view that prior notice is not entirely appropriate for every case. Where there is a hard-core cartel investigation in process, and there is at least some reason to suspect the likelihood of destruction of documents, or the possibility that a key witness will disappear, the giving of prior notice would certainly jeopardize the investigation. On the other hand, let us assume that knowledge of the investigation already exists and it is just a matter of doing some clean-up work on some documents: in this case, would prior notice jeopardise the investigation? Probably not. The point is: except in the Netherlands, there is no timely court review procedure available for instances of transfer of information. One proposal brought before the ICC by Canada is that of a ‘post-talk’ notice; in other words, the authority providing the information should inform the interested parties that a transfer of information has occurred and the information is about to be used in proceedings against them. This kind of notice would at least provide some opportunity for the review of the giving authority’s accountability. Having said this, unless there is some willingness on both sides to address the problem, possibly through guidelines as to what will or will not jeopardise the integrity of an investigation, etc., all that I have done is put a fairly intractable problem on the table. At least in a forward-looking pragmatic sense, I think that such guidelines may provide some kind of flexibility.

314

Panel Discussions

CLAUS-DIETER EHLERMANN — My experience at the Appellate Body of the WTO leads me to believe that the issue of the exchange of confidential information is extremely important in practice, and therefore needs to be tackled. In a certain sense, our discussion in relation to the competition network might be a paradigm for other contexts. All attempts at cooperation between the US and the EC in Geneva — whether in the ambit of anti-dumping or safeguards — have failed when it came to the exchange of confidential information. I remember Barry Hawk once explained that US rules sanction the breach of a promise to keep confidential information secret. Therefore, if the rules on exchange of confidential information are not equivalent in different jurisdictions, there will always be a fundamental reluctance to exchange confidential information. That seems to be very much the case between the EC and the US, where the difference of positions is sharpened by institutional differences. Yet you cannot adjudicate a case conveniently without dealing with this issue. Also, competition agencies are generally more careful about these issues than those dealing with anti-dumping procedures. I also have the feeling that the courts are not as sensitive with respect to issues in the antidumping field as they are in the area of competition.



ANNE WILLEM KIST — As it seems that I have given rise to a misunderstanding, I would like to clarify my point about notice on the exchange of confidential information under the Dutch legislation. Dutch law establishes a general duty to inform the parties concerned about the exchange of information, but this is without prejudice to the power of the Director General of the Dutch competition authority to exchange confidential information, subject to certain conditions. Having said this, I do believe that this is a very important issue, and we have to find ways of ensuring a fair balance. I do appreciate that conducting an investigation might not be unduly jeopardized by an exchange of confidential information but, on the other hand, I also believe there are legitimate interests of the parties concerned that need to be safeguarded.



䉴 MARC VAN DER WOUDE — I have to confirm what Director General Kist has just said: indeed, the Dutch provision concerning the exchange of information does not explicitly call for such a notice procedure, but I think it could be applied by analogy. It would not be completely unfeasible and it could also take place within very short deadlines (I think the deadline for a reaction is reached after eight days). So I do not think that introducing such a procedure would be difficult. In reply to Céline Gauer, I do not think you can argue that there is a common standard concerning the right of defence, or that there are common rules about gathering evidence; there are objective differences. The same holds true for the treatment of documents. I am not speaking about the concept of confidentiality, but the rules concerning access to public documents differ. So, I think we have an issue here. Mr Rill’s point about the a posteriori notice was interesting. In that respect, we could also think about the possibility of damage claims; an authority that

Technical Problem Areas

315

decided not to inform the parties concerned about the exchange of information in order not to jeopardize an investigation would assume the risk of damage claims arising later on. Another suggestion would be to identify the differences between the various pieces of national legislation on these issues. Assuming that it is neither workable nor advisable to establish common standards through the new regulation, one could still think of preparing (for example, during the period preceding the entry into force of the new regulation) a kind of chart of the existing differences. For example, are there countries in which information is gathered under criminal procedural rules? Can this information be sent to another country? Such a reference guide would allow an authority to know in advance the risks that it is taking by receiving the information. JOHN FINGLETON — The new Irish competition law, which was enacted this week, contains a very general provision concerning the sharing of information with other national competition authorities. In principle, this provision will allow us to exchange information with other authorities under our national law. There are probably similar legislative developments in other EC countries, probably mainly driven by a growing concern amongst competition authorities about how to tackle international hard-core cartels. Otherwise, it must be said that there is nothing in our national legislation to prevent us from sharing information about every other aspect of antitrust enforcement. 䉴

HELMUT BROKELMANN — I think that the topic of the exchange of confidential information should be addressed by the Commission. We cannot leave things as they are now, otherwise there is a serious risk that undertakings will disrupt the new enforcement system. Our clients are not prepared to have their confidential documents circulating freely throughout the Community. They need to have the possibility of taking legal action to prevent the exchange, and they need to be given notice of such exchanges. I acknowledge the fact that there are different needs to be balanced here, but, like Marc van der Woude, I also think that it is possible to introduce some sort of interim procedure whereby it could be assessed whether the conditions for the exchange to take place have been met. It would also be useful if the Commission adopted a notice interpreting the concepts of professional secrecy, business secrets, internal documents, and so on. As a matter of principle, the concerned parties should be able to appeal in the jurisdiction of origin of the information against the exchange of information a priori; exceptionally, in order not to obstruct the investigation of the receiving authority, it could also be acceptable if the undertaking had the right to appeal in the country of destination of the information. Finally, even if the Commission were to adopt a notice interpreting the concept of confidentiality, the practical application of this notion by a host of national authorities will probably lead to very different results. This is why there is a very clear need for a procedure allowing the parties to appeal these kinds of decisions by the competition authorities.



316

Panel Discussions

LAURENCE IDOT — On the issue of confidentiality, I believe we can distinguish between differences in the definition of ‘confidentiality’ (which gives rise to what we call in international private law ‘conflicts of qualification’) and differences concerning the content of the provisions on confidentiality (which gives rise to the traditional ‘conflict of laws’). In order to solve conflicts of qualifications, the only solution is to adopt common definitions. In this respect, I am not sure that the safeguard we find in the draft regulation is clear enough to solve such conflicts. In order to solve the conflicts of law, the solution is either to adopt a rule of conflict or to adopt a common rule. I am not sure that the draft regulation offers a common, EC-wide rule; this is because it makes reference to the national law, and I agree with the concerns expressed by Marc van der Woude on this point.



CALVIN GOLDMAN — A possible three-fold programme dealing with these issues would involve, first, establishing common and transparent safeguards, including on the issues of legal privilege (as reflected in the written contributions of Margaret Bloom and Anne Willem Kist). Secondly, the issue of notice would have to be dealt with; a possible solution that would not prejudice investigations would be the a posteriori notice (as explained by James Rill). At that point, there is some accountability on behalf of the competition authorities. Thirdly, in order to protect the authority’s ability to carry out investigations, the right to review the notice should be limited by a very strict time period so that the competition authority can fix any irregularities and continue its investigation.



CÉLINE GAUER — To be clear, the Commission does not say that companies should not be informed about the transfer of information. Of course companies have the right to be informed that information has been exchanged or will be exchanged, but we also have to have some flexibility in order to safeguard the effectiveness of investigations. As for sanctions: there are sanctions if, for example, EC officials breach the obligation to respect confidentiality. There are certainly similar rules in the Member States. However, I take the point that we could create a sort of reference guide in order to make these rules known and to show how stringent they are; this would make it clear that we are all bound by the same concept. I insist that Article 27 of the draft regulation is very clear: we are all bound to respect professional secrecy. The issue here is what sanctions apply for the breach of this obligation. I also fully take your point with regard to guidelines, or a notice, interpreting the concept of professional secrecy. We have to make sure that the application of this concept will be the same in all the Member States. But I would simply advise companies not to push too much for that. The reason is simply that, if you have severe sanctions, you also have a greater likelihood that the competition authorities will err on the side of caution with regard to broader concepts, so as not to risk being held liable. In other words, companies may be better protected 䉴

Technical Problem Areas

317

by a broad concept of professional secrecy than by one more that is more specific and limited. IAN FORRESTER — I have a comment on James Venit’s intervention, and another on pharmaceuticals agencies and judicial review. Most of my practice involves looking at really tricky issues about which it is not obvious which way the law should be interpreted. In these situations, whether a practice or an agreement or the system is good or bad depends upon how you look at it. In the new enforcement system, these issues will be argued in 15 — and actually up to 27 countries — instead of just in Brussels. This is a very good thing. The Commission has given up its theoretical monopoly, and this is a good thing too. This means that we are going to have more decisions, more discrepancies, more errors, and more inconsistencies. In the early days of the reform, solving those discrepancies or preventing the inconsistencies must be an important part of what the Commission will do. The notice on cooperation with the national courts14 failed because the Commission was too cautious, too conservative, and not wanting to give advice for fear of being unfair, while the courts were reluctant to consult the Commission. That did not work well. Roundtable discussions like this one, with people sitting around a table and discussing a common problem, are obviously right and have got to be a good way forward. This is especially necessary for the competition authorities of the new Member States from the East, who lack experience in using the principles of law in a flexible way in a dynamic market-place. I do not think anyone could have a problem with agencies discussing among themselves the best way of approaching problems. The vice of pharmaceuticals, published agencies and anti-dumping officials is that their decisions are not very well reasoned; to the contrary, they are very brief, unintelligible, and rarely presented with frankness. This makes judicial review of such decisions very difficult: the European courts are reluctant to look behind how an administrative agency has exercised its discretion. By contrast, in the competition field the Commission’s decisions are more reasoned. It would be very desirable if roundtables of the kind that I referred to before produced decisions reasoned on the merits and subject to proper judicial review. Otherwise we would once again have non-transparent discussions producing a common view that cannot be reviewed by the courts. And even I — though usually very tolerant and optimistic — would not be willing to accept a regime in which the experts discussed something among themselves and reached a conclusion, if that conclusion were not susceptible to proper judicial review. 䉴

CLAUS-DIETER EHLERMANN — We are now moving towards the subject of consultations. In the view of some among us, this can apparently result in a decision, although the notion of consultation per se does not seem to suggest this.



14 European Commission (1993): Notice on Cooperation Between National Courts and the Commission in Applying Articles 85 and 86 of the EEC Treaty, OJ C 39 of 13.2.1993, pp. 6–11.

318

Panel Discussions

I think that the answer to the question raised by Ian Forrester — namely, whether there should or should not be some kind of judicial review following consultations within the network — depends very much on whether consultation is followed by a decision within the network, or only by a decision by the authority that requested consultations. MARIO SIRAGUSA — I honestly do not see how the network could make decisions on issues subject to consultation. This is not provided for in the draft regulation, inter alia because the power to make decisions is linked to a specific procedure for the adoption of such decisions. My comment is more related to the issue of judicial review, so it is linked to the subject of consultations. James Venit suggested that the task of avoiding an outcome in which different competition authorities take different positions in the application of the same EC competition rule, such as in the Visa example mentioned by Dennis Waelbroek, should be left to the courts. Under the current enforcement system, Visa has the choice of going to the Commission to ask about the legality of its practice, and this ensures the uniform application of the EC law throughout Europe. However, the Commission is practically abandoning cases of this kind even before the implementation of the reform. In the new enforcement system such cases will be the subject of consultations within the network, but I suppose that the result of the consultations will again be individual decisions made by the individual members of the network. Individual decisions made by members of the network can be appealed in court, which provides a safeguard against this kind of outcome. And, since the members of the network apply Community law, the national courts will refer requests for preliminary rulings to the ECJ. This is what I think James Venit had in mind when he referred to harmonisation via the court system. This underlines once again the importance of Article 3 in the draft regulation, because it is this provision that enables overall judicial control. On the other hand, I thought that there was a contradiction in Dennis Waelbroeck’s very interesting intervention. He was criticising Article 11(6) in the draft regulation, but doesn’t Article 11(6) establish the mechanism to avoid conflicting decisions? The Commission has the power to intervene and take cases over so as to decide them in a uniform way if the members of the network continue to apply the same principle of Community law in different and conflicting ways after consultations concerning a particular legal problem have taken place within the network. To conclude, I think that we have at hand two ways of dealing with the issue of conflicting decisions: judicial review (and I am very much in favour of increasing the number of cases referred to the ECJ for preliminary rulings) and Article 11(6) of the draft regulation.



䉴 JAMES VENIT — To continue on the same topic, I think that there are two arguments in favour of judicial review as a mechanism for ensuring consistency of decision-making. First, I think that courts may ultimately be better suited to deal with complex economic issues than the regulators, as the latter are sometimes driven by policy considerations. Secondly, I find very interesting the extent to

Technical Problem Areas

319

which one can invoke EC law in the judicial review of inconsistent decisions taken under national law affecting companies that operate throughout the Community. In the current enforcement system, the way out of such situations is to notify the Commission of the agreement or practice, relying on the primacy of EC law and the fact that national decisions based on national law cannot contradict an individual exemption obtained from the Commission. This possibility does not disappear in the new enforcement system: all that goes away is the Commission’s exemption monopoly. If an appeal against such a decision were to be denied in the national court, I think that there is still room to make a final argument in a reference to the ECJ that the national law has been applied incorrectly because the case had a Community dimension. This could be a way to replace the lacuna generated by the Commission’s withdrawal from its central role in guaranteeing the consistent application of EC competition rules. JOHN FINGLETON — I earlier made a point about institutional resources, and I would like to clarify it. What I meant to say this morning is that institutional resources are a problem for the network in so far as they create externalities across the Member States. In other words, if the Irish government does not allocate sufficient resources to the Irish competition authority in view of its participation in the network, this can become a common problem for the network, as all members of the network are expected to contribute to its good functioning in some way. I think that, besides the obvious case of consultations within the network, there are other more subtle ways in which resources matter. First, it is a fact that the national competition authorities will participate in the network whether they are well informed or not, so there is always the danger that some authorities will give opinions within the network without having actually examined the facts and the issues at stake. Of course, it would be even more worrying if the uninformed opinion were to come from the Commission. Secondly, each of the members of the network must find time to read others’ decisions, and this is very important in terms of what both James Venit and Ian Forrester have stressed about consistency in decision-making. Consistency is much more likely to occur if we do read each other’s decisions. At the moment we probably do not have the resources necessary to have two or three of our staff reading the decisions of the other competition authorities, and language is an important factor to consider here. I think that at least the important decisions made by other authorities should be distributed and read within the network. Monica Widegren said she did not think that resources were a problem; I think that resources are not a problem at present, insofar as it is only individual Member States that are affected and the shortage does not have consequences for the others.



䉴 JACQUES BOURGEOIS — I would like to react to some points made by Mario Siragusa and James Venit. I think we should acknowledge the fact that in Europe the enforcement of competition rules is primarily administrative, and judicial review is limited and late. I do not think that we should do away with the

320

Panel Discussions

administrative enforcement system, but we should build into it some of the guarantees for the rights of private parties that exist in the judicial system. For instance, if consultations within the network end up in some form of common resolution — not to mention decision — the positions of the private parties are affected. As such, the resolution should be challengeable; somebody should take responsibility for this resolution or decision and give the private parties the opportunity to be heard and to explain their positions. The judicial review system is not satisfactory in this respect because of the long periods of time that it takes in order for a decision to be reached. PAOLO SABA — I work for the Italian competition authority, but I am speaking in my personal capacity. I wanted to make a point concerning the functioning of the network and consultations within the network. It seems to me that little attention has been paid so far to the scope of application of the consultation mechanism. Nobody has raised the issue of which cases should go under the consultation procedure, as stipulated in Article 11(4) of the draft regulation. The Commission proposes basically that only prohibition cases be subject to this form of prior consultation; I find this sort of counter—intuitive, for different reasons. First of all, one of the major innovations of the reform proposal is that Article 81(3) EC will become directly applicable by the national competition authorities. One would expect that the major risk of the decentralisation process would be the relaxed application of Community rules to cases that, for example, raise problems of market foreclosure. In other words, since the national competition authorities will adopt decisions whose effects exceed the national borders, the degree of independence of the national competition authorities becomes even more important than in the past. It is astonishing to me that there seems to be no particular concern about the possibility of national competition authorities making decisions that are biased in favour of national interests. 䉴

䉴 CLAUS-DIETER EHLERMANN — I can only support what you have said. I remember the discussions we had at the edition of the workshop of two years ago, when we all agreed that it was absolutely indispensable to have Commission guidelines or a notice on the interpretation of Article 81(3) EC. I suppose that such a text would be discussed within the network before it officially emerges. Your second point reminds me of some personal reflections about international antitrust enforcement. I think that, at the international level, the problem is not one of over-enforcement, but rather of under-enforcement. It could very well be that some areas within the Community have a problem of under-enforcement. Under the new enforcement system, the problem will be solved within the network by means of peer pressure and bench-marking (instead of the Commission stepping in and solving this problem itself). In other words, the phenomenon of under-enforcement will be discussed within the network, and one would normally expect the network members in question to react. I believe that this is will be much better than re-centralising enforcement.

Technical Problem Areas

321

䉴 MONICA WIDEGREN — I think that Claus-Dieter Ehlermann is absolutely right: one of the positive effects of the network will be precisely that it gives its members incentives to apply EC rules. It is true that there are no safeguards against lax administrative enforcement (as Paolo Saba mentioned before). But I think that close cooperation within the network will involve all the members assisting each other to apply the EC competition rules in a professional way. On a different note, John Fingleton is right: there is of course an issue of resources. The network cannot function well if we do not have the right numbers of skilled staff. What I meant to say before is that I foresee that the application of Article 81 and 82 EC will take priority in our work in the future. 䉴 BARRY HAWK — It seems to me that several people around this table believe that judicial review of what the network is doing remains important. In so far as this is about the protection of individual rights, things are quite clear. My question is: what kind of judicial review can be performed with respect to the internal functioning of the network? I think there are two main reasons why the courts will be hesitant to review what the network is doing. One is ripeness: the courts will avoid performing judicial reviews of cases if they consider that it is too early to intervene with respect to procedures that are under way. In other words, the courts will examine only preliminary issues in such circumstances. The other is the administrative discretion enjoyed by the competition authorities in applying Article 81(3) EC. So I think a lot of thought should be given to the legal status and effects of actions and decisions taken within the network. 䉴 ULF BÖGE — Just to clarify a point: the network will not take any decisions about cases. Such decisions will be taken by the individual members of the network, and of course they are subject to judicial review. The main purpose of the network is to facilitate the exchange of views and information among its members. Maybe the network will also develop some guidelines, but decisions made according to these rules will not subject to judicial review in so far as they are not legally binding. As I mentioned before, our internal (or infra-national) network of competition authorities in Germany has prepared a joint paper concerning our approach to abuses of a dominant position in the electricity sector. It is in our interest to pass this information to the companies, so that in designing their business strategies they can take into account what the competition authorities are likely to do in the future. However, in some cartel cases it makes no sense to make this information public. The information exchanged within the network in relation to such cases will be confidential, but not with respect to the evidence concerning the companies. It is the exchange of views between the competition authorities with respect to the application of the law that will be confidential.

IAN FORRESTER — I fully agree with Dr Böge. What I was instead asking for is that individual decisions about cases, which are made by individual competition authorities and addressed to particular companies, be fully reasoned. And the



322

Panel Discussions

reasoning should draw on the law and its interpretation in view of the facts; it should not simply turn on an internal decision of the network. CÉLINE GAUER — Initially I wanted to make exactly the same point as Dr Böge: the network is not an institution, it is simply a group of people talking to each other. There will be no decisions of the network as such. This is why I think that information exchanges within the network cannot be subject to judicial review; if only for the simple fact that these are informal internal acts. I would just add a comment: he was referring to the German network of competition authorities, which seems to work very well, but Germany is a federal system, whereas the EU is not. For instance, we do not have the ability to challenge national decisions before the Community courts; Article 11(6) of the draft regulation would not be necessary if we had such an option. Indeed, the courts are responsible for maintaining consistency in the application of the law in a federal system but, I repeat, the EU is not one. Therefore we need to create a sort of a path leading to the European Court of Justice, and this is why the Commission has to step in.



䉴 ULF BÖGE — However, let us remember that when the national courts have doubts as to the application of EC law, they can ask the ECJ for a preliminary ruling. So I do not think that the federal system is ultimately indispensable for obtaining a clear and final opinion from the ECJ. In this sense, it is not obligatory that the Commission step in. This is not to say that the Commission should not have the ability to step in, but this should occur only in special cases; it should not be the general rule. I would think that, at least in principle, a final decision of the ECJ gives more certainty to everybody.

ALEXANDER SCHAUB — I fully agree with this view: it is not ‘a must’ for the Commission to intervene in every case, and the Commission will be selective in its interventions in the future, as it has been in the past (the Commission has not acted not acted upon every suspicion of a violation of the Treaty rules in the Member States). By the same token, the Commission will have to decide whether to intervene within the network and try to direct developments towards a correct application of the Community rules, or to rather leave the problem to be settled at the level of the courts. Indeed, there will be many cases in which it will be quite clear that the case will go to court for appeal, regardless of what is happening at the administrative level. I believe that the Commission can reasonably come to the conclusion that it is better if the matter goes directly to the court in such situations, as it would arrive there anyway even if the Commission were to intervene. 䉴

䉴 DAVID GERBER — What comes out clearly in our discussion is that there is a considerable cost involved in running the network. The network will not itself

Technical Problem Areas

323

make final decisions, but there will be numerous decisions made by the individual members and each of these has a cost. Also, so far we have not paid enough attention to the disparate impact of participating in the network on the individual members. I think we should reflect further on the funding of the network, i.e. where the resources will come from. If it is in fact a Community structure, it may be reasonable to think of Community funding for it. JOHN FINGLETON — I would like to make a last point about what we called initially ‘consultation’ within the network. Yesterday I asked, at what point does a consultation actually become a decision? I am still not convinced by the arguments that have been put forward to the effect that the network makes no decisions. On the other hand, I think there are good arguments for believing that consultations are preliminary steps in a decision-making process. The European courts will probably adopt a purposive interpretation of the consultation procedure, bearing in mind that the substantive decisions will ultimately be taken by individual authorities and will then be subject to several layers of judicial review. These elements would allow you to get away with the idea that the network decisions are preliminary, and are therefore not justiciable acts. But I think that these arguments are unconvincing, and it is not the best way for the Commission to proceed with this position. 䉴

䉴 WOUTER WILS — I would like to return to the issue of double jeopardy. Take the example of a European-wide cartel for which fines are imposed by both the French and Italian competition authorities. Can we say there is no problem of double jeopardy because the French authority protects the French consumers and the Italian authority protects the Italian consumers? I think this raises two issues that should not be confused. One is the notion of what is the same offence within Article 4 Protocol 7 ECHR or Article 50 CFR. The other is, what are the rules applicable to the different members of the network — be they Community or national rules — that determine which type of cases are being prosecuted by which authority? The right not to be prosecuted twice for the same offence is a fundamental right of individuals, and I think we should put ourselves in their position. The representatives of the companies I mentioned as an example have met in smoke-filled rooms only once, but they are to be fined twice for a violation of Article 81 EC. There is only one culpable action and only one applicable rule. How can we then say that they are not being prosecuted and punished twice for the same offence? Another matter is that the rules of distribution of work between the Italian and French competition authorities are possibly rules of French and Italian administrative law; these exist to determine when the national authorities can become active. This may lead to the result that the French competition authority will become active only when French consumers are affected by the infringement of the law, and the Italian competition authority only when the Italian consumers are affected. I believe

324

Panel Discussions

that the second matter is irrelevant for judging whether the companies concerned are prosecuted twice for the same offence within the meaning of Article 4 Protocol 7 ECHR and Article 50 CFR. 䉴 MARC VAN DER WOUDE — I wanted to go back to the issue of judicial protection; this is, after all, also a fundamental right. I am very sensitive to what has been said by Ian Forrester, Mario Siragusa, James Venit and Jacques Bourgeois. Yet judicial review is indeed a difficult area, particularly if we are considering preliminary rulings from the ECJ. I must say that I have been a practicing lawyer for 23 years now, and I have only once succeeded in obtaining a preliminary ruling. This was after ten years of legal proceedings, and on a minor issue that had nothing to do with my case. The Treaty of Nice provides for the establishment of jurisdictional chambers at the European court. We are now talking about jurisdictional chambers for trade-mark law and employment cases, and it has already been suggested that perhaps something should also be done in relation to competition law. I wonder if this is not the solution to our problem: an adversarial system as suggested by Barry Hawk, with several prosecutions by national authorities before one jurisdictional chamber. This would be easy to set up, and it would give us the benefit of coherence in decision-making. Furthermore, there would be an independent body to decide cases brought by the competition authorities of the Member States. This would also ensure a fair trial. For example, in such a system you could discuss Visa International’s nondiscrimination rule in a much fairer way than in a preliminary ruling in which the Court will not hear experts or the evidence.

I Céline Gauer* Does the Effectiveness of the EU Network of Competition Authorities Require a Certain Degree of Harmonisation of National Procedures and Sanctions?

I.

Introduction

The proposal for a new Council regulation abolishes the exemption monopoly of the Commission and puts into place a decentralised system in which national authorities and courts will play an essential role in the enforcement of EC competition law. National competition authorities and the Commission will together form the ‘network of competition authorities’ — that is, a network of public enforcers of the same law. The concept covers a number of cooperation and assistance mechanisms between the various members. The network is a key feature of the future enforcement system. It will allow an efficient division of work between its members and a consistent application of Community law. The members of this network already exist, or it could even be said that the network is already in place. It will not be created ex nihilo, following a given model. The members are the existing competition authorities, some of which have a long history behind them. They are the ‘competent authorities’ of Regulation 17/62. They have been created by national legislators. Their structure, their powers, the rules of procedure they are bound to follow, and the sanctions they can impose, express the variety of traditions of national systems. The draft regulation harmonises certain elements and foresees certain powers for national authorities, but it largely leaves the determination of procedures and sanctions to national law. It does not interfere with the structure of the authorities. A lot of comments about the White Paper and the draft regulation relate to the heterogeneity of the members and to the limited harmonisation of procedures. They question the ability of the network to function effectively without a minimum degree of harmonisation, and they call upon the Commission to propose a harmonisation directive based on Article 83 EC. The European Parliament with regard to sanctions, and the Economic and Social Committee more generally, have also insisted on this point. * Administrator, Unit Regulatory and Legislative Projects, DG Competition, European Commission. All opinions expressed in this Article are personal. The author thanks Dorothé Dalheimer, Eddy De Smijter, Lars Kjølbye and Emil Paulis for their contributions to the preparation of this Article.

188

Working Paper I

The purpose of this paper is to examine whether a certain degree of approximation between the different systems is needed for the network to function efficiently, particularly with regard to the structure of the authorities, their procedures and their sanctions.

II.

Structure of the Competition Authorities

Whereas Regulation 17/62 referred to ‘the competent authorities of the Member States’, the draft regulation uses the word ‘competition authorities’ throughout.

1.

The Network will be Composed of Authorities Structured in Different Ways

Public enforcement in Europe is ensured by three main types of bodies. The first type of authority comprises integrated administrative bodies that are in charge of both the enquiries and the decision-making. They enjoy investigative powers and are empowered to adopt decisions for the application of competition law. Examples of this type of authority are the German Bundeskartellamt, the Italian Autoritá Garante and the UK Office of Fair Trading. The second type of competition authority includes administrative authorities, which are split into two different parts: one body, often a division of the Ministry of Economic Affairs, is in charge of the enquiries and another body, generally a quasi-judicial authority, makes the decision. Examples of this model exist in France, Belgium and Spain. The third type of enforcement structure comprises an administrative authority that is in charge of the enquiry and of making certain decisions (generally, positive decisions), and a court to which the authority can bring a case for prohibition of an agreement or practice with or without fines. The dividing line between the second and the third types is not static. The Conseil de la concurrence in Belgium is treated by law as ‘an administrative jurisdiction’. The Conseil de la concurrence in France, although it is composed in the same manner as its Belgian equivalent, is not considered to be a Court but an independent administrative authority. Article 36 of the draft regulation foresees that all Member States designate authorities that are responsible for ensuring the public enforcement of Articles 81 and 82 EC. The draft regulation uses the term ‘competition authorities’ to refer to the authorities to be designated and empowered as public enforcers of the EC competition rules in accordance with Article 36 of the draft regulation. The term ‘competition authority’, as used in the draft regulation, is an autonomous concept of Community law. It applies uniformly in the whole of the Community; its meaning can differ from the meaning conferred on the same words in national law.

Broad Systemic Issues

189

In designating and empowering their authorities, Member States must ensure that the various functions attributed to competition authorities by the Regulation are effectively implemented, otherwise they would be in breach of Community law. Under Article 36 of the draft regulation, Member States are free to divide up the different functions foreseen for competition authorities between different administrative and/or judicial authorities responsible for the application of Articles 81 and 82 EC as long as the effective implementation of the provisions of the new regulation will be ensured.

2.

The Proposal Takes Account of the Variety of Structures of the Competition Authorities

The draft regulation contains a distinction between national courts and national competition authorities, but it does not distinguish between the various types of authorities. Within the meaning of Article 6 of the draft regulation, ‘national courts’ refers to courts deciding on litigation between private parties. It does not encompass national courts deciding about the request of an administrative authority to prohibit an agreement or to impose a fine. The various mechanisms of cooperation set out in the proposal take due account of the existence of several types of structures. One could indeed have feared that certain obligations could not be fulfilled by a court or by a collegiate body in the countries in which they are in charge of the decision-making.

2.1.

Obligations to Inform (Articles 11(3) and (4) of the Draft Regulation)

The ‘obligation to inform’ at the beginning of proceedings pursuant to Article 11(3) in the draft regulation does not raise any specific problem. In all Member States, this obligation will be fulfilled by the body in charge of the investigation. In the Member States in which a court or a collegiate body decides (whereas the competition authority has public prosecution functions), this obligation will be performed by the competition authority. This will always happen before the court or the college proceedings are opened. The question is more delicate for the consultation obligation provided for in Article 11(4) in the draft regulation. This provision states: No later than 30 days before adopting a decision requiring that an infringement be brought to an end, accepting commitments or withdrawing the benefit of a block exemption regulation, the competition authorities of the Member States shall inform the Commission. To that effect, they shall provide the Commission with a summary of the case, the envisaged decision or, in the absence thereof, any other document indicating

190

Working Paper I

the proposed course of action. This information may also be made available to the competition authorities of the other Member States. At the request of the Commission or the competition authority of a Member State, the acting competition authority shall make available other existing documents necessary for the assessment of the case.

This wording allows national authorities to consult at the stage of the statement of objections, that is, before bringing an action before the court or seizing the college. This is sufficient because the statement of objections describes the ultimate scope of the prohibition. For decisions accepting commitments (which currently do not exist in any national system), Member States will be able to define a procedure for their adoption that allows the network to be usefully consulted.

2.2.

Article 11(6) of the Draft Regulation

It has been suggested that Article 11(6) should not apply to national courts when they make decisions in public enforcement cases. The argument goes that this situation should instead be covered by the same rules as those applying to national courts acting in the framework of private litigation — that is, national proceedings should continue with the obligation not to contradict the envisaged decision of the Commission.1 The Court of Justice decided in 1974 in the BRT/SABAM case2 that Article 9(3) of Regulation 17/62 applies to the authorities of the Member States referred to in Article 84 (ex-88) EC, including courts deciding competition cases and courts reviewing the decisions of competition authorities. The rationale of this judgement is a very reasonable jurisdictional rule. It provides companies with a onestop shop whenever the Commission opens proceedings. This makes a lot of sense, as the Commission’s decision will have effect in the whole Community. It is also an instrument to promote consistent application of Community rules. For these two reasons, it would be both inefficient and inequitable in the new system — as in the present one — to exclude from the application of Article 11(6) national proceedings in which the first decision is made by a court. It would be inefficient with regard to the function of Article 11(6) as a means by which to promote coherent application. The preventative nature of the Article 11(6) mechanism would be lost: the Commission would no longer be able to prevent inconsistent application of Community law by opening proceedings itself. The only way for the Commission to bring a problematic case into the Community system would be the adoption of a contradictory decision. The latter course of action could be detrimental both to companies that are parties to the national proceedings, and to the

1 See Case C–344/98 Masterfoods [2000] ECR I–1412; Case C–234/89 Delimitis [1991] ECR I–935; and Case 314/85 Foto-Frost [1987] ECR 419. 2 Case 127/73 Belgische Radio en Televisie et société belge des auteurs, compositeurs et éditeurs ./. SV SABAM et NV Fonior (BRT-I) [1974] ECR 5, points 18 and 19.

Broad Systemic Issues

191

image of the network. It would also be inequitable because integrated authorities would be submitted to a stricter level of monitoring by the Commission than authorities involving a court. It is important to avoid such discrimination between Member States. It has been argued that there would be a conflict between Article 11(6) of the draft regulation and Article 6(1) of the European Convention on Human Rights (ECHR), which protects the right to a fair trial. The argument goes that the mere existence of Article 11(6) would deprive courts deciding competition cases of their independence, and would thereby violate Article 6 ECHR, which stipulates: In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.

I will argue instead that the proposed system does not violate Article 6 ECHR, for three reasons. First, Article 6 ECHR protects the right of any person to a fair trial before an independent tribunal. It does not protect the right of courts to come to a final decision on cases pending before them. When the Commission opens proceedings and relieves national authorities of their competence, it simply shifts the case from the national system to the Community system. But in both systems, a person who is a subject of the proceedings enjoys the right to a fair trial before a tribunal meeting the requirements of Article 6 of the Convention. Secondly, Article 11 of the draft regulation does not impinge on the independence3 of courts deciding competition cases. As explained above, the administrative authority in charge of the inquiries can very well carry out the consultation pursuant to Article 11(4) of the draft regulation. Therefore, the Commission does not interfere with the decision-making process of the Court, and the latter can adopt whatever decision it considers appropriate. This is very different from the situation dealt with by the Court of Human Rights, for example in the Beaumartin case.4 In that case, the French Conseil d’Etat, when facing a difficulty in the interpretation of an international Treaty, was obliged to request clarification of the meaning of the provision from the Minister of Foreign Affairs, and was bound by the interpretation given by the Minister. Finally, even if the mere existence of Article 11(6) were to deprive courts deciding competition cases of their independence in interpreting the meaning of Article 6 ECHR, this would not constitute a breach of the Convention. If the existence of Article 11(6) were to lead to the conclusion that courts deciding

3 Pursuant

to the case-law of the Court of Human Rights, independence is to be assessed with regard inter alia to the manner of appointment of its members and to their term of office, to the existence of guarantees against outside pressures, and to the question of whether the body presents an appearance of independence. 4 Beaumartin v. France, judgment of 24 November 1994, Series A no. 296–B.

192

Working Paper I

competition cases no longer comply with the requirement of Article 6 of the Convention, then the existence of a right of appeal before review courts not subject to Article 11(6) would certainly be sufficient to comply with the requirements of the Convention. The case law5 of the Court of Human Rights recognises the authority of Member States to confer the prosecution and punishment of offences on administrative authorities, provided that the person concerned is empowered to take any decision thus made against him before a tribunal that does offer the guarantees of Article 6 ECHR. Thus, the power of, for example, the Bundeskartellamt or the Autoritá Garante to adopt decisions prohibiting and sanctioning anticompetitive behaviours is compatible with the Convention.

3.

National Competition Authorities Must be Independent and Have the Resources Necessary to Carry Out Their Duties

The draft regulation does not interfere with the designation of national authorities and does not prescribe a given model. But there is an underlying requirement: all these bodies have to be fully independent. Decentralisation implies that national bodies will apply Community competition rules to protect competition within their respective Member States, in pursuit of the general Community interest. The criteria for case allocation will naturally give a predominant responsibility to national authorities for enforcing the rules when the effects of an agreement or practice are mainly in the territory of one Member State. National authorities will therefore have to be fully independent. That means they must be independent from business, and also independent from national influences (ex: foreclosure cases). This is a key issue for the credibility of case allocation. There is no such independence requirement in the draft regulation, and some voices have spoken in favour of its introduction. This would help the authorities in the Member States in which the issue is being debated. They would be able to argue for their independence, basing their claim on an obligation under Community law. On the other hand, independence is very difficult to establish objectively. The law in itself is a poor proxy for independence; some national laws may provide for interference at a political level that never occurs in practice. In general, the new system will help all authorities to acquire more independence. First, they will be applying Community law and will submit to the case-law of the European Court of Justice. Secondly, they will be part of the network, and will find support for their independence in discussions with the other members. Finally, competition authorities will have to have sufficient resources to carry out their duties and to ensure that competition is protected in their respective markets. They will also have to be able to assist other authorities in collecting

5 See

e.g. Öztürk v. Germany, judgment of 21 February 1984, Serie A no. 73, para 56.

Broad Systemic Issues

193

evidence, and to follow their activities. The Commission can always compensate for a lack of enforcement at national level, but in an enlarged Community it will be essential that national authorities play an active role in the enforcement of Community competition law. The network has to be composed of independent authorities that can ensure that competition rules are obeyed within their territories. The system is designed to accommodate the various existing structures: it ensures an efficient allocation of cases and a consistent application of the rules without unduly interfering with the independence of the national authorities.

III.

Procedures before the Competition Authorities

It goes without saying that the members of the network have to be empowered to apply Community law. This is ensured by Article 36 of the draft regulation, and this is already the case in all Member States, except Finland, Ireland, Luxembourg and the UK. In all the Member States except Finland, reforms are being considered to empower the authorities to apply Community law. Rules of procedure will remain national. The draft regulation introduces some elements of harmonisation necessary for the functioning of the system, but some, mainly in industry circles, call for the fully-fledged harmonisation of national procedures. Article 83 EC would provide a sufficient legal basis for such a measure. This provision states that: The appropriate regulations or directives to give effect to the principles set out in Articles 81 and 82 shall be laid down by the Council, acting by a qualified majority on a proposal from the Commission and after consulting the European Parliament.

A directive harmonising national procedures for the application of Community law could be adopted on this basis. However, the question is whether such a far-reaching measure is necessary, or what adverse consequences the absence of harmonisation could have.

1.

Investigative Powers

National authorities need to have sufficient investigative powers to efficiently enforce competition rules within their territories. Some have therefore proposed that the powers given to the Commission in the Council regulation should establish a minimum standard; national authorities should have at least the powers of the Commission (and not only when they assist the Commission, as provided for in Article 20(5) of the draft regulation).

194

Working Paper I

In practice, most of the authorities are empowered to inspect the premises of companies, to collect documents and information, and to ask questions. The powers may differ in different jurisdictions, but their ability to collect information is assured. For example, the German Bundeskartellamt cannot compel companies to answer any question in an administrative procedure leading to fines (Bußgeldverfahren), and the French authority does not have the power to send written requests for information, but both authorities can efficiently collect evidence. There is therefore no immediate need to envisage harmonisation of national laws in this respect. The draft regulation establishes a common rule only in order to allow a national authority to collect information on behalf of another authority. Thus, Article 21(1) states: The competition authority of a Member State may in its own territory carry out any fact-finding measure under its national law on behalf and for the account of the competition authority of another Member State in order to establish whether there has been an infringement of Article 81 or Article 82 of the Treaty. It shall transmit the information collected to the requesting authority in accordance with Article 12 of this Regulation.

This allows for mutual assistance for the collection of evidence. But it is a fact that the degree of this assistance will depend on each national law. The wider the fact-finding powers of an authority, the greater its assistance can be. If the powers of an authority are limited, the extent of its cooperation will also be limited. The absence of harmonisation of the powers of investigation can be regretted but it certainly does not cause insurmountable problems for the functioning of the network provided that all authorities have sufficient fact-finding powers.

2.

Types of Decisions

The proposed regulation abolishes the authorisation system. This shift is mirrored by the types of decisions that it is envisaged the Commission will make. The emphasis is on prohibition decisions and decisions accepting commitments. ‘Positive decisions’ are only envisaged in exceptional circumstances, for reasons of policy setting and consistency. Companies cannot request them. Article 5 of the draft regulation sets out the types of decisions that can be adopted by national competition authorities. It is drafted along the same lines, and reads: The competition authorities of the Member States shall have the power to apply Articles 81 and 82 of the Treaty in individual cases. For this purpose, acting on their own initiative or on a complaint, they may take the following decisions: • requiring that an infringement be brought to an end; • granting interim measures;

Broad Systemic Issues

195

• accepting commitments; • imposing fines, periodic penalty payments or any other penalty provided for in their national law. Where on the basis of the information in their possession the conditions for prohibition are not met they may likewise decide that there are no grounds for action on their part.

This wording excludes Member States from adopting positive decisions with a constitutive effect for the application of Community law. However, the question remains as to whether national authorities can introduce or maintain notification systems for the application of Community competition law. As explained above, they cannot subject the application of Article 81(3) EC to the notification of authorities of an agreement. Article 81(3) EC will be directly applicable and companies will be able to invoke it as a defence. However, they can have a notification system that the companies would be free to use in order to obtain an assessment of the legality of their agreement by a competition authority. Nothing in the draft regulation prevents them from doing so. However, the only decisions that they will be able to adopt are those stating that they have no grounds for action (that is, decisions equivalent to negative clearance decisions). It remains to be seen whether this is attractive enough for companies to notify national competition authorities of their agreements. Furthermore, as it will be a burden for national authorities to deal with these inoffensive cases instead of prosecuting infringements, it is likely that these notification systems will gradually disappear. This is the reason why the draft regulation does not contain any provision harmonising national systems in this respect.

3. 3.1.

Rights of Companies Rights of complainants

There is also no harmonisation as far as the rights of companies are concerned. Some have argued that this could be problematic in the field of complaints, rights of defence and protection of confidential information. Some systems provide for extensive rights for complainants. In the Community system, complainants have a right to a decision on their application, and they can challenge this decision before the Community Courts. In other systems, complaints can be rejected by a simple letter or even through informal contact with the complainant. It has been argued that this discrepancy entails the risk of a certain degree of forum shopping on the part of complainants. They might tend to lodge their complaint with the authority that provides the most favourable treatment. This risk would be particularly high for the Commission, as its action

196

Working Paper I

can be a substitute for national action and its procedure is very protective of complainants. A solution would be to harmonise national and Community procedures for the treatment of complaints. However, such a far-reaching measure is not necessary at this stage. The real interest of a complainant is not to be able to challenge a decision rejecting its complaint. Its aim is rather to bring about a hasty end to the competition problem from which it is suffering. The complainant’s interest is therefore to lodge its complaint with the authority that is best placed to deal with it. Article 13 of the draft regulation allows an authority to reject a complaint on the grounds that another authority is dealing with the case. This will considerably simplify the task of the authorities that have to reject complaints by a formal decision. Furthermore, the Commission has already announced in the White Paper its intention to publish a notice elaborating on the notion of ‘sufficient Community interest’. Paragraph 119 of the White Paper reads: The Commission should improve the information available to potential complainants regarding its likely course of action. The case law accepts that the Commission is entitled to lay down priorities governing the action it takes against infringements, on the basis of the concept of sufficient Community interest.6 It now seems advisable to publish a notice clarifying this concept, so that complainants can more easily determine whether they would be better advised to address their complaint to the national or to the Community authority. This notice could thus guide complainants’ choice of the forum in which to pursue their complaint.

This notice, combined with the rigorous practice of the Commission, could be sufficient to ensure that complaints are directed to the authority best able to solve the competition problem at hand.

3.2.

Rights of defence

The rights of defence are also different in the various national systems. The time defendants are granted to answer a statement of objections, or the extent to which they have a right to be heard, can differ. However, the European Convention of Human Rights binds all Member States, and the Commission is bound by the case-law of the Court of Justice. This has consistently held7 that: Fundamental rights form an integral part of the general principles of Community law whose observance the Court ensures. For that purpose, the Court draws inspiration from the constitutional traditions common to the Member States and from the guidelines

6 Case T–24/90 Automec v. Commission [1992] ECR II–2223. 7 See, in particular, Opinion 2/94 [1996] ECR I–1759, para

33; Case C–260/89 ERT [1991] ECR I–2925, para 41; Case C-299/95 Friedrich Kremzow v. Republik Österreich [1997] ECR I–2629.

Broad Systemic Issues

197

supplied by international treaties for the protection of human rights on which the Member States have collaborated or of which they are signatories. The Convention has special significance in that respect.

There is thus a minimum common standard. A further harmonisation of national procedures does not seem necessary at this stage and was therefore not proposed by the Commission.

3.3.

The protection of confidentiality

Finally, the notion of confidentiality might not be the same in all Member States. The provision in the draft regulation on exchange of confidential information requires that all national authorities protect confidentiality in an adequate manner. To that effect, Article 27(2) creates an obligation on all members of the network not to disclose exchanged information: Without prejudice to Articles 11, 12, 14, 15 and 26, the Commission and the authorities of the Member States, their officials and other servants shall not disclose information acquired or exchanged by them pursuant to this Regulation and of the kind covered by the obligation of professional secrecy. This obligation also applies to all representatives of Member States attending meetings of the Advisory Committee pursuant to Article 14.

This establishes a Community obligation for all information that is collected by the Commission or exchanged via the network, and that is considered to be covered by ‘professional secrecy’. Professional secrecy is a Community concept and the level of protection will therefore be the same throughout the Community. A greater homogeneity of national and Community procedures does not seem to be a prerequisite for the network to function efficiently. The instruments on human rights and the case-law of the Court of Justice provide a de facto minimum standard of harmonisation for the rights of companies. A minimal harmonisation could be envisaged in the field of investigative powers, but the current state of the law in the Member States shows that it already exists.

IV. 1.

Sanctions Sanctions Applied are Different in Nature and in Amount

The sanctions imposed for violations of competition law vary from one Member State to another. First, they can be of a different nature: in Community law, they are considered to be of an administrative nature. They are considered to be administrative in nature in Austria (except for collusive tendering), Belgium,

198

Working Paper I

Germany (except for collusive tendering), Spain, Italy, the Netherlands, Portugal, Finland, Sweden, and the UK. On the other hand, they have a criminal nature in Denmark, Ireland and Luxembourg. In the UK, a reform is being considered to introduce criminal sanctions for serious breaches of competition law. A combination of both types of sanctions exists in Greece and France. There are three main types of sanctions for violations of competition law in the Community: • fines on companies; • fines on individuals; • imprisonment. In all Member States,8 violations of competition law can be sanctioned by fines imposed on companies. In a majority of Member States, the law sets a ceiling on the percentage of the turnover. However, this ceiling is sometimes to be calculated on the basis of the national market (as in Belgium or the UK), sometimes it relates to the relevant product (as in Italy), and sometimes it does not (as in Ireland). In other Member States, an absolute figure is set by law, ranging from around 550.000m in Sweden to 2.320.000m in Ireland. In Germany, the ceiling is calculated as three times the additional revenue made as a result of the infringement. These maximum amounts do not give any indication of the actual fining practice of authorities. Fines can also be imposed on individuals in eight Member States (Germany, Denmark, Greece, Spain, France, Ireland, Luxembourg and Austria). In five Member States (Germany and Austria for bid rigging; France, Ireland, Luxembourg), imprisonment can be ordered for breaches of competition law. It is a fact that these sanctions are very rarely imposed. These discrepancies have given rise to several comments and calls for harmonisation. In its resolution on the proposal for a new regulation, the European Parliament adopted several amendments in order to ensure that, when applying Community law, Member States impose the same sanctions as the ones provided by the Council regulation. The Economic and Social Committee has also spoken in favour of a certain degree of harmonisation of sanctions.

2.

The Possible Impact on the Functioning of the Network

The differences in the sanctions could create two types of problems: one for the exchange of evidence, and another regarding the efficiency of leniency programmes. 8 In Luxembourg, the system is an abuse control system. No sanction can be imposed for the breach of

substantive provisions. Sanctions can only be imposed for breaches of decisions that have found an infringement.

Broad Systemic Issues

199

For the same infringement of the same rules, companies and individuals will face different sanctions in different jurisdictions. This could be perceived as certain discrimination. However, this will not be very different from the current system. Today, when they conclude restrictive agreements, companies also face different sanctions depending on which Member State brings the case. But the application of the same law to these infringements and the cooperation between authorities for the collection of evidence makes this discrepancy more apparent. Where a case is reallocated from one authority to another, there could be consequences for the parties as to the sanction encountered. There is generally a link between rights of defence and possible sanctions. The more stringent the sanctions are, the more the law has to protect the rights of a defendant. In particular, rights of individuals are generally better protected than rights of companies, especially when there is a risk of imprisonment. It would be unfair to collect information on the basis of a law conceived for companies and to rely on that information to imprison individuals. In order to solve this problem, the draft regulation introduces a safeguard in Article 12(2): Information provided under paragraph 1 may be used only for the purpose of applying Community competition law. Only financial penalties may be imposed on the basis of information provided.

The functioning of the network will also have to take account of different leniency programmes. Pursuant to Article 12 of the draft regulation, information can be exchanged between competition authorities. If appropriate, cases can be reallocated from one authority to another. No exception is provided for information brought to the attention of an authority by a leniency applicant. This means that leniency applicants will only come forward if they would do so under the least favourable leniency programme available. This could be detrimental to the efficiency of well-designed leniency programmes. On the other hand, to exclude the reallocation of cases for leniency applications would be to invite forum shopping. In systems in which personal sanctions exist, leniency programmes have to cover both corporate sanctions and personal sanctions. Companies would only apply for leniency after the managers have secured their own situation. A decision addressed to the company could subsequently be used against them personally. Without harmonisation, companies will only approach authorities having an administrative system if they have been able to obtain a guarantee of leniency from systems with personal sanctions. The efficiency of all programmes will depend on the level of certainty offered by leniency programmes in the countries that apply sanctions to individuals.

200

3.

Working Paper I

A Minimum Level of Harmonisation Might be Required in the Medium-Long Term

A proposal to harmonise sanctions for violations of competition rules might have to be considered. Three main options could be envisaged: • only fine companies, in conformity with the sanctions that can be imposed by the Commission. (This is the proposal made by the European Parliament in its resolution.) • provide for fines on both companies and individuals at national and Community levels; • provide for fines on companies and individuals, and for imprisonment of individuals, at both national and Community levels. Under the first option, the Commission and the national competition authorities would be empowered to fine companies up to 10 per cent of their total turnover. In addition, in order to ensure that the rule is applied consistently, guidelines applicable to all authorities could be adopted. This would be desirable for reasons of equity, and it does not seem to raise insurmountable difficulties. But this would lead a number of Member States to increase the fines currently provided for by their national laws, and it would deprive Member States that currently apply individual sanctions of the authority to apply them. The latter is not necessarily desirable. Individual sanctions are an effective deterrent. Managers are more careful to avoid breaking the law if they are at risk of receiving a penalty themselves. The second option would only exclude imprisonment sanctions. Personal fines could be created at the Community level. Article 83(2)(a) EC provides a sufficient legal basis to introduce personal fines. However, they would deeply modify the nature of the Community system and would probably have an impact on the rights of defence in competition procedures. If individual sanctions were to be imposed, there would be a risk that the powers of the Commission to request information and ask questions would be limited by a more extensive application of the right not to incriminate oneself. One basis of the Community system, which is the obligation for companies to cooperate in the Commission’s procedure, could be put in question. In this context, it is interesting to note that companies cannot invoke the Fifth Amendment of the Constitution in the United States, which allows an accused person to remain silent. A balance will therefore have to be struck between the deterrent effect of individual sanctions and the risk that they would impinge on the Commission’s investigative powers. In this option, individual sanctions could remain, as they are at present, a complement to sanctions on companies. To be workable, this option requires that Member States with such a system create a very efficient leniency program for individuals. This would preserve the efficiency of enforcement throughout the Community.

Broad Systemic Issues

201

The third option would be not only to harmonise fines that can be imposed on individuals and companies, but also to introduce imprisonment sanctions in every Member State and at the Community level. This would most probably require a modification of the Treaty9 and it is not certain that all Member States would regard violations of competition rules to be serious enough to justify imprisonment.

V.

Conclusion

A certain degree of harmonisation of sanctions and procedures will probably have to be envisaged in the medium to long-term perspective, but it is not a precondition for the adoption of the present reform and the effective functioning of the network. The Commission’s proposal takes account of the variety of the structures of competition authorities, and it makes sure that cooperation mechanisms can work efficiently in all systems. It also includes the necessary elements for cooperation between authorities. As far as sanctions are concerned, it should not be forgotten that the proposal does not change the current situation. In the current system, companies face different competition laws applied without any coordination by national authorities pursuant to different procedures leading to different sanctions. The creation of a level playing field as to the substance by Article 3 of the draft regulation makes the divergence more obvious, but does not create it. The system can work as it stands in the draft regulation. However, this does not exclude the possibility that the Commission will study the opportunity to make proposals in order to introduce a certain degree of harmonisation in the field of procedures and sanctions.

9 Article

83(2)(a) EC only mentions fines as a means to ensure compliance with EC rules. It could however be argued that Article 83(1) EC, combined with Article 307 EC (ex Article 234 EEC), is a sufficient legal basis.

II Frédéric Jenny* Does the Effectiveness of the EU Network of Competition Authorities Depend on a Certain Degree of Homogeneity within its Membership (With Respect to Status, Structure, Powers, Responsibilities, etc.)?

I particularly liked Dr Ulf Böge’s written contribution for this Workshop, ‘The Commission’s Position within the Network: The Perspective of the NCAs’ (Working Paper VI). I agree with most of his remarks. In my own intervention, I would like to address briefly the following three issues: • What is a network? • Voluntary cooperation versus commitment to cooperate between the NCAs. • Is harmonisation a prerequisite for the establishment of the EC network of competition authorities?

I.

What is a Network?

The most important feature of a network is that its output is greater, in quantitative and qualitative terms, than the sum of the outputs of its various components. In other words, the concept of network is intimately linked to the idea of externality among the network members. In the area of high technology, for example, ‘network effects arise when the value of the products to consumers increases with the number of buyers both because of demand side economies of scale and because complementary products will be better and cheaper the larger the number of buyers of the basic product’. If we think about a retail network (such as a selective distribution system), what makes it valuable to the manufacturer is the fact that the sales efforts of the retailers will complement its own sales efforts, thus leading to an outward shift of the demand curve for the product which the manufacturer could not obtain on his own.

* Vice-Chair, Conseil de la Concurrence, France.

204

Working Paper II

From this standpoint, it seems that the Commission’s proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 EC (hereafter ‘the draft Regulation’) is not really a proposal to create a network composed of itself and the NCAs. Indeed, the architecture of the network proposed by the Commission does not allow for many interactions between NCAs and the Commission itself. The Explanatory Memorandum of the draft Regulation states that ‘the network will provide an infrastructure for mutual exchange of information, including confidential information, and assistance, thereby expanding considerably the scope for each member of the network to enforce Articles 81 and 82 effectively’. This definition of the network emphasises the decentralisation of enforcement but does not mention the qualitative contribution that the NCAs can make to the enforcement of EU competition law. The Explanatory Memorandum continues: ‘The second way in which the proposal will increase the protection of competition is by allowing the Commission to concentrate on the detection of the most serious infringements’. Thus, NCAs will make a contribution, not by increasing the quality of the enforcement of existing EU law or taking up important cases, but by lightening the Commission’s load. It would be an indirect contribution at best. The mechanism proposed by the Commission would severely restrict the autonomy of NCAs to enforce Articles 81 and 82 EC. While it is natural that the Commission retains some power, one must draw a distinction between two ways in which the Commission could ensure that the enforcement of EU law is consistent and predictable. The first way would be by having clear rules concerning the allocation of cases between the Commission and NCAs, and concerning cooperation mechanisms within the network. This solution would vest NCAs with the role of enforcing EU law in a meaningful way. The second way would be to put NCAs under the constant supervision of the Commission when enforcing EU law. Under this second system, the Commission would reserve the right to interfere with national proceedings once the initial allocation of the case had been made. In other words, the allocation of cases would depend on what the NCAs thought about the substance of cases. Article 3 of the draft Regulation states that, where EU law applies (that is, where a concerted action or abuse of dominant position affects trade between Member States), national law shall not apply. If such a case is referred to a NCA — and assuming that it is not simultaneously referred to another NCA — the latter must, according to Article 11(3) of the draft Regulation, inform the Commission at the outset of the proceedings. Then, according to Article 11(4) of the draft Regulation, if the NCA intends to adopt a decision requiring that an infringement be brought to an end, accepting commitments or withdrawing the benefit of a block exemption, it must first consult with the Commission (at least one month before it plans to adopt the decision). Pursuant to Article 14(6) of the draft Regulation, the Commission may on its own initiative put the case dealt with by the NCA on the agenda of the Advisory Committee for discussion before the final NCA decision is adopted. Finally, the Commission can withdraw the

Broad Systemic Issues

205

case from the NCA at any time by initiating its own proceedings in accordance with Article 11(6) of the draft Regulation. This mechanism is not a meaningful, ‘cooperative’ way for the Commission and the NCAs to interact with each other. It is contrary to Article 11(1) in the draft Regulation, which states that ‘the Commission and the competition authorities of the Member States shall apply the Community rules in close cooperation’. Rather, it is a decentralisation mechanism whereby NCAs are invited to handle Artsicle 81 and 82 EC cases (either those that have been referred to them or those that they have taken up on their own initiative) in strict accordance with the Commission’s view of these cases and under the constant threat of losing the cases if they diverge from the Commission’s view. The situation just described does not leave any degree of freedom for NCAs to influence the substantial interpretation of EC law or the interpretation of facts in specific cases. It is unsatisfactory for NCAs. But it is justified by the Commission desire to ensure a consistent application of EC law, which in itself is a reasonable goal. The question thus arises whether any better method might exist by which to reconcile the search for legal certainty in individual cases and the active participation of NCAs in the interpretation of EU law. Dr Böge has suggested that the allocation of cases between the Commission and the NCAs — and the possible use of the Commission’s power to take over a case in accordance with Article 11(6) of the draft Regulation — should be decided within the first two or three months of the proceedings at the national level. Subsequently, the Advisory Committee should be the body to decide whether or not the Commission could take a case over from an NCA. The merit of this proposal is that it gives a substantial role to NCAs as represented in the Advisory Committee. However, the way in which NCAs would intervene and contribute to the interpretation of EC law would be by taking a stand on procedural questions. What is more, such procedural questions would arise only with regard to cases handled by NCAs (and in the case of a conflict between an NCA and the Commission). Thus the NCAs’ role in the elaboration of EC law would remain quite limited, and it is not certain that the impression of undue interference of the Commission in national proceedings would be eliminated. Some Commission interference in national proceedings is inevitable if we want to limit the risk of inconsistent interpretation of EC law by NCAs. But such interference might be more acceptable if the relationship between the NCAs and the Commission in the interpretation of EC law were more balanced than it would be under the scheme that is proposed by the Commission, and if NCAs had the impression that they could usefully contribute to the Commission’s proceedings on possible infringements of Articles 81 or 82 EC. Another proposal, which could usefully be combined with Dr. Böge’s proposal, would be to modify Article 14 of the draft Regulation by giving the Advisory Committee a more important role to in the elaboration of Commission decisions on Articles 81 and 82 EC cases.

206

Working Paper II

The Commission’s proposal regarding the Advisory Committee is contained in Article 14(5) of the draft Regulation, which states: The opinion of the Advisory Committee shall be delivered in writing and appended to the draft decision. The Advisory Committee may recommend publication of the opinion. The Commission may carry out such publication. The decision to publish shall take account of the legitimate interest of undertakings in the protection of their business secrets. (emphasis added)

This proposal does not allow NCAs to meaningfully contribute to the elaboration of EC case-law and does not provide for the establishment of genuine cooperation between the Commission and the NCAs. Two amendments to this Article would go a long way toward improving the situation. First, Article 14 could provide for the automatic publication of the (reasoned) opinions of the Advisory Committee with the Commission’s decisions. Secondly, the Advisory Committee’s opinions could be binding on the Commission in certain circumstances (for example, whenever they result from a qualified majority vote). If such amendments were adopted, the NCAs would feel that their cooperation with the Commission in the enforcement of EC law was meaningful and that the attempt by the Commission to ensure the consistency of individual decisions at the national level was justified. The amendments therefore seem to be fully consistent with the goal of the Commission as stated in the Explanatory Memorandum: ‘It is a core element of the Commission’s proposal that the Commission and the national competition authorities should form a network and work closely together in the application of Articles 81 and 82’.

II.

Voluntary Cooperation Versus a Commitment to Cooperate

The second issue I would like to address concerns the degree of cooperation within the network. The rules to be applied for the allocation of cases are not clearly established in the draft Regulation. The Commission merely states: ‘The network will also ensure an efficient allocation of cases based on the principle that cases should be dealt with by the best placed authority’, and adheres to the principle of ‘one-stop shop’. Whatever the system of allocation of cases, it is clear that if the best-placed authority is an NCA, it might need information located outside its area of territorial jurisdiction. For example, in a case of possible infringement of Article 82 EC, the best placed (national) authority might need information about the structure of the European market in order to decide whether or not the firm that engaged in the practices under examination has a dominant position. The acquisition of such information might require investigations in other Member States. Similar investigations about the structure of the European market may be

Broad Systemic Issues

207

necessary to decide whether or not there is a presumption of legality with regard to a selective distribution system. In the case of a transnational cartel involving firms located in two Member States, the NCA in charge of the case might also need help to find elements establishing that participants located in the territory of the other Member State have conspired with domestic firms to restrict competition in its national market. The Commission states in the Explanatory Memorandum: ‘The network will provide an infrastructure for mutual exchange of information, including confidential information, and assistance, thereby expanding considerably the scope for each member of the network to enforce Articles 81 and 82 effectively’ (p 6) (emphasis added). And: The scope for effective intervention at the national level is substantially increased by the cooperation mechanisms contained in Articles 12 and 21 of the proposed regulation which empower national competition authorities to exchange confidential information and to assist each other in respect of fact-finding. As a result of market integration evidence and information will increasingly be located in several Member States. Enhanced horizontal cooperation will make it easier for national competition authorities to obtain all the relevant facts’ (p 11) (emphasis added).

Article 12(1) of the draft Regulation states: ‘Notwithstanding any national provision to the contrary, the Commission and the competition authorities of the Member States may provide one another with and use in evidence any matter of fact or of law, including confidential information’ (emphasis added). Article 21(1) of the draft Regulation states: The competition authority of a Member State may in its own territory carry out any fact-finding measure under its national law on behalf and for the account of the competition authority of another Member State in order to establish whether there has been an infringement of Article 81 or Article 82 of the Treaty. It shall transmit the information collected to the requesting authority in accordance with Article 12 of this regulation (emphasis added).

However, para. 2 of the same Article 21 states: at the request of the Commission, the competition authorities of the Member States shall undertake the inspections which the Commission considers to be necessary under Article 20 (1) or which it has ordered by decision pursuant to Article 20(4) (emphasis added).

These provisions create a basis for cooperation between the NCAs through exchange of information or through the carrying out of fact-finding measures by an NCA at the request of another, but they do not commit either the Commission

208

Working Paper II

or any NCA to answer positively to a request for information or investigation by another NCA dealing with an Articles 81 or 82 EC case. To the extent that ‘the network will ensure an efficient allocation of cases on the principle that cases should be dealt with by the best placed authority’, this means that a NCA may choose not to cooperate if it is requested to help ‘the best placed authority’ deal with a case. It is questionable whether a satisfactory level of spontaneous cooperation is likely to emerge between NCAs if there is no obligation for them to engage in such cooperation. By definition, Articles 81 and 82 EC apply to cases where trade between the Member States is likely to be affected. It is thus possible that some NCAs will be reluctant to provide information to other NCAs or to investigate on their behalf if there is a possibility that such cooperation will lead to decisions that are contrary to the interests of local firms. This is particularly so if the objectionable practices have most of their effects in the country of the requesting NCA but very few effects in the country of the requested NCA. Needless to say, the requested NCAs could hide their true motivations by invoking their limited resources and by exercising their prosecutorial discretion in the use of these resources to decline the requests. According to Article 21(2) of the draft regulation (as quoted previously), NCAs cannot decline to cooperate with the Commission when the latter has ordered inspections under Article 20(1) or (4). It thus would be advisable to ensure that NCAs must also cooperate with each other.

III. Is Harmonisation a Prerequisite for the Establishment of the Network and for its Effectiveness? Differences exist between the national competition regimes in Europe. These include the degree of independence of the NCAs, their powers and their level of resources, the type and level of sanctions that can be imposed against anticompetitive practices, and the ways in which the rights of defendants are protected. Also, some NCAs are more experienced than others. It is clear that cooperation between NCAs, as well as the cooperation between NCAs and the Commission, would be made easier if there were a higher degree of procedural and substantive convergence among the domestic laws and competition regimes. The lack of harmonisation among national competition regimes and institutions is precisely the reason for which the Commission proposal would make national law inapplicable to cases that fall under the EU law. It is also the reason for which the Commission proposes to limit the NCAs’ enforcement of EC law. Yet, as I also argued, the Commission is not entirely consistent in its approach: it does not provide a mechanism to ensure that cooperation among NCAs will take place when requested by one of them. Whereas there is less risk

Broad Systemic Issues

209

that independent NCAs will refuse to cooperate with other NCAs for strategic reasons, NCAs that are divisions of a ministry may be under more domestic pressure to refuse to cooperate in cases that could prove detrimental to the firms of their country. Beyond that, the major problem raised by the heterogeneity of national competition law systems is that of the protection of the rights of defendants. The business community often raises a question concerning the allocation of cases among NCAs within the network (bearing in mind that some can impose stiffer penalties than others) and the extent to which defendants will have a say in this allocation. Another question is whether the NCAs whose countries have given them limited powers of investigations (or stricter standards for the protection of rights of defendants) will use requests for cooperation from NCAS that have more extensive powers of investigation (or lower standards for the protection of the rights of the defendants) to accumulate evidence for their proceedings. The Commission’s approach raises at least three issues. First, it is very vague on the crucial issue of allocation of cases between the members of the network that has important implications both for the functioning of the network and for the exercise of the rights of defendants. Secondly, the EU Commission’s proposal will be difficult for some Member States to accept if they already have a fully developed competition law regime with adequate resources, vast experience, and so on. The NCAs of these countries, which are the most likely to contribute usefully to the decentralisation process of the enforcement of EC law, are also the ones that are most likely to resent the heavy-handed intrusion of the Commission in their proceedings and the modest role of NCAs in the network. This is a potentially crucial problem for the success of the network, as has been pointed out by many contributors to this Workshop (particularly by David Gerber, who states: ‘The proposal will create extensive routine obligations to transmit, store and review information, and there is a risk that the weight of bureaucratic routine will overshadow and even blunt the necessary commitment’). There are at least two ways in which this problem could be alleviated. One would be to create a two-tier system in which the more experienced and efficient NCAs would have a greater say (or responsibility) in the functioning of the network. However, this approach has two serious flaws. First, differentiating between NCAs on the basis of their experience or efficiency is sure to create political controversy. In addition, such a proposal is essentially static and does not include a mechanism for the convergence of national competition regimes. The second approach would be to include in the new Regulation the principle that a review of the concept and the functioning of the network will take place, perhaps after ten years, in view of progress made in the enforcement of EC law in each country, the experience acquired in the functioning of the network, and the possible convergence of national competition law systems. Thirdly, by limiting the scope for cooperation between the Commission and the

210

Working Paper II

NCAs and the cooperation between the NCAs themselves, the EC proposal may slow the process of harmonisation of competition law regimes in Europe (at least if one believes that cooperation is a factor that is favourable to harmonisation and the convergence of national competition regimes). Thus, a network in which a greater role would be assigned to NCAs would contribute to hastening the establishment of a level playing field in Europe and the emergence of the unique market.

III Laurence Idot* A Necessary Step Towards Common Procedural Standards of Implementation for Articles 81 and 82 EC Within the Network

Back when the Modernisation White Paper was published in May 1999, the absence of any debate over the complex issue of the diversity of national competition authorities seemed surprising. Though the Commission’s position — giving priority to the adoption of the reform — may be understood and shared, the Social and Economic Committee’s and European Parliament’s focus on the need to think about a minimum harmonisation of procedures is equally valid. Almost three years later, we are asked to adopt a position as to whether the effectiveness of the EU network of competition authorities will depend on a certain degree of homogeneity among its membership with respect to status, structure, power (responsibilities), and so on. Although I was requested to approach this subject from a practitioner’s point of view, mine will mainly be an academic perception based on limited practice. The topic is huge and twofold. It can be dealt with from an institutional and structural perspective (I) or from a strictly procedural perspective (II).

I.

The Institutional Perspective

The structural and institutional perspective is of the greatest interest. As has already been noted, there is a kind of consensus in Europe about referring cases to administrative bodies for the enforcement of competition rules, rather than to ordinary courts. This contrasts with the United States antitrust enforcement system. However, the European administrative authorities vary considerably from one country to another. The purpose of this paper is not to make a thorough comparative study of the different choices made by the EC Member States in this respect(1), but to express what is supposed to be the ‘practitioner’s point of view’ on the issue (2).

1.

The choices of the Member States

There is a strong tendency to limit the study to the first level (what the competition authority is), but the second level (who reviews the decisions of the competition authorities) is also very important to the legal practitioner. * Professor

at the University of Paris I-Panthéon Sorbonne.

212

1.1.

Working Paper III

First Level

At the first level, the issue in most Member States is not to choose between the ordinary court and an administrative body, but to choose between the traditional governmental department (in most Member States, this is the Ministry of the Economy or equivalent) and an ‘independent administrative authority’. The national classification of these bodies is of no practical consequence as they all have decision-making powers; for the application of EC law, and mainly for the use of the preliminary ruling procedure under Article 234 EC, they are considered to be courts or tribunals. Another — complicated — issue is to determine if the tribunals are independent and impartial within the meaning of Article 6 ECHR and Article 47 CFR. Without going further into that discussion, it is important to note that the French Cour de Cassation adopts a very strict interpretation of Article 6 ECHR, which is noticeably stricter than the position adopted by the CFI and the ECJ. Returning to the essential choice — independent body or traditional governmental department — we observe that there are two main models in Europe. Some of the EC Member States (mainly Germany, the United Kingdom and, more recently, Italy) have set up ‘integrated’ administrative bodies that, like the Commission, are often responsible for both antitrust enforcement matters and merger control. With respect to antitrust enforcement, there is an overlapping of functions: investigation, examination of the cases, and decision-making. In other EC Member States, it was very difficult to give full enforcement powers in this area of law to an administrative body (whether an administrative court or not) for historical, cultural and other reasons. Thus, the competition authorities of the latter Member States have to share enforcement competencies with government departments even in antitrust matters, which are politically less delicate than merger control. The Belgian, French and Spanish are the best examples in this sense. This model has also been chosen by the EC Member States that recently modernised their competition legislation, such as the Nordic states. There are, of course, some differences between the models.

1.2.

Second Level

At the second level there is greater homogeneity because the principle of jurisdictional control is generally accepted. From a structural and procedural perspective, two issues must be dealt with. The first one is to determine which entity should exercise the judicial control: should it be a specialised judiciary body, like the Appeals Tribunal of the Competition Commission in the UK, or an ordinary court? If the second solution is chosen — as is generally the case in Europe — those Member States with both administrative and ordinary courts must decide between the two institutions. France and Belgium, for instance, have decided to give competence in this area to ordinary courts (Court of Appeal of Paris or Court of Appeal

Broad Systemic Issues

213

of Bruxelles) but the administrative courts retain their natural jurisdiction in most Member States because an administrative entity makes the decisions at the first level. The second issue is to decide the scope of the judicial control: full review, or only annulment review, as under Article 230 EC.

2.

A Practitioner’s Point of View

The institutional and structural differences discussed above are significant from the perspective of the business community. For the legal practitioner, these concerns are mainly of practical. The main issue is efficiency (2.1); the issue of the competition authority’s independence is only secondary in the current European context (2.2).

2.1.

Efficiency

Efficiency, or effectiveness, of enforcement is the main objective. It goes without saying that this is especially the case when companies are the complainants, but it is also true when they are defendants. In the business world, we need to know what is happening. Uncertainty and delays give rise to suspicion. It is for this reason, more than the institutional structure, that the business community is concerned about the resources that are allocated to the competition authority. It is sufficient to compare the resources of the German Bundeskartellamt, the Italian Autoritá Garante della Consorrenza e del Mercato, or the UK Office of Fair Trading, on one side, with those of the Belgian or the French Competition Councils on the other, to understand the importance of the issue. In France, the Competition Council has notably fewer resources than sectoral regulators such as the ART (Autorité de Régulation des Télécommunications — telecommunications regulator) or even the CRE (Commission de Régulation de l’Energie — electricity regulator). The Competition Council has just four permanent members out of a total of seventeen. In addition, it has only 38 rapporteurs — most of them have been seconded from the specialised governmental department (the Competition, Consumer and Fraud Division — DGCCRF) to examine cases at the Council. If we compare the human resources and budgets of these different authorities, it will give us a good idea of how the national authorities work and are perceived by the firms. The study published every year by the Global Competition Review is a very good illustration in this sense.

2.2.

Independence

The above is not to say that legal practitioners are indifferent to the issue of the national competition authority’s independence. However, since national laws in

214

Working Paper III

Europe guarantees independence of action, the issue of institutional independence is of secondary importance regardless of which enforcement model is chosen. The issue is mainly structural in the Member States where enforcement competencies are divided between the departments of the Ministry of the Economy and an independent body. It will be necessary to decide who will lead the network. The European Parliament is aware of this difficulty and has therefore proposed an amendment to the draft Regulation, according to which it should be the authority that has decision-making power to be represented on the Consultative Committee. This amendment has to be approved even if it will lead to significant changes in the enforcement framework of some Member States, such as France. Having said this, more than to institutional independence stricto sensu, practitioners and companies will be especially attentive to the fairness of the proceedings. In conclusion, on this first point everybody agrees that the diversity cannot be eliminated, as it is linked to the history of each Member State, and to the perception that each State has of what we now call the ‘competition culture’. The resources allocated to the competition body, and even the rank of its President in the State protocol, are good indicators of the ranking of freedom of competition in the hierarchy of national values. Nobody denies that it would not be realistic to go further in the future Regulation than has already been provided for in Article 36 of the draft, which asks the Member States to designate a national authority to enforce both national and EC laws. This is merely the application of a general principle of EC law, namely institutional and procedural autonomy. However, it is precisely because it is not possible to go any further on the institutional and structural levels that it becomes necessary to think about a minimum harmonisation or approximation of national procedural laws with respect to antitrust matters. The Commission nevertheless asserts the contrary: ‘It is not necessary for the implementation of the reform to embark on full-scale harmonisation of national procedural laws’.

II.

The Procedural Perspective

‘Not necessary to embark on full-scale harmonisation’: it is possible to agree with this assertion, but between full-scale harmonisation and nothing, there is some room for a midway solution. This solution can be found by identifying the issues, following the patterns of the procedure, and distinguishing between three stages: initiation, development and the end of proceedings. However, it appears that most questions arising in relation to what I call ‘the development of proceedings’ (that is, the investigation stage) are not very problematic. The constraints that derive from the imperious need to respect the rights of the defendant are such that the solutions are very similar. There is always a

Broad Systemic Issues

215

document equivalent to the statement of objections, in which the facts and the relevant law are presented. The defendants always have the right to be heard, even if this right takes different forms (one phase, under EC law; two phases under French law), and they consequently have access to the file. The variations, if any, are not of such a nature to impede the correct functioning of the future network. The main problem is confidentiality (one of the subjects discussed in Session III of this Workshop). Difficulties become acute when firms are confronted with issues that are on the borderline between strictly procedural aspects and substantive legal ones. This line may be difficult to draw in the first (1) and last (2) stages.

1.

Initiation of the Proceedings

During the initial stage of the proceedings, the first issue to clarify is who is entitled to submit a case to the national competition authority. The solutions are generally similar: like the Commission, national competition authorities may initiate proceedings ex officio; or if a case is brought to the attention of the authority by another public body (for instance, the Ministry of the Economy, which is the rule when the competition authority is not fully integrated); or if firms submit a complaint. Here the differences are not essential, and therefore they should not affect the functioning of the future network. From the companies’ standpoint, the main issues arise with respect to the competition authorities’ right to choose among the cases brought before them according to their ability to deal effectively with the case (1.1), which in turn is linked to their investigative powers (1.2).

1.1.

Selection of Cases

Two questions need to be answered in this respect: • Is the competition authority entitled to examine all cases in which there is an alleged breach of competition law? • Does the competition authority have the right not to examine all the cases that are submitted to it? The answer to the first question is negative, because a general principle of law concerning legal certainty imposes certain limitations on enforcement action. This issue was debated in the context of the application of EC law at the beginning of the 1970s. Following a period during which contradictory positions were sustained, the ECJ eventually ruled that the general principle of legal certainty requires the application of a time-limit with respect to bringing actions involving the application of EC competition rules. Regulation 2988/74 was adopted

216

Working Paper III

subsequent to this case-law, and it lays down two limitation periods: the first in proceedings, the second for the enforcement of sanctions. The Commission suggests transposing these rules into the new Regulation, with a few changes, due mostly to the existence of the network. Thus, action taken by a national authority could re-start the period of time within which action must be taken. This reform has a limited scope, as the rules are applicable only to cases brought by the Commission. At the same time, national rules concerning limitation periods will remain valid in so far as actions by the national authorities are concerned. These may be different from the EC rules. This is particularly noticeable if we compare, for instance, the EC rules to the French rule of Article L 462–7 Commercial Code (former Article 27 of the Ordinance of 1 December 1986). The French rule is stricter: not only is the deadline of three years, instead of five, but the case is completely time-barred before the Competition Council once the limitation period has expired. By contrast, the only limit under EC law is that fines cannot be imposed after the expiry of the limitation period. The application of the French provision governing limitation of actions has led to a great deal of case-law and, as a consequence, some cases must inevitably be dropped. A comparison of national laws on this technical but very important issue will probably display a great variety of approaches. Both firms and their lawyers find it difficult to accept such differences. The rules concerning the limitation of actions should be the same for the application of Articles 81 and 82 EC, regardless of which national authority is responsible for the case. The answer to the second question is also very sensitive, both for the companies and the competition authority itself. The status of the complaints also gives rise to a wide range of solutions. In some national systems, complaints can be rejected through a simple informal procedure. On the opposite, the French Competition Council has to examine the admissibility of each referral and to adopt a formal decision, which can be reviewed by the Court of Appeal of Paris. If the complaint is admissible, the Competition Council has no choice and must examine the case. This diversity is acceptable as long as national competition law is applied, but for the claimant companies this leads however to discrimination. Let us imagine two similar cases concerning abuses of dominant position without a Community dimension. In the first Member State, the competition authority has to deal with the complaint as soon as it receives it; in the second Member State, the competition authority deliberately chooses not to act. As companies are becoming more and more sensitive to the status of their complaints, it becomes necessary to impose some form of judicial control over the decisions that reject them, furthermore in cases where EC rules apply. However, the Commission argues that it is not necessary to harmonise national procedures for the treatment of complaints, and that Article 13 in the draft Regulation — which allows a competition authority to reject a complaint on the grounds that another authority is dealing with the case — provides a sufficient solution to this problem. This solution is not bad, but

Broad Systemic Issues

217

the Commission only considers the hypothesis of what can be called a ‘positive conflict’, that is, a case in which several complaints about the same agreement or behaviour are referred to different national authorities and at least one of them has already agreed to deal with the case. The converse hypothesis, in which no national authority acts upon the case, is ignored. A Commission notice establishing exactly what a ‘Community interest’ entails will not be sufficient to solve this ‘negative conflict’. Without going as far as harmonising the national status of the complaints, it would be useful to have some clear indication of the acceptable grounds for a national competition authority to reject a complaint based on Articles 81 or 82 EC when no other national authority intervenes, and to impose control on such decisions. The right to select cases is also a thorny problem for the competition authorities themselves, especially when their resources are limited. Accepting such a power may contribute to the efficiency of EC competition law enforcement. In France, the Competition Council does not have this power, though it has claimed it for some time now.

1.2.

Investigative Powers

The competition authority may itself exercise the investigative powers. However, this solution is not widespread because few national competition authorities are ‘fully integrated’ in this respect. In most Member States, investigative powers are still in the hands of a governmental department. This dual structure may in itself raise some difficulties for the functioning of the network. But, more generally, the question is twofold: do the national authorities have sufficient investigative powers to efficiently enforce EC competition rules? Is it necessary to harmonise national laws in this respect? That the competition authorities of the Member States have sufficient investigative powers is a fact that can be agreed upon without difficulty. Even if there are some differences between national laws in this respect, the powers of the national competition authorities are very similar in practice. Let us take an example: it is true that the French DGCCRF (Competition, Consumer and Fraud Division, which is responsible for carrying out investigations in this field) cannot send written requests for information, but in practice its agents may ask questions when they visit the premises of companies. Furthermore, the DNE (Direction Nationale des Enquêtes, which is the part of the DGCCRF that has special responsibility for investigations within the whole of the French territory) even summons to its offices members of staff of the companies that are under investigation. Therefore, this is again a problem of resources rather than a legal issue, and it cannot be dealt with by an EC regulation. The fundamental issue from the EC perspective is the territorial limit of the national competition authorities’ investigative powers. A solution to this problem can be found in Article 21(1) of the draft Regulation, which establishes the

218

Working Paper III

grounds for mutual assistance in the collection of evidence. At this stage it seems difficult to go any further than thinking about possible harmonisation, because what can be admissible in one Member State is not necessarily admissible in another Member State. The Commission’s position on this issue is clearly negative: harmonisation is not necessary because the problem of territorial limitation to the competition authorities’ investigative powers does not cause serious difficulties for the functioning of the future network. However, the situation is more complex in reality, because the issue of the competition authorities’ investigative powers clearly overlaps with that of companies’ fundamental rights to argue a defence. The Commission and the national competition authorities will have to face two types of constraints in the future: claims from companies related to equivalent treatment, and the need for a system that is compatible with fundamental rights. The following points are particularly sensitive. • With respect to the authorities’ powers: the powers of the investigative authorities when carrying out inspections on business premises, and the need for, and extent of, prior judicial authorisation to carry out inspections on the premises of individuals. • With respect to the rights of the companies: the right to remain silent, the protection of confidentiality in relation to either legal professional privilege or business secrets. These questions have already been discussed in other contexts. On all these points, it is not the diversity of legal situations that is the main issue. More important is the fact that information obtained through different means and under various standards of protections will circulate within the network.

2.

End of Proceedings

The final stage of the proceedings also raises some difficult issues, which mainly concern the content and the control of the decisions.

2.1.

The content of the Decisions

Content varies according to the type of decision adopted. There seems to be agreement about the exact typology of decisions adopted by competition authorities, but the harmonisation of the sanctions applicable to an infringement is a more difficult issue. Different types of decisions can be identified. The first distinction is between interim measures and definitive measures. For a complainant, it is fundamental to have access to interim measures. The principle of access to interim relief

Broad Systemic Issues

219

seems to have been incorporated into the legislation of most Member States, but this point has to be checked further from the perspective of its practical application. With respect to definitive decisions, the draft Regulation imposes some kind of harmonisation, because, with the exception of the new category of ‘positive decisions’ that are monopolised by the Commission, the national authorities will have to adopt the same types of decisions as the Commission. Two of these types — that is, decisions when there are no grounds for intervention, and infringement decisions (either interim, or definitive) — have already been harmonised at the level of the Member States. The innovation for some national authorities is the introduction of a new category of decisions — decisions accepting commitments — that will require changes in their national procedural systems. This type of decision might be introduced without major difficulties when the competition authority is fully integrated. However, when the authority is not fully integrated, as in the French system, the situation will be more difficult. Who will negotiate the commitments and decide whether they are admissible? Will it be the governmental department (the DGCCRF), the rapporteurs in charge of investigating the case under the authority of the General Rapporteur, or the Competition Council itself? For the companies and their lawyers, the introduction of this type of decision at the national level is good news. However, we will have to wait for the details of implementation to confirm this positive judgement. The most difficult aspects arise in relation to infringement decisions and, above all, with respect to the variety of sanctions available. In most Member States — as under EC law — the sanctions are mainly administrative, and they take the form of fines imposed on companies. However, this identical principle may be enforced in very different ways. First, it is necessary to compare the legal conditions governing the application of fines in each national system. The upper ceiling on the level of fines, which is generally expressed as a percentage of the turnover, may vary. Furthermore, this limit may be calculated on the basis of the national market or on the basis of the worldwide market. It may be limited to the relevant product, or it may take into consideration the entire activity of the group. The criteria that have to be observed when imposing and calculating the fine may also vary: relevant factors might be the duration and gravity of the infringement under EC law, damage caused to the economy in national law, and so on. Furthermore, the practices of the authorities when imposing fines may be as diverse as the law. Companies will not easily accept discrimination with respect to the sanctions imposed for infringements of EC law. It is nevertheless true that no protest has yet been voiced on this point, and there could be two reasons for this. First, the fines imposed by the national authorities for infringement of EC law were lower that the fines imposed by the Commission for the same type of infringement, and so it was advantageous for the case to be brought before the national authorities. Secondly, the enforcement of EC law by the national authorities was very rare.

220

Working Paper III

In the context of a decentralised enforcement of EC law reinforced through the application of Article 3 of the draft Regulation, companies will become more attentive to this new situation, particularly when the case is transferred from one authority to another. Another difficult issue relating to sanctions is to determine whether or not it would be possible for some Member States to maintain criminal sanctions, such as fines on individuals responsible for the infringement and even custodial sentences. These difficulties are reinforced because of the constraints of the ECHR. Even if criminal sanctions do not seem to have been very effective so far, their existence must be taken into account. It is not certain that the safeguard introduced by Article 12(2) of the draft Regulation, which states that only financial penalties may be imposed on the basis of information provided, will be sufficient. The wording ‘financial’ seems too broad, as it includes fines on individuals, and there are no means for controlling whom they will be applied to, or the magnitude of the fines. It is obvious that the Commission is now conscious of these difficulties and does not exclude harmonisation, but harmonisation would only come about in the medium to long-term. The European Commission’s proposal to oblige the national competition authorities to impose fines on companies up to 10 per cent of their total turnover, in conditions similar to those followed by the Commission, could be adopted immediately. There would be no automatic increase in fines as a result of this harmonisation, and any increase in the level of fines could be controlled through guidelines indicating the criteria to be used in establishing the quantum of fines. One must not forget that 10 per cent of the total turnover is only a ceiling. However, the Commission’s 1998 guidelines on fines set a bad example because there was a serious mistake in the calculation. The basis of the calculation was not defined and the amount fixed according to the gravity of the infringement was left entirely to the discretion of the Commission. This has led to very serious distortions between cases invloving multinational groups on the one hand, and small and medium size companies on the other, and these provisions are currently contested before the Court of First Instance. The second argument — that harmonisation will deprive the Member States of their autonomy to apply individual sanctions — is not convincing. Harmonising the rules of administrative sanctions does not affect the possibility that Member States impose secondary criminal sanctions. However, the Parliament’s proposal in this sense seems preferable to that of the Commission. Furthermore, the harmonisation of leniency programs is another reason for stepping up work on this issue. The current differences between the functioning of national leniency programs are such that, in a future system in which information can be exchanged between the authorities, nobody will use a leniency programme if there is a risk that the case will be transmitted to a competition authority in another jurisdiction where such a programme either does not exist or is less protective — consider for example the differences between the very rigid British leniency programme and the new French one.

Broad Systemic Issues

2.2.

221

Control of the Decisions

One last point seems surprising: so far there has been no debate about the fundamental aspect of the control of the decisions handed down by the national competition authorities. Regardless of which judge is entitled to exercise this control, the following two points should be considered for harmonisation purposes. The first concerns the possible suspensory effect of an appeal: an appeal does not have a suspensory effect in most Member States. This is not cause for criticism as there is the ability to ask for a stay of execution. The second is the question of the extent of judicial control: either there is limited review, as under Article 230 EC, or the courts must have unlimited jurisdiction in every Member State. Even if this will not impede the functioning of the network, major differences will be very difficult to accept when the same substantive rules will have to be applied. In conclusion, the strict distinction between the substantive rules determined at the EC level and the procedural rules fixed by Member States did not create major difficulties when the application of Articles 81 and 82 EC by national competition authorities was limited and generally more lenient for firms. As this situation will change following the implementation of the reform, and as integration of European markets increasingly becomes a reality, it will be very difficult to explain major differences in the way that the same EC rules are being enforced. I fully agree with the Commission: full harmonisation is not necessary. Moreover, it will be contrary to the principle of institutional and procedural autonomy. However, we cannot wait for spontaneous convergence. At the very least, immediate intervention would seem to be needed concerning the limits on actions and sanctions.

IV Imelda Maher* Networking Competition Authorities in the European Union: Diversity and Change

I.

Introduction

The reform of the enforcement of Articles 81 and 82 EC exemplifies the fragmentation of regulatory responsibility within the policy space of competition (Scott, 2001: 331). EC competition law will be enforced by at least 25 national competition authorities (NCAs) as well as the EC Commission but those NCAs will only enforce the EC rules within their own jurisdictions, thus fragmenting the geographical scope of the competition rules. The response to this fragmentation of authority and jurisdiction is a robust and stable network of national competition authorities with effective enforcement powers. The greater formalisation of a network, that to some extent existed already, (see the written contribution of D Gerber for this volume) is predicated on the assumption that there is sufficient similarity of purpose (the effective enforcement of the EU competition rules) underlying it. Given the draft Regulation (EC Commission, 2000), this is a fairly safe assumption if viewed in a static manner. However, the enforcement of the law is only effective when it implements the policy underlying it and, as Gerber’s work shows, neither policy nor law stand still (Gerber, 1994). The dynamic nature of EU competition policy and the way the law has been able to respond to that dynamism is one of its key strengths. However, in the context of formalising a network of NCAs this dynamism needs to be managed so that there is agreement over time as to the policies underlying the competition norms. In short, while the text of Articles 81 and 82 EC has remained constant over time, and the Commission’s Directorate-General for Competition (DG Comp) is viewed as generally insulated from the political fray, competition nonetheless is politics in the sense of the day-to-day bureaucratic politics of decision-making and negotiation; in the contests that can occur between DGs in relation to e.g. energy and telecommunications; and at the level of adoption of decisions which go before the full College where in difficult cases there may be compromises (Wilks and Bartle, 2002: 164, 169; Wilks and McGowan, 1996: 254). * Director, Centre for Competition and Consumer Policy, Regulatory Institutions Network (RegNet), Australian National University and Senior Lecturer, Law Department, London School of Economics. Earlier versions of this paper were presented at the Socio-Legal Studies Association Conference, Abersytwyth, April 5 2002, and at Sydney University Law Faculty seminar series. I thank the participants and Colin Scott for useful comments.

224

Working Paper IV

Most fundamentally, the extent to which competition policy favours one particular economic theory over another is in itself a political choice — economic theory is not neutral (Fraser, 1992: 5). Thus, one important function of the network of competition authorities will be to ensure there is a clear understanding of the direction of policy at the supranational level and that, in so far as possible, that policy is itself clearly articulated. This question of policy formation and its communication from the top down highlights the extent to which institutions matter. Institutions here is not limited to the agencies responsible for implementing the rules but also extends to the norms and procedures which constrain and reflect the policies being implemented (Steinmo and Thelen, 1992: 2). Institutions matter not only because they determine policy (as the Commission clearly does in the sphere of competition policy), but also because they mediate those economic and political forces which also shape that policy — often in unanticipated ways (Armstrong, 1995: 165; Maher, 1999: 599). Institutions as mediators of policy, facilitate the incorporation of changes in policy (such as the modernisation of competition law), and the way policy is applied into the existing legal culture. It is this role of institutions as mediators of policy change and their subsequent impact on the operation of the law that makes the organisation of the network of NCAs so important and such a challenging task. This is especially so because just as institutions influence the implementation of policy, so too are they shaped by the context in which they operate. Across the Union, in different sectors of the economy, in different states, among different interest groups, there will be different expectations about the purpose of competition regulation, about what activities and economic operators should be regulated and how. These expectations in turn are shaped by cultural references such as location, time and the legal culture within which these rules operate (Hancher and Moran, 1989: 3; Maher, 1996: 230). The extent to which institutional influences can shape the implementation of EU competition law at the national level in a manner that reflects its context as well as (or, more drastically, instead of), EU policy raises what is the central question of this session: to what extent does the function of the EU network of competition authorities depend on a certain degree of homogeneity of membership? This paper uses an institutional analysis in order to explain the patterns and characteristics (particularly organisation, practices and informal rules), which limit the scope of adaptation and change by NCAs to the decentralisation plans. It in effect provides a check list against which to evaluate difference among NCAs. Moving beyond these largely formal and organisational differences and drawing on the theory of responsive regulation, (Ayres and Braithwaite, 1992; Braithwaite, 2002) it then argues that the extent to which enforcement strategies vary between NCAs in relation to sectors, size of undertaking and across states is crucial to the credibility of the new framework of enforcement. The key function of the network will be to ensure an effective balance between flexibility and fairly fixed guidelines for action which run the risk of a one-size-fits-all model and the attendant dangers of inappropriate regulation and a lack of responsiveness.

Broad Systemic Issues

II.

225

Institutional Endowments of NCAs

Levy and Spiller (1996: 4) suggest in their work on telecommunications regulation that the choices about regulatory governance are constrained by the specific institutional endowments of a nation. They identify six endowments: legislative and executive institutions, judicial institutions, customs and other norms that restrain action, the character of social interests in a society, and administrative capabilities. Drawing on their work but adapting it to take account of formal and informal characteristics specifically relevant to competition authorities, we focus on independence, (an over-arching consideration in relation to competition agencies), administrative capabilities, social interests, and customs and norms that restrain action. If a reductionist approach is taken, the nature of a competition authority can be gleaned from the level of its independence and the other characteristics can be subsumed within that. Instead, we will briefly discuss independence — why it is important and what are its consequences, before disaggregating it and exploring it in the light of other characteristics.

1.

Independence

Independence of competition agencies from government is deemed important for the credibility of competition policy as it ensures that the agency is in a better position to act in the public interest without interference from party politics (Peters, 1996: 48). Decisions are not influenced by the approach of elections allowing for longer-term perspectives and great consistency and neutrality in decision-making as between different interests and over time. Like monetary policy, competition policy is seen as technical in nature and requiring a high level of expertise by those implementing the law. As recent experience in relation to the European Central Bank has shown, an independent agency creates a highly visible commitment to a policy goal (e.g. price stability in the case of EMU or effective competition in relation to competition policy), which is easy to articulate and hence can garner relatively wide political support, but is vague and difficult to apply in practice thus leaving a wide margin of discretion to the relevant agency (Hodson and Maher, 2002). Where the policy goal has a quasi-constitutional status, this in turn enhances the status of the agency responsible for giving effect to it and increases its margin of discretion, provided its actions are seen to achieve the articulated goal. The status of competition policy in EC law is arguably quasiconstitutional given its centrality to the creation of the internal market. At the national level the status of competition policy varies, with compliance with some of the recently repealed but longstanding UK laws deemed optional, while it has a very high constitutional status in Germany (Gerber, 1998: 207, ch 8). To the extent to which the policy goal is vague, the greater the delegation of power to the agency, the greater the pressure for additional accountability to ensure the

226

Working Paper IV

policy objective is being realised (Majone, 1998: 24). Agencies exercising public power need to be accountable in order to ensure that they hold with the public will; promote fairness and act rationally in their decision-making (Scott, 2000: 39). Accountability can be assessed through inputs (how democratic is the decision-making process?); outputs (is the agency achieving its goals?), and process (is there transparency — in so far as possible — of decision-making processes such that the independence of the agency from influence is without doubt?) (Hodson and Maher, 2002). In the enforcement of competition policy, the perceived need for the independence of competition agencies minimises accountability through inputs, leaving outputs and process as the primary accountability mechanisms within traditional public law models. In relation to the creation of any network of public agencies, the extent to which they are accountable for that particular aspect of their work to parliament, the courts, other agencies and auditors, is important. Thus the NCAs will need to be accountable specifically for their work within the network. From an EU perspective, the network is a form of intensive trans-governmentalism (Wallace, 2000: 33) and thus it is not strictly speaking a supranational body with the need for concomitant Union accountability mechanisms. Picciotto has been critical of the emergence of policy networks (of which the proposed network would be one) albeit in the international context (Picciotto, 1997: 263; Maher, 2002: 135). The concern is that fragmentation of government erodes traditional accountability mechanisms. Slaughter, on the other hand, argues that at least in relation to technocratic networks, they are insulated from some politics but ultimately remain accountable through existing national channels (Slaughter 2000: 195). In relation to the European network, it can be seen to some extent as shoring up the state by delegating enforcement responsibility back to the national level in a manner entirely in the spirit of subsidiarity. It is also designed to ensure more effective compliance and enforcement with any changes in enforcement policy having been discussed by the NCAs. Despite this optimism, and in the light of the debate concerning legitimacy and the European Central Bank (de Grauwe, 2000: 164), it behoves the NCAs to remain ever conscious of the importance of legitimising the network through greater public accountability which can be put in place by the network ensuring at least some transparency in its practices. The extent to which NCAs are independent from government varies in the EU. The longest established competition agencies in the EU (the Bundeskartellamt, the EC Commission and the UK Competition Commission and Office of Fair Trading), are not fully independent agencies as the term is used in US literature (Horn, 1995: 55), with some potential for majoritarian influences, but there is little evidence of regulatory capture (Wilks and Bartlett, 2002: 151). Using a principal-agent analysis, Wilks and Bartlett show how the establishment of an independent competition authority often has a strong symbolic element. Early symbolism, where competition agencies were not seen as having much of a policy impact, has given way over time to entrenched competition agencies with extensive technical expertise having influence and a sense of priorities shaped by

Broad Systemic Issues

227

their own experience in a policy climate where governments now express an overt preference for stronger competition laws (2002: 149 and 170). The time frame is significant here as it allows for the embedding of competition policy within the polity, shaping expectations of those subject to the law and of the general public. The agency can develop a body of rules and ensure its processes are legitimate and consistent with the norms of natural justice. The longevity of an agency alone is however not sufficient. Its status, resources and the exact nature of its powers (in particular the credibility of the sanctions it can deploy), are significant in ensuring that the agency is effective in its allocated role (see table 1). Table 1.

Institutional Endowments of NCAs

Status

Resources

Powers

Constraints

Location

Equipment

Dependency on courts

Appointment/ human rights Media/public perception Position v. other NCAs and other domestic agencies

Personnel

Role in policy debates (advocacy) Arsenal of decisionmaking and sanctions

2.

Budget

Constitution/human rights Ideological constraints (social interests)

Status

Status is something that is dependent on powers and resources. It is important in determining and augmenting the agency’s influence in the policy domain and in achieving its role. The degree of independence, the expertise of staff, its presence in the media and its position in relation to other agencies are important formal and informal factors in determining its status. The location of the agency may be symbolically important. For example the UK Office of Fair Trading (a government department) is located near the edge of the City, and the Competition Commission has vacated premises which it shared with other government agencies for new premises — both a couple of kilometres from Whitehall, the home of government. The Irish authority initially had offices rented from a major insurance company located near one of the main shopping districts not far from the main travel hubs. It is now located about 1km from government at a more prestigious and central address. The move to some extent reflected the increase in its powers following statutory changes in 1995. Recruitment and appointment are important for two reasons. First, if appointment is by government, it is an important way of exercising some control over the direction of the agency at key junctures, although it can be difficult to predict how a particular appointee will perform once within the agency especially if the basis for dismissal is tightly controlled. Secondly, the terms of appointment (remuneration, duration, additional benefits) and the perceived status of the

228

Working Paper IV

agency are relevant factors in attracting suitably qualified and widely respected personnel to the agency and in particular, to the Head of the agency. The reputation of the Head in turn is important in attracting other highly qualified staff, as is the reputation of the agency in relation to other regulators. If there are other regulatory agencies with an ambience of power that is not present in the competition authority, then it will find it more difficult to recruit and retain staff who may be more interested in working with other regulators. In Ireland the authority experienced staffing crises in the 1990s where the Authority was not quorate for some months due to the failure of the Minister to appoint sufficient members and later, due to the resignation of the Chair and a delay in replacing him (Maher, 1999: 75; OECD, 2001: 62). The more limited criminal powers of the Authority and its smaller staff compared with e.g. the Commission for Communications Regulation could arguably make it a less attractive destination than that regulator. This phenomenon, in so far as it exists for NCAs, will dissipate when the NCA has power to enforce the European norms. Association with the highly successful DG Competition and the ability to enforce EU rules, and membership of a European-wide network are all important factors in improving status in relation in particular to other utility regulators at the national level. Sometimes, personnel are drawn from the civil service bringing with them values and practices imbued within a permanent closed-career structure where there are received ways of conducting business and mutual rating by other civil servants (of the valued qualities of cleverness, reliability and influence) are important (Hall et al, 2000:35). These ‘loans’ are part of the career path of the civil servants concerned and they are motivated to perform well so their status within the government bureaucracy can be enhanced and they will proceed to interesting and dynamic work when they move on from the agency. These loans also facilitate links between the agency and other parts of government as civil servants have their own networks and understand the modus operandi of various government departments. This information can be useful but at the same time, the independence of the agency can be undermined if it is part of the civil service or employs a very large number of civil servants on its staff. All of these factors affect public perception of the agency and the kind and amount of coverage it gets in the national media. Where the agency enjoys broad public support for its work, is seen as advocating the public benefit and as doing ‘a good job’, this increases its status making it more difficult for its independence to be eroded (should government be so minded), and in fact underpins that independence as it is seen as having an identity separate from government.

3.

Resources

Linked to status is the issue of resources. Limited budgets are not a bar to an agency developing its role and its effectiveness (as can be clearly seen in the

Broad Systemic Issues

229

history of DG Competition), but it can impede effective enforcement in particular when there are insufficient resources to deal with the number of cases arising. Opportunities for policy innovation are also curtailed, as staff attention remains focused on day-to-day management of case-load. Where the creation of the agency is symbolic and is not intended to lead to real policy change, then restricting resources ensures the legislation and agency remain largely symbolic (Wilks and Bartlett, 2002: 155, quoting Mitnick). Limited resources can reflect an uncertain commitment on the part of government to the policy. Even if the agency enjoys wide formal powers, if they have limited resources which affects office space, IT and administrative support, it will be difficult to attract dynamic personnel respected in their field and with relevant expertise which in turn will reduce the status of the agency further. In general, where an agency is dependent on a government department for its budget and personnel it can experience indirect ministerial control over priorities (OECD, 2001: 62).

4.

Powers

Formally, the scope of the agency’s powers is central to the effectiveness of its role. In the EU context, the substantive scope of national competition rules is largely consistent with Articles 81 and 82 EC (Maher, 1996; Drahos, 2001: 199). For competition agencies these powers fall under two main headings: • policy formation: to what extent are government and other agencies required to seek its opinion in relation to relevant policy formation? • rule enforcement: how effective is it? The extent to which an NCA can influence emerging policy debates is an indicator of the commitment of the government to competition policy and its position within the wider policy framework. Effective advocacy in relation to competition policy requires a high level of agency independence and adequate resources so this overview function can be undertaken without jeopardising day-to-day enforcement issues. Such advocacy also provides some safeguards against the whittling away of the impact of competition law and policy. It can advocate the removal of anomalies such as that found in Greece, where competition legislation applied to publicly-owned undertakings but government intervention created and preserved state monopolies — initially at least (Vainanidis and Pournara, 1992: 229). It provides the agency with an opportunity to point at anomalies in the law and to advocate removal of the more blatant, e.g. the amendment of legislation that allows blatant anticompetitive activity by professions governed by older statutes. More fundamentally, the specific enforcement powers available to the authority are central to its effectiveness. It needs its sanctions to be credible such as to

230

Working Paper IV

encourage compliance with the law and when it does decide to impose sanctions that they are sufficient, and can be secured in a timely manner so as to act as an effective deterrent and to create an example for other subjects of the law. For competition agencies, the key requirements are: an ability to impose fines; adequate investigative powers including the ability to secure search warrants; the power to impose interim measures; and, where there is non-compliance with their investigations, the power to impose periodic fines for non-compliance. Whether criminal sanctions are necessary as well as civil remedies is a much debated topic (OECD, 2002; Hammond and Penrose, 2001; Wils, 2002). The ability to imprison directors of a company for breach of the competition rules can bring the rules into the boardroom and focus minds in a way a financial penalty on the company cannot. One aspect of the process surrounding criminal actions is whether they can be brought directly by the competition authority or must be brought through a government body (like a public prosecutor) with responsibility for bringing criminal actions more generally. If and when such actions are brought, the question of whether they arise before a specialist court or not is also important, given the technical nature of the dispute. While criminal sanctions may focus the mind, substantial fines that can be imposed by the competition authority without recourse to a court of law may be less cumbersome and hence more effective. At the same time, if within a hierarchy of sanctions the competition authority can point to the possibility of imprisonment, a criminal record for the company and/or its liable officers, then this may provide additional incentives for compliance and improve the overall effectiveness of the law even if rarely invoked in practice. Whether sanctions are civil or criminal, where the authority lacks the power to impose fines and is reliant on persuading another government body of the importance of pursuing prosecutions, this more cumbersome process makes the costs and risks associated with enforcement greater for the authority.

5.

Constraints

Even where there are adequate formal powers of action there remain constraints — formal and informal. Formally, the constitution and human rights law (in particular the European Convention on Human Rights — hereafter ‘the ECHR’), may impose considerable procedural constraints on action even allowing for the doctrine of supremacy. Thus in Ireland, the Competition Act 2002 gave the Authority the power to enforce Articles 81 and 82 EC, but it can only do so by pursuing undertakings through the civil or criminal courts, and in relation to the latter, it can do so only through the agency of the Public Prosecutor. This is because the imposition of fines could be regarded as criminal and under the constitution, and criminal sanctions can only be imposed by a court of law (Maher, 1999: 53; Jones, 2001: 406). It could be argued that the Authority should be able to impose fines itself in relation to Articles 81 and 82 EC, but it would have to be shown first that the effectiveness of the competition rules is being undermined by this

Broad Systemic Issues

231

particular means of enforcement. There may be human rights constraints on procedures, either under the ECHR or domestic law, on the way fines are imposed and/or if they are deemed to constitute criminal charges, and the nature of Commission decisions whether they are civil or criminal (Wils, 2002). A related issue is the degree of protection an undertaking is entitled to when under investigation and defending itself against the Commission (Jones, 2001: 415; Commission Notice on Cooperation within the Network of Competition Authorities, OJe 104/43, 27.04.04, para. 23). Finally, there are informal constraints — customs and culture — that may restrict action. Some sectors may be more difficult to tackle, even where there is no regulatory capture, because of the lack-lustre commitment of government to the imposition of competition norms on what are often viewed as traditional sectors or because of a grass roots revolt to the operation of the law. For example, an Irish farmers association chose to pay a sizeable contempt of court fine when it continued a boycott of meat factories in demand of higher prices, in defiance of an injunction secured under competition law (Irish Times, 2000). The extent to which competition is recognised and endorsed as an important policy by the general public is another important constraint. In the UK, a recent survey of experts showed only 10 per cent of them believed competition policy to be important to the UK public (Price, Waterhouse, Coopers: 2001). This may reflect the limited effectiveness of the previous laws and, with the emphasis on education under the new regime this may change (Maher, 2000).

III.

Enforcement Strategies

Within the proposed EU enforcement model, the nature of the substantive rules is (relatively) clear. Article 5 of the draft Regulation gives NCAs the power to apply Articles 81 and 82 EC in their entirety. While sanctions are listed — interim measures, fines, accepting commitments — there is no obligation to harmonise sanctions across Member States. Instead the Commission Notice on Cooperation within the Network of Competition Authorities expressely states that the Member States recognise the standards of each other’s systems as the basis for cooperation (at para. 2). Article 35 requires states to designate NCAs (there may be more than one especially if sectoral regulators are included), and to take the measures necessary to empower those authorities by May 2004, elaborating the general duty of loyalty under Article 10 EC. Given the explicit obligation, the Commission could pursue a Member State through the European courts if it is tardy in introducing appropriate sanctions under Article 226 EC, once having exhausted less formal attempts to secure compliance with the new Regulation. The difficulty will be proving that procedures are ineffective — while there may be extreme cases, effectiveness is very much a matter of degree. There will also be a role for the network to undertake ex post review of the availability of sanctions, and how they are being applied to allow some comparison to be drawn and to identify best practice.

232

Working Paper IV

The Information is available through the ex ante consultation requirements in Articles 11(3) and (4) of the Regulation where the Commission must be informed by an NCA when acting under Articles 81 and 82 and prior to the adoption of a decision under the EU rules. And under Article 14(7) of the Regulation the Advisory Committee can give opinions on decisions of NCAs in certain circumstances (Commission Notice on Cooperation, section 3.1.). This information needs to be collated and reviewed say annually to compare and contrast enforcement practice. Ideally, the information should be available publicly e.g. through an appendix to the Annual Competition Report or as a freestanding document. Discrepancies (if they exist), could then be explored. Uniformity should not be expected given differences between national economies, but the factors influencing different outcomes in similar decisions could be usefully identified. The table below shows the variation in sanctions for breaches of domestic competition laws that are expected to apply to the EU rules when applied at the national level. Table 2 Procedures for Enforcement of National Competition Laws in EU States (Drawing on Jones, 2001) State

Fines (max) In Euro

Prison (max)

Austria Belgium Denmark Greece Spain Finland France

720,000 (u) and 130,000 (I) 10% of annual turnover for Belgian sales & exports No limit 15% of world-wide group annual turnover 10% of turnover 10% of world-wide annual turnover 3250,000 (I); 5% French annual turnover (U) or 65 million (non profit U) More than the financial benefit of the breach to the U. 3.8 million or 10% of annual turnover 10% of Italian annual turnover for relevant market 454,000 1 million 578,000 or 10% of world-wide annual turnover 10% of aggregate turnover for last 3 yrs in UK re relevant market

3 yrs (price fixing)

Germany Ireland Italy Netherlands Portugal Sweden United Kingdom

4 yrs 5 yrs

5 yrs hard core cartels (Enterprise Act, 2002)

U – Undertaking; I – Individual

Research has shown that the perceived fairness of procedures in decision-making processes affects the general satisfaction with a decision, support for the regulator and acceptance of and compliance with the decision (Wenzl, 2001: 6; Braithwaite, 2002: 84 discussing Tyler and Blader, 2000). Procedural fairness here refers to neutrality, trustworthiness and respect. Procedural justice and the fairness of the decisions-making process are significant for subjects of the law because such treatment is seen as indicative of their inclusion within the social

Broad Systemic Issues

233

group (i.e. the law-abiding, competitive business community). Where there is perceived procedural fairness, cooperative behaviour with regulators increases because the subjects receive the message that they are respected within the group (Braithwaite, 2002: ch 4). These concerns are important in relation to the reform of the EC competition law enforcement system because they highlight not so much the importance of uniformity (something of a chimera at any rate), but of consistency and fairness of process. This further highlights the issue of different outcomes in relation to application of the EC competition rules by different NCAs — and the important advocacy role that the network (preferably) or the Commission will have to play in explaining how differences to some extent are predicated on economic differences. The imposition of sanctions is however only one aspect of effective enforcement. Arguably the key characteristic of an effective competition policy is one where a competition culture is embedded in business practice and all government policy. Hence an advocacy role is important for competition authorities to ensure that competition policy infuses government policy (Dabbah, 2000: 371). Effective advocacy shores up the commitment of the state to competition policy, ensures consistency in the scope of competition law and facilitates compliance. A second aspect of advocacy is focused not on government but on those subject to the law. Here, competition authorities can fulfil a vital role in educating and informing both potential complainants, victims and perpetrators of anticompetitive practices of the exact scope of the law, what it is designed to achieve, how this is beneficial, how the law is applied and what the consequences of breach can be. To fulfil such a function, agencies require resources to allow them e.g. to go on information road shows, present papers, publish leaflets for display in public libraries, supermarkets, local chambers of commerce etc. Resourcing is going to be a problem for NCAs, who, it seems, are unlikely to get many extra resources to carry out their new role (Kingston, 2001: 346). This renders it all the more important to secure compliance without recourse to formal (and expensive) enforcement strategies. Compliance can be encouraged in a number of ways. An excellent recent example was the way the UK Office of Fair Trading used the lead time between the signing of the 1998 Competition Act into law and its actual implementation to launch a well targeted and highly effective education campaign that included a well organised web site, conference papers, useful booklets on various aspects of the Act, an emphasis on compliance programs (Rodger, 2000: 253) and more formally, the introduction of a large number of guidelines (Office of Fair Trading, 2001: ch. 2; Maher, 2000: 549). This information campaign seem to have had some success in particular, the campaign to discourage notification led to only seven requests for decision and 15 requests for guidance in the first 18 months of the Act’s operation (Vickers, 2001). An emphasis on securing compliance before having recourse to formal sanctions reflects two features of competition regulation. First, the authority is dependent on business for information and thus it has to engage to some degree, in a cooperative relationship with those being regulated. This cooperation is

234

Working Paper IV

qualified of course when the authority is supported by strong and effective sanctions either for breach of the substantive norms or for failing to assist in any formal investigations conducted. In practice the presence of these powers make it difficult for firms not to offer some level of cooperation if and when approached. Secondly, no competition agency can have a proverbial policeman at every street corner or in every office. Resources will never be such as to ensure complete command-and-control. Thus, it pays to secure cooperative compliance by companies so that the competition enforcer’s personnel can be directed to the serious breaches of the law that warrant sanction. Similarly, it pays for firms as voluntary adaptation of behaviour removes the (real) risk of investigation and sanctions e.g. a compliance programme is cheaper than investigation and/or litigation. This approach — called responsive regulation (Ayres and Braithwaite, 1992; Braithwaite, 2002) and adopted to some degree by the Australian Competition and Consumer Commission — operates within the shadow of the law in that compliance can be best achieved where there is a substantial threat of detection and sanction (with substantial sanctions up to and including ‘incapacitation’ which in the competition sphere would mean e.g. imprisonment). Responsive regulation where cooperative compliance is encouraged minimises regulatory interference so as to maximise freedom of action within the law (Braithwaite, 2002: 33). Detection is more likely if compliance becomes more widespread and firms are less willing to tolerate breaches of the law by others so complaints and whistle blowing become important. The status of the authority, its access to the media and the wide reporting of its decisions, all add to the impression of a powerful agency capable of detecting and widely publicising breaches of the law. The strategy is represented by what is called a regulatory pyramid.

Figure 1 The Regulatory Pyramid (Ayres and Braithwaite, 1992: ch 2; Braithwaite, 2002: ch 2)

Broad Systemic Issues

235

The extent to which individual NCAs exercise a compliance strategy in relation to the EU norms or domestic law significantly affects the extent to which competition policy is embedded within national culture, which in turn affects the extent to which anticompetitive behaviour is viewed as unacceptable. The extent to which there are differences in relation to the attitudes to competition policy and the way NCAs approach compliance outside the realm of strict legal enforcement will necessarily impact on the more visible and more easily measurable standard of how many decisions are taken a year and what were the fines imposed. And yet, these compliance strategies can have a fundamental impact not just on competition culture with in the state but also on perception as to the consistency of the operation of European norms throughout the EU. One aspect of enforcement which will be central to the operation of the network is the exchange of information — an issue discussed at length under Panel III of this Workshop. Article 12 of the Regulation does not require NCAs to exchange information — nor to assist each other in investigations — they are simply empowered to do so. This provision is problematic on two fronts (at least). First, it does not require authorities to transfer information — it merely provides a legal power to do so that overrides national law (Kingston, 2000: 345; Commission Notice on Cooperation, para. 29). This emphasises the cooperative nature of the relationship. Secondly, while functionally the exchange of information is crucial to the effective operation of decentralised enforcement to ensure consistency and to avoid jurisdictional disputes, the regime as yet is so new, that there is little to underpin the proposed exchange apart from the Commission Notice, which has been endorsed by all Member States except Greece through the signing of a statement (Commission Notice, para. 72). In short, the existing national competition law systems may be intolerant of such exchanges. It will be for the network to create a climate of confidence to facilitate trust and mutuality such that transfer of information — even where proceedings are regarded as criminal in nature — will be engaged in without undue difficulty.

IV.

Conclusions

The network under the new Regulation is in many ways resonant with another soft policy coordination mechanism in the EU — that of the open method of coordination (Lisbon, 2001; Hodson and Maher, 2001). Conditions similar to those that saw the development of the open method in the sphere of economic policy are also present in relation to competition policy. There is a common dominant theoretical framework; common goals; key elites who coordinate in the policy area; and a commitment by the Commission (in this case) to the project. The characteristics of the open method: peer review, surveillance, bench-marking and best practice, could well serve the tasks of the network especially in relation to non-formal enforcement strategies and institutional endowments. The network will be more supranational and hierarchical than the analogy with the open

236

Working Paper IV

method would suggest. The Regulation sets out a specific agenda backed by 40 years of law enforcement and is supported by a DG of the Commission, i.e. DG Competition, which exercises considerable influence and power. Lesson-drawing — where lessons are drawn from other places or times, but do not necessarily lead to any policy change (Dolowitz and Marsh, 1996: 344) — will hardly be enough for NCAs or an effective competition policy. Instead, what may best characterise the network is some form of policy transfer with an emphasis on compliance and enforcement strategies. Policy transfer is a process where ‘governments study each other’s different methods, gauge the success of various policy alternatives and mimic best practices employed elsewhere, with successful policies transferred deliberately and willingly’ (Bomberg and Peterson, 2000, 6). One form of transfer that seems most pertinent in relation to the network is indirect coercive transfer where the presence of externalities or functional interdependencies leads to joint action (Dolowitz and Marsh 1996). Functional interdependencies arise not only at the level of the Heads of NCAs but lower down the administrative structure. Just as the Sutherland Report found that a key issue in the completion of the internal market was an increase in trust among national officials at all levels (High Level Group, 1992; Council, 1994), so too in the competition sphere will mutual trust between officials at different levels in NCAs be necessary to ensure the effective operation of the network. The network is one where there is a hierarchy with the Commission in charge though acting in a cooperative spirit with the NCAs. Thus while the success of the network is dependent on the NCAs — including but not only the their Heads — there is also a responsibility on the Commission to ensure its smooth operation both in shaping the agenda of the network and facilitating the sort of policy learning that will best ensure the common goal of an effective competition policy.

V Alexander Schaub* The Commission’s Position within the Network

I.

Introduction

Three years ago, the Commission adopted a White Paper in which it proposed to abolish its monopoly over the application of Article 81(3) EC. A spirit of decentralisation — some preferred the term ‘subsidiarity’ — inspired this proposal. It is not often that a monopolist decides, on its own initiative, to get rid of its prerogative and to share its powers with others. There were two reasons for the Commission to make this change: first, it believes that it is desirable to involve more national bodies in the application of Community law in order to bring that law closer to the citizen and to increase its acceptance. And, second, it remains convinced that, in an enlarged Community, the really efficient protection of competition cannot rely on a single institution. This proposal, which has been rather broadly accepted in the meantime, has been the subject of criticism from two angles. Some critics — mainly from the European Parliament and industry circles — warned against a ‘re-nationalisation’ of EC competition law. They feared that one of the pillars of the Treaty was being given back to Member States. They understood the proposal as indicating the Commission’s intention to withdraw from the field of EC competition law enforcement and to limit itself to issuing block exemption regulations and guidelines. Others heavily criticised the so-called ‘control’ the Commission would have over the national competition authorities (hereafter ‘NCAs’) in the new system. They denounced the ‘supervision’ under which the Commission would like to place NCAs. The purpose of this paper is to examine the Commission’s position within the network. It will show that the Commission is by no means retreating from the enforcement of EC competition rules, but that its intervention will be better focussed and more effective, and that it will work in much closer cooperation with NCAs. In the new system, the Commission will be one enforcer amongst others, but it will retain particular responsibility for the functioning of the network.

* Director-General, DG Competition, European Commission. The author thanks Céline Gauer, DG Competition, for her contribution to the preparation of this paper.

238

Working Paper V

II. The Commission will be Part of a Network of Public Enforcers The Commission will remain an active enforcer of the EC competition rules. Article 85(1) EC entrusts the Commission with the task of ensuring the application of the principles laid down in Articles 81 and 82 EC. It states: Without prejudice to Article 84, the Commission shall ensure the application of the principles laid down in Articles 81 and 82. On application by a Member State or on its own initiative, and in cooperation with the competent authorities in the Member States, who shall give it their assistance, the Commission shall investigate cases of suspected infringement of these principles. If it finds that there has been an infringement, it shall propose appropriate measures to bring it to an end.

The Commission’s enforcement role was recently underlined by the European Court of Justice in the Masterfoods case.1 The judgment recalls the content of Article 85(1) EC and concludes that ‘the Commission is entitled at any time to adopt individual decisions under Articles 81 and 82 of the Treaty.’ It follows that the Commission can never be precluded from taking a decision in a given case. This judgment is based on the Treaty itself. The reform of Regulation 17/62 could not and will not harm the role of the Commission in this respect. The aim of the reform is, on the contrary, to step up the enforcement of EC competition rules by eliminating the routine examination of notifications and thus allowing the Commission to focus on the prosecution of serious infringements. However, in a system of parallel competencies, the Commission will deal with cases when it is better placed to do so than other members of the network.

1.

A Greater Contribution of NCAs to the Application of EC Law

In the current system, NCAs do not contribute as much as they effectively could to the enforcement of EC competition rules. First, they are not yet all empowered to apply EC law.2 Secondly, in the countries in which competition authorities can apply EC law, they very rarely do so.3 This can largely be explained by the existence of the exclusive competence of the Commission to apply Article 81(3) EC. Complainants lodge their complaints with the Commission in order to avoid a dilatory notification, or they bring their case to the national authority but under national competition law. 1 Judgment of the Court of Justice of 14 December 2000 in Case C–344/98 Masterfoods [2000] ECR I–1412. 2 UK, Ireland, Luxembourg, Finland and Austria are not yet empowered to apply EC law. 3 Since the notice on cooperation of 1997, NCAs inform the Commission when they take decisions that apply EC competition law. On average, a total of ten decisions applying Community law are made per year by all NCAs combined.

Broad Systemic Issues

239

Where EC competition law is applied by NCAs, it is mainly in parallel with national law, or to supplement national competition law if the national law does not apply to certain sectors. This explains why the enforcement of EC competition law relies essentially on the Commission. This will significantly change in the new system. NCAs will all be fully empowered to apply Community law and will have an obligation to apply Community law when agreements or practices affect trade between Member States. Moreover, applying Community law and being part of a network of competition authorities will reinforce their independence. The pressure that certain political or economic interest groups could try to exercise on them will be discouraged by the fact that consultation will take place within the network, and the interpretation of the law given by the authorities will have to comply with the case law of the European Court of Justice. The NCAs’ more active contribution to the enforcement of Community rules will be of great importance for the protection of competition. They are close to the markets and generally have a good knowledge of them. They will act in order to protect consumers in their territory; that is, they will generally prosecute infringements that have substantial effects on their territory. The Commission would therefore deal with cases that affect a single Member State only to compensate for a lack of enforcement or if proceedings are unduly drawn out.

2.

Parallel Action by NCAs or Action by the Commission?

The Commission will generally be considered to be better placed to deal with a given case where, in order to solve a competition problem, it is required to adopt a decision that is enforceable in a number of Member States. In the new system, Commission decisions are still the only decisions that can be enforced throughout the Community. In order to bring an end to infringements that have substantial effects in more than one Member State and to sanction them appropriately, there are two different options: either several NCAs act in parallel, or the Commission intervenes. The draft Regulation and the draft Notice on the functioning of the network take a pragmatic approach to this aspect. Some flexibility has to remain in the system. Companies argue in favour of an automatic one-stop shop — that is, they want the exclusion of parallel action as a matter of principle. However, parallel action can be an efficient way to solve certain competition problems, particularly when they are regional in scope. In an enlarged Community, it will be neither desirable nor feasible to allocate all cases to the Commission that cannot be dealt with by a single national authority. On the other hand, where parallel action involves more than three NCAs, it is very unlikely that it will be efficient. The Commission should generally deal with these cases.

240

Working Paper V

In order to allow the Commission to enforce the rules efficiently, the draft regulation creates several types of decisions. First, it adopts Regulation 17/62’s power to bring an end to an infringement. Article 7 of the draft Regulation also spells out clearly that the Commission has the power to impose any remedy necessary, including structural ones. The latter aspect has given rise to many discussions. However, it is not — as many have argued — a new tool. The Commission already had this power under Regulation 17/62, and it has sometimes made use of it.4 As in the current system, it will have to be exercised under the control of the European Court of Justice and will be subject to the principle of proportionality. Changes to the structure of an existing company, which most commentators had in mind, will happen only in extreme cases: the Commission would only impose such a remedy if no other solution with equivalent effect exists. Second, Article 9 of the draft Regulation creates a new tool for the Commission, which was inspired by the ‘consent decrees’ — decisions accepting commitments — of the American system. These decisions allow the Commission to close proceedings by accepting commitments offered by the parties. This can be an efficient way of restoring competition to a market without having to hear a case through to the point at which a decision is awarded. It would solve the problems identified more speedily and would save the costs of a proceeding. These types of decision are only appropriate in cases in which the Commission does not impose fines.

3.

The Commission as Policy-Maker

Other cases that are suited to Community competition policy should also be dealt with at the European level. The Commission has always used individual decisions to determine policy directions in new areas. In the field of economic law, cases are essential to the drawing of policy lines. It is only when sufficient experience has been gathered that the adoption of a horizontal can be envisaged. The Court of Justice recently recalled the special responsibility of the Commission for the establishing of competition policy. In the Masterfoods case, the Court held that: the Commission, entrusted by Article 85(1) of the EC Treaty with the task of ensuring the application of the principles laid down in Articles 81 and 82 of the Treaty, is responsible for defining and implementing Community competition policy (emphasis added).

The new enforcement system cannot put into question the role of the Commission as policy-maker. Being a Community institution, the Commission

4 See

Commission Decision Warner-Lambert/Gillette and BIC/Gillette, OJ L 1993/116, p. 21. See also judgment of 17.11.1987, Joined Cases 142/84 and 156/84 British-American Tobacco and R.J. Reynolds [1987] ECR 4487.

Broad Systemic Issues

241

retains its special responsibility for the definition of Community competition policy. However, this does not imply that NCAs cannot deal with cases that raise new issues of law. It simply means that, irrespective of the geographic scope of a case, the Commission could be persuaded to deal with a given case in order to set policy for the whole Community by the adoption of a Community act. To that effect, the draft Regulation includes a special instrument: decisions finding inapplicability of Article 81 or 82 EC to an agreement or practice. These decisions, foreseen in Article 10 of the draft Regulation, will be taken in the Community public interest. They will be made by the Commission on its own initiative and not, like exemption decisions under Regulation 17/62, upon notification of an agreement by companies. They will complement guidelines and block-exemption regulations. Article 7 of the draft Regulation also clarifies that the Commission can have an interest in taking a prohibition decision, even if an infringement has already been brought to an end and if the Commission does not impose sanctions. This is particularly the situation when there is a need to set policy vis-à-vis a certain type of restriction. Prohibition decisions are very important to draw the line between what is prohibited and what is not. All these decisions, being Community acts, will have the effects described by the European Court of Justice in the case law.5 National Courts or authorities will not be allowed to contradict Commission decisions without, as far as national courts are concerned, making a reference for a preliminary ruling to the Court of Justice. For the sake of clarity, the draft Regulation codifies this case-law in Article 16. The numerous comments made on these judgments show how necessary it is to have Article 16 in the new Regulation.

4.

An Enhanced Role for the Advisory Committee

All Commission decisions will be made in close cooperation with the NCAs. The Advisory Committee will be consulted as it is at present, but its contribution will be given more weight. This is, first, because it will be composed of members of competition authorities that bring with them the daily experience of enforcing Community competition rules, making their contributions even more valuable. Secondly, the regulation itself will strengthen the Committee’s position. Article 14(5), as drafted in the Presidency compromise text, states that: The Commission shall take the utmost account of the opinion delivered by the Advisory Committee. It shall inform the Committee of the manner in which its opinion has been taken into account.

5 See Case C–344/98 Masterfoods [2000] ECR I–1412, Case C–234/89 Delimitis [1991] ECR I–935, and Case 314/85 Foto-Frost [1987] ECR I–4199.

242

Working Paper V

The text also provides for the publication of the opinion at the request of the Committee. These elements will give more weight to the opinion of the Advisory Committee.

II. The Commission Will Have Particular Responsibility for Ensuring the Efficient Functioning of the Network New issues arise in a system of parallel competencies: cases have to be dealt with efficiently and in a consistent manner. The Commission, as the guardian of the Treaty, has a special responsibility in this respect and therefore enjoys certain powers to fulfil its duties. However, these powers will be exercised in close cooperation with the NCAs, which will themselves contribute both to an efficient allocation of cases and to a consistent application of the rules.

1.

The Commission’s Role in the Allocation of Cases

Case allocation will not be mechanical but flexible in the new system. Criteria for allocation will be qualitative in nature. In most cases, the authority receiving the complaint or the application for leniency will deal with it, and there will be no discussion within the network about the allocation of the case. But it could occur that the same case is brought to the attention of several authorities simultaneously or successively. The draft Regulation creates a mechanism in order to avoid unnecessary parallel proceedings and to ensure, to the extent possible, a one-stop shop. Article 13 states: 1.

Where competition authorities of two or more Member States have received a complaint or are acting on their own initiative under Article 81 or Article 82 of the Treaty against the same agreement, decision of an association or practice, the fact that one authority is dealing with the case shall be sufficient grounds for the others to suspend the proceedings before them or to reject the complaint. The Commission may likewise reject a complaint on the ground that the competition authority of a Member State is dealing with the case. 2. Where the competition authority of a Member State or the Commission has received a complaint against an agreement, decision of an association or practice which has already been dealt with by another competition authority, it may reject it.

Article 13 forms a legal basis by which to reject a complaint on the grounds that another authority is dealing with, or has dealt with, the same case (that is, the same practice or agreement in the same market by the same companies).

Broad Systemic Issues

243

However, it is only a power and there is no rule defining which authority has to close or suspend its proceedings. This leaves scope for appreciation of the peculiarities of each individual case. This flexibility is important; if a national authority rejected a complaint after thorough examination, another national authority may not want to redo the assessment. On the other hand, if a complaint was rejected because the national authority was unable to gather evidence of the infringement, another national authority might wish to carry out its own investigation. Given the voluntary character of the provision, disagreement may arise between two authorities. In that case, the relevant competition authorities should endeavour to find a workable solution by engaging in direct exchanges. This should nonetheless be limited in time. In order to give certainty to companies involved and to ensure an effective and swift protection of competition, case allocation should be settled as quickly as possible. The network could work with an indicative deadline (say, up to three months) during which cases should normally be allocated. Of course, this would not prevent the reallocation of the case if new facts were to arise and justify such a reallocation (for example, if a cartel that was thought to be national is of a wider scope, or vice versa). If discussions between these authorities do not lead to a satisfactory solution, there is a need for a referee. In the proposed system, that referee can only be the Commission. If there is sufficient Community interest in doing so, the Commission will open proceedings on the basis of Article 11(6) and bring an end to the disagreement over case allocation. This will rarely happen, but it is needed as a safety valve in the system.

2.

The Commission’s Duty to Ensure a Consistent Application of the Rules

The existence of several decision-makers necessarily raises the question of the consistency of the decisions. As the new system commences, it cannot be excluded that decisions will diverge to some degree, given that each authority has its own traditions and legal culture. Given that there is no integrated judicial system in the Community, the responsibility for ensuring consistent application of the rules will have to be ensured by the authorities themselves. To that effect, the proposal sets out an obligation on the NCAs to inform the Commission prior to making a decision. The information sent to the Commission should be made available to all NCAs at the same time. Article 11(4) of the draft Regulation defines the scope of the obligation to inform the Commission of proposed decision. This provision states: No later than 30 days before adopting a decision requiring that an infringement be brought to an end, accepting commitments or withdrawing the benefit of a block

244

Working Paper V

exemption regulation, the competition authorities of the Member States shall inform the Commission. To that effect, they shall provide the Commission with a summary of the case, the envisaged decision or, in the absence thereof, any other document indicating the proposed course of action. This information may also be made available to the competition authorities of the other Member States. At the request of the Commission or the competition authority of a Member State, the acting competition authority shall make available other existing documents necessary for the assessment of the case.

The obligation to inform is limited to three types of decisions: prohibition decisions, decisions accepting commitments, and decisions withdrawing the benefit of a block-exemption regulation. Decisions ordering interim measures are not subject to an obligation to inform because they are not definitive and cannot prejudge the outcome of the proceedings. Decisions rejecting complaints or closing ex officio procedures are not subject to consultation because they do not change the legal position of the parties. They only state that a given authority will not proceed with a case. This does not preclude other authorities from investigating the matter. Prohibition decisions, on the other hand, have severe consequences for companies and create situations that are difficult to undo. Decisions withdrawing the benefit of a block exemption regulation are also important because they allow a national body to reverse a Community act. In practice, NCAs will have to send to the Commission — one month before the adoption of the decision, at the latest — a document setting out the content of the envisaged course of action. For prohibition decisions, this will often be the national equivalent of a statement of objections (which defines the ultimate scope of the prohibition, in accordance with the general principle that the parties must be given the opportunity to answer the objections before the final decision is made). For decisions accepting commitments, it will be a document setting out the competition concerns identified by the national authority and the proposed commitments to solve the competition problems. All authorities within the network will share the information. They will all be able to read decisions and make their contributions to consistent application. They will be well placed to do so as they might have dealt with similar cases before. In this context, there are two possible risks of inconsistency: several authorities acting in parallel may come to different outcomes, or a national authority may adopt a decision contrary to the established case law. To assess this later aspect, the case law of the European Court of Justice, block exemption regulations, and Commission decisions should serve as a yardstick. Where necessary, following the information provided pursuant to Article 11(4), a discussion can be held in the Advisory Committee. This discussion may take place at the request of either a national authority or the Commission. It will not lead to a

Broad Systemic Issues

245

formal opinion but simply to an exchange of views. This should, in principle, be sufficient to ensure the consistent application of the law. However, if no agreement can be found and if there is a persistent risk of inconsistent application, it will ultimately be for the Commission to ensure that the law is applied consistently. It will be able to do so by opening proceedings and removing a case from NCAs pursuant to Article 11(6) of the draft Regulation. However, this should only happen after the Commission has duly consulted with the Member State(s) concerned, has explained the need for intervention, and has given an opportunity for discussion in the Advisory Committee. The Commission will not allow companies to turn it into an appeal body for case allocation or for decisions of national authorities. The essence of the mechanism of consultation is to avoid having companies play the Commission off against NCAs. It is already the case that companies and their lawyers, when they face proceedings at national level, try to involve the Commission. However, companies should not have any right to force the Commission to start proceedings pursuant to Article 11(6) in order to seize the case from the NCAs. This decision will only be made on the Commission’s own initiative, on the basis of the information received from the authorities and the exchanges carried out. After the adoption of a decision by a national authority, the Commission will normally not contradict national decisions if consultation has taken place. Given the scope of the obligation to inform, this means that the Commission should normally not adopt a decision that is more favourable to the companies concerned than the one adopted by the national authority. On the other hand, it will not have to take account of rejections of complaints made by NCAs. In a system of decentralised application of Community law, the Commission will have a new role to play: it will need to ensure that case allocation functions properly and that the rules are applied consistently. This will have to be done in close cooperation with the NCAs.

IV.

Conclusion

The proposal abolishes the Commission’s monopoly under Article 81(3) EC, but it does not limit its role in the field of competition policy. Nor does it re-nationalise competition law. The aim of the whole reform is to bring Community law closer to the citizen and to improve the protection of competition by involving more enforcement bodies. The abolition of the monopoly will change the position of the Commission. The Commission has a prominent role by law in the current system, because the monopoly in itself ensures that cases are brought to its attention. In the new system, the Commission’s strength will depend directly on the quality of its

246

Working Paper V

intervention. If it is credible as an enforcer, detects and successfully prosecutes infringements, and maintains or efficiently restores competition in the market, it will be respected and feared by companies engaging in unlawful activities. If it plays its role of referee with moderation and equity, it will not be contested by the NCAs. This is the new challenge that the Commission will have to face.

VI Dr Ulf Böge* The Commission’s Position Within the Network: The Perspective of the NCAs

I.

The Commission’s Special Position

The EC Member States and the public received their first information about the European Commission’s concrete plans to reform Regulation 17/62 through the White Paper on Modernisation of the Rules Implementing Articles 85 and 86 (now 81 and 82) of the EC Treaty.1 The White Paper included among its arguments those of increased decentralisation in the application of EC competition rules and the creation of an EC network of competition authorities based on partnership. The White Paper states, inter alia, that ‘[T]he competition authorities are well placed to take effective action in certain types of case: they are normally well acquainted with local markets and national operators … .’. In another context, the Commission says that: ‘[T]he proper functioning of the network … clearly implies a reinforcement of the role of the Advisory Committee on Restrictive Practices and Dominant Positions.’ From the very beginning, the Bundeskartellamt has expressly welcomed and supported these elements of the modernisation project. At the same time, although the Commission’s White Paper announced comprehensive measures to ensure the coherent and uniform application of the law in the new enforcement system, the business community and legal practitioners expressed doubts regarding the compatibility of the new enforcement system with the Treaty as well as concerns about legal certainty. These doubts and concerns may have contributed to the fact that the rules concerning decentralisation and the network that were finally included in the draft regulation were different from what the White Paper had originally led observers to expect. In the view of the Bundeskartellamt and other national competition authorities and Member States, the draft Regulation contains various provisions that instead lead to essential enforcement roles being centralised in the hands of the Commission. This would put into question the reform approach as a whole. Decisive motives for deriving this impression are, in particular, the interaction or combination of the obligatory application of EU law under Article 3 of the draft regulation, the obligation to inform (and originally also the obligation * President of the Bundeskartellamt. 1 OJ C 132 of 12.5.1999, pp. 1–47.

248

Working Paper VI

to consult) on behalf the Member States vis-à-vis the Commission under Article 11(3) and (4) of the draft Regulation, and the Commission’s extensive power to take over cases in the course of national proceedings under Article 11(6) of the draft regulation. The termination of the Commission’s monopoly over the application of Article 81(3) EC and the extension of national competencies resulting from this is largely neutralised by the simultaneous introduction of the legal exception system, that is, the general abolition of decisions granting exemptions. However, the Commission reserves for itself the sole competence to make ‘positive decisions’ under Article 10 of the draft Regulation, and thus re-establishes its de facto special position. Moreover, the Member States’ current participation rights in the process of issuing block exemption regulations are to be abolished by Article 28 of the draft Regulation. Apart from the actual text of the draft Regulation, the impression that there is an intention to centralise the process is reinforced by the discussions within the Council Working Group, as well as by working papers that have been submitted in the aftermath of the publication of the draft regulation. The Commission intends, for example, to lay out rules concerning the functioning of the future network of competition authorities in a special Commission Notice — thus excluding a genuine opportunity for the directly concerned national authorities to participate — and claims extensive competencies for itself in the allocation of cases. It is undisputed that the European Commission has a special role due to its responsibility in the development of EC competition policy and the necessity to establish coherence within the network of competition authorities in Europe. However, the Commission’s special rights should be confined to what is actually necessary, as any further restriction of the national competition authorities’ independence, competencies and opportunities to influence the process would have a counterproductive effect on the protection of competition in Europe. The following is an analysis of three areas in which, according to the present state of discussions, this precondition is not fulfilled and that, particularly if regarded as a whole, counteract the aim of effectively decentralising the application of EC competition law and creating a network based on partnership.

II.

Allocation of Cases

In contrast to the European merger control, the Commission and the Member States have concurrent competencies in the area of Articles 81 and 82 EC. As Article 3 of the draft Regulation would significantly increase the number of cases in which Articles 81 and 82 EC are applied, and the Member States are to be involved more intensively in the application of the provisions, it is undoubtedly necessary to establish general principles clarifying which authorities are to handle which cases. The relevant questions are, first, where or how should such principles be laid down? And, second, what substantive criteria should they include?

Broad Systemic Issues

249

As to the first question, Germany and other Member States had originally called for the inclusion of the essential principles of case allocation in the text of the Regulation itself. The Member States directly concerned by these provisions can only exercise genuine participation rights within the framework of a Council decision. As experience shows, increased opportunities for influence will also lead to a greater degree of active participation by the parties concerned, thus strengthening the parties’ identification with the rules they jointly agreed upon, and finally resulting in a more effective binding or commitment effect. These effects are indispensable if the network of competition authorities is to work successfully. However, the Commission refuses to have general criteria for case allocation included in the text of the Regulation, arguing that if amendments to these principles were to become necessary they could only be decided upon with the Council’s consent, thus rendering the system too inflexible. Moreover, the Commission argues that there is a danger that predetermined criteria for case allocation would justify claims by the parties to the respective proceedings. Instead, the Commission wants to regulate the decisive case allocation criteria in a Notice. This will be discussed with the Member States within the framework of expert meetings and finally adopted by the Commission ‘in consensus’ with the Member States, following the same procedure as with other Commission guidelines. From the German point of view, there are concerns about such a procedure. Experience has shown that the Member States’ opportunities to influence the outcome of expert meetings are even smaller than in the Advisory Committee, and very often neither a consensus nor the above-mentioned binding effect for national competition authorities can be achieved. As a compromise, it has been suggested that general principles on case allocation be adopted through a joint declaration of the Council and the Commission. However, if such a document were ultimately to consist of merely vague political statements and fail to include any significant substantive criteria, the purpose of genuine participation by the Council and the Member States would be totally missed. As regards the criteria for case allocation, the Member States and the Commission have agreed to follow the one-stop shop principle as far as possible. The Dutch and German delegations have submitted to the Council Working Group their own concrete proposal on how this principle could be put into practice. According to this proposal, a national competition authority should be regarded as the ‘best placed authority’ to handle a case if a restriction of competition has a substantial effect only in its own Member State and the cause for the restriction of competition is to be found in this particular Member State. If two or three Member States are substantially concerned, their respective national authorities will be regarded as being ‘best placed’ unless one particular national authority would be able to solve a case for the other Member States as well. According to the proposal, the Commission is to be considered the ‘best placed’ authority if more than three Member States are affected by a restriction of competition, if a Community-wide binding decision is required, or if there is a close relation to

250

Working Paper VI

other EU provisions that only the Commission can apply. Apart from this, the Dutch/German proposal includes a general commitment by the network members to allocate cases as fast and transparently as possible in order to contribute to the effective implementation of European competition rules. In its own proposal for a Notice, the Commission has altered these principles so that a national authority’s responsibility for national cases is no longer formulated as a principle, but merely as a possibility. However, the Commission’s competence is extended and given a certain priority over the others by using generous terms, such as ‘Community interest’ or ‘particularly well placed’. The idea of a network based on partnership can hardly be found in this proposal. In the majority of cases, decisions about case allocation will be totally unequivocal. Moreover, experience has also shown that the cause and effect of a competition restriction are often to be found in one single Member State, and therefore the national competition authority concerned will have the greatest interest, the best market knowledge and all necessary instruments to handle the case. These facts cannot and will not be changed by a reform of Regulation 17/62. If, despite these facts, competencies were to be transferred from the national competition authorities to the Commission, the motivation and initiative of the national authorities would be drastically reduced as a consequence. From the German point of view, the establishment of the general principles for case allocation described above is indispensable for a well-functioning and transparent network. This particularly applies to cases with a national focus. The ability to diverge from these general principles in cases of justified exceptions remains unaffected.

III.

Consistency

In an enforcement system in which several distinct authorities apply the same legal provisions, it is desirable to ensure a certain degree of consistency and coherence in the application of the rules, particularly from the perspective of the companies affected. In the European network of competition authorities, the Commission (as guardian of the EC Treaty) undoubtedly plays a particularly responsible role in this process. However, with its proposals concerning this issue laid down in Article 11 of the draft Regulation, the Commission once again goes far beyond what is necessary and tolerable. The original formulations in Article 11(3) and (4) of the draft Regulation lay forth comprehensive and largely one-sided duties on the part of the Member States to inform and consult the Commission whenever they apply Article 81 or 82 EC. These rules would not only have restricted the independence and motivation of the national competition authorities, but would also have involved additional time and effort in legal proceedings and possibly bottlenecks as a result of the additional stage of examination by the Commission. During the discussion

Broad Systemic Issues

251

within the Council Working Group, the extent and one-sided nature of these information duties were relativised and the consultation concept deleted entirely from the text. At the same time, at least the possibility to inform the other national competition authorities of the respective proceedings was explicitly included in the text of the Regulation. The latter is important, as one of the main reasons for the Bundeskartellamt to accept the information obligations is that they serve an important role in the promotion of cooperation within the network (for example, as regards case allocation). However, even with the developments described above, the core function of Article 11 in the draft Regulation is still that of an instrument of control by the Commission vis-à-vis the national competition authorities. Probably the most far-reaching power that the Commission has reserved for itself is the possibility (embodied in Article 11(6) of the draft Regulation) to take over a case from the national competition authorities at any time, and to consequently end the national proceedings. Although the Commission is in principle already vested with this power, this provision — in combination with Article 3 of the draft Regulation, which drastically raises the number of proceedings potentially affected — gains a completely new dimension. The implications are even more serious if one considers that the Commission will not subject its power to withdraw a case to any kind of conditions or restrictions in the text of the Regulation itself. Even in a possible joint declaration with the Council the Commission will, at most, lay down some general broad criteria for the applicability of this provision. In the Bundeskartellamt’s opinion, the rules proposed by the Commission for ensuring a consistent application of the law lead to uncertainty and even mistrust at the level of the Member States in respect of their future role within the network. The Commission’s assertions that it only intends to make use of its competencies in individual cases stand in contrast with its refusal to include restrictive conditions in the text in consensus with the Member States. Even if the Commission bears special responsibility in this connection, the responsibility for controlling the application of European competition law and ensuring coherence ultimately lies not with the Commission as executive organ, but with the European courts, which should continue to exercise this function in future. After all, absolute uniformity can never be achieved in a network in which different decentralised bodies apply the law. The same considerations apply to the German network of competition authorities, and even in the case of a single authority. If the Commission is indeed pursuing the aim of decentralising application of EC competition rules, then it should accept this fact and allow room for a contribution to the development and improvement of legal practice by the national competition authorities. The German and Dutch delegations have proposed a two-phase model to tackle the issues of case allocation and consistency. Case allocation should generally be concluded in the first phase, in view of the above principles. During this phase, the Commission would be free to take over a case in accordance with

252

Working Paper VI

Article 11(6) of the draft Regulation. The first phase should be limited to two or three months. In the second phase, case allocation should not in principle be altered, thus limiting to a few clearly defined exceptional situations the Commission’s ability to withdraw a case. The Advisory Committee should function as the potential controlling body of the Member States within this system, and it should also be able to take an opposing view to the Commission when it comes to taking over a case. This model should either be included in the Regulation itself or (otherwise) be made part of the above-mentioned joint declaration of the Council and the Commission. As more detailed rules for the practical cooperation of the competition authorities will be needed in addition to these basic principles, the new network should also be given the competence to issue common declarations.

IV.

Concretisation of the Law

A third area in which the Commission plans to extend its special position at the expense of the Member States is related to the issuing of block exemption regulations; this area is regulated by Article 28 of the draft Regulation. Whereas existing rules allow a block exemption regulation to be adopted only on the basis of a special enabling Council regulation that defines the area of application and the content of the regulation concerned, the new Regulation grants the Commission a ‘general power’ in this regard. Such a Regulation would eliminate the existing participation rights of the Council and the Member States, as well as the consultation rights of the European Parliament. The Commission justifies its proposal by arguing that block exemption regulations are ultimately nothing other than individual exemptions (which, however, will no longer exist under the new system) and could thus be decided by the Commission alone. However, the effects of the new block exemption regulation on vertical agreements, and the public interest in the new block exemption regulation on motor vehicles distribution, already show that such general regulations establish fundamental guiding rules which often form the most important legal basis for those concerned. Obligatory application of EU law under Article 3 of the draft Regulation would result in the national authorities increasingly examining and applying the Commission’s block exemption regulations. Thus, their substantive provisions would directly affect them. For this reason, such a massive restriction of the participation rights of the Member States is unacceptable from the German perspective. The only possible alternative to restricting the Council’s powers would be to accordingly extend the rights of the Advisory Committee. Under current practice, the Commission does consult the Advisory Committee before it adopts a new regulation or issues possible notices, and to some extent it takes into account the concerns and proposals expressed during the consultative process.

Broad Systemic Issues

253

However, the Advisory Committee does not have to consent to new legal texts. In addition, it does not have the ability to give a public ‘opinion’ about the proposals. Ultimately, this gives the Commission the exclusive power to decide on the content of the new regulations, irrespective of whether a consensus with the Member States is reached (which, after all, is often not the case). Naturally, the effectiveness of the discussions in the Advisory Committee also depends crucially on the active participation of the Member States. This participation would undoubtedly be enhanced if their opportunities for influence were extended. However, according to Article 14 of the draft Regulation, the existing rights of the Advisory Committee will remain largely unchanged. This means that, thus far, there are no plans for a possible ‘compensation’ to Article 28 of the draft Regulation. As in the areas of case allocation and consistency in the application of the law, the Commission’s special responsibility is also undisputed with respect to the question of the (binding) concretisation of EU competition law. Nevertheless, it would be in keeping with the decentralisation and network concept if the Member States’ position in the development of the interpretation of Articles 81 and 82 EC were to be improved. This applies particularly in view of Article 3 of the draft Regulation. Therefore, the Council’s participation rights should not be weakened for any reason, and the Advisory Committee’s role should be strengthened in the medium term.

V.

Conclusions

The future network of competition authorities in the EU will only be effective if the rules on which its functioning is based are well accepted and respected by all the parties that are involved. Moreover, effective protection of competition in Europe can only be ensured with the support of the national competition authorities and their knowledge of the local, regional and national markets. Therefore, on no account can the rules of the new system be permitted to suppress the national authorities’ own initiative and competence. This applies to questions of case allocation, consistent application of the law and the concretisation of EU competition law (as analysed here), as well as to other areas of the new Regulation. Some of the issues that are still problematic from the German point of view indeed seem to involve substantial differences in evaluation, while others could be improved by simply reformulating the wording. In any case, the Member States cannot and should not rely merely on the oral statement of positive intentions on the part of the Commission. The Commission has repeatedly preached the community ‘spirit’ of the new network during Council Working Group meetings. The Bundeskartellamt would like to believe in this new form of cooperation, and it is prepared to make every effort to contribute to its success. However, in order to give the network a

254

Working Paper VI

reliable and workable framework of rules, the trust-based cooperation between its participants must begin well before this framework is completed. Ultimately, the prime objective of the new system must be to create as much cooperation based on partnership as possible, and as little centralisation by the Commission as necessary.

VII Mario Siragusa* The Commission’s Position Within the Network: The Perspective of the Legal Practitioners

The purpose of this paper is to discuss some aspects of the proposed Regulation for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty and amending Regulations (EEC) No 1017/68, (EEC) No 2988/74, (EEC) No 4056/86, (EEC) No 3975/87 (hereafter ‘the draft Regulation’).1 After a brief overview of the main changes introduced by the draft Regulation, the paper will discuss the importance of Article 3 within the system envisaged by the draft Regulation. Attention will then be focused on the provisions of the draft Regulation that aim to regulate the allocation of cases between the Commission and the National Competition Authorities (hereafter ‘the NCAs’), and among the different NCAs. Finally, the paper suggests some possible changes to further improve the new decentralised system of application of EC competition law.

I.

The Core of the Draft Regulation

In a nutshell, the so-called ‘decentralisation’ or ‘modernisation’ of EC competition law essentially consists of moving from a system whereby the Commission holds a monopoly over the application of Article 81(3) EC towards a system in which the provision can be directly applied by any NCA or national court (a ‘legal exemption system’).2 Following the implementation of the reform, the enforcement of Articles 81 and 82 EC will be entrusted to a network composed of all the NCAs, the national courts and the Commission. The Commission will be a member of the network and its centre of gravity; a sort of ‘primus inter pares’. It will be the Commission’s

* Cleary, Gottlieb, Steen & Hamilton, Brussels and Rome. The author wishes to acknowledge the invaluable contribution of his colleague Michele Piergiovanni. 1 COM(2000) 582 final, 27.9.2000. 2 Article 1 of the draft Regulation. According to the Commission, the reform will decrease its workload and allow it to concentrate its efforts and resources on detecting major infringements of the EC rules of competition. In addition to this, it will reduce the burden on the undertakings, especially as far as the costs of a notification are concerned, and it will start the process of adapting the structures devoted to the enforcement of EC competition law to the new dimensions the Community will reach when the enlargement of the European Union takes place.

256

Working Paper VII

responsibility to guide the development of EC competition policy and to ensure the uniform application of EC competition law throughout the EC Member States. The idea of abandoning the Commission’s monopoly in the application of Article 81(3) EC has generally been welcomed. Allowing different NCAs and the national courts to apply the whole of EC competition rules is, on the whole, considered to be a positive evolution, and another step towards a more ‘federal’ system of enforcement of Community law as a whole. However, the simple fact of granting NCAs and national courts the power to apply not only Article 81(1) and (2) EC and Article 82 EC, but also Article 81(3) EC, is not in itself sufficient to achieve a decentralised application of EC competition law as envisaged by the Commission. The real core of the reform is Article 3 of the draft Regulation. This provision defines the relationship between Articles 81 and 82 EC, on the one hand, and national competition laws, on the other hand, by stating that: where an agreement, a decision by an association of undertakings or a concerted practice within the meaning of Article 81 of the Treaty or the abuse of a dominant position within the meaning of Article 82 may affect trade between Member States, Community competition law shall apply to the exclusion of national competition law.

In the absence of a duty to exclusively apply Community law when a case involving competition issues affects trade between Member States, it is doubtful whether NCAs and national courts would apply Article 81 and 82 EC rather than their national laws, even if they were empowered to do so. Indeed, in a scenario in which the whole body of EC competition rules is directly applicable, but a provision like Article 3 of the draft Regulation does not exist, the existing rules governing the relationship between national and Community law would continue to apply. NCAs and national courts will still be allowed to apply their national competition laws — even when dealing with a case that has a transnational dimension — as long as they do not conflict with Community provisions. Furthermore, without a duty to apply EC competition law whenever trade between Member States is likely to be affected by an agreement or a practice, NCAs and national courts may prefer to examine any such case on the basis of their national provisions and limit their inquiry to the effects of the agreement or the practice in their respective national markets. In this way, they might also avoid any interference by other NCAs, the Commission and the European Courts. The Commission would therefore give up its monopoly over the application of Article 81(3) EC, and lose one of the instruments through which it has been able to apply EC competition law throughout the different Member States. The application of the provision would be left to NCAs and national courts that, in turn, would not be inclined to do apply EC competition law unless required to do so by a precise provision. If this were the case, there would be a risk that the

Broad Systemic Issues

257

reform would result more in a nationalisation of competition law than in a decentralisation of the application of EC competition rules.3 On the other hand, if a binding provision obliges NCAs and national courts to apply Community competition law to the exclusion of their national provisions whenever they are called upon to assess a case affecting trade between Member States, the scenario changes radically. Faced with an agreement or a practice that has such transnational effects, NCAs and national courts would no longer have the choice to limit their analysis to the national aspects of the case and to thus commence proceedings under national law. They would be obliged to apply EC competition law exclusively, while the application of national laws — which are likely to give rise to inconsistencies, and to threaten the ‘level playing field throughout the European Union’ that the reform intends to create4 — would be ruled out. By guaranteeing that NCAs and national courts will effectively apply Article 81(3) EC and the rest of Articles 81 and 82 EC, the existence of such an obligation would somehow counterbalance the Commission’s loss of its exclusive power to apply Article 81(3) EC. Moreover, the absence of a provision obliging NCAs and national courts to apply EC competition law when dealing with cases having transnational effects would deprive the proposed system of its foundations. The reform is based on the assumption that, following its implementation, the application of EC competition law will be effectively decentralised. It is on this understanding that all the provisions contained in Chapters II, III and IV of the draft Regulation have been conceived. A provision like that articulated in Article 3 of the draft Regulation must be seen as a necessary cornerstone of the system. Of course, this does not mean that its current wording cannot be improved. But it must be borne in mind that any changes to Article 3 should not affect its basic role of regulating the relationship between EC competition law and national laws, and obliging NCAs and national courts to apply the former in certain circumstances. In light of the above, and given the importance of the provision, some of the criticisms of the current text of Article 3 of the draft Regulation will now be examined. Moreover, it will be considered whether and eventually how, Article 3 could be improved without detracting from its function.

II.

Article 3 of the Draft Regulation

An interesting criticism of the current text of Article 3 comes from the German Monopolies Commission (‘GMC’), which attacks the total displacement of 3 In

the Explanatory Memorandum preceding the draft Regulation, the Commission itself recognises that NCAs and national courts may be more inclined to apply their national provisions. It further adds that such provisions, though not conflicting with EC competition law, are not the same body of rules whose application the Commission intends to decentralise through the reform. 4 See the Explanatory Memorandum preceding the draft Regulation.

258

Working Paper VII

national competition laws in favour of EC competition law that such article would cause. The counter-draft Regulation it puts forward involves modifying the current wording of Article 3 of the draft Regulation to exclude the application of national competition law only when it conflicts with the EC provisions, and to allow NCAs and national courts to apply their national rules when such rules are more stringent than the Community ones. In such a way, it is argued, (i) the implementation of competition law would be more effective due to the assessment of individual cases under more specific and differentiated national laws, and (ii) the dialogue and comparison between EC competition law and different national regulations would stimulate the development of EC competition policy. Greater legal certainty would be also ensured through the avoidance of the time and money-consuming discussions that would be likely to arise concerning the issue of whether, in each specific case, an agreement or practice is likely to affect trade between Member States. In light of the above, the GMC welcomes the fact that the Commission has recently shown its willingness to reach a compromise with respect to Article 82 EC, according to which, ‘more stringent national regulations’ may remain applicable in the control of abuse of dominant positions. On the other hand, it strongly criticises the unwillingness of the Commission to open a debate with respect to Article 81(3) EC. The Commission’s reluctance to modify the text of Article 3 of the draft Regulation, at least as far as Article 81 EC is concerned, is especially justified in light of what has been argued above. As for those cases falling under Article 81 EC, the possibility that NCAs and national courts would be able to apply national competition laws that are more stringent than EC competition rules, should not be welcomed. It is true that an agreement or practice to which only EC competition law should be applied under the current version of Article 3 may at times be better analysed in accordance with more stringent and detailed national legislation than in accordance with the Community body of rules. However, the greater number of negative implications that such approach would entail outweighs this advantage. If such approach was upheld: (i)

to the extent that more restrictive national provisions exist, the application of EC competition law by NCAs and national courts will be endangered. For the reasons set out above, if they were allowed to choose between applying EC competition law or their national provisions, it is likely that they would elect the latter; (ii) the Commission’s ability to fulfil its role of ensuring the uniform application of EC competition law and to guide the development of EC competition policy would be seriously endangered. The possibility that NCAs and national courts would reach conflicting judgments would increase as they would no longer be bound to apply the same body of

Broad Systemic Issues

259

rules. After losing its monopoly over the application of Article 81(3) EC, the Commission will undoubtedly find it more difficult to draw the lines along which EC competition law should develop by issuing Article 81(3) EC decisions, especially if NCAs and national courts were allowed to disregard them simply because of the existence of more stringent national rules; (iii) the outcome of a case will strongly depend on whether an NCA or a national court hears it. Paradoxically, if national legislation that is more restrictive than EC competition law is in force in just one Member State, the very same agreement could be considered to be ‘not prohibited’ or ‘exemptible’ by the NCAs or the national courts of all but that State. As a consequence, forum shopping and disputes concerning the issue of competence would increase; (iv) finally, legal certainty would be jeopardised for undertakings operating at a pan-European level. Undertakings would indeed be obliged to proceed to a preliminary autonomous assessment (no longer being entitled to notify the Commission of the agreement in order to obtain clearance for any agreement they wish to enter into. This has transnational dimensions in light of the relevant legislation of all the Member States in which the agreement will have effects. In light of the above, the approach recently adopted by the Commission — that is, to allow the application of more stringent national laws to those cases falling under Article 82 EC, but not to those to be dealt with under Article 81 EC — seems to be a good compromise. Indeed, the potential lack of uniformity in the application of Article 82 EC that I outlined above is less likely to occur, and it would certainly entail less dangerous consequences than the application of Article 81 EC. It is doubtful whether there are a great many fields or sectors in which national laws may adopt a more stringent approach than the one adopted by the Commission and the European Courts in the control of abuses. Even if this were the case, the differences are likely to be marginal. As a consequence, it is less likely that EC competition rules will not be applied and that their uniform interpretation and legal certainty will be jeopardised. Furthermore, the application of Article 82 EC is less politically sensitive than that of Article 81 EC, especially as far as its third paragraph is concerned, so if NCAs and national courts apply their more restrictive national rules to the control of abuses, this would create fewer difficulties for the Commission in achieving its task of guiding the development of EC competition policy. Finally, the Commission’s new approach is mirrored by a new version of the text of Article 3 of the draft Regulation — recently circulated informally by the Commission — pursuant to which, as a general rule, NCAs and national courts would be allowed to apply both national competition law and EC competition

260

Working Paper VII

law in parallel to the same set of facts when trade between Member States is affected. Nonetheless, a distinction is drawn between cases falling under Article 81 EC and cases to be dealt with under Article 82 EC.5 Indeed, as far as Article 81 EC cases are concerned, a caveat is introduced by the second paragraph of the proposed text of the provision. This states: The application of national competition law may not lead to the prohibition of agreements, decisions of associations of undertakings or concerted practices which may affect trade between Member States that do not restrict competition within the meaning of Article 81(1) of the Treaty, that fulfil the conditions of Article 81(3) of the Treaty or that are covered by a regulation for the application of Article 81(3) of the Treaty.

NCAs and national courts would therefore be prevented from applying more stringent national rules to Article 81 EC cases, whereas — as anticipated above — they would be allowed to do so in Article 82 EC cases. To summarise, the Commission is trying to be receptive towards the criticisms expressed, especially by the different NCAs, to the original text of Article 3 of the draft regulation. At the same time, it is trying to retain the instruments it needs to ensure application of EC competition law at a national level and the uniformity of such application, and to guide the development of EC competition policy.

III. The Powers of the Commission In the New Enforcement System This part of the paper discusses whether, assuming that a provision ensuring an effective application of EC competition law exists within the system, the powers the draft regulation grants the Commission are suitable to enable it to achieve the tasks it is supposed to carry out within the new EC competition law enforcement system (that is, to (i) guarantee a uniform application of EC competition law, and (ii) guide its development). In other words, are the 5 The

new version of the text of Article 3 of the draft regulation is the following: ‘Where the competition authorities of the Member States or national courts apply national competition law to agreements, decisions of associations of undertakings or concerted practices within the meaning of Article 81(1) of the Treaty which may affect trade between Member States, they shall also decide upon the legality of such agreements, decisions or practices under Article 81 of the Treaty.’ The application of national competition law may not lead to the prohibition of agreements, decisions of associations of undertakings or concerted practices which may affect trade between Member States that do not restrict competition within the meaning of Article 81(1) EC, that fulfil the conditions of Article 81(3) EC or that are covered by a regulation in the application of Article 81(3) EC. ‘Where the competition authorities of the Member States or national courts apply national competition law to any abuse prohibited by Article 82 of the Treaty, they shall also apply Article 82 of the Treaty’.

Broad Systemic Issues

261

powers the draft Regulation grants to the Commission an adequate substitute for its loss of the exclusive power to apply Article 81(3) EC, which has always constituted a privileged tool for the Commission to achieve the aforementioned tasks? Very briefly, the powers with which the Commission is asked to guide the development of EC competition law and ensure its uniform application are: • the power to adopt an individual decision finding an infringement of Article 81 or 82 EC, acting on a complaint or on its own initiative (Article 7 of the draft Regulation); • the power to adopt decisions finding that Article 81 EC is not applicable, either because the conditions of Article 81(1) EC are not fulfilled or because those of Article 81(3) are satisfied, or that Article 82 EC is not applicable, acting on its own initiative and in accordance with the public interest (Article 10 of the draft Regulation); • the right to be informed of the commencement of any proceeding involving the application of Article 81 and 82 EC started by any NCA (Article 11(3) of the draft Regulation); • the right to be consulted where any NCA intends to adopt a decision under Article 81 or 82 EC that requires an infringement be brought to an end, or that accepts commitments or withdraws the benefit of a block exemption regulation (Article 11(4) of the draft Regulation); • the power to withdraw a case from an NCA and to deal with it (Article 11(6) of the draft Regulation); • the power to adopt block exemption regulations (Article 28 of the draft Regulation);6 • the power to adopt notices and guidelines.7 The rules laid down by the draft Regulation are a good starting point to ensure that the Commission is able to keep its central position within the network.8 However, the existing body of rules can be further improved and clarified. One of the most crucial and debated aspects of the draft Regulation is the allocation of cases within the network: between the Commission and the NCAs (the ‘vertical division of competencies’), and among the different NCAs (the ‘horizontal division of competencies’).

6 It

is likely that the Commission will adopt an increasing number of block exemption regulations in order to guide the actions of the different NCAs applying EC competition law. 7 For the same reasons stated above with respect to block exemption regulations, it is also likely and desirable that the Commission will, at the same time, issue more notices and guidelines interpreting EC competition law. 8 The strong institutional position of the Commission, even without the exemption monopoly, is underlined by CD. Ehlermann, Reform of European competition law. Coherent application of EC competition law in a system of parallel competencies, Freiburg, 10.11.2000, available at http://www.europa.eu.int/ comm/competition/antitrust/others.

262

1.

Working Paper VII

The Vertical Division of Competencies

The provisions regulating the vertical division of competencies are Article 11(6) and Article 13(1) of the draft Regulation. According to the former, the Commission’s initiation of proceedings for the adoption of a decision under this Regulation shall relieve the competition authorities of the Member States of their competence to apply Articles 81 and 82 EC. Pursuant to the latter, if a case is already being dealt with by a NCA, the Commission may refrain from acting and may reject a complaint. The centrality of the role of the Commission within the network is already evident from the above-mentioned provisions. The Commission’s simple gesture of opening a proceeding on the basis of Article 81 or 82 EC automatically deprives any NCA of the power to apply EC competition rules to the same facts. On the other hand, when an NCA commences a proceeding on the basis of EC competition law, the Commission must choose between refraining from acting and letting the NCA deal with the case, or opening a proceeding itself (with the consequence that the case is withdrawn from the NCA). In this respect, the first issue that needs to be analysed is whether the power entrusted to the Commission by Article 11 (6) of the draft Regulation should remain as it is (that is, unlimited power) or whether the introduction of some kind of criterion to guide the exercise of this power should be welcomed. Again, the question should be examined bearing in mind the role envisaged for the Commission within the network. There is no doubt that the Commission should be allowed to withdraw a case from a NCA when the issues at stake are of crucial importance from the point of view of the development of EC competition policy and the uniform interpretation of the law. On the other hand, it should not be forgotten that the aim of the draft Regulation is to decentralise the application of EC competition law. The intervention of the Commission pursuant to Article 11(6) should be an exception to the general rule, according to which EC competition law is uniformly and consistently applied by the different NCAs. If this were not the case, the whole system would collapse. From this perspective, the system does not require legally binding criteria that precisely define the competence of the Commission and the NCAs from a vertical standpoint. In any event, the Commission is not likely to exercise its power to withdraw a case from an NCA in too pervasive a fashion. As mentioned above, it would be against the spirit of the reform, but also against its practical aims (which are to decrease the Commission’s work-load and allow it to concentrate on the major infringements of EC competition law). In addition, NCAs would feel less motivated to pursue infringements on the basis of EC law if they knew that the Commission is active in withdrawing cases from NCAs when the issues at stake are important, complex and new. Furthermore, NCAs may be tempted to pursue these cases only under national law.

Broad Systemic Issues

263

In any event, in order to increase legal certainty, the Commission should issue a non-legally binding notice whereby it states how it will exercise its power of withdrawal. A non-binding instrument is preferable to a binding one, because: (i) it ensures flexibility within the system, and (ii) it reduces actions by private parties who may prefer to have their case dealt with at a national level rather than at a Community one, or the other way around. As for the content of such a notice, quantitative criteria such as turnover thresholds are not the best indicator of either the centre of gravity of a case, its importance to the development of competition policy, or for the uniform interpretation of EC competition law. Qualitative criteria should be preferred. These could include of the geographical scope of the relevant market involved in a specific case (the Commission may intervene when an issue affects competition in three or more Member States, and no national market is mainly affected). They could also — as stated above — comprise the importance of the issues involved in the case. Examples might include: new legal questions that have not yet been decided by any Community precedent, questions about which different NCAs have different views, or cases involving public undertakings of a Member State in which an NCA may be alleged to lack institutional independence or where there is a serious danger that NCAs will deviate from the fixed principles of interpretation and application in EC competition law. The exercise of the power to withdraw a case from the NCAs, in situations such as those mentioned above, can be seen to be an application of a general principle of efficiency, according to which the best placed authority should deal with each case. Such a system would allow the Commission to keep full control of the development of EC competition law on the one hand and, on the other hand, NCAs to fully cooperate with the decentralised application of this body of law. As to the possibility that private parties might challenge the Commission decision to withdraw a case from a NCA: in the absence of legally binding criteria regulating the vertical division of competencies, it is arguable whether such a decision can be considered to be an act open to challenge under Community law. It is suggested that such a decision should not be open to challenge if the procedural rights of the private parties and the substantive rules by which the case is examined are the same both at the European and at the national level.9 EC rules and the different national legislation governing the rights of private parties in such administrative proceedings should be harmonised. In the absence of the envisaged harmonisation, a private party may therefore appeal to the Court of First Instance (‘CFI’) to overturn a Commission decision that withdraws an NCAs competence to deal with a specific case. If the private party is granted interim measures that suspend the effects of the Commission’s decision, the proceeding risks being blocked for an indefinite period. If such interim 9 Temple Lang J.: The new approach to European Competition law, Notes for a lecture to the Comité Européen des groupements de constructeurs du mécanisme agricole.

264

Working Paper VII

measures are not granted, it is probable that the investigation by the Commission and the appeal to the CFI will proceed on parallel tracks. The risk is that, if the CFI upholds the appeal and annuls the Commission decision, the contents of the proceeding before that time would be deprived of any effect. The scenario would be even worse if the ECJ were to reach such a conclusion after a further appeal. Absent the harmonisation, the envisaged problems could be solved through the provision of a short time limit to appeal against the Commission’s decision, and a fast-track appeal procedure directly to the European Court of Justice.

2.

The Horizontal Division of Competencies

2.1.

The Need for Qualitative Criteria

In order to coordinate the action of the members of the network and to avoid parallel proceedings, the draft Regulation establishes principles by which to regulate the horizontal division of competencies. These are as follows: (i) when two or more NCAs are investigating the same agreement or practice under Articles 81 or 82 EC, the fact that one authority is dealing with the case shall be sufficient grounds for the others to suspend the proceeding or to reject the complaint (Article 13(1) of the draft Regulation), and (ii) the case should be always dealt with by the ‘best placed authority’. Much criticism has been brought in response to the absence from the draft Regulation’s text of any kind of criterion to identify which NCA should be deemed to be the ‘best placed’ to deal with a case under EC competition law. Costly parallel procedures, forum shopping, and a lack of legal certainty — both for complainants and undertakings against whom a proceeding is opened — are the main disadvantages envisaged in this respect. Some NCAs strongly recommend including clear criteria of a binding nature in the draft Regulation. However, many authors envisage that, if this were to be done, endless and time-consuming litigation would arise concerning their application.10 Again, the adoption of non-binding criteria may give clear guidance about which NCA may be considered to be the best placed to deal with a specific case. This would benefit NCAs, complainants and undertakings against whom a proceeding is opened. It seems preferable to sacrifice part of the legal

10 Temple

Lang J. (2000): ‘Decentralized application of Community competition law’, in Rivas J. and Horspool M., eds., Modernization and Decentralization of EC Competition Law, Kluwer Law International, The Hague and Boston.

Broad Systemic Issues

265

certainty arising from the provision of legally-binding criteria to the advantage gained by reducing the possible claims based on the correct allocation of a case and, therefore, to recommend the adoption of such criteria through a ‘soft law’ instrument. Such criteria should be general and valid for proceedings brought under Articles 81 and 82 EC, and for those started ex officio and in response to complaints. Moreover, they should be of a qualitative nature and attribute the competence to deal with a specific case to the NCA of the Member State on whose territory an agreement or practice mainly produces its effects (the criterion of the ‘national focus’ of an agreement or practice). A Commission Notice in this respect would be welcome. The horizontal division of competence among the NCAs should therefore be structured as follows: (i) when two NCAs open a proceeding based on EC competition law and dealing with the same case, the best placed NCA should deal with the case; (ii) in order to identify which is the best placed NCA, the ‘national focus’ criterion should be adopted; and (iii) once the best-placed NCA has been identified, the other(s) NCAs should terminate their proceedings, pursuant to Article 13 of the draft Regulation.

2.2.

Conflict of Competencies

Conflicts may arise within a system such as that outlined above. In particular, two or more NCAs may claim to be ‘the best placed authority’ to deal with a single case. Under the current text of the draft Regulation, such a conflict would be solved either by the Commission (which may withdraw the case and it decide itself) or by the interested NCAs reaching an agreement that all of them but one step back and suspend their proceedings. The present situation does not seem satisfactory. As for the solution involving the Commission acting pursuant to Article 11(6) of the draft Regulation, it may be the case that the issues involved do not require its intervention. As for the NCAs reaching agreement, a situation may well arise whereby none of them has the intention to step back, or the matter may become the object of political negotiations which do not lead to the case being entrusted to the best placed NCA. The task of solving such a conflict among NCAs should be entrusted to the Commission. When two or more NCAs believe themselves to be the best placed to deal with a case and they cannot reach agreement within a reasonable time-limit, they should be obliged to refer the issue to the Commission. If the Commission is not willing to deal with the substance of the case itself, it should indicate, within a short time-limit, the NCA that it considers to be the best placed to deal with the matter.

266

Working Paper VII

Such conflicts should be prevented by the draft Regulation’s provision of a duty of cooperation among the different NCAs, and (possibly) by the Commission’s future provision of the necessary criteria to identify which authority is the best placed to deal with a case. It cannot be said that entrusting the Commission with this additional competence will increase its work-load, but it should be seen as an emergency tool to de-block potentially critical situations that might occur in the functioning of the system. In this respect, doubt might also arise as to whether a Commission decision that solves a conflict over competence between two or more NCAs is open to challenge. In this case, and in the absence of legally binding criteria that define the horizontal division of competence, the principles discussed above (in relation to a decision adopted by the Commission under Article 11(6) of the draft Regulation) should apply. If the case is transferred from one NCA to a different NCA where it will be dealt with in accordance with the same procedural and substantive rules, the Commission’s decision should not be open to challenge. The harmonisation of the procedural rights of the private parties in the different national proceedings and, in its absence, the provision of a fast-track appeal procedure — possibly directly to the Court of Justice — are urgent matters to deal with.

IV.

Concluding Remarks

De iure condendo, the uniform application of EC competition law in an environment of decentralised enforcement could be greatly improved if the Commission were to be empowered to delegate cases to one or more of the NCAs. If this were the case, the constant presence of a Commission representative within each NCA would constitute a better guarantee that the decisions adopted by the different NCAs involving the application of EC competition law will be consistent with the Commission’s practice and the jurisprudence of the European Courts. In addition, the experience gained by a Commission representative while working within a NCA will make him more aware of the specific competition issues existing within the territory of a Member State and more familiar with the national competition legislation. The possibility that a Commission representative could work within a NCA, together with the existing opportunity for officials of the different NCAs to spend certain periods within the Commission, will undoubtedly contribute to the development of a common culture, both at a Community and national level. Finally, a concluding remark: it is worth noting that the ‘regulated industries’ sector currently seems to be following a different path to that taken in the evolution of EC competition law enforcement. The latter appears to have proceeded from a centralised system towards a decentralised one, in which the different authorities that have been created at a national level are increasingly entrusted with its application.

Broad Systemic Issues

267

The different industry sectors have had different national regulatory authorities (‘NRAs’), usually one for each industry sector. This has occurred since the start of the process of liberalisation, and was mainly promoted through Community legislation (Regulations, Directives, Article 86 EC Directives and Decisions) within each of the Member States. However, no Community regulatory authority was established, with the obvious exception of the role played by the Commission itself. The necessity for pan-European authorities is progressively emerging, and the degree of support for the draft Regulation that has been received in recent Commission studies (in the field of telecommunications) provides evidence of a push towards a centralised regulatory authority. Indeed, obstacles to intra-Community trade that could arise from different authorisation and control procedures existing at the national level, together with the distortion of competition among Member States that could arise from different regulatory policies being pursued by the NRAs within each Member State, might constitute serious obstacles towards the creation of an integrated European market. A possible short-term solution might involve the Commission itself re-focusing its activities to accomplish the tasks of a pan-European regulator. The reduction of the Commission’s workload that would be brought about by the decentralisation of the application of EC competition law might allow the institution to dedicate more time and resources to the application of the competition rules to such sectors, and to coordinate the action and the policies of the different NRAs.

VIII Giuseppe Tesauro* The Relationship Between National Competition Authorities and Their Respective Governments in the Context of the Modernisation Initiative

The nature of the existing relationship between National Competition Authorities (NCAs) and their respective Governments is among the factors which may substantially affect the proper functioning of the future EC network (‘the Network’), which is about to be created in response to the so-called ‘modernisation project’. Within the new enforcement architecture, NCAs are deemed to share with the Commission the responsibility of enforcing EC competition rules across the EU. Accordingly, it is self-evident that distortions at the national level have immediate repercussions on the consistency of the whole system. Thus, this relationship cannot be simply dismissed as an internal matter of no relevance to the functioning of the network. In order to fully measure the impact that this relationship might have on the success of the overall reform, it is important to remember why a network has been established in the context of the modernisation project. First, the network is designed to enhance coordination between NCAs and the Commission in view of the fact that the former will be granted more extensive powers to enforce Community competition rules. In other words, the reform will mean that NCAs will become the main actors enforcing Community competition rules, including Article 81(3) EC, and there is therefore a need to create a coordinated system in order to avoid duplications and conflicts of jurisdiction. In fact, the goal of securing a coordinated allocation of Community cases between NCAs and the Commission does not seem crucial to the network. I even dare say that, in the light of the past experience, this would probably be the least relevant task of the network. It is sufficient to remember that no case has ever raised a conflict of jurisdiction between NCAs. Accordingly, I assume that most cases would not raise issues of jurisdiction in the future. More specifically, I expect that most cases will be examined by one NCA or the Commission, depending on the prevailing national or trans-border effects of the alleged anticompetitive practice. It is possible that there will be a certain number of cases (though not the most important cases) over which both the Commission and one NCA would claim jurisdiction. I refer to those cases that have an essentially national impact, but in which a relevant Community interest is also at stake * President

of the Italian Antitrust Authority.

270

Working Paper VIII

(such as abuse cases that involve former monopolists in newly liberalised sectors, like telecommunication, energy or transport firms, or postal services). Such cases have arisen in the past and they have always been accommodated without any difficulty. By contrast, there will probably be a small number of cases over which more than one NCA would claim jurisdiction. The network would therefore have different — and perhaps more important — functions, which are to secure cooperation between its members, the consistent application of EC rules, and effective antitrust enforcement across the EU territory. Each actor is indeed required to provide a roughly equivalent level of enforcement in terms of consistency and effectiveness within the network, so to maintain a level playing field. This implies stronger links between NCAs and the Commission, in order to develop a common antitrust culture and a sense of mutual trust. In summary, it seems to me that the main objective of the network is to reach such a convergence. NCAs will ultimately be enforcing Community competition rules just like the Commission, and the Commission will regard them as being on an equal footing. From this perspective, it is clear that the institutional framework of NCAs will be of utmost importance. This is especially the case in the context of the effectiveness of competition enforcement, as the latter depends primarily on how independently a competition agency exerts its role. In light of my personal experience, I believe that the Italian decision to establish a competition agency that is independent from the government has been successful. While this does not imply that different models of enforcement are unsatisfactory, it seems that independence from the government gives more guarantees. With the Italian experience in mind, I believe that independence from the government is beneficial for a number of reasons: i) Independence places the competition agency in a position from which it can enforce antitrust rules impartially. The agency action is aimed exclusively at enforcing the law, in a judicial-like manner. ii) An independent agency seeks no accommodation of interests and values other than competition rules. In a situation of dependence of the antitrust agency on the government, the latter also pursues other goals besides competition policy. Thus, the balance of conflicting interests (for instance, social and industrial policy objectives versus competition) could ultimately upset and impair effective antitrust enforcement if the antitrust agency is dependent on the government. iii) An independent agency is more insulated and less exposed to pressures from lobby and other interests groups; accordingly, there are less chances of it being captured. I don’t intend to imply that government agencies are more prone to capture than are independent agencies. Rather, there can be a risk of capture at the level of government. Indeed, the political process is somehow conditioned by interest groups in any democratic

Broad Systemic Issues

271

society. A government is the expression of a political majority that results from a coalition of more or less structured interest groups. As a consequence, the possible political capture of the government could interfere with the competition agency’s decision-making. iv) An independent agency eradicates the risk of conflict of interest that arises when a government is required to play conflicting roles because the government has different and opposing roles at the same time. This risk typically arises in two situations in antitrust matters. First, the most obvious conflict arises whenever the government maintains a strong presence in the economy by way of state-owned industries. How does a government, as an antitrust enforcer, accommodate respect of antitrust rules with its interest in maximising the profits of its owned firm? Second, a conflict arises when a competition agency happens to scrutinise private anticompetitive conduct that is somehow covered or favoured by public measures. Again referring to the Italian experience, this situation is even more problematic than the first one. The Autorità has extensively enforced competition rules in regulated sectors, pursuing anticompetitive conducts committed by undertakings or associations of undertakings that claimed to act in compliance with all sorts of public measures. As a general approach, the Autorità systematically rejects any claim of immunity from compliance to antitrust rules based on the existence of a legislative or regulatory act imposing or allowing such conduct. In essence, the line of arguments generally put forward by the Autorità is to disregard statutory measures that limit firms’ autonomy in the market. It considers them responsible for conduct that restricts competition whenever it is proved that the public measures being invoked as a shield leave a margin, however little, of autonomy and discretion. In many of these cases the Autorità had faced strong opposition from Government representatives who had enacted the statutory measures that would restrict competition. v) Only an independent agency can exert advocacy powers with credibility, criticising the government and the parliament’s legislative choices in matters related to competition policy. In Italy, the Autorità has extensively used its consultative powers to urge both the government and parliament to remove regulatory constraints that have an adverse effect on competition. In addition, through these interventions, the Autorità has also tried to spell-out general guidelines that could inspire the legislator when re-determining the competitive structures of the markets in a pro-competitive way. In such a way, over time the Autorità has become a sort of guardian of the overall liberalisation process, in so far as it has carefully monitored the activity of the parliament and the government in the implementation of EC directives. Against this background, it is true that the new enforcement system will contain mechanisms designed to correct inconsistent application of Community rules by

272

Working Paper VIII

NCAs. More specifically, in order to permit a timely allocation of cases within the network and an efficient treatment of multiple complaints, the draft Regulation provides that, whenever they apply Articles 81 or 82 of the Treaty, NCAs will have to inform the Commission at the outset of the proceedings and at least 30 days before adopting a decision. The latter prior consultation obligation is only meant to cover cases in which the relevant NCA intends to adopt a decision under Articles 7, 9 or 29(2) of the new Regulation — that is, where it considers that there is sufficient evidence of an infringement of Community competition rules. By contrast, the new system envisages no prior consultation mechanism if NCAs decide to reject a complaint concerning alleged infringements of Articles 81 or 82 EC. This appears to be even more important when due account is taken of the fact that such decisions are much less likely to be appealed than decisions finding a violation. In exceptional circumstances, the Commission also has the power to open proceedings and withdraw a case from the hands of an NCA that is already dealing with it, with a view to ensuring consistent application and effective enforcement of Articles 81 and 82 EC. In summary, while a mechanism of monitoring within the network is envisaged to prevent inconsistent application of Community rules, no controls exist to avoid the risk of a lax and nationally biased application of the same provisions that is aimed at protecting specific domestic economic interests or businesses. In fact, no serious control could be realistically set up to tackle a problem of failure to act, or a more relaxed attitude of an NCA when enforcing competition rules. Nor would it be appropriate to give the Commission a monitoring power in such matters. To these arguments, it could be answered that giving the task of antitrust enforcement to an agency that is under government control is precisely intended to make other more valuable general interests prevail over typical competition based values. However, I am not convinced that such an argument would stand careful scrutiny in the light of the radical changes that have occurred in our economies over the last years. In this respect, the only area of intervention in which there might be scope for serving different public interests (say industrial policy reasons versus competition) in the area of competition enforcement is merger control. However, it is a matter of fact that the importance of the EU merger control system has dramatically increased as a result of the ever-growing globalisation of the world economy and the mergers of a large number multinational companies. As a consequence, merger control is destined to play a marginal role at a national level. Very few cases of mergers that put the national interest at stake are scrutinised at a national level. And while it is true that, under Article 9 of the EC Merger Regulation, the Commission can refer the assessment of a merger to a Member State in presence of some circumstances, such a procedure is not aimed at allowing a Member State to circumvent Community competition rules.

Broad Systemic Issues

273

As a consequence, it is not very realistic to imagine that it still possible to influence the industrial policy by merger control rules. To conclude, within a networking system in which antitrust law and policy are regulated and enforced at both Community and National level, consistency and effectiveness must be a top priority. Consistency implies that Community competition rules are applied coherently within the EU territory, irrespective of the agency that applies them. Effectiveness implies that each NCA attains an equivalent level of competition enforcement in the territory under its own jurisdiction, so as to create a level playing field. Effective enforcement of competition rules depends to a great extent on the enforcement model of a competition agency. In the long run, different competition enforcement models that result from different cultural approaches might jeopardise the proper functioning of the whole system. That is why it might be appropriate to promote a certain degree of convergence towards the model of the independent agency.

IX Roland Sturm* Networking in an Uncharted Territory: The Relationship between the Members of the Network and their National Governments

I.

Introduction

The traditional task of National Competition Authorities (‘NCAs’) is to protect economic competition in the legal set up provided by national and EU law. Decision-making in a supranational network of NCAs, as was envisaged by the Commission’s White Paper on the decentralisation of antitrust enforcement, will not change the legal obligations under which NCAs act. NCAs will, however, be confronted with the double challenge of defining case by case the suitable strategy for challenging market distorting behaviour. They will also have to find a way to communicate this strategy to other NCAs, so that companies in the EU find relatively uniform rules for the economic game in all Member States and can make plausible assumptions about legal obstacles that they might meet when they choose to cooperate with competitors. Decision-making about competition control in transnational networks greatly depends on informality, cooperation, the balancing of national economic interests, and the compatibility of philosophical approaches concerning the role of the state in the economy. Each NCA has its own institutional, historical and cultural background and reputation in its respective society and among national decision-makers. The institutional strength of an NCA also depends to some extent on ministerial support, especially when it comes to the political independence of the NCA. In addition, the legal frameworks of national competition rules are embedded in national ‘competition cultures’, which define acceptable degrees of intervention that competition authorities may make vis-à-vis the autonomy of economic actors. National governments have priorities with regard to the role of competition control. These range from the most general (the securing of a more or less central role for competition in their national economy) to the specific (decisions about cases that affect the future of certain markets, which are seen as being of strategic value for national economies). It is safe to assume that neither general nor specific decisions concerning economic behaviour in certain markets will automatically follow the same path in every country. Differences in national competition cultures can become an obstacle to the successful cooperation of NCAs in a European network. * Professor, Universität Erlangen-Nurnberg, Germany.

276

Working Paper IX

Two questions need to be asked to evaluate the quality of this obstacle to the strategic use of NCA networks for antitrust enforcement in the EU. The first is: when do national decision-makers find it acceptable for the public interest to override competition rules? And the second, more general and related, question is: how much acceptance do markets find as mechanisms for allocating goods in national societies? Answers to both questions provide information about the persistence of national differences with regard to confidence in the benevolent effects of free markets. It is assumed that EU-wide rules for antitrust enforcement will not alone suffice to give coherence to a network of NCAs. In addition, a convergence of national ‘competition philosophies’ is needed — a reduction of diversity of ‘competitive capitalism’ (Milton Friedman) — if similar interpretations of the common European rules are to be secured throughout the EU. Otherwise there is: the danger that decision-making bodies, located in different Member States with different judicial traditions and with differing understanding of and emphasis on competition policy, might decide similar cases in dissimilar ways. The basis for this concern is present already today, and it will likely only increase in the future in connection with the enlargement of the EU (Goldschmidt and Lanz, 2001, 17).

When we know more about the restrictions under which NCAs operate today, we can proceed to investigate the support a policy network may provide for strengthening the autonomy of NCAs vis-à-vis national governments. And we will be in a position to ask whether this network will be able to make a contribution to a European competition culture. In this context it will be necessary to distinguish between different types of networks defined by the degree of integration of the NCAs.

II.

‘Capitalist Diversity’

A simple but helpful distinction for an analysis of the reputation of economic competition as a social force in EU Member States was made by Michel Albert (1993), who juxtaposed models of Rhenish and Anglo-Saxon capitalism. In a more elaborate form we find a distinction of different systems of capitalism in Will Hutton’s (1996) contributions, which can also serve as a first introduction into the debate. For Albert and Hutton, competition as an unrivalled ‘Darwinian’ principle of free markets finds its strongest support in Britain. In other EU Member States, competition as a principle that shapes economic policies is much less dominant. ‘At the heart of the European model’, Hutton (1996, 263) argues, ‘is the notion of a rule-governed, competitive market whose power to generate wealth is intimately linked with social cohesion.’ Competition has a role to play, but prosperous markets are not — as is assumed in Britain — ‘simply the product of atomistic individuals and firms competing fiercely in competitive markets with no state intervention’ (Hutton, 1996, 256). From a EU perspective, the economic

The Network Concept, Competition Authority Networks

277

role ascribed to competition adds up to a diversity of national approaches (Smith, 2001); this diversity is often hidden behind the general terms of the EC Treaty. With regard to the future of social policies — but also in the context of the liberalisation of public utilities, for example — the EU Commission frequently referred to the European social model. Among others, Member State governments echoed this in France (Jospin, 2001) and Germany (Bundestagsdrucksache 14/6249, 3). They stressed that they, as governments, see the defence of their national economic models as their task in the process of a worldwide liberalisation of markets. One focus of this debate has recently been the interpretation of Article 86(2) EC. For France, the future of the concept of sérvice public in an environment of liberalising markets is of crucial national interest. The lack of conceptual clarity when it comes to the definition of the role of markets in national economies, historical legacies, and the range of degrees of trust national governments have in the forces of competition, have so far prevented an EU-wide convergence of economic strategies and identical definitions of the role of competition in national economies. This is especially the case with regard to specific policies such as industrial policies or social policies. In practical terms, this means that regulatory interventions into the markets of EU Member States still take different forms. And the threshold of what is an acceptable exception to competition rules varies. We find public enterprises operating services d’intérêt économique général, regulatory agencies, and self-regulatory quasi-corporatist or traditional corporatist arrangements. There are also marked differences with regard to subsidy levels, although they do not always correspond exactly to the distinction between countries following either the Rhenish or the Anglo-Saxon model of capitalism. Britain has the least subsidised European economy, and Germany the highest level of subsidies both as percentage of all subsidies paid in the EU and per capita (of those citizens in employment). With regard to the level of subsidies per capita, Italy follows Germany, and the next countries down the list are Belgium and France (Bundestagsdrucksache 14/6748, 46). Most economic policy decisions that concern the free market principle are not covered by the competencies of NCAs. Still, they influence the general political climate and are therefore not to be ignored. In a society that is sceptical about the performance of markets, and in which political interventions into the economy can be justified by the assumption that markets are regularly failing, it is difficult to find support for a strong and active role for a NCA whose major task is the defence of competition in its markets. Though it would be an exaggeration to model Britain and Germany as opposite ends of a spectrum of pro- and anticompetitive philosophies, some telling observations can be made through a comparison of both countries. One is, for example, that the liberalisation of markets in Britain in the 1990s was accompanied by a public debate about a reform of competition law, and the result was ‘a truly comprehensive and revolutionary transformation’ (Wilks, 1999, 351). The 1998 Competition Act ended ‘the ‘old’ model of British competition policy’, which was ‘an aberrant hangover from an earlier model of industrial politics.’

278

Working Paper IX

(Wilks, 1999, 350). Another observation is that the regulation of utilities in Britain was given legitimacy by the argument that a major task of regulation was the creation of competitive markets. Though one can criticise the practical significance of such rhetoric (Burton, 1997, 184), its very presence hints at the central role competition has as economic principle, at least in expert discourses. Germany had two opportunities in the 1990s to give competition a more prominent role in public discourse. One opportunity was — as in Britain — utility liberalisation; the other was the privatisation of the state-owned GDR economy by the Treuhand. In the public mind, the work of the Treuhand became above all an exercise of rescuing firms and saving jobs through the guidance and financial support of a government agency. The Cartel Office even had to defend its occasional objections to Treuhand decisions against the argument that merger control in East Germany was itself an obstacle to investment. It had to tolerate (among other things), for example, the use of German taxpayers’ money to subsidise firms so that they could beat their competitors, or so that some firms — especially SMEs — were the object of discrimination (Sturm, 1996, 204ff). Competition as an economic principle did not become a focus, or at least a central criterion, of policies to transform the East German economy. This stood in marked contrast to the expectations raised by the Treaty on Social and Monetary Union (1990). In a protocol that is part of this Treaty between the GDR and the FRG, the first principle mentioned concerning the economic union between East and West Germany is that the German economy shall be primarily based on private property and competition. Another missed opportunity to promote the principle of economic competition to the German general public was the liberalisation of public utilities. Awareness of the benefits of competitive behaviour only exists in the telecommunications market, however this is an asymmetrical market (as in the British case). In the gas, electricity, water and railway markets, the federal government either backs existing champions (railways) and/or hopes that industry negotiations will produce sufficient self-regulation (Stellungnahme 2001a). Negotiated markets seem to be much more convincing for the corporatist leanings of German governments than competitive markets (OECD, 2001, 86ff). One can conclude that it is far from obvious that the EU’s NCAs operate in broadly the same environment when it comes to defining the acceptable limit of state interventionism and the role of competition in national economies. Of course, much of the pressure of public opinion can be ignored, and most of the time the role of the courts is much more important to the work of the NCAs in deciding specific cases then are the pressures of public opinion. However, there are two important consequences of the current state of affairs in the EU: (a) public opinion is only in rare cases (and not in every country) a reliable partner that can support NCAs in cases of conflict with the corporatist leanings of national governments, and (b) political interference is easier to justify today, because competition is seen largely in the context of specific cases to be decided. These cases are connected to national interests and are therefore not seen to be demonstrating a general market principle, as was the case in some countries in the past. The best

The Network Concept, Competition Authority Networks

279

example of this great historical change is found in Germany. The Cartel Office was originally intended to protect both the constitutive and regulatory rules of competition that governed markets (Ortwein, 1998, 67ff). Ludwig Erhard and Alfred Müller-Armack regarded competition as part of the economic constitution of Germany and as the most efficient mechanism to steer markets. When the federal government today reports on the ground rules of competition policy (Grundsätze der Wettbewerbspolitik), it refers to new legislative developments (Stellungnahme 2001) and not to the role of competition in the national and European economies. In its latest report, the Cartel Office mentions that it debated a ‘more economic approach’ when looking at cooperation between competitors (Bericht 2001, 9). This seems to be one way of saying that, in an age of globalisation, competition is of less importance as a general principle. More important is the new flexibility of the cartel law, which allows the strategic partnerships that are now deemed to be acceptable even if they reduce the number of competitors and thereby the level of competition.

III.

Public Interest v. Competition Rules

It is not only in Germany that competition policy is not solely directed towards keeping the principle of competition alive. NCAs act in accordance with a wide range of motives. They may be political, social, cultural or economic. Discretion is left to the respective NCAs. Moreover, even where the goals are economic in nature, this does not necessarily mean that policy strives to achieve a higher degree of ‘competitiveness’. Economic efficiency can also be defined as the strengthening of big business by enhancing its potential for mergers and acquisitions in order to create global players and national champions that are able to use synergies and to produce at low cost. EU countries have a history of incorporating strategies of industrial policy into their competition policies. At least their respective legal frameworks provide for ‘escape clauses’ that allow industrial policy and other considerations to supersede plain competition-oriented policies (Sturm and Müller, 2001, 11f). These escape clauses might take the form of the last word of the responsible minister. They might also take the form of a more general reference to the public interest in legal texts. The definition of ‘the public interest’ remains a national exercise. Even if the first step for the Europeanisation of competition control — namely, a general agreement about the role of the state in the economy — were to become a reality (maybe because of the pressures of globalisation that affect all EU Member States in a similar way), the reference to ‘public interest’, which is so far not a European interest, remains divisive from the point of view of the international cooperation of NCAs. As long as this reference allows decisionmakers to ignore competition rules, a harmonisation of the outcomes of national competition control policies is not guaranteed. It is probably not sufficient to change legal texts in order to exclude interventions based on national special interests. Stephen Wilks (1999, 353) explains how such

280

Working Paper IX

precautions can even be perverted. The ‘public interest’ test was replaced as a formal criterion for decisions concerning the British Competition Act by the (admittedly fuzzy, but well-chosen) test of ‘effect on competition’. It is telling for the power of national competition cultures that one observer was quick to point out that ‘effective competition’ is best understood as competition that serves the public interest. In order to enable the NCAs to play a constructive role in a European network of NCAs, they need to be given as much freedom as possible from political interference so as to avoid conflicts about how to address the cases before them. One might hope that the Article 81 or 82 EC cases are less spectacular for the general public than the merger cases, and that they are of less interest to national governments. The current arrangements and the practice of government interventions seem to point in this direction. However, the special permission to engage in anticompetitive behaviour that can be granted by the German economics minister, for example, covers both antitrust policies (Article 8 GWB) and merger cases (Article 24 GWB). It might be useful to exclude a second (political) stage of competition control in all EU Member States. The NCAs should be given the last word (excluding court decisions, of course) about the cases. What should be ruled out is any kind of procedural involvement by other actors. NCAs will have the chance to develop converging practices of antitrust policies only if their autonomy is guaranteed.

IV.

Policy Networks and the Independence of Network Members

As policy analysts tell us, networking is more than an additive process (see, for example: Jansen and Schubert, 1995). Fifteen NCAs are not automatically a network. Cooperation between members on a regular or ad hoc basis is also less than cooperation in a policy network. Some theoretical articles assume that partners in networks never cooperate voluntarily, but only because they need the abilities and resources of others to solve their problems (Messner, 2000, 111). This may be disputed. But, at the very least, networks do need a purpose and a role. These can be defined by Articles 81 and 82 EC. Networks also need to be structured, they need rules for efficient cooperation, and they need a strategy to deal with conflicts (in a network, non-cooperation — the exit option — no longer exists as a useful strategy for conflict avoidance). Supranational networks will be able to create greater independence from the constraints of the national polity for their members only if they succeed in developing an organisational identity and a functional legitimacy for themselves. It is necessary to guarantee the core competencies of the network in order to enable the network of NCAs to develop an organisational identity. In other words, the Commission should not devolve Article 81 or 82 EC responsibilities to the network on the one hand, while also reserving for itself the right to deal with devolved matters whenever it believes this is necessary. It is also important to find an organisational expression for the network on the European level that allows the regular exchange of information, and debate about principles of

The Network Concept, Competition Authority Networks

281

competition policy and common strategies against political interference. National competition cultures enter into the supranational cooperation of NCAs, but they can also be shaped by joint decisions of NCAs, and in this way they can become part of a European competition culture. A similar process is under way in another economic sector — budgetary policy-making — where the Maastricht criteria have helped to make a culture of financial stability acceptable in all EU countries. Alternative Concepts of NCA Networks Role definition:

(1) steering

(2) moderating

(3) informing

EU level: Member States level:

Integration Convergence

Structured dialogue flexible adaptation

cooperation status quo

The alternative concepts of NCA networks discussed here are not mutually exclusive. However, it is obvious that the greater the degree of ‘integration’ among NCAs, the easier it will be to overcome the pressures of capitalist diversity in the EU and the restraints of national competition cultures. The choice of a concept for a NCA network depends crucially on the amount of stable exchange among decision-makers one wants to institutionalise for the interpretation of Articles 81 or 82 EC. If predominance of a joint European approach is to be achieved, a high degree of institutional and intellectual integration is needed (for example, the regular exchange of staff). In this way, it is probable that there will be a convergence of NCAs’ views with regard to competition control. Jointly developed ideas can become norms for NCAs’ decisions. Regular interaction at the European level produces a steering effect. Regular interaction in a network of NCAs is not to be confused with a nucleus for a European Cartel Office. Integration in a network is voluntary and case-driven, both with regard to the NCAs involved and the branches of industry affected. However, there should be a common interest in finding general rules that can guide the behaviour of NCAs and that can be presented to national governments as a form of acquis communautaire. Below the threshold of self-regulatory steering, a network of NCAs could also function as a clearinghouse that moderates between different national views on antitrust cases. NCAs could meet whenever conflicts arise with regard to the interpretation of antitrust cases or when an NCA takes the initiative to clarify more general problems of competition control. An organised and structured dialogue between the 25 NCAs would be necessary in order to secure comparable standards of competition control. At the very least, a certain level of constant and common awareness of the activities and philosophies of the other NCAs should accompany conflict moderation. Though these efforts would produce a ‘weaker’ network than the one that results from a convergence of competition cultures based on an integrative approach, the flexible adaptation of their antitrust decisions to European standards would become an option for the work of NCAs. The third alternative for the NCAs is to work in the context of the status quo. Bilateral and multilateral cooperation can hardly claim to have the quality of a

282

Working Paper IX

network. Ad hoc cooperation will not help to transcend national boundaries efficiently or to embed NCAs adequately into a supranational context that increases the independence of NCAs from national governments or that makes them part of a common decision-making culture. Though one could argue that the principle of subsidiarity implies the possibility of, and may be even the need for, national diversity, de facto competition with regard to the interpretation of EU antitrust rules is counterproductive in a single market.

V.

Perspectives

The debate on the effects of a decentralisation of EC antitrust enforcement is difficult to separate from stocktaking with regard to the significance of the capacities for merger control that remain at the national level. Of course, antitrust and merger control decisions can be separated for technical reasons, but the standing of a NCA vis-à-vis its national government can not be divided in the same way. Either the NCA is regarded as having an important role to play, and then its voice will be heard as the guardian of the economic constitution of a country, or it is seen as an ornament to the substantive process of decision-making, which is controlled by Brussels or by national governments. In other words, whatever beneficial effects a supranational network of NCAs may have, this will not change the problem they currently experience, which is that they lose influence because they decide the less significant cases compared to the work done by the Commission. In addition, as long as the Commission decides and there is no decision-making in a ‘non-political body’ such as a European Cartel Office, national governments see a natural political partner in the Commission, who — in their view — is as open to political horse-trading as they are used to in other policy fields. This has the effect that governments are less interested in enforcing market rules and strengthening competition in general, and are much more interested in defending a diffuse public interest that is often defined for them by industrial lobbyists. In this political game, NCAs end up as specialist institutions who are no longer respected as guardians of the economic constitution and whose advice is only of marginal influence when structural reforms to national economies are decided upon. Whether the optimists are right and a network of NCAs gives them a second chance to improve their standing in the EU and vis-à-vis their national governments not only depends on the reputation of the NCAs, but also on the development of a European competition culture. Convergence of national interpretations of the relationship between the state and the market is only one aspect of this. Even more important are a greater respect for competition as a constructive force in all European markets, and a body of competition law (including maybe a less active role by the courts) that does not unduly restrict NCAs freedom of action.

Technical Problem Areas

285

Panel Three: Technical Problem Areas

JOHN FINGLETON — When talking about a large European project, it seems appropriate to quote Bismarck. Asked how to solve the Irish problem, Bismarck said that the solution would be to swap the populations of Ireland and Holland: the Dutch would make Ireland into an industrious and prosperous nation that would feed the rest of the world, and the Irish would be careless and fail to maintain the dykes, and would drown. I think we have all moved on a little since then, but a question that remains valid is, how is the network going to function in light of the cultural diversity in Europe? I personally think that Europeans work very well together nowadays. Returning to the topic of my presentation: since I am not a lawyer, but an economist, I have a lot of questions about how this network is going to operate. However, that is not to say that I am not enthusiastic about this project. My first question is about the legal status of the decisions concerning case allocation. As I said yesterday, there are two issues that I see coming up here: who does what? And how do they do it? If the outcome of a case can vary depending on who deals with it, then the rights of the undertakings concerned may also be affected. We need to clarify this. It is not apparent to me whether the parties will be able to challenge case allocation decisions. Yet whether they can or cannot do so will affect how the network itself operates. Unless one authority knows something more than the complainant does, a case ultimately has to be allocated on the basis of what the complainant argues in his petition. I think that we need greater clarity, particularly with respect to the complainants’ rights, and we should even see whether a harmonisation of the complainants’ rights across the national jurisdictions within Europe is necessary. Ultimately, this is a question about procedure, and about the harmonisation or non-harmonisation of procedures across the Member States. With regard to the rights of the defendants, I think that the issue of sanctions is probably the most important. In other words, if the main issue for complainants is one of procedure, for defendants the main issue is one of sanctions. Returning to case allocation, in practice a case may have be allocated either between the Commission and the competition authority of a single Member State, or between the competition authorities of several Member States, or between a group of national competition authorities and the Commission. With respect to the first type of case allocation, between the Commission and the competition authority of one Member State, the Commission has had the power to take up a case from the national competition authorities since 1962 (as Alexander Schaub mentioned yesterday). This is an issue that may be of greater concern to larger Member States than to smaller ones: if an anticompetitive agreement or behaviour has an effect in Germany, it might be of more interest to the Community than if it had a similar effect in Ireland. Clearly, its effect on intra-community 䉴

286

Panel Discussions

trade is going to be the determining issue. The cross-border effects doctrine is not relevant here, because the effect is felt in one Member State. The effects doctrine is relevant to the allocation of cases among the EC Member States, not between a Member State and the Commission. An issue about which I have a question mark is whether the Irish competition authority, for example, could actually ask the Commission to take up a case simply because it cannot itself impose sanctions. If you take the Irish Sugar case as an example:1 after the implementation of the reform, the Irish competition authority would be able to look at the case but not to impose a fine. (The fine imposed by the Commission in Irish Sugar under Article 82 EC — then Article 86 EEC — was 8,8 million Euro). So my question is: could we investigate the case and then pass the file on to the Commission, asking them to make a decision? Is that an option for national competition authorities wanting to deal with the lack of effective sanctions in their national legal systems? The Irish competition authority is often asking the Commission to take up cases, and not because it does not find them interesting, but because of lack of effective remedies available in its own jurisdiction. It takes a long time to get a case through court in Ireland, and we do not have the ability to impose fines or other remedies comparable to those available to the Commission. Similar situations might occur under Article 86 EC: we might run into a state restriction that the Commission would be in a much stronger position to challenge. As to the allocation between the competition authorities of the Member States: here the effects doctrine applies. I think this will be the most exciting area of enforcement in the future, when a group of national competition authorities will have the opportunity to work together on a case. I think that there are already subgroups of Member States whose markets are already much more closely integrated than in the rest of the EU; for example, Ireland and the UK, or the Czech and Slovak Republics. In some of those groupings the procedures and sanctions may also be very similar. Consequently, when we talk about groupings or pairings of competition authorities, it is quite likely that these will not happen in a random way, but rather within some fairly coherent group of competition authorities. And I think that this is probably how we are going to start cooperating within the network. For example, the Irish and UK authorities will start working more closely, or the Belgian, Dutch and Luxembourg authorities will do the same. In my written contribution I also made a note on the four Nordic States, questioning whether in their case we should not allow for greater flexibility. I also have some questions about how these joint teams will operate; these are technical questions that we need to clarify. I think that the Commission and the Member States should agree on a new notice on market definition. Market definition is going to be crucial in the new enforcement system, and if the national competition authorities do not share 1 Commission Decision of 14 May 1997, IV/34.621, 35.059/F–3 Irish Sugar plc, OJ L 258, 22/09/1997, p.1 et seq.

Technical Problem Areas

287

the same fundamental principles about market definition, the issue of case allocation will become more complicated as well. We do not want to see situations in which one national competition authority says it does not take into account supply substitutability — that would be an invitation for undertakings for whom supply substitutability is essential to seek to have their cases heard there rather than somewhere else. I would very much hope that the first sign of our ability to deliver within the network would be the adoption of a joint notice on market definition. I do not see big problems with the existing Commission notice on market definition, but we might need to change a few things here and there in order to take our collective experience into account. As to the third type of case allocation, between the Commission and a subgroup of Member States, I think that decisions will depend on how well the subgroup works together. This is a type of decision-making that will become smoother over time, and it is not going to be a big issue. In my written contribution I raised the issue of conflicting decisions. I wonder whether that is going to be interpreted conservatively, meaning that the same mind could not possibly have reached the two conflicting decisions. Or will it be interpreted broadly, meaning that incompatible decisions cannot be imposed upon the same undertaking(s). I prefer the broader interpretation, as it leads to more experimentation and learning, and I think it allows the national competition authorities to reason independently from the Commission, perhaps even challenge the Commission’s reasoning, find innovative remedies, and so on. Another issue that I touched upon in my written contribution is that of resources. Resources will be required for doing a whole range of things in the new enforcement system. The first of these is the actual investigations. This touches upon a point that I made yesterday, about whether the whole purpose of the reform is to do what we are currently doing at a lower cost — which I do not think is the case — or whether it is to improve the quality and quantity of what we are doing. We need resources in order to improve the quantity of the decisions. The national competition authorities are going to have more work to do in the new enforcement system, and not all Member States seem to be aware of this. Resources will also be needed for the operation of the network itself. For example, who is going to read all the information circulated within the network? For example, when information is sent to the Commission, will it just be filed or will somebody actually read it? It would be good if the information were actually to be read. Another additional cost would result from translations, albeit into a single language. Another issue is that of staff exchanges: I would be very surprised if any competition authority had staff to spare for exchanges. To sum up, these are some of the technical issues that I mentioned in my written contribution, and which I hope will be discussed within the Council Working Group. I would conclude by saying that, ultimately, there is a self-equilibrating mechanism in the functioning of the network. I also believe that the system is going to converge towards efficiency, but that may require some harmonisation of sanctions, and possibly of procedures. As an economist, I would put emphasis

288

Panel Discussions

on the harmonisation of sanctions because they have an important overall economic effect on the functioning of the system. The more we can do up front in sorting out these problems, the faster the convergence will occur. As I said before, convergence will occur anyway, but the solving of these problems would accelerate the process. I believe that, even in the absence of the modernisation project, we would have assisted with the development of competition law enforcement at the Member State level. IAN FORRESTER — I must first confess that I had some difficulty in composing my intervention for this workshop, and I do not offer my thoughts with great confidence. Subject to that warning, I would observe, first of all, that the old rules functioned adequately and did their task in their time. They were not logical, they were not coherent, but they were made to work by talented officials and energetic lawyers. The new regime will certainly have imperfections, and I assume that at least twenty percent of what has been proposed in the draft regulation is inadequate or wrong, and will have to be changed. The same happened with the merger regulation adopted twelve years ago: some of it was good, and a little bit of it had to be altered. However, I am not troubled by the idea that you won’t get it right the first time. I do think that building flexibility into the new system is highly desirable. The biggest news is that the reform is under way. Some people around this table, such as Prof. Ehlermann, Karel van Miert, Alexander Schaub, and others, deserve the credit for a process that has led to a bloodless reformation without the Inquisition. This is to make an analogy with one of my old remarks about the laicisation of Community competition law, with DG IV as the Vatican pushing out onto the laity the enforcement of the EC competition rules. Yesterday we heard two very diverging points of view: on the one hand, Mario Siragusa said that the Commission is losing its control and we need order and predictability. On the other hand, we heard Frédéric Jenny and Ulf Böge say that the national competition authorities will be placed in a subordinate position by the Commission. I hope both these views are wrong. I presume that reality lies somewhere in the middle. I also presume that the Commission will use the power to take up cases from the national competition authorities, and will exercise its powers with respect to case allocation in an intelligent way. At the same time, I do not worry too much about legal certainty in the new enforcement system. Competition lawyers have done fine for forty years without legal certainty. I do worry about quality control and homogeneity, for reasons that I have set forth in my written contribution. We can already see instances of naked use of political power and problems of quality control in the current Member States, and those will be amplified after enlargement. The competition authorities in the Eastern European countries apply a formalistic, mechanical competition law that does not have much to do with what is going on inside the EU. They need the assistance of the competition authorities in the EU in resisting political pressure. But sometimes not even the competition authorities inside the EU succeed in resisting political pressure. 䉴

Technical Problem Areas

289

So, let us get the network functioning as quickly as possible, so that the competition authorities of the new Member States will have the support of its members. It is easy to conceive clever legal points about inconsistency and so on. The new rules are not perfectly clear, but I don’t think they could be, given the current state of affairs. It seems to me that, when confronted with difficult decisions, the Commission is telling the national competition authorities: ‘Do your best. Work on it. We are not going to intervene quite yet.’ I think this is an appropriate and understandable approach. Lawyers will take every chance they can get to make procedural trouble. So do not assume that the lawyers will give you an easy ride; they will not. But that is fine. Controversy will make decisions, and out of the decisions will come adequate results. I did a little bit of anecdotal research during the course of preparing my written contribution, and I spoke to officials of about ten national competition authorities. They all said they meet with each other as if they were members of a family, and so on, but when asked if they ever cooperated in allocating a case, not one said ‘yes’. I may be wrong, and I am sure there are instances where the national competition authorities have agreed on an efficient allocation of cases, but it is not enough to just meet. It is necessary to do things together. Family weddings are nice, but families don’t (always) run a business. Lawyers also meet and discuss things — for example, in the context of practising law, when a conflict arises between clients of a law firm — and the issue is then discussed by looking at the relative advantages and making a rational decision. This model can also work, I believe, for competition agencies. There is lots of ground for hope. Practitioners are appearing regularly before the national competition authorities, and there is a desire to make the reform succeed. In my personal experience, the national administrations also have great respect for the decisions taken by other national administrations. Mario Siragusa, Frédéric Jenny and Ulf Böge raised beautiful questions to which I do not know the answers. But I believe that, on the basis of common sense and Europe’s capacity to deal with the illogical, we will have a successful reform. 䉴 ANNE WILLEM KIST — I empathise with Ian Forrester’s views on the subject, so I would like to start with an anecdote: there was a legendary figure in my university, who had the reputation of being what we call in Holland an ‘eternal student’. He was to become a medical professional. It took him years and years to graduate, and in the end, he started his own practice. He chose to locate his practice next door to a very highly reputed general practitioner. On the door of his own practice he put a plate: ‘For difficult cases, go next door.’ This is exactly how I feel when dealing with this subject: for the difficult technical issues, go to others. Now allow me to make a few observations about the issues discussed in this session. First of all, I think that we should consider in more detail what the exact scope and effects of Article 12 in the draft regulation will be, and particularly its likely effects on national procedures. This is, of course, an issue that will develop through case-law, but I think that we should nevertheless give it some more consideration before the adoption of the new regulation.

290

Panel Discussions

Secondly, it is clear to me that we need to have as much clarity as possible when the new enforcement system is introduced. It would be impossible to arrive at the necessary degree of clarity if we want to implement the reform soon. Therefore we should find a balance between the objective of having as much clarity as possible as to the terms of cooperation within the network, on the one hand, and relying on emerging case-law and best practice procedures, on the other. Maybe practitioners will not be very happy with such a compromise but, as Ian Forrester said before, did we ever have it any better? Thirdly, what will be the position of undertakings in the event of an information exchange within the network? One possibility would be to make an administrative decision as a preparatory phase: if one does not agree with it, one can always challenge it in court. Another would be to adopt a solution similar to what we have in the Dutch legal system with respect to the application of both EC and national competition rules. The Dutch system allows the interested parties to go to a specialized competition court in summary proceedings. Furthermore, the Dutch competition authority is under a legal duty to announce to the parties involved that it is about to exchange a certain kind of information, this gives them the opportunity to go to court. I am not sure whether this is the best solution, but a solution must nevertheless be adopted. Finally, a very complicated and thorny issue arises from the fact that nowadays we see various European countries moving in the direction of criminalising certain aspects of antitrust law. Other countries, such as the Netherlands, are determined to maintain antitrust in the administrative sphere. This oncoming divergence may very well raise thorny and difficult issues with respect to the exchange and use of information within the network. For example, I am not sure that a Dutch court would accept evidence obtained under criminal law proceedings in another EC country, or where the other competition authority has powers that the Dutch authority does not have. Some may argue that any national legal obstacles to the use of evidence obtained through exchange within the network should be set aside in order to give full effect to Article 12 in the draft regulation. But this is an arguable position. 䉴 CLAUS-DIETER EHLERMANN — This is indeed a technical issue of great practical relevance: during my six years at the WTO Appellate Body, I came to appreciate that it is essential to have rules concerning the exchange of confidential information, otherwise the WTO dispute settlement system can be brought to a halt. However, there are no such rules at the WTO. 䉴 MARC VAN DER WOUDE — I would like to start with some general observations. First, under the present rules, as defined in the Spanish Banks case,2 the

2 Case

C–67/91 Dirección General de Defensa de la Competencia (DGDC) v. Asociación Espanola de Banca Privada (AEB) and others (‘Spanish Banks’) [1992] ECR I–4785.

Technical Problem Areas

291

information gathered in the context of one proceeding may not be used as evidence in another one. However, such information may be used as a reason to initiate new proceedings under either EC or national law. This has important consequences in terms of designating the authority competent to deal with a case: one should be careful not to send information to those authorities that may, in the future, be put in a conflict of interest situation. Furthermore, problems may arise if a competition authority uses the information received for purposes other than the enforcement of competition law (for example, for commercial negotiations). My second observation also concerns conflicts of interests. In antitrust enforcement we have, on the one side, the interests of the competition authorities, which want to obtain the maximum amount of relevant information for solving a case, and who do not want to be disturbed by anybody — and certainly not by lawyers — in the gathering of such information. On the other side we have the interests of the companies concerned: those from which the evidence originates, and whose interests are obviously different from those of the companies against which these documents are going to be used. Somewhere a balance has to be reached between these conflicting interests. Thirdly, I would like to say a few words about differences between the present and the new enforcement system in terms of exchange of information. In particular, the new enforcement system will be characterised by horizontal flows of information. This contradicts the current rules: in Spanish Banks, the ECJ specifically ruled out the exchange of information among competition authorities. The draft regulation not only envisages such an exchange of information as essential for the functioning of the new system, but also requires that any rule of national law that would prevent such an exchange be set aside. This sets aside the Spanish Banks case-law. However, the draft regulation speaks about a right to exchange information, and not about a duty to do so. The limits to the exchange of information are set out in Article 20 of the draft regulation. They re-confirm the existing case-law on the subject; see Dow Benelux.3 Curiously enough, the current limits will continue to apply with respect to the Commission services, but not in relations between the national competition authorities that are part of the network. Of course, for the latter there is another limit, insofar as the information obtained may be used only for the purposes of applying EC competition law. But what happens if the information is sent to another authority that intends to apply national competition law? May the latter authority use the information obtained in order to initiate proceedings under national competition law? Here Article 3 of the draft regulation also comes into the debate. What is its interface with Article 12 in the draft regulation? Finally, the draft regulation establishes that the information obtained through the network may be used only for applying financial sanctions.

3 Case

85/87 Dow Benelux v. Commission [1989] 3150.

292

Panel Discussions

And now to the more provocative part, which concerns technical problem areas as seen from the perspective of a practicing lawyer. The first problem that I identified is as follows: is Article 12 of the draft regulation a valid provision? In both Spanish Banks and Post Bank4 — the latter being a case that concerns the exchange of information with national courts outside the scope of Regulation 17/69 — the European courts do not make reference to Regulation 17/69, but to primary Community principles: the rights of defence and confidentiality obligations. One could easily say that Article 12 in the draft regulation is not valid, but that would go too far. There are means to reconcile the need to protect the rights of defence with the objective of promoting a free exchange of information. I would say that the interested companies have three rights: first, they have a right to the highest standard of defence. This includes, as Mr Kist has already suggested, the principle that a competition authority may not rely upon evidence that it would not have been authorised to gather in its system. Secondly, the company from which the documents originate should have the highest level of protection. This is not only an issue of confidentiality, but also one of access to public documents. In some countries, such as the Netherlands, the system of access to public documents is very generous. Does this mean if I were to send some documents to the Dutch competition authority under the Dutch law, a third party could get access to the documents, which are perhaps are not confidential in the Community’s legal conception, but which nevertheless I would not like to be made public? Here again, I would support establishing a right to the highest level of protection. (I feel similarly with regard to the issue of legal professional privilege, but I think that Mr Santiago Martínez Lage will speak about that later on). Also important is the right to having the most favourable criminal law applied, especially when we speak about leniency programs. For the moment there is no harmonisation of these programmes, and I do not think it will be possible to harmonise them in the future either. In summary, if someone sends information to one national competition authority, s/he should be entitled to the same conditions under which the information is granted in the country of destination. One could even push it as far as to say that, if the level of protection in the country of destination is higher than in the country of origin, the information provider should benefit from the higher protection available in the country of destination. Finally, I would support the introduction of a right to the complete transfer of the file. In complex cartel cases, particularly when the prosecution relies on circumstantial evidence, exculpatory documents can be very important. So, if a file is sent from one competition authority to another, it should include both incriminating and exculpatory documents. To conclude, if these rights — the right to the highest standard of protection, the right to the most favourable conditions, and the right to the complete transfer

4 Case

T–353/94 Post Bank v. Commission [1996] II–926.

Technical Problem Areas

293

of the file — are respected, Article 12 of the draft regulation would be compatible with the ECHR on the rights of defence. Then the question arises: how can we safeguard these rights? I think that Mr Kist has already given the answer: before documents are sent, the company from which the documents originate should be informed about the fact that this information will be sent and given the opportunity to oppose the transmission of the documents. This is nothing new; it is a procedure that already exists (we saw it applied in Post Bank). I do not see any reason why this should be changed, and I think it can be made compatible with a smooth exchange of information by imposing very strict deadlines on the companies involved. I think that in many cases the rights of defence will be protected in the country of destination, but the interests of the company from which the documents originate will not necessarily be protected in the country of destination. FRANCES BARR — The written contributions prepared for this workshop and the discussions we have had so far show that this is an area raising a great many issues. The ones that I would like to cover in my presentation include:



• the type of information covered by the provision in the draft regulation on the exchange of information within the network; • possible practical problems in the application of Article 21 of the draft regulation; • possible implications of the exchange of information for the rights of defence; • the impact of the exchange of information within the network on the effectiveness of leniency programmes; and • possible consequences of the existing differences between national procedures and sanctions in the new enforcement system. This is long list, and obviously I shall only set out some preliminary and personal thoughts that are intended to highlight some possible practical problems and consequences of the exchange of information within the network. These issues are covered in more detail in the paper prepared by my colleague Margaret Bloom for this workshop. As noted, these views are all personal. As to the type of information covered by the exchange of information within the network, Article 12 of the draft regulation refers to the Commission and the Member States having the power to provide one another with information, including confidential information, and exchange evidence on any matter of fact or law. Article 21 in the draft regulation refers to investigations, and in this context the reference is simply to ‘information’. The question is therefore: what type of information is actually covered by these provisions? And, what is actually protected? The draft regulation uses the concept of confidential information and makes some qualifications to this concept: first, Article 27 sets out protection for

294

Panel Discussions

information of the kind covered by the obligation of professional secrecy. In addition, the protection of business secrets in the process of access to the file is recognised in Article 26. The obligation to maintain professional secrets is established by the EC Treaty itself. The idea that particular types of information may require special treatment in administrative proceedings has also been developed in Community case-law. Similar concepts are also dealt with in national laws, and these may provide a different type of coverage and protection. In the UK, for example, the Competition Act establishes a prohibition system modelled on Articles 81 and 82 EC, and the regime defines confidential information as commercial information, the disclosure of which would, or might, significantly harm the legitimate business interests of the undertaking to which it relates. This is very different from the concept of professional secrecy in the EC Treaty, which refers to information about undertakings, their business relations, or their cost components. The differences between national laws in this regard could have an impact on the actual nature of the information that can be collected and exchanged within the network. In addition, there are also differences in national laws with respect to the powers of investigation of the competition authorities. These include, for example, differences as to the power to gather different types of information, the ability to ask those under investigation for information, the scope of the protection provided in areas such as legal professional privilege, the type of premises that can be inspected, and so on. In the UK, for example, the Office of Fair Trading can carry out on-site investigations on domestic premises provided that these are used in connection with the business of an undertaking, or documents relating to the business are actually kept on the domestic premises. Article 21 in the draft regulation requires that the competition authorities of the Member States assist each other as far as national procedures allow. This may mean, for example, that if an investigation is carried out in the UK on behalf of another Member State, our investigators would not take documents that had been prepared by an in-house lawyer. These would be protected by legal-professional privilege under UK law. I think that this may well have an impact on the actual effectiveness of the mutual assistance envisaged by Article 21 in the draft regulation. Information obtained under Article 21 can be exchanged and used in accordance with Article 12 of the draft regulation, which allows the receiving authority to use the information even in situations in which it could not have been obtained by its own on-site investigation. In this respect, the only limitation appears to be in terms of the sanctions that can be imposed. Taking the example of a document produced by an in-house lawyer: this would mean that this document could be used by the Office of Fair Trading if it was information received under Article 12 of the draft regulation following an investigation carried out by the competition authority of another Member State at the request of the Office of Fair Trading based on Article 21. The same document could not be used if the Office of Fair Trading’s investigators had themselves come across it during an investigation in the UK.

Technical Problem Areas

295

This is my interpretation of how Articles 12 and 21 interact. However, looking at the written contributions prepared for this session, and taking into account the discussion we have had so far, it seems to me that others have a contrary view. So I think that this is an issue we could examine in more detail. If the issue is not settled in the new regulation, I fear that lawyers will challenge our (the Office of Fair Trading’s) acts relying on these provisions. Furthermore, lack of clarity in this respect this may also have implications for the rights of defence. For example, will undertakings have a right to make representations about the use of such information during national legal proceedings? If so, when would those rights actually arise? Would this be when the investigation is carried out on behalf of the requesting Member State in accordance with Article 21, or when the information is being used by that requesting Member State to pursue the case? These are all issues that arise when the information obtained through the exchange is being used for the actual purpose for which it was acquired under Article 21. I have not considered additional concerns, that may arise when information exchanged under Article 12 is used for another investigation, albeit in relation to the application of competition law. However, the issue has been tackled in other written contributions for this session. I have also not considered the nature of the rights of defence, which I believe to be a very complicated area. I would like to turn instead to an area that is slightly more straightforward, namely whether the exchange of information within the network will have an impact on leniency programmes. Leniency programmes operate on the basis of confidentiality of the information that an undertaking has provided in return for a grant of immunity from sanction. It is generally accepted that this confidentiality can only be respected for a limited period during the investigation. In the UK we have a new leniency system that has been operating for a couple of years. It is extremely effective but, at every single stage of the proceedings, companies argue about the confidentiality of every single aspect of their application. I think that the ability of the competition authorities to exchange information within the network might mean that an undertaking would apply for leniency in every Member State that has a leniency program. This will happen because undertakings know that the information provided as part of one leniency application could be exchanged with other Member States in accordance with Article 12, even though no immunity would arise in any other jurisdiction in the absence of an application in that particular Member State. If a case has already been allocated by the network, the undertaking will know who will be handling its case, and therefore the choice of where to make any application for leniency will be fairly straightforward. But if a case has not been allocated, the undertaking will need to make a choice of where to apply for leniency, taking into account the considerations I have just outlined. For these reasons it may actually decide to apply first to the Commission. I think it will be very interesting to see what impact, if any, this choice — which is prompted by potential concerns about information exchange — may have on the operation and effectiveness of national leniency programmes.

296

Panel Discussions

Considering the issues that arise in relation to information exchanged within the network, one might conclude that significant differences between the laws of Member States cannot actually be sustained in the new EC competition law enforcement system. Most significant in this respect are differences between civil and criminal law procedures, and those concerning the scope of sanctions. I do not think that these differences are necessarily a problem right now, so I agree with Céline Gauer and I share Professor Ian Forrester’s optimism about the new system. But it may be likely that pressure to harmonise national procedures will increase as organisational reforms take effect. And this could occur by ‘soft harmonisation’, for example, by adopting practices and procedures developed as the experience of the network grows, and through continued discussions such as the one here, and other discussions within the ‘family’ of the national competition authorities. But it may be that only formal changes to national laws will produce a practical and workable system, or that there will be a need for further reform of EC competition law. 䉴 SANTIAGO MARTÍNEZ LAGE — My intervention will focus on four points related to the exchange of information. First, my colleague Helmut Brokelmann and I think that there is a need for a common notion of confidentiality and business secrecy. Under Article 26(2) of the draft regulation, access to the Commission file is limited by business secrecy principles and excludes confidential documents and internal documents. This provision refers only to procedures before the European Commission, so it does not apply to procedures before national competition authorities. We think that it would be necessary to harmonise this concept in national laws in order to avoid the loss of information’s confidential character as a result of the transfer of documents from one national competition authority to another. Furthermore, even if the notion of confidentiality is harmonised, a rule will still be needed to maintain the classified or confidential character of a document. We suggest that something similar to a mutual recognition principle should be applied in this case. That is, the assessment of the document by the authority of the country of origin should be respected by the authority of the country of destination. Probably one exception is needed from this rule: the country of destination’s authority should be allowed to make such a decision only if the declassification is based on a new fact, such as voluntary disclosure by the party that claimed confidentiality, or if the document has been made public by the other party. The second point concerns limits on the exchange of information within the network. Here we share the views already expressed by Marc van der Woude. Under Article 27 of the draft regulation, information obtained through an exchange within the network is to be used only for the purpose for which it has been acquired. However, Article 27 also applies ‘without prejudice to Article 12’, which stipulates that the information obtained can be used for the application of EC competition law. In the latest version of the regulation (presented as ‘Progress Report to the Council’) this provision is completed by ‘and in parallel, national

Technical Problem Areas

297

competition law’. In our opinion, the information obtained through an exchange should not be used for the application of national competition law alone and, even when used for the application of national competition law together with Community law or for the application of Community competition law only, it should be used with the limits established by the Dow Benelux and Spanish Banks case law.5 The Dow Benelux decision says very clearly that ‘the right of the defence of the undertaking concerned would be seriously endangered if the Commission could rely on evidence against undertakings, which was obtained during an investigation but was not related to the subject matter of the purpose thereof.’ In Spanish Banks, the Court established that ‘the right of the defence requires that the information thus obtained should not subsequently be used outside the legal context in which the request was made’. The third point has already been touched on by Willem Kist and Marc van der Woude; it concerns the introduction of a procedure whereby the undertakings concerned about an exchange of information should be heard before the competition authority of the country of origin makes a decision to transfer the information. In particular, the undertaking should be heard on two points: whether there are circumstances that justify the transfer of information, and the limits to this transfer (that is, what documents, or parts of documents, should be considered confidential or business secrets?). In our opinion, the competition authority of the country of origin should issue a reasoned decision on this matter, and there should be an opportunity to review it in court. The fourth point concerns legal privilege. It is a bit disappointing that the draft regulation does not make any reference to the existing case-law on legal privilege, in particular to the AM&S and Hilti doctrine.6 Nevertheless, it is obvious that this case-law should always apply to procedures before the European Commission, so that the Commission is prevented from obtaining — either through a request for information or by confiscation — any document that is the object of an exchange of correspondence between an undertaking and external lawyer in the context of an actual, or a potential, procedure before the European Commission. This is the case law, but we think that we must go even further if we want the legal privilege to be effectively protected; here I disagree with Frances Barr’s position. I think that here we have two different situations. In the first, national competition authorities are obliged to respect the legal privilege attached to documents generated in proceedings before the European Commission. In other words, if a company receives legal advice from external lawyers in an actual or potential proceeding before the European Commission, the national competition authorities cannot try to obtain this document, as this would be contrary to Article 10 EC. Even if legal privilege is not regulated under national law, the national competition authorities cannot obtain the documents 5 See supra notes no. 2 and 3. 6 Case 155/79 AM&S Europe v.

Commission [1982] ECR 1575; Order of the Court of First Instance, Case T–30/89 Hilti v. Commission [1990] ECR II–163.

298

Panel Discussions

generated — I insist — in procedures before the European Commission. The same holds true if the national competition authorities apply Articles 81 or 82 EC: when applying Community law, the Member States are bound by the general principles of Community law. In relation to the examples made by Frances Barr: it is difficult for me to imagine that the OFT is authorised under UK law to obtain from a foreign country information that it could not itself have obtained in its own jurisdiction, thereby taking advantage of the weaknesses of less developed foreign legislation that does not protect legal privilege. At the same time, I share Marc van der Woude’s suggestion that the highest standard of protection should be applied in both the country of origin and that of destination. Probably the best way to deal with these matters is through a directive. 䉴 MONICA WIDEGREN — I will turn now to a different subject — consultations within the network — which, along with the exchange of information, is one of two pillars necessary for a robust functioning of the network. In the legal exception system, the network will have to deal with more cases than at present (leaving aside, of course, applications for individual exemptions or negative clearance). We shall also have to deal with the newcomers in the network — the competition authorities of the new Member States — which have less experience in cooperating with the Commission or between themselves. These factors taken together will have an impact on the consultation process. The concept of consultation is not defined in either the existing Regulation 17/69 or the draft regulation. I suppose it could be defined as the initiative of a national competition authority to exchange information in order to obtain a reaction from another national competition authority and, in turn, take further action in a given case. We are, of course, talking about formal consultations among the members of the network. On the basis of the draft regulation, one can identify three types of consultation process: (1) consultation regarding the allocation of cases; (2) consultation prior to making decisions; and (3) assistance with the information gathering process. Following the implementation of the reform, the Commission will retain more or less the same powers with respect to consultation as it has in the current system. The Commission’s powers to provide information to the national competition authorities about new cases before it, and to consult within the Advisory Committee prior to adopting decisions, will essentially remain the same. With regard to Article 11(6) in the draft regulation, I was pleased to hear Alexander Schaub say that the Member States will be consulted before such a measure is adopted. The Member States were mentioned in plural, so I take it that Mr Schaub was referring to such consultations taking place within the network. I think that the consultation procedures should be explicitly addressed in the text of the regulation itself in order to underline the horizontal nature of cooperation within the network. The national competition authorities are required to submit information about new cases to the network. This provision was discussed in the Council Working

Technical Problem Areas

299

Group, and we have now specified that such information should be submitted after the first formal investigative measure is taken, whatever its nature. As I understand it, this means that the members of the network must ensure, for example, that a complaint has some substance before it is disclosed within the network. In order to facilitate consultation, we would also have to elaborate a standard format for this kind of information-sharing. As to consultations prior to the adoption of a decision, we agreed at the Council Working Group that every Member State needs to clarify for itself what a draft decision would mean and by whom it should be submitted to the network. For instance, would the national court or the national competition authority submit it? The answer to these questions depends on the particularities of each national decision-making system and the kind of information involved. The Commission says that a statement of objection would be fine, but of course it depends on whether all the Member States’ national legal systems allow for procedures similar to a statement of objection. In my view, we are on safer ground if referring to draft decisions. Finally, with respect to consultations following a request for assistance to obtain information: I prefer to see this procedure as a consultation process. When a member of the network requires assistance to gather information, that the members of the network must do their best to cooperate, or the network will never function. I am now referring to what Frédéric Jenny said yesterday: perhaps the new regulation should establish an obligation for the national competition authorities to cooperate. Finally, the functioning of the network will require a lot of informal consultations, which of course we already have. The members of the network will need to discuss competition policy problems: actual cases and also how to investigate distinct competition problems. A last point on language: the draft regulation makes no mention about this, but I think it is quite obvious that we will need a single working language. We could not possibly continue the current system, whereby we work in eleven languages; the upcoming enlargement makes the need for a single language more important. A network intranet will also have to be set up. I do not think that will be very difficult; we already use an internal system of transmitting encrypted information from the Commission to the national competition authorities. What we have to do instead — and here I agree with John Fingleton — is to design and set up a workable internal organisation within the national competition authorities with a view to their new tasks ahead. I do not see this as a problem of resources, but perhaps we could discuss that later on. It is all a matter of priorities, and I foresee that most of the cases before the Swedish competition authority will be Article 81 and 82 EC cases. 䉴 JACQUES BOURGEOIS — I have two short introductory remarks to make, and four brief questions to ask. First, I think there are two fundamental underlying policy choices to make when we talk about consultation within the network. The first is to say: ‘Well, let’s let the chips fall where they may’. In other words, let’s live with the

300

Panel Discussions

inconsistencies or lack of uniformity of application, because it will all sort itself out over time. If I may say so, this is somewhat like the approach in the US. The other choice is to say: ‘No, uniformity and consistent application is of paramount importance’, and the only way to achieve them is to allow the Commission to step in whenever there is a risk of inconsistent application and non-uniform application of the EC rules. This is, in fact, the option advocated by a good number of practicing lawyers. Could the solution be somewhere in between? This mid-way would be a form of cooperative joint management of the cases within the network, to be achieved through reinforced consultation and allowing everybody to have a say about things. In the end, if there is no consensus and a decision is to be made, that decision could be made by qualified majority. If we were to opt for such a system, a certain number of consequences flow in terms of consultation; more precisely, with regard to the scope and consequences of consultation. My second remark is that I find myself in agreement with many of the views put forward by Monica Widegren in her written contribution. She defined consultation as the initiative taken by one or several competition authorities to share information within the network in order to get reaction from other network members that will lead to further action. In her paper she actually extends the concept of consultation to the request for assistance and other matters. She insists very much on the horizontal nature of consultation, and I think that is an important point. And now the first question: ‘Consultation about what?’ In other words, should consultation occur with regard to competition policy issues that arise in general, or about cases? In my view, consultation about cases would, of course, involve considerable costs. How could we handle this matter? I think that, to start with, the information to be brought before the network should be limited to the essential elements of each case. Then, if the national competition authority that submitted this information also wants to request assistance from other members in relation to the case, a consultation process would be triggered. Or vice versa, other members of the network might request consultation about the case. My second question is related to Monica Widegren’s proposal that information about cases should be submitted to the network after the first formal investigative measure. Isn’t that a bit late? If we look at the Commission’s enforcement practice, we see that quite a few things would have happened before the first formal investigative measure is taken. I do not think that this is a good solution. It would be more sensible and efficient if the other members of the network were consulted before the first formal steps of investigation were taken. What should be the object of such consultations? The answer depends again on what one wants to achieve, but certainly this should include advice from the Commission, interpretation of the law, mutual practical assistance and, finally, the draft decision or statement of objection. My last question concerns the legal effect of the consultation process. On this issue too, the views appear to be very mixed. On one view, the outcome of the

Technical Problem Areas

301

consultation should not necessarily be given legal effect as the new regulation will give the Commission enough power to intervene to avoid major inconsistencies in the application of EC law. On another view, the results of the consultation process should be legally-binding on the national competition authority that is handling the case, as this would be a way of ensuring consistency and uniformity in the application of the EC competition rules. Finally, some maintain that there is no need to take a formal position on this issue because the outcomes of the consultation are a form of ‘soft law’. In other words, it would be difficult for the national competition authorities to depart from the line of action suggested as a result of the consultation unless there were very serious reasons for doing so. Finally, I think we should also take into account how these consultations are going to affect private parties. I am not going to come back to everything that has already been said on this. One extreme view is that the parties should have a say during the consultation process on how a particular rule should be applied to a specific case. I think that this would probably be very difficult to manage in practice. On the other hand, it is not very satisfactory to have the national competition authority dealing with the case imposing the result of the consultation process on the parties as a decision that cannot be challenged. 䉴 WOUTER WILS — I would like to start with a cautious disclaimer: I will not speak for the Commission or the national competition authorities; I will just express my personal views on this subject. Also, I will base my remarks on the Commission’s original proposal for a new regulation (other texts that have been circulated in the Council have never been approved by the Commission, and have been even less approved by the Council). My assignment for this workshop was to talk about the network from the perspective of the European Convention of Human Rights (ECHR). When I received the task I did not really see the link between the two, but in the meantime I have discovered that this is an extremely interesting subject. In my written contribution I have actually broadened it, in the sense that I also took into account the Charter of Fundamental Rights of the European Union (CFR), a document that has been ‘solemnly declared’ by the European Parliament, the Council and the Commission on 7 December 2000 in Nice (whatever that means in terms of its status in the architecture of the EC legal system). In any event, it would be unthinkable for the new system not to fully respect the provisions of both the ECHR and the CFR. The Commission’s proposal for a new regulation envisages three functions for the network: the efficient allocation of cases between its members, assistance between the members in terms of fact-finding, and coordination of action in order to ensure the consistent application of EC competition rules. Within the latter I would distinguish between exchange of information, consultations, and then the mechanism of control by the Commission over the national competition authorities on the basis of Article 11(6).

302

Panel Discussions

There are two sets of provisions in the ECHR and the CFR that are relevant in this context. On the one hand, there is the right to fair trial and other guarantees for defendants in criminal proceedings. You may ask yourselves: ‘What does this have to do with the network? EC competition law is not criminal law.’ The concept of ‘criminal’ has its own autonomous meaning in the ECHR. Maybe a more appropriate term would be ‘penal’ because it has a wider sense than ‘criminal’ and I think it is impossible to deny that EC competition rules are of criminal law nature within the meaning of the ECHR. I think that the meaning of the concept of ‘criminal law’ is the same in the CFR, because if you look at its preamble and the last but one article, they take over — possibly even extending — the defence rights established by the ECHR. In the second half of my written contribution I went through all the different elements of the network looking for possible problems of compatibility with the ECHR and CFR. After discussing them, I concluded that most of them are not real problems. I would like to mention now two of the seven problem areas that I identified in my written contribution. The first is the problem of what American legal jargon would call ‘double jeopardy’: the right not to be prosecuted and tried for the same offence twice. This right is laid down in Article 4 of Protocol VII ECHR, respectively Article 50 CFR. Back in 1969, in Walt Wilhelm, the ECJ held that there was no problem of double jeopardy if a company was prosecuted both by the Bundeskartellamt under German competition law and by the European Commission under EC competition rules, because these laws were not the same.7 The difference laid in the EC competition law’s orientation concerning effects on trade between the Member States. If I compare this case-law with a recent case from the Strasbourg Court, I am not sure that this idea that the two sets of law are different still holds true. The principle that seems to have emerged in Strasbourg is that two prosecutions are for the same offence when their essential elements overlap;8 vice versa, two offences will be separate when each contains an essential element that is not found in the other. The same principle is applied by the US Supreme Court in the interpretation of the 5th Amendment to the US Constitution. To cite the US Supreme Court: ‘Each provision should require proof of an additional fact, which the other does not, for it to be separate offences.’9 So, we may potentially have a problem with the Walt Wilhelm doctrine. Article 4 Protocol 7 ECHR, prohibits prosecution of the same offence twice within the same jurisdiction. One could say that two prosecutions, taking place in two distinct Member States, do not infringe the ECHR. The situation is distinct if the same case is prosecuted by the Commission and one Member State. Community law is part of the national law, and therefore not a different jurisdiction. At any 7 Case 14/68 Walt Wilhelm v. Bundeskartellamt [1969] ECR 3. 8 Judgment of the European Court of Human Rights of 29 August

37950/97. 9 Blockburger v. US, 284 US 299 (1932).

2001, Fisher v. Austria, case

Technical Problem Areas

303

rate, Article 3 of the draft regulation establishes the primacy of Community competition rules. There is no doubt that in such cases it will be the same offence for which there is a risk of double prosecution. While the ECHR prohibits double prosecution of the same offence within the same jurisdiction, the CFR prohibits double prosecution of the same offence within the European Union. I think the risk of double jeopardy is augmented in this situation. However, in recital 16 of the draft regulation, the Commission proposes that there should always be one single competition authority to deal with each case, and that should be the authority best placed to handle the matter. We do not risk double jeopardy if this proposal is embraced. The second problem area is that of who is going to be a member of the network. The problem is that in some Member States the courts have a role to play with respect to the decision to start prosecuting a case. What is relevant here is Article 6 ECHR and the case-law of the Strasbourg Court on the conditions of independence, impartiality and full jurisdiction that courts have to satisfy. I think that there is an obvious tension between independence, impartiality and full jurisdiction, on the one hand, and having to cooperate with the Commission on the other; this tension involves consultation, the transmission of documents in response to a request from the Commission, and possibly even the case being removed by the Commission under Article 11(6) of the draft regulation. For most Member States there is no problem in this respect because their antitrust prosecution systems do not involve the courts, and the decisions of the competition authorities can be appealed before courts. There are three Member States for which I do see a problem: Sweden, Ireland, and Finland. So, I believe that this issue should be clarified in the new regulation. The draft regulation is not clear on this point, as it does not define what constitutes a ‘competition authority’. The Commission has not taken a position on this. I think it should be made clear that the national bodies, which we sometimes call competition authorities but that qualify as independent and impartial courts within the meaning of Article 6 ECHR and Article 47 CFR, should not be brought under the scope of Articles 11 and 14 of the draft regulation. At the same time, and in order to achieve consistency, they should be brought under the scope of Article 15; that is, they should be brought under the same rules covering courts that apply Community law in litigation between private parties. And, of course, because these entities are courts, they should also be able to ask for preliminary rulings from the ECJ. I fully agree with Dr. Böge, who said yesterday that the Community body mainly responsible for ensuring coherence in the application of EC competition rules should be the ECJ. To my mind, a possible counter-argument — which was made yesterday by Céline Gauer — is that, just as there is no problem if the Bundeskartellamt applies EC competition rules, because its decisions can be appealed in court, there should be no problem if the Irish courts were to cease to be independent, because there would still be an appellate process and the Irish appeal courts would still be independent. I do not think that my definition would go against De Cubber v. Belgium and Findlay v. United

304

Panel Discussions

Kingdom case law, 10 where the Strasbourg Court said that its acceptance of ‘decisions taken in first instance by administrative authorities open for further appeal’ does not apply to traditional criminal law. I think that this definition certainly applies to Ireland, and could also be transposed to the Swedish and Finnish jurisdictions. In my view, the essence of the Strasbourg case law is to accept the historical situation of administrative authorities whose decisions are open to appeal. But I do not think that the Strasbourg Court would accept a situation in which a court that was once independent became dependent. DENNIS WAELBROECK — Yesterday we spent a lot of time trying to figure out whether the Commission was the sun or the moon in the new system and, more generally, where is the centre of the new universe? Since we are in the city of Galileo Galilei, I wonder whether we have put the centre of the system in the right place. If you read the Bible, at the centre of the creation are Adam and Eve, and it is for them that the rules are made. So, to return to our world, I think that the new rules are not for the competition authorities but for the business community. After this short introduction, I would like to make two preliminary remarks. The first is that competition law is not only about cartels. As I see it, competition law is about trying to strike the right balance between economic efficiency on the one hand, and workable competition on the other. It is there to define the sound limits within which companies should conduct their daily business. I emphasise this because I have the feeling that too often when we discuss the modernisation project we have a very repressive vision of the purpose of competition rules. This distorts the perspective because, if the law is only about cartels, there is almost no limit to what enforcers should do in order to fight them. I would like to think that not all companies are wrongdoers, and that they try to know what the law is about and observe that law. This area of law is extremely complex and difficult to apply, and what I would therefore expect from any reform project is clarity and simplification. My second preliminary remark is that the network is not some sort of uncertain ghost in the future; it is a daily reality we already live, with or without the modernisation project. What we are seeing now in our daily practice is that, instead of having only one interlocutor — as, in the good past, the Commission — cases emerge everywhere. As Giuseppe Tesauro said yesterday, this state of affairs is probably very lucrative for the lawyers, so we should not complain. But I am not sure this is necessarily the best way to operate. Just to give you a few examples: in the petrol station business there is an interesting question that emerges now, after the adoption of the new regulation on vertical restraints, which is whether a system of guaranteed margins for the retailers is legal or not. In Italy, the position 䉴

10 De Cubber v. Belgium, judgment of 26 October 1984, Series A no. 86, paras 31–32; Findlay v. United Kingdom, judgment of 25 February 1997, Reports 1997–I, para 79.

Technical Problem Areas

305

appears to be that they are not; in Spain, the position appears to be the contrary. The problem is, which rule would be applied within the network? And finally, is having all these procedures operating in parallel the most efficient way of handling cases? Returning to the issue of the network and the ECHR, I would like to make three points. I may be a bit provocative, but this is why we are here, so this should not be seen as a criticism of the modernisation project. The first point is that, as Wouter Wils said before, in this area of law we are speaking of imposing sanctions that may be of a criminal nature. Article 7 ECHR establishes the principle of the legality of the sanction applied. In other words, the law and the sanction should be predictable. There is a long line of case-law from the Strasbourg Court on the predictability of the law. In simple terms: it establishes that one should be able to know in advance what the law asks of him/her, and what sanctions will be applied if s/he infringes it. I agree, of course, with Dr Böge and Ian Forrester that there cannot be absolute legal certainty, particularly in this area of the law. This is not the point; I just wonder how much legal uncertainty is acceptable, or desirable, in a system where about 26 competition authorities will very soon apply the same law but may often go in different directions. Just to give a few examples: in Michelin, the European Commission imposed a fine of 20 million Euro for a rebate system in place in France, even though that system had been reviewed every year and approved by the French competition authority.11 So, one national competition authority says your behaviour is legal, and then the Commission comes in and imposes a fine of 20 million Euro for the same behaviour. Is this compatible with Article 7 ECHR? Another recent example is Visa International:12 following the Commission’s decision, that had found no restriction of competition in the non-discrimination rule, Sweden and the Netherlands prohibited the non-discrimination rule on the basis of national competition rules. And Belgium is now starting a procedure against the same company, threatening to impose a major fine for the same behaviour. I could go on and on with such examples, because lately I feel like all my practice is going in that direction. I recently notified the Commission of a selective distribution system for a major software company: while the Commission issued a comfort letter, the French competition authority imposed a fine for the same system. This is the reality of our practice, and I am not sure that the modernisation project fully addresses these issues. What I see instead is that we are going to lose the benefit of one-stop shopping in the current system — that is, application of Article 81 (3) EC by the Commission — and, frankly, I do not see any compensation for the companies. Once again, is it acceptable to have no rule on the allocation of cases and mutual recognition? Is this compatible with Article 7 ECHR? 11 COMP/36.041 – PO/Michelin, Commission Decision of 20.06.2001. 12 Case COMP/29.373 Visa International — Multilateral Interchange Fee,

24 July 2002, OJ L 318 of 22.11.2002, pp. 17 et seq.

Commission Decision of

306

Panel Discussions

If I go back to my starting hypothesis, when Adam and Eve breach the rule that forbids eating the apple, as far as I know, there is only one authority to sanction them, and not 26. My next question is: was it impossible to introduce another system? If I look at the Community legal system, I see everywhere mutual recognition where regulation is done by networks. For example, in the medical, pharmaceutical, veterinary, pesticide sectors there are networks of national agencies working together; they appoint one competent authority to deal with a case, and the others are consulted. There is a procedure of mutual recognition, and everybody accepts the ruling of the designated authority. We also heard yesterday about the NAAG in the US, whereby various authorities bring joint actions before one single court. There is also the alternative of having a system of joint management of cases, as proposed by Jacques Bourgeois. I find that the chosen system is minimalist, and it creates problems under Article 7 ECHR insofar as there is uncertainty about the forum in which one’s case will be decided. Very briefly, with respect to the network and the CFR: the reform does not envisage any harmonisation of the national procedural rules whatsoever, and I think that this is really a mine-field of problems. Finally, the EU is still not a signatory to the ECHR. So, if the Commission will apply Article 11(6) of the draft regulation, I wonder if this in itself is not a violation of the ECHR, because it denies companies the defence rights they would enjoy under the ECHR if prosecuted by a national competition authority. Having said this, I repeat that I support the modernisation project generally; I have just tried to make some constructive suggestions for the future. 䉴 CLAUS-DIETER EHLERMANN — Your remarks remind me of the fact that monopolies are traditionally easier to run than deregulated industries, for which you need rules, and here we are discussing the new rules for the post-monopoly period. My second point is that mutual recognition requires mutual trust, and if there is no mutual trust there is no mutual recognition. Having said this, the effect of further integration in the internal market should normally be at least a minimum of harmonisation.

CÉLINE GAUER — A lot of interesting critiques and remarks were made this morning, and I would like to make a few comments on two of the aspects touched upon in our discussion. On the exchange of confidential information under Article 12 of the draft regulation, I would like to give some answers to the points and criticisms raised by Marc van der Woode and Santiago Martínez Lage. First of all, Article 12 applies notwithstanding any national provision to the contrary. This means that no national rule can prevent the national competition authorities from exchanging information or using the information received. Even if the information has been collected in the course of criminal law proceedings, this would not prevent a national competition authority — which is traditionally an administrative authority — from



Technical Problem Areas

307

using the information. This is because Article 12 empowers national competition authorities to use the information received. Article 12 ultimately creates a new legal ground on the basis of which information may be obtained. Some of the speakers this morning called for the introduction of a procedure whereby we should have a legal decision, and the possibility of challenging it, before information is exchanged. If this were to happen, we might as well get rid of Article 12; the time possibly needed to have such a decision reviewed in court, with a risk of appeal, would be prohibitive. One may have to wait for two years before the court decides that, for instance, the Commission could take up a case and require the necessary information from the national competition authority. By then the case would be dead anyway. Furthermore, it could be that we need to exchange information even before a company becomes aware that the receiving authority will be carrying out an investigation. If the company became aware at this stage, it would probably be delighted to destroy all evidence that the receiving authority could later seek to find in its own jurisdiction. This is the reason why I do not think it is possible or desirable to have such a procedure. Furthermore, I do not think it is necessary; the level of confidentiality, which was mentioned several times this morning, will be the same. Here again (I mentioned this yesterday), I think that Article 27 of the draft regulation harmonises the level of confidentiality within the Community for all information exchanged pursuant to Article 12. It is not really important to establish who is going to decide the status of the information transmitted — be it the transmitting or the receiving authority — because the rules and standards to be applied will be the same anyway. Moreover, the receiving authority will generally be in a better position to assess the status to be given to the received information. At the beginning of an investigation, companies would normally claim that everything is confidential. Then, when we prepare the file for access at the Commission, we have more a precise discussion with the companies about each and every document that is contested, and we find a solution. So, for me, it is more for the authority that is going to make the final decision on the case (rather than the transmitting authority) to have this kind of constructive discussion with the companies. The second argument I wanted to comment upon is related to defence rights, and I refer in particular to the legal privilege for in-house lawyers. I do not agree with Marc van der Woude that the highest standard of protection should always prevail. I think we have a common standard of protection within the EU, which is already high, and that derives from the ECHR. The Commission is not a signatory to the ECHR, but it nevertheless respects it because of the case law of the European courts. If some Member States give legal privilege to in-house lawyers, I do not believe this is an expression of a fundamental right; it is simply a political choice that was made in a particular country. Companies carrying out their activities in Europe know that they are exposed to the possibility of having advice given by their in-house lawyers used in competition proceedings. I fully share the opinion expressed by Frances Barr this morning that the OFT could, for example, use an advice given by an in-house lawyer that it receives via the French

308

Panel Discussions

competition authority (which does not recognise the privilege). Again, this does not impinge on a fundamental right. For these reasons I do not think that we really need the sort of procedure proposed by Marc van der Woude. Article 12 of the draft regulation should be left to the competition authorities. It is for the deciding authority to make all the necessary assessments, and for the company to contest the use of this sort of evidence before the court. Of course, if the information were to be illegally collected by a national competition authority, then companies would have the normal recourse to the courts of the country where the collecting authority is located. But this is a completely different issue. Thirdly, I would like to comment very briefly on Wouter Wils’ intervention. He made it clear that he was expressing personal views, and not the official position of the Commission. I would like to contradict him on two aspects, also from a personal standpoint. If Wouter Wils is right on the issue of double jeopardy, then everything we have been talking about this morning with respect to teamwork within the network is worth nothing. The members of the network will not be able to work together if we can have only one decision-making body. Furthermore, if we allow only one decision to made, the result is that we will under-punish infringers. By contrast, I believe we should allow each national competition authority to take into account the effects of the infringement in their jurisdiction and to impose sanctions. The same infringement does not necessarily have the same effects in different jurisdictions. For example, with respect to an infringement producing effects in Italy and France, the Italian competition authority will take into account mainly the effects on Italian consumers, and the French competition authority will look mainly at the damage it causes to French consumers. I believe that — at least from the horizontal perspective — parallel action is perfectly sustainable and does not violate the principle of double jeopardy. Having said this, the issue is of course more complicated from the vertical perspective. If a national competition authority has adopted a decision concerning a certain infringement and the Commission then wants to make a more restrictive decision, this would probably create problems from the point of view of the principle of double jeopardy. Finally, Wouter Wils gave an answer to my comment yesterday about the possibility of making Article 11(6) applicable to courts. I take note of his answers, but I do not think it really convinces me. I do not think that competition rules, when applied to companies, fall within the scope of what the Strasbourg Court defines as ‘traditional criminal law matters’. Furthermore, I am not sure of the fact that, if the Commission could use Article 11(6) up to the first level of decision-making, this would disqualify the Irish court (for example) as a fair and impartial tribunal. However, that issue will probably have to be clarified by the court. ALEXANDER SCHAUB — First, I hope that everybody has appreciated the rich diversity and the liberty of opinion within the European Commission, to which I



Technical Problem Areas

309

will not immediately add further examples. Then, while listening to Dennis Waelbroeck’s critical remarks and the cases of conflicting decisions under the existing enforcement system that he mentioned, I was thinking that one cannot find better arguments for the reform. In response to the suggestion that we should introduce the principle of mutual recognition: this is basically what we are doing. The new system is based on the application of Community rules in all Member States, and the replacement of the Commission’s exemption monopoly by a system of parallel enforcement competencies. De facto, this means that there is the confidence among the members of the network that cases can be decided by their colleagues. Technically speaking this might not be ‘mutual recognition’, but I would say it goes even further that that. By the way, Dennis Waelbroeck’s optimistic description of regulation in the pharmaceutical sector is certainly well intentioned, but things are much more complicated in the real world. It so happens that I have personally been involved in the launch of the European pharmaceutical agency — and I continue to be very proud of this — but the realities of this sector are exactly the opposite of what you were describing. This is a case where you see very clearly the limits of the mutual recognition principle. What we are proposing for the competition area actually goes much further than what is currently practiced in the pharmaceutical sector. On a different note, I fully agree with John Fingleton on the fundamental importance of adopting pragmatic attitudes to the launch and application of the new rules, and avoiding by all means any obsession with perfectionism. I also agree with those who said areas of the reform still need to be clarified and there are rules that need further improvement. We have already seen this in the merger control area. At the same time, this reform has been prepared and thought over for years now, and I believe there are few other European projects to have been so carefully prepared. Of course, there will be room for further improvement, but those who expect everything to be absolutely clear and perfect from day one are not the right persons to make the new system a success. Another point that emerged clearly from this discussion is that the new system will trigger the need for convergence (which is the term that I prefer to ‘harmonisation’). I think we should avoid falling again into the harmonisation obsession of the past. As Commissioner Monti said yesterday, the mere fact that there will be such close cooperation between the competition authorities should serve as an eye-opener. The areas that have been identified here as problematic — and that were sometimes depicted in frightening colours — already exist under the current system, but perhaps there is less awareness of their existence. An example is the need for extra resources and staffing. The first question is: do we use our staff in the most efficient way? I am sure that close cooperation between the competition authorities has already improved our working and training methods and our enforcement priorities. For instance, the next ECA meeting will discuss common training programs, through which best practices are transferred much more easily, and which encourage peer pressure and peer reviews. Those who have

310

Panel Discussions

followed the developments over the last three years will remember quite a few examples of changes already induced by the prospective reform. A last remark concerns timing: at the last Council meeting (in Barcelona), the European heads of state and government expressed their hope that the reform will be approved by the Council of Ministers before the end of 2002.13 If this timing is respected, then certainly another full year will be reserved for the preparation of the effective start of the new system. CLAUS-DIETER EHLERMANN — I fully agree with you with respect to the pharmaceutical sector: this is an example where mutual recognition did not work, and where it was therefore necessary to establish a central agency.



ALEXANDER SCHAUB — The initial effort to apply the principle of mutual recognition was a total failure. Then the concept of a European agency came into being, and it is not a great success either. How does it work? All the cases that are ‘controversial’ by nature (for instance, those concerning biotech products) go directly to the European agency, and so do other cases in which some Member States are hesitant to recognise decisions previously made in other Member States. This mechanism is very close to that established by Article 11(6) of the draft regulation, although the Commission has never used its power to take up a case from a national competition authority against the will of the Member State concerned, and it does not intend to do so in the future. However, this difference of approach is fully justified in the pharmaceutical sector, because issues of health are at stake.



CLAUS-DIETER EHLERMANN — I totally agree. If the principle of mutual recognition is not workable and there is a need for a certain degree of uniformity in the application of certain decisions, then a mechanism of central control is indispensable. By the way, following the discussion we had yesterday, it is still not very clear to me what the current situation is in the telecom and energy sectors (that is, now the principle of mutual recognition no longer applies, and the Commission is no longer the decision-making centre). Personally, I do not believe in the sustainability of the telecom regulators working together without a very strong element at the centre, and that is absent the moment.



JAMES RILL — I think that Dennis Waelbroeck’s remarks focused on a very important split in the decentralisation process. It is something that people should really understand because it frames the whole discussion. Basically everybody accepts that a cartel is bad thing and should be punished, so there is not a whole lot of legal uncertainty in this area of antitrust law; there are some procedural issues to be dealt with, but these are not subtleties of the law. One of the objectives of the



13 The

Regulation was adopted 16 December 2002, and will become applicable from the 1 May 2004. (See Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ L 1 of 4.1.2003, pp. 1–25.)

Technical Problem Areas

311

modernisation project is to render more effective this area of law enforcement. So long as the procedural rules are respected, I think it is largely their own business how the Commission and the competition authorities of the Member States work this out among themselves. The issues are less clear-cut with respect to an abuse of dominant position. The non-discrimination rule example invoked by Dennis Waelbroek is very telling. In such cases the application of the law involves economic assessments that are very different from and more complex than those to be made in cartel cases. Prima facie it may be concerning to see different rules about the same kind of anticompetitive behaviour in different EC Member States, but instead we should see this as a laboratory to test the impact of the rules on the consumer. The rules can be harmonised afterwards. Then the question is: should that be done by the Commission and the Member State authorities? Or should it be left to the courts? Moreover, should the court system eventually evolve, or does it already have the ability to bring about a solution to the apparent irrationality of different rules in different countries? It seems to me that, in a decentralised enforcement system, the Commission and the competition authorities of the Member States should discuss together certain important enforcement and policy issues. For example, is resale price maintenance really a bad thing or not? Should it be prohibited with a per se rule or not? Other issues may be better left to the courts (such as whether there is a need for a non-discrimination rule or not), if only because they would otherwise overload the public enforcement system. To conclude, I see first an institutional issue about who does what. As to the more subtle questions of competition law, I believe that the courts should be given a bigger role to play. One should not lose sight of the main objective of the reform proposal, which is to liberate the Commission from the tedium of routine enforcement issues in order to allow it to focus on the big enforcement issues. In a certain sense this is also an answer to Mario Siragusa’s criticism about the fragmentation of the EC competition law system, and to doubts about whether the Community’s judicial system is robust enough to ensure coherence in the application of EC competition rules. As a practicing lawyer, I have greater confidence in the courts than in the administrative authorities. They may provide a more impartial resolution mechanism for some legal issues and, due to the adversarial process, they certainly have the ability to arrive at informed decisions. 䉴 CALVIN GOLDMAN — Listening to the remarks made this morning concerning the very interesting and challenging subject of the exchange of confidential information, I thought this is a discussion that is quite parallel to one that a number of us in this room were engaged in very recently at the OECD. And, of course, the issues addressed today are parallel to those that we wrestle with in the area of international cooperation between competition authorities. Most recently, I was asked — along with James Rill and a few others — to put together a response to the OECD’s very long list of questions in the area of exchange of international

312

Panel Discussions

confidential information in the context of trans-border cartel investigations. We attempted to address the two topics of safeguards and process. The issues that emerged are almost identical with those arising within the EU, and with those that arise in the context of bilateral cooperation between the EU and the US, Canada, Australia, and so on. To summarise them very briefly: on the safeguards side, the thorniest issues are those of legal privilege and immunity, and testimony protections. Then there is the question of giving notice about the exchange of information. Today I heard some very interesting remarks from the head of the Dutch competition authority, in the sense that in the Dutch legal system the parties have access to a special procedure that allows them to go to a special court for a right to be heard in advance of the exchange of information. I am not sure that a similar procedure exists anywhere else. On the contrary, the US Department of Justice very strongly opposes any notice about the exchange of information whatsoever. In Europe, the business community has asked for one. In fact, in our work for the OECD we suggested a compromise solution that tries to bridge the gap between these two positions in the sense that there ought to be notice only when it will no longer prejudice the trans-border investigation. So the authorities themselves can decide when to give notice, but they need not give notice if there is a risk of prejudicing a dawn-raid or any other instruments of investigation. So far this proposal has not been sufficient to close the gap. Nevertheless, the issue must be addressed, because we will otherwise not see effective cooperation in the pursuit of cartels. This is an issue of balance and of fair trial. If we can reach some consensus on this, the competition authorities will have a set of safeguards through which cooperation in enforcement becomes more transparent. JAMES RILL — On the same point of the issuing of notice, I think there is a difference between what might be the unofficial position of some in the management of the Department of Justice’s Antitrust Division, and the official position of the same. I would not want to contradict Calvin Goldman on the informal position, but if one wants to understand the official position of the Department of Justice’s Antitrust Division, one first needs to understand the circumstances under which information may be shared. Apart from the court procedures concerning the exchange of information — the delivery of information and aid of formal proceedings under the Grand Jury rules — there are only two legal instruments in the US that cover the exchange of confidential information. One is the Mutual Legal Assistance Treaty with Canada, which expressly refers to the exchange of information in antitrust proceedings. The other is the US–Australian Agreement, that was entered into under the International Antitrust Enforcement Agreement (commonly known as ‘the IAEAA’), which authorises the exchange of confidential information. With respect to notice prior to the exchange of that information, it was a lively topic of debate during the legislative proceedings in 1994, leading up to the enactment of the IAEAA. In its Committee Report, the Committee on Judiciary of the House



Technical Problem Areas

313

of Representatives provided that notice should be given prior to the exchange of confidential information with foreign agencies operating under an IAEAA agreement, except where that notice might prejudice the progress of investigation. That particular provision was endorsed and actually put into the report at the urging of the US Council for International Business (USCIB), and was subsequently endorsed by the then Assistant Attorney General in a formal response to a request for information from the USCIB. This is the formal position of the Department of Justice’s Antitrust Division. Again, I would not like to contradict Calvin Goldman’s perception of the informal position as possibly being something along the following lines: ‘No way we are going to give advance notice.’ Regrettably, the prejudice to the course of an investigation is somewhat in the eye of the beholder. What qualifies as confidential information is not always so easy to determine, and what will prejudice the process of an investigation can be very subjective, very nebulous. My guess is that, in any kind of exchange of information taking place in relation to a cartel investigation, the enforcement agency tends to believe that prior notice will jeopardise the investigation in almost every circumstance. Whose investigation is jeopardised — the one of the authority requesting the information, or the one of the authority providing it — is another question. They will probably both take the same position. So the likelihood of the prior notice being given does not provide an enormous hope for a safeguard (unless there is some negotiation prior to the information being given to the authority in the first place, which is something that has happened). The question is, what should be done about it? I think that there is a strong argument in favour of the view that prior notice is not entirely appropriate for every case. Where there is a hard-core cartel investigation in process, and there is at least some reason to suspect the likelihood of destruction of documents, or the possibility that a key witness will disappear, the giving of prior notice would certainly jeopardize the investigation. On the other hand, let us assume that knowledge of the investigation already exists and it is just a matter of doing some clean-up work on some documents: in this case, would prior notice jeopardise the investigation? Probably not. The point is: except in the Netherlands, there is no timely court review procedure available for instances of transfer of information. One proposal brought before the ICC by Canada is that of a ‘post-talk’ notice; in other words, the authority providing the information should inform the interested parties that a transfer of information has occurred and the information is about to be used in proceedings against them. This kind of notice would at least provide some opportunity for the review of the giving authority’s accountability. Having said this, unless there is some willingness on both sides to address the problem, possibly through guidelines as to what will or will not jeopardise the integrity of an investigation, etc., all that I have done is put a fairly intractable problem on the table. At least in a forward-looking pragmatic sense, I think that such guidelines may provide some kind of flexibility.

314

Panel Discussions

CLAUS-DIETER EHLERMANN — My experience at the Appellate Body of the WTO leads me to believe that the issue of the exchange of confidential information is extremely important in practice, and therefore needs to be tackled. In a certain sense, our discussion in relation to the competition network might be a paradigm for other contexts. All attempts at cooperation between the US and the EC in Geneva — whether in the ambit of anti-dumping or safeguards — have failed when it came to the exchange of confidential information. I remember Barry Hawk once explained that US rules sanction the breach of a promise to keep confidential information secret. Therefore, if the rules on exchange of confidential information are not equivalent in different jurisdictions, there will always be a fundamental reluctance to exchange confidential information. That seems to be very much the case between the EC and the US, where the difference of positions is sharpened by institutional differences. Yet you cannot adjudicate a case conveniently without dealing with this issue. Also, competition agencies are generally more careful about these issues than those dealing with anti-dumping procedures. I also have the feeling that the courts are not as sensitive with respect to issues in the antidumping field as they are in the area of competition.



ANNE WILLEM KIST — As it seems that I have given rise to a misunderstanding, I would like to clarify my point about notice on the exchange of confidential information under the Dutch legislation. Dutch law establishes a general duty to inform the parties concerned about the exchange of information, but this is without prejudice to the power of the Director General of the Dutch competition authority to exchange confidential information, subject to certain conditions. Having said this, I do believe that this is a very important issue, and we have to find ways of ensuring a fair balance. I do appreciate that conducting an investigation might not be unduly jeopardized by an exchange of confidential information but, on the other hand, I also believe there are legitimate interests of the parties concerned that need to be safeguarded.



䉴 MARC VAN DER WOUDE — I have to confirm what Director General Kist has just said: indeed, the Dutch provision concerning the exchange of information does not explicitly call for such a notice procedure, but I think it could be applied by analogy. It would not be completely unfeasible and it could also take place within very short deadlines (I think the deadline for a reaction is reached after eight days). So I do not think that introducing such a procedure would be difficult. In reply to Céline Gauer, I do not think you can argue that there is a common standard concerning the right of defence, or that there are common rules about gathering evidence; there are objective differences. The same holds true for the treatment of documents. I am not speaking about the concept of confidentiality, but the rules concerning access to public documents differ. So, I think we have an issue here. Mr Rill’s point about the a posteriori notice was interesting. In that respect, we could also think about the possibility of damage claims; an authority that

Technical Problem Areas

315

decided not to inform the parties concerned about the exchange of information in order not to jeopardize an investigation would assume the risk of damage claims arising later on. Another suggestion would be to identify the differences between the various pieces of national legislation on these issues. Assuming that it is neither workable nor advisable to establish common standards through the new regulation, one could still think of preparing (for example, during the period preceding the entry into force of the new regulation) a kind of chart of the existing differences. For example, are there countries in which information is gathered under criminal procedural rules? Can this information be sent to another country? Such a reference guide would allow an authority to know in advance the risks that it is taking by receiving the information. JOHN FINGLETON — The new Irish competition law, which was enacted this week, contains a very general provision concerning the sharing of information with other national competition authorities. In principle, this provision will allow us to exchange information with other authorities under our national law. There are probably similar legislative developments in other EC countries, probably mainly driven by a growing concern amongst competition authorities about how to tackle international hard-core cartels. Otherwise, it must be said that there is nothing in our national legislation to prevent us from sharing information about every other aspect of antitrust enforcement. 䉴

HELMUT BROKELMANN — I think that the topic of the exchange of confidential information should be addressed by the Commission. We cannot leave things as they are now, otherwise there is a serious risk that undertakings will disrupt the new enforcement system. Our clients are not prepared to have their confidential documents circulating freely throughout the Community. They need to have the possibility of taking legal action to prevent the exchange, and they need to be given notice of such exchanges. I acknowledge the fact that there are different needs to be balanced here, but, like Marc van der Woude, I also think that it is possible to introduce some sort of interim procedure whereby it could be assessed whether the conditions for the exchange to take place have been met. It would also be useful if the Commission adopted a notice interpreting the concepts of professional secrecy, business secrets, internal documents, and so on. As a matter of principle, the concerned parties should be able to appeal in the jurisdiction of origin of the information against the exchange of information a priori; exceptionally, in order not to obstruct the investigation of the receiving authority, it could also be acceptable if the undertaking had the right to appeal in the country of destination of the information. Finally, even if the Commission were to adopt a notice interpreting the concept of confidentiality, the practical application of this notion by a host of national authorities will probably lead to very different results. This is why there is a very clear need for a procedure allowing the parties to appeal these kinds of decisions by the competition authorities.



316

Panel Discussions

LAURENCE IDOT — On the issue of confidentiality, I believe we can distinguish between differences in the definition of ‘confidentiality’ (which gives rise to what we call in international private law ‘conflicts of qualification’) and differences concerning the content of the provisions on confidentiality (which gives rise to the traditional ‘conflict of laws’). In order to solve conflicts of qualifications, the only solution is to adopt common definitions. In this respect, I am not sure that the safeguard we find in the draft regulation is clear enough to solve such conflicts. In order to solve the conflicts of law, the solution is either to adopt a rule of conflict or to adopt a common rule. I am not sure that the draft regulation offers a common, EC-wide rule; this is because it makes reference to the national law, and I agree with the concerns expressed by Marc van der Woude on this point.



CALVIN GOLDMAN — A possible three-fold programme dealing with these issues would involve, first, establishing common and transparent safeguards, including on the issues of legal privilege (as reflected in the written contributions of Margaret Bloom and Anne Willem Kist). Secondly, the issue of notice would have to be dealt with; a possible solution that would not prejudice investigations would be the a posteriori notice (as explained by James Rill). At that point, there is some accountability on behalf of the competition authorities. Thirdly, in order to protect the authority’s ability to carry out investigations, the right to review the notice should be limited by a very strict time period so that the competition authority can fix any irregularities and continue its investigation.



CÉLINE GAUER — To be clear, the Commission does not say that companies should not be informed about the transfer of information. Of course companies have the right to be informed that information has been exchanged or will be exchanged, but we also have to have some flexibility in order to safeguard the effectiveness of investigations. As for sanctions: there are sanctions if, for example, EC officials breach the obligation to respect confidentiality. There are certainly similar rules in the Member States. However, I take the point that we could create a sort of reference guide in order to make these rules known and to show how stringent they are; this would make it clear that we are all bound by the same concept. I insist that Article 27 of the draft regulation is very clear: we are all bound to respect professional secrecy. The issue here is what sanctions apply for the breach of this obligation. I also fully take your point with regard to guidelines, or a notice, interpreting the concept of professional secrecy. We have to make sure that the application of this concept will be the same in all the Member States. But I would simply advise companies not to push too much for that. The reason is simply that, if you have severe sanctions, you also have a greater likelihood that the competition authorities will err on the side of caution with regard to broader concepts, so as not to risk being held liable. In other words, companies may be better protected 䉴

Technical Problem Areas

317

by a broad concept of professional secrecy than by one more that is more specific and limited. IAN FORRESTER — I have a comment on James Venit’s intervention, and another on pharmaceuticals agencies and judicial review. Most of my practice involves looking at really tricky issues about which it is not obvious which way the law should be interpreted. In these situations, whether a practice or an agreement or the system is good or bad depends upon how you look at it. In the new enforcement system, these issues will be argued in 15 — and actually up to 27 countries — instead of just in Brussels. This is a very good thing. The Commission has given up its theoretical monopoly, and this is a good thing too. This means that we are going to have more decisions, more discrepancies, more errors, and more inconsistencies. In the early days of the reform, solving those discrepancies or preventing the inconsistencies must be an important part of what the Commission will do. The notice on cooperation with the national courts14 failed because the Commission was too cautious, too conservative, and not wanting to give advice for fear of being unfair, while the courts were reluctant to consult the Commission. That did not work well. Roundtable discussions like this one, with people sitting around a table and discussing a common problem, are obviously right and have got to be a good way forward. This is especially necessary for the competition authorities of the new Member States from the East, who lack experience in using the principles of law in a flexible way in a dynamic market-place. I do not think anyone could have a problem with agencies discussing among themselves the best way of approaching problems. The vice of pharmaceuticals, published agencies and anti-dumping officials is that their decisions are not very well reasoned; to the contrary, they are very brief, unintelligible, and rarely presented with frankness. This makes judicial review of such decisions very difficult: the European courts are reluctant to look behind how an administrative agency has exercised its discretion. By contrast, in the competition field the Commission’s decisions are more reasoned. It would be very desirable if roundtables of the kind that I referred to before produced decisions reasoned on the merits and subject to proper judicial review. Otherwise we would once again have non-transparent discussions producing a common view that cannot be reviewed by the courts. And even I — though usually very tolerant and optimistic — would not be willing to accept a regime in which the experts discussed something among themselves and reached a conclusion, if that conclusion were not susceptible to proper judicial review. 䉴

CLAUS-DIETER EHLERMANN — We are now moving towards the subject of consultations. In the view of some among us, this can apparently result in a decision, although the notion of consultation per se does not seem to suggest this.



14 European Commission (1993): Notice on Cooperation Between National Courts and the Commission in Applying Articles 85 and 86 of the EEC Treaty, OJ C 39 of 13.2.1993, pp. 6–11.

318

Panel Discussions

I think that the answer to the question raised by Ian Forrester — namely, whether there should or should not be some kind of judicial review following consultations within the network — depends very much on whether consultation is followed by a decision within the network, or only by a decision by the authority that requested consultations. MARIO SIRAGUSA — I honestly do not see how the network could make decisions on issues subject to consultation. This is not provided for in the draft regulation, inter alia because the power to make decisions is linked to a specific procedure for the adoption of such decisions. My comment is more related to the issue of judicial review, so it is linked to the subject of consultations. James Venit suggested that the task of avoiding an outcome in which different competition authorities take different positions in the application of the same EC competition rule, such as in the Visa example mentioned by Dennis Waelbroek, should be left to the courts. Under the current enforcement system, Visa has the choice of going to the Commission to ask about the legality of its practice, and this ensures the uniform application of the EC law throughout Europe. However, the Commission is practically abandoning cases of this kind even before the implementation of the reform. In the new enforcement system such cases will be the subject of consultations within the network, but I suppose that the result of the consultations will again be individual decisions made by the individual members of the network. Individual decisions made by members of the network can be appealed in court, which provides a safeguard against this kind of outcome. And, since the members of the network apply Community law, the national courts will refer requests for preliminary rulings to the ECJ. This is what I think James Venit had in mind when he referred to harmonisation via the court system. This underlines once again the importance of Article 3 in the draft regulation, because it is this provision that enables overall judicial control. On the other hand, I thought that there was a contradiction in Dennis Waelbroeck’s very interesting intervention. He was criticising Article 11(6) in the draft regulation, but doesn’t Article 11(6) establish the mechanism to avoid conflicting decisions? The Commission has the power to intervene and take cases over so as to decide them in a uniform way if the members of the network continue to apply the same principle of Community law in different and conflicting ways after consultations concerning a particular legal problem have taken place within the network. To conclude, I think that we have at hand two ways of dealing with the issue of conflicting decisions: judicial review (and I am very much in favour of increasing the number of cases referred to the ECJ for preliminary rulings) and Article 11(6) of the draft regulation.



䉴 JAMES VENIT — To continue on the same topic, I think that there are two arguments in favour of judicial review as a mechanism for ensuring consistency of decision-making. First, I think that courts may ultimately be better suited to deal with complex economic issues than the regulators, as the latter are sometimes driven by policy considerations. Secondly, I find very interesting the extent to

Technical Problem Areas

319

which one can invoke EC law in the judicial review of inconsistent decisions taken under national law affecting companies that operate throughout the Community. In the current enforcement system, the way out of such situations is to notify the Commission of the agreement or practice, relying on the primacy of EC law and the fact that national decisions based on national law cannot contradict an individual exemption obtained from the Commission. This possibility does not disappear in the new enforcement system: all that goes away is the Commission’s exemption monopoly. If an appeal against such a decision were to be denied in the national court, I think that there is still room to make a final argument in a reference to the ECJ that the national law has been applied incorrectly because the case had a Community dimension. This could be a way to replace the lacuna generated by the Commission’s withdrawal from its central role in guaranteeing the consistent application of EC competition rules. JOHN FINGLETON — I earlier made a point about institutional resources, and I would like to clarify it. What I meant to say this morning is that institutional resources are a problem for the network in so far as they create externalities across the Member States. In other words, if the Irish government does not allocate sufficient resources to the Irish competition authority in view of its participation in the network, this can become a common problem for the network, as all members of the network are expected to contribute to its good functioning in some way. I think that, besides the obvious case of consultations within the network, there are other more subtle ways in which resources matter. First, it is a fact that the national competition authorities will participate in the network whether they are well informed or not, so there is always the danger that some authorities will give opinions within the network without having actually examined the facts and the issues at stake. Of course, it would be even more worrying if the uninformed opinion were to come from the Commission. Secondly, each of the members of the network must find time to read others’ decisions, and this is very important in terms of what both James Venit and Ian Forrester have stressed about consistency in decision-making. Consistency is much more likely to occur if we do read each other’s decisions. At the moment we probably do not have the resources necessary to have two or three of our staff reading the decisions of the other competition authorities, and language is an important factor to consider here. I think that at least the important decisions made by other authorities should be distributed and read within the network. Monica Widegren said she did not think that resources were a problem; I think that resources are not a problem at present, insofar as it is only individual Member States that are affected and the shortage does not have consequences for the others.



䉴 JACQUES BOURGEOIS — I would like to react to some points made by Mario Siragusa and James Venit. I think we should acknowledge the fact that in Europe the enforcement of competition rules is primarily administrative, and judicial review is limited and late. I do not think that we should do away with the

320

Panel Discussions

administrative enforcement system, but we should build into it some of the guarantees for the rights of private parties that exist in the judicial system. For instance, if consultations within the network end up in some form of common resolution — not to mention decision — the positions of the private parties are affected. As such, the resolution should be challengeable; somebody should take responsibility for this resolution or decision and give the private parties the opportunity to be heard and to explain their positions. The judicial review system is not satisfactory in this respect because of the long periods of time that it takes in order for a decision to be reached. PAOLO SABA — I work for the Italian competition authority, but I am speaking in my personal capacity. I wanted to make a point concerning the functioning of the network and consultations within the network. It seems to me that little attention has been paid so far to the scope of application of the consultation mechanism. Nobody has raised the issue of which cases should go under the consultation procedure, as stipulated in Article 11(4) of the draft regulation. The Commission proposes basically that only prohibition cases be subject to this form of prior consultation; I find this sort of counter—intuitive, for different reasons. First of all, one of the major innovations of the reform proposal is that Article 81(3) EC will become directly applicable by the national competition authorities. One would expect that the major risk of the decentralisation process would be the relaxed application of Community rules to cases that, for example, raise problems of market foreclosure. In other words, since the national competition authorities will adopt decisions whose effects exceed the national borders, the degree of independence of the national competition authorities becomes even more important than in the past. It is astonishing to me that there seems to be no particular concern about the possibility of national competition authorities making decisions that are biased in favour of national interests. 䉴

䉴 CLAUS-DIETER EHLERMANN — I can only support what you have said. I remember the discussions we had at the edition of the workshop of two years ago, when we all agreed that it was absolutely indispensable to have Commission guidelines or a notice on the interpretation of Article 81(3) EC. I suppose that such a text would be discussed within the network before it officially emerges. Your second point reminds me of some personal reflections about international antitrust enforcement. I think that, at the international level, the problem is not one of over-enforcement, but rather of under-enforcement. It could very well be that some areas within the Community have a problem of under-enforcement. Under the new enforcement system, the problem will be solved within the network by means of peer pressure and bench-marking (instead of the Commission stepping in and solving this problem itself). In other words, the phenomenon of under-enforcement will be discussed within the network, and one would normally expect the network members in question to react. I believe that this is will be much better than re-centralising enforcement.

Technical Problem Areas

321

䉴 MONICA WIDEGREN — I think that Claus-Dieter Ehlermann is absolutely right: one of the positive effects of the network will be precisely that it gives its members incentives to apply EC rules. It is true that there are no safeguards against lax administrative enforcement (as Paolo Saba mentioned before). But I think that close cooperation within the network will involve all the members assisting each other to apply the EC competition rules in a professional way. On a different note, John Fingleton is right: there is of course an issue of resources. The network cannot function well if we do not have the right numbers of skilled staff. What I meant to say before is that I foresee that the application of Article 81 and 82 EC will take priority in our work in the future. 䉴 BARRY HAWK — It seems to me that several people around this table believe that judicial review of what the network is doing remains important. In so far as this is about the protection of individual rights, things are quite clear. My question is: what kind of judicial review can be performed with respect to the internal functioning of the network? I think there are two main reasons why the courts will be hesitant to review what the network is doing. One is ripeness: the courts will avoid performing judicial reviews of cases if they consider that it is too early to intervene with respect to procedures that are under way. In other words, the courts will examine only preliminary issues in such circumstances. The other is the administrative discretion enjoyed by the competition authorities in applying Article 81(3) EC. So I think a lot of thought should be given to the legal status and effects of actions and decisions taken within the network. 䉴 ULF BÖGE — Just to clarify a point: the network will not take any decisions about cases. Such decisions will be taken by the individual members of the network, and of course they are subject to judicial review. The main purpose of the network is to facilitate the exchange of views and information among its members. Maybe the network will also develop some guidelines, but decisions made according to these rules will not subject to judicial review in so far as they are not legally binding. As I mentioned before, our internal (or infra-national) network of competition authorities in Germany has prepared a joint paper concerning our approach to abuses of a dominant position in the electricity sector. It is in our interest to pass this information to the companies, so that in designing their business strategies they can take into account what the competition authorities are likely to do in the future. However, in some cartel cases it makes no sense to make this information public. The information exchanged within the network in relation to such cases will be confidential, but not with respect to the evidence concerning the companies. It is the exchange of views between the competition authorities with respect to the application of the law that will be confidential.

IAN FORRESTER — I fully agree with Dr Böge. What I was instead asking for is that individual decisions about cases, which are made by individual competition authorities and addressed to particular companies, be fully reasoned. And the



322

Panel Discussions

reasoning should draw on the law and its interpretation in view of the facts; it should not simply turn on an internal decision of the network. CÉLINE GAUER — Initially I wanted to make exactly the same point as Dr Böge: the network is not an institution, it is simply a group of people talking to each other. There will be no decisions of the network as such. This is why I think that information exchanges within the network cannot be subject to judicial review; if only for the simple fact that these are informal internal acts. I would just add a comment: he was referring to the German network of competition authorities, which seems to work very well, but Germany is a federal system, whereas the EU is not. For instance, we do not have the ability to challenge national decisions before the Community courts; Article 11(6) of the draft regulation would not be necessary if we had such an option. Indeed, the courts are responsible for maintaining consistency in the application of the law in a federal system but, I repeat, the EU is not one. Therefore we need to create a sort of a path leading to the European Court of Justice, and this is why the Commission has to step in.



䉴 ULF BÖGE — However, let us remember that when the national courts have doubts as to the application of EC law, they can ask the ECJ for a preliminary ruling. So I do not think that the federal system is ultimately indispensable for obtaining a clear and final opinion from the ECJ. In this sense, it is not obligatory that the Commission step in. This is not to say that the Commission should not have the ability to step in, but this should occur only in special cases; it should not be the general rule. I would think that, at least in principle, a final decision of the ECJ gives more certainty to everybody.

ALEXANDER SCHAUB — I fully agree with this view: it is not ‘a must’ for the Commission to intervene in every case, and the Commission will be selective in its interventions in the future, as it has been in the past (the Commission has not acted not acted upon every suspicion of a violation of the Treaty rules in the Member States). By the same token, the Commission will have to decide whether to intervene within the network and try to direct developments towards a correct application of the Community rules, or to rather leave the problem to be settled at the level of the courts. Indeed, there will be many cases in which it will be quite clear that the case will go to court for appeal, regardless of what is happening at the administrative level. I believe that the Commission can reasonably come to the conclusion that it is better if the matter goes directly to the court in such situations, as it would arrive there anyway even if the Commission were to intervene. 䉴

䉴 DAVID GERBER — What comes out clearly in our discussion is that there is a considerable cost involved in running the network. The network will not itself

Technical Problem Areas

323

make final decisions, but there will be numerous decisions made by the individual members and each of these has a cost. Also, so far we have not paid enough attention to the disparate impact of participating in the network on the individual members. I think we should reflect further on the funding of the network, i.e. where the resources will come from. If it is in fact a Community structure, it may be reasonable to think of Community funding for it. JOHN FINGLETON — I would like to make a last point about what we called initially ‘consultation’ within the network. Yesterday I asked, at what point does a consultation actually become a decision? I am still not convinced by the arguments that have been put forward to the effect that the network makes no decisions. On the other hand, I think there are good arguments for believing that consultations are preliminary steps in a decision-making process. The European courts will probably adopt a purposive interpretation of the consultation procedure, bearing in mind that the substantive decisions will ultimately be taken by individual authorities and will then be subject to several layers of judicial review. These elements would allow you to get away with the idea that the network decisions are preliminary, and are therefore not justiciable acts. But I think that these arguments are unconvincing, and it is not the best way for the Commission to proceed with this position. 䉴

䉴 WOUTER WILS — I would like to return to the issue of double jeopardy. Take the example of a European-wide cartel for which fines are imposed by both the French and Italian competition authorities. Can we say there is no problem of double jeopardy because the French authority protects the French consumers and the Italian authority protects the Italian consumers? I think this raises two issues that should not be confused. One is the notion of what is the same offence within Article 4 Protocol 7 ECHR or Article 50 CFR. The other is, what are the rules applicable to the different members of the network — be they Community or national rules — that determine which type of cases are being prosecuted by which authority? The right not to be prosecuted twice for the same offence is a fundamental right of individuals, and I think we should put ourselves in their position. The representatives of the companies I mentioned as an example have met in smoke-filled rooms only once, but they are to be fined twice for a violation of Article 81 EC. There is only one culpable action and only one applicable rule. How can we then say that they are not being prosecuted and punished twice for the same offence? Another matter is that the rules of distribution of work between the Italian and French competition authorities are possibly rules of French and Italian administrative law; these exist to determine when the national authorities can become active. This may lead to the result that the French competition authority will become active only when French consumers are affected by the infringement of the law, and the Italian competition authority only when the Italian consumers are affected. I believe

324

Panel Discussions

that the second matter is irrelevant for judging whether the companies concerned are prosecuted twice for the same offence within the meaning of Article 4 Protocol 7 ECHR and Article 50 CFR. 䉴 MARC VAN DER WOUDE — I wanted to go back to the issue of judicial protection; this is, after all, also a fundamental right. I am very sensitive to what has been said by Ian Forrester, Mario Siragusa, James Venit and Jacques Bourgeois. Yet judicial review is indeed a difficult area, particularly if we are considering preliminary rulings from the ECJ. I must say that I have been a practicing lawyer for 23 years now, and I have only once succeeded in obtaining a preliminary ruling. This was after ten years of legal proceedings, and on a minor issue that had nothing to do with my case. The Treaty of Nice provides for the establishment of jurisdictional chambers at the European court. We are now talking about jurisdictional chambers for trade-mark law and employment cases, and it has already been suggested that perhaps something should also be done in relation to competition law. I wonder if this is not the solution to our problem: an adversarial system as suggested by Barry Hawk, with several prosecutions by national authorities before one jurisdictional chamber. This would be easy to set up, and it would give us the benefit of coherence in decision-making. Furthermore, there would be an independent body to decide cases brought by the competition authorities of the Member States. This would also ensure a fair trial. For example, in such a system you could discuss Visa International’s nondiscrimination rule in a much fairer way than in a preliminary ruling in which the Court will not hear experts or the evidence.

I John Fingleton1 The Distribution and Attribution of Cases Among the Members of the Network: The Perspective of the Commission/NCAs

I.

Introduction

Modernisation, or the reform of Regulation 17/62, is about to happen. How it will work depends on the functioning of the EC network of competition authorities about to be set up, and on the allocation of cases among the network members — which are the issues examined in this paper. Relatively little is known about the latter aspect. My remarks throughout this paper rely on two documents:2 the Draft Joint Statement of the Council and the Commission on the Function of the Network of Competition Authorities3 (hereafter ‘the Draft Statement’); and the Draft Notice on Cooperation within the Network of Competition Authorities4 (hereafter ‘the Draft Notice’). The issues that I raise come under three broad headings: • the criteria for allocating cases; • technical questions; • policy issues. When I was asked to examine this subject, I was faced with several dilemmas. First, the distinction between distribution and attribution was intriguing. Distribution is defined as ‘the act of dispersing’, ‘apportionment’, ‘the geographic occurrence or range of a custom, usage, or other feature’, or in law, ‘the division of an estate or property among rightful heirs’. Attribution is: ‘the act of attributing or ascribing, as a quality, character, or function, to a thing or person, an effect to a cause’, ‘assigning to a cause or source’ or ‘the act of attributing, especially the act of establishing a particular person as the creator of a work of

1 Chairman,

Competition Authority, Ireland. This paper represents my personal views, and not those of the Irish or any other Competition Authority. The paper has benefited from discussion with, and comments from, Damian Collins, Declan Purcell, John Temple Lang and Wouter Wils. Any errors are mine. Comments are welcome to [email protected]. 2 These documents may not be in the public domain. 3 Presidency Working Document, dated 14 March 2002, Brussels. 4 Commission Staff Working Paper, EU Restricted, Submitted to the Council Working Group, issued February 2002.

328

Working Paper I

article’ Not wishing to engage in a discussion on whether competition cases are works of art or have rightful heirs, I shall address the simple question of how cases get allocated among NCAs. Secondly, I am not clear where the boundary of technical issues lies, or if a clear boundary exists between so-called technical issues and broader policy issues. One of the themes of this paper is that any system of case allocation needs to take account of not just existing behaviour, but also how behaviour will change in response to the system that is designed. Only a system that takes full account of the behaviour that it induces in others (whether the EC Member States, the Commission, or private parties) will be fully stable. Because broader policy issues get entwined with technical issues, I address some of these. Thirdly, I am not sure that I can fully represent the perspective of the Irish Competition Authority, let alone that of 14 other NCAs, and the Commission. Indeed, it is not clear that a common perspective yet exists, although hopefully one will emerge.

II.

Criteria for Allocating Cases

1.

Case Characteristics

The allocation of cases will presumably involve a two-stage process: • Stage I: determining which cases are suitable to be allocated; and • Stage II: allocating those cases to NCAs. At both stages, decisions will be made based on observable characteristics that cases possess. I attempt to list (in Table 1) the common or standard characteristics that a competition case might have, so as to see which, how, and at what stage these characteristics might be used. The Table focuses primarily on how the characteristic is relevant to the Stage II allocation. Table 1:

Characteristics of Cases

Characteristic

Elements

Relevance to Stage II Allocation

Size

Turnover affected

Complexity

Number of parties Both vertical and horizontal Abuse of Dominance Cartel Other agreement

Not unless via other characteristics. May be relevant for Stage I determination? Not unless via other characteristics. May be relevant for Stage I determination? Not unless via other characteristics. May be relevant for Stage I determination? Behaviour that segments national market might be particularly relevant at Stage I.

Type

Continued.…

Technical Problem Areas

329

Table 1 Continued.… Characteristic

Elements

Relevance to Stage II Allocation

Reaction required Urgent action (injunction) Unclear. Some Member States can bring Sanction: fines or prison infringement to an end more quickly or apply a more effective sanction, but not clear if allocation will be based on this. Relevant Regional or national This characteristic will determine whether geographic market Supranational national or Community law applies and may be more relevant at Stage I. If the market is supranational, then an effect on trade will exist. If the market is narrower, it is likely, but not certain, that there will be no effect on trade. Geographic Regional or national This characteristic will determine the set of markets affected5 Supranational NCAs that have an interest because consumers (not clear if also buyers or indirect consumers) in that Member State are affected. Where more than three Member States are affected, the Commission will automatically be best placed. Source Where evidence is located This affects which NCA is best equipped to carry out investigation. This information may not be known until the investigation is underway, and initial indications may differ from subsequent outcomes. Additional State Aid, State Regulation In these cases, the Commission is Background Previous or related cases better placed. Not clear that these matter.

2.

Questions about Criteria and Characteristics

A number of questions arise relating to how characteristics will be used in the criteria for deciding which cases get allocated and how they are allocated. Is there a clear distinction between the Stage I and Stage II decisions and, if so, a similarly clear distinction between the characteristics that are used at each stage? It appears that the size, complexity and type of case may be used in the Stage I decision on whether the case is suitable for allocation. It is less clear how the relevant geographic market, as opposed to the geographic markets affected, will be treated. If the relevant markets were national, then national law would apply, unless the anticompetitive act itself was responsible for or contributed to the segmentation of national markets. 5 This

characteristic needs to be distinguished from the relevant geographic market because several distinct geographic markets could be affected by a single action, particularly if the effect was to further segment national markets.

330

Working Paper I

It is clear that the first NCA to become aware of the case will make the assessment as to allocation.6 However, it is less clear how disagreements in regard to this can be addressed. To whom will cases be allocated?7 This issue is fundamental as to whether the network exists simply to allocate cases, or whether it exists to work on them jointly. A case can clearly be allocated to a single National Competition Authority (NCA) or the Commission, and this is seen as the desired option.8 It also appears that one NCA could take a lead role with others assisting (presumably in gathering evidence).9 Other permutations are less clear. Could the Commission assist when a NCA is in the lead role? Could a case be allocated jointly to two or three of these, with virtual teams working in close cooperation? Could one NCA take the lead in the investigation, while another takes the lead in bringing the infringement to an end? Could the Commission lead an investigation, but pass the completed file onto NCAs?10 These various options depend, in part, on how closely NCAs cooperate in practice and on technical issues (see Section III below). Some characteristics may not be known when the decision is made. For example, we may not know (a) where evidence is located or (b) the geographic effects11 of the case. If only a minority of cases have the property that relevant allocation information is not known at the outset and could change significantly, then the allocation of cases should be earlier rather than later. However, it is also necessary to have a mechanism to deal with the minority where the information does change. Re-allocation is one option, but this may be highly inefficient. The other option is greater cooperation among NCAs so that the investigative team is sufficiently flexible to deal with changing information. This raises questions about (a) whether we can predict in advance the type of cases where relevant

6 The Draft Notice, at para 7, states: ‘Normally, allocation will be assessed by the first authority to be seized with the case’. 7 The Draft Statement, at para 16, states: ‘Cases will be dealt with by an authority, or by authorities, able to restore or maintain competition in the market’. 8 The Draft Statement, at para 17, states: ‘Cases will be dealt with by a single competition authority as often as possible’. 9 The Draft Statement, at para 19, seems to suggest that cases should always be led by one NCA: ‘In cases where single action is not possible (when several national markets are affected and no NCA can deal with the case alone successfully), the network members affected should seek to designate one competition authority as the lead institution’. On the other hand, para 16 suggests a simultaneous approach by several in cooperation: ‘After the initial allocation period, when the same case (same market, same parties, same conduct/agreement) is being dealt by more than one NCAs well placed to do so, one national competition authority will take a formal decision, whilst others stay their proceedings or, if this is not possible, the NCAs will deal with the case in close cooperation’. 10 A relevant example is the SACEM case, where the Commission undertook an investigation but passed the matter on to the relevant Member States. See Marenco G. (1994): ‘The uneasy enforcement of Article 85 EEC as between Community and national levels’, in Hawk B., ed., Annual Proceedings of the Fordham Corporate Law Institute 1993, New York, Juris, pp. 605–27. 11 A case in point is the alleged cartel in medical and industrial gases (Comp 36.700 of 2002) that was treated as a case solely within the Dutch market but, as the investigation proceeded, it was realised that it might extend beyond the Dutch market.

Technical Problem Areas

331

allocation information will change and (b) how virtual teams will operate. It also raises questions about how disagreements will be resolved.12 Taking up one of these latter themes, can we engage in useful sector profiling that will give rise to greater clarity and predictability around case allocation? Are there some sectors or markets where the fact that firms’ (e.g., pricing) behaviour is either supranational or based around national markets can be used to make predictions about the outcome of the Stage I or Stage II processes. One could imagine a useful once-off ex ante exercise where information is collected to develop pre-indicia of sectors. While such a thought is highly appealing, I have doubts as to whether it would be effective. Such doubts are based on the fact that such behaviour could change in response to published criteria. In other words, firms could ‘game’ the system. This question of private parties, NCAs and the Commission ‘gaming’ the system will be a recurrent theme in this paper. A further question is what information on case characteristics should be provided by each NCA to the network. It is likely that there will be some discretion in what information is provided, if only because there will be some discretion in what information the NCA decides to collect initially and how much effort the NCA devotes to its task. Immediately, one sees the possibility for gaming by the NCAs. If the system of case allocation is seen to work efficiently and fairly, then NCAs will be happy to have full disclosure. If not, then information is likely to be biased. There are questions about what constitutes the first investigative action. It is not clear that there is a shared and agreed understanding of this across Member States. Market definition will be crucial, possibly at both State I and Stage II. It is particularly important for practical purposes, for trust within the network, and for external confidence in the process, that each network member adopts a similar methodological approach to market definition. The NCAs and the Commission should therefore consider whether it would be appropriate to adopt and publish a common notice on market definition to which they would all adhere. There is a particular issue that arises for smaller Member States whose markets are closely integrated with one or two larger neighbours. In such cases, the application of criteria could result in the cases automatically going to the larger economy in all cases because the effects on consumers would be greater there. One option is to have some interdependence between cases, so that they can be shared proportionately and cooperatively. 12 See

the Draft Statement, at para 16: ‘Case allocation will be completed as quickly as possible. An indicative time limit will be used within the Network (up to 3 months). Normally, this allocation will remain definitive to the end of the proceedings provided that the facts known about the case remain substantially the same. If so, this implies that the competition authority which has notified the case to the network, will normally remain the responsible competition authority if it is well placed to deal with the case and no other competition authority raises objections during the indicative time period’.

332

Working Paper I

It is not just the characteristics of cases that matter, but also the characteristics of the Member States and national procedures and sanctions. If an important element of efficient allocation is that the case goes to the Authority best placed (a) to end the infringement and (b) to apply an appropriate sanction, there are important policy implications. I give two examples peculiar to Ireland: • Irish law (like the US and unlike many other Member States) does not permit civil fines. In other words, fines can only be imposed on companies or individuals following a criminal trial and with a ‘beyond reasonable doubt’ burden of proof. In effect, this means that no fines apply to non-cartel (i.e., Article 82 and much of Article 81 EC) violations. Would it be possible, as a general rule, that the Irish Authority would undertake an investigation and then pass the file to the Commission to impose a sanction? Could this only happen if the complaint was first made to the Commission? Again, forum shopping by private parties may be possible. • Irish law allows (and UK law is likely to allow) for imprisonment as a sanction for engaging in hard-core cartels. Where a hard-core cartel has effects across a number of national markets, a subset of whom have such criminal sanctions, would it be appropriate that the lead role be taken by one of those Member State with appropriately tough sanctions. Finally, it is important to avoid gaming by parties in the form of forum shopping. In the case of Ireland, it would be important that parties could not rush to the Commission or another NCA as a way to avoid criminal sanctions in Ireland.

3.

The Commission and the NCAs

An important question, worthy of individual treatment, is when the Commission is best placed to deal with a case. This question is addressed in the Draft Statement at para 20: the Commission will be particularly well placed to deal with a case if more than three Member States are substantially affected by an agreement or practice, if it is closely linked to the application of other Community provisions (e.g. state aid, Article 86 or Article 10 EC), if Community public interest is served (a case that requires a decision that is binding throughout the whole Community, this normally not being a case that only affects one national market), or if several competition authorities that consider themselves well placed to act do not reach an agreement on how to proceed with the case. Para 22 continues: after the initial allocation period, when a case is being dealt with by one or several competition authority (-ies) which is (are) well placed to do so, the Commission will normally not open proceedings with the effects of

Technical Problem Areas

333

deseizing them pursuant to Article 11(6) of the draft Regulation unless one of the following situations arises: • Network members envisage conflicting decisions in the same case. • Network members envisage a decision which is obviously in conflict with consolidated case-law; the standards defined in the judgments of the Community courts and in previous decisions and regulations of the Commission should serve as a yardstick; concerning facts, only a significant divergence will trigger an intervention of the Commission. • Network member(s) is (are) unduly drawing out proceedings. There is a need to adopt a Community decision enforceable throughout the Community in order to develop Community competition policy. The national competition authority does not object: again, this raises a number of issues. An issue that will determine in large part whether Modernisation is about real decentralisation is the interpretation of the term ‘conflicting decision’. This could be interpreted conservatively to mean that the same mind could not reasonably have reached two particular decisions, or liberally to mean that two different decisions must not impose on an undertaking a set of obligations that are (commercially) inconsistent with each other.13 There are, I believe, several reasons to favour the liberal interpretation. First, the conservative interpretation would create the perception, if not the reality, that the NCAs are simply automaton satellites of DG Competition. This could undermine the cooperation of the NCAs in the process generally. Second, experimentation is extremely valuable in the evolution of legal thought, especially within the constraint that the single market is not undermined and two NCAs do not order firms to behave inconsistently. That is precisely the type of thing we are currently seeing with regard to the merger test, and it gives rise to constructive and healthy debate about improving the design and implementation of competition law. A second, and more pedantic issue, is why more than three NCAs cannot routinely deal with a case? It could be argued that enormous cost and effort will be devoted to establish a network, which will not then work as a full network. Perhaps the network is only intended as a forum for allocating cases, and a limited role is envisaged for joint casework within the network. Instead of a criterion that gives a specific number, it might be more useful to point out that the Commission will be more likely to take over a case the more member states affected. The role of the advisory committee will be crucial in all of this, and I address this in Section III below. 13 Even

if obligations are only national in nature, they could still be inconsistent if they prevented the undertaking from pursuing an EU wide strategy. Indeed, such inconsistency could undermine the internal market, as it would force firms to adopt different pricing strategies in different member states.

334

Working Paper I

III.

Technical Questions

This section deals with a number of questions that might be called technical, and that relate to the operation of the network in respect of the allocation of cases. These technical issues are not simply back-room loose ends that need to be sorted out by technical experts, but instead are fundamental to the success of the entire project. Technical decisions will determine what information is seen by whom, how it is processed, and ultimately how decisions are made.

1.

Inputs

The first question concerns the inputs to the network from the NCAs. This has a number of dimensions: • What information precisely will be sent — should there be a minimum necessary to enable others to assess the matter, and indicate what other documentation exists? Should there be a standardised initial file report? • What triggers a data entry? • How far can an NCA go with a matter, whether it arises from a complaint or is own initiative, before it communicates it to the network? • Where an NCA determines that a case does not have an effect on trade and that national law applies (Stage I process above), will this be notified to the network? If so, this means that all domestic cases would be entered on the network. If not, it means that Stage I determinations would not be open to challenge.

2.

Timing

It is envisaged that all cases should be notified immediately,14 but it is also acknowledged that there may be exceptions to this rule.15 The Draft Notice identifies clearly the trade-off in cartel cases between the NCA that first becomes 14 The

Draft Notice, at para 16: ‘The Council Regulation creates a mechanism for the competition authorities to inform each other in order to ensure an efficient and quick allocation of cases. Article 11(3) of the Regulation lays down an obligation to inform the Commission before or without delay after commencing the first investigative measure and states that this information may also be made available to the competition authorities of the other Member States. This obligation covers all cases where Article 81 and Article 82 EC are applied. It is parallel to the obligation imposed on the Commission by Article 11(2) of the Regulation’. 15 The Draft Notice, at para 17: ‘Given the extremely sensitive and confidential nature of certain information before the investigation has effectively started, there is some flexibility in the system: if the authority wants to carry out surprise inspections on the spot, the other competition authorities can be informed only after the inspection has been carried out’.

Technical Problem Areas

335

aware of a cartel acting quickly to protect evidence at home and at the same time not acting so quickly that evidence can be destroyed in other jurisdictions. Another area that the Irish Authority has had to address is that of blockades where a group of competitors, often trade associations, attempts to disrupt trade in order to gain higher prices. Such cases often involve an effect on trade, but both cause and effect have been entirely within Ireland. Urgency is of the essence; the Authority tries to obtain injunctive relief without delay. It is arguable that it should not be necessary to inform the network (in advance) in cases such as this. This raises a minefield about possible injunctive relief against mergers that is beyond the scope of this paper.

3.

Information Technology

There are questions about the technical operation of the network that have to be addressed. These include the security of the network, its capacity, its stability (and what happens if (or when?) it crashes), and whether our systems can be integrated. For example, the Irish Competition Authority has not hitherto had control over its own IT systems and, as a result, has not been able to get the OLIS system of the OECD working properly in its office. Given that many of these issues will take time to resolve, especially if a computer network system must be put out to public tender in the Official Journal, it may be appropriate that a technical working group should now be started on this question.

4.

Language

It is not clear what language will be used both for documents and for cases. For written documents, translation will be required, but it is not clear whether this will be done by the authority entering data (which would imply a single language) or by the authority reading the data off the network (which would enable multiple languages over the network). Translations at both ends could happen, for example, if there were two network languages. There are some indications that the burden would be at the source.16 On the other hand, the practice in the (criminal?) courts in Ireland is that the State provides translators for non-nationals. Ultimately, decisions on translation will affect how much information is put onto the network, and how much is downloaded. If a single language is used, there is a danger of asymmetric involvement by NCAs. 16 See para 10 of the Draft Statement: ‘All members of the Network will manage information efficiently. They will also minimize the administrative burden of participating in the Network on the understanding that any mandatory information sent will be made available and easily accessible to all Network members’.

336

Working Paper I

Once cases are allocated, there are important questions about the rights of the parties, and the NCAs to relevant documentation, and whether documents would be translated routinely and whether meetings would be interpreted.

5.

Conflicts of Interest and Hiring Policies

The network may also raise questions about conflicts of interests. At present, the Irish Competition Authority (and some other NCAs) routinely hires lawyers and economists from the private sector and lawyers and economists subsequently go on to do private sector work. This is very similar to the American practice, and we see such mobility as hugely desirable. At present, we are able, within Ireland, to identify and address possible conflicts of interest. The existence of the network may make it more difficult to control this issue and, importantly, to create both the reality and the perception that conflicts of interest are addressed.

6.

Teams and Virtual Teams

There are important practical questions about how NCAs will cooperate within the network. There are several scenarios envisaged (see Point 2 on page 3 above). One NCA may take the lead, with others assisting or inputting evidence. Another is that two or three (not more) NCAs may work in parallel.17 This suggests the existence of some type of virtual team. How would such teams operate?

7.

Resources

The big question is whether the governments of member states will provide the resources necessary to enable NCAs to undertake these investigations. This issue is too broad for treatment here. A second concern is the network cost itself, by which I mean the resources needed to establish, support and operate the network. If the fixed network cost is invariant to the size of the Member State or NCA, then it would represent a much larger proportion of the budget in smaller NCAs. There is no guarantee that national governments will provide resources for this, raising the prospect that the

17 The

Draft Notice, at para 16, states: ‘Cases will be dealt with by a single competition authority as often as possible. However, if the action of one NCA is not sufficient to bring an end to the infringement in a wider geographic area, two or three NCAs may be able to act in parallel against that infringement’. Para 11 goes on to state: ‘When several NCAs act in parallel, they have to cooperate closely in order to ensure that the analysis and the outcome are consistent’.

Technical Problem Areas

337

network cost could result in less investigation within smaller Member States.18 This would be a highly undesirable outcome. In addition, there is no indication that Member States will provide these resources in advance, and they are needed in advance to get the system up and running. I make this point because the current process of discussing modernisation at the Commission and Council over past years has put enormous pressure on the Irish Authority. As no allowance for this has been made in our budget, core competition work has been displaced. A third resource issue is what happens if the Member State best placed to take a case does not have the resources so to do? On a related point, is there a danger that certain types of boring cases might not get taken by anybody? Would the Commission end up as claimant of last resort? Would this depend on whether the original complaint was made to the Commission? Will the resource constraints within a Member State become a factor for the network to consider in allocating cases? Another issue is whether the existence of the network will increase the burden on some member states under national competition law. The very fact that knowing there was a cartel in Market X in Portugal, Greece, Belgium and Finland might require the NCA to examine Market X at home. If it did not do so, it might subsequently risk the criticism that it should have known about a cartel but did not act.

8.

Advisory Committee

It is envisaged that the Advisory Committee will meet as a narrow and extended group.19 The narrow group will look at cases and the broader group at horizontal issues. This leaves a number of questions. First, will the members of the broader group be political representatives? It is not clear from the wording whether the additional members that the Member State can appoint to the Advisory Committee to discuss horizontal issues should be from the NCA. If not, then it is important to recognise that this broader group would be less politically independent. It is not clear what defines ‘horizontal issues’ for the purposes of the distinction. If it includes all matters other than cases, then the broader group would play a very large role in determining policy in the form of notices, guidelines and block exemptions, and could indirectly determine how cases are treated.

18 If so then taking up my theme on gaming the system, smaller member states would become safer havens for evidence. 19 The Draft Notice, at para 62, states: ‘The Advisory Committee is the forum where experts from the various competition authorities discuss general horizontal competition issues and individual cases. It is therefore important to ensure that each member of the committee belongs to the competition authority of that person’s Member State. When other issues than individual cases are being discussed, Member States may also appoint an additional representative competent in competition matters’.

338

Working Paper I

Para 65 of the Draft Notice states that: ‘The Commission must take the utmost account of the opinion of the Advisory Committee’, but this is in the context of Commission decisions, and it is not clear if it applies to network decisions on allocating cases. Could the role of the Advisory Committee be strengthened to deal with possible disagreements in case allocation? Suppose a dispute arises between two NCAs or between an NCA and the Commission. Would it be appropriate for the Advisory Committee, or a subgroup of disinterested NCAs, to do this? There is a serious concern that the Advisory Committee that examines cases could be reduced to a technocratic forum, with important policy decisions being taken outside of the forum of NCAs. This would, in my view, lead to excessive politicisation of the network. Generally, there may be merit in examining experience elsewhere to see if particular approaches have worked well, or if there are clear pitfalls to be avoided. For example, UK law allocates parallel competition enforcement competence to both the OFT and the sectoral regulators. It would be interesting to know more about whether case allocation has been dealt with and, if so, how and what factors have made it work better. Similarly, there may be lessons to be learned from the United States where the Federal Government and States have concurrent powers. Neither of these examples however will inform us about some of the legal, cultural and language differences that the European network may need to address.

IV.

Policy Issues

I now turn to address one or two more policy topics. I raise these because the answers will determine important decisions and choices over the technical questions just discussed.

1.

Is Case Allocation a Legal Decision?

A central question concerns whether the decision to allocate a case amounts to a legal decision. The decision to allocate a case may well have consequences for how the case is investigated, appraised and ultimately dealt with. The question of whether this is a legal decision will need to be clarified. The ongoing work has begun to address this issue. This Joint Statement is political in nature and does therefore not create any legal rights or obligations. It is limited to setting out common political understanding shared by all Member States on the principles of the functioning of the Network [Draft Statement, Item 3]. If a case is reallocated to a given competition authority, it is because there is an anticompetitive effect on its territory and because that authority can bring an

Technical Problem Areas

339

end to the infringement in question. This means that the competition authority to which the case is reallocated would have been in a position, in any event, to commence ex officio proceedings against the infringement. The reallocation of the case does therefore not change the position of the parties. [Draft Notice, Item 41] Paras 41–44 of the Draft Notice deal with the position of complainants. These documents overall suggest that the view is that case allocation is not a legal decision, but at most an administrative one. It is in everybody’s interest that doubt on this point be eliminated before the system goes live. Complainants and parties subject to proceedings, and their legal advisors, need to know where they stand and what are the possible treatments of their cases. The NCAs need to know the legal value and standing of their actions within the network, as this could affect their legal basis to act. There are a number of subsidiary questions: • Could a party or complainant challenge a case allocation decision either at the time of allocation (or as soon as it is known) or ex post (when the decision has been made)? • What happens if a case is re-allocated to a member state that has criminal sanctions, from one that is not? • What happens if a case is allocated to a Member State that conducts proceedings in a different language than one or more defendant? If the system is to work, it is essential that the end result be substantially the same no matter where the case is sent. Otherwise we risk either that case allocation gets bogged down in legal challenges or that private parties game the system. In both cases, effectiveness would be undermined. The NCAs and Commission may have a strong preference to treat case allocation as an administrative decision, in order to reduce gaming by private parties. However, we need to be clear that such an approach is legally valid.

2.

Will the Network Have Ancillary Effects?

Decentralisation means that each NCA and the Commission will know more about the cases within each other Member State. This raises the potential concern that interesting ‘national’ cases might be removed from the national jurisdiction. Hitherto, the Commission and neighbouring NCAs would not have been aware of such cases. The reform of Regulation 17/62 intends that NCAs should use European law in preference to national law where both apply (Article 3 of the draft regulation) in the interests of consistency. However, if NCAs or Member States are concerned that case disclosure could lead to re-allocation abroad, they treat such cases under national law. This would undermine the aim of consistency.

340

Working Paper I

A related issue is whether the network will lead to more Article 86 EC cases being brought by the Commission. The existence of the network may make the Commission aware of a wider set of cases where state restrictions prevent competition and, given the statement that the Commission will be more likely to take over a case if Article 86 EC applies, it is possible that the number of such cases may rise. A similar point could be made about state aids.

V.

Conclusion

This paper has asked more questions than it has answered. That, appropriately, may best represent the position of the NCAs and the Commission, as many of the answers have yet to be formulated. There are, I think, three areas of conclusion. First, we need to sort out some broad policy and legal questions before it will be clear that the allocation of cases can work efficiently in practice. Foremost among these are whether case allocation will constitute legal decisions, and the related question of the rights of parties. Resolution of such issues, and resolution in the right direction, may be essential to ensuring that the network can operate. Clarity is also essential to ensure that the design of the network is correct. If case allocation is an administrative decision, the possible mechanics for case allocation could be very different than if it is a legal decision. The big question is whether such legal doubts can be resolved before the system is put in place. Secondly, there are a number of technical areas that are critical to the successful operation of case allocation within the network. Central among these are the language, resource and information technology issues. Not enough attention has yet been given to these, and they should be identified as urgent priorities within the Council working group or other fora as appropriate. Thirdly, for the remaining issues, I am optimistic that they can be sorted out as the network evolves. That is not to say that they are not critical to its success, but rather that there will be an inbuilt form of self-correction within the system. The myriad of possibilities for gaming by public and private actors within the system mean that there will be constant pressure within the system to address shortfallings and to improve how the network allocates cases. The NCAs will ultimately have considerable discretion in the allocation of cases, and if the system does not work efficiently and fairly from their perspective, then this will distort the working of the network so that the information going into it may be biased, delayed or incomplete. The more flexibility in both the structure of the network and the high-level rules (e.g., constitutional rules in the Regulation) on case allocation, the more quickly and fully will the system converge towards efficiency. It is also important at the outset in the design of the system to try to anticipate how the various actors will behave so as to reduce the need for convergence.

II Ian S Forrester1 Diversity and Consistency: Can They Cohabit?

I.

Introduction

I was asked to give the view of the academic and of the practising lawyer concerning the allocation of cases among the members of the EC network of competition authorities once the decentralisation of enforcement of European competition rules has been effected. These are two distinct assignments. A teacher of law would favour clarity and predictability, so that the procedural consistency of the structure would emerge steadily, through the application of clear procedures by administrations operating under judicial review. An advocate’s task is to pursue a successful result for the client within whatever constitutional structure exists. Lawyers reacted to the curious enforcement structure applicable to the EC competition rules during the 1980s and 1990s, exploiting the opportunities for the benefit of their clients. To the extent that some legal advantage, real or theoretical, could notionally be obtained by making a notification, lawyers made such notifications. The system worked irrationally and reality did not correspond to theory, but the system did function. Thus I conceive my role as ‘legal practitioner’ to be that of a practitioner interested in law reform, where the practitioner’s and the academic’s views may indeed coincide. I begin by making some acknowledgements and suggestions. 1.

1 White

The current system of EC competition law enforcement did not deliver enough answers to the legal questions which arose in daily commercial life. The dominance of the Commission over decision-making had advantages, in that its decisions were thoroughly debated, thoroughly reasoned and (in most cases) transparently adopted, and the body of case-law was broadly coherent. There were disadvantages, however, in that the Commission took too few decisions to resolve the theoretical uncertainties created by its doctrine. Any regime that promises a richer case-law through recruiting the help of fifteen other agencies to reinforce the central agency will present problems of quality control and inconsistency.

& Case, Brussels; Visiting Professor in European Law, Glasgow University. The author gratefully acknowledges the generous assistance and research offered by his Danish colleague, Jacob Borum, and his Latvian colleague, Aleksandra Melesko, as well as a number of other helpful persons who have answered technical, policy and legal questions.

342

Working Paper II

The reservations expressed below do not betoken scepticism about the desirability of reform. 2. There is a lack of confidence in the quality and consistency of the decisionmaking to be expected from the Member States’ competition authorities, especially after enlargement. It is honest, not disloyal or discourteous, to acknowledge the existence of such concerns. During the early debate on the reform of Article 85(1) and (3) EEC in the 1980s, very similar arguments were voiced about national courts which — so said critics — would not be capable of economic analysis and might damage the purity of European competition law as it had emerged in Brussels. The perceived weaknesses relate partly to resources and partly to capacity to resist political pressure. 3. There is an evident danger of inconsistency of approach, procedure, and outcome, even if each authority were well equipped, totally immune to governmental lobbying, and would decide every case in a magisterial manner. The proposed principles by which such controversies will be resolved are quite imprecise, and will themselves be controversial in their application. 4. Trusting that the Commission will sort out crises, discrepancies and errors is understandable, but we will have achieved only a limping reform if the Commission’s intervention becomes routine, in effect an extra layer of bureaucracy, equivalent to an appellate instance. If the Commission’s intervention is public and rather formal, then such intervention probably requires the opportunity for both sides to be heard; and if it is private and informal, without all parties having been heard, there will be problems of procedural fairness and ‘clubbiness’. 5. There is today very little cooperation in concrete cases between national competition authorities. Most have little experience of even considering with other authorities whether to renounce or to claim ‘proprietorship’ of a specific matter. There is plenty of fraternal goodwill, and little evidence of hostility or obstructiveness. This fraternity needs reinforcement. This paper will consider whether diversity is intrinsically undesirable; the choices of fora available to the practitioner; the hard and soft rules for reconciling conflicts; the alleged weaknesses of the authorities; the remarkable lack of practical cooperation in concrete cases between them; and the weight to be given to other authorities’ decisions.

II.

Diversity of Approach, Outcome and Procedural Rules

Is legal diversity in itself bad? There are plenty of fields of legal practice where national rules differ significantly. One is labour law, where Belgium and Italy offer

Technical Problem Areas

343

employees a level of protection which used to be regarded as barely believable in other countries. Perfecting a security interest over a moveable in Scotland will require different steps than in England. Differences of legal tradition and approach, and even differences of result, are therefore familiar to European lawyers.2 They are not, however, desirable if we aspire to a consistent continentwide regime where the same rules are applied in parallel. There is an important difference between other areas of law and competition law when it comes to adapting to local legal constraints. The company which wishes to reduce or increase its workforce in several countries or grant a security interest over railway carriages in four countries will pursue its goal in function of local legal rules which are different for historically verifiable reasons. The company which wishes to launch a new franchise for, say, gardening centres, and wants to avoid competition law difficulties may have to confront differences in substantive result despite the identical substantive principles being applied. Such discrepancies of approach and of outcome are therefore not the same as the normal patterns of legal diversity which characterise Europe. If the decentralisation exercise is to work well, the discrepancies of result should not be numerous or large. Are there mechanisms which can reduce the discrepancies? We are familiar with the fact that a criminal investigation, accusation, trial, conviction and sentence will be handled very differently in different countries. It seems inevitable that there will be considerable differences in style, rules of evidence, rights of the defence, formality of the procedure, extent of access to the dossier, openness of the proceedings to third party interveners. For example, should statements of objections, notifications or complaints be available to third parties for their use in collateral proceedings before national courts?3 Anecdotal experience and enquiries suggest that these procedural differences are today large. I suspect they will diminish only slowly.

III.

Preliminary Advice and Reality Checks

Any practitioner wishes to give solid advice to clients. Formulating legal advice about competition law (like some other fields of law, notably tax law) necessarily involves an element of intelligent prediction. If the administration responsible for enforcing the rules were to know of the transaction, what would be its attitude? Is the period of the non-compete clause too long? Are the proposed topics of the trade association meeting inappropriate? Is it legitimate for the technology licence to be territorially limited? Such questions often cannot be answered

2 Forrester

I. (2001): ‘The reform of the implementation of Articles 81 and 82 following publication of the Draft Regulation’, 28 Legal Issues of Economic Integration 2, pp. 173–94. 3 Case C–67/91 Direccion General de Defensa de la Competencia v. Asociacion Espanola de Banca Privada (AEB) [1991] ECR I–4785.

344

Working Paper II

confidently and practically without reflecting on the likely approach of the administration. Making a prediction about probable administrative reaction will be more difficult when fifteen national administrations, as well as the European Commission, must be taken seriously into consideration. There will frequently be a choice of authority for the seeking of advice. Since the new regime contemplates that there will be no affirmative — and binding — decisions, such multiplicity of advice should not in itself present problems of principle, even though it is not desirable. Officials are likely to express themselves cautiously and it will probably become common for the target of an enquiry to ask whether other agencies have been contacted on the same topic.

IV.

Initiation of the Case: Which Authority will Take the Lead?

The practitioner may well have a choice of remedies inasmuch as he has a choice of fora. Maybe (but only maybe) a weak case may best be argued first before the putatively sympathetic ‘home’ authority, especially if it is a relatively young one. Maybe a complaint will be best made in the country where the complainants are likely to be the most vocal. Which authority should follow up a complaint, or on its own initiative order the initiation of an investigation, is a delicate matter of procedure which can have considerable consequences. Presumably binding negative clearances will disappear since notifications will have disappeared, unless the Commission’s views on this subject change. The law will advance through condemnations, not through approvals. The cases which affect competition in several Member States are the ones which must be handled successfully if decentralisation is to work. The reform contemplates informal mutual respect, with cooperation to settle problems of allocation of jurisdiction. It is a core element of the proposed Regulation that the Commission and the national competition authorities should form a network and work closely together in the application of Articles 81 and 82 EC. The draft Regulation states in the Explanatory Memorandum that the network is meant to provide an infrastructure for mutual exchange of information, including confidential information, and assistance, and also to ensure an efficient allocation of cases based on the principle that cases should be dealt with by the best placed authority. In several Member States the competition authority, once seized of a case, is obliged to come to a formal decision. Such obligations may hinder reallocation of cases to a better-placed authority. To overcome this problem in respect of the application of Articles 81 and 82 EC the draft Regulation empowers a competition authority to suspend a proceeding or reject a complaint on the grounds that another competition authority is dealing with or has dealt with the case. Parallel proceedings can be avoided if the authorities choose to avoid them.

Technical Problem Areas

345

Further guidance can be found in the Draft Joint Statement of the Council and the Commission on the Function of the Network of Competition Authorities, which says in para 20 (this may not be the current version) that the Commission will be … particularly well placed to deal with a case if more than three Member States are substantially affected by an agreement or practice, if it is closely linked to the application of other Community provisions (eg state aid, Article 86 or Article 10 of the Treaty), if Community public interest is served (a case that requires a decision that is binding throughout the whole Community, this normally not being a case that only affects one national market), or if several competition authorities that consider themselves well placed to act do not reach an agreement on how to proceed with the case.

I submit that this threshold risks being too low. I suspect that few significant competition law problems will not affect three or more Member States. The Commission may do better to propose its services as a mediator about which agency is best placed to carry matters forward, rather than offering to take over responsibility for prosecuting the case where several Member States are involved. Para 22 of the Draft Joint Statement offers more formal criteria: After the initial allocation period, when a case is being dealt with by one or several competition authority (-ies) which is (are) well placed to do so, the Commission will normally not open proceedings with the effects of deseizing them pursuant to Article 11(6) unless one of the following situations arises: a) Network members envisage conflicting decisions in the same case. b) Network members envisage a decision which is obviously in conflict with consolidated case-law; the standards defined in the judgments of the Community courts and in previous decisions and regulations of the Commission should serve as a yardstick; concerning facts, only a significant divergence will trigger an intervention of the Commission. c) Network member(s)is (are) unduly drawing out proceedings. d) There is a need to adopt a Community decision enforceable throughout the Community in order to develop Community competition policy. e) The national competition authority does not object.

Each of these subparas can be defended as rational, but each presents questions. If there were a formal set of criteria for deciding which would be the lead authority, this would not prevent arguments over the application of these criteria. If there are no formal criteria, and instead disputed ‘title’ to a case were settled informally, there might be objections of non-transparency. A combination of indicative criteria and common sense would appear to be the most promising. However, for as long as legal practitioners exist, and for as long as there is the chance of a better outcome somewhere, we will see sequential pleadings of similar arguments before successive authorities. Put more plainly, lawyers will not give up for as long as there is a forum before which they might achieve a better outcome than the one before which they have been unsuccessful.

346

Working Paper II

The Commission will be likely to intervene only reluctantly, when there is an evident danger of a constitutionally improper outcome. That one agency seems to have taken a surprising, even aberrant theory, has not sufficed to induce Commission intervention. Commission officials have said in effect ‘Good luck. Try to convince them on your own. We do not want to pre-empt authorities’ discretion at this early stage’. A senior Commission official responded recently to an expression of concern about a national authority’s preliminary taking of position in a specific case by stating that I am sure you will agree with me that the decentralised enforcement of the competition law is bound to be a failure in the absence of mutual trust between the Commission and the national competition authorities.

Such a policy is very understandable at this stage. Experience will indicate whether more interventionist measures become appropriate in future cases. They will not be if the process goes as well as the Commission hopes.

V. Different Resources for National Authorities and Different Levels of Maturity The concern, which deserves to be put plainly, is that some competition authorities may do a poorer job than others. It is sure that some have small resources, and they may lack experience. These are facts. They may be regrettable, but they do not mean that the quality of the output will be bad. Small countries can function well. Moreover, decentralisation of the EC competition rules will not collapse due to these imperfections. National competition authorities have varying levels of prestige, money, staff, experience and productivity. Belgium and Austria have been ill served, so far. Germany and Italy are very well served. Countries have to make choices about how they spend government money: courts, health care, police, education, prisons, roads are competing causes. Competition law enforcement is one of many claimants for priority. In some countries, membership of the authority is a part-time job. That is a perfectly understandable situation, and may indeed be wholesome in that it makes available in senior positions the talents of professional economists and lawyers who might otherwise not be tempted. However, arranging the calendars of nine people to sit as a plenary tribunal (as in Greece) can be very difficult, especially if the case is a complex one requiring several days of hearing. More troublesome is the danger of favouritism, and political influence. One reason why I do not favour a European Cartel Agency is that an agency needs energy and courage and history to resist political pressure. The European Commission is accustomed to coping with such pressure, even if it does not always succeed. If Member States applied the state aid rules conservatively, the

Technical Problem Areas

347

Commission could double the resources it deploys in enforcing Articles 81 and 82 EC. We have seen that France and Belgium (to name only two) have in recent history tried to avoid complying with the EC state aid rules. These governments have deployed policy and political arguments, and in specific cases sought to convince DG Competition that only a tiny stretching of established practice was necessary for particular aid schemes to be blessed. We cannot therefore be shocked if we hear that newer participants in the European Union deploy political and policy lobbying in domestic competition matters. In Denmark, a medium-sized Member State, the Parliament modified the Danish Competition Act in 2000. The changes included interpretative notes, submitted by the Ministry of Justice to the Parliament before the Bill was considered, which stated that when defining the relevant market, one should not consider ‘exclusive design’ to belong to a separate market. It is widely recognised that this insertion was made in order to prevent the Danish competition authority from deeming Bang & Olufson and Montana to have a dominant position on the market for exclusive hi-fi equipment and exclusive furniture respectively.4 In at least one candidate country, the Prime Minister’s department requested rapid endorsement of the legality of a major merger of two large enterprises with an important state aid component. The officials of the competition agency worked at high speed, and clearance was granted within a few days. In a small Member State a controversial decision on the subject of parallel trade (compelling the furnishing of unlimited quantities!) was justified in the decision by the likely beneficial effect on national exports to other Member States. I submit that we cannot have competition law enforcement without political influence being deployed by governments to influence it. Any competition agency should regularly play the role of ‘turbulent priest’ (Henry II so described Thomas Becket, who had been promoted from the post of Chancellor (perhaps 4 Betænkning

over Forslag til lov om ændring af konkurrenceloven — L 242 , Folketingstidende 1999–2000, Supplement B. p. 1283, p. 1288: ‘The principles on supply substitution and potential competition will entail that a product which separates itself only with regard to its design and the prestige connected with the product will not constitute a separate product market. The reason being that such competitive advantages are considered to be surmountable by competitors through their own design development, marketing etc.’ In Danish it reads: ‘Principperne om udbudssubstitution og potentiel konkurrence vil indebære, at et product, der alene adskiller sig fra andre produkter med same funktionalitet ved eksklusivt design eller prestige tilknyttet produktet, ikke vil udgøre et særskilt produktmarked. Årsagen hertil er, at sådanne konkurrencefordele må betragtes som mulige for konkurrerende virksomheder at overvinde gennem deres egen designudvikling, marketsføring m.v.’ Any doubt left as to the purpose behind this clarification was eliminated during the Parliament’s debate in connection with the adoption of the amendments in May 2000. One example among many was Svend Erik Hovmand (The Liberal Party) who said: ‘We are furthermore delighted that the market definitions now have been changed so that undertakings, who makes good design, functionality and quality, such as B&O, Montana Furniture and Le Klintskærme, no longer are punished’. In Danish it reads: ‘Vi har endvidere grund til at glæde os over, at markedsdefinitionen nu er blevet ændret, så virksomheder, der laver god design, funktionalitet og kvalitet, som vi har set det med f.eks. B&O, Montana Møbler og Le Klintskærme, ikke strafes’. See FF 1999–2000, Tuesday 23. May 2000 (L242) p. 8389.

348

Working Paper II

analogous to Secretary-General of the Commission) to that of Archbishop of Canterbury (Head of DG Competition?) and who suffered martyrdom because of his inflexibility on matters of principle). Governments will never be pleased by everything their agencies do and they will always press for changes in policy. I expect that this fact of life will not change rapidly. Remedies will include the passage of time; experience in resisting lobbying; acceptance nationally of the utility of the voice of competition; more cooperation to assist agencies under pressure identify the best practice, and more involvement of practising lawyers in politically-sensitive matters at an early stage, at the invitation of governments. The likelihood, indeed the certainty, that government departments will try to persuade competition authorities is not a reason for abandoning decentralisation. It is a reason to be alert during the process of hand-over to the need fraternally to pursue neutral, accurate decisions. Previous gatherings in Florence have noted that in some Member States court proceedings can be extremely slow, and in others impressively swift. Italy and the Netherlands are often contrasted. It is to be hoped that there will be no such serious disparities of speed and efficiency between the national competition authorities. If disparities were to emerge, the Commission should publicly take account of these in its interventions and decisions on allocation of responsibility.

VI.

Cooperation between National Authorities

I am struck by the remarkably modest level of cooperation between national authorities at this moment. With my Danish and Latvian colleagues, I made enquiries to national competition authorities in Scandinavia, Lithuania, Estonia and Latvia, Italy and Greece, mentioning the present gathering in Florence. We expected to find more indications of cooperation than we did. Even in Scandinavia, we found very modest evidence of current cooperation in concrete cases. No one disagreed that cooperation was desirable, and no one disagreed that the rules by which conflicts of jurisdiction would be governed should be clear. However, there were few reports of cooperation in past cases, and there were significant theoretical differences about how the allocation of cases should work, notably the degree of flexibility in allocating cases and taking jurisdiction. In order to avoid disputes over jurisdiction, Latvian and Estonian officials appear to favour strict criteria governing the allocation of cases, thus making the whole system of allocation of cases as predictable and inflexible as possible, the choice of jurisdiction being dependent solely upon objective factors, such as the degree of effect of the specific infringement on the national market. Both competition authorities are ready to accept that both can deal with a dispute simultaneously (especially in cases concerning mergers and cartels), but both lack practical understanding of how to approach the problem of whose decision should prevail.

Technical Problem Areas

349

While Estonian and Latvian officials vote for less discretion, Lithuanian officials would like to be more flexible. Their preference would be for choice based on the ‘best-placed authority’ principle, experience and skill and resources being the determining factor. As Lithuania is the largest of the three countries, it would be likely that the Lithuanian authority would be best placed in most cases (not a criticism, merely an observation). Cooperation between the three competition authorities has been rather modest, although they all acknowledge the value of mutual cooperation. Under the Trilateral Baltic Free Trade Agreement of 1994 parties undertook to cooperate within the competition area. However, this instrument requires to be supplemented with more detailed rules on how cooperation should be enforced procedurally. Baltic competition authorities have discussed the results of each other’s cases and investigations during meetings and workshops, but have never pursued common investigations. Exchange of information is limited to publicly known and generally available information. Under the Memorandum of Understanding of 1996 between the Competition Authorities of Latvia, Estonia and Lithuania, different kinds of nonconfidential information can be exchanged. However, there are no procedural rules for exchange of confidential information, and procedurally such information cannot therefore be used as proof before other agencies. There is talk of new rules to permit exchange of confidential information. The Nordic countries5 have a long-standing tradition of cooperation in various legal matters. In the field of competition law, close informal cooperation can be traced back to 1956.6 Yearly meetings, hosted by each country in turn, offer case handlers from all the Nordic competition authorities7 the opportunity to discuss issues of common interest; the Director-Generals of the agencies also meet annually; and the countries exchange information on cases of common interest. However, this cooperation has remained informal and can best be described as an ‘exchange of experience’. There has been no form of allocation of cases between the authorities. We have not been able to find any example of Nordic countries agreeing between themselves to transfer jurisdiction over a pending case. In 1998, the Nordic competition authorities decided to investigate the possibility of closer cooperation. In 1999 the Nordic cooperation committee published a Report on reinforced cooperation between the Nordic competition authorities.8 One reason for the Report was the increased risk of contradictory decisions in 5 ‘Scandinavian

countries’ will refer to Denmark, Norway and Sweden. ‘Nordic countries’ will refer to the Scandinavian countries, plus Finland and Iceland. 6 In 1956 the Nordic Economic Cooperation Counsel set up a committee whose task was to look into differences in price and monopoly regulation. This led to regular meetings between the agencies. 7 Konkurrencestyrelsen (Denmark), Konkurransetilsynet (Norway), Konkurrensverket (Sweden), Samkeppnisstofnun (Iceland), and Konkurrensverket (Finland). 8 Utvidgat Nordiskt Samarbete vid Handläggning av Konkurrensärenden, Report from the Nordic competition authorities No. 1/1999.

350

Working Paper II

cases involving several Nordic countries. There are examples of the authorities reaching different results in apparently identical or similar cases.9 The Report recommends increased cooperation in various ways, but does not mention possible allocation of cases between the authorities. With regard to extraterritorial application of national competition laws, the Report considers that there would be little political will for extending cooperation to this extent. The Nordic countries have a remarkable cultural historical and social record of cooperation. If they, with so much in common, have not handled many cases together and have not agreed who should handle which matter in case of overlap, consider the challenge which such cooperation would likely present to agencies in countries which have been political — or even military — rivals in the twentieth century. In Greece, the story is similar. Although the Hellenic Competition Committee has existed for nearly 30 years, and although it has acquired a considerable body of case-law in a wide range of industries, it has concentrated on domestic matters and has had little experience of cross-border cooperation. Its officials do, however, meet regularly with officials of other authorities, and the Committee certainly favours a reinforcing of such links. The Italian story is slightly different, reflecting the bold policies taken by the Italian Antitrust Authority (IAA). The majority of investigations have related to facts and behaviour which affect the Italian territory. In many decisions, however, the IAA determined the existence of a dominant position in a European or worldwide market, or fined agreements which affected not only Italian territory. From a theoretical point of view, an in-depth investigation opened by the IAA could be in conflict with other investigations handled by other national authorities. Until now there has been no provision in the Italian Antitrust Law which regulates a conflict of competence between the IAA and other national authorities. The Commission’s involvement would be indispensable to address such a conflict if one arose which could not be solved bilaterally. The investigative powers of the IAA do not extend beyond Italy’s borders. If the IAA opens an investigation against a non-Italian company (and it has done so), the IAA notifies such decision by means of the competent Italian foreign services. It cannot conduct a dawn raid outside of Italy (and likewise, no other competition authority of a Member State has conducted a dawn raid in Italy). Thus cooperation with other national competition authorities is based on exchange of information, but there is no formal agreement which regulates such exchange. The IAA is not entitled to provide other national competition 9 For

example in the Øresunds Consortium case, the Danish and Swedish competition authorities reached different results. This case, decided before the current Danish competition act under a ‘control of abuse’ law where that which was not prohibited was lawful, related to the terms of the consortium governing a bridge between Denmark and Sweden. The European Commission granted a negative clearance to cooperation between cement producers. However, the Danish authority and the Swedish authority then issued conflicting decisions (albeit pursuant to different legislations), one condemning and one approving. After some well-mannered Nordic wrestling, consensus was achieved.

Technical Problem Areas

351

authorities with confidential information provided by the parties during another investigation or a merger proceeding.

VII.

Conducting the Proceedings in a ‘Foreign’ Language

Languages, whose multiplicity forms part of Europe’s richness and forms an administrative curse for the European Institutions, are being handled in a practical manner. English is accepted in many small countries: in March 2002, in both the Czech Republic and Greece, advocates have to my personal knowledge been heard by the national competition authority speaking English, with interpreters.

VIII.

Problems of Extra-Territoriality and of Fines

It is to be expected that national competition laws will gradually become less and less important substantively, although some authorities (notably the Bundeskartellamt) regret this.10 Correspondingly, we expect more application of the EC competition rules where the matter affects several Member States. Many national authorities have so far never applied EC law, although they have the right to do so. As a majority of laws use language parallel or identical to Articles 81 and 82 EC, this is of little substantive consequence. However, there is a question of territorial jurisdiction. When a national authority takes up a case affecting the territories of two or more Member States, it necessarily gathers evidence about events occurring outside its country. If it imposes a fine, will the fine reflect the effect or extent of the infringing conduct outside its country? If not, and if the case is serious, then the penalty would be much lower than what the European Commission might consider appropriate. If yes, then more attention needs to be given to whether such a practice is constitutionally compatible with the national and Community legal orders. It is my impression that the Italian authority has gone further than most of its peers in arriving at a point which may be controversial.

IX.

Reducing the Risks of Inconsistency

Would it be feasible to formalise a rule that the decision of one national competition agency stops another national competition agency taking up the same 10 Statement

on the reform of the European competition rules at the Hearing of the European Parliament Committee on Economic and Monetary Affairs on 27 February 2001 by Dr Ulf Böge, President of the Bundeskartellamt.

352

Working Paper II

matter? Probably not realistic, though it would be consistent with the principle in civil litigation of res judicata, that a final judgment of a competent court is conclusive upon the parties in any subsequent litigation involving the same cause of action. All European states recognise this principle. Furthermore, Article 4 of Protocol 7 to the European Convention on Human Rights protects the right not to be tried or punished twice under the criminal law: No one shall be liable to be tried or punished again in criminal proceedings under the jurisdiction of the same State for an offence for which he has already been finally acquitted or convicted in accordance with the law and penal procedure of that State.

Similarly, could one formalise a rule whereby the rejection of a complaint by one national competition agency stops another national competition agency receiving the same complaint? If we had 15 equally matched, equally experienced, homogeneous competition authorities, there would be a better chance of agreement on the proposition that once a matter has been decided by one authority, the same question could not thereafter be examined by another authority. Given the existing disparities in seniority, size, resources and experience, especially with accession drawing near, it seems most unlikely that this would be acceptable today. In addition, even if such a rule were to be adopted, it would not be difficult to find arguments for the proposition that the facts or legal issues or market shares or barriers to entry were different in the second country. Similar difficulties would await the perhaps milder proposition that the rejection of a complaint or the closing of an own-initiative investigation into a matter by one authority should prevent an identical matter being taken up by another authority. It would appear to be more promising to consider what level of deference ought formally to be accorded by one authority to the decisions of another. There will never be a strict rule of stare decisis in the English sense, but it ought to be possible to formulate a principle whereby each authority is obliged to give careful attention to decisions of other authorities and national courts on matters touching the same issue. We may note that in A v. National Blood Authority,11 Burton J of the English High Court reviewed the jurisprudence of several Member States when considering how to interpret the 1985 EC Product Liability Directive: was it appropriate to impose strict liability on the suppliers of tranfused blood which conferred benefit to the patient but which also caused an infection due to the presence in the blood of an allegedly undetectable hepatitis virus? Many difficult questions of law arose, and their analysis was helped by looking at ‘foreign’ authorities. Particular attention was given to the judgment of the Bundesgerichthof on injuries caused by an exploding bottle, a

11 [2001]

3 All ER 289–416.

Technical Problem Areas

353

judgment which was accorded considerable weight by the English judge, not just because of its conclusion but of how it reached that conclusion. There should be no difficulty in reaching consensus that foreign competition law precedents are not to be neglected.

X.

Conclusion

Decentralisation of enforcement of the competition rules is desirable, and should have been started earlier. It will be a success if practitioners and officials are willing to trust the quality of the national decisions which apply EC law locally. If there is no such trust, then a perfect constitutional structure will not achieve the desired reform, as the Commission will remain omni-present. The above difficulties are genuine and foreseeable. They do not detract from the ultimate merit of reform, but they illustrate how delicate will be the process. There are real, practical difficulties associated with transferring to national authorities responsibility for policing a law which jurisdictionally requires crossborder effect, and where the relevant facts will usually be found in several Member States. It is very likely that identical or very similar questions will be presented to several authorities in parallel. Different outcomes have occurred. The hard and soft rules for addressing this problem are an adequate basis for addressing problems, but success will require further circulation of the proposed rules for practical input; and it will require the growth of confidence and mutual trust which events like these gatherings in Florence foster. There is surprisingly limited experience of cooperation in concrete cases by national competition authorities today. The Commission is trying to encourage us to look across the valley to the new territory where the law is enforced rationally and consistently. It should give attention to the obscurities occupying the slopes of the valley. It needs to avoid the situation arising where it is the automatic ‘appellate instance’ by any party which has been unsuccessful locally. It seems doubtful if the Commission’s intervention should be offered when an agreement or behaviour touch three or more Member States, as too many will then fall under its jurisdiction. The Commission’s fostering of regular consultation between national competition authorities about common problems is very highly desirable. It should be borne in mind that the Commission’s Notice on cooperation with the national courts,12 although an attractive idea, was little used. Practitioners and courts were perhaps insufficiently aware of its existence, and the Commission’s responses to questions put to it by courts were cautious and slow to emerge. It is to be hoped that as the individual officials and agencies involved are quite small in number (far smaller in number than national courts), there will be more success in creating from them a reliable network. 12 Notice on Cooperation between National Courts and the Commission in applying Article 85 or 86 of the EC Treaty, OJ [1993] C 39/6.

III Anne Willem Kist* Exchange of Information: Scope and Limits Seen From the Perspective of the National Competition Authorities

I.

Introduction

The proposed new system for enforcing European competition law is based on the notion that full cooperation yields more than the sum of the parts.1 By using the resources of the various competition authorities more efficiently, more infringements of competition rules can be detected, and the offenders can be fined. To achieve this, agreements about the allocation of work should be made within the network.2 If such a system is to function, a number of principles about which all members can agree will have to be formulated. The basis should be the equality of all members of this network. The network may be viewed as a web, in which the Commission is at the centre due to its special role as the guardian of the Treaty. The network should therefore have both a horizontal and a vertical character. Several instruments have been included in the draft Regulation to ensure that the network functions effectively.3 One of these instruments is the authority of the competition authorities to exchange information with each other. The question is whether these instruments are adequate or whether further rules have to be introduced and, if so, at what level these rules should be determined. This paper will discuss in more detail the possibilities for exchanging information, as set out in Article 12 of the draft Regulation. In Section II, attention will be paid to the present methods of exchanging information. Since the ability of the various competition authorities to exchange information currently depends on national legislation, Dutch legislation will be discussed. In Section III, the possibilities for and obstacles to exchanging information will be examined in the light of the draft for a Council Regulation on the implementation of the rules laid down in Articles 81 and 82 EC. It is essential for the functioning of the network that this issue is properly regulated.

* Director-General of the Netherlands Competition Authority (Nma). 1 Draft Regulation for a Council Regulation on the implementation of the rules laid down in Articles 81 and 82 of the Treaty, OJ C 365/284 [2000]. 2 See the written contributions by Profs. J. Fingleton and I. Forrester in this volume. 3 See Article 11 of the draft Regulation.

356

Working Paper III

It will be argued that, because Article 12 of the draft Regulation grants a right and does not impose an obligation to exchange information, it is necessary to reach further agreement within the network. It is necessary, for instance, to draw up further regulations within the network relating to the acceptance of one another’s confidentiality regimes and the scope of information exchange.

II.

Exchange of (Confidential) Information: The Present Regime

Under the OECD Recommendation on anticompetitive practices affecting international trade in relation to cooperation between member countries, Member States should cooperate in developing or applying mutually satisfactory and beneficial measures for dealing with anticompetitive practices in international trade.4 In this regard, they should supply each other with such relevant information on anticompetitive practices as their legitimate interests permit them to disclose. They should allow, subject to appropriate safeguards — including those relating to confidentiality — the disclosure of information to the competent authorities of Member States by the other parties concerned, unless this would be contrary to significant national interests. Such disclosure could be unilateral or in the context of bilateral or multilateral agreements.

1.

Exchange of Information Between the Commission and the National Competition Authorities

Regulation 17/62 provides for information flows between the Commission and national authorities (vertical flow). Several articles of Regulation 17/62 deal with the Commission’s obligation to provide national authorities with information.5 When received by the national authorities, this information may only be used for the proceedings for which that information was requested. In other words, this information may not be used as evidence in national proceedings, or any other proceedings for that matter.6

2.

Exchange of Information Between National Competition Authorities

The possibility of exchanging information between national competition authorities therefore depends on national legislation and varies from one state 4 OECD

(1995): ‘Revised Recommendation of The Council Concerning Co-operation between Member Countries on Anticompetitive Practices Affecting International Trade’, 27 and 28 July 1995 — C(95)130/FINAL. 5 In this regard, see also the written contribution of Marc van der Woude for this volume. 6 Article 20 Regulation 17/62 and Case C–67/91 Spanish Banks [1992] ECR I–4785.

Technical Problem Areas

357

to the next. Regulation 17/62 does not provide for the possibility of exchanging information between national competition authorities (horizontal exchange).

3.

Exchange of Information: The Netherlands

In principle, information received by the NMa (the Dutch competition authority) from companies or other parties is made public during proceedings in relation to the granting of an exemption or the imposition of a sanction. The criteria applicable to the nondisclosure of information are contained in section 10 of the Government Information (Public Access) Act [Wet Openbaarheid Bestuur, WOB, referred to below as the ‘Government Information Act’].7 The exchange of information is governed by section 91 of the Dutch Competition Act.

3.1.

The Dutch Competition Act

Information or data concerning an undertaking that is obtained in the course of any activity in relation to the implementation of the Dutch Competition Act may be used solely for the purpose of the application of that Act.8 By way of departure from this rule, the Director-General is authorised to provide information or data obtained in the performance of the duties assigned to him under this Act to: • a foreign institution that is responsible for the application of competition rules, pursuant to national statutory provisions, in so far as such information or data are, or could be, of significance for the performance of the tasks of that institution and, in the Director-General’s opinion, the provision of such information or data is in the interests of the Dutch economy; • an administrative body that is responsible for tasks relating, or partly relating, to the application of competition rules, pursuant to statutory provisions other than the Act, in so far as such information or data are, or could be of significance for the performance of the tasks of that body. It is also necessary that the confidentiality of the information or data is sufficiently protected, and there is sufficient assurance that the information or data will not be used for any purpose other than that for which they are provided. 9 In verifying these provisions, the Director-General shall consider not only the statutory provisions regarding confidentiality and the use of data in the country concerned, but also their application in practice.10 7 Section 18(4), Competition Act; Section 65(1), 8 Section 90 of the Dutch Competition Act. 9 Section 91 of the Dutch Competition Act. 10 TK 24707, no. 3, para 12.2.

Competition Act.

358

3.2.

Working Paper III

Government Information (Public Access) Act

Section 10 of the Government Information Act provides an exception to the principle that information should be made public during proceedings. The criteria applicable to this exception may be subdivided into absolute and relative. The absolute exception relevant to competition law relates to confidential operating and production data.11 From case-law it appears that this exemption criterion, pursuant to section 10(1)(c), should be interpreted restrictively. According to the Judicial Division of the Council of State, data is deemed to be operating data and production data if and in so far as information in relation to the technical operations or the production process, or with regard to the sale of products or the supplier’s circle of customers, may be obtained or derived from such data.12 Section 10(2) of the Government Information Act contains a non-exhaustive list of the relative exception criteria. With regard to the relative exception criteria, the interests in question and the importance of providing information are weighed against each other. For example, the second paragraph of section 10 of the Government Information Act provides the following summary of interests: • international relations;13 • inspection, monitoring and supervision by or by order of public authorities;14 • respect for the privacy of individuals.15 This may include, for instance, the data that the NMa has received from other competition authorities, information regarding investigations, or the identity of a natural person. The relative exception criteria apply if it is concluded that the data are not operating data or production data or, if they may be deemed to be such, that they are not deemed to be confidential. In assessing the interests, what is important are not the personal interests of the person who submitted the request, but the public interest in providing the information.

3.3.

Exception Criterion: Information Received within the Framework of International Relations

The NMa may appeal to this exception criterion with regard to disclosure on the grounds that it has to honour international relations if the relations between the competition authorities would become less cooperative if the confidentiality of the data were not to be honoured. It is not necessary that there should, in fact, be a deterioration in relations. 11 Section 10(1)(c) of the Government Information (Public Access) Act. 12 See for instance, rulings of the Judicial Division, 30 September 1985;

Law Division, 1987/43. 13 Article 10(2)(a) of the Government Information (Public Access) Act. 14 Article 10(2)(d) of the Government Information (Public Access) Act. 15 Article 10(2)(e) of the Government Information (Public Access) Act.

rulings of the Administrative

Technical Problem Areas

359

A fairly lenient line has been followed with regard to this exception. If and when a foreign government or an international organisation declares that documents are not for publication, or if they have not yet expressly been released for publication, it is an obstacle to disclosure. No case-law has yet been developed with regard to the exchange of information between competition authorities. However, if this line is extended, the NMa would have to honour the fact that data are considered to be confidential in the country from which they originate. By allowing a dossier to be inspected or by issuing a public decision, no problems are expected to arise from the present situation, even if the NMa regards less information to be confidential than the competition authority that initially handled the case.

3.4.

Practical Experience

In the past four years, the Director-General of the NMa has made little use of section 91 of the Competition Act. The information exchanged with OPTA (the regulator of the post and telecommunications sector) is worth noting in this regard. In addition, the issue of exchanging information with foreign competition authorities has occurred on two occasions. The NMa applies five criteria when deciding whether it is possible to give information to other authorities in accordance with Section 91 of the Competition Act. Section 91 of the Competition Act gives the NMa the discretionary power to exchange information. Even if all the criteria have been met, the DirectorGeneral of the NMa may nevertheless decide not to exchange information. These criteria are: • the institution to which the information is sent must be a foreign institution responsible for the application of competition rules in accordance with national statutory regulations; • the data or information must or must potentially be of importance in carrying out the duties of the institution in question (in other words, in applying the competition rules); • the provision of such information must be in the interest of the Dutch economy, in the opinion of the Director-General; • the confidentiality of the data or information must be sufficiently guaranteed; • there are sufficient guarantees that the data or information will not be used for a purpose other than that for which it was provided. In assessing the guarantees that the data that is deemed confidential by NMa will be kept confidential, it appears that the Dutch and other Member States’ systems with regard to business secrets are fairly similar– as far as our experience goes — so this need not be an obstacle to the exchange of information. Aditional agreements could possibly be made to solve any problems. The latter point — that the

360

Working Paper III

information may not be used for other purposes — could cause problems if a foreign competition authority is obliged to provide information to other public authorities in certain circumstances (for example, the foreign competition authority might be under an obligation to provide information to its nation’s Department of Internal Revenue or Public Prosecution Service). Additional bilateral agreements may be entered to prevent this from happening. If a decision to exchange information is made, a further question arises as to whether the parties involved are or must be informed. Three options may be identified: • the provision of data/information relating to a particular company, without informing it of this; • the provision of data/information relating to a particular company, subject to (written) notification of this to the company in question. In this notice, the company is given a period during which it can try to prevent the issuing of the data or information by instituting civil proceedings; • the provision of data/information relating to a certain company on the basis of a decision in relation to the company. The question that needs to be asked in this case, and that is not easy to answer, is whether this is a decision/ruling in terms of section 1(3) of the General Administration Law Act and not a preparatory procedure, as referred to in section 6(3) of the General Administration Law Act. In the case in which information was exchanged with OPTA, option 2 was applied. The advantage of this procedure is that the legality of such a decision is more or less determined early in the procedure. On the basis of Nma’s limited experience, it has become clear that the exchange of information is a time-consuming activity. This is due to the need for the competition authority that wishes to receive the information to determine the procedure, and due to the conditions applied by NMa pursuant to Section 91 of the Competition Act. To prevent companies from unduly hindering an investigation by appealing against an exchange of information, it must be stressed that the legality of this exchange can only be assessed by a court at the end of the proceedings. There is, of course, the possibility of a judicial review, but only at the end of the procedure after the NMa has made a formal decision. A situation in which the Director-General of NMa spontaneously decides to make information available to a foreign competition authority (in other words, without a formal request) has not yet arisen.

3.5.

Lessons to be Learned

The greatest obstacles to the exchange of information up until now have been the issue of whether the confidentiality of the data or information is sufficiently

Technical Problem Areas

361

guaranteed and whether there are sufficient guarantees that the data or information will not be used for any purposes other than those for which they were provided. In this regard, the NMa has to carry out extensive research into the legislation of the Member State to which the information would be provided. The first obstacle may be resolved by agreeing to accept the confidential status of each other’s information. For this purpose, the competition authority that provides the information may issue a list in which it expressly states which documents are confidential under its legislation. The recipient authority will then respect this notice. In addition, it appears that each request for the exchange of information has raised different questions, and the procedure is therefore anything but as simple as it seems at first glance.

III. 1.

Exchange of Information: Article 12 of the Draft Regulation Article 12

In the decentralised system, a competition authority is obliged to apply Articles. 81 and/or 82 EC if interstate trade is affected.16 If this is the case, notification of the case has to be given within the network17 so that a decision can be made about which authority is best placed to deal with it. A dossier consists of various types of information. It relates to both confidential and non-confidential data. Article 12 of the draft Regulation includes regulations in this regard, which explicitly state that competition authorities may provide one another with any matter of fact or law including confidential information. Article 12 of the draft Regulation reads as follows:18 1.

2.

For the purpose of applying Articles 81 en 82 of the Treaty, the Commission and the competition authorities of the Member States shall have the power to provide one another with and use in evidence any matter of fact or of law, including confidential information. Where national competition law is applied in parallel with Community competition law pursuant to Article 3, information exchanged under this Article may also be used for the application of national competition law. No sanction of imprisonment can be imposed on the basis of information provided under paragraph 1, except if the transmitting Member State’s law foresees such sanctions for the same kind of infringement.

16 Article 3 of the draft Regulation. 17 Article 11(3) of the draft Regulation. 18 Progress Report to the Council of the

European Union, Draft Regulation for a Council Regulation on the Implementation of the Rules on Competition Laid Down in Articles 81 and 82 of the Treaty and Amending Regulation (EEC) No 1017/68, (EEC), No 2988/74, (EEC) No 4056/86 and (EEC) No 3975/87, SN 4467/01, 7 November 2001.

362

Working Paper III

If it is clear at an early stage in the case that interstate trade is affected and the case is transferred, very limited information will be exchanged. On the other hand, it is also possible that the fact that interstate trade is affected will only emerge during the investigation — and therefore at a later stage in the handling of a case — and that a different competition authority will handle the case or participate in the handling of the case. The dossier that is transferred in this instance would be much more extensive. In this case, a competition authority may transfer the entire dossier, including the confidential information.19 Problems will only emerge in relation to the exchange of information in so far as it is confidential.

2.

Confidential Information in the Draft Regulation

Article 27 of the draft Regulation makes provisions in relation to confidential information. In principle, this stipulates that information collected by the Commission in accordance with its powers shall only be used for the purpose for which it was acquired.20 In addition, this information may be exchanged in accordance with Articles 12 and 15. Information covered by the obligation to maintain professional secrecy and information obtained on the basis of an investigation or the exchange of information will not be disclosed. Article 13(1) of Regulation 2842/98 in relation to hearing interested parties and third parties in certain proceedings, pursuant to Articles 81 and 82 EC,21 contains regulations with regard to the confidentiality of certain information. The Commission neither discloses nor allows inspection of information, including documents, that contain the business secrets of, for instance, persons against whom the Commission has raised objections, applicants and plaintiffs or other third parties, nor other confidential information, nor internal documents of the authorities. The Commission will take suitable measures to provide access to the dossier and in doing so will ensure that the business secrets, internal documents of the Commission and other confidential information are protected. The concept of ‘confidential information’ has been specified in more detail in the judgments of the Court of Justice.22 This has also been worked out in further detail in the Commission Notice on the internal rules of procedure for processing requests for access to the dossier in cases pursuant to Articles 81 and 82 EC, Articles 65 and 66 of the ECSC Treaty and Council Regulation (EEC) No 4064/89.23

19 Explanatory

Memorandum to the Draft Regulation for a Council Regulation Implementing Articles 81 and 82 of the Treaty, COM (2000) 582 final of 27.09.2000. 20 Article 27(1) of the draft Regulation. 21 Regulation of 22 December 1998, OJ L 354/18 [1998]. 22 See, for instance Case 53/85 AKZO [1986] ECR 1965, and Joined Cases 142/84 and 56/84 BAT v. Reynolds [1987] ECR 487. 23 Commission Notice of 23 January 1997, OJ C 23/3 [1997].

Technical Problem Areas

363

In the light of the above, it appears that the Commission will not make information that it considers to be confidential available to third parties. However, this implies nothing about the acceptance of the confidentiality of data that are considered to be confidential under the regime of the Member State, and it does not provide a scheme for the confidentiality of data in the event of the exchange of information between the competition authorities of two or more Member States.

3.

Scope and Limits Regarding the Exchange of Information

As Article 12 of the draft Regulation does not impose an obligation to exchange information, a consequence may be that not all competition authorities will adopt the same policy, for instance with regard to the exchange of confidential information. Although one might assume that all competition authorities aim to achieve the same objectives — preventing and penalising infringements of competition rules — I believe it is necessary that the scope of information exchange within the network should be regulated. This would consequently give rise to some obligation to exchange information.

3.1.

Independent Competition Authorities

Not all competition authorities are independent of the Ministries of Economic Affairs in their countries. This might present problems with regard to the exchange of information in relation to companies that compete with state enterprises in these Member States. The competition authority, and consequently the Ministry, might acquire sensitive competitive information as a result of the exchange of information. This could result in problems comparable to those that emerged in SEP.24 A solution will have to be found to this problem within the network. In the most extreme cases, a competition authority might decide not to exchange information. It would be entitled to do so in light of the text of Article 12 in the draft Regulation. Should the authorities decide to exchange information, a question remains as to whether the company that has submitted this information may lodge an appeal against the exchange. After all, Article 12 gives rise to a right and not an obligation to exchange information, so it provides no legal grounds that make the exchange of information obligatory. On the basis of the ruling in the SEP case, Article 27 of the draft Regulation offers insufficient assurance that the information that is obtained will not be used for purposes beyond the application of competition law.

24 Case

C–36/92P SEP [1994] ECR I–1911.

364

3.2.

Working Paper III

Mutual Recognition of Confidential Information

With regard to the exchange of information: it might give rise to problems if the competition authority handling the case accepts more information on a confidential basis than the competition authority that makes the final decision. An example of this is the scope of the legal privilege principle. This principle includes the correspondence of an in-house lawyer in some legal systems,25 while in others it does not.26 After all, what is considered to be confidential is determined by national legislation and, as far as Dutch law is concerned, an administrative authority has considerable room to exercise discretion. This uncertainty might cause undertakings to be cautious about the extent to which they provide the competition authority with information. This is an obstacle to the proper execution of a competition authority’s duties and is therefore undesirable. To ensure that it is not necessary to reach agreement about whether information is confidential or not each time information is exchanged, it is essential that the concept of ‘confidential information’ be defined in more detail within the network. This will also reduce legal uncertainty amongst undertakings with regard to the status of information that they submit to a competition authority under a particular national regime. One solution is mutual recognition of each other’s regimes with regard to confidential information. In principle, this has been included in the provisions relating to the exchange of information contained in the Competition Act, as information may only be made available if the confidentiality of the information is adequately protected. This implicitly assumes that the fact that information is deemed to be confidential in the Netherlands means that it will also be deemed to be confidential by the other competition authority and may therefore not be disclosed. The NMa may therefore only exchange information when the recipient authority guarantees adequate protection of data that is deemed to be confidential under Dutch law, and that the information may not be used for purposes other than ensuring compliance with competition law.

3.3.

Information Received on the Basis of a Leniency Programme

In addition, there are a number of specific issues that might lead a competition authority to decide not to provide another competition authority with information. These may include, for example, information that was acquired within the framework of the leniency programme of a competition authority. Not every EC Member State has a leniency programme. For instance, the Dutch NMa is in the process of establishing one. Setting up a national leniency programme is only useful if those who wish to make use of it can be sure that 25 For example in the UK — see the written 26 Case 155/79 AM & S [1982] ECR 264.

contribution of Margaret Bloom for this volume.

Technical Problem Areas

365

they will be punished in accordance with the programme. On the basis of the principles applied within the network, a dossier may be transferred to another national competition authority because the latter is the more appropriate authority in terms of the allocation criteria. If this competition authority does not have a leniency programme or a similar arrangement, the consequences for the undertaking concerned will be considerable. Although one competition authority grants a (considerable) reduction in the fine payable, this may not extend to the investigating competition authority. This will result in a situation in which the national leniency programme is not effective, a situation in which an authority refuses to exchange information, or a situation in which the authority requires guarantees that any sanctions will take the leniency programme into account. This makes it clear that this is one of the topics that will have to be discussed if the scope of information exchange is to be determined.

3.4.

Investigation

The following problem may also arise: in cases in which the authority starts an investigation, there is always a chance that during the investigation the authority will find that interstate trade is affected and that, according to the criteria of case allocation, another authority is better placed to deal with it. However, if the competition authority has gathered information on the basis of investigative powers that the competition authority that receives the information does not have, the question arises as to whether the latter competition authority may use this information. Article 12(1) of the draft Regulation stipulates that the competition authorities of the Member States not only have the power to provide one another with information, but also that they may use in evidence any matter of fact or of law, including confidential information. The question that arises is what a Dutch court would do with material acquired in accordance with Article 12 from, for instance, Ireland. In Ireland, competition law is enforced within the framework of criminal law and investigations are carried out with the use of the investigative powers granted under criminal law. In the Netherlands, it was explicitly decided to withdraw competition law from domain of criminal law and not to grant the Dutch competition authority criminal investigative powers.27 This still leaves aside possible problems relating to the ECHR [the European Convention of Human Rights]. 28 The following is a further point that plays a role in relation to the exchange of information. In accordance with Article 21 of the draft Regulation, a competition authority may carry out verification on behalf of another competition authority. The information obtained as a result may be passed to the competition authority that requested the information in accordance with Article 12 of the draft 27TK 24 707, No. 3, 11.2: Enforcement of administrative law. 28 See the written contribution of Wouter P.J. Wils to this volume.

366

Working Paper III

Regulation. The question that thus arises is whether the verifying authority may decide not to hand over information that has been gathered after a request for verification was issued. As the text reads at present, it appears that a competition authority is free to decide whether or not to carry out the verification: The competition authority of a Member State may in its own territory carry out any fact-finding measure under its national law on behalf and for the account of the competition authority of another Member State in order to establish whether there has been an infringement of Article 81 or Article 82 of the Treaty (emphasis added).

If it decides to undertake the verification, the competition authority is then expected to hand over the information: ‘It shall transmit the information collected to the requesting authority in accordance with Article 12 of this Regulation’. However, as Article 12 does not carry an obligation to exchange information, the competition authority may still decide not to pass on the information. Agreements will have to be made within the network in this regard. Finally, the question may arise as to how one should deal with information that is obtained from another competition authority in accordance with Article 12 of the draft Regulation, if during the proceedings the information proves to be flawed. In which Member State and against which competition authority should proceedings be instituted if an appeal is filed on the grounds that the evidence was obtained illegally? This is another topic about which agreements will have to be reached within the network.

IV.

Conclusion

In accordance with the above arguments, it should be clear that the provisions of Article 12 of the draft Regulation provide a basis for the exchange of information that is essential in a system in which the application of competition law is decentralised. A wide range of topics may be mentioned that should be worked out in more detail by the competition authorities, which could then exchange information. In my opinion, this means that further rules will have to be agreed within the network. To ensure that policy differences do not arise between the competition authorities with regard to the exchange of information, further agreements will have to indicate when competition authorities will exchange information or make the exchange of information obligatory when certain conditions are met. This may be the case, for instance, if agreement is reached that the recipient competition authority will honour the confidentiality of the information. It is also necessary to agree that the sanctions imposed will take into account the leniency policy of the competition authority that provided the information. Nevertheless, a number of questions still remain. A solution will have to be found for information that is transferred, but that is flawed. The same applies

Technical Problem Areas

367

to the use of information obtained on the basis of investigative powers that the recipient competition authority does not possess. It is questionable whether the text of Article 12 of the draft Regulation is sufficient for a Court to allow the information to be admitted as evidence: the Commission and the competition authorities of the Member States shall have the power to provide one another with and use in evidence any matter of fact or of law.

Some of these questions will be resolved in practice but as we have learnt from our experience with the Dutch Competition Act, new questions will certainly arise that have not been considered in the new Regulation.

IV Marc van der Woude* Exchange of Information Within the European Competition Network: Scope and Limits

I.

Introduction

One of the basic ideas underlying the Commission’s proposal for a new procedural regulation on the application of Articles 81 and 82 EC is its intention to focus its resources on the fight against cartels and other hard-core restrictions.1 As business is increasingly international in nature, restrictive practices tend to be so too. The Commission therefore intends to concentrate its activities on cartels and abuses that have a European dimension, whereas the competition authorities of the EC Member States are expected to deal with infringements affecting one or more national markets.2 Community and national resources should be allocated in the most effective manner: a case should be handled by the authority or authorities that are best placed to deal with the practice in question. Coordinated action requires that authorities communicate with each other and exchange information to that effect. However, the present procedural system does not guarantee a free flow of information between the European competition authorities. Rules governing the exchange of information differ from one jurisdiction to another and do not necessarily allow such an exchange. For example, Article 91 of the Dutch Competition Act explicitly authorises the Dutch Competition Authority to pass information to the Commission. The reverse is not true: Article 20(1) of Regulation 17/62 states that information supplied to the Commission must only be used for the purposes of the relevant Community investigation. This means, according to the AEB case law,3 that national authorities must not use the documents they receive from the Commission as evidence in

* Erasmus University of Rotterdam, and Nauta Dutilh, Brussels. 1 Proposal for a Council Regulation on the implementation of the rules laid down in Articles 81 and 82 of the Treaty, presented by the Commission on 27 September 2000, OJ C 365/284 [2000]. 2 See inter alia, Schaub A. (2001a): ‘The direction of competition policy: reconciling national and international objectives’, paper presented at the Annual Fall Conference on Competition Law, Ottawa, 21 September 2001; Schaub A. (2001b): ‘The global competition forum: how it should be organised and operated’; EPC Brussels, 14 March 2001; Monti M. (2000): ‘Nouveaux enjeux pour la politique européenne de concurrence’, Commission des affaires étrangères de l’Assemblée nationale, Paris, 17 October 2000. 3 Case C–67/91 Direccíon General de Defensa de la Competencia v. Asociacíon Española de Banca Privada (AEB) a.o. [1992] ECR I–4785.

370

Working Paper IV

national proceedings, even if these proceedings relate to the application of EC competition rules. National authorities wanting to take over a case that is being heard by the Commission, or wishing to act in parallel with it, have to re-collect that evidence using their own means. From the standpoint of the national competition authorities, this duplication of work is an obstacle to effective and coordinated law enforcement throughout the Community. By contrast, the present system offers a clear procedural safeguard from the undertakings’ point of view. When they supply information, whether or not voluntarily, they know in advance what the possible consequences could be. Allowing the competition authority to freely dispose of the information may expose undertakings to an unknown number of investigations, which in certain cases may even lead to criminal sanctions. As a rule, undertakings will not have any interest in waiving their right to oppose the transmission of information from one competition authority to another.4 The present rules on information exchange are therefore hard to reconcile with the two major objectives of the Commission’s reform proposal: a concentration of their own resources concerning hard core restrictions of competition, and more reliance on national enforcement resources.5 In these circumstances, the Commission has proposed to modify the existing system by inserting in its draft Regulation a new provision that should guarantee a free flow of information between all jurisdictions concerned. Article 12 of the draft Regulation reads as follows: 1.

2.

Notwithstanding any national provision to the contrary, the Commission and the competition authorities of the Member States may provide one another with and use evidence on any matter of fact or of law, including confidential information. Information provided under paragraph 1 may be used only for the purpose of applying Community competition law. Only financial penalties may be imposed on the basis of information provided.

Although the idea does not seem to be far advanced at this stage, it is intended that establishing a special intranet linking EC and Member State competition authorities should facilitate the free flow of information under Article 12 of the draft Regulation. The object of the present contribution is to examine whether this provision will indeed change the existing situation. Will it offer more possibilities for coordinated

4 See by analogy, for the fight against restrictive practices overarching the Atlantic, Report of 29 January 2002 from the Commission to the Council and the European Parliament on the application of the Agreement between the European Communities and the Government of the USA regarding the application of their competition laws — 1 January 2000 to 31 December 2000, COM (2002) 45 final. 5 Wils W. P. J. (2000): ‘The modernisation of the enforcement of Article 81 and 82 EC: a legal and economic analysis of the Commission’s proposal for a new Council regulation replacing Regulation 17’, in Hawk B., ed., Annual Proceedings of the Fordham Corporate Law Institute 2000, pp. 313 et seq.

Technical Problem Areas

371

antitrust enforcement than are available under present rules? What changes is it likely to bring about? Will it negatively affect the position of undertakings confronted with antitrust investigations? Finally, if the answer to this latter question is positive, do undertakings benefit from any kind of judicial protection? The present contribution is written from the perspective of an academic who is also a private practitioner. Although the author is conscious of the need for effective antitrust enforcement, he has concentrated on identifying legal issues that undertakings can raise in order to oppose the information exchange contemplated in Article 12 of the draft Regulation. In order to stimulate debate, he deliberately chose a rather provocative approach leading to the conclusion that this new provision is to be considered compatible with primary Community law only if additional safeguards are developed to protect the rights of defendants and the confidentiality rights of the undertakings concerned. The first section of this article gives an overview of the existing rules contained in Regulation 17/62. The second deals with the new rules and identifies the major changes. The third is concerned with identifying the legal arguments that undertakings involved in antitrust proceedings might challenge under the new enforcement procedures. It should be noted that the present contribution only deals with relations within the network of authorities envisaged by the draft Regulation. It will not deal with relations with authorities of third states, or with relations with national courts, even though these may be relevant when discussing the source of the information that the members of the network want to exchange. Similarly, the contribution will focus on the enforcement of market conduct rules, and it will not deal with merger control.

II.

Information Exchange Under the Present Rules

Regulation 17/62 provides for a centralised system for the application of EC competition law. Article 9(1) grants the Commission exclusive jurisdiction in applying Article 81(3) EC, and Article 9(3) gives it the power to block the application of Articles 81(1) and 82 EC by national authorities. These authorities are supposed to assist the Commission in assuming its central role, but not to replace it. Regulation 17/62 puts the Commission at the top of a system in which vertical relations prevail. It does not envisage any form of horizontal cooperation between national authorities. In other words, Regulation 17/62 does not concern the application of EC competition law by national authorities; the Member States are not obliged to appoint authorities capable of applying Articles 81 and 82 EC. Luxembourg, for example, does not have a competition authority. Nor does Regulation 17/62 pay any attention to enforcement of national competition law, as it is not based on Article 83(2) EC. Even despite this centralised system, the Commission has sought to establish some form of cooperation and case allocation with national authorities. In 1997

372

Working Paper IV

it published a notice on cooperation with national authorities (hereafter ‘the cooperation notice’).6 Paragraph 7 of the cooperation notice states that cooperation can only be fruitful if national authorities and the Commission stay in close and constant touch. However, the limits of this relation are set by Regulation 17/62; information may circulate between national authorities and the Commission only insofar as it does not conflict with its provisions.

1.

Which are the ‘Authorities’ of the Member States?

The first issue to be clarified with regard to Regulation 17/62 is the identity of the national competition authorities, as this determines with whom the Commission may or may not exchange information. Article 10 is the basic provision in this respect. It is entitled ‘Liaison with the Authorities of the Member States’, and it distinguishes two types of national authorities: authorities that are called ‘the competent authorities’, and the officials appointed as representatives in the advisory committee.7 These two categories can, but need not, be the same. The European Court of Justice has not given any formal indication as to the nature of these authorities. In BRT/SABAM and Nouvelles Frontières, it stated that this concept referred either to administrative authorities in charge of applying national competition law or to judicial authorities entrusted with a similar task.8 The lack of compelling guidance in EC law probably explains the wide variety in the identity and powers of national authorities within the meaning of Article 10. The discrepancies and differences are even greater when one examines which of these authorities have the power to apply EC competition rules.9 This variety makes the coordinated enforcement of EC competition law almost impossible at a national level. In some Member States there are no authorities capable of participating in a system of information exchange, whereas in others the procedural rules governing these authorities are clearly opposed to a free flow of information. This finding particularly applies to those countries in which criminal courts apply competition rules.

2.

Exchange of Information

Given its centralising nature, Regulation 17/62 provides only for vertical flows of information between the Commission and the national competition authorities. 6 OJ C 39/6 of 13.2.1993. 7 See, respectively, Articles 10(1), 11, 13 and 14, and Article 10(4) of Regulation 17/62. 8 Case 127/73 BRT/SABAM [1974] ECR 51, and Cases 209–213/84 Nouvelles Frontières

[1986] ECR–1425, para 55. 9 See Jones T. (2001): ‘Regulation 17: the impact of the current application of Articles 81 and 82 by national competition authorities on the European Commission’s proposals for reform’, 22 European Competition Law Review 10, pp. 405–15.

Technical Problem Areas

373

In addition, the national authorities must not use this information in an autonomous manner, but only to assist the Commission by applying Articles 81 and 82 EC, and, in accordance with Community practice, try to influence the Commission’s policy. First, the Commission must keep national authorities informed about its activities. Article 10(1) of Regulation 17/62 provides that it must transmit copies of applications and notifications to the competent authorities, as well as copies of the most important documents. In addition, these authorities must receive copies of requests for information sent to the undertakings located in their Member State pursuant to Article 11. Finally, when the Commission intends to organise inspections within the meaning of Article 14 in a given Member State, it must first consult the competent national authority. Conversely, the Commission receives input from the competition authorities of the Member States. Article 10(2) provides that the competent authorities have the right to express their views about the way that they receive documents. In addition, the Commission must consult the Advisory Committee before making any final decision. Regulation 17/62 also allows the Commission to require national authorities to act in certain circumstances. For example, Article 11(1) provides that the Commission may obtain all necessary information from the governments and competent authorities of the Member States. In accordance with Article 13, the Commission can also request such competent authorities to carry out inspections on its behalf. With the cooperation notice, the Commission tried to broaden the scope of the information exchange in order to promote some form of case allocation. Para 46 of the cooperation notice stipulates that the Commission will supply the relevant documents in its possession to the national authority to which the case is referred. In return, the Commission also expects some additional input from the national authorities. According to para 49 of the notice, it is desirable that these authorities inform the Commission of any proceedings they initiate. The Commission will ensure that this information is passed on to the authorities of the other Member States. This dispatching of information is the first step towards horizontal information flows between these authorities.

3.

Limits

Article 20 of Regulation 17/62 sets limits on the exchange of information as described above. Its first paragraph states that information acquired by the Commission using its fact-finding powers under Articles 11 to 14 must only be used for the purposes of the relevant investigation. This means that information received in the context of one investigation cannot, as a rule, be used in another investigation. If the Commission finds documentary evidence of a cartel infringement involving a subject-matter other than that for which the procedure was initiated, it must not use

374

Working Paper IV

that evidence to establish that infringement.10 Procedures within the Commission have to be kept separate.11 The same applies to parallel procedures pending before both the Commission and national authorities. The information received by national authorities from the Commission pursuant to Regulation 17/62 may, according to the AEB case law,12 be used for the purposes of the procedure for which it was requested. It must not be used as evidence by national authorities to establish the existence of infringements of national or Community competition rules. This safeguard applies not only to information collected pursuant to Articles 11 to 14, but also to other types of information exchanged under Regulation 17/62 — such as copies of notifications and complaints — to which Article 20(1) does not explicitly refer. The second limit to a free exchange of information is contained in Article 20(2) of Regulation 17/62, which prohibits both the Commission and the national authorities from disclosing information of the kind covered by the obligation of professional secrecy. This provision is an expression of the general confidentiality obligation imposed by Article 287 (ex 214) EC. It is not very clear what the concept of ‘information covered by professional secrecy’ covers in the context of Article 20(2) of Regulation 17/62. At any rate, this provision implies that the secrecy obligation relates to a certain type of information, whereas information not belonging to this category can be divulged. In the absence of any guidance from the Court’s case law, it could be argued that the category of confidential information includes the information that the Commission must keep secret when granting access to files in antitrust proceedings. According to the Commission’s notice concerning access to files,13 this information relates to business secrets, the identity of persons wishing to remain anonymous, and internal documents. The confidentiality obligation laid down by Article 20(2) probably provides a degree of protection that goes beyond the standards imposed by that notice. This is because the notice relates to the specific situation in which an undertaking has to defend itself against charges brought against it by the Commission. It therefore strikes a balance between the interests of that undertaking and the undertaking from which the information originates. In the absence of such a conflict of interests, the authorities should abstain from spontaneously disclosing information that could affect the interests of the latter undertaking. In accordance with Article 4(4) of Regulation 1049/2001, national and Community officials must not disclose ‘third party documents’ without first consulting those third parties.14 Moreover, pursuant to Article 4(2) and (3) of the same Regulation, they must refrain from 10 Case 85/87 Dow Benelux v. Commission [1989] ECR 3150. 11 See, mutatis mutandis, Article 17 of Regulation 4064/89. 12 Case C–67/91 Direccíon General de Defensa de la Competencia

v. Asociacíon Española de Banca

Privada (AEB) a.o. [1992] ECR I–4785. 13 Notice on access to files, OJ C 25/3 [1997]. 14 Regulation 1049/2001 regarding public access

documents, OJ L 145/43 [2001].

to European Parliament, Council and Commission

Technical Problem Areas

375

disclosing preparatory documents such as requests for information and statements of objections, because such disclosure could both affect commercial interests and undermine the institutions’ decision-making processes.15 These findings imply that confidentiality standards may differ from case to case. Information that is not confidential in respect of one undertaking, such as a defendant in an antitrust investigation, might be confidential in respect of others, such as complainants.16 The ECJ even went so far as to stress that information received from the Commission by a national authority must not be communicated to other national authorities.17 Two factors undermine the efficiency of the safeguards granted by Article 20. The first is the mental well-being of EC and national officials. According to the Court, one should not expect these persons to suffer from ‘acute amnesia’ when they receive information. Article 20(1) of the Regulation does not allow them to use that information for other procedures, but they are not required to forget what they read. Both the Commission and the national authorities may use the information to justify the initiation of a new procedure.18 In the context of that new procedure, the Commission or the national authority may verify whether this information constitutes evidence for an infringement of Community or national competition rules. The above case-law implies that the Commission can, under certain circumstances, be prevented from transmitting information to national authorities. In the SEP case,19 the Court ruled that the Commission had to bear in mind the fact that Dutch officials from the Ministry of Economic Affairs who received certain business secrets could not be expected to forget this information when entering into commercial negotiations. The secrets in question related to Statoil, a gas supplier, which directly competed with Gasunie, a company jointly controlled by the Dutch State and represented by officials of the same Ministry. The Court held that, before deciding to transmit this confidential information to the national authorities, the Commission should have given the undertaking concerned the opportunity to challenge that decision before the Court of First Instance.20 The second factor affecting the protection granted by Article 20 of Regulation 17/62 relates to the Commission’s duties under Article 10 EC. Pursuant to this provision the Commission is obliged to actively assist national courts in applying EC competition law. This obligation implies that the national courts may request information. In dealing with these requests the Commission is not bound by 15 In

addition, a statement of objections may be seen as a criminal charge. These charges can fall under the category of personal data (Cf Article 2 of Directive 95/46 on the protection of individuals with regard to the processing of personal data and on the free movement of such data, OJ L 281/31 [1995]). 16 Case T–17/93 Matra v. Commission [1994] ECR II–595. 17 See supra note no. 13, at para 41. 18 See respectively, Dow Benelux, paras 18 and 19, and AEB, para. 39. 19 Case T–39/90 Sep v. Commission [1991] ECR II–1497, and Case C–36/92 P [1994] I–1911. 20 Compare Case C 53/85 Akzo [1986] ECR 1965 and Article 9 of Commission Decision 2001/462/EC on the terms of reference of hearing officers in competition proceedings, OJ L 162/21 [2001].

376

Working Paper IV

Article 20 of Regulation 17/62, because this provision does not cover relations with national courts. The Court of First Instance therefore considered that the Commission was entitled to authorise a complainant to use a copy of a statement of objections in national court proceedings. However, before granting such an authorisation, the Commission should give the undertaking concerned an opportunity to express its views about whether or not the document contains business secrets and, if need be, to challenge the Commission’s authorisation decision under Article 230 EC. 21

4.

Preliminary Conclusions

Regulation 17/62 provides for a centralised system for the enforcement of EC competition rules. It provides for vertical flows of information only. Article 20(2) even prohibits a national authority from passing information to another national authority. Information can be exchanged freely within vertical relations, except in situations in which such an exchange would place national authorities in a conflict of interest. In that case, this information cannot be freely used. Direct transfers of evidence from one procedure to another are not possible. However, information exchanged in the context of one procedure may be used to initiate another. Finally, Article 20(2) does not preclude the exchange of confidential information between the Commission and a national authority. Both are bound to the same confidentiality obligation. The scope of this duty may differ from one situation to another; information that may be disclosed to a defendant in antitrust proceedings may not necessarily be revealed to other undertakings. In any event, before actually transmitting information that could contain business secrets, the Commission should give the undertaking concerned an opportunity to challenge its decision.

III.

Information Exchange under the Proposed Rules

In order to achieve its objective of using both national and Community enforcement resources optimally, the draft Regulation intends to set up a network of competition authorities to cooperate closely in the application of Articles 81 and 82 EC.22 The system focuses on the decentralised application of these rules. Unlike Regulation 17/62, the draft Regulation addresses the relationship between

21 Case T–353/94 Postbank v. 22 See p. 6 of the Explanatory

Commission [1996] ECR II–926. Memorandum.

Technical Problem Areas

377

national and Community competition rules: where trade between the Member States is affected, national competition rules will cease to apply.

1.

What is the Network?

The first digression from the current enforcement system is the obligation of Member States to appoint national authorities. Article 36 of the draft Regulation obliges them to grant those authorities the power to apply Articles 81 and 82 EC in full. This obligation means that the members of the network will be clearly defined and that each member will have equivalent powers under EC law. It is unclear what levels of discretion Member States will have in meeting their obligations under Article 36. According to inside information, the authorities should preferably be those dealing with competition issues in the public interest. It excludes authorities that intervene in matters between private persons. Another condition would be that these authorities intervene in first instance hearings. Their decisions should be subject to appeal or judicial review before another independent body. Finally, the same authorities should be represented on the advisory committee. It is questionable whether these requirements can be imposed on Member States by virtue of their duties under Articles 10 and 249 EC.23 Insofar as information exchange is concerned, it is submitted that these provisions do not allow Member States to appoint authorities that are likely to get entangled in conflicts of interests like that in SEP.24 Officials of national authorities should be separated from other parts of the national administration. It seems most likely that Member States will appoint authorities that will retain their powers to apply national competition rules in situations in which interstate trade is not affected. This situation is not ideal, as will be explained below.

2.

Rules on Info-Exchange within the Network

The Commission will continue to assume a leading role within the network; it will remain ‘the spider in the web’. Vertical flows of information will continue to play a role. Remarkably, the Commission will have fewer obligations to provide information to the national authorities than it has under the present system. It will have to continue to send copies of the most important documents to the national authorities

23 Pursuant to their obligations under the ECHR and Article 6 of the EU Treaty, Member States will have to ensure the rights to fair trial. See infra under Section III.3. 24 See supra note no. 20.

378

Working Paper IV

(Article 11(2) of the draft Regulation), and consult them if it intends to organise inspections within the meaning of Article 20. But it will no longer be obliged to send a copy of each request for information to the authority of the Member State in which the addressee of that request is located. The reverse is not true. National authorities will have to supply more information under the new rules. Pursuant to Article 11(3) and (4) they will have to inform the Commission of their intention to act. Before making a decision, they must send a summary of the case as well as copies of the most important documents. Article 11(4) establishes that the Commission may also request a copy of any further document relating to that case. These new powers complement the existing rules that entitle the Commission to request all necessary information from the Government and national authorities (Article 18). Nor will the Commission’s right to request national inspections on its behalf be altered (Article 21(2)). The major novelty of the draft Regulation concerns the horizontal exchange of information envisaged by Article 12 (cited above). According to the Explanatory Memorandum, Article 12 (1) creates a legal basis for the exchange of any information within the network and for its use as evidence in proceedings pending before the Commission or national authorities. It purports to facilitate an optimal case allocation within the network. Article 12(1) is a very broad provision covering any type of information, including confidential documents. Entire case files, which may contain manuscripts and internal notes, may be exchanged. This means that a Commission file may physically be transferred to a national authority. Also, it sets aside any provision of national law that opposes such an exchange. However, one might wonder whether this conflict rule will always prevail. To a certain extent, this issue will depend on Member States’ policies in applying Article 36. If the national procedural rules relating to the appointed authorities are criminal in nature, an unrestricted exchange of information might clash with the overriding requirements of the European Convention of Human Rights.25 Finally, Article 12(1) states that the exchanged information can be used directly as evidence. It intends to remove the obstacles raised by Article 20(1) of Regulation 17/62, as interpreted by the Court in AEB.26 However, Article 12(1) establishes a right, and not a duty. This should not be a problem, according to DG Comp Director-General Alexander Schaub:27 although there is no obligation to assist another member of the network, such help should become natural in a network based on partnership and solidarity. Even so, from a formal point of view, national authorities cannot be compelled to pass on information to

25 See the written contribution of Wouter P.J. Wils for this volume. 26 See supra note no. 13. 27 Schaub A. (2001): ‘Continued focus on reform; recent developments

in EC competition policy’, paper presented at the 28th Annual Conference of the Fordham Corporate Law Institute, October 2001.

Technical Problem Areas

379

another national authority. This use of Article 12(1) will remain a matter for discretion. Another novelty can be found in Article 21(1) of the draft Regulation; one national authority may carry out inspections on behalf of another. The information collected during these inspections must be transmitted in accordance with the procedure laid down by Article 12(1). The interplay between Articles 21 and 12 gives rise to several questions. Does it mean that the authority that carried out the investigation may decide not to pass on the information to the requesting authority? Or, does it imply that the first authority foregoes its discretionary power under Article 12(1) once it has decided to assist another national authority under Article 21(1)? It should be noted in this respect that Article 21(2), which deals with national inspections carried out on behalf of the Commission, does not refer to Article 12(1). This could be put forward as an argument in favour of the first position. Finally, page 12 of Explanatory Memorandum mentions that the members of the network should inform each other of all new cases, and should exchange all relevant case-related information. It is unclear whether this information is subject to compulsory communication as provided for in Article 12(1). Such an obligation may possibly flow from Article 16, concerning the need to avoid conflicting decisions. In any event, no explicit provision requires national authorities to inform their sister authorities of their intention to act. It is more likely that such prior consultation will be channelled through the Commission pursuant to Article 11(3) and (4). The Explanatory Memorandum states that the Commission will include the other members of the network in this consultation process.

3.

Limits

Just like Article 20 of Regulation 17/62, the draft Regulation contains two general rules concerning professional secrecy. According to the Explanatory Memorandum, the first rule — contained in Article 27(1) — is supposed to replace Article 20(1) of Regulation 17/62. The text of the new provision is worded differently in two respects. Thus the new text stipulates that the information must be ‘used only for the purpose for which it was acquired’ instead of ‘used only for the purpose of the relevant request for investigation’.28 This textual difference is not sufficiently relevant to justify a different interpretation of the substance. As a rule, information obtained during the course of one investigation cannot be used for another investigation. In other words, the Dow Benelux case-law continues to apply;29 if the Commission collects documents during the course of one investigation, it

28 In para 35 of AEB, the Court refers to the use of information ‘for purposes other than for which it was requested’. 29 See supra note no. 11.

380

Working Paper IV

may not use those documents as evidence in another investigation. It may only use those documents as a reason to start a new investigation in which the same documents can be requested again. By contrast, the second textual difference does matter: Articles 12 and 15 derogate from the general rule laid down by Article 27(1). Insofar as Article 12(1) is concerned, this derogation implies that information exchanged within the network can be used as direct evidence. The AEB case law will cease to apply within the network. However, it should be noted that Article 12(1) has its own limits. These are contained in the two sentences of Article 12(2). Pursuant to the first sentence, the information can only be used for the purposes of applying Community competition law. It may not be used for any other purposes. It must therefore be assumed that the information cannot be used to justify the initiation of proceedings under national competition law. Here again, the AEB case law will cease to apply. Even so, the limits imposed by Article 12(2) are not entirely clear. Community competition law relates not only to Articles 81 and 82 EC, but also to merger control, state interference and Articles 87 and 88 EC. These provisions of Community competition law can be directly relevant for national competition authorities; for example, information obtained from the Commission or a national authority could be used by another national authority when assessing mergers after a referral within the meaning of Article 9 of Regulation 4064/89. Similarly, such information may be used to examine financial relations between the State and public undertakings within the meaning of the transparency directive.30 It is also unclear how Article 12(2) relates to Article 36. If, pursuant to Article 36, Member States appoint authorities that are also entrusted with the application of national competition rules, the exchange of information under Article 12(1) will create informed national officials who might be tempted to use this information for initiating national competition proceedings. There is no reason to assume that the entry into force of the new Regulation will cause national officials to acquire ‘acute amnesia’. Does this mean that Article 12(2) prohibits the sending of information to national authorities that are also competent to apply national competition rules? The second sentence of Article 12(2) states that the information exchanged under Article 12(1) can be used only as a basis for financial penalties. The information may not be used for obtaining sanctions like imprisonment. The objective of this rule is, according to the Explanatory Memorandum, to ensure an appropriate balance between the right to a defence that the undertakings concerned could invoke in the Member State in which the information originates, and the penalties they may face in the prosecuting Member State. As will be discussed below, this 30 Under

Dutch competition law the agency is authorised to examine whether the Community rules on financial transparency are respected; Commission Directive 80/723 on the transparency of financial relations between Member States and public undertakings, OJ L 195/35 [1980], amended by Directive 93/84, OJ L 254/1 [1993] and Directive 2000/52, OJ L 193/78 [2000].

Technical Problem Areas

381

safeguard is probably not adequate to protect these rights. At any rate, the fact that Article 12(2) implicitly envisages criminal sanctions confirms that antitrust proceedings may be governed by criminal law in some Member States. Finally, the second rule under Article 27 replaces the old Article 20(2) of Regulation 17/62. The Explanatory Memorandum mentions that this provision, which does not add anything to the existing protection, complements the safeguards laid down in Article 12(2). This remark is puzzling, as Article 12 as a whole grants a lower degree of protection to undertakings than the existing rules. It is not an additional protection, but a restriction.

4.

Preliminary Conclusions

Under the new rules, the network in which the information will circulate will be better defined. In addition, it will be possible to exchange information horizontally; that is, between national authorities. Unlike the vertical exchange of information with the Commission, this horizontal exchange will be optional. National authorities cannot be required to transmit information. In deciding whether or not to exercise this right, national authorities may, as a rule, not invoke national rules opposing such an exchange. The other major innovation concerns the use of the information. Pursuant to Article 12 of the draft Regulation, the information received can be used directly as evidence to establish the existence of an infringement of Community competition rules; the AEB case law is put aside. By contrast, the Dow Benelux rule continues to apply between the Commission’s services; information collected in the context of one investigation can be used only to initiate a new investigation. Article 12 also repeals the AEB case law in another important respect: information exchanged on the basis of this provision must only be used for the application of Community competition law. It must not be used for initiating procedures under national competition law. This finding might complicate the exchange of information as authorities become competent to apply both Community and national competition law. This is because such authorities are unlikely to forget the information received under Article 12 when exercising their national tasks.

IV. 1.

Article 12 of the Draft Regulation Applied Is Article 12 Valid?

In AEB, the Court interpreted, inter alia, Article 20(1) of Regulation 17/62: information supplied by the EC Commission to a national authority cannot be used in

382

Working Paper IV

national procedures. As with any rule of secondary law, the Council is free to modify that provision or to derogate from it by inserting a new provision, such as Article 12 of the draft Regulation. The Court explicitly foresaw that possibility in para 37 of the AEB judgment. However, the ECJ ruling relates not only to Article 20 of Regulation 17/62, but also to general principles of Community law, and particularly to the rights of the defendant.31 The Court’s reasoning is, in fact, oriented along two lines: professional secrecy and the rights of the defendant. The latter imply that undertakings must know the purpose for which the information is requested.32 The twopronged approach in the Court’s reasoning is confirmed by the fact that it also relies on the rights of defendant to limit the use of information, to which the professional secrecy provision contained in Article 20(1) of Regulation 17/62 does not explicitly refer.33 In addition, primary Community law applies even where professional secrecy is concerned. Indeed, the Court not only refers to Article 20 of Regulation 17/62, but also to Article 214 EEC (now 287 EC).34 The Council cannot derogate from primary EC law when adopting new legislation. This raises the question of whether the Council may depart from the Court’s findings in AEB? The rights of the defendant will also have to be safeguarded under the new enforcement system. Although the Council has the power to organise the exercise of these rights at the Community level, it is not competent to do so at the national level. The Court’s finding in AE — that Regulation 17/62 does not govern national procedures — remains valid for the new Regulation.35 The new Regulation will not determine how national authorities have to apply Articles 81 and 82 EC. That remains a question of national procedural law. Hence, no Community guarantee exists to safeguard the defendant’s rights at the national level. This raises doubts as to the legality of Article 12 of the draft Regulation. The CFI ruling in Postbank reinforces these concerns.36 In that case, the CFI established that Regulation 17/62 could not prevent the transfer of information to national courts, because the regulation does not concern relations with such jurisdictions. However, it noted that, when relying upon information received from the Commission, national courts have to protect the rights to mount a defence of the parties concerned.37 Moreover, the Postbank case law has to be seen in the context of the Commission’s duties under Article 10 EC. It is hard to see how that provision could be interpreted as authorising a free flow of information without proper guarantees of the defendants’ rights. 31 See 32 See

paras 30, 35 and 36, 50 of AEB. paras 36 of AEB; compare with Joined Cases 46/87 and 227/88 Hoechst [1989] ECR 2859 and Dow Benelux, para 18. 33 See paras 49 to 50 of AEB. 34 See paras 41 and 55 of AEB. 35 See paras 31 and 32 of AEB. 36 See supra note no. 22. 37 See supra note no. 22, para 72.

Technical Problem Areas

383

If its application does not take proper account of the rights of the defence of the undertakings concerned, it must therefore be concluded that Article 12 of the draft Regulation is contrary to the EC Treaty. These rights require that adequate measures be adopted to deal with the issues identified below.

2. 2.1.

Issues and Rights The Right to the Highest Standard of Confidentiality

The draft regulation does not harmonise national procedural rules. Each authority will act according to its own rules, which may differ in at least two respects. First, differences relate to fact-finding powers. For example, the authority is not required to disclose the subject matter of the investigation under Dutch competition law.38 This could give rise to ‘fishing expeditions’, which are considered to infringe on the defendant’s rights under Community law.39 Similarly, oral statements may be used as evidence to establish infringements of competition rules in some jurisdictions (like Belgium) but not in others (including the Community) under the current enforcement system. Finally, the exchange of information can differ within national administrations. Some national authorities may receive confidential information from other parts of the national administration, such as sector-specific regulators and tax authorities, or from third states. Secondly, national rules concerning disclosure may differ. Information that cannot be disclosed in one Member State is not necessarily confidential in another. It might also occur that certain information benefits from legal professional privilege in one jurisdiction, but not in another (for example, legal advice given by an in-house counsel). If the application of Article 12 does not take account of these two differences, a free exchange of information will lead to a gradual erosion of procedural safeguards. National authorities that cannot collect certain information by their own means might be tempted to request authorities with stronger powers to collect that information on their behalf. Similarly, the transfer of information from a jurisdiction with high confidentiality standards to a jurisdiction with lesser standards implies a loss of protection for the undertaking from which the information originates. It is submitted that Article 12 should not be used as a tool to circumvent procedural safeguards. As a rule, the highest standard should prevail.40 First, the 38 Slotboom

M. (1997): ‘De NMa staat voor de deur; toezicht en onderzoek in het kader van de Mededingingswet’, TVVS, p. 322; van der Woude, M. (2000): ‘De eerste contouren van de procedure voor de NMa’, in Slot P. J. and Swaak C. R., eds., De Nederlandse Mededingingswet in perspectief, Deventer, Kluwer, at p. 141. 39 Article 14(2) and (3) of Regulation 17/62 establishes that the Commission should define the subject-matter of its investigation; see also Joined Cases 46/87 and 227/88 Hoechst [1989] ECR 2859. 40 In any event, such an exchange may never be contrary to the rights granted by the ECHR.

384

Working Paper IV

receiving jurisdiction must not establish the existence of an infringement of EC competition law by relying on information that it could not have acquired by its own means or that it would not be entitled to use under its own procedural rules. Secondly, the receiving jurisdiction shall not disclose information that has the benefit of confidential treatment in the originating jurisdiction. A practical example might be helpful in order to illustrate the practical functioning of this principle. If a document is subject to legal professional privilege in one jurisdiction, the national authority of that jurisdiction must not use that document, whether under its own procedural rules or under Article 12. Conversely, if a document does not have legal professional privilege in the jurisdiction of origin, and is sent pursuant to Article 12 to a jurisdiction in which such protection does exist, the receiving authority is barred from using that document for the purposes of applying Community competition law.

2.2.

The Right to the Most Favourable Conditions

The application of the principle proposed above might be complicated even further if the information in question was not actively requested by a competition authority, but was supplied to that authority at the initiative of the undertaking itself. The undertaking could have made this supply subject to certain conditions. Several types of conditions can be imagined. First, the undertaking giving the information could have requested that its identity should not be disclosed. This condition does not in itself oppose transmission of the information under Article 12. The authority sending the information would simply have to make sure that the information is made anonymous by the removal of identifying details, and the undertaking has a legitimate interest in verifying that this condition is met. Secondly, an undertaking could have supplied information subject to an explicit condition that it must not be passed on to other competition authorities. This condition directly conflicts with the objectives underlying Article 12. In addition, it is to hard to see what legitimate interest an undertaking would be trying to protect in imposing such a condition. It is therefore submitted that national authorities should not accept such conditions and that, even if they do, such conditions should not prevent them from relying on Article 12. Thirdly, information may be supplied under a leniency programme. Here again, rules may differ from one jurisdiction to another. Not all member States have leniency programmes, and conditions and incentives may differ in those that do. By analogy with the general principle explained above, leniency programmes should not be an obstacle to information exchange, but the application of Article 12 should not imply a loss of advantages. The most generous leniency programme should prevail. This could either be the programme in the jurisdiction of origin or the programme in the receiving jurisdiction.

Technical Problem Areas

2.3.

385

The Right to Complete Transfer

The existence of a cartel infringement can be established by relying on circumstantial or indirect evidence. The Court accepts this type of evidence; the existence of an infringement can be inferred from market circumstances.41 However, if the accused undertaking is in a position to put forward a plausible explanation for the conduct in question, the Commission must prove that this explanation is not valid. In such situations, access to all relevant documents may be of crucial importance to the defence of the undertakings concerned.42 A document might appear to a competition authority to be irrelevant, but it might contain elements that an undertaking wishes to present as proof of its alternative explanation for the conduct under investigation. Such a document is therefore highly relevant for its defence. It is submitted that Article 12 must not undermine the rights of defendants in such situations. If documents are transferred from one authority to another, the accused undertakings should be entitled to a complete transfer. Exculpatory evidence should not be lost as a result of a free flow of information.

2.4.

Enforcement of these Rights

It follows that the free flow and use of information within the network, as provided for by Article 12, can be reconciled with the requirements of primary Community law if the rights of the undertakings concerned are respected. These rights concern both the undertaking from which the information originates (for example, the right to remain anonymous, other confidentiality rights, the right to leniency, legal professional privilege) and the rights of the undertaking against which the exchanged information is going to be used (for example, the right to complete transfer, the right to leniency). These undertakings are not necessarily the same entity; for example, confidential information from company X established in Member State A can be used in cartel proceedings initiated against company Y in Member State B. Company X might require guarantees that its documents will not unduly be disclosed in Member State B, whereas Company Y might want to have the certainty that Member State A transfers not only incriminating, but also exculpatory documents. This example shows that the rights mentioned above should be protected in both Member States. The authority in the country of origin should safeguard the confidentiality rights of the undertaking where the documents were found, whereas the authority in the country of destination should protect the accused undertaking’s right to mount a defence. With regard to the protection of the rights of the defendant, Member States are required to respect the minimum standards set by the ECHR as reflected 41 C–89/85 Woodpulp [1993] 42 Case T–25/95 Cimenteries

ECR I–1307. CBR [2000] ECR II–491.

386

Working Paper IV

in Article 6 of the EU Treaty. Indeed, when the Commission and national authorities apply Community competition rules, they operate within the Community legal order that demands that these fundamental rights be guaranteed. Relying on these principles, it can be assumed that the accused undertaking will be able to invoke its rights in the Member State of destination during the course of the antitrust proceedings that are brought against it. In that Member State, the competition authority and the controlling judicial bodies should safeguard the defendant’s rights to the highest standards, and should honour the defendant’s right to receive the most favourable conditions available. These bodies should also ensure that the complete file is transferred pursuant to Article 12 of the draft Regulation. A question arises as to whether judicial protection in the country of destination will always be sufficient to protect all the rights identified above. Indeed, the principles flowing from Article 6 EC not necessarily relate to confidentiality issues. Will an undertaking enjoy judicial protection in the country of destination if the competition authority in that country discloses information in accordance with its own rules concerning access to publicly held documents? And even if it does, will the judicial bodies in that country accept the argument that the competition authority should not have disclosed a document because it benefits from confidential treatment in the country of origin? If it is not the object of these proceedings itself, can one expect the undertaking from which the information originates to verify the conduct of the prosecuting national authority? Would it not be more equitable and practical to solve any issue concerning diverging confidentiality standards before the information is actually passed on to another authority? Similarly, one might wonder whether adequate judicial protection can be granted in the country of destination in order to ensure compliance with all the conditions of Article 12(2) of the draft Regulation, and particularly with the condition that the information will only be used for the purposes of Community competition law. In situations in which the authority in the country of destination is also entrusted with the application of national competition law, it might be tempted to use the transferred information to initiate national proceedings. Should undertakings have a right to oppose transfer in order to avoid such situations? In the light of these uncertainties, it is submitted that the competition authority in the jurisdiction of origin should hear undertakings before it decides to transfer information on the basis of Article 12 in the draft Regulation. Undertakings should have a right to inform that authority of the problems they might encounter in the country of destination if the information is transferred unconditionally. In addition, they should have a right to have a transfer decision reviewed by an independent judge before this decision is put into effect. In other words, competition authorities should apply by analogy the same procedure that the Commission has to follow when it intends to transfer information to national courts.43 There is no

43 See

supra note no. 22.

Technical Problem Areas

387

reason why the protection granted by the Commission in its relations with national courts should not apply in relations within the network of competition authorities. Both scenarios concern a proper application of EC competition law. In addition, it is a fundamental principle of Community law that individuals are entitled to be heard before the administration makes a decision that may adversely affect their interests, and that they may challenge this decision before an independent judge before the decision is enforced.44

V.

Conclusion

The major difference between the existing and proposed rules concerning information exchange is the new provision for horizontal cooperation between national authorities. Article 12 of the draft Regulation creates a directly applicable right that sets aside any national rule that would oppose such an exchange of information. Even so, Article 12 creates a right, but not a duty, to transmit information. The success of the information exchange process, and therefore the success of an optimal system of case allocation, will largely depend on the political will of these authorities and on the Commission’s coordination efforts. Article 12 sets aside the AEB case law in two respects. First, it enables the members of the network of competition authorities to use the exchanged information as evidence in their own proceedings. This information may be confidential in nature. Complete files can be exchanged. Secondly, the proposed rule imposes a limit that does not exist under the present system. The received information can only be used for the purposes of applying EC competition law. This means that the information cannot be used to initiate proceedings under national competition law. It can be argued that Article 12 alone conflicts with primary Community law, and particularly with the fundamental rights of the defendant and Article 287 EC, as interpreted by the Court in AEB. However, possibilities exist to reconcile the need for a proper functioning of the network of competition authorities with these fundamental rights. When information is exchanged, undertakings should always be entitled to the highest level of protection in the jurisdictions concerned. This level of protection implies that the undertakings should benefit from the rules offering the strongest safeguards for their rights, with regard to both the confidentiality of their documents and the fact-finding powers of the authorities in question. Article 12 must not lead to an erosion of standards. Nor should it enable an authority to acquire evidence that it would not have been able to collect by its own means. As all the jurisdictions concerned have to comply with the minimum standards deriving from Article 6 EC, most of these additional safeguards can be offered in the jurisdiction of destination. Any violation of these safeguards should imply that 44 Case

C–222/86 Heylens [1987] ECR 4097.

388

Working Paper IV

the exchanged information could not be used as evidence against the undertakings concerned. However, the judicial bodies of the country of destination are not always in a position to offer adequate protection. Difficulties might arise in situations in which the country of destination applies confidentiality standards that are less stringent than those in the country of destination. Difficulties might also occur if information is sent to national authorities that are tempted to use it to initiate procedures under national competition law. In order to safeguard these rights, the dispatching authority should hear undertakings before it decides to apply Article 12. In addition, the undertakings should have an opportunity to appeal against such decisions before they are enforced. In any event, Article 12 must not lead to a situation whereby the rights of undertakings are less well protected at the national level than they currently are at the Community level.

V Margaret Bloom1 Exchange of Confidential Information among Members of the EU Network of Competition Authorities: Possible Consequences of a Relatively Broad Scope for Exchange of Confidential Information on National Procedural Law and Antitrust Sanctions

This paper first considers briefly the need for a broad scope of exchange of confidential information between competition authorities generally. It then outlines the proposals for exchange of confidential information in the draft Regulation to implement Articles 81 and 82 EC and the safeguards to be applied to such exchange. It summarises current restrictions in EC and national laws on exchange of information and some of the differences in national procedures and sanctions. This sets the stage for a consideration of the possible consequences for national procedural law and antitrust sanctions of a broad scope for exchange of confidential information. Finally, the paper concludes that changes in national procedural laws and sanctions alone may well not be sufficient for fully effective enforcement by the EU network. A further reform of EC law covering procedures and sanctions may, in due course, be desirable.

I. The Need for a Broad Scope of Exchange of Confidential Information The importance of information exchange between competition authorities for effective antitrust enforcement should not be underestimated. This applies not just to the EU network of competition authorities, but more widely to authorities throughout the world. Gathering information can be a particular problem for competition authorities: as business becomes increasingly global, the likelihood of a national authority requiring information which is located outside its jurisdiction increases correspondingly.

1 Director of Competition Enforcement, OFT, UK. The views expressed are personal. I am grateful to my colleague Frances Barr for her comments, which are also personal, in preparing this paper.

390

Working Paper V

A problem is that jurisdictional rules developed in the nineteenth century are not particularly well-suited to the business context or the information technology of the twenty-first century.2

The need to exchange information is particularly important for dealing with international cartels. Cartellists can share whatever information they like; but competition authorities are constrained by numerous rules. However, some of those rules are in the course of being relaxed and cooperation between authorities is growing both quantitatively and qualitatively. This augurs well for increasingly effective enforcement. The improved cooperation with foreign law enforcement authorities has provided us with increased access to foreign-located evidence and witnesses that has proven to be instrumental in the cracking of a number of international cartels. While I am constrained as to what I can reveal about the nature of this assistance, I will offer one compelling statistic to demonstrate the breadth of this cooperation. In the last year alone, foreign authorities from five different countries have executed search warrants at our request in more than a half-dozen of our international cartel investigations. For anyone who is familiar with the history of foreign assistance in antitrust matters, this is a remarkable advancement in international cooperation.3

A broad scope of information exchange is necessary for effective working of the EU network of competition authorities. A key element of the functioning of the network is the possibility for the Commission and the national authorities to exchange and use information and documents that have been collected by any one of them. This power is a precondition for effective antitrust enforcement including efficient allocation of cases and consultation on them.

II.

What is Covered by ‘Exchange of Information’?

Both Articles 12 and 21 of the draft Regulation cover aspects of exchange of information. The current text of Article 12 reads as follows: Exchange of information 1. For the purpose of applying Articles 81 and 82 of the Treaty, the Commission and the competition authorities of the Member States shall have the power to provide one another with and use in evidence any matter of fact or of law, including 2 Richard Whish (2001): Competition Law, 4th edition, London, Butterworths. 3 Scott D. Hammond, Director of Criminal Enforcement, Antitrust Division,

US Department of Justice: ‘When calculating the costs and benefits of applying for corporate amnesty, how do you put a price tag on an individual’s freedom?’, speech delivered at the US National Institute on White Collar Crime, 8 March 2001.

Technical Problem Areas

2.

391

confidential information. Where national competition law is applied in parallel with Community competition law pursuant to Article 3, information exchanged under this Article may also be used for the application of national competition law. No sanction of imprisonment can be imposed on the basis of information provided under paragraph 1, except if the transmitting Member State’s law foresees such sanctions for the same kind of infringement.4

As the Explanatory Memorandum to the draft Regulation explained: Paragraph 1 creates a legal basis for the exchange of any information between the Commission and the Member States’ competition authorities and its use as evidence in proceedings applying Community competition law. It covers the transfer of information from the Commission to a Member State’s authority and vice versa, and the transfer from one Member State authority to another. This provision also allows the transfer of entire case files, including confidential documents, the objective being to render possible the transfer of a case from one authority to another in the interest of effective case allocation.5

Article 12 allows for free movement of evidence within the European Union for the purposes of the application of the EC competition rules. The current text of Article 21 reads as follows: Investigations by competition authorities of Member States 1. The competition authority of a Member State may in its own territory carry out any fact-finding measure under its national law on behalf and for the account of the competition authority of another Member State in order to establish whether there has been an infringement of Article 81 or Article 82 of the Treaty. It shall transmit the information collected to the requesting authority in accordance with Article 12 of this Regulation. 2. At the request of the Commission, the competition authorities of the Member States shall undertake the inspections that the Commission considers to be necessary under Article 20(1) or which it has ordered by decision pursuant to Article 20(4). The officials of the competition authorities of the Member States who are responsible for conducting these inspections shall exercise their powers in accordance with their national law. If so requested by the Commission or by the competition authority of the Member State in whose territory the inspection is to be conducted, the officials of the Commission may assist the officials of the authority concerned.6

4 Progress Report to Council of the European Union Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty and amending Regulation (EEC) No 1017/68, (EEC) No 2988/74, (EEC) No 4056/86 and (EEC) No 3975/87 SN4467/01, 7 November 2001. 5 European Commission (2000): Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, COM(2000)582 of 27.09.2000. 6 See supra note no. 4.

392

Working Paper V

As the Explanatory Memorandum to the draft Regulation explained: This Article is based on Article 13 of the existing Regulation 17, adapted to the new implementing system. While continuing to allow Member States to conduct investigations on their territory on behalf of the Commission, it also enables them to carry out fact-finding measures on behalf of a competition authority of another Member State. This provision is necessary in order to allow effective cooperation between the competition authorities of the Member States. Such cooperation enables national competition authorities to deal with cases where some evidence is to be found in other Member States. Without such mechanisms a real decentralisation of the application of Community competition rules would be seriously hampered.7

III.

Safeguard Measures

What safeguards will be applied to the exchange of confidential information under the new Regulation? All exchanges are to be subject to safeguards concerning: — — —

the protection of confidentiality the use of the information; and the imposition of sanctions.

These are considered briefly below.

1.

Protection of Confidentiality

Undertakings will have a right to protection of their business secrets and of confidential information they submit to any member of the network. The draft Regulation guarantees this protection in Article 27. This states that the Commission and the national authorities, their officials and other staff shall not disclose information acquired or exchanged between them pursuant to the Regulation which is of the kind covered by the ‘obligation of professional secrecy’. This obligation is set out in Article 287 EC that refers to such information as ‘in particular information about undertakings, their business relations or their cost components’. The concept of ‘professional secrecy’, including the scope of business secrets and confidential information is a Community concept which has been developed in case-law of the European courts. Exceptions to this protection would be if disclosure were required to prove the infringement and/or to observe the rights of defence of the parties involved 7 See

supra note no. 5.

Technical Problem Areas

393

in the procedure. Even here, for example, the right of access to file is subject to the protection of business secrets (Article 26(2) of the draft Regulation). The obligation of professional secrecy is also without prejudice to various activities that are central to the efficient operation of the network. These include the exchange of information between the national competition authorities and the Commission (Article 12), in relation to the operation of the Advisory Committee (Article 14) and cooperation with the national courts (Article 15).

2.

Use of the Information

The second guarantee to be given to undertakings relates to the use of information and documents exchanged within the network. Under Article 12 of the draft Regulation, information so exchanged can only be used for the application of Community competition law or for the application of national competition law where it is applied in parallel with Community competition law for the same infringement. No use for other purposes is possible.

3.

Imposition of Sanctions

The penalties that can be imposed on the basis of information exchanged are limited to pecuniary sanctions — i.e. fines — and those such as disqualification of company directors. This is unless the transmitting Member State permits imprisonment for the same kind of infringement that is involved in the receiving Member State. It remains to be seen precisely what is covered by ‘the same kind of infringement’

IV. Current Restrictions on Exchange of Confidential Information There are currently two obstacles to information exchange between EU competition authorities concerning: — —

the flow of information from the Commission to national authorities; and the flow of information from national authorities to the Commission or to other national authorities.

These will be overcome by the new Regulation as long as the national authorities are applying Community law or national law in parallel with Community law for

394

Working Paper V

the same infringement. The Treaty does not provide a basis for establishing a system of exchange of confidential information solely for the application of national law.

1.

Flow of Information from the Commission to National Authorities

In the Spanish Banks judgment,8 the European Court of Justice interpreted Article 20 of Regulation 17/62 in such a way that a Member State may not use as direct evidence unpublished information, which the Commission has obtained under its investigative powers pursuant to Regulation 17/62, to apply national competition law or even Community competition law. National authorities are not precluded, however, from using this information to justify the initiation of a national investigation.

2.

Flow of Information from National Authorities to the Commission or to other National Authorities

Most national laws currently prohibit the transmission of confidential information to foreign authorities. However, a few Member States already have powers that permit the transmission of information to ‘foreign authorities’ or to specified authorities, e.g. the Netherlands and Denmark (as described below) and the April 2001 Nordic cooperation agreement between Denmark, Iceland and Norway.9 The Dutch Competition Act of 1997 provides in Article 91 that … the director general is authorised to provide information or data obtained in the performance of the tasks assigned to him by this Act, to: 1. a foreign institution which is responsible, pursuant to national statutory provisions, for the application of competition rules, in so far as such information or data are, or could be of significance for the performance of the tasks of that institution and, in the director general’s opinion, provision of that information or data is in the interests of the Dutch economy, provided that: 2…….

a. the confidentiality of the information or data is sufficiently protected and b. there is sufficient assurance that the information or data will not be used for any purpose other than that for which they are provided.10

8 Case C–67/91 Associacion Espanola 9 The agreement between Denmark,

de Banca Privada [1991] ECR I–4785. Iceland and Norway on cooperation in competition cases permits exchange of confidential information between these three countries. See http://www.konkurransetilsynet.no/velkommen.html. Sweden hopes to join this agreement soon. 10 See http://www.nma-org.nl/english for an English version of the Dutch Competition Act 1997.

Technical Problem Areas

395

The Danish Consolidated Competition Act of 2000 provides in Section 18a: Subject to reciprocity, the Competition Authority may divulge information covered by the rules of professional secrecy to other countries’ competition authorities, if such information is necessary to improve the enforcement of the competition rules of those countries, including the fulfilment of any bilateral or multilateral agreement entered into by the Kingdom of Denmark. When divulging information pursuant to subsection (1), the Competition Authority must make the divulgement subject to the following conditions: i. The addressee must have a similar obligation of professional secrecy ii. The information divulged may exclusively be used for the purposes set forth in a bilateral or multilateral agreement, if the divulgement takes place according to such agreement, or iii. The information divulged may only be passed on with the express consent of the Competition Authority and only for the purpose covered by the consent. The Minister for Trade and Industry may lay down further rules on the divulgement of information to foreign authorities covered by the Competition Authority’s professional secrecy.11

Other Member States are making changes or proposing to allow exchange of confidential information with foreign authorities. For example, the French New Economic Regulation Act of 15 May 2001 provides that it will be possible to communicate information and documents held or received by the Conseil de la Concurrence to the European Commission and to foreign competition authorities. The Conseil may also conduct — or ask the Minister in charge of the Economy to conduct — inquiries at the request of foreign competition authorities. The UK is legislating in the Enterprise Bill to enable information — other than that obtained in merger or market investigations — to be disclosed to overseas competition authorities for civil proceedings and also to any overseas authorities for criminal proceedings.12 The Bill was introduced to the United Kingdom Parliament in March 2002. Suitable safeguards will, of course, apply to disclosure made for both civil and criminal proceedings. Also, in Spring 2001 the UK and US Governments deleted the provisions excluding antitrust cases from the UK/US Mutual Legal Assistance Treaty. This enables the UK to provide assistance to the US authorities in the use of 11 See http://www.ks.dk for an English version of the Danish Consolidated Competition Act 2000. 12 See Part 9 of the Bill http://www.publications.parliament.uk/pa/pabills.htm. For the policy back-

ground, see paragraphs 4.44–4.45 of A World Class Competition Regime, DTI, CM5233, The Stationery Office, July 2001; see http://dtiinfo1.dti.gov.uk/cp/news.htm. Paragraphs 79–80 of Government’s response to consultation, December 2001; see http://dtiinfo1.dti.gov.uk/enterprisebill/ response.pdf. It is at least arguable that the Competition Act 1998 already contains provisions that would allow for disclosure to other competition authorities following modernisation of EC competition law. Section 55(3)(d) applies to any disclosure ‘which is required to meet a Community obligation.’ This would cover disclosure required in accordance with the new Regulation once this is in operation.

396

Working Paper V

UK courts to take testimony from witnesses, the service of documents and assistance in the imposition of fines. 13 Section 46 of the Irish Competition Act, 2002 enables the Irish Competition Authority, with the consent of the Minister, to agree arrangements for the exchange of information with similar bodies abroad.14

V. Differences in Member State National Procedures and Sanctions for Breaches of National Competition Law With the modernisation of EC competition law, procedures and sanctions will remain national even where the national authorities are applying Articles 81 and 82 EC. There are some significant differences between Member States in their national procedures and sanctions both in law and in practice. A summary of the national competition laws and procedures for the EEA countries, which has been compiled by the national antitrust authorities, can be found on the website of the Italian Competition Authority.15 Some of the differences between national procedures are apparent from this summary. Annex A to this article summarises national sanctions that apply in EU Member States for breaches of substantive prohibitions and for procedural breaches. It shows that all Member States can impose fines on undertakings for breaches of substantive prohibitions — although in the case of Luxembourg fines are only imposed for breach of prohibition decisions. Five Member States can impose fines on individuals for breaches of substantive prohibitions (Austria, France, Greece, Ireland and Spain). In addition, such fines are, at least, theoretically possible in a further three jurisdictions under certain circumstances (Denmark, Germany and the Netherlands). Imprisonment for breach of a substantive prohibition is available in three Member States (Austria, France and Ireland). Also, in Germany, imprisonment can be imposed for collusive tendering. Austria is proposing to repeal its criminal powers, while the UK is legislating to introduce criminal sanctions — imprisonment and/or fines — for individuals convicted of involvement in hard-core cartels.16

13 http://dtiinfo1.dti.gov.uk/cp/cooperation.htm. 14 For text of the Act see http://www.gov.ie/oireachtas/frame.htm. 15 http://www.agcm.it/eng/tema0140.htm. 16 See Part 6 of the Enterprise Bill http://www.publications.parliament.uk/pa/pabills.htm.

For the policy background, see DTI press release, 28 November 2001, and Chapter 7 of A World Class Competition Regime, DTI, CM5233, The Stationery Office, July 2001. See http://dtiinfo1.dti.gov.uk/ cp/news.htm for both references. Paras 51–69 of Government’s response to the consultation, December 2001. See http://dtiinfo1.dti.gov.uk/enterprisebill/response.pdf.

Technical Problem Areas

397

Procedural breaches in Annex A include breaches of restrictions on disclosure of information. For procedural breaches, fines on both undertakings and individuals are generally available. Five Member States also provide for imprisonment for such breaches (Finland, France, Greece, Ireland and the UK). While Annex A summarises the sanctions that can be imposed, it does not indicate what sanctions are applied in practice. In addition, even where Member States can, for example, all apply fines there are differences in the level of these either in law and/or in practice. The above paragraphs have described some of the differences between Member States in sanctions and referred to the fact that procedures also differ. Some further examples concerning differences in powers of investigation are described in the following section. Clearly, some of the differences in procedures and sanctions are significant.

VI. Possible Consequences for National Procedural Law and Antitrust Sanctions of a Relatively Broad Scope for Information Exchange A number of important issues arise, including those in relation to the following issues: — — — — — —

The type of information that may be covered by the network of EU competition authorities. The effect of Article 21 in the draft Regulation. Implications for protection of the rights of defence. Implications for the successful operation of leniency programmes. Is there a real need to exchange confidential information when applying national competition law only? Are significant differences in national procedures and sanctions sustainable in the medium term?

These are considered below.

1.

The Type of Information that may be Covered by the Network of EU Competition Authorities

Are there specific issues that are raised by the concepts of confidentiality, business secrets and professional secrecy? What is the scope of each? Are there,

398

Working Paper V

for example, different concepts and practices in national laws? The UK Competition Act provides certain safeguards against the disclosure of certain types of information. This is more precise than the position established under EC case-law and may differ from the law and practice in other Member States. In addition, what are the implications of the differing approaches to legal professional privilege? In the UK, legal professional privilege for in-house lawyers is recognised in contrast to most other Member States and EC law. However, changes in procedures in Member States may reduce some of these differences over time.

2.

The Effect of Article 21 of the Draft Regulation

A national authority can currently send written requests for information to undertakings located outside its territory but it cannot carry out inspections outside the territory of its own Member State. In most cases, national authorities are able to collect the evidence necessary to prove the infringement they are prosecuting. However, they may need to collect information located outside their territory in order to deal effectively with a case. Under Article 21 one national authority will be able to ask another for assistance in order to collect information on its behalf including carrying out an inspection on its behalf. Article 12 of the draft Regulation empowers the assisting national authority to transmit the information it has collected to the requesting one. Where one national authority acts on behalf of another, it acts pursuant to its own rules of procedure and under its own powers of investigation. Litigation concerning the collection of information is subject to national law. Although members of the network will use their best efforts to assist each other this will only be possible as far as national procedures permit such efforts. The powers of investigation differ in some significant aspects between Member States; for example, the Irish Competition Authority has powers to interview individuals and record their statements. Whereas in the UK, individuals can only be required to provide explanations of documents. As another example, only some of the Member States — including the UK — permit on site investigations of domestic premises (under certain circumstances). Differences in the recognition of professional privilege have already been mentioned above. Hence, Article 21 will only be able to work effectively — thereby ensuring the necessary exchange of information — in so far as this is possible with the differences between national powers of investigation. It is assumed that national authorities will use their powers to assist each other as far as this is legally and practically possible with regard to the resources, workloads and priorities of each authority. This is a sound assumption given the increasingly close working relations between the EU national authorities.

Technical Problem Areas

3.

399

Implications for Protection of the Rights of Defence

Differences in the scope of the various national competition laws may raise particular issues concerning the rights of defence.17 Because the powers of investigation differ to some extent from one Member State to another, it is conceivable that one national authority would be able to collect information that the receiving authority would not itself have been in a position to collect. It could be, for example, that the Commission or a national competition authority transmits information to the Office of Fair Trading that would have been privileged under UK law. This issue raises a question as to whether the receiving authority should be entitled to use such information. In principle, this question is answered by Article 12 itself. This Article allows national authorities to use in evidence information sent by another competition authority. If national law could prevent them from doing so, Article 12 would be deprived of its useful effect. One of the issues that may need to be considered is the rights of defence, particularly if information is used that could not have been obtained by the receiving authority under its national powers. The above paragraph considers one particular issue raised in relation to the rights of the defence. Are there other issues which may be raised where the information has been obtained by another competition authority, for example the process by which it was obtained such as differences in the scope of investigations (referred to earlier) and any rights to make representations? In the UK for example parties are entitled to indicate which parts of information submitted to the Office of Fair Trading are confidential.18 Consideration may need to be given to the treatment of any such information that has been obtained from other Member States. Another fundamental issue will be at what stage a party is informed that information has been obtained from another Member State: will there be an automatic right to be told or might it only emerge, for example, at the time of the statement of objections/access to file? Will the position for third parties be the same as for parties to an investigation? These issues will need to be considered carefully by Member States.

17 This paper does not address the European Convention on Human Rights, as this is the subject of other papers. 18 Rule 30(1)(c) says that ‘information is confidential if it is:

(i) commercial information the disclosure of which would, or might, significantly harm the legitimate business interests of the undertaking to which it relates; or (ii) information relating to the private affairs of an individual the disclosure of which would, or might, significantly harm his interests;’ See http://www.hmso.gov.uk/si/si2000/20000293.htm#27 for Director’s rules. Statutory Instrument 2000 No 293.

400

4.

Working Paper V

Implications for the Successful Operation of Leniency Programmes

Leniency applications are proving very effective at bringing cartels to the attention of the competition authorities that might otherwise go undiscovered and facilitating the punishment of cartels that might otherwise escape sanction for the lack of sufficient evidence. Applicants for leniency will usually require some form of confidential treatment for the information they provide and also for the fact that they have applied for leniency. In the long run, a delay in information reaching the public domain rather than complete secrecy may be the best that undertakings can expect in jurisdictions with a procedure culminating in a written, reasoned decision. Such decisions include explanations of why there has been a full or partial reduction in any financial penalty imposed. When an undertaking decides to apply for leniency it is well advised to do so in all jurisdictions where it may have infringed the law and there is a leniency programme. Where this involves a cartel case that has already been allocated through the EU network of competition authorities the applicant will know which authority or authorities are handling the case — and will apply to them. Information obtained would be handled in the same way as for other cases. Where a potential leniency applicant is considering making an application concerning a cartel that has not yet been uncovered they will weigh a number of factors before deciding which authority or authorities to approach. The cartel may cover two or more Member States where leniency policies differ or do not exist in some cases. If so, the leniency applicant may chose to approach the Commission rather than national authorities seeking the benefit of the EU-wide leniency policy.

5.

Is there a Real Need to Exchange Confidential Information when Applying National Competition Law Only?

Will there be any cases where national competition law only should be applied and there is relevant information that is held by another EU national competition authority? In a good number of cases EC and national law are likely to be applied in parallel to the same infringement. 19 Information obtained under either Article 12 or 21 of the draft Regulation can only be used to apply national law where Community and national law are applied in parallel in the same infringement. If the case was subsequently pursued only under national law because there was no effect on trade between Member States, the information obtained from another Member State could not be used.20 19 Article

3 of the draft Regulation requires that where national authorities apply national law to anticompetitive behaviour prohibited by Articles 81 or 82 EC they shall also apply EC law. 20 Assuming the precedent of Spanish Banks [see supra note no. 8] is applied, the information could be used only to initiate a domestic investigation, but not itself as evidence in pursing that domestic investigation.

Technical Problem Areas

401

If a case involves information in more than one EU jurisdiction it is likely that Community law should be applied — possibly in parallel with national law. However there may well be some cases where it will be valuable to be able to exchange confidential information between EU authorities when applying only national competition law. For example, two British companies might have discussions in the UK to enter the French market and divide it between themselves. Would French law be the appropriate one to apply if there was no effect on trade between Member States? But relevant information could be in the UK.

6.

Are Significant Differences in National Procedures and Sanctions Sustainable in the Medium Term?

What should be done to help resolve differences in national procedures and sanctions? Soft harmonisation will increase, as the network grows stronger to the extent that such harmonisation is possible under national law. A good example of soft harmonisation is the Principles for Leniency Programmes21 agreed by the European Competition Authorities22 at a meeting in September 2001. These Principles cover desirable high level principles for leniency programmes, cooperation between authorities and good administrative practice. The Principles ‘are put forward to advance the debate on the structure of leniency programmes that is taking place within Europe. Recognising the differences in national situations … [they] do not bind the authorities to adopting these principles.’23 However, there are clear limits over the extent to which soft harmonisation can reduce differences between national procedures and sanctions in the absence of changes to national laws. But, given the safeguards described above over the exchange of confidential information, do these differences matter? It is the differences between civil and criminal powers and between sanctions on individuals and undertakings that are most obvious. Although, other differences, such as in the powers of investigation, may be at least as significant in practice. There are, of course, at least as great differences between Member States in areas totally outside competition law. However, modernisation of EC law, the working of the network and wide exchange of information will all increase pressure for harmonisation in antitrust enforcement. The strengthening of the EU as a community and globalisation will both add to this pressure. It is suggested that in the medium term, reform will be necessary to reduce the differences in national 21 See http://www.tca.ie for the Principles for Leniency Programmes. 22 Participation in the activities of the European Competition Authorities

(ECA) is open to authorities entrusted with competition law enforcement within the European Economic Area. See http://www.konkurransetilsynet.no/velkommen.html for the Memorandum of Understanding of the ECA. 23 See supra note no. 21.

402

Working Paper V

procedures and sanctions. This reform could either be through changes to national legislation or through a further reform of EC law. The latter seems the more likely vehicle to achieve the changes necessary.24

24 To

the extent that national procedures serve to apply Articles 81 and 82 EC, Article 83 EC should provide a sufficient legal basis for this reform. To the extent that national procedures serve to apply national competition law, Article 308 EC could serve as a legal basis.

Technical Problem Areas

403

ANNEX A: THE POWERS OF MEMBER STATES TO IMPOSE SANCTIONS ON UNDERTAKINGS AND ON INDIVIDUALS FOR BREACHES OF NATIONAL COMPETITION LAW

P⫽yes

Member State

x⫽no

Fines on undertakings26 B D28 DK29 EL ES F I IRE L30 NL31 O P SUO SVE UK

Procedural breaches25

Breaches of substantive prohibitions

P P P P P P P P X P P P P P P

Fines on individuals27 X x x P P P X P X X P X X X X

Imprisonment (max) X X X X X 4 years X 2 years X X 3 years X X X X

Fines on undertakings P P P P P P P P P P P P P P P

Fines on individuals P P X P P P P P P X7 P P P P P

Imprisonment (max) X X X Unlimited X 6 months X 6 months X X X X 6 months X 2 years

25 Includes breach of restrictions on disclosure of information. 26 For the purpose of this table, the term ‘undertakings’ includes natural persons only where the under-

taking has no legal personality separate from a natural person, i.e. the fine is imposed on the economic actor. the purpose of this table, ‘fines on individuals’ means fines which are imposed on individual natural persons acting on behalf of an undertaking. 28 Exception: fines and sentences of imprisonment up to 5 years can be imposed on individuals for collusive tendering. 29 In theory, fines could be imposed on individuals, but in practice, fines are imposed on undertakings. 30 Luxembourg operates a ‘control of abuse’ system, and therefore there are no sanctions for breach of the substantive prohibitions in themselves: sanctions are only imposed for breach of prohibition decisions or definitive orders. 31 Sanctions are primarily imposed on undertakings, however, the Dutch Competition Act provides for sanctions to be imposed on the ‘natural person to whom, or the legal person to which the infringement can be attributed’. 27 For

VI Santiago Martínez Lage and Helmut Brokelmann1 The Possible Consequences of a Relatively Broad Scope for Exchange of Confidential Information on National Procedural Law and Antitrust Sanctions

I.

Introduction

The Commission’s draft Regulation,2 proposed to replace the current procedural Regulation 17/62, contemplates the creation of a ‘network of public authorities applying the Community competition rules in close cooperation’. When there is an infringement of Articles 81 or 82 EC the competition authorities of the Member States are to take effective action against the conduct in question, acting in accordance with the draft Regulation and applicable national procedural rules.3

Hence, while the network composed of the Commission and national competition authorities (hereafter ‘NCAs’) will enforce a set of common substantive rules (Articles 81 and 82 EC), the draft Regulation does not contemplate a harmonisation or coordination of national procedural rules within this network. The same is true of administrative sanctions for infringing Articles 81 and 82 EC, as the draft Regulation does not foresee their harmonisation,4 although in Article 12(2) it provides that ‘only financial penalties may be imposed on the basis of information provided’ within the network of competition authorities. As the ECJ held in Spanish Banks,5 ‘[e]ven in cases where they apply the substantive provisions of Article 85 (1) and 86 of the Treaty, it is incumbent upon the NCAs to implement them in accordance with national rules.’ Hence, NCAs will apply Articles 81 and 82 EC in the context of the new network primarily on the basis of their respective national procedural laws. The question therefore arises: what will be the consequences of the new Regulation on national procedural rules governing

1 2

Martínez Lage & Asociados, Madrid. Draft Regulation for a Council Regulation implementing Articles 81 and 82 of the Treaty, COM (2000) 582 final, of 27.9.2000. 3 Explanatory note to Article 5 of the draft Regulation. 4 Ibid. 5 Case C–67/91 Dirección General de Defensa de la Competencia v. Asociación Española de Banca Privada [1992] ECR I–4785, para 32.

406

Working Paper VI

investigatory powers, procedural safeguards, access to files, confidentiality of documents, fines, and so on? This article will specifically analyse the consequences on national procedural law of the exchange of confidential information within the EC network of competition authorities. We will also discuss whether the issues that the Commission’s draft Regulation has left unresolved should be addressed by way of further implementing and/or harmonising measures at the Community level, or whether they should instead be dealt with individually in their respective procedural laws of the Member States.

II. The Framework for Exchange of Information in the Draft Regulation The draft Regulation contains various references to the exchange of information — including confidential documents — between the competition authorities that comprise the network. Thus, Article 11(2) — mirroring Article 10(1) of Regulation 17/62 — provides that the Commission shall transmit ‘copies of the most important documents’ to the NCAs in the context of the proceedings it conducts with a view to adopting a decision in application of Articles 81 or 82 EC. Conversely, Article 11(4) provides for the transfer of ‘copies of the most important documents’ from the NCAs to the Commission if the former intend to adopt a decision pursuant to Articles 81 or 82 EC. In respect of information exchanged in this manner, Article 26(2) restricts the right of access to the Commission’s file, excluding access to confidential information (such as correspondence between the Commission and the NCAs, or between the latter) particularly with respect to documents drawn up pursuant to Article 11. However, this provision is not addressed to the NCAs of the Member States, which have the access to their files regulated by national procedural laws. Article 12 of the draft Regulation covers the exchange of information in general terms, making it clear that the Regulation not only covers the (vertical) transfer of information — including confidential documents — to and from the Commission, but also (horizontally) between the different NCAs. When an inspection is carried out by a NCA on behalf of and for the NCA of another Member State, the former ‘shall transmit the information collected to the requesting authority in accordance with Article 12’ (Article 21(1)). Any exchanged information may be used as evidence in Community or national proceedings (Article 12(1)), although Article 27(1) provides that information collected in the context of requests for information or inspections ‘shall be used only for the purpose for which it was acquired’. Although this provision mirrors Article 20(1) of Regulation 17/62, it is ‘without prejudice to Article 12’, the second paragraph of which significantly relaxes this limitation by allowing the use of information exchanged pursuant to Article 12(1) ‘for the purpose of applying Community competition law’. The repercussions of this important relaxation of

Technical Problem Areas

407

the prohibition to use information for purposes other than those for which it was acquired — and also in light of the European Courts’ rulings in cases such as Dow Benelux, Spanish Banks and SEP — will be discussed below. These provisions are complemented by Article 27(2), which imposes an obligation of professional secrecy on the Commission and NCAs in respect of information acquired or exchanged pursuant to the Regulation. This provision extends the current obligation of professional secrecy under Article 20(2) of Regulation 17/62 to all confidential information exchanged by the NCAs. Furthermore, Article 12(2) limits the use of such information to impose financial penalties. The exchange of information is not limited to the network formed by the Commission and NCAs. It also includes the courts of the Member States, which will be empowered to apply Articles 82 EC and 81 EC in their entirety (Article 6 of the draft Regulation). Thus, for the purposes of submitting written or oral observations to national courts that are hearing a case concerning the application of Article 81 or 82 EC, both ‘the Commission and the competition authorities of the Member States may request the national courts to transmit to them any documents necessary’ (Article 15(3)). This definition arguably includes the transmission of confidential documents. Conversely, national courts may ‘ask the Commission for information in its possession’ (Article 15(1)). Here too, the limitation provided for in Article 27(1), which confines the use of such information to the purpose for which it was acquired, would not seem to apply, as Article 15 contains more specific rules in this respect.

III. Unresolved Issues in the Exchange of Information within the Network The draft Regulation establishes the legal framework for the exchange of information between the different authorities and courts that may be involved in the application of Articles 81 and 82 EC. Yet several issues have not been addressed in the Commission’s Draft Regulation and therefore need to be dealt with either in further legislative measures adopted at the Community level or by national procedural rules. Article 34(c) of the draft Regulation leaves ‘the practical arrangements for the exchange of information and consultations provided for in Article 11’ to the implementing regulation that will lay down more detailed rules. The rules applicable to the procedures carried out by the NCAs pursuant to their procedural laws will have to be included in national procedural laws, which the govern the application of Articles 81 and 82 EC by the NCAs.6 6See

Case C–67/91 Dirección General de Defensa de la Competencia v. Asociación Española de Banca Privada, supra note 5, para 32.

408

Working Paper VI

So, which issues concerning the exchange of confidential information need to be addressed in the implementing regulation and in the applicable national procedural rules?

1.

Common Notions of ‘Confidentiality’ and ‘Mutual Recognition’

If the exchange of information between the different members of the proposed network of competition authorities is to function smoothly, it would be helpful to at least ensure that the classification of information into different categories, depending on its accessibility to third parties, is according to rules that are common to all NCAs within the network. If this does not occur, the divergent classification of documents as ‘business secrets’, ‘internal’ or ‘confidential’ could lead to disruption. An example of such a problem might arise if companies oppose the transfer of their business secrets to the competition authority of another Member State because the latter grants a lower level of protection to such confidential information. It is therefore advisable that the Member States adopt common standards concerning which documents are communicable and which are not. There must be common criteria for determining what constitutes a ‘business secret’, ‘internal’ information, and what other information deserves protection as ‘confidential’. NCAs will gather information to apply Articles 81 and 82 EC on the basis of their national procedural laws and will hence also classify such information as ‘communicable’ or ‘non-communicable’ in accordance with their national laws. Nonetheless, these criteria should be ‘harmonised’ at the Community level in order to establish common criteria for the protection of such confidential information throughout the network, so as to ensure that there is a uniform standard of protection. Otherwise, the protection granted to confidential information by a NCA in its own jurisdiction could be lost in another jurisdiction after the transferral of the information to another NCA. In this context, Article 26(2) of the draft Regulation limits access to the Commission’s file by excluding ‘confidential information and internal documents of the Commission or the competition authorities of the Member States’ (in line with the Commission’s Notice on access to the file7). Obviously the same limitations must apply if third parties seek access to a file before a NCA, and national procedural laws must have provisions equivalent to Article 26(2) of the draft Regulation in order to protect confidential and internal information. Existing provisions granting a right to access will have to be broadened to exclude confidential documents, which stem from the Commission or other NCAs, and that have been transferred to a NCA. 7 Commission Notice on the internal rules of procedure for processing requests for access to the file in

cases under Articles 85 and 86 of the EC Treaty, Articles 65 and 66 of the ECSC Treaty and Council Regulation (EEC) No 4064/89, OJ C 23/3 [1997].

Technical Problem Areas

409

Even if a common standard were adopted (in line with our above-mentioned suggestion) concerning the nature of a confidential or internal document, there will nevertheless remain a need for a rule to maintain the degree of protection granted by the NCA ‘of origin’ — that is, where the information was gathered and classified as confidential — before other members of the network, once such confidential information has been exchanged with other authorities. A question then arises as to whether the NCA from which a party is seeking access to file has the power to re-classify the documents obtained from other members of the network as communicable or non-communicable. In spite of the existence of common substantive criteria to determine what constitutes a ‘business secret’, ‘internal’ documents or other ‘confidential information’, the interpretation of these concepts by a host of authorities is certainly liable to lead to divergent results in its practical application. Thus, the classification of a specific document as ‘confidential’ by one authority could well be ‘overruled’ by another authority that has obtained the document through the network, but that finds the document does not deserve protection as a confidential document. To ensure that companies cooperate and do not oppose the exchange of confidential information, national laws should foresee some kind of ‘principle of mutual recognition’ with respect to the confidentiality awarded to a specific document by the authority from which it originates. If the original authority that has obtained the information from an undertaking (irrespective of whether it has obtained it during an inspection, by way of a request of information, or by any other means) grants confidential treatment to a document, such confidentiality must be preserved throughout the entire network in which such documents are exchanged. Alternatively, where the confidential nature of an exchanged document is disputed, the authority facing a request to grant access to such a document in its possession could refer any such request to the authority of origin, which would be the competent authority to decide on any request to have access to confidential information (for example, if third parties claim that the document does not deserve such protection or should no longer be regarded as confidential). In a way, the idea of giving the authority of origin the responsibility for classifying a document as confidential for the entire network can be found in Article 5 of Regulation 1049/2001. This Article concerns public access to European Parliament, Council and Commission documents,8 and states: Where a Member State receives a request for a document in its possession, originating from an institution, unless it is clear that the document shall or shall not be disclosed, the Member State shall consult with the institution concerned in order to take a decision that does not jeopardise the attainment of the objectives of this Regulation. The Member State may instead refer the request to the institution.

8 OJ

L 145/43 [2001].

410

2.

Working Paper VI

The Use of Information Only for the Purpose for which it was Acquired

Article 12 of the draft Regulation allows for the vertical and horizontal exchange of information between the different competition authorities comprising the network. Any such information may be used as evidence in Community or national proceedings (Article 12(1)), although Article 27(1) provides that information collected in the context of requests for information or inspections ‘shall be used only for the purpose for which it was acquired’. However, this provision, which mirrors Article 20(1) of Regulation 17/62, is ‘without prejudice to Article 12’, the second paragraph of which significantly relaxes this limitation by allowing the use of information exchanged pursuant to Article 12(1) ‘for the purpose of applying Community competition law’. To the extent that Article 12(2) constitutes the more specific rule vis-à-vis Article 27(1), it must be understood that the limitation provided for in Article 27(1) does not exclude the use of information exchanged by either the Commission or the NCAs from application of Articles 81 and 82 EC in any competition proceedings. Thus, Article 12(2) significantly relaxes the previous rule under Article 20(1) of Regulation 17/62, which limited the use of the information acquired by the Commission ‘only for the purpose of the relevant request or investigation’. This apparent relaxation of the previous limitation contained in Article 20(1) of Regulation 17/62 raises important questions as regards the undertakings fundamental procedural rights. Where, for example, confidential information has been obtained from a third party that is not under investigation by a member of the network, the question arises as to whether the Commission or another NCA can use the information in an infringement proceeding against other companies, or even against the company that delivered the information as a third party, without being incriminated in the first place. The wording of Article 12(2) would suggest that such information could be used against the undertaking that provided it if the enforcement of Community competition rules is at stake. However, this interpretation would run counter to fundamental principles of both Community law and the national constitutional laws of the Member States, as it would amount to using potentially self-incriminating material in infringement proceedings contrary to principles developed by the European Courts in Orkem and Mannesmannröhren-Werke.9 The right to use information in such circumstances would therefore undermine the undertaking’s fundamental rights to mount a defence, the scope of which cannot be amended through secondary legislation. Similar issues would arise if an NCA obtains information from another authority that the latter had acquired for a specific purpose (for example, to investigate vertical restraints in a distribution system, or in the context of a 9 Cases 374/87 Orkem v. Commission [1989] ECR 3283; and T–112/98 Mannesmannröhren-Werke v. Commission [2001] ECR II–729.

Technical Problem Areas

411

voluntary submission), and the NCA were then to use the information to investigate a different infringement that had been discovered as a result of that information. As the ECJ held in Dow Benelux,10 professional secrecy and the undertaking’s rights to mount a defence would be seriously endangered if the Commission could rely on evidence against undertakings which was obtained during an investigation but was not related to the subject-matter or purpose thereof.

And in Spanish Banks the Court confirmed that [t]he rights of defence […] require […] that the information thus obtained should not subsequently be used outside the legal context in which the request was made.11

If NCAs and the Commission are barred from directly using such information for purposes other than that for which it was obtained, it also means that they cannot use such information as evidence if they wish to impose sanctions pursuant to their national competition laws. This is a rule that can be derived from a joint reading of Articles 27(1) and 12(2) of the draft Regulation. The above-mentioned guarantees should not only be enshrined in the new Regulation’s final version to limit the Commission’s powers, but they must also be a part of national procedural laws as the new Regulation (like the earlier Regulation 17/62) will not govern proceedings conducted by the NCAs, even where such proceedings are for the implementation of Articles 81 and 82 EC. As the ECJ stressed in Spanish Banks: proceedings conducted by the authorities in the Member States are distinct from those conducted by the Commission and the gathering of evidence by those authorities is in conformity with the provisions of national law, provided Community law is complied with. Even in cases where they apply the substantive provisions of Articles 85 (1) and 86 of the Treaty, it is incumbent upon the NCAs to implement them in accordance with national rules (emphasis added).12

Thus, national procedural law should protect the undertakings’ rights of defence by constraining the powers of their respective competition authorities. It should do this by prohibiting the use of any information obtained through the network for purposes other than those for which it was originally acquired. In any event, pursuant to the ECJ’s rulings in Spanish Banks and Dow Benelux, the competition authority that has obtained any such information through the

10 Case 11 Case

85/87 Dow Benelux v. Commission [1989] ECR 3137, para 18. C–67/91 Dirección General de Defensa de la Competencia v. Asociación Española de Banca Privada, supra note 5, para 36. 12 Case C–67/91 Dirección General de Defensa de la Competencia v. Asociación Española de Banca Privada, supra note 5, para 32.

412

Working Paper VI

network is not barred from initiating a new inquiry to verify or supplement such information if that information indicates the existence of conduct contrary to the EC competition rules, but the undertakings’ rights of defence certainly impede the direct use of such information.

3.

Procedure to Exchange Information

The Commission’s draft Regulation establishes clear obligations on the part of the Commission, NCAs, and even national courts, to transmit information between them. However, no procedural safeguards to protect the rights of the companies concerned can be found in the draft Regulation, except the obligation of professional secrecy contained in Article 27(2). Similar problems to those currently arising in the context of the passing of confidential information to the NCAs pursuant to Article 10(1) of Regulation 17/62 — addressed by both Courts in the SEP case13 — will arise when the Commission or an NCA transmits documents from its file to another authority. For what purpose is the information being transmitted to the other authority? That is, what use will the recipient authority make of such information in light of the limitations provided for in Articles 12(2) and 27(1) of the draft Regulation (discussed above)? Will the ‘authority of destination’ maintain the confidentiality that was granted to specific documents by the authority that originally obtained the documents? If a third party seeks access to a transferred file containing confidential documents, to which authority should it request such access and argue, if relevant, that certain documents should not be treated as confidential? These issues, which directly affect the fundamental procedural rights of the companies concerned, show that the Commission and NCAs should not be entitled to exchange information among themselves at their free discretion. As mentioned above, a first problem could lie in the degree of protection granted to confidential information by the different members of the network, which might vary between different Member States. Furthermore, there should be common rules defining the circumstances in which a member of the network may request information and documents obtained from undertakings in another Member State. These circumstances will be particularly concerned with whether to allocate the entire case to the requesting authority if it is in a better position to deal with the case. Inasmuch as the exchange of information governed by Article 12 of the draft Regulation concerns the application of Community law, it cannot be interpreted in a way that runs counter to the general principles of Community law. It is particularly important that it respects the fundamental rights granted by the 13 Cases T–39/90 SEP v. Commission [1991] ECR II–1497; and C–36/92 P SEP v. Commission [1994] ECR I–1911.

Technical Problem Areas

413

Community legal order. It should be remembered that the Community’s case-law treats the rights of the defendant as fundamental rights that deserve protection through administrative procedures that may lead to the imposition of penalties from the preliminary investigation stage.14 In this framework, it is necessary to establish the necessary procedural safeguards in the process of exchange of information within the network in order to protect such rights. The procedure established by the ECJ in the AKZO and SEP cases provides a useful model by which to protect the rights of undertakings that have provided information to NCAs, if the latter exchange the information with other competition authorities. In light of this case-law,15 if an NCA that bows to a request to transmit the documents to another authority, it should be under a duty to verify the concurrence of the above-mentioned circumstances that justify an exchange of information. It should then adopt a reasoned decision after hearing the views of the undertaking concerned. In particular, affected undertakings must be offered the opportunity to indicate the documents and passages whose disclosure might cause damage if no precautions were taken, and the nature and scope of the possible damage. Before the information is actually transferred, national law should provide a right enabling an affected undertaking to appeal the NCA’s decision, in order to have the assessments reviewed and to prevent disclosure. In exceptional cases in which it would not be possible to fully guarantee the protection of the affected undertakings even if the transmitting authority were to take the necessary precautions, it may be justified to refuse to transmit the information in question. The possibility of an appeal would serve the purpose of ensuring that the undertaking concerned is able to claim either that the requirements for transferral of documents have not been met (perhaps because the exchange is not connected to the application of Community law in the recipient State), or that its request for the confidential treatment of its documents has not been satisfied by the NCA and that there is no justification for this omission. In the latter case, it seems essential to safeguard the rights of the undertaking concerned by having a right to appeal, which suspends the transfer pending the decision. Otherwise, the (possibly) confidential information would be transferred as non-confidential, and it would be accessible to third parties in the Member State of destination. The ability to appeal would also allow an undertaking to raise other specific problems concerning the transfer of information to a national authority in circumstances similar to the one raised in the SEP case (in which the document in question was a contract that could have influenced commercial relations between that company and the Dutch authorities, through the monopoly supplier of gas).

14 See, for instance, Case 85/87 15 Case 53/85 AKZO Chemie

Dow Benelux v.Commission, supra note 10, paras 25–27. v. Commission [1986] ECR 1965, para 29; Case 36/92P SEP v. Commission, supra note 13, paras 38–39.

414

Working Paper VI

A right to appeal against the decision of the NCA of origin seems essential. The affected undertaking cannot be expected to appeal the decision of the NCA of destination to grant access to such documents to third parties, as this would require it to litigate in a Member State in which the undertaking might not even be established and where it will normally not be familiar with the legal system. If no effective remedy were awarded to the affected undertaking in the Member State in which the information originates, the undertaking would probably have no other option than to safeguard its rights before the requesting authority, although this solution would itself infringe on the undertaking’s fundamental rights. In the above-mentioned case-law regarding the exchange of information in the framework of Regulation 17/62, the ECJ and the CFI have acknowledged the need to apply an AZKO-type procedure concerning both the disclosure of confidential documents vis-à-vis other parties in a procedure (such as complainants), and the transmission of information to NCAs (as was the case in SEP) or to national courts (as in Postbank).16 In the SEP case, in the context of Article 10(1) of Regulation 17 (Article 11(2) of the draft Regulation), the ECJ referred to a company that claims before the competent NCA that a given document is of a confidential nature. The ECJ declared that, if such information is to be transmitted to the national authority, the company is entitled to a reasoned decision by the Commission open to judicial review.17 It should also be noted that the ‘SEP proviso’, as it is known, will remain good law after the adoption of the new regulation, as it is founded on the fundamental rights of the defendant. The guarantees enshrined in the SEP procedure derive from a general principle of Community law that applies to administrative procedures and protects the legitimate interest of undertakings in protecting their business secrets and confidential information against the extremely serious damage that could result from improper communication of these details.18 Fundamental principles like this cannot be amended by secondary legislation. In these circumstances, national procedures will govern the activities of the NCAs with regard to the exchange of information within the network, provided they are not discriminatory vis-à-vis the application of Community law and that they do not make it impossible or excessively difficult to exercise rights conferred by Community law. In a number of cases, the ECJ has already created remedies under national law in order to ensure the effectiveness of Community law.19 In the present case, this principle would require NCAs to adopt the guarantees

16 Case T–353/94 Postbank v. Commission [1996] ECR II–921. 17 Case 36/92P SEP v. Commission, supra note 13, para 39. 18 Case 36/92 P SEP v. Commission, supra note 13, para 36; see

also Case 53/85 AKZO Chemie v. Commission, supra note 15, paras 28–29. 19 See, for instance, Case C–213/89 The Queen v. Secretary of State for Transport ex parte Factortame (No. 2) [1990] ECR I–2433, where the ECJ ruled that, in a case involving Community law, when a

Technical Problem Areas

415

enshrined in the AKZO/SEP procedure, if their national legislation does not already provide for the level of protection awarded by that procedure.

4.

Protection of Legal Privilege within the Network

Against the background of the significantly expanded investigatory powers of the Commission in the draft Regulation, it is certainly disappointing that the Commission has not incorporated all the procedural achievements developed by the Community Courts in the last 40 years of application of Regulation 17/62 in the text of the new Regulation. This is particularly apparent with respect to the right to legal privilege that protects communications between undertakings and their independent lawyers. Although developed by the ECJ as a fundamental principle of Community law in the 1982 AM&S judgment,20 the draft Regulation contains no word about the protection of legally privileged information. Where documents enjoy legal privilege under AM&S and Hilti21 because they have been produced for the purpose of defending undertakings against a procedure carried out by the Commission, they cannot be made the object of a request of information or confiscated during an inspection. Consequently, they cannot be the object of an information exchange between competition authorities and/or courts. But to be effective, the protection of legally privileged communications must go further in the context of the network of competition authorities: a member of the network must not request from an undertaking documents that have been produced for defence purposes in proceedings before any other competition authority (either the Commission or an NCA). It is clear that NCAs are obliged to respect the legally privileged character of documents that are generated in procedures before the European Commission when they qualify for such protection in accordance with the above-mentioned case-law of the European courts. In other words, they cannot request correspondence exchanged between undertakings and external lawyers in the framework of proceedings before the European Commission. This would be the case even if they were simply applying national competition law. Otherwise, in spite of the protection attached to lawyer-client communications by Community law, if companies ran the risk that information exchanged with their outside lawyers in the context of Community proceedings would be requested and used by NCAs, they would no longer enjoy the necessary freedom to mount an appropriate

national court considers that the sole obstacle to granting interim relief necessary to ensure the full effectiveness of the judgment law is a rule of national law, it is obliged to set aside that rule. 20 Case 155/79 AM&S Europe v. Commission [1982] ECR 1575. 21 Order of the Court of First Instance, Case T–30/89 Hilti v. Commission

[1990] ECR II–163.

416

Working Paper VI

defence in a procedure before the European Commission in the terms undoubtedly guaranteed by Community law. Therefore, the possible violation by a NCA of the protection that Community law confers on lawyer-client communications would be contrary to Article 10(2) EC, which provides that ‘[Member States] shall abstain from any measure which could jeopardise the attainment of the objectives of this Treaty.’ This will be the case if an NCA applies Community competition law under the Draft Regulation. In these circumstances, the ECJ has explicitly declared that the fundamental rights recognised in its case-law — and derived from the constitutional traditions common to the Member States, as well as from the guidelines supplied by international treaties for the protection of human rights applicable to the matter in question — are an integral part of the general principles of the Community legal order. These bind the Member States when they apply Community law, as was established in the Wachauf judgment of the ECJ.22 It is less clear whether, should it be applicable to the case, NCAs are obliged to accept the legally privileged character of documents in accordance with the domestic law of a different Member State.23 If they are applying Articles 81 and 82 EC, as will be the case under the new Regulation, they will be obliged to respect at least the Community law standard, pursuant to the Wachauf case-law. But, in our opinion, any higher standard of protection in either the country of destination or origin should also be respected in the context of exchanges within the network. With respect to the country of destination, this is because the fundamental right of defence should prevent any authority from taking advantage of a ‘weakness’ in foreign legislation in order to obtain information that its legal order prevents it from obtaining domestically. With respect to the country of origin, it is because otherwise undertakings that benefit from a high degree of protection in their relations with lawyers would see this protection betrayed; in other words, the country granting the lowest degree of protection would impose its standard across the whole EU.

V.

How to Address the Issues Raised in this Paper?

The procedural issues arising in the context of the exchange of confidential information discussed above could, in principle, be addressed either individually by the national procedural laws of the Member States or by legislative measures adopted at the Community level. NCAs are primarily bound by their national procedural laws and will apply Articles 81 and 82 in the context of the new network according

22 Case 5/88 Hubert Wachauf v. Federal Republic of Germany [1989] ECR 2609, para 19. See also case C–2/92 Bostock [1994] ECR I–955, para 16. 23 It is difficult to ascertain whether Community or domestic law governing legal privilege should be applicable if no procedure has been initiated.

Technical Problem Areas

417

to these national procedural rules (cf Article 5 of the draft Regulation24). Therefore, it does not seem possible that the issues raised here could be dealt with in the implementing regulation that will detail ‘the practical arrangements for the exchange of information and consultations provided for in Article 11’, as envisaged in Article 34(c) of the draft Regulation. As quoted above, NCAs’ application of Articles 81 and 82 EC according to their national procedural rules was recognised by the Court in Spanish Banks.25 Thus, the question remains whether the procedural issues arising in the context of the new network’s functioning should be individually addressed by each Member State or whether these issues could be ‘better’ resolved (in the sense of Article 5(2) EC) by adopting a Directive with harmonised procedural rules on the basis of Article 83 EC. As set out in this paper, the divergent interpretation of notions such as ‘confidential information’, ‘internal documents’ or ‘business secrets’ by a host of competition authorities, as well as different procedural arrangements to safeguard the rights of the defence of undertakings in the Member States, could soon lead to the disruption of the network’s functioning. Therefore, in our opinion, it would be advisable to harmonise national procedures when it comes to the application of Articles 81 and 82 EC, particularly with regard to the requirements that must be fulfilled to exchange (confidential) information; the limits on the use that the requesting authority can make of exchanged information; the procedural steps to be adopted in this respect by the requesting and the requested authority, and the legal provisions for undertakings to appeal any decision to transmit information to another authority before the courts; and the procedural rules pertaining to the respect of legally privileged communications. As the Economic and Social Committee put it in its Opinion on the Commission’s Draft Regulation of March 2001:26 The importance of procedural provisions cannot be ignored, however, and the Committee cannot support the Commission’s position on this matter. The last paragraph of point 3 of the Explanatory Memorandum reads: ‘Thus, the draft Regulation does not purport to harmonise national procedural law, except that it grants the Commission and the NCAs the power to make submissions on their own initiative’. 24 The

note to Article 5 of the Explanatory Memorandum states that if NCAs ‘find that there is infringement of Article 81 as a whole, or of Article 82, the competition authorities of the Member States are to take effective action against the conduct in question, acting in accordance with the draft Regulation and applicable national procedural rules.’(emphasis added). 25 Case C–67/91 Dirección General de Defensa de la Competencia v. Asociación Española de Banca Privada, supra note 5, para 32. More specifically with respect to the use of information, in the same paragraph the Court declared that ‘in both cases [where the NCAs use the information concerned for the purpose of applying Community law and cases where those authorities take action to apply national competition law], the proceedings conducted by the authorities of the Member States are distinct from those conducted by the Commission and the gathering of evidence by those authorities is in conformity with the provisions of national law, provided that Community law is complied with.’ 26 CES 410/2001 — 2000/0243 (CNS) IT/CAT/ht. Some of the Committee’s concerns in this regard were contained in its Opinion on the ‘White Paper on modernisation of the rules implementing Articles 81 and 82 of the EC Treaty — Commission programme No 99/02’, OJ C 51/55 [2000].

418

Working Paper VI

The Committee can only reiterate its concern that consistent application of the principles, which all parties agreed on, will be compromised owing to the wide discrepancies in practice between the Member States. Procedures (or, at least, administrative procedures) should to a certain extent reflect the unity of the principles. In this regard the Committee would also recommend that the Commission bear in mind that Article 83 of the Treaty also provides for directives to be used as an instrument ‘to give effect to the principles set out in Articles 81 and 82’. A directive is a more flexible instrument because it generally offers the choice of different options and allows a suitable period of time for provisions to be implemented. It is thus an instrument which can be adopted in order to start taking practical legislative steps -albeit only prospective ones- to harmonise complex fields such as procedures. (points 2.10.5 and 2.10.6. Emphasis added.)

We share the Committee’s view that the procedural measures to be adopted to make the network function smoothly and to safeguard the fundamental rights of undertakings would be best addressed in a Directive pursuant to Article 83 EC.

VII Monica Widegren* Consultation Among Members Within the Network

The network to be formed by the national competition authorities (NCAs) and the European Commission is a core element of the proposed reform for the application of Articles 81 and 82 EC. According to the Commission’s reform proposal, the network will provide an infrastructure for the mutual exchange of information, including information of a confidential nature. Close cooperation between the competition authorities will be the foundation of the network. Information exchange and consultation are the main cornerstones of cooperation within the network. The mechanisms for sharing of information and consultation shall be the tools for an efficient and consistent application of Articles 81 and 82 EC within the European Union. These tools must be robust enough to be managed effectively by a large number of competition authorities, including those of the future Member States. In addition, the obligation in Article 3 of the draft Regulation to apply Articles 81 and 82 EC in all cases to which they are applicable will lead to a large number of ‘new’ cases at the national level (that is, cases that today are dealt with under national competition laws). Even if the network will not deal with applications for exemptions or negative clearance, the network will no doubt have to deal with a large number of cases. Yet this is not only a question of network dimensions. The enforcement tools must also be properly designed in order to function well in a network of competition authorities in which perhaps one third of the members will not have the experience of cooperating in competition cases with either the European Commission or other NCAs. This circumstance may be worth underlining. Concepts that seem clear to current Member States and their authorities, because they have experience of how those concepts are used in practice under the current Regulation 17/62, may be understood in quite different ways by the competition authorities of the new Member States. There is no definition of the concept of ‘consultation’ in Regulation 17/62 or the draft Regulation. For the purpose of the discussion below, the concept of consultation is understood as the initiative taken by one or several authorities to share information in the network in order to get a reaction from the other network members that will lead to further action. This paper deals with the consultation between competition authorities in the network. It does not cover the consultation with national courts, or consultation between national courts and the Commission or the European Court of Justice. *Swedish Competition Authority. All views expressed here are personal.

420

Working Paper VII

As will be discussed in this paper, consultations that are envisaged or may be derived from the draft Regulation may be formal or informal and may have different objectives and characteristics. The paper focuses on the various situations in which a consultation will take place in order to ensure an efficient and consistent application of Articles 81 and 82 EC. At least three types of consultation are of crucial importance for the functioning of the network and are laid down in the Commission’s draft Regulation: • consultation for the proper allocation of new cases in the network; • consultation prior to the making of decisions, an obligation imposed on the competition authorities to ensure a consistent application of Articles 81 and 82 EC; • consultation requesting members of the network to gather information or to carry out inspections in their territory. The mechanisms for consultation are primarily found in Articles 11 and 14 of the draft Regulation. In addition to those provisions, more detailed rules will be laid down in an implementing Commission Regulation in accordance with Article 34. Furthermore, the interpretation of the rules with regard to criteria for the allocation of cases, and details about how to handle cases in the network, will be explained in a notice concerning cooperation between members of the network. A draft notice of this kind is being discussed in the Council Working Group, and it is foreseen that the notice will contain some details about the consultations in the network.

I. The Obligation for the Commission to Consult the Authorities of the Member States The draft Regulation retains largely the same provisions as Regulation 17/62 for consultation with the competition authorities of the Member States. According to Article 10, the Commission shall transmit copies of the most important documents to the authorities of the Member States, which have the right to express their views about the procedure. According to Articles 11 to 14, the Commission may request information and assistance from the Member States. The only formal type of consultation is stipulated in Article 10, which imposes an obligation on the Commission to consult the Member States prior to making decisions about infringements, exemptions or negative clearances. It also stipulates that the Member States shall be consulted within the framework of an Advisory Committee. Article 11(2) of the draft Regulation imposes a similar obligation on the Commission to transmit copies of the most important documents to the members of the network with a view to making decisions about infringements, including

Technical Problem Areas

421

interim decisions, decisions with commitments or decisions concerning the inapplicability of Articles 81 or 82 EC. According to Article 14 of the draft Regulation, the Commission will also be obliged to consult an Advisory Committee prior to making any of the abovementioned decisions, as well as decisions about fines as laid out in Articles 22 and 23. Council Working Group discussions about strengthening the role of the Advisory Committee in view of the closer cooperation between the members of the network have resulted in a proposal to amend the text. However, it should be noted that no obligation of consultation has been imposed on the Commission regarding the application of Article 11(6). This provision corresponds to the rule of Article 9(3) in Regulation 17/62 and gives the Commission the power to initiate proceedings in a case that ‘shall relieve the competition authorities of the Member States of their competence to apply Articles 81 and 82 EC’. The aim of this rule is to ensure efficient case allocation and consistent application of Community competition law. At least three situations can thus be envisaged that could motivate the application of this provision: • Competition authorities in the network cannot reach a rational agreement about the most efficient allocation of a particular case. There could be a dispute between authorities that cannot be settled even after discussion by the Advisory Committee, or no decision about which authority will hear the case can be reached within a ‘decent’ period of time. • The Commission may consider that proceedings before an NCA are being wrongly handled, and the authority will therefore probably adopt a decision that is not in conformity with EC rules and case-law. The Commission may come to that conclusion on the basis of the authority’s investigation or as a result of the authority’s draft decision that has been submitted to the network. • The Commission may take the view that a case dealt with by a national authority is of a particular Community interest, even though this was not identified when the case was initiated. During discussions within the Council Working Group, the Member States have expressed the view that consultation with the authority or authorities concerned must take place prior to the application of such a strong and exceptional measure. It could also be argued that the consultation should not be limited only to the authority(ies) affected by the Commission’s intervention. Application of this provision could instead be considered to be a matter of importance for all network members, who should also be given the opportunity to discuss the matter in the Advisory Committee. Considering the importance of this provision for the functioning of the network, it seems appropriate to state explicitly in the draft Regulation that the Commission shall consult the national authorities prior to using the powers awarded to it by Article 11(6).

422

II. 1.

Working Paper VII

Obligations of Consultation Imposed on NCAs Allocation of Cases

Article 11(3) of the draft Regulation imposes an obligation on the NCAs to inform the Commission at the outset of their own proceedings whenever they intend to apply Articles 81 and 82 EC on their own initiative. This obligation largely corresponds with the Commission’s obligation to inform the NCAs of new cases. However, for the purpose of efficient case allocation, it is important that NCAs also inform each other without delay of new cases dealt with within the network. It is outside the scope of this paper to discuss what is meant by ‘case’. However, it is worth noting that the discussions in the informal workshop for the members of the Council Working Group held under the Swedish presidency last year clearly showed that there are different notions and legal interpretations about what constitutes a new case before an NCA. By submitting information to the network about a new case, the authority in question also initiates a consultation procedure aimed at gaining agreement between the network members about the allocation of the case. The information must be brief and concise in order to facilitate the consultation. But it must, of course, provide information that is sufficient for an assessment of the allocation. A standard format for this allocation consultation will probably have to be elaborated. When should this allocation consultation take place? Article 11(3) of the draft Regulation stipulates that the information shall be submitted at the outset of the proceedings. As a result of the discussions in the Council Working Group, the concept of ‘after the first formal investigative measure’ has now been added to the text. The interpretation of this concept will inevitably vary between the members of the network, as the Member States have different national procedural laws that regulate how cases are to be opened, investigated and handled by the authorities. The authorities may also have internal case handling procedures that have an impact on their interpretation of ‘after the first formal investigative measure’. It remains to be seen whether this discretion for the national authorities to decide when the information should be submitted and the consultation started will lead to any real distortions in the functioning of the network. Timeliness will be important for an effective allocation procedure. It might very well be the case that a complaint about a certain market practice is lodged more or less simultaneously with several competition authorities. It will then be important for the network to identify this common complaint, in order to allow a timely consultation about which authority is the best placed to handle the case. The time frame of this consultation process about the allocation of cases is also important. From the point of view of efficiency, it is essential that a final agreement about the allocation can be reached as speedily and diligently as possible. Companies also have a right to be informed without undue delay about

Technical Problem Areas

423

which authority will handle the case. In the discussions between the Member States, a time frame of two or three months has been mentioned for this process. It goes without saying that the time for consultation must be kept at a minimum in those cases in which an interim measure may be called for. A special form of consultation will also have to be applied to those cases in which an authority intends to initiate the proceedings by carrying out a ‘dawn-raid’ inspection at the premises of certain undertakings. Due to the highly confidential character of the measure, it may seem very natural to limit the circle of authorities that will be informed before the inspection takes place. However, it can be argued that it is equally important that at least informal consultation occurs. Cartels are frequently not limited to the national territory, and a joint, parallel action by two national authorities or an action by the Commission may be the most efficient method of intervening against a serious infringement. In most cases there will probably be no dispute or even doubt about which authority is the best placed to handle the case. However, the matter may be discussed in the Advisory Committee if no satisfactory agreement is reached. This type of consultation is not envisaged in Article 14 of the draft Regulation, which deals with the role of the Advisory Committee. But it is foreseen that the Advisory Committee will not only deal with draft decisions but also with procedural issues and guidelines etc., as is the case today. Therefore, there does not seem to be any reason for not bringing discussions on interpretation of the network provisions to an Advisory Committee, if one or more competition authorities so request.

2.

Consultation Prior to Making Decisions

In the new enforcement system, NCAs will have to inform the Commission of their intentions before adopting a decision to stop an infringement (Article 7) or to accept commitments (Article 9). The provisions concerning this obligatory consultation are laid down in Article 11(4). It is further stipulated that, for the purpose of this consultation, NCAs shall submit a summary of the case together with copies of the most important documents drawn up in the course of their own proceedings. Further, the Commission shall be given one month to react to the consultation, if necessary. The Commission’s proposal does not contain any rules about consulting all the members of the network; NCAs are only obliged to consult the Commission. However, the information should also be shared with the other members of the network. Other NCAs could make useful contributions to the decision-making process, as is often the case when the Commission submits its draft decisions for consultation in the Advisory Committee. Article 14(6) establishes that the Commission may put a national case on the agenda for discussion in the Advisory Committee on its own initiative or on

424

Working Paper VII

request from the NCA concerned. However, this provision unnecessarily limits the scope for consultation. It is certainly in the interest of the network that important cases can be discussed among the competition authorities. Any competition authority should therefore be entitled to request that a case be put on the agenda of the Advisory Committee. This would require that all members of the network receive summaries of draft decisions from the national authorities.

3.

Requesting Information

It may also be seen as a consultation to request assistance from members of the network to gather information and to carry out inspections on behalf of another authority as provided for in Article 21(1). This kind of cooperation is an important efficiency element in the network. In order to be able to apply Articles 81 and 82 EC in an efficient way, NCAs will require much more information from outside their own national territories than they do today. This situation will particularly arise when a case is allocated to one NCA but the restriction on competition also affects other countries, or if the companies involved are located in other Member States. The reason for considering requests for information as a consultation lies in the nature of the request. In contrast to the provision in Article 21(2) that obliges NCAs to undertake inspections on the Commission’s orders, there is no explicit obligation to respond positively to a request from another national authority. It must of course be assumed that all members of the network will make their best efforts to assist each other in the application of the Community rules. Yet, as inspections are very resource consuming and require careful planning, NCAs may find it difficult to provide the assistance at the time requested by the investigating authority. Consultation between the authorities concerned may also be needed to clarify details of the information-gathering activities. As each authority applies its national procedural rules, this may lead to differences in what can be achieved in different jurisdictions.

III.

Informal Consultations

In addition to the consultations of a more formal character that have been discussed above, the close cooperation between the members of the network will also give rise to a need for frequent informal consultations. Many minor issues related to making the network function smoothly will have to be resolved through informal consultations. Authorities may also find it useful to consult each other about substantive questions, such as their experiences dealing with certain competition problems or markets.

Technical Problem Areas

425

A specific issue for consultation with the Commission will probably be the applicability of Articles 81 and 82 EC, which concern whether the Commission considers that there is an appreciable effect on trade between the Member States. However, the importance of consultation about this dividing line between Community law and national law is related to the final wording of Article 3 of the draft Regulation. If the obligation to apply Articles 81 and 82 EC also allows national law to be applied in parallel, the importance of establishing the division will be significantly reduced. The Commission invites consultations ‘on any case involving the application of Community law’ (Article 11(5) of the draft Regulation). Informal consultations with the Commission on substantive, as well as on procedural, issues will certainly be needed. The Commission must be prepared to dedicate sufficient resources to fulfil this important task as the central point of the network.

IV. 1.

Technical Aspects of Consultations Languages

One general aspect that will be important for the functioning of the network is the question of the language or languages used for consultation between the members of the network. To allow for proper consultations, all information that is put into the network must necessarily be in a language that can be managed by all competition authorities. The system utilised under Regulation 17/62 will no longer be acceptable. Today the Commission sends copies of complaints or notifications in the original language to the competent authorities of the Member States. In practice, this means that information may be received in all eleven languages of the European Union. When new members are added, new languages will also be part of the family. It is evident that information sharing and consultation in the network cannot accommodate a large number of different languages, nor can every national authority cope with all languages. In order to avoid a situation in which language would become an important barrier for the proper functioning of the network, it will be necessary to introduce the ‘working language’ concept into the network. And one working language seems to be the most rational solution.

2.

Intranet

It is foreseen that consultations in the network will be based on electronic communications within an Intranet. As consultations in the network have to be speedy, a system of electronic communication seems to be the only reasonable solution. However, the Intranet system between the national authorities must be

426

Working Paper VII

as reliable as possible. Software programmes must be installed to allow for the exchange of encrypted information. In addition, each competition authority will have to build an internal organisation to handle the consultations in an efficient way. Closer ‘horizontal’ cooperation and time-limits that must be met for the various consultations within the network are two aspects of the new system that will probably require changes in the internal routines that national authorities currently apply for their ‘vertical’ cooperation with the European Commission under Regulation 17/62.

VIII Jacques HJ Bourgeois* Consultations between National Competition Authorities and the European Commission in a Decentralised System of EC Competition Law Enforcement

I.

Introduction

Consultation between National Competition Authorities (NCAs), and between the European Commission and NCAs, is one of the principal elements of the complex decentralised system of EC competition law enforcement that is proposed by the European Commission.1 A division of genuine responsibilities in enforcing competition rules is not without precedent in Europe. The enforcement system of the competition rules of the European Economic Area offers an interesting example in this sense.2 Here, the risk of conflicting decisions was recognized and a series of cooperation measures was introduced ‘with a view to developing and maintaining a uniform surveillance throughout the European Economic Area in the field of competition and to promoting a homogeneous implementation, application and interpretation of the provisions of their Agreement to this end’.3 Protocol 23 of the EEA Agreement provides for a number of cooperation measures, such as exchange of information and consultations on general policy issues at the request of either of the surveillance authorities,4 ‘switching’ notifications and complaints, 5 consultation before publication of intention to give negative clearance or to take an exemption decision and before addressing a Statement of Objections, 6 participation at hearings, 7 access to

* Partner,

Akin Gump Strauss Hauer & Feld (Brussels) and Professor at the College of Europe (Bruges). 1 Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, COM (2000) 582 (September 27, 2000). 2 For comments, see i.a. Haagsma A. (1993): ‘The competition rules of the EEA and the European Agreements: lawyer’s paradise or user’s safe harbour’ in Slot P.J. and McDonnell A., eds., Procedure and Enforcement in EC and US Competition Law, Sweet & Maxwell, London, pp. 243–66. 3 Agreement on the European Economic Area, Article 58 (OJ 1994 L 1/3). 4 Article 1. 5 Article 10, para 2. 6 Article 3. 7 Article 5.

428

Working Paper VIII

documents lodged with the other surveillance authority, and the ability to make observations.8 Thus, consultation within the network is not terra incognita for the European Commission and the EC Member States. Surely what was acceptable under the EEA Agreement should be the base line for a consultation procedure within the EU? When putting forward its formal proposal for a ‘new Regulation 17’, the Commission said that ‘maintaining consistent application is essential in an enforcement system in which parallel powers to apply Articles 81 and 82 EC are exercised by the Commission, national competition authorities and national courts’.9 Some would argue that, when opting for a decentralised system of enforcement, one should allow the chips to fall where they may fall, and that the Commission ought not to have its cake and eat it too. Others take the view that consistent application of competition rules to conduct that has cross-border effects is of paramount importance. The Commission proposes several devices in order to achieve this objective. One is consultation with three requirements. First, NCAs would need to ‘consult the Commission prior to [the] adoption of prohibition decisions, decisions accepting commitments and decisions withdrawing the benefit of a block exemption regulation’. In addition, ‘in case of substantial disagreement within the network, the Commission retains the power to withdraw a case from a national competition authority by itself initiating proceedings in the case’.10 Secondly, NCAs may consult the Commission on any pending case.11 Thirdly, where necessary, a case may be put on the agenda for discussion by the Advisory Committee.12 This paper does not take these proposed measures for granted. Instead, it attempts to deal with six practical and straightforward questions relating to consultations from the perspective of practicing lawyers.13 Practicing lawyers and their clients are of two minds about consultations within the network, leaving aside the issue of protection of confidential information. On the one hand, such consultations compensate to some extent for the likely absence of legal effects of an NCA decision in other Member States. Consultation should normally contribute to avoiding different outcomes. On the other hand, consultation is likely to lead to increased exposure, as consultations about a case handled by one NCA may trigger proceedings by another.

8 Article 7. 9 Explanatory Memorandum of the draft Regulation, at p. 23. 10 Article 11(6). 11 Article 11(5). 12 Article 14(6). 13 I thank Pierre Bos (Dorsey & Whitney), Cani Fernández

(Cuatrecasas), Olivier d’Ormesson (Linklaters) and John Temple Lang (Cleary, Gottlieb, Steen & Hamilton) for their input.

Technical Problem Areas

429

II. Should Problems and/or Cases be the Subject of Consultations within the Network? The distinction between problems and cases appears appropriate in that a case may be comprised of a number of issues of fact and law, only some of which may be relevant for consistent application. Requiring consultation on cases would overburden the system. Leaving it to the relevant NCA to filter out problems may be insufficient. A better solution would be for NCAs to inform the network of the essential elements of the case. The other NCAs and the Commission can then request consultation. Some practicing lawyers consider that there should be no consultation about cases unless parties could participate in the ‘decision-making process’, as otherwise the rule audi et alteram partem would not be respected. This position is inspired by the system of preliminary rulings by the ECJ. Others consider that the parallelism is inappropriate, particularly if the results of the consultation are not binding on the NCA that takes the decision. Moreover, it is difficult to see how some form of participation by private parties in this consultation could be arranged in practice. Another separate question is whether parties should be told what views the other NCAs and the Commission have communicated to the initiating NCA. Parties should be made aware if these views insofar as they concern comments about the case or other cases that have been contemplated by the commenting NCA and the Commission.

III. Should Every Problem of EC Competition Law/Every Case of Application of EC Competition Law be the Subject of Consultations? If Not, Should there be Some Criteria or Threshold for Triggering Consultation? In line with the reply to the first question, the essential elements of every case should be communicated to the network, and any NCA and the Commission should have the ability to seek consultations. At any rate, the scope of the (mandatory) consultations proposed by the Commission is too narrow. Consultations should also be mandatory concerning the application of Article 81(3) EC, the applicability of Articles 81 and 82 EC to a given factual situation (including, for example, determining whether a certain practice may be subject to one or the other, or both) and the application of Article 81 EC to de minimis situations. Some practicing lawyers wonder whether the consultations should be limited to cases raising the same issue(s) in several Member States, as consultations are

430

Working Paper VIII

only necessary where there is a risk of contradiction between the decisions of several NCAs. Others consider that, at least during a transitional period and until practice within the network has reached a certain degree of harmonization, there should be no criteria or thresholds. Moreover, one would expect NCAs and the Commission to seek consultations when it really matters.

IV. Should there be Consultation Ex-officio? Or Should the Network Simply be Informed about New Cases/Problems of EC Competition Law, Leaving it to Another Member of the Network (NCA or Commission) to Request a Consultation? Conceivably, the question of whether consultations should be ex officio arises only if the essential elements of every case are not automatically communicated to the network. If the essential elements of every case are communicated to the network, it may be left to the initiating NCA to request consultations or to other NCAs and the Commission to do so.

V.

Should Consultations be Given any Legal Effect?

This question arises in the event that the Commission will not propose, or that the Council will not accept, that a decision given by an NCA in a Member State will be binding in another Member State. One should be clear about what the consultation process entails or may entail: the initiating NCA informs the network; the initiating NCA asks for advice, information, similar cases, and so on; the other NCAs and the Commission respond or ask for coordination, so that all NCAs treat similar cases similarly; the other NCAs disagree with the approach being adopted by the initiating NCA; the consultation gives rise to disagreement, which is brought before the Advisory Committee, or the Commission pulls the case in; the case is transferred to another NCA; the initiating NCA has other cases transferred to it. According to some, the system of decentralised enforcement would be ineffective in light of the rationale of the network if the results of consultations have no legal effect. Where the consultation leads to findings on a particular issue, the findings should be binding on the NCA requesting the consultation as well as on the other NCAs. If the consultation is about the same case being investigated by two or more NCAs, the result of the consultation should be binding on all of the NCAs involved. A variation on this view is that joint findings generated by the consultation will be soft law in that it will be difficult for the Commission and NCAs to deviate from them without good reasons.

Technical Problem Areas

431

Another view is that consultations do not need to be given legal effect, provided that the new Regulation gives the Commission enough power to avoid major inconsistencies.

VI.

Should Consultations Take Place Ex Ante or Ex Post?

To be consistent with the reply to the first question, consultations should take place ex ante (that is, prior to the adoption of the decision by an NCA). Such consultation already exists within the Advisory Committee with respect to Commission proceedings. Consultation ex post would be no more than giving information to the network about the outcome of a given NCA proceeding. If it is not made possible for parties to participate in the consultation, they should be given the ability to make comments to the NCA in charge on the result of the consultation. The final decision should be communicated to the network.

VII. Should the Result of Consultations within the Network be made Accessible to the Public, for Example, on the DG COMP Website? As the main purpose of the consultation is to ensure a high level of consistency in the application of Articles 81 and 82 EC, it would seem appropriate to make the results of the consultation publicly available. They constitute a valuable source of information on the application and interpretation of Articles 81 and 82 EC, to which all parties involved in the enforcement of these provisions should have access. Business confidential information would be likely to be involved in the consultation process. Parties to the proceeding should be given the ability to subtract such information from the published text on the result of the consultation. For reasons of legal certainty, either the Commission or the NCA in charge or should be given the responsibility of publishing the result of the consultation. There is a preference for publication on the DG COMP website.

IX Wouter P.J. Wils* The EU Network of Competition Authorities, The European Convention on Human Rights and the Charter of Fundamental Rights of the EU

I.

Introduction

This paper addresses the question of the compatibility of the network of competition authorities as envisaged in the European Commission’s proposal for a new Council Regulation implementing Articles 81 and 82 EC1 (hereafter also: ‘the network’ and ‘the draft Regulation’) with the Convention for the Protection of Human Rights and Fundamental Freedoms2 (hereafter also: ‘the European Convention on Human Rights’, ‘the Convention’, or ‘ECHR’) and with the Charter of Fundamental Rights of the European Union3 (hereafter also: ‘the Charter of Fundamental Rights’, ‘the Charter’ or ‘CFR’). The paper is structured as follows: Part II describes the network as envisaged in the draft Regulation, pointing out the three objectives or functions of the network (efficient allocation of cases, assistance in respect of fact-finding and coordination to ensure the consistent application of Articles 81 and 82 EC) and

* Member

of the Legal Service of the European Commission, lecturer in European law and economics at the University of Utrecht. The author gratefully acknowledges comments and contributions received from Peter Oliver, Katja Arenmaa, Richard Lyal and Kirsi Leivo. All views expressed in this paper are strictly personal and should not be construed as reflecting the opinion of the European Commission or the national competition authorities. The author can be reached at [email protected]. 1 Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty and amending Regulations (EEC) No 1017/68, (EEC) No 2988/74, (EEC) No 4056/86 and (EEC) No 3975/87 (‘Regulation implementing Articles 81 and 82 of the Treaty’), [2000] OJ C365E/284; for the Explanatory Memorandum, see COM(2000)582 final, available at: http://europa.eu.int/eur–lex/en/com/pdf/2000/en_500PC0582.pdf. For an analysis of this proposal, see my paper entitled ‘The Commission’s proposal for a new Council Regulation replacing Regulation No 17’, presented at the Fordham Corporate Law Institute 27th Annual Conference on International Antitrust Law & Policy (New York, 19–20 October 2000) and published in Hawk B., ed., Proceedings of the Fordham Corporate Law Institute 2000, New York, Juris, 2001, pp. 313–363, as well as in 24 Fordham International Law Journal 5 (2001), pp. 1655–1717 and in Wils W. P. J. (2002): The Optimal Enforcement of EC Antitrust Law: Essays in Law and Economics, London, Kluwer Law International, pp. 105–62. 2 Convention for the Protection of Human Rights and Fundamental Freedoms as amended by Protocol No. 11 with Protocol Nos. 1, 4, 6 and 7, available at: http://www.echr.coe.int/Eng/ BasicTexts.htm. 3 OJ C 364/1 [2000], also available at http://ue.eu.int/df/docs/en/CharteEN.pdf.

434

Working Paper IX

the regulatory provisions proposed to achieve these objectives. Parts III and IV introduce respectively the ECHR and the CFR, pointing out their relevant provisions in the present context, and describing briefly their legal status. Parts V, VI and VII analyse, respectively with regard to the three objectives or functions of the network, possible problems under the ECHR and the CFR, followed by a brief conclusion in Part VIII.

II.

The Network of Competition Authorities

1.

General Provisions

Article 11(1) of the draft Regulation provides generally: The Commission and the competition authorities of the Member States shall apply the Community competition rules in close cooperation’.4

According to the Explanatory Memorandum: It is a core element of the Commission’s proposal that the Commission and the national competition authorities should form a network and work closely together in the application of Articles 81 and 82. The network will provide an infrastructure for mutual exchange of information, including confidential information, and assistance, thereby expanding considerably the scope for each member of the network to enforce Articles 81 and 82 effectively. The network will also ensure an efficient allocation of cases based on the principle that cases should be dealt with by the best placed authority.5

The Explanatory Memorandum also specifies that ‘the principle of close cooperation [covers] cooperation between the Commission and the Member States competition authorities on the one hand and between the latter on the other hand’.6 We thereby distinguish three objectives — or functions — of the cooperation within the network: (1) the efficient allocation of cases between the members of the network, (2) assistance between the members of the network in respect of fact-finding, and (3) coordination within the network so as to ensure consistent application of Articles 81 and 82 EC by all its members.

4 As

the draft Regulation has Article 83 EC as its sole legal basis, ‘the Community competition rules’ can here only mean Articles 81 and 82 EC and their implementing rules. 5 COM(2000)582 final, p. 6. 6 Idem, p. 19.

Technical Problem Areas

2.

435

Efficient Allocation of Cases between the Members of the Network

According to the Explanatory Memorandum: The network will […] ensure an efficient allocation of cases based on the principle that cases should be dealt with by the best placed authority.7

Recital 16 of the draft Regulation confirms what is already apparent from the use of the term ‘best placed authority’ in the singular, namely that ‘the objective [is] that each case should be handled by a single authority’. A number of provisions of the draft Regulation are instrumental to achieve this objective of allocating each case to the single best-placed authority. First, Article. 36 requires the Member States to designate the competition authorities responsible for the application of Articles 81 and 82 EC, and to take the measures necessary to empower those authorities to apply those provisions. According to the Explanatory Memorandum, the full empowerment of the national competition authorities is the precondition for the proper functioning of the network of competition authorities. Without it, case allocation could not take place as envisaged, and the Commission might be forced to take a disproportionate share of cases concerning the markets of a Member State whose authority is unable to apply Articles 81 and 82.8

Second, Article 11(3) provides: Where a matter involving the application of Article 81 or Article 82 of the Treaty is referred to the competition authorities of the Member States or where they act on their own initiative to apply those Articles, they shall inform the Commission accordingly at the outset of their proceedings.

According to the Explanatory Memorandum, ‘the information will be made accessible to all Member States authorities via the network’.9 Similarly, Article 11(2) provides that ‘the Commission shall forthwith transmit to the competition authorities of the Member States copies of the most important documents it has collected […]’. Article 12 further provides that ‘the Commission and the competition authorities of the Member States may provide one another with and use in evidence any matter of fact or of law, including confidential information’. As pointed out in the Explanatory Memorandum, ‘effective case allocation […] makes it necessary that the members of the network should inform each other of all new cases and exchange relevant case-related information’.10 Article 12 ‘also allows the transfer 7 Idem, 8 Idem, 9 Idem, 10 Idem,

p. 6. p. 30. pp. 19–20. p. 12.

436

Working Paper IX

of entire case files […], the objective being to render possible the transfer of a case from one authority to another in the interest of effective case allocation’.11 Thirdly, Article 13 of the draft Regulation empowers the national competition authorities and the Commission to suspend or terminate proceedings on the ground that another member of the network is, or has been, dealing with the same case. As stated in the Explanatory Memorandum, this provision ‘removes risks of duplication of work and incentives for multiple complaints’.12 This provision empowers the national competition authorities and the Commission to suspend or terminate duplicative proceedings, but it does not oblige them to do so. According to the Explanatory Memorandum, it is ‘neither necessary nor appropriate to oblige other competition authorities to suspend or terminate their proceedings. It is the task of the network to ensure in practice that resources are used efficiently’.13 Indeed, in the absence of a rule determining which authority is the single best placed to deal with the case, a mandatory rule would not make sense, as only those authorities which are not the best placed should suspend or terminate their proceedings. Fourthly, Article 11(6) provides that the national competition authorities are automatically relieved of their competence if the Commission initiates its own proceedings. As pointed out in the Explanatory Memorandum, this provision ‘serves to ensure efficient case allocation […]’.14 Whereas all these provisions are instrumental to achieve the objective of allocating each case to the single best placed authority, the draft Regulation does not contain any rules or criteria identifying which authority is the best placed to deal with a case. The Explanatory Memorandum points out, however, that ‘the functioning of the network will be further elaborated upon in a notice on cooperation between competition authorities’.15

3.

Assistance between Members of the Network in Respect of Fact-Finding

Article 21(2) of the draft Regulation provides that ‘at the request of the Commission, the competition authorities of the Member States shall undertake the inspections which the Commission [has decided]’. Article 21(1) reads as follows: The competition authority of a Member State may in its own territory carry out any fact-finding measure under its national law on behalf and for the account of the competition authority of another Member State in order to establish whether there has been an infringement of Article 81 or Article 82 of the Treaty. It shall transmit the information collected to the requesting authority in accordance with Article 12 of the Regulation. 11 Idem, 12 Idem, 13 Idem, 14 Idem, 15 Idem,

p. 21. p. 21. pp. 12–13. p. 21. p. 12.

Technical Problem Areas

437

According to the Explanatory Memorandum, such cooperation enables national competition authorities to deal with cases where some evidence is to be found in other Member States. Without such mechanisms a real decentralisation of the application of Community competition rules would be seriously hampered.16

As already indicated above, Article 12 allows more generally for the transfer of information between the members of the network, and its subsequent use as evidence by the receiving authority.

4.

Coordination to Ensure the Consistent Application of Articles 81 and 82 EC

According to the Explanatory Memorandum: Maintaining consistent application is essential in an enforcement system in which parallel powers to apply Articles 81 and 82 are exercised by the Commission, national competition authorities and national courts. If significant differences in the application of these provisions were to develop the consistency of Community competition law and the proper functioning of the single market would be put at risk. It is therefore necessary to adopt measures addressing the danger of inconsistent application effectively.17

A number of provisions in the draft Regulation allow coordination within the network so as to ensure consistent application of Articles 81 and 82 EC by all its members, by providing for (4.1) exchange of information between the members of the network, (4.2) consultation within the network and (4.3) supervision by the Commission over the competition authorities of the Member States.

4.1.

Exchange of Information between the Members of the Network

As already indicated above, Article 11 (2) and (3) and Article 12 of the draft Regulation provide for information to be exchanged between the members of the network.

4.2.

Consultations within the Network

Article 11(4) and (5) and Article 14 of the draft Regulation provide for consultations within the network. 16 Idem, 17 Idem,

p. 26. p. 23.

438

Working Paper IX

As to cases dealt with by the Commission, Article 14 provides for the consultation of the Advisory Committee on Restrictive Practices and Dominant Positions, composed of representatives of the competition authorities of the Member States, prior to the taking of a final decision by the Commission. As to cases dealt with by competition authorities of the Member States, Article 11(4) reads as follows: Where the competition authorities of Member States intend to adopt a decision under Article 81 or Article 82 of the Treaty requiring that an infringement be brought to an end, accepting commitments or withdrawing the benefit of a block exemption regulation, they shall first consult the Commission. For that purpose, they shall no later than one month before adopting the decision provide the Commission with a summary of the case and with copies of the most important documents drawn up in the course of their own proceedings. At the Commission’s request, they shall provide it with a copy of any other document relating to the case.

Furthermore, Article 11(5) states that ‘the competition authorities of the Member States may consult the Commission on any other case involving the application of Community law’. According to the Explanatory Memorandum, ‘the Commission will associate the other members of the network in the consultation process’.18 Article 14(6) further provides that, ‘acting on its own initiative or at the request of a Member State, the Commission may include a case being dealt with by the competition authority of a Member State on the agenda of the Advisory Committee for discussion before the final decision is adopted’. According to the Explanatory Memorandum, the purpose of this provision ‘is to allow the Committee to serve as a forum for discussion of all cases that may be of common interest, in particular cases raising issues of consistent application of Articles 81 and 82. If requested by a Member State, the Commission will normally put a national case on the agenda’.19

4.3.

The Commission’s Powers of Control

Article 11(6) of the draft Regulation provides that the initiation by the Commission of proceedings for the adoption of a decision under this Regulation shall relieve the competition authorities of the Member States of their competence to apply Articles 81 and 82 of the Treaty.

As explained in the Explanatory Memorandum, this provision serves to ensure consistent application of Articles 81 and 82 EC: ‘In case of substantial 18 Idem, 19 Idem,

p. 12. p. 22.

Technical Problem Areas

439

disagreement within the network, the Commission retains the power to withdraw a case from a national competition authority by itself initiating proceedings in the case’.20 Article 16 further provides: In accordance with Article 10 of the Treaty and the principle of the uniform application of Community law, national courts and the competition authorities of the Member States shall use every effort to avoid any decision that conflicts with decisions adopted by the Commission.

As pointed out in the Explanatory Memorandum, ‘national competition authorities can avoid taking conflicting decisions by consulting the Commission and — in cases where a Commission decision is pending before the Community courts — by suspending their own proceedings’.21

III. 1.

The European Convention on Human Rights Relevant Provisions

In the context of competition law enforcement, two sets of provisions in the European Convention on Human Rights appear relevant: (1.1) the right to a fair trial and the other guarantees protecting defendants in criminal proceedings, and (1.2) the protection of private life and of property.

1.1.

a.

Right to a Fair Trial and other Guarantees Protecting Defendants in Criminal Proceedings Right to a Fair Trial (Article 6 ECHR)

The first sentence of Article 6(1) ECHR reads as follows: In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.

The remaining provisions of Article 6 ECHR list some further guarantees protecting defendants in criminal proceedings, as does Article 7 ECHR.

20 Idem, 21 Idem,

page 10; see also page 21. page 24.

440

b.

Working Paper IX

Right not to be Tried or Punished Twice (Article 4 Protocol No. 7)

Article 4 of Protocol No. 7 to the Convention states: 1.

No one shall be liable to be tried or punished again in criminal proceedings under the jurisdiction of the same State for an offence for which he has already been finally acquitted or convicted in accordance with the law and penal procedure of that State. 2. The provisions of the preceding paragraph shall not prevent the reopening of the case in accordance with the law and penal procedure of the State concerned, if there is evidence of new or newly discovered facts, or if there has been a fundamental defect in the previous proceedings, which could affect the outcome of the case.

1.2. a.

Protection of Private Life and of Property Right to Respect for Private Life (Article 8 ECHR)

Article 8 ECHR reads as follows: 1.

Everyone has the right to respect for his private and family life, his home and his correspondence. 2. There shall be no interference by a public authority with the exercise of this right except such as is in accordance with the law and is necessary in a democratic society in the interests of national security, public safety or the economic well-being of the country, for the prevention of disorder or crime, for the protection of health or morals, or for the protection of the rights and freedoms of others.

b.

Protection of Property (Article 1 Protocol No. 1)

Article 1 of Protocol No. 1 to the Convention provides: Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law. The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties.

2.

Legal Status

Neither the European Community nor the European Union are parties to the ECHR, which is not as such part of Community law. According to the case-law

Technical Problem Areas

441

of the European Court of Justice, however, fundamental rights are part of the general principles of Community law, and the EC courts thus ensure respect of these rights on the basis of Article 220 EC.22 In defining and applying fundamental rights, the EC courts draw from the constitutional traditions common to the Member States and from international treaties on which Member States have collaborated and to which they are signatories. In that respect the EC courts have accorded particular significance to the ECHR.23 The core of this case-law is enshrined in what is now Article 6(2) EU: ‘The Union shall respect fundamental rights, as guaranteed by the European Convention for the Protection of Human Rights and Fundamental Freedoms signed in Rome on 4 November 1950 and as they result from the constitutional traditions common to the Member States, as general principles of Community law’. By virtue of Article 46(d) EU, the ‘powers’ of the European Court of Justice with respect to Article 6(2) EU are confined to ‘action of the institutions, insofar as the Court has jurisdiction under the Treaties establishing the European Communities and under this Treaty’.

IV. 1.

The Charter of Fundamental Rights Introduction: The CRF and its Relationship with the ECHR

The Charter of Fundamental Rights of the EU was solemnly proclaimed by the European Parliament, the Council and the Commission on 7 December 2000.24 According to its preamble, this Charter reaffirms […] the rights as they result, in particular, from the constitutional traditions and international obligations common to the Member States, the Treaty on European Union, the Community Treaties, the European Convention for the Protection of Human Rights and Fundamental Freedoms, the Social Charters adopted by the Community and by the Council of Europe and the case-law of the Court of Justice of the European Communities and the European Court of Human Rights. 22 Case 29/69 Stauder v. Stadt Ulm [1969] ECR 419, para 7. 23 See for example Case C–309/96 Annibaldi [1997] ECR I–7493,

Case C–185/95 P Baustahlgewebe [1998] ECR I–8417 and Case T–112/98 Mannesmannröhren–Werke [2001] ECR II–732. All 15 Member States of the EU are parties to the Convention as amended by its Protocol No. 11, as well as to Protocol No. 1 thereto. However, only nine of the 15 Member States have ratified Protocol No. 7 (Austria, Denmark, Finland, France, Greece, Ireland, Italy, Luxembourg and Sweden; Germany, the Netherlands, Portugal and Spain have signed the Protocol, without ratification; Belgium and the United Kingdom have not signed; see http://conventions.coe.int). Insofar as this situation might have raised doubts as to the status as general principle of Community law of the right not to be tried or punished twice as laid down in Article 4 of Protocol No. 7, this doubt has been removed by the proclamation of the CFR, given that Article 50 of the latter contains the corresponding right. 24 See supra note no. 3; see also Lenaerts K. and De Smijter E. (2001): ‘A “Bill of Rights” for the European Union’, 38 Common Market Law Review 2, pp. 273–300.

442

Working Paper IX

As to the scope of the Charter, its Article 51(1) provides: The provisions of this Charter are addressed to the institutions and bodies of the Union with due regard for the principle of subsidiarity and to the Member States only when they are implementing Union law. They shall therefore respect the rights, observe the principles and promote the application thereof in accordance with their respective powers.

Article 52(3) CFR provides: In so far as this Charter contains rights which correspond to rights guaranteed by the Convention for the Protection of Human Rights and Fundamental Freedoms, the meaning and scope of those rights shall be the same as those laid down by the said Convention. This provision shall not prevent Union law providing more extensive protection.

It follows that, if a fundamental right enshrined in the Charter corresponds to a right guaranteed by the Convention, the latter serves as a minimum in determining the scope and meaning of the right in question. The Charter, which is part of EU law, may however provide more extensive protection.25

2.

Relevant Provisions

2.1.

a.

Right to a Fair Trial and other Guarantees Protecting Defendants in Criminal Proceedingsa Right to a Fair Trial (Article 47 CFR)

The first sentence of the second alinea of Article 47 CFR reads as follows: ‘Everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.’ According to the Explanatory Memorandum provided by the Secretariat of the body which drafted the Charter,26 this provision corresponds to Article 6(1) ECHR: […] In Community law, the right to a fair hearing is not confined to disputes relating to civil law rights and obligations. That is one of the consequences of the fact that the Community is a community based on the rule of law […]. Nevertheless, in all respects other than their scope, the guarantees afforded by the ECHR apply in a similar way in the Union.27 25 Lenaerts and De Smijter (2001), supra note no. 24, at p. 293. 26 Council of the EU, Charter of Fundamental Rights of the European

Union — Explanations relating to the complete text of the Charter (December 2000), available at http://ue.eu.int/df/docs/en/ EN_2001_1023.pdf. This memorandum is expressed to ‘have no legal value’ and is ‘simply intended to clarify the provisions of the Charter’. 27 Idem, pp. 65–66.

Technical Problem Areas

b.

443

Right not to be Tried or Punished Twice (Article 50 CFR)

Article 50 CFR reads as follows: No one shall be liable to be tried or punished again in criminal proceedings for an offence for which he or she has already been finally acquitted or convicted within the Union in accordance with the law.

According to the Explanatory Memorandum, this provision ‘corresponds to Article 4 of Protocol No 7 to the ECHR, but its scope is extended to European Union level between the Courts of the Member States’.28

2.2. a.

Protection of Private Life and of Property Respect for Private Life (Article 7 CFR)

Article 7 CFR provides: ‘Everyone has the right to respect for his or her private and family life, home and communications.’ According to the Explanatory Memorandum, ‘the rights guaranteed in Article 7 correspond to those guaranteed by Article 8 of the ECHR. To take account of developments in technology the word “correspondence” has been replaced by “communications’’’.29 b.

Right to Property (Article 17 CFR)

Article 17 CFR reads as follows: 1. Everyone has the right to own, use, dispose of and bequeath his or her lawfully acquired possessions. No one may be deprived of his or her possessions, except in the public interest and in the cases and under the conditions provided for by law, subject to fair compensation being paid in good time for their loss. The use of property may be regulated by law in so far as is necessary for the general interest. 2. Intellectual property shall be protected.

According to the Explanatory Memorandum, ‘Article 17 corresponds to Article 1 of the Protocol to the ECHR’.30

2.3.

Limitations (Article 52 CFR)

Article 52(1) CFR reads as follows: Any limitation on the exercise of the rights and freedoms recognised by this Charter must be provided for by law and respect the essence of those rights and freedoms. 28 Idem, 29 Idem, 30 Idem,

p. 76. p. 25. p. 75; see also p. 35.

444

Working Paper IX

Subject to the principle of proportionality, limitations may be made only if they are necessary and genuinely meet objectives of general interest recognised by the Union or the need to protect the rights and freedoms of others.

As already indicated above, Article 52(3) CFR provides that, if a fundamental right enshrined in the Charter corresponds to a right guaranteed by the Convention, the latter serves as a minimum in determining the scope and meaning of the right in question.

3.

Legal Status

The inter-governmental conference which was meeting at the time when the CFR was drafted could have decided to insert the CFR into the Treaties on which the Union is founded or to add a reference to it in Article 6(2) EU. Instead, the CFR was merely the subject of a solemn proclamation by the European Parliament, the Council and the Commission. As Lenaerts and De Smijter (2001) have pointed out In practice, however, the legal effect of the solemn proclamation of the Charter of Fundamental Rights of the European Union will tend to be similar to that of insertion into the Treaties on which the Union is founded. Indeed, to the extent that the charter is to be regarded as an expression of the constitutional traditions common to the Member States, the Court will be required to enforce it by virtue of Article 6(2) juncto Article 46(d) EU ‘as general principles of Community law’. The holistic interpretation given by the Court to the term ‘common constitutional traditions’, the composition and the functioning of the Convention which drafted the Charter as well as the unanimous acceptance of the text of the Charter by the three EU institutions with legislative powers, may reasonably lead to the conclusion that the Charter is to be regarded as an emanation of those common constitutional traditions, in the substantive sense of the term ‘constitutional traditions’. The Charter is thus part of the acquis communautaire, even if it is not part yet of the Treaties on which the Union is founded.31

V.

Possible Problems Related to Case Allocation

As to the allocation of cases between the members of the network, two possible problems can be identified: (1) possible differences in applicability of 31 Supra

note no. 24, at pp. 298–99. See also para 27 of the Opinion of Advocate–General Tizzano of 8 February 2001 in Case C–173/99 BECTU, para 126 of the Opinion of Advocate–General Mischo of 20 September 2001 in Joined Cases C–20/00 and C–64/00 Booker, and para 48 of the Judgment of the Court of First Instance of 30 January 2002 in Case T–54/99 max.mobil.

Technical Problem Areas

445

the ECHR or the CFR to the different members of the network and (2) double jeopardy.

1.

Differences in Applicability of the ECHR or the CFR

Under the draft Regulation, a case will be allocated either to the competition authority of a Member State or to the Commission, depending on which authority is the best placed to deal with the case. Given that all fifteen Member States are parties to the ECHR,32 whereas the European Community and the European Union are not, the allocation of the case to a national authority or to the Commission will determine whether or not the ECHR applies to the treatment of the case. It could be argued that the Member States would be in breach of their obligations under the ECHR whenever they allow a case to be transferred from their national competition authorities, which are bound by the ECHR, to the Commission, which is not. This problem does however not appear to be a real one. Under Article 1 ECHR, the Member States ‘shall secure to everyone within their jurisdiction the rights and freedoms defined in [the] Convention’. This obligation is fulfilled as long as an equivalent level of protection of fundamental rights as required under the ECHR is also guaranteed when the case is dealt with by the Commission.33 As explained above,34 the case-law of the European Court of Justice, and the provisions of the EU Treaty, provide for an equivalent protection via the general principles of Community law. As to the CFR, according to Article 51(1) thereof, its provisions are addressed not only to the institutions of the Union, such as the Commission, but also ‘to the Member States […] when they are implementing Union law’. The Charter thus applies equally when a case is dealt with under Articles 81 or 82 EC by the Commission or by the competition authority of a Member State.

2.

Double Jeopardy

As explained above, the objective of the draft Regulation is that each case should be handled by a single authority.35 If this objective were not achieved and some 32 See supra note no. 23 as to the Protocols to the Convention. 33 See Decision of the European Commission of Human Rights

of 9 February 1990, M & Co v. Germany, Case 13258/87, Y.ECHR, 1990, at 46; see also Judgment of the European Court of Human Rights of 18 February 1999, Matthews v. United Kingdom, Case 24833/94, and the currently pending Case 56672/00, DSR Senator Lines v. 15 Member States of the European Union. 34 Supra note no. 23. 35 Supra note no. 7.

446

Working Paper IX

cases were dealt with by several authorities within the network, problems could arise under the ne bis in idem principle.

2.1.

The Current Situation Under Regulation 17/62

Under Regulation 17/62 the competition authorities of the Member States have only rarely applied Articles 81 and 82 EC.36 Instead they normally apply their national competition laws. a.

The Case-Law of the European Court of Justice

In Wilhelm v. Bundeskartellamt, the European Court of Justice held that the possibility of one and the same agreement being the object of two sets of parallel proceedings, one before the Commission under Article 81 EC, the other before the national authorities under national law, is not contrary to the ne bis in idem principle, given the differences between the two laws, notably because Community law focuses on the effect on trade between Member States.37 The Court did however add that ‘if […] the possibility of two procedures being conducted separately were to lead to the imposition of consecutive sanctions, a general requirement of natural justice, such as that expressed at the end of the second paragraph of Article 90 of the ECSC Treaty, demands that any previous punitive decision must be taken into account in determining any sanction which is to be imposed’.38 b.

The ECHR

The judgment in Wilhelm v. Bundeskartellamt does not appear to be incompatible with Article 4 of Protocol No. 7 to the ECHR. It has to be recognised that the European Court of Human Rights has ruled that Article 4 of Protocol No. 7 is breached not only when a person is tried or punished twice for nominally the same offence, but also when he or she is prosecuted twice for two offences the essential elements of which overlap.39 Articles 81 and 82 EC, on the one hand, 36 As

to the reasons for this lack of application, see Marenco G. (1994): ‘The uneasy enforcement of Article 85 EEC as between Community and national levels’, in Hawk B., ed., Annual Proceedings of the Fordham Corporate Law Institute 1993, Kluwer, p. 605; Von Stoephasius H.–P. (1993): ‘Enforcement of EC competition law by national authorities’, in Slot P. J. and McDonnell A., eds., Procedure and Enforcement in EC and US Competition Law, London, Sweet & Maxwell, p. 32; ‘Editorial Comments: Subsidiarity in EC competition law enforcement’, in 32 Common Market Law Review 1 (1995), and Wils W. P.J. (2002): The Optimal Enforcement of EC Antitrust Law: Essays in Law and Economics, London, Kluwer Law International, pp. 140–41. 37 Case 14/68 Wilhelm v. Bundeskartellamt [1969] ECR 3, paras 3 and 11. 38 Idem, para 11. 39 Judgment of the European Court of Human Rights of 29 August 2001, Fischer v. Austria, Case 37950/97, discussing earlier judgments of 23 October 1995, Gradinger v. Austria and of 30 July 1998, Oliveira v. Swizerland. In the United States, the Supreme Court has similarly held in Blockburger v. US, 284 US 299 (1932) that punishment for two statutory offences arising out of the same criminal act

Technical Problem Areas

447

and national competition laws, on the other hand, may very well constitute an example of such overlap. Indeed, whereas Articles 81 and 82 EC contain an element not embraced by national competition laws, namely the effect on trade between Member States, the essential elements of national competition laws would often appear to cover the same ground as Articles 81 and 82 EC. Article 4 of Protocol No. 7 to the ECHR, however, only applies to double prosecution ‘under the jurisdiction of the same State’, a condition which does not appear to be fulfilled in a situation where one prosecution happens in a Member State and the other before the European institutions.40 c.

The CFR

The situation under the CFR appears very different. Whereas Article 4 of Protocol No. 7 ECHR only applies to double prosecution ‘under the jurisdiction of the same State’, Article 50 CFR prohibits double prosecution ‘within the Union’. It thus also covers the situation of parallel prosecutions by the Commission and a national competition authority, or by the competition authorities of several Member States.

2.2.

The Situation Under the Draft Regulation

Article 3 of the draft Regulation provides: Where an agreement, a decision by an association of undertakings or a concerted practice within the meaning of Article 81 of the Treaty or the abuse of a dominant position within the meaning of Article 82 may affect trade between Member States, Community competition law shall apply to the exclusion of national competition law.

This means that under the draft Regulation the competition authorities of the Member States will be obliged to apply Articles 81 and 82 EC to any case falling under these provisions, and will no longer be allowed to apply their national competition law. Whenever the same case is dealt with by the Commission and a national competition authority, or by the authorities of several Member States, these authorities will thus be applying nominally the same law, and the applicability of the ne bis in idem principle becomes all the more obvious. As is currently the case under Regulation 17/62, multiple proceedings by the Commission and a national competition authority, or by the authorities of several or transaction does not violate the double jeopardy clause in the 5th Amendment to the US Constitution only if ‘each provision requires proof of an additional fact which the other does not’; see also more recently US v. Dixon, 113 S.Ct. 2849, 2856 (1993). 40 Similarly,

in the United States, the courts have held that the state and federal governments are separate sovereigns and that successive prosecutions based on the same underlying conduct do not violate the double jeopardy clause if the prosecutions are brought by separate sovereigns; see US v. Lanza, 260 US 377, 382 (1922) and more recently US v. Koon, 34 F.3d 1416, 1438 (9th Cir.’94).

448

Working Paper IX

Member States, would not be contrary to Article 4 of Protocol No. 7 ECHR, which only applies to double prosecution ‘under the jurisdiction of the same State’, but there would a problem under Article 50 CFR, which prohibits double prosecution ‘within the Union’.

2.3

The Application of Articles 81 and 82 EC is of a Criminal Law Nature

The above conclusion that multiple proceedings under Articles 81 or 82 EC by the Commission and a national competition authority, or by the authorities of several Member States, concerning the same agreement or practice would raise a problem under Article 50 CFR is dependent on such proceedings being qualified as ‘criminal’. Indeed, Article 50 CFR only contains a right not to be tried or punished twice ‘in criminal proceedings’. The Charter does not contain any definition of what is ‘criminal’. Given the close relationship between the Charter and the Convention, the obvious approach would be to give ‘criminal’ in the Charter the same meaning as this term has in the Convention.41 Article 22(5) of the draft Regulation provides, like Article 15(4) of Regulation 17/62, that decisions by which the Commission imposes fines on undertakings pursuant to that regulation ‘shall not be of a criminal law nature’. However, this provision is not decisive in determining whether proceedings based on that Regulation are of a criminal law nature within the meaning of the Convention. Indeed, according to the case-law of the European Court of Human Rights, the indications furnished by domestic law as to the criminal nature of the offence ‘have only a relative value’, the term ‘criminal’ within the meaning of Article 6 ECHR being ‘autonomous’.42 For Article 6 ECHR to apply by virtue of the words ‘criminal charge’, ‘it suffices that the offence in question should by its nature be ‘criminal’ from the point of view of the Convention’, because it relates to ‘a general rule, whose purpose is both deterrent and punitive’, ‘or should have made the person concerned liable to a sanction which, in its nature and degree of severity, belongs in general to the ‘criminal’ sphere’.43 As I have argued in detail elsewhere,44 it appears difficult, if 41 In

the ECHR, the term appears in several Articles. The case-law of the European Court of Human Rights on the concept of ‘criminal’ focuses on Article 6 ECHR, but there are no reasons to assume that the concept would not be exactly the same in other provisions of the ECHR and its Protocols. Gradinger v. Austria and Fischer v. Austria (supra note no. 38) appear to confirm that the term ‘criminal’ has the same meaning in Article 4 of Protocol No. 7 ECHR as in Article 6 ECHR, as one of the two Austrian proceedings in both of these cases was of an administrative nature according to Austrian law, but nevertheless considered ‘criminal’ for the purposes of Article 4 of Protocol No. 7 ECHR. 42 Öztürk v. Germany, judgment of 21 February 1984, Series A no. 73, paras. 52 and 50. 43 Lutz v. Germany judgments of 25 August 1987, Series A no. 123, para 55, and Bendenoun v. France, judgment of 24 February 1994, Series A no. 284, para 47. 44 Wils W. P. J. (1996): ‘La compatibilité des procédures communautaires en matière de concurrence avec la Convention européenne des droits de l’homme’, 32 Cahiers de droit européen 329.

Technical Problem Areas

449

not impossible, to deny that the application of the criteria set out in the case-law of the European Court of Human Rights leads to the conclusion that proceedings based on Regulation 17/62 or the draft Regulation, leading or possibly leading to decisions in which the Commission finds a violation of Articles 81 or 82 EC, orders its termination and imposes fines relate to ‘the determination of a criminal charge’ within the meaning of Article 6 ECHR.45 The same would hold true for national proceedings leading to or possibly leading to findings of violations of Articles 81 and 82 EC and imposing penalties for such violations.

2.4.

a.

Some Further Issues Concerning the Scope of the Prohibition on Multiple Prosecutions and Punishments Parallel Proceedings are Temporarily Possible

According to the terms of Article 50 CFR, ‘no one shall be liable to be tried or punished again in criminal proceedings for an offence for which he or she has already been finally acquitted or convicted within the Union in, accordance with the law’. The prohibition of double prosecution and punishment thus does not exclude the possibility of parallel proceedings being conducted by several members of the network of competition authorities at the same time, as long as none of these proceedings has already reached the stage of final acquittal or conviction. This means that at a preliminary stage, when it may not yet be clear which member of the network is the best placed to deal with the case, several competition authorities may investigate the same agreement or practice. However, at the latest at the moment when the first proceeding is closed by a final acquittal or conviction, the other authorities have to discontinue their proceedings. b.

The Types of Decision Precluding a Second Prosecution or Punishment

Article 50 CFR precludes further prosecution or punishment for an offence for which the person concerned ‘has already been finally acquitted or convicted within the Union in accordance with the law’. This raises the question as to what types of decision qualify as ‘final acquittal or conviction’. • Prohibition and fining decisions There can be no doubt that decisions pursuant to Article 7 of the draft Regulation, by which the Commission may find infringements of Articles 81 or 82 EC, and 45 See

also Lenaerts K. and Vanhamme J. (1997): ‘Procedural rights of private parties in the Community administrative process’, 34 Common Market Law Review , 531 at 557, Opinion of Judge Vesterdorf acting as Advocate–General in Case T–1/89 Rhône–Poulenc v. Commission [1991] ECR II–867 at 885, Opinion of Advocate–General Léger in Case C–185/95 P Baustahlgewebe v. Commission [1998] ECR I–8422 para 31, and Judgment of the Court of Justice of 8 July 1999 in Case C–199/92 P Hüls v. Commission [1999] ECR I–4383 paras 149–150.

450

Working Paper IX

decisions pursuant to Article 22(2)(a) of the draft Regulation, by which the Commission may impose fines for such infringements, constitute ‘convictions’ within the meaning of Article 50 CFR. If the addressees of these decisions bring an application for annulment before the Community courts, these decisions will however only be ‘final’ when confirmed by the latter. Similarly, decisions of competition authorities of the Member States requiring that infringements of Articles 81 or 82 EC be brought to an end or imposing fines, pursuant to Article 5 of the draft Regulation, constitute ‘convictions’. Again, they would only become ‘final’ at the end of possible applications for judicial review or appeals. • Non-infringement decisions Equally, there can be no doubt that decisions taken pursuant to Article 10 of the draft Regulation, under which the Commission may find that Articles 81 or 82 EC are not applicable to an agreement or practice, constitute ‘acquittals’ within the meaning of Article 50 CFR. As to the competition authorities of the Member States, Article 5 of the draft Regulation provides that ‘where on the basis of the information in their possession the conditions for prohibition are not met they may likewise decide that there are no grounds for action on their part’. Such a decision can probably not be qualified as a ‘final acquittal’ within the meaning of Article 50 CFR. • Consent decisions Article 9 of the draft Regulation reads as follows: 1.

Where the Commission intends to adopt a decision requiring that an infringement be brought to an end and the undertakings concerned offer commitments such as to meet the Commission’s objections, the Commission may by decision make those commitments binding on the undertakings. Such a decision shall be adopted for a specified period. 2. Irrespective of whether or not there has been or still is an infringement of Article 81 or Article 82 of the Treaty, such a decision shall terminate the proceedings. 3. The Commission may reopen the proceedings: (a) where there has been a material change in any of the facts on which the decision was based; (b) where the undertakings concerned act contrary to their commitments; or (c) where the decision was based on incomplete, incorrect or misleading information.

On the one hand, the second paragraph expressly provides that the question of whether there has been or still is an infringement is left open. This would suggest that a consent decision is not to be regarded as an acquittal or conviction. On the other hand, the same paragraph makes it clear that the decision ‘shall terminate the proceedings’, subject only to the possibility of reopening proceedings

Technical Problem Areas

451

pursuant to the third paragraph. This rather suggests that a decision pursuant to Article 9 should preclude further prosecutions or punishments by other authorities.46 c.

Deducting the First Penalty from a Second One will not Suffice

Article 50 CFR clearly establishes the right for any person not ‘to be tried or punished again’ in criminal proceedings for an offence for which he or she has already been finally acquitted or convicted. A first acquittal or conviction thus precludes any subsequent proceedings. Thus it does not suffice to deduct from the penalty imposed in the second proceeding any penalty imposed in the first proceeding. The second proceedings are precluded as such. This has also been held by the European Court of Human Rights with regard to Article 4 of Protocol No. 7 ECHR in Fischer v. Austria: The Court is not convinced by the Government’s argument that the case was resolved due to the reduction of the applicant’s prison term by one month, being equivalent to the fine paid in the administrative proceedings. The reduction of the prison term […] cannot alter the above finding that the applicant was tried twice for essentially the same offence, and the fact that both his convictions stand.47

d.

Limitations Under Article 52 CFR

Article 4 of Protocol No. 7 to the ECHR does not provide for any possible derogation or limitation to the right not to be tried or punished twice. By extending this right to situations where the prosecutions take place in different jurisdictions within the Union, Article 50 CFR extends the scope of Article 4 of Protocol No. 7 ECHR. It could thus be argued that limitations would still be possible under Article 52 CFR in multi-jurisdictional cases. On the other hand, it could be argued that, by extending Article 4 of Protocol No. 7 ECHR to multi-jurisdictional cases, Article 52 CFR has extended the scope of application but not altered the substance of the Article 4 of Protocol No. 7 ECHR, including the impossibility of derogations. In any event, it is difficult to see what objectives of general interest that do not apply within a single jurisdiction could justify limitations to the right not to be tried or punished twice in multi-jurisdictional cases, particularly in the context of the network of competition authorities, as envisaged by the draft Regulation, given the close cooperation which is to characterise the functioning of this network.

46 In Cases C–187/01 Hüseyin Gözütok and C–385/01 Klaus Brügge (currently pending) the European

Court of Justice will be called upon to decide a similar question, namely whether the ne bis in idem principle as laid down in Article 54 of the Convention of 19 June 1990 applying the Schengen Agreement precludes a second prosecution where the first prosecution has been terminated in the form of a settlement by the public prosecutor before the case is brought to court. 47 Supra note no. 39, para 30.

452

Working Paper IX

VII.

Possible Problems Related to Assistance in Fact-Finding

Two problems may arise as to the assistance between members of the network in respect of fact-finding: (1) the extension of the possible use of evidence to members of the network other than the one collecting the evidence and (2) different levels of protection of the rights of defence in the jurisdictions concerned.

1.

Extension of the Possible use of Evidence to Members of the Network Other than the One Collecting the Evidence

1.1.

The Current Situation Under Regulation 17/62

Article 20(1) of Regulation 17/62 provides that information acquired as a result of requests for information or verifications pursuant to this Regulation ‘shall be used only for the purpose of the relevant request or investigation’. In Spanish Banks the European Court of Justice has made it clear that in consequence the competition authorities of the Member States, which may receive this information pursuant to Article 10 of Regulation 17/62, which provides for the involvement of the Member States in the Commission’s proceedings, may not use themselves this information in evidence in their own proceedings applying Community or national competition law.48 As to information collected by national competition authorities, whether or not this information can be transferred to and subsequently used in evidence by the competition authorities of other Member States, depends on the provisions of the applicable national law. The Commission could always request the national authority pursuant to Article 11 of Regulation 17/62 to hand over the information, and subsequently use it in evidence.

1.2.

The Situation Under the Draft Regulation

Article 12 of the draft Regulation provides that ‘the Commission and the competition authorities of the Member States may provide one another with and use in evidence any matter of fact or of law’. Article 21 further provides for the possibility for one member of the network to carry out fact-finding measures at the request and on behalf of another member of the network. Each member of the network will thus be able to use in evidence information collected by any member of the network.

48 Case

C–67/91 Asociación Española de Banca Privada and Others v. Commission [1992] ECR I–4820. National authorities may however use the information as intelligence, prompting their own investigations.

Technical Problem Areas

1.3.

453

Assessment Under the Charter and the Convention

Restrictions on the possible use of evidence collected by a competition authority, such as the restriction laid down in Article 20(1) of Regulation 17/62 as interpreted by the European Court of Justice, could be considered as responding to two kinds of concern relating to fundamental rights.49 First, to the extent that the investigative measures used interfere with privacy, use restrictions serve to prevent unnecessary or abusive interference, in that they preclude the investigating authority from using its investigative powers for purposes unrelated to those for which the investigative powers have been granted. Secondly, use restrictions could also contribute to guaranteeing a fair trial in that the undertakings or persons being investigated know, when subjected to investigative measures, by whom and for what type of offences they are being prosecuted. Restrictions on the possible use of evidence collected by a competition authority could thus be valuable with respect both to the right to respect for private life, as guaranteed by Article 8 ECHR and Article 7 CFR, and to the right to a fair trial, as guaranteed by Article 6 ECHR and Article 47 CFR. It would not appear, however, that the extension to all members of the network of the possible use in evidence of information collected by any member of the network would lead to a breach of any of those fundamental rights. Indeed, as long as the possible use of the information collected remains limited to the application of Articles 81 and 82 EC — as foreseen by Articles 12(2) and 27(1) of the draft Regulation, the guarantee against the use of investigative powers for purposes unrelated to those for which the investigative powers have been granted remains in place. As to the second concern underlying use restrictions, the extension does not appear problematic either. It is true that the undertakings or persons being investigated, when subjected to investigative measures, face somewhat more uncertainty as to the question by whom they are being prosecuted, in that the prosecuting authority may later turn out to be another member of the network than the one investigating initially, but this uncertainty would appear to remain within acceptable limits, given the closed number of members of the network. As an additional guarantee, it could be considered to provide that, when members of the network make requests for information and carry out investigations, they would indicate to the undertakings or persons being investigated the possibility of the information collected being transferred and subsequently used by other members of the network. It could also be provided that the undertakings or persons concerned would be informed when information is transferred to another member of the network for use in evidence. 49 Compare with paras 36 and 37 of the Spanish Banks judgment, supra note no. 48, with points 206 to 208 of the Opinion of Advocate–General Mischo in Joined Cases 46/87 and 227/88 Hoechst v. Commission [1989] ECR 2909 and with point 317 of the Opinion of Advocate–General Mischo of 25 October 2001 in Case C–244/99 P DSM v. Commission (not yet published).

454

2.

2.1.

Working Paper IX

Different Levels of Protection of the Rights of Defence in the Jurisdictions Concerned The Problem

Community law, which governs the collection and use of evidence by the Commission, and the fifteen national laws, which — in the absence of applicable Community law rules — govern the collection and use of evidence by the competition authorities of the Member States, may differ with respect to the type of evidence which can be legally collected, or the way in which it can be collected, or with respect to what evidence can be adduced. When the members of the network assist each other in collecting evidence, it could happen that one member of the network, either on its own initiative or at the second competition authority’s request, collects, in accordance with the law governing the investigating powers of the first authority, evidence which the second authority could not have lawfully collected under its own law, and hands over this evidence to the second authority so as to allow it to use this evidence. For instance, the Commission could lawfully collect and subsequently transmit to the Belgian competition authority information which under Belgian law would be protected by in-house counsel legal privilege, a privilege not recognised in Community law. A second example would be where the Commission collects evidence through a request for information to which the undertaking concerned is obliged to respond under Community law,50 and transmits this information for use in evidence to the German competition authority, which might not have been able to collect this information itself because of the right not to incriminate oneself as laid down in Paragraph 136(1) of the German code of criminal procedure.51 To analyse this problem, it may be helpful to draw a distinction according to the source of the difference in protection levels in the two jurisdictions concerned, namely whether the difference reflects a requirement to respect fundamental rights as guaranteed in the ECHR or the CFR, or whether the difference is due to a national law or Community law providing for a level of protection exceeding the requirements of the ECHR and the CFR.

2.2.

Where the Difference Reflects a Requirement to Respect Fundamental Rights as Guaranteed in the ECHR or the CFR

As explained above, the CFR equally applies to the Commission and to the competition authorities of the Member States when the latter are implementing

50 See judgment of the Court of First Instance in Case T–112/98 Mannesmannröheren–Werke v. Commission [2001] ECR II–732. 51 See idem, paras 80–81.

Technical Problem Areas

455

Community competition law. The fundamental rights laid down in the Convention also have to be respected by the competition authorities of the Member States, since they are parties to the ECHR, as well as by the Commission, as general principles of Community law. The ECHR and the CFR may nevertheless impose different requirements on different members of the network, reflecting relevant differences between the various national laws and Community law. For instance, whereas Community law only provides for fines on undertakings for violations of Articles 81 or 82 EC, national law may provide for individual decision-makers within those companies being sentenced to prison. It appears that Article 6 ECHR grants natural persons faced with the risk of a prison sentence a broad right of silence, allowing them to refuse to answer requests for information.52 At present, there is no case-law of the European Court of Human Rights indicating that Article 6 ECHR would also imply a similarly broad right of silence for legal persons faced with the risk of being fined. In this example, a problem under the Convention would arise if the competition authority of a Member State in which prison sanctions lie for violation of Articles 81 or 82 EC were to use in evidence information collected by another member of the network which itself could have obtained and used this evidence without violating Article 6 ECHR given that under its law no prison sanctions lie. The rather obvious solution to this problem is that the competition authority of the Member State in which prison sanctions lie should abstain from using this evidence, or preferably even avoid receiving it. Article 12(2) of the draft Regulation, which provides that ‘only financial penalties may be imposed on the basis of information provided’ by one member of the network to another under Article 12(1) of the draft Regulation, is clearly instrumental in achieving this solution.

2.3.

Where the Difference is Due to a National Law or Community Law Providing for a Level of Protection Exceeding the Requirements of the ECHR and CFR

A national law or Community law may provide for some protection not required by the ECHR or the CFR. For instance, legal privilege for in-house counsel provided for in some national laws such as Belgian law does not appear to correspond to any right guaranteed by the ECHR or the CFR. It logically follows that the ECHR and the CFR do not stand in the way of the Belgian competition authority using in evidence information which it could not have 52 See

Funke v. France judgment of the European Court of Human Rights of 25 February 1993, Series A no. 256–A, John Murray v. United Kingdom judgment of 8 February 1996, Reports of Judgments and Decisions 1996–I, p. 49, Saunders v. United Kingdom, judgment of 17 December 1996, Reports 1996–VI, p. 2064, Servès v. France, judgment of 20 October 1997, Reports 1997–VI, p. 2173, and J.B. v. Switzerland, judgment of 3 May 2001, Case 31827/96.

456

Working Paper IX

collected itself under Belgian law because of the in-house counsel legal privilege but which it obtained under the draft Regulation from the Commission or from another member of the network under whose national law a similar privilege is not recognised.

VII. Possible Problems Related to Coordination to Ensure Consistency Three possible problems may arise in relation to the coordination within the network: (1) different levels of protection of the confidentiality of information exchanged, (2) the consultations within the network and the right to adversarial proceedings and (3) the Commission’s powers of control and the independence of national competition authorities.

1.

Different Levels of Protection of the Confidentiality of Information Exchanged

When information is exchanged within the network, the confidentiality of this information may be protected to varying degrees by the different members of the network. As to the protection of business secrets and other confidential information, two sets of provisions of the ECHR and CFR may come into play. First, the right to respect for private life and for correspondence or communications, as guaranteed by Article 8 ECHR and Article 7 CFR, could be relevant. Indeed, according to the case-law of the European Court of Human Rights, ‘private life is a broad term not susceptible to exhaustive definition. […] It may include activities of a professional or business nature’.53 Secondly, business secrets may benefit from protection under the right to property, as guaranteed by Article 1 Protocol 1 ECHR and Article 17 CFR. According to the case-law of the European Court of Human Rights, ‘the concept of ‘possessions’ in Article 1 of Protocol No. 1 has an autonomous meaning which is certainly not limited to ownership of physical goods: certain other rights and interests constituting assets can also be regarded as ‘property rights’, and thus as ‘possessions’ for the purposes of this provision’.54

53 Judgment of 25 September 2001, P.G. and J.H. v. United Kingdom, Case 44787/98, para 56, referring to the Niemitz v. Germany judgment of 16 December 1992, Series A no. 251–B, para 29, and the Halford v. United Kingdom judgment of 25 June 1997, Reports 1997–III, para 44. 54 Gasus Dosier– und Fördertechnik GmbH v. the Netherlands judgment of 23 February 1995, Series A no. 306–B, para 53 and Iatridis v. Greece judgment of 25 March 1999, Case 31107/96, para 54; see

Technical Problem Areas

457

Article 27(2) of the draft Regulation provides that, without prejudice to the provisions on cooperation within the network and with national courts and to the provisions on the hearing of parties, complainants and other interested parties, ‘the Commission and the competition authorities of the Member States, their officials and other servants shall not disclose information acquired or exchanged by them pursuant to this Regulation and of the kind covered by the obligation of professional secrecy’. As pointed out in the Explanatory Memorandum, this provision ‘makes the obligation of professional secrecy laid down in Article 20(2) of the existing Regulation 17/62 applicable to all confidential information exchanged by the national competition authorities under the proposed Regulation’.55 There are no reasons to believe that this level of protection, which will apply as a uniform minimum standard to all members of the network, would be insufficient in the light of the ECHR and CFR.56 The possibility that some national law may provide an even higher level of protection does not raise any problem under the ECHR or CFR.

2.

2.1.

Consultations within the Network and the Right to Adversarial Proceedings

The Problem

Under the draft Regulation, when a competition authority intends to adopt a decision finding an infringement of Articles 81 or 82 EC, ordering its termination or imposing penalties, it first has to consult other members of the network. According to Articles 11(4) and 14 of the draft Regulation, the Commission is required to consult the Advisory Committee, composed of representatives of the competition authorities of the Member States, whereas the competition authorities of the Member States are obliged to consult the Commission, which will associate the other members of the network in the consultation process.

also Tre Traktörer AB v. Sweden judgment of 7 July 1989, Series A no. 159, para 53 and Fredin v. Sweden judgment of 18 February 1991, Series A no. 192, para 40. Article 17 (2) CFR explicitly protects intellectual property, which could arguably include business secrets. 55 Supra note 5, p 28. 56 According to the case–law

of the European Court of Human Rights, neither the right to respect for private life and correspondence nor the right to property are absolute. A fair balance must be struck between the demands of the general interest and the requirements of the protection of the individual’s fundamental rights; see Sporrong and Lönnroth v. Sweden judgment of 23 September 1982, Series A no. 52, para 69, Brumarescu v. Romania judgment of 28 October 1999, Case 28342/95, and Z v. Finland judgment of 25 February 1997, Reports 1997–I, para 99.

458

Working Paper IX

The question arises whether the undertakings or persons concerned should be entitled to have knowledge of and comment on the opinions given by other members of the network.57

2.2.

The Case-Law of the European Court of Human Rights

According to the case-law of the European Court of Human Rights, the right to a fair trial, as guaranteed by Article 6 ECHR, includes the fundamental right that proceedings should be adversarial.58 In several cases concerning the role of the Advocate-General or similar officers at the Court of Cassation or Supreme Court in Belgium, Portugal, the Netherlands and France,59 and at the Conseil d’Etat in France,60 the European Court of Human Rights has held that this right to adversarial proceedings ‘means in principle the opportunity for the parties to a criminal or civil trial to have knowledge of and comment on all evidence adduced or observations filed, even by an independent member of the national legal service, with a view to influencing the court’s decision’.61 In reaching the conclusion that parties should be entitled to have knowledge of and comment on the opinion of the Advocate-General or similar officer, the European Court of Human Rights has regard to what is at stake for the person being tried and to the nature of the opinion. For instance, in the case concerning the procedures before the Portuguese Supreme Court, the following relevant elements were identified: [T]he duty of the Attorney-General’s department at the Supreme Court is mainly to assist the court and to help ensure that its case-law is consistent. […], the department’s intervention in the proceedings was more particularly justified for the purposes of upholding the public interest. […] […] great importance must be attached to the part actually played in the proceedings by the member of the Attorney General’s department, and more particularly to the content and effects of his observations. These contain an opinion which derives its

57 Article

26(2), second alinea, of the draft Regulation appears to exclude such a right, at least in the context of access to the file in proceedings conducted by the Commission: ‘The right of access to the file shall not extend to confidential information and internal documents of the Commission or the Member States. In particular, any correspondence between the Commission and the competition authorities of the Member States, or between the latter, inter alia, documents drawn up pursuant to Articles 11 and 14 are excluded.’ 58 Ruiz–Mateos v. Spain, judgment of 23 June 1993, Series A no. 262, para 63. 59 Borgers v. Belgium, 30 October 1991, Series A no. 214–B; Vermeulen v. Belgium and Lobo Machado v. Portugal, 20 February 1996, Reports 1996–I; Van Orshoven v. Belgium, 25 June 1997, Reports 1997–III; J.J. v. the Netherlands and K.D.B. v. the Netherlands, 27 March 1998, Reports 1998–II; and Reinhardt and Slimane–Kaïd v. France, 31 March 1998, Reports 1998–II. 60 Kress v. France, 7 June 2001, Case 39594/98. 61 Lobo Machado v. Portugal, supra note no. 59, para 31.

Technical Problem Areas

459

authority from that of the Attorney General’s department itself. Although it is objective and reasoned in law, the opinion is nevertheless intended to advise and accordingly influence the Supreme Court.62

Opinions of the Commission (in association with the network) to a national competition authority pursuant to Article 11(4) of the draft Regulation and opinions of the Advisory Committee pursuant to Article 14 of the draft Regulation would appear to be comparable in nature to the observations of the Advocate-Generals or similar officers considered by the European Court of Human Rights. Before jumping to the conclusion that Article 6 ECHR and Article 47 CFR thus appear to give the undertakings or persons concerned the right to have knowledge of and comment on the opinions pursuant to Articles 11(4) and 14 of the draft Regulation, we should however draw a distinction between the members of the network that constitute and those that do not constitute ‘independent and impartial tribunals’ within the meaning of Article 6 ECHR and Article 47 CFR.

2.3.

Administrative Authorities Versus Independent and Impartial Tribunals

Article 6(1) ECHR provides that ‘in the determination of […] any criminal charge against him, everyone is entitled to a fair and public hearing […] by an independent and impartial tribunal […]’. The European Court of Human Rights has however ruled that, for reasons of efficiency, the prosecution and punishment of offences which are ‘criminal’ within the wider meaning of Article 6 ECHR can be entrusted to administrative authorities which do not or not entirely meet the requirements of independence and impartiality, provided that the persons concerned are enabled to take any decision thus made before a judicial body that has full jurisdiction and does provide the full guarantees of Article 6(1) ECHR.63 Since it combines the functions of prosecutor and judge when applying Articles 81 and 82 EC, the Commission can manifestly not be regarded as an independent and impartial tribunal, but its decisions can be challenged before the Court of First Instance, which provides the full guarantees of Article 6(1) ECHR and which undertakes a comprehensive review of the Commission’s decisions.64 62 Idem, paras 28 and 29. 63 Öztürk v. Germany, judgment

of 21 February 1984, Series A no. 73, para 56, and Bendenoun v. France judgment of 24 February 1994, Series A no. 284, para 46. This possibility does however not exist for ‘serious charges classified as “criminal” under both domestic and Convetion law’, which should be dealt with already in first instance by a tribunal fully meeting the requirements of Article 6(1) ECHR: De Cubber v. Belgium, judgment of 26 October 1984, Series A no. 86, paras 31–32 and Findlay v. United Kingdom, judgment of 25 February 1997, Reports 1997–I, para 79. 64 Wils W. P. J., supra note no. 44, at pp. 337–338, and Lenaerts K. and Vanhamme K., supra note no. 45, at 559–562.

460

Working Paper IX

The same applies to those national competition authorities, such as the German Bundeskartellamt or the Italian Autorità garante della Concorrenza e del Mercato, which combine the functions of prosecutor and judge, but whose decisions can be challenged before a court fulfilling the requirements of Article 6(1) ECHR. In some other Member States, decisions finding violations of Articles 81 or 82 EC or imposing penalties for such violations are taken by independent and impartial tribunals. For instance, in Sweden, the Stockholm City Court is the body which has the power to impose fines for violations of Articles 81 and 82 EC, in proceedings brought by the Competition Authority. In Ireland, the national law provisions equivalent to Articles 81 and 82 EC are enforced through criminal procedures. The Competition Authority refers the case to the Director of Public Prosecutions, which decides whether or not to take proceedings before the criminal courts.

2.4.

Administrative Authorities

In the case of the Commission and of those national competition authorities which are not independent and impartial tribunals within the meaning of Article 6(1) ECHR but whose decisions are subject to full review by a judicial body meeting all requirements of the said provision, it can be argued that the right to adversarial proceedings, and the resulting right to have knowledge of and comment on all observations filed with a view to influencing the court’s decision, does not apply at the level of the decision by the competition authority, but only at the level of the court reviewing the decision. The fact that, in proceedings before the Commission, undertakings do not have the right to comment on the opinion of the Advisory Committee is thus not incompatible with Article 6 ECHR and Article 47 CFR. For the same reason, national competition authorities such as those in Germany and Italy would thus not be obliged under Article 6 ECHR and Article 47 CFR to give the undertakings concerned the possibility to comment upon the opinion of the Commission received pursuant to the consultation under Article 11(4) of the draft Regulation.

2.5.

Independent and Impartial Tribunals

Where decisions are taken by independent and impartial tribunals, such as the Stockholm City Court or the Irish criminal courts, it would appear to follow from the case-law of the European Court of Human Rights on Article 6 ECHR discussed above65 that, if such tribunals were to receive observations from the 65 Supra

text accompanying notes no. 56 to 60.

Technical Problem Areas

461

Commission (or from other members of the network) pursuant to Article 11 of the draft Regulation, the undertakings or persons being prosecuted should be entitled to have knowledge of and to comment on these observations. There are several possible means of ensuring that the consultations within the network do not lead to violations of the right to adversarial proceedings as guaranteed by Article 6 ECHR and Article 47 CFR: The first possibility is to exclude independent and impartial tribunals from the obligation of consultation provided for in Article 11(4) of the draft Regulation. This could be done either through a specific exclusion added to this provision, or more generally by specifying in the draft Regulation that the notion of ‘competition authorities of the Member States’ in that regulation does not include independent and impartial tribunals. At present, the draft Regulation does not define the concept of ‘competition authorities of the Member States’. In the absence of a definition in the regulation, this expression is likely to be interpreted in the same way as the term ‘authorities of the Member States’ in Article 88 EC and in Article 9(3) of Regulation 17/62, which, according to the case-law of the European Court of Justice, appears to include tribunals such as the Stockholm City Court or the Irish criminal courts when deciding on prosecutions under Articles 81 or 82 EC.66 The second possibility is for the independent and impartial tribunals which receive observations from the Commission (or from other members of the network) following consultation pursuant to Article 11(4) of the draft Regulation to provide the undertakings or persons being prosecuted the opportunity to comment on these observations.

3.

The Commission’s Powers of Control and the Independence of National Competition Authorities

It has been argued, notably by Fourgoux,67 that the obligation for the competition authorities to consult the Commission provided for in Article 11(4) of the draft Regulation together with the Commission’s power pursuant to Article 11(6) of the draft Regulation to withdraw the case from the national competition authority removes the independence of national competition authorities and thus raises a problem under Article 6 ECHR.68 66 Case 127/73 BRT v. SABAM [1974] ECR 51, para 19 and Joined Cases 209 to 213/84 Ministère public v. Asjes [1986] ECR 1425, para 55: ‘the term “authorities in the Member States” in Article 88 refers to either the administrative authorities entrusted, in most Member States, with the task of applying domestic legislation on competition subject to the review of legality carried out by the competent courts, or else the courts to which, in other Member States, that task has been especially entrusted’. 67 Fourgoux J. C. (2001): ‘Un nouvel antitrust européen: espoirs et craintes’, conference held at the Tribunal de Commerce de Paris, 22 October 2001. 68 Questions have also been raised concerning Article 16 of the draft Regulation, which obliges the competition authorities of the Member States to ‘use every effort to avoid any decision that conflicts

462

Working Paper IX

According to the case-law of the European Court of Human Rights: Only an institution that has full jurisdiction and satisfies a number of requirements, such as independence of the executive and also of the parties, merits the designation “tribunal” within the meaning of [Article 6(1) ECHR]’.69

In Beaumartin v. France, for instance, the European Court of Human Rights found that the French Conseil d’Etat did not meet these requirements under a practice, discontinued in 1990, which ‘meant that, when the administrative court encountered serious difficulties in interpreting an international treaty, it was obliged to request the Minister for Foreign Affairs to clarify the meaning of the impugned provision and it then had to abide by his interpretation in all circumstances’.70 When applying this case-law to the members of the EU network of competition authorities, a distinction should again be made between those authorities that do and those that do not qualify as independent and impartial tribunals within the meaning of Article 6(1) ECHR. In the case of those national competition authorities, such as those in Germany and Italy, which are not independent and impartial tribunals but whose decisions are subject to full review by a judicial body meeting all the requirements of Article 6(1) ECHR, the requirements of full jurisdiction and independence only apply at the level of the reviewing court. The application of Article 11 of the draft Regulation to national competition authorities of the German and Italian type is thus unproblematic. It could however be argued that the application of Article 11 of the draft Regulation to independent and impartial tribunals, such as the Stockholm City Court or the Irish criminal courts, is problematic under Article 6 ECHR and Article 47 CFR. Again, the problem can be avoided either by specifically excluding independent and impartial tribunals from the provisions of Article 11 of the draft Regulation, or by specifying in the draft Regulation that the concept of ‘competition authorities of the Member States’ in that regulation does not include independent and impartial tribunals. This should not substantially alter the functioning of the network, as the provisions of Article 11 of the draft

with decisions adopted by the Commission’. As I have argued elsewhere, this provision does not appear contrary to fundamental principles of judicial independence in that it only obliges the national bodies to give deference to the Commission’s decisions without being bound; see Wils W. P. J. (2001): ‘The modernization of the enforcement of Articles 81 and 82 EC: A legal and economic analysis of the Commission’s proposal for a new Council Regulation replacing Regulation No. 17’, 24 Fordham International Law Journal 1655, at pp. 1710–1711. 69 Beaumartin

v. France, judgment of 24 November 1994, Series A no. 296–B, para 38, referring to Ringei v. Austria judgment of 16 July 1971, Series A no. 13, para 9, Le Compte, Van Leuven and De Meyere v. Belgium, judgment of 23 June 1981, Series A no. 43, para 55, and Belilos v. Switzerland, judgment of 29 April 1988, Series A no. 132, para 64. 70 Idem, para 38.

Technical Problem Areas

463

Regulation would in any event still be applicable to the authorities investigating possible violations Articles 81 and 82 EC, such as the Competition Authority in Sweden and Ireland.

VIII.

Conclusion

Provided that the stated objective that each case should be dealt with by a single authority is achieved, and provided that it is made clear that the concept of ‘competition authorities of the Member States’ does not include independent and impartial tribunals within the meaning of Article 6 ECHR and Article 47 CFR, or that such tribunals are otherwise excluded from Article 11 of the draft Regulation, the structure and functioning of the EU network of competition authorities as envisaged in the draft Regulation does not appear, in the current state of the caselaw of the European Court of Human Rights, incompatible with the requirements of the ECHR and of the CFR.

X Denis Waelbroeck* ‘Twelve Feet All Dangling Down and Six Necks Exceeding Long’: The EU Network of Competition Authorities and the European Convention on Human Rights and Fundamental Freedoms

Against her there is no defence; flight is the bravest way. For if thou tarry to do on thine armour by the cliff, I fear lest once again she sally forth and catch at thee with so many heads, and seize as many men as before. So drive past with all thy force. Homer, Odyssey, Book XII

Who will not remember — with a shudder — mythology’s many-headed beasts: the Hecatoncheires, the Lernean Hydra, or Odysseus’ Scylla? Scylla, it was said, had once been a beautiful maiden, but now: she hath twelve feet all dangling down; six necks exceeding long, and on each a hideous head, (…) and there she fishes swooping round the rock, for dolphins or seadogs, or whatso greater beast she may anywhere take. Homer, Odyssey, Book XII

Is there a surer way — the conceivers of the ‘modernisation’ project must have thought — of fishing for the modern ‘beasts’ (whose names are cartels, abuses of dominant position, and so on) than relying on a modern Scylla, whose heads cannot be chopped off without two more growing in their place? By relying on the multi-headed forces of 15 (soon to be 25, and even 27) national competition authorities plus the Commission, cartels — the modernisation proponents must have thought — will soon belong to the past. And obviously nobody, including us, will contest the accuracy of the analysis, nor the appropriateness of this project, at least as far as the fight against cartels is concerned. However, it is clear that this ‘close cooperation’,1 particularly with respect to fact-finding, allocation of cases,2 and so on, is useful to wage a war against cartels, but it also raises number of delicate issues. For if ‘there is no defence’ against

* Avocat, Professor of EC competition law at the Free University Brussels and the College of Europe in Bruges. 1 See Article 11(1) of the draft Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty and amending Regulations (EEC) No 1017/68, (EEC) No 298/74 and (EEC) No 3975 (‘Regulation implementing Articles 81 and 82 of the Treaty), COM/2000/0582 final — OJ C 365/284 E [2000]. 2 See Explanatory Memorandum of the draft Regulation, at p. 6.

466

Working Paper X

Scylla, is there not a risk that the same can be said of the new ‘EC network of competition authorities’? In particular, one of the main problems of this new network will lie in the parallel application of the EC competition provisions by sixteen, and up to twenty-six, different competition authorities, which: • may become active in parallel, make inconsistent decisions and apply multiple sanctions (Section I below); • follow different procedural rules, and thus apply different protection standards (Section II); and moreover, • are not all bound by the European Convention on Human Rights (Section III).

I. Parallel Application of Competition Rules by Multiple Competition Authorities and the Risk of Inconsistent Decisions and Double Jeopardy In recent years, the emergence of national antitrust agencies throughout the Community, all applying largely similar rules, has often lead to contradictory outcomes in terms of enforcement, such as: • a given behaviour being condemned by one authority — sometimes even by imposing criminal sanctions — while cleared by another authority;3 • a given behaviour being condemned by one authority while exempted by another;4 or • the same behaviour being condemned in different jurisdictions and thereby being imposed multiple sanctions.

3 In

one case, in which I was involved, the French competition authority imposed in 1996 on an American company (Autodesk) a fine for a selective distribution system that had been previously notified to the EC Commission and which shortly thereafter benefited from a comfort letter from the Commission. In Visa International, the non-discrimination rule was first condemned by the Swedish and Dutch Authority and then exempted by the Commission. Now the Belgian authority attacks Banksys — the main debit card company in Belgium — for a non-discrimination rule. 4 In its recent Michelin decision of 20 June 2001 (case COMP/E-2/36041), the Commission imposed a 19,76 million @ fine on Michelin for a rebate system which had previously been reviewed and cleared by the French competition authority. And vice versa, behaviour exempted by the Commission has sometimes been condemned by national authorities (see e.g. the British rules on beer supply agreements, which are stricter than those of the Commission). Another example can be found in the differing interpretation of the vertical block exemption regulation as regards the prohibition of resale price maintenance. In the petrol station business, several companies provide for a guaranteed margin system for the retailers. In a decision of 8 June 2000, the Italian competition authority found this system to be blacklisted. (In the words of the Italian authority, the modalities of determination of the margins stimulate the resellers not to undertake an independent policy on prices.) The Spanish competition authority came to the opposite conclusion: in a decision of 23 February 2000 (File R 348/98) it took the view that this system was in the interest of petrol stations. More recently, the Dutch authority found that the system was illegal. According to its decision, since all oil companies use systems for supporting filling station owners in almost identical ways, filling station owners are not given an incentive to charge prices below their suppliers’ national recommended

Technical Problem Areas

467

Moreover, there have not only been conflicts between national authorities, but also between national courts and the Commission.5 All these problems obviously arose because no serious binding mechanism has been established in the past with a view to ensuring consistency of decision-making.6 In the draft Regulation, the Commission itself recognises that ‘maintaining consistent application is essential in an enforcement system in which parallel powers to apply Articles 81 and 82 are exercised by the Commission, national competition authorities and national courts’.7 However, the question is whether the rules contained in the draft Regulation are sufficient to avoid such inconsistencies in the absence of stringent case allocation criteria.8 The rules indeed appear to be minimal. Thus, Article 11 of the draft Regulation merely requires national competition authorities to ‘consult the Commission prior to the adoption of prohibition decisions, decisions accepting commitments and decisions withdrawing the benefit of a block exemption regulation’, and repeats the current rule — contained in Article 9(3) of Regulation 17/62 — that ‘in case of substantial disagreement within the network, the Commission retains the power to withdraw a case from a national competition authority by itself initiating proceedings in the case.’9 (Nevertheless, there is no definition of what should be considered to be a ‘substantial disagreement’.10) prices. As a result, the Director General of the Dutch authority notified its intention to render the Block Exemption inoperative in this case (Press release 01-46, The Hague, 18 December 2001). 5 See e.g. the Irish Ice Cream case (Case C–344/98 Masterfoods Ltd v. HB Ice Cream Ltd., [2000] ECR I–1214), or the IMS Health case (Case T–184/2001 R), [2001] ECR II–0000. 6 So far reference can only be made to the loose mechanism envisaged in the Commission Notice on the cooperation between the national competition authorities and the Commission in handling cases falling under the scope of Articles 85 and 86 of the Treaty; OJ C 313/3 [1997]. It should in this respect also be emphasised that national competition authorities are normally mere administrative bodies who cannot avail themselves of Article 234 EC. 7 See Explanatory Memorandum, p. 23. 8 The draft Regulation does not contain rules on case allocation. In the framework of the proposal, the Commission prepared a draft paper dealing with the cooperation within the network of competition authorities, and establishing rules for allocating cases among the national competition authorities. This paper considers three options: (i) cases that can be dealt by a single authority with the assistance of other authorities for gathering information; (ii) cases that can be dealt by parallel action of national authorities where the action of one single authority would not be sufficient to bring the infringement to an end, and with the need to cooperate closely in order to avoid the risk of inconsistency; and (iii) cases that are best dealt at the Community level, where the practice or the agreement has appreciable effect on consumers in more than three Member States. Proper allocation of cases (and the need to impose on national competition authorities some self-restraint so as not to open ex officio procedures in cases where a national competition authority is already seized and has already started investigations) is, however, a sensitive issue. See also on this issue the Commission Notice on the cooperation between national competition authorities and the Commission in handling cases falling under the scope of Articles 85 and 86 of the Treaty, OJ C 313/3 [1997]. The Notice should be reviewed in the context of the decentralisation (see Explanatory Memorandum, p. 12). 9 See Explanatory Memorandum of the Proposal for a Council Regulation, p 9 Article 11(6) allows the Commission — as does the current Article 9(3) of Regulation 17/62 — to initiate proceedings on its own and thereby relieve the national authorities of their competence to apply Articles 81 and 82 EC. 10 In addition, it has not been considered ‘necessary for consistency purposes to provide for prior consultation in respect of other types of decisions adopted by national competition authorities, such as rejections of complaints and decisions to take no action. These decisions bind only the deciding authority, and do not preclude subsequent action by any other competition authority or before national courts.’ (see Explanatory Memorandum, p. 10).

468

Working Paper X

Moreover, Article 11(5) of the draft Regulation merely confirms the existing rule that national competition authorities may consult the Commission on any pending case (a rule that, to my knowledge, has so far failed to produce significant results). Finally, Article 14(6) of the draft Regulation stipulates that a case may be put on the agenda of the Advisory Committee for discussion, where necessary. As these various mechanisms show, the system is still largely based on flexibility without any stringent criteria concerning case allocation.11 Therefore, the system is not very different in practice from the one used, for example, in the context of the EC–US cooperation agreement. The risk of inconsistent decision-making (that is, of different authorities arriving at different conclusions when applying the same rule to the same behaviour and to the same companies) is therefore not negligible. If such situations arise — as they have in the past12 — this will, in my view, raise serious questions concerning both the principle of legality of sanctions contained in Article 7 ECHR and the non bis in idem principle.

1.

Legality of Sanctions

1.1.

Different Interpretations of the Same Rules

The principle of the legality of sanctions is contained in Article 7 ECHR,13 which not only prohibits the retrospective application of more severe sanctions, but also requires ‘predictability’ or ‘foreseeability’ (that is, offences should be clearly defined by the law).14 This condition is satisfied when an individual is in a position to know in advance how the courts will interpret the acts and/or omissions that can make him liable for infringing the law. Thus, according to the European Court of Human Rights, when speaking of law, Article 7 [ECHR] alludes to […] a concept which comprises statutory law as well as case-law and implies qualitative requirements, notably those of accessibility and foreseeability.15

11 See supra note no. 8. 12 See supra notes no. 5 and 6. 13 ‘1. No one shall be held guilty

of any criminal offence on account of any act or omission which did not constitute a criminal offence under national or international law at the time when it was committed. Nor shall a heavier penalty be imposed than the one that was applicable at the time the criminal offence was committed. 2. This article shall not prejudice the trial and punishment of any person for any act or omission which, at the time when it was committed, was committed, was criminal according to the general principles of law recognised by civilised nations’. 14 ECHR, Kokkinakis v. Greece case, judgment of 25 May 1993; Series A, No. 260–A, para 52. 15 ECHR, Cantoni v. France, judgment of 15 November 1996.

Technical Problem Areas

469

It should indeed be recalled that competition law does not comprise only straightforward rules concerning cartels, but also complex rules designed to strike the right balance between economic efficiency and workable competition, and to define delicate limits guiding companies in their daily business decisions. Any excessively ‘repressive’ view of the rules does not, in my view, sufficiently take into account the fact that competition law enforcement has more functions than merely prohibiting cartels, and that competition rules are often far from clear and can easily lead to conflicting decisions. In the new enforcement system, multiple authorities will be entrusted with the application of the same (EC) competition rules. One can only wonder how far it is likely to be compatible with the legality of sanctions requirement contained in Article 7 ECHR, that one national authority could make a decision concerning a fact previously examined by another authority under the same set of rules and thereby considered not to constitute an offence. For this reason, I believe that, in order to avoid conflicting decisions and legal uncertainty, it would have been desirable for the new Regulation to include some more stringent mechanism for ensuring that a single authority deals with each case. This is particularly so because, under the new enforcement system, companies will lose the possibility of notifying the authorities of agreements about whose legality they are uncertain. Moreover, they will lose the benefits of the ‘one-stop shop’ system under Article 81(3) EC, and will be left only with the possibility of obtaining ‘opinions’ from the Commission, which are not subject to judicial control.16 It remains to be determined how far a system whereby several authorities apply the same set of rules without any clear case allocation criteria and mutual recognition is compatible with the requirement of ‘foreseability’ of criminal law enforcement, as enshrined in Article 7 of the Convention.

1.2.

Different Sanctions for One and the Same Infringement

Another problem to be considered results from the different levels of sanctions applied in the different Member States. Indeed, while national competition

16 It is unclear why it was not possible to insert in the draft Regulation more stringent rules for ensuring

coherence in enforcement, as in other areas of EC law. For instance, in the pharmaceutical sector (but also for pesticides, veterinary products, biocides, etc) the Community also established an enforcement system built on a network of national medicinal agencies, whereby one authority in each case is appointed rapporteur, others and the Commission (or EMEA) are consulted, and the decision is subject to mutual recognition. One could also have conceived a system as the NAAG in the US, where actions are brought jointly by the various authorities before one single court (e.g. before one of the new jurisdictional chambers that can be set up after the Treaty of Nice). More modestly, one could have conceived some form of joint management of cases. All this would have been preferable to the lose consultation system currently envisaged, whereby companies cannot anticipate by which authority they will be sanctioned, and whether they will face one sanction applied by the Commission or — theoretically — even 25 sanctions by 25 national authorities.

470

Working Paper X

authorities are empowered to determine whether a certain form of behaviour is compatible with uniformly set EC competition rules, the draft Regulation leaves the Member States free to determine the consequences of any infringement. While the sanction will typically be a fine imposed on the infringing undertaking, the level of the fine varies greatly among the Member States.17 The legislation of some Member States even provides for the possibility of fining an individual who has participated in the infringement, or applying the sanction of imprisonment.18 Moreover, fines are tax-deductible in certain Member States, but not in others. Thus, the same facts, appraised under the same substantive rules, may give rise to completely different sanctions. Given that case allocation among national competition authorities will often be arbitrary in practice, in the absence of clear rules concerning case allocation it is unclear how far such a system is at all compatible with the fundamental rights of legality of the sanction and of equality of treatment.19

2.

The Non Bis in Idem Principle

Under the enforcement system governed by Regulation 17/62, the European Court of Justice accepted parallel application of EC competition law and national competition law to the same facts,20 as long as the national procedure did not jeopardise the full and uniform application of EC law.21 Article 3 of the draft Regulation establishes that EC competition law will be applied to the exclusion of national competition law whenever trade between Member States is affected. Consequently, the possibility of parallel application of EC and national rules to the same facts should be excluded (although there will obviously be scope for discussion about when trade between Member States is affected or not). 17 See

Zinsmeister U. and Lienemeyer M. (2002), infra note no. 29, and their overview of sanctions across different Member States. As an illustration, fines amount to up to 15% of a company’s turnover in Greece, whereas most of the other Member States limit the amount to 10% of the turnover (i.a. Spain, Ireland, the Netherlands, Finland, Belgium, Sweden, and recently France). The base of calculation for the fine may also vary: in Italy for example fines are calculated based on the part of the turnover deriving from the products/services forming the subject-matter of the anticompetitive agreement or conduct; in Belgium fines are calculated on the basis of sales on the national market added to exports. 18 See, for France, Article L 420–6 of Nouveau Code de Commerce, and for Ireland, Article 3 I Competition (Amendment) Act 1996. See also the recently-adopted Enterprise Bill in the United Kingdom, under which price fixing will become a criminal offence, giving rise to the possibility of applying the sanction of imprisonment. In several Member States, bid-rigging can give rise to imprisonment. 19 Articles 7 and 14 ECHR. 20 Case 14/68 Walt Wilhelm and others v. Bundeskartellamt [1969] ECR 1. 21 By virtue of the principle of primacy, the national competition authority should take ‘proper account’ of Commission decision already issued and ‘appropriate measures’ where national decision are to be taken before Commission decision and inconsistency may be feared.

Technical Problem Areas

471

This does not mean, however, that the same infringement could not be sanctioned twice, by two different authorities, for violation of the same EC rule. In this sense, Article 13 of the draft Regulation merely establishes that ‘the fact that one authority is dealing with the case shall be sufficient grounds for the others to suspend the proceedings before them’. In such circumstances, an authority ‘may’ thus decide not to act, but it is not obliged to do so.22 Whenever an authority decides not to act, it follows from the case-law of the European Court of Justice that: if the possibility of two procedures being conducted separately were to lead to the imposition of consecutive sanctions, a general requirement of natural justice (…) demands that any previous punitive decision must be taken into account in determining any sanction which is to be imposed.23

It remains to be seen what the European Court of Human Rights will decide about this issue in the light of Article 4 in Protocol 7 ECHR, which stipulates that: no one shall be liable to be tried or punished again in criminal proceedings under the jurisdiction of the same State for an offence for which he has already been finally acquitted or convicted in accordance with the law and penal procedure of that State.

In the words of the European Court of Human Rights, the aim of Article 4 in Protocol 7 ECHR ‘is to prohibit the repetition of criminal proceedings that have been concluded by a final decision’ and therefore, this article ‘is not confined to the right not to be punished twice but extends to the right not to be tried twice’.24 Under the case-law of the European Court of Human Rights, if different offences based on one act are prosecuted consecutively, one after the final decision of the other, the Court has to examine whether or not such offences have the same essential elements.25 Following this line of thought, the Court notes in the

22 See Explanatory Memorandum (pp. 12–13) according to which it is ‘neither necessary nor appropriate

to oblige other competition authorities to suspend or terminate their proceedings’. 23 See Walt Wilhelm, supra note no. 20, para 11. 24 See ECHR Franz Fischer v. Austria, judgment of 25 In the Gradinger v. Austria case (judgment of 23

29 May 2001, paras 22 and 29. October 1995, Series A no. 328–C) the European Court of Human Rights, while emphasising that the offences at issue differed in nature and aim, found a violation of Article 4 Protocol 7 to the ECHR, as both decisions were based on the same conduct (paras 54–55). In the Oliveira v. Switzerland case (ECHR judgment of 30 July 1998, Reports of Judgments and Decisions 1998–V) the Court found no violation of the same provision, considering that it presented a typical example of a single act constituting various offences (‘concours idéal d’infractions’) which did not infringe Article 4 Protocol 7 to the ECHR since that provision only prohibited people being tried twice for the same offence (para 26). This view is supported by the decision of the European Court of Human Rights in the case of Ponsetti and Chesnel v. France (nos. 36855/97 and 41731/98 ECHR 1999-VI, [14.9.99]), relating to separate convictions for two tax offences arising out of the failure to submit a tax declaration, where the respondent Government also argued that this was an example of one act constituting more than one offence. Nevertheless, the Court examined whether the offences in question differed in their essential elements.

472

Working Paper X

Franz Fischer case that ‘the question whether or not the non bis in idem principle is violated concerns the relationship between the different [two] offences at issue and can, therefore, not depend on the order in which the respective proceedings are conducted’.26 Each sanction will in theory concern an infringement in the territory of the State in which the sanction is imposed. No authority has the legal power — unless the EC Treaty is changed — to impose criminal sanctions for behaviour that occurs outside its territory, but it should be stressed that all authorities will penalise the infringement of the same basic rule (EC Article 81 or 82). Moreover, in many Member States, the fine is the same as defined in Article 15 of Regulation 17 (that is, 10 per cent of worldwide turnover: see, for example,. Finland or Sweden, or even 15 per cent of worldwide turnover in the case of Greece). In providing for the — theoretical — possibility of ‘case allocation’, the draft Regulation implicitly recognises that one authority can take the lead and ‘act for others’; that is, one authority imposes sanctions ‘on behalf of’ other States. In other words, if several authorities impose parallel sanctions, there is a risk that this could effectively be regarded as a case of double jeopardy.27 The problem is even more acute with regard to Article 50 of the Charter of Fundamental Rights of the European Union. Indeed, whereas Article 1 of Protocol 7 to the ECHR only applies to double prosecution ‘under the jurisdiction of the same State’, Article 50 of the Charter prohibits double prosecution ‘within the Union’.

II. Parallel Application of Competition Rules by Multiple Authorities Following Different National Procedural Rules The second main problem related to the ‘network of competition authorities’ and fundamental rights concerns the parallel application of competition rules by multiple authorities that follow different national procedural rules. Indeed, while the substantive competition law is largely harmonised throughout the EC, procedural requirements remain exclusively regulated at the national level, and in practice often diverge widely among the Member States. As the Commission itself stated in the draft Regulation: The present proposal is based on the premise that national competition authorities will apply Article 81 and 82 in accordance with their respective national procedural rules. It is not necessary for the implementation of the reform to embark on a full-scale harmonisation of national procedural laws.28

26 ECHR, Franz Fischer v. Austria, judgment of 29 August 27 See Wouter Wils’ written contribution for this volume. 28 See draft Regulation.

2001, para 29.

Technical Problem Areas

473

The question that arises is to what extent these divergences may cause concerns, particularly vis-à-vis the requirements of the ECHR, and vis-à-vis the stricter standards that Member States may occasionally adopt to protect of fundamental rights. Given the numerous differences between the Member States’ legal and procedural systems,29 there is an obvious risk of conflicting decisions and double jeopardy situations, and also that the standards of legal protection be different, depending on the national competition authority to which the case will be (randomly) allocated.30 It therefore seems that decentralisation — in so far as it involves not one but 16 (or soon up to 27) authorities, which may all be adopting conflicting decisions and applying different standards of protection — must necessarily enhance the risks of conflict with the ECHR. These risks are also likely to appear at each and every stage of the procedure (investigation, request for information, statement of objections, access to file, hearing, decision, sanction, appeal, and so on). Without pretending to be exhaustive, I hereafter refer to some of these possible issues of conflict.

1.

Investigations by National Competition Authorities

Following decentralisation, the national competition authority deciding on a case will not necessarily be located in the same Member State as the authority that has been requested to make the inspection. Particularly where an infringement of competition rules occurs in several Member States but only one national competition authority is in charge of the case, the competent authority is entitled to request the assistance of other national competition authorities in order to proceed to the investigation.31 Despite Article 8(2) ECHR (and the requirement under that Article that inspections take place where, in accordance with the law, they are necessary and pursue a legitimate aim), not all Member States provide for the same level of legal protection when inspections are made by their competition authorities.32

29 See e.g.: Zinsmeister U. and Lienemeyer M. (2002): ‘Die Verfahrensrechtlichen Probleme bei der dezentralen Anwendung des europäischen Kartellrechts’, 52 Wirtschaft und Wettbewerb 4, pp. 331–39; Jones T. (2001): ‘Regulation 17: the Impact of the current application of Articles 81 and 82 by national competition authorities on the European Commission’s proposals for reform’, 22 European Competition Law Review 10, pp. 405–15. 30 See supra note no. 9. 31 See Article 21 of the draft Regulation. 32 As is well known, there is some divergence in the approach of the European Court of Justice and of the European Court of Human Rights as to the applicability of Article 8 ECHR to undertakings. The European Court of Human Rights interpreted Article 8 ECHR (protection of privacy) as extending to search in offices. See Case 72/1991/324/396 ECHR, judgment of 16 December 1992, Niemietz v. Germany, paras 27–33. Some years before, the European Court of Justice had taken the opposite view in Hoechst (Joined Cases 46/87 and 227/88, [1989] ECR 2859).

474

Working Paper X

There are differences as to, for example, whether the private homes of employees can be inspected or not, whether methods such as phone taps can be used or not, how much control can be exercised by a national court over the appropriateness of an inspection, and so on.33 Indeed, the very requirement of a prior judicial authorisation to proceed with an inspection is subject to various national rules.34 The intervention of a judge is necessary in some Member States,35 whereas an authorisation from the national competition authority or ministry of economic affairs is sufficient in others.36 Sometimes no authorisation is required at all.37 However, it is common to all Member States that prior judicial authorisation is required when inspections of employees’ private homes are ordered.38 It follows that, without a minimal degree of harmonisation, the standard of legal protection for the undertakings will differ between the Member States, thus giving rise to possible litigation as to the appropriate standard of protection under Article 8 ECHR. How far can an authority in one Member State rely on information obtained in another Member State that does not respect the same standard of protection? Can a Member State that has a higher standard of protection impair the efficiency of investigations that it is requested to make on behalf of another national competition authority?39 Could a company refuse an investigation because it does not satisfy the fundamental rights standard applied in the state ‘on behalf of’ which the inspection is made?

2.

Requests for Information by National Competition Authorities and the Right to Remain Silent

Difficulties may also be anticipated when requests for information are made by national competition authorities. For instance, the right to remain silent (that is, the ‘privilege against self-incrimination’) could be interpreted differently

33 See e.g. most recently in the Opinion of Advocate General Mischo in Case C–94/00 Roquette Frères SA v. DGCCRF, judgment of 21 September 2001. 34 See for a comparative analysis, the Opinion of Advocate General Mischo in Joined Cases 46/87 and 227/88 Hoechst [1989] ECR 2859. 35 See Germany, Ireland, Portugal, Sweden and Spain, U. Zinsmeister and Lienemeyer M. (2001), supra. note no. 29. 36 See Belgium and Italy, for example. 37 See in the Netherlands, Algemene Wet Bestuursrecht, Article 5:15. 38 See Ziensmeister U. and Lienemyer M. (2002) , supra note no. 29. As is well known, this latter possibility may arise in case of adoption of the proposal for a new Regulation 17/62, introducing the possibility to ‘enter homes of directors, managers and other members of staff of the undertakings concerned, in so far as it may be suspected that business records are being kept there’ (Article 20(2)(b) of the proposal). 39 See mutatis mutandis the difficulties faced by the Commission in the Hoechst case due to the high standards applied by Germany and the Court’s ruling that ‘the Member States are required to ensure that the Commission’s action is effective, while respecting the general principles [of Community law]’ (see Joined Cases 46/87 and 227/88 Hoechst v. Commission [1989] ECR 2859, point 33).

Technical Problem Areas

475

within different Member States. Could companies, as a result, object to national competition authorities relying on evidence obtained in other Member States in so far as the latter contravenes (what they perceive to be) the right standard of protection under the ECHR? Could companies rely on excessively strict standards to refuse to give any reply to requests for information? Are the Member States required — on the basis of the ‘principle of efficacy’ and the ‘principle of loyalty’ (Article 10 EC) — to remove stricter standards of protection of fundamental rights? The fact that differences of interpretation are likely is demonstrated by the different interpretation that the right to remain silent has received in the case-law of the European Court of Justice and that of the European Court of Human Rights. In the Saunders case,40 the European Court of Human Rights gave a broad interpretation to the principle when it held that: in any event, bearing in mind the concept of fairness in Article 6, the right not to incriminate oneself cannot reasonably be confined to statements of admission of wrongdoing or to remarks which are directly incriminating. Testimony obtained under compulsion which appears on its face to be of a non-incriminating nature — such as exculpatory remarks or mere information on questions of fact — may later be deployed in criminal proceedings in support of the prosecution case, for example to contradict or cast doubt upon other statements of the accused or evidence given by him during the trial or to otherwise undermine his credibility. Where the credibility of an accused must be assessed by a jury the use of such testimony may be especially harmful. It follows that what is of the essence in this context is the use to which evidence obtained under compulsion is put in the course of the criminal trial.41

The European Court of Justice took a narrower view: an undertaking in receipt of a request for information pursuant to Article 11(5) of Regulation No 17 can be recognised as having a right to silence only to the extent that it would be compelled to provide answers which might involve an admission on its part of the existence of an infringement which it is incumbent upon the Commission to prove.42

and: the mere fact of being obliged to answer purely factual questions put by the Commission and to comply with its requests for the production of documents already in existence cannot constitute a breach of the principle of respect for the rights of defence or impair the right to fair legal process. There is nothing to prevent the addressee of such questions or requests from showing, whether later during the administrative procedure or in proceedings before the Community courts, when exercising 40 ECHR, Saunders v. UK, 17 December 1996, Application 00019187/91. 41 See ECHR, Saunders case, as above, para 71. 42 Case T–112/98 Mannesmannröhren-Werke v. Commission [2001] ECR II–729,

para 67.

476

Working Paper X

his rights of defence, that the facts set out in his replies or the documents produced by him have a different meaning from that ascribed to them by the Commission.43

The differences of interpretation that may arise are also demonstrated, for instance, by the fact that, while the European Court of Human Rights found that the rights of defence and of non-self incrimination also extended to ‘mere information on questions of fact’,44 the European Court of Justice has refused to see any difficulty with a lenient policy that rewards companies that renounce their right of defence concerning issues of fact.45 It is clear therefore that the system of exchange of information contained in Article 12 of the draft Regulation can lead to serious discussion about the compatibility of its implementation in certain cases concerning fundamental rights. This will be particularly relevant if employees face prison sentences as a result of information that is released in infringement of fundamental rights in other Member States. This might explain why Article 12(2) of the draft Regulation provides that ‘only financial penalties may be imposed on the basis of information provided’. However, given the wide discretion of authorities in defining the sanction, there will never be any way to discover — when no financial penalty is imposed — whether this is the result of the information exchange system of Article 12 or of other information. So Article 12(2) does not remedy this problem, in my view.

3.

Confidentiality / Legal Professional Privilege

Another difficulty might emerge as a result of the Member States’ different approaches to confidentiality and the legal professional privilege. The correspondence between undertakings and their lawyers is not regarded to be confidential in all EU jurisdictions.46 The degree of protection appears to be more complete in the UK, Ireland and Belgium. In other Member States (and at the EC level), the legal privilege applies

43 See 44 See 45 See

CFI, Mannesmannröhren-Werke Case, as above, at para 78. ECHR, Saunders case, supra note no. 40. most recently ABB v. Commission, Case T–31/99, [2002] ECR II–1881, para 242, according to which ‘the Commission should have differentiated the reduction for cooperation to be granted to the applicant from the reduction granted to Løgstør Rør and to Tarco on the ground that the applicant, after receiving the statement of objections, no longer disputed the findings of fact or their interpretation by the Commission’. 46 For a detailed and comprehensive study of the respective powers of ‘outside’ and ‘in house’ lawyers across the Member States, see the comparative table made out by the ECLA, from which it results that discrepancies are serious, with some Member States (UK, Ireland, Denmark and Portugal) recognising legal privilege for both category of lawyers in all circumstances (before Court and antitrust authorities), some denying any legal privilege in all circumstances (Sweden and Luxemburg), and other Member States still having intermediary solutions. See http://www.ecla.org/gb/frmpaper.htm.

Technical Problem Areas

477

only where the correspondence emanates from an independent lawyer who is entitled to practice in one of the Member States.47 The question therefore arises as to which level of protection will prevail in the case of documents seized in a Member State in which legal privilege does not apply to the same standards as in another (and vice versa). When the information is able to be shared between the different Member States, it is indeed unclear to what extent information obtained legally in one Member State can be used in another Member State in which it would have been illegal to obtain it. Moreover, if an inspection is conducted jointly by EC Commission officials and officials of the competition authority of a Member State that recognises legal privilege over documents emanating from internal lawyers, which rule should prevail?48 According to the prevailing case-law of the European Court of Justice, the principle of confidentiality is not breached if the information received is used by the national authority as ‘circumstantial evidence that may, if necessary, be taken into account to justify initiation of national procedure’.49 Article 12 of the draft Regulation (on exchange of information) establishes that, for the purpose of applying Community competition law: Notwithstanding any national provision to the contrary, the Commission and the competition authorities of the Member States may provide one another with and use in evidence any matter of fact or of law, including confidential information.

Moreover, Article 27 of the draft Regulation (dealing with professional secrecy), states generally that: Without prejudice to Articles 12 and 15, information collected pursuant to Articles 17 to 21 (powers of investigations) shall be used only for the purpose for which it was acquired.

These provisions appear prima facie difficult to reconcile with the previous caselaw. They allow national competition authorities and the Commission to use any evidence collected for the purpose of applying Community competition law (including confidential information) to a case concerning a suspected infringement other than the one for which the information was gathered. The question is obviously whether European Court of Human Rights will accept the extensive use of information obtained by another national authority 47 See, for the position at EC level, Case 155/79 AM&S v. Commission [1982] ECR 1575. 48 If an incident is raised concerning such an issue during an inspection, the only solution

could be to propose to the investigation team to seal the said document until a decision on this matter is taken by the EC Commission/Court of First Instance or by the national competition authority/court. 49 Case C–67/91 Dirección General de Defensa de la Competencia v. Asociación Española de Banca Privada and others [1992] ECR I–4785, point 39. The European Court of Justice, however, also held that it contravenes the undertakings’ rights of defence to ‘rely on evidence against undertakings which was obtained during an investigation but was not related to the subject-matter or purpose thereof’. See Case 85/87 Dow Benelux v Commission [1989] ECR 3137, point 18.

478

Working Paper X

that is envisaged in the draft Regulation. Article 10(2) of the Convention indeed recognises the necessity, in a democratic society, of protection against disclosure of information received in confidence. In any case, it is clear that the sheer multiplication of the authorities that are entitled to receive the information greatly increases the risk that disclosures of confidential information to third parties will be made.50

4.

Access to the File

Another possible problem concerns access to the file. Indeed, the scope of the right of access to the file (and the interpretation of what is deemed to be confidential) is different across the Member States. Depending which the authority the case is allocated to, there may be very different standards of the rights of the defendant to access the file. If different authorities are involved simultaneously, can documents obtained from one authority be used in a procedure before another authority? Information will circulate freely within the network. Mightn’t parties or third parties apply for access to certain documents in one Member State concerning proceedings taking place in another Member State, and to which access would have been denied in the investigating country? All this raises complex issues about confidentiality and also about the proper standard of rights of the defendant (access to the file being a necessary part of a proper defence). Here again, I believe it would have been desirable to have a common standard throughout the EU, as otherwise there might be different levels of protection that depend on which authority the case is (often randomly) allocated.

5.

Nature of the Sanction

Fines are viewed as being of merely administrative in nature in some Member States, in which case the guarantees provided for by Article 6 ECHR do not apply. Other Member States consider that fines have a criminal character.51 In Stenuit, the 50 Already

in the present situation, it is often difficult to ensure that confidentiality is maintained. In one case, for instance, where I filed with the Commission a confidential complaint on behalf of a company alledging abuses of a dominant position by a competitor, the Commission decided to refer the case to the Finnish competition authority, which, according to its general policy, reports all pending cases in a registry accessible to the public. Despite the explicit request of confidentiality made to the Commission, the undertaking suspected to abuse of its dominant position became thus aware that a complaint was lodged against it, as well as of the identity of the company which lodged the complaint, contrary to the principle of confidentiality which prevails at the Community level in such a case (see Case 145/83 Adams [1985] ECR 3539). 51 See in Denmark, Ireland and France. In France, criminal penalties are however imposed only on individuals who have played a personal and decisive role in the conception, organization or implementation of the anti-competitive practices. See Jones T. (2001), supra note no. 29, at p. 408.

Technical Problem Areas

479

European Court of Human Rights52 decided that fines imposed in competition procedures have a criminal character, and that the guarantees provided by Article 6 ECHR therefore had to be respected. This was further confirmed by the ECJ in the Hüls case.53 Although it is unclear how far Member States have already accepted all the consequences of this case-law,54 it will be essential that Member States accept the same rule in this area when they decentralise.

6.

Leniency Programmes

The content and application of leniency programs also varies among Member States, with some not even having any. The protection granted to an undertaking admitting an infringement can thus depend on the Member State(s) to which it applied for leniency. Moreover, where a national authority to which a request for leniency was made is discharged from the case because of re-allocation to another competition authority, it is unclear whether the undertaking can claim the benefit of ‘whistle-blowing’ before the former authority. Furthermore, where different undertakings have applied to various national competition authorities across different Member States, it is unclear which can be considered to be the first and thus entitled to a higher ‘rebate’. Nor is there any uniformity as to the kind of information that has to be provided in order to benefit from leniency programmes. It will have to be clarified how far the very principle of rewarding a company for not defending itself, which is applied differently in the different national legal orders, can be reconciled with the right to mount a defence. It appears to me that all these uncertainties are difficult to reconcile with the principle of legality of sanction enshrined in the ECHR.55

52 Report

of the Commission of 30 May 1991, Application No 11598/85. The Court did not issue a judgment in this case. 53 Case C–199/92 P Hüls v. Commission [1999] ECR I–4287, points 149–150, according to which: ‘given the nature of the infringements in question and the nature and degree of severity of the ensuing penalties, the principle of the presumption of innocence (resulting in particular from Article 6(2) of the ECHR) applies to the procedures relating to infringements of the competition rules applicable to undertakings that may result in the imposition of fines or periodic penalty payments’. 54 France appears to have drawn the most far-reaching consequences from this case-law. See Waelbroeck D. and Griffiths M. (2000): ‘The right to a fair hearing — recent French judgments on EC competition law’, 37 Common Market Law Review 6, pp. 1465–1476. 55 So far, the Commission merely noticed the existence of the difficulties raised by the existence of different leniency programmes and the allocation of cases among the national competition authorities in a paper on cooperation within the network of competition authorities submitted to the Council Working Group. Aware that such disparities risk impairing the efficiency of leniency programs, the Commission drew up three options in order to remedy to this situation: harmonisation of the main features of leniency programmes, mutual recognition, or a common leniency programme if the case is reallocated.

480

6.

Working Paper X

Judicial Control of Decisions by the National Competition Authorities

The institutional status of the authority competent to impose an antitrust sanction also varies among the Member States, some of which may entrust it to an administrative body and others to a true court.56 Under the current case-law of the European Court of Human Rights, fines imposed by a national administrative authority for infringements of competition law must be subject to a full jurisdictional review in order to comply with Article 6 ECHR.57 This Article imposes full jurisdictional control with respect to all disputes concerning civil rights obligations (that is, all cases in which a private right is at issue58) or criminal charges (imposition of fines59). In the words of the Court, full jurisdictional control implies a review of ‘questions of fact just as much as questions of law’. It therefore cannot consist of a mere review of legality, and even less a marginal review of legality. As the authorities entitled to award fines against undertakings violating competition law vary greatly between the Member States, fundamental issues about rights are to be anticipated. Once again, undertakings may be very dependent for the protection of their fundamental rights on the way the case is allocated between the national competition authorities.

7.

Suspensive Effect of an Appeal

The consequences of an appeal against a decision of a competition authority that has found an infringement of competition rules are also different among the Member States. In some Member States there is an automatic suspension of the decision and its consequences (fines, interim measures, implementation of remedies) as soon as the undertakings challenge its validity. 60 In others, suspension is possible only upon specific application by the addressee of the decision.61 Suspension is accepted only in exceptional circumstances in other cases.62

56 In France, see Waelbroeck D. and Griffiths M. (2000), supra note no. 54. 57 See Waelbroeck D. and Fosselard D. (1994): ‘Should the decision-making

power in EC antitrust procedures be left to an independent judge?’, 14 Yearbook of European Law, pp. 111 et seq. 58 See ECHR, Le Compte, judgment of 23 June 1981, paras 44 and 48. 59 See Commission HR Report, Stenuit, supra note no. 52, para 63. 60 Germany (§64 GWB), Portugal (Article 24 Decreto Lei 371/93), and Italy (Article 588 Codice di procedura penale). 61 France (Article L464–8 Nouveau Code de Commerce), Belgium (Article 43bis al. 4 Loi sur la protection de la concurrence économique), Ireland (Article 9 III Competition Act 1991), Sweden, Greece and Spain (Article 44 Real Decreto 16/1989). 62 In The Netherlands, for example.

Technical Problem Areas

481

If the company concerned is unable to pay the fine and might be forced into bankruptcy before its case has even been heard by a judge, the non-suspensive effect of an appeal against a decision pronouncing fines — which, as developed above, are at least likely to be generally subject to Article 6 ECHR — may be viewed as infringing the principle of Article 6(2) ECHR. That principle establishes the right to be presumed innocent until proven guilty, as well as the right to have an independent tribunal hear one’s case.63

8.

Obligation for the Authority to Act Upon Complaints

When a complaint is lodged with a national competition authority, the complainant does not enjoy in all Member States the right to obtain a decision.64 The ability to lodge an action for failure to act is not recognised in all Member States. Thus, the degree of judicial protection given to the complainant also diverges among the Member States. More generally, the rights of the complainant differ as to access to the file or the right to participate in the hearing, for example.65

III. Parallel Application of Competition Rules by Multiple Authorities, of which One (the Commission) is not Bound by the ECHR It is clear that many of the above-mentioned problems are likely to have some ‘fundamental right’ dimension. They may therefore give a new role to the European Court of Human Rights in the development of a common minimum standard in the application of EC competition rules. One particular challenge will result from the fact that national competition authorities will partly act on the basis of EC law, and partly on the basis of

63 See

Case Senator Lines currently pending before the ECHR subsequent to an order of the President of the CFI (Order of 21 July 1999, Case T–191/98 R DSR Senator Lines v. Commission [1999] ECR II–2531 confirmed by the President of the European Court of Justice, Order of 14 December 1999, Case C–364/99 P (R) DSR-Senator Lines v. Commission [1999] ECR I–8733) refusing to suspend a Commission decision imposing fines leading the condemned company to bankruptcy before its main case on the merits has ever been heard by the Court. 64 See Zinsmeister U. and Lienemeyer M. (2002), supra note no. 9. In Greece, Article 14 of the Competition Act (No. 703/1997) provides for this possibility. In Germany, this possibility is explicitly excluded (Schmidt in Immenga-Mestmäcker, GWB, Kommentar zum Kartellgesetz, § 63, Nr 33). 65 According to Zinsmeister U. and Lienemeyer M. (2001), supra note no. 29, complainants have a right of access to the file in Ireland, Portugal and Finland, while the rights are slightly more restricted in Italy, Austria, Sweden and Spain.

482

Working Paper X

national law and procedures. This leads to our third main problem: that is, that while the Member States are individually bound by the ECHR, the EC as such is not. Moreover we know also that the European Court of Justice and the European Court of Human Rights have occasionally expressed diverging views as to the interpretation of some provisions of the Convention. Under these circumstances, confusion could be created by the question of whether a national court should rely on the case-law of the European Court of Justice — insofar as it is acting ‘within the scope of Community law,’66 which requires it to ensure the ‘efficacy’ of procedures67 — or of the European Court of Human Rights. In principle, the interpretation of EC rules is left solely to the European Court of Justice, and national authorities are not allowed to deviate from the standards it sets. However, as procedural aspects remain national under the proposal, national authorities should respect the ECHR, to which they are signatories. National judges may thus hesitate to follow the interpretation given by the European Court of Justice on the one hand68 (as they risk being sanctioned by the European Court of Human Rights), but they risk jeopardising the uniform interpretation of EC law among the Member States and the ‘efficacy’ of procedures (as interpreted by the European Court of Justice) if the judges stick to the latter’s interpretation of the law. Moreover, as long as the EC is not part to the ECHR, one might even wonder whether every allocation of a case to the Commission — instead of to a national authority — is an infringement of the ECHR because it deprives the parties of the judicial review mechanism set up by the Convention. Although this would be a very unsatisfactory result (and irreconcilable with my arguments under Section I above), it seems to me that the correct remedy to this problem is to ensure the early accession of the EC to the ECHR.

IV.

Conclusions

In this article I tried to show that, while the creation of a new ‘multi headed hydra’ for the application of EC competition rules will certainly enhance the efficiency of antitrust enforcement across the EU, it is likely to raise a 66 See e.g. Case C–260/89 ERT [1991] ECR I–2925, point 41. 67 According to the European Court of Justice, there is an obligation

for Member States ‘to ensure that the Commission’s action is effective, while respecting the general principles [of Community law]’ (see Joined Cases 46/87 and 227/88 Hoechst [1989] ECR 2859, at point 33). 68 In Mannesmannsröhren-Werke (supra note no. 42) the applicant relied on the right arising from German law whereby natural or legal person may be required to incriminate himself or itself before an investigating authority. The ECJ answered at para 84 that ‘in the field of competition law, the national laws of the Member States do not, in general, recognise a right not to incriminate oneself. It is, therefore, immaterial to the result of the present case whether or not, as the applicant claims, there is such a principle in German law’.

Technical Problem Areas

483

few lawyers’ eyebrows. This is at least with regard to the ‘fight against cartels’ (the Commission will be able to count on the additional forces of 15 to 25 other authorities). First, the parallel application of competition rules by multiple authorities entails a major risk of inconsistent decision-making and double jeopardy. This invokes questions concerning the principles of sanction legality and non bis in idem. This situation could ideally have been avoided through a more far-reaching reform that implied concrete case allocation and mutual recognition (or, alternatively, joint prosecutions before one central court). Secondly, the application of different procedural rules by the different authorities is likely to raise procedural problems for the companies concerned, at all stages of the procedure (investigation, requests for information, statement of objections, access to the file, decision, sanction, appeal, and so on). The list of examples given above is only illustrative of the types of situations that can be anticipated. If decentralisation is to succeed, it therefore appears highly desirable for us to consider a minimal harmonisation of procedural standards throughout the EU in the medium term, at least in those situations in which problems are most likely to occur. As stressed by S. Kon:69 If national competition authorities are empowered to apply Articles 81 and 82, there seems to be no good reason why they should each follow separate procedures in matters as fundamental as powers of search and seizure, rights of inspection, rights of defence, duration of procedures, interim measures and fining powers. The very difficulties in removing or at least narrowing such disparities only go to emphasise that they must be addressed for decentralised enforcement to be effective.

Leaving harmonisation on an ad hoc basis to the development of the respective case-law of the European Court of Human Rights and the European Court of Justice would be a lengthy process. It would occur, not only at the cost of the companies concerned, but also at the cost of the efficacy of the procedures. Problems such as those identified in the present contribution could, in my view, most appropriately be avoided — ex ante — through appropriate legislative action at EC level. Thirdly, it appears to me to be more urgent than ever that the EU should sign the ECHR, as the risk of the European Court of Justice and the European Court of Human Rights giving diverging interpretations of its provisions only adds to the confusion. Moreover, there is no reason to deprive the interested parties of the procedural mechanisms established by the ECHR when a case is allocated to

69 Kon

S. (2000): ‘The Commission’s White Paper on Modernisation: the need for procedural harmonisation’, in Hawk B., ed., 1999 Fordham Corporate Law Institute: International Antitrust Law and Policy, New York, Juris, at p. 252.

484

Working Paper X

the Commission. This accession would appear to follow the logic of the recent Matthews case, in which the European Court of Human Rights held that transfers of competencies to international organisations could only be made provided that Convention rights continue to be ‘secured’. It is therefore clear that Member States’ responsibilities continue even after such a transfer.70

70 ECHR,

Matthews v. UK, 18 February 1999, Application #24833/94, para 32.

BIBLIOGRAPHY

I.

Publications

Albert M. (1993): Capitalism Against Capitalism, Whurr, London. Amato G. (1998): Il Gusto della Libertà: Italia e l’Antitrust, Laterza, Roma-Bari. Ansell C. (2000): ‘The networked polity: regional development in Western Europe’, 13 Governance 3, 303 et seq. Armstrong K.A. (1995): ‘Regulating the free movement of goods: institutions and institutional change’ in Shaw J. and More G., eds., New Legal Dynamics of European Union, Clarendon Press, Oxford. Ayres I. and Braithwaite J. (1992): Responsive Regulation: Transcending the Deregulation Debate, Oxford University Press, Oxford/New York. Benson J. K. (1975): ‘The inter-organizational network as a political economy’, 20 Administrative Science Quarterly, 229 et seq. Bergeron J. H. (2001): ‘Antitrust federalism in the European Union after the modernization initiative’, Antitrust Bulletin, 513 et seq. Bericht des Bundeskartellamtes über seine Tätigkeit in den Jahren 1999/2000 sowie über die Lage und Entwicklung auf seinem Aufgabengebiet (2001), in Bundestagsdrucksache 14/6300. Bernitz U. (1993): Den Nya Konkurrenslagan, 2nd edition, Stockholm. Bomberg E. and Peterson J. (2000): ‘Policy transfer and europeanization: passing the Heineken test?’, published in Queen’s Online Papers on Europeanization, No. 2/2000, available at http://www.qub.ac.uk/ies/onlinepapers/poe2.html. Börzel T. (1997): What’s so special about policy networks? An exploration of the concept and its usefulness for studying European governance’, 1 European Integration On-line Papers 16; available at http://eiop.or.at/eiop/texte/1997-016a.htm. Braithwaite J. (2002): Restorative Justice and Responsive Regulation, Oxford University Press, Oxford. Burley A.-M. and Mattli W. (1993): ‘Europe before the Court: A political theory of legal integration’, 47 International Organization 1, 41 et seq. Burton J. (1997): ‘The competitive order or ordered competition? The “UK model” of utility regulation in theory and practice’, 75 Public Administration, 157 et seq. Christiansen T. and Kirchner E. (2000): ‘Introduction’, in Christiansen T. and Kirchner E., eds., Committee Governance in the European Union, Manchester University Press, Manchester. Coen D. and Doyle C. (2000): ‘Designing economic regulatory institutions for European network industries’, 9 Current Politics and Economics of Europe 4, 455 et seq. Dabbah M. (2000): ‘Measuring the success of a system of competition law’, 21 European Competition Law Review 8, 369 et seq.

486

Publications

Dahl R. A. (1994): ‘A democratic dilemma: system efectiveness versus citizen participation’, 109 Political Science Quarterly, 23 et seq. De Grauwe P. (2000): Economics of Monetary Union, 4th edition, Oxford University Press, Oxford. Dehousse R. (1997): ‘Regulation by networks in the European Community: the role of European agencies’, 4 Journal of European Public Policy 2, 246 et seq. DiMaggio P. J. and Powell W. W. (1991): ‘The Iron Cage revisited: institutional isomorphism and collective rationality in organizational fields’, in Powell W. W. and DiMaggio P. J., eds., The New Institutionalism in Organizational Analysis, University of Chicago Press, Chicago, 63 et seq. Djelic M. (1998): Exporting the American Model, Oxford University Press, Oxford. Doern B. and Wilks S., eds. (1996): Comparative Competition Policy: National Institutions in a Global Market, Clarendon Press, Oxford/Oxford University Press, New York. Dolowitz D. and Marsh D. (1996): ‘Who learns what from whom: a review of the policy transfer literature’, 44 Political Studies 2, 343 et seq. ——. (2000): ‘Learning from abroad: the role of policy transfer in contemporary policymaking’, 13 Governance 1, 5 et seq. Donativi V. (1990): Introduzione della Disciplina Antitrust nel Sistema Legislativo Italiano, Giuffrè, Milano. Dorf M. C. and Sabel C. (1998): ‘A constitution of democratic experimentalism’, 98 Columbia Law Review 2, 267 et seq. Drahos M. (2001): Convergence of Competition Laws and Policies in the European Community, Kluwer, The Hague. Eberlein B. (2000a): ‘Configurations of economic regulation in the European Union: the case of electricity in comparative perspective’, 9 Current Politics and Economics of Europe 4, 407 et seq. ——. (2000b): ‘Institutional change and continuity in German infrastructure management: the case of electricity refom’, 9 German Politics 3, 81 et seq. ——. (2001): ‘To regulate or not to regulate: explaining the German Sonderweg of Electricity Regulation’, 2 Journal of Network Industries 3, 1 et seq. Eberlein B. and Grande E. (2000): ‘Regulation and infrastructure management: German regulatory regimes and the EU framework’, 1 German Policy Studies 1, 39 et seq. ——. (2002): ‘Die Europäische Union als Regulierungsstaat’, in Jachtenfuchs M and Kohler-Koch B., eds., Europäische Integration, 2nd edition, UTB, Opladen. Eberlein B. and Kerwer D. (2002): ‘Theorising the new modes of European Union governance’, 6 European Integration On-line Papers 5; available at http://eiop.or.at/eiop/ texte/2002-005a.htm. Ehlermann C.-D. (1986): ‘Implementation of EC competition law by national antitrust authorities,’ 17 European Competition Law Review, 88 et seq. ——. (2000): ‘Reform of European competition law. Coherent application of EC competition law in a system of parallel competencies’, speech delivered at a conference in Freiburg, 10.11.2000, available at http://www.europa.eu.int/ comm/competition/ antitrust/others. ——. (2000): ‘The modernization of EC antitrust policy: a legal and cultural revolution,’ 37 Common Market Law Review 3, p. 537 et seq. Ehlermann C.-D. and Atanasiu I., eds. (2001): European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, Oxford and Portland, Oregon.

Bibliography

487

Eichener V. (1997): ‘Effective European problem-solving: lessons from the regulation of occupational safety and environmental protection’, 4 Journal of European Public Policy 4, 591 et seq. Eisner M. A., Worsham J. and Ringquist E. J. (2000): Contemporary Regulatory Policy, Lynne Rienner, Boulder, Colorado. Fingleton J., Fox E., Neven D. and Seabright P. (1996): Competition Policy and the Transformation of Central Europe, CEPR, London. Forrester I. (2001): ‘The reform of the implementation of Articles 81 and 82 following publication of the Draft Regulation’, 28 Legal Issues of Economic Integration 2, 173 et seq. Fourgoux J. C. (2001): ‘Un nouvel antitrust européen: espoirs et craintes’, conference held at the Tribunal de Commerce de Paris, 22 October 2001. Frazer T. (1992): Monopoly, Competition and the Law, 2nd edition, Hemel Hempstead, Harvester Wheatsheaf. Gerber D. J. (1994): ‘Constitutionalizing the economy: German neo-liberalism, competition law and the ‘New Europe’’, 42 American Journal of Comparative Law, 25 et seq. ——. (1994): ‘The transformation of European Community competition law’, 35 Harvard International Law Journal 1, 97 et seq. ——. (1998): ‘System dynamics: toward a language of comparative law?,’ 46 American Journal of Comparative Law, 719 et seq. ——. (1998): Law and Competition in Twentieth Century Europe: Protecting Prometheus, Clarendon Press, Oxford. Goldschmidt P. I. B. and Lanz C. (2001): ‘Maybe definitely — definitely maybe? EC competition law — is the time ripe for reform?’, Eipascope 2, 16 et seq. Goyder D. G. (1998): EC Competition Law, 3rd edition, Clarendon Press, Oxford. Graupner F. (1973): ‘Commission decision-making on competition questions’, 10 Common Market Law Review, 291 et seq. Greene T., O’Connor K. and Hubbard R. L. (2001): ‘State antitrust law and enforcement’, Practicing Law Institute, 42nd Annual Antitrust Law Institute, 1252 PLI/ Corp 1129. Haagsma A. (1993): ‘The competition rules of the EEA and the European Agreements: lawyer’s paradise or user’s safe harbour’, in Slot P.J. and McDonnell A., eds., Procedure and Enforcement in EC and US Competition Law, Sweet & Maxwell, London, 243 et seq. Haas P. M. (1992): ‘Introduction: epistemic communities and international policy coordination’, 46 International Organization 1, 1 et seq. Hall C., Scott C. and Hood C. (2000): Telecommunications Regulation, Routledge, London. Hall P. (1986): Governing the Economy. The Politics of State Intervention in Britain and France, Cambridge University Press, Cambridge. Hammond A. and Penrose R. (2001): ‘Proposed criminalisation of cartels in the UK’, OFT365 of November 2001, Office of Fair Trading, London. Hammond S. D. (2001): ‘When calculating the costs and benefits of applying for corporate amnesty, how do you put a price tag on an individual’s freedom?’, speech delivered at the US National Institute on White Collar Crime, 8 March 2001. Hancher L. and Moran M. (1989): Capitalism, Culture and Economic Regulation, Clarendon Press, Oxford.

488

Publications

Hawk B. E. (1972): ‘Antitrust in the EEC — The first decade,’ 41 Fordham Law Review, 229 et seq. Hawk B. E. and Laudati L. (1996): ‘Antitrust federalism in the United States and decentralization of competition law enforcement in the European Union: a comparison’, 20 Fordham International Law Journal, 18 et seq. Heclo H. (1978): ‘Issue networks and the executive establishment, in King A., ed., The New American Political System, Washington DC, The American Enterprise Institute, 87 et seq. Heclo H. and Wildavsky A. (1974): The Private Government of Public Money, London, Macmillan. Héritier A. (1999): Policy-Making and Diversity in Europe: Escape from Deadlock, Cambridge University Press, Cambridge. Hix S. (1999): The Political System of the European Union, St. Martin’s Press, New York. Hodson D. and Maher I. (2001): ‘The open method as a new mode of governance’, 39 Journal of Common Market Studies 4, 719 et seq. ——. (2002): ‘Economic and Monetary Union: balancing credibility and legitimacy in an asymmetric policy-mix’, 9 Journal of European Public Policy 3, 391 et seq. Holley D. L. (1993): ‘E.E.C. competition practice: A thirty-year retrospective,’ in Hawk B. E., ed., 1992 Fordham Corporate Law Institute, Huntington, New York, 669 et seq. Hooghe L. (1996): ‘Introduction: Reconciling EU-wide policy and national diversity’, in Hooghe L., ed., Cohesion Policy and European Integration: Building Multi-Level Governance, Oxford University Press, Oxford, 1 et seq. Horn M. J. (1995): The Political Economy of Public Administration: Institutional Choice in the Public Sector, Cambridge University Press, Cambridge. Hutchins E. (1995): Cognition in the Wild, MIT Press, Cambridge, Massachusets. Hutton W. (1996): The State We’re In, Vintage, London. Immergut E. M. (1992): Health Policies. Interests and Institutions in Western Europe, Cambridge University Press, Cambridge. Jansen D. and Schubert K., eds. (1995): Netzwerke und Politikproduktion. Konzepte, Methoden, Perspektiven, Schüren, Marburg. Jones T. (2001): ‘Regulation 17: the Impact of the current application of Articles 81 and 82 by national competition authorities on the European Commission’s proposals for reform’, 22 European Competition Law Review 10, 405 et seq. Kenis P. and Schneider V. (1991): ‘Policy networks and policy analysis: scrutinising a new analytical toolbox’, in Marin B. and Mayntz R., eds., Policy Networks: Empirical Evidence and Theoretical Considerations, Campus Verlag, Frankfurt. Keohane R. O. and Nye J. S. (2001): ‘Between centralization and fragmentation: the club model of multilateral cooperation and problems of democratic legitimacy’, John F. Kennedy School of Government, Harvard University, Faculty Research Working Paper Series, February 2001. Kettl D. F. (2000): ‘The transformation of governance: globalization, devolution, and the role of government’, Discussion Paper prepared for the Spring Meeting of the National Academy of Public Administration, June 1–3, 2000, Albuquerque, NM. Kickert W. J. M. et al. (1999): ‘Introduction: a management perspective on policy networks’, in Kickert W. J. M., Klijn E.-H. and Koppenjan J. F. M., eds., Managing Complex Networks: Strategies for the Public Sector, London, Sage, 6 et seq.

Bibliography

489

Kingston S. (2001): ‘A new division of responsibilities in the proposed Regulation to modernise the rules implementing Articles 81 and 82? A warning call’, 22 European Competition Law Review 8, 340 et seq. Knoke D. (1990): Political Networks. The Structural Perspective, Cambridge University Press, Cambridge. Knoke D., Pappi F. U., Broadbent J. and Tsujinaka Y. (1996): Comparing Policy Networks: Labour Politics in the U.S., Germany, and Japan, Cambridge University Press, Cambridge. Kohler-Koch B. (1999): ‘The evolution and transformation of European governance’, in Kohler-Koch B. and Eising R., eds., The Transformation of Governance in the European Union, Routledge, London, 14 et seq. Kon S. (2000): ‘The Commission’s White Paper on Modernisation: the need for procedural harmonisation’, in Hawk B., ed., 1999 Fordham Corporate Law Institute: International Antitrust Law and Policy, New York, Juris. Kovacic W. E. (1990): ‘The Sherman Act: the first century; comments and observations,’ 59 Antitrust Law Journal 119. ——. (2000): ‘Lessons of competition policy reform in transition economies for U.S. antitrust policy’, 74 St. John’s Law Review 361. Krause H. (1995): ‘EC merger control: An outside view from inside the Merger Task Force’, 39 Journal of Business Law, 627 et seq. Kreher A. and Mény Y. (1997): ‘European Agencies’, Special Section, 4 Journal of European Public Policy 2, 225 et seq. Laumann E. O. and Knoke D. (1987): The Organizational State. Social Choice in National Policy Domains, University of Wisconsin Press, Madison. Lenaerts K. and De Smijter E. (2001): ‘A ‘Bill of Rights’ for the European Union’, 38 Common Market Law Review 2, 273 et seq. Lenaerts K. and Vanhamme J. (1997): ‘Procedural rights of private parties in the Community administrative process’, 34 Common Market Law Review, 531 et seq. Levy B. and Spiller P. T. (1996): Regulation, Institutions and Commitment, Cambridge University Press, Cambridge. Lynch J. (2001): ‘Federalism, separation of powers, and the role of State Attorneys General in multistate litigation,’ 101 Columbia Law Review, 1998 et seq. MacConnell S. (2001): ‘IFA-Meat Plants dispute has roots in veterinary levy’, Irish Times, January 17, 2001, Dublin. Macey J. (1992): ‘Organizational design and political control of administrative agencies’, 8 Journal of Law, Economics and Organization 1, 93 et seq. Maher I. (1996): ‘Alignment of competition laws in the European Community’, 16 Yearbook of European Law, 223 et seq. ——. (1999): ‘Competition law and intellectual property rights: evolving formalism’, in Craig P. and de Búrca G., eds., The Evolution of EU Law, Oxford University Press, Oxford. ——. (2000): ‘Juridification, codification and sanction in UK competition law’, 63 Modern Law Review 4, 544 et seq. Majone G. (1991): ‘Cross-national sources of regulatory policy-making in Europe and the United States’, 11 Journal of Public Policy 1, 79 et seq. ——. (1994): ‘Paradoxes of privatization and deregulation’, 1 Journal of European Public Policy 1, 53 et seq.

490

Publications

——. (1994): ‘The rise of the regulatory state in Europe’, in Müller W. and Wright V., eds., The State in Western Europe. Retreat of Redefinition, Frank Cass, Illford, 77 et seq. ——. (1996): Regulating Europe, Routledge, London. ——. (1997): ‘The new European agencies: regulation by information’, 4 Journal of European Public Policy 2, 262 et seq. ——. (1998): ‘Europe’s “democratic deficit”: the question of standards’, 4 European Law Journal 1, 5 et seq. ——. (2000): ‘The evolution of regulatory policies and institutions in the European Union’, paper presented to the Symposium on European Integration, Center for International Studies of El Colegio de Mexico, Mexico City, November 30, 2000. Marenco G. (1994): ‘The uneasy enforcement of Article 85 EEC as between Community and national levels’, in Hawk B., ed., Annual Proceedings of the Fordham Corporate Law Institute 1993, Juris, New York, 605 et seq. Marin B. and Mayntz R., eds. (1991): Policy Networks: Empirical Evidence and Theoretical Considerations, Campus Verlag, Frankfurt. Marsh J. G. and Olson J. P. (1998): ‘The institutional dynamics of international political orders’, 52 International Organization 4, 943 et seq. May J. (1987): ‘Antitrust practice and procedures in the formative era: The constitutional and conceptual reach of state antitrust law, 1889–1918’, 35 University of Pennsylvania Law Review, 495 et seq. McGowan L. and Cini M. (1999): ‘Discretion and politicization in EU competition policy: the case of merger control’, 12 Governance 2, 175 et seq. Messner D. (2000): ‘Gesellschaftliche Determinanten wirtschaftlicher Entwicklung in der Weltmarktwirtschaft. Markt, Netzwerksteuerung und soziale Gerechtigkeit als Elemente einer Entwicklungsstrategie jenseits des Neoliberalismus’, in Brunkhorst H. and Kettner M., eds., Globalisierung und Demokratie. Wirtschaft, Recht, Medien, Suhrkamp, Frankfurt am Main, 17 et seq. Mestmäcker E.-J. (1989): ‘Merger control in the Common Market: Between competition policy and industrial policy’, in Hawk B., ed., 1989 Fordham Corporate Law Institute, Huntington, New York. Monti M. (2000): ‘Nouveaux enjeux pour la politique européenne de concurrence’, Commission des affaires étrangères de l’Assemblée nationale, Paris, 17 October 2000. Moravcsik A. (1998): The Choice for Europe: Social Purpose and State Power from Messina to Maastricht, Cornell University Press, Ithaca. Müller W. and Wright V., eds. (1994): The State in Western Europe. Retreat of Redefinition, Frank Cass, Ilford. Noll R. G. (1985): ‘Government regulatory behavior. A multidisciplinary survey and synthesis’, in Noll R. G., ed., Regulatory Policy and the Social Sciences, University of California Press, Berkely, 9 et seq. O’Connor, K. J. (1997): Working Paper V — Panel II (Decentralisation of Enforcement of Community Law, in Ehlermann C.-D. and Laudati L., eds., Robert Schuman Centre Annual on European Competition Law 1996, Kluwer Law International, London/The Hague/Boston, 147 et seq. Ortwein E. (1998): Das Bundeskartellamt. Eine Politische Ökonomie deutscher Wettbewerbspolitik, Nomos, Baden-Baden. Peters G. B. (1996): ‘United States competition policy institutions: structural constraints and opportunities’, in Doern G. B. and Wilks S., eds., Comparative Competition Policy: National Institutions in a Global Market, Clarendon Press, Oxford.

Bibliography

491

Picciotto S. (1997): ‘Fragmented states and international rules of law’, 6 Social and Legal Studies 2, 259 et seq. Pollack M. A. and Shaffer G. C. (2001): ‘Transatlantic governance in historical and theoretical perspective’, in Pollack M. A. and Shaffer G. C., eds., Transatlantic Governance in the Global Economy, Rowman & Littlefield, Lanham, 3 et seq. Price Waterhouse Coopers (2001): Peer Review of the UK Competition Policy Regime, final report, 18 April 2001, London, Department of Trade and Industry. Radaelli C. M. (1997): The Politics of Corporate Taxation in the European Union: Knowledge and International Policy Agenda, Routledge, London. Radaelli C. M. (2000): ‘Policy transfer in the European Union: Institutional isomorphism as a source of legitimacy’, 13 Governance 1, 25 et seq. Rhinard M. (2002): ‘The democratic legitimacy of the European Union committee system’, 15 Governance 2, 185 et seq. Rhodes R. A. W. and Marsh D. (1992): ‘New directions in the study of policy networks’, 21 European Journal of Political Research 1/2, 181 et seq. Richardson J. and Jordan G. (1979): Governing Under Pressure, Oxford, Martin Robertson. Rill J. F. (2001): ‘Recent trends and developments in the role of federal and State authorities in US antitrust enforcement: is cooperation working?’, in Ehlermann C.-D. and Atanasiu I., eds., European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, Oxford and Portland, Oregon, 429 et seq. Rill J. F. and Chambers C. S. (1997): ‘Federalism in antitrust enforcement: the United States’ experience with a dual enforcement regime,’ in Ehlermann C.-D. and Laudati L., eds., Robert Schuman Centre Annual on European Competition Law 1996, Kluwer Law International, London/The Hague/Boston, at pp. 205 et seq. Risse-Kappen T. (1995): ‘Introduction: Bringing transnational relations back in’, in Risse-Kappen T., ed., Bringing Transnational Relations Back In: Non-State Actors, Domestic Structures, and International Institutions, Cambridge University Press, New York, 3 et seq. Rodger B. J. (2000): ‘Compliance with competition law: a view from industry’, Commercial Liability Law Review, 249 et seq. Rose J. (1994): ‘State antitrust enforcement, mergers, and politics,’ 41 Wayne Law Review, 71 et seq. Sabel C. (2000): ‘A quiet revolution of democratic governance: towards democratic experimentalism’, in OECD, Governance in the 21st century, OECD, Paris, 121 et seq. Sandholtz W. and Stone Sweet A. (1997): ‘European integration and supranational governance’, 4 Journal of European Public Policy 3, 297 et seq. Schaap L. and van Twist M. J. W. (1999): ‘The dynamics of closedness in networks’, in Kickert eds., Managing Complex Networks: Strategies for the Public Sector, London, Sage. Scharpf F. W. (1997): ‘Introduction: the problem-solving capacity of multi-level governance’, 4 Journal of European Public Policy 4, 520 et seq. ——. (1999): Governing in Europe. Effective and Democratic?, Oxford University Press, Oxford. ——. (2001): ‘European governance: common concerns vs. the challenge of diversity’, MPIfG Working Paper 01/6, September 2001, Max Planck Institute for the Studies of Societies, Cologne.

492

Publications

Schaub A. (2000): ‘Modernization of EC competition law: reform of Regulation 17’, 23 Fordham International Law Journal, 752 et seq. ——. (2001): ‘Continued focus on reform; recent developments in EC competition policy’, paper presented at the 28th Annual Conference of the Fordham Corporate Law Institute, October 2001. ——. (2001): ‘The direction of competition policy: reconciling national and international objectives’, paper presented at the Annual Fall Conference on Competition Law, Ottawa, 21 September 2001. ——. (2001): ‘The global competition forum: how it should be organised and operated’; EPC Brussels, 14 March 2001. Scott C. (2001): ‘Analysing regulatory space: fragmented resources and institutional design’, Public Law, 329 et seq. Slaughter A.-M. (1997): ‘The real new world order’, 76 Foreign Affairs 5, 183 et seq. ——. (2000): ‘Governing the global economy through government networks’, in Byers M., ed., The Role of Law in International Politics: Essays in International Relations and International Law, Oxford University Press, Oxford. Slotboom M. (1997): ‘De NMa staat voor de deur; toezicht en onderzoek in het kader van de Mededingingswet’, TVVS, 322 et seq. Smith M. P. (2001): ‘In pursuit of selective liberalization: Single Market competition and its limits’, 8 Journal of European Public Policy 4, 519 et seq. Steinmo S. and Thelen K. (1992): ‘Historical institutionalism in comparative politics’, in Steinmo S., Thelen K. and Longstreth F., eds., Structuring Politics, Cambridge University Press, Cambridge. Stellungnahme der Bundesregierung zum Dreizehnten Hauptgutachten der Monopolkommission 1998/1999 (2001), Bundestagsdrucksachen 14/4002 and 14/4003. Stellungnahme der Bundesregierung zum Tätigkeitsbericht 1999/2000 des Bundeskartellamts (2001), Bundestagsdrucksache 14/6300. Sturm R. (1996): ‘The German cartel office in a hostile environment’, in Doern G. B. and Wilks S., eds., Comparative Competition Policy. National Institutions in a Global Market, Clarendon Press, Oxford, 185 et seq. Sturm R. and Müller M. M. (2001): ‘Governing capitalist economies: beyond the regulation debate’, in Sturm R., Dieringer J. and Müller M. M., eds., Rediscovering Competition. Competition Policy in East Central Europa in Comparative Perspective, Hvgorac, Budapest, 9 et seq. Temple Lang J. (2000): ‘Decentralized application of Community competition law’, in Rivas J. and Horspool M., eds., Modernization and Decentralization of EC Competition Law, Kluwer Law International, The Hague and Boston. Thatcher M. (2001): ‘The Commission and national governments as partners: EC regulatory expansion in telecommunications 1979–2000’, 8 Journal of European Public Policy 4, 558 et seq. ——. (2002): ‘Delegation to independent regulatory agencies: pressures, functions, and contextual mediation’, 25 West European Politics 1, 125 et seq. Turetsky D.S. (1997): Working Paper IX–Panel II (Decentralisation of Enforcement of Community Law, in Ehlermann C.-D. and Laudati L., eds., Robert Schuman Centre Annual on European Competition Law 1996, Kluwer Law International, London/The Hague/Boston, 227 et seq. Tyler T. and Blader S. (2000): Cooperation in Groups: Procedural Justice, Social Identity, and Behavioral Engagement, Psychology Press, Phyladelphia.

Bibliography

493

Vainanidis C. D. and Pournara J. (1993): ‘Greek anti-trust law: a critical appraisal’, 14 European Competition Law Review 5, 226 et seq. Van der Woude M. (2000): ‘De eerste contouren van de procedure voor de NMa’, in Slot P. J. and Swaak C. R., eds, De Nederlandse Mededingingswet in perspectief, Kluwer, Deeventer, 141 et seq. Van Waarden F. (1992): ‘Dimensions and types of policy networks’, 21 European Journal of Political Research, 29 et seq. Vasconcelos J. (2001): ‘Cooperation between energy regulators in the European Union’, in Henry C. et al, eds., Regulation of Network Utilities. The European Experience, Oxford University Press, Oxford, 284 et seq. Verdun A. (1999): ‘The role of the Delors Committee in the creation of EMU: an epistemic community?’, 6 Journal of European Public Policy 2, 308 et seq. Vickers J. (2001): ‘The Competition Act — The inside story’, 2 November 2001, a speech by the DGFT to the Linklaters/DTI Annual Seminar on Competition Law. Vogel S. K. (1996): Freer Markets, More Rules. Regulatory Refom in Advanced Industrial Countries, Cornell University Press, Ithaca. Von Stoephasius H.-P. (1993): ‘Enforcement of E.C. competition law by national authorities’, in Slot P. J. and McDonnell A., eds., Procedure and Enforcement in EC and US Competition Law, Sweet & Maxwell, London. Waelbroeck D. and Fosselard D. (1994): ‘Should the decision-making power in EC antitrust procedures be left to an independent judge?’, 14 Yearbook of European Law, 111 et seq. Waelbroeck D. and Griffiths M. (2000): ‘The right to a fair hearing — recent French judgments on EC competition law’, 37 Common Market Law Review 6, 1465 et seq. Wallace H. (2000): ‘The institutional setting’, in Wallace H. and Wallace W., eds., Policy Making in the European Union, 4th edition, Oxford University Press, Oxford. Wenzl M. (2001): ‘The impact of outcome orientation and justice concerns on tax compliance: the role of taxpayers’ identity’, Centre for Tax System Integrity Working Paper No. 6, Research School of Social Sciences, Australian National University, Canberra. Whish R. (2001): Competition Law, 4th edition, Butterworths, London. Wilks S. (1999): In the Public Interest. Competition Policy and the Monopolies and Merger Commission, University Press, Manchester. Wilks S. and Bartle I. (2002): ‘The unanticipated consequences of creating independent competition agencies’, 25 West European Politics 1, 148 et seq. Wilks S. and McGowan L. (1996): ‘Competition policy in the European Union: creating a federal agency?’, in Doern G. B. and Wilks S., eds., Comparative Competition Policy: National Institutions in a Global Market, Clarendon Press, Oxford. Wils W. P. J. (1996): ‘La compatibilité des procédures communautaires en matière de concurrence avec la Convention européenne des droits de l’homme’, 32 Cahiers de droit européen, 329 et seq. ——. (2001): ‘The modernisation of the enforcement of Article 81 and 82 EC: A legal and economic analysis of the Commission’s proposal for a new Council regulation replacing Regulation 17’, in Hawk B. E., ed., Annual Proceedings of Fordham Corporate Law Institute 2000, Juris, New York, 313 et seq. (also published in 24 Fordham International Law Journal 5 [2001], 1655 et seq.). ——. (2002): The Optimal Enforcement of EC Antitrust Law: Essays in Law and Economics, London, Kluwer Law International, 105 et seq.

494

Legislation, Official Documents and Reports

——. (2003): ‘Does the effective enforcement of Articles 81 and 82 EC require not only fines on undertakings but also individual penalties, in particular imprisonment?’, in Ehlermann C.-D. and Atanasiu I., eds., European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland, Oregon, 411 et seq. Zinsmeister U. and Lienemeyer M. (2002): ‘Die Verfahrensrechtlichen Probleme bei der dezentralen Anwendung des europäischen Kartellrechts’, 52 Wirtschaft und Wettbewerb 4, 331 et seq.

II.

Legislation, Official Documents and Reports

Agreement on the European Economic Area, OJ L 1/3 [1994]. Charter of Fundamental Rights of the European Union — Explanations relating to the complete text of the Charter, available at http://ue.eu.int/df/docs/en/ EN_2001_1023.pdf. Charter of Fundamental Rights of the European Union (2000/C 364/01) EN 18.12.2000, OJ C 364/1 [2000], text available at http://ue.eu.int/df/docs/en/CharteEN.pdf. Commission Directive 80/723 on the transparency of financial relations between Member States and public undertakings, OJ L 195/35 [1980], amended by Directive 93/84, OJ L 254/1 [1993] and Directive 2000/52, OJ L 193/78 [2000]. Commission Directive 90/388/EEC on competition in the markets for telecommunications services, OJ L 19/22 of 24.7.90. Commission Directive 96/2/EC of 16 January 1996 amending Directive 90/388/EEC with regard to mobile and personal communications, OJ L 20/59. Commission Directive on a common regulatory framework for electronic communications networks and services, COM(2000)393 final. Commission Directive on competition in the markets in telecommunications terminal equipment, Commission Directive 88/301/EEC, OJ L 131 of 27.5.88. Convention for the Protection of Human Rights and Fundamental Freedoms, as amended by Protocol No. 11 with Protocol Nos. 1, 4, 6 and 7, text available at: http://www.echr. coe.int/Eng/BasicTexts.htm. Council Directive of 13 December 1995 on the application of open network provision (ONP) to voice telephony, Directive 95/62/EC, OJ L 321/6 of 30.12.1995. Council Directive of 28 June 1990 on the establishment of the internal market for telecommunications services through the implementation of open network provision, 90/387/EEC, OJ L 192/1 of 24.07.1990. Council of the EU (2000): Presidency Conclusions, Lisbon European Council 23–24 March, Press Release No. 100/1/00 of 24.03.2000, Lisbon. Council of the EU, Economic and Social Committee (2000): Opinion on the White Paper on modernisation of the rules implementing Articles 81 and 82 of the EC Treaty — Commission programme No 99/02, OJ C 51/55 [2000]. Council Regulation (EEC) 4064/89 of 21 December 1989 on the control of concentrations between undertakings, OJ L 257/14 [1990] (corrected version of OJ L 395 [1989])]. Council Resolution on the Development of Administrative Cooperation in the Implementation and enforcement of Community Legislation in the Internal Market, OJ C 179/1.

Bibliography

495

Directive 95/46 on the protection of individuals with regard to the processing of personal data and on the free movement of such data, OJ L 281/31 [1995]. Draft Joint Statement of the Council and the Commission on the Function of the Network of Competition Authorities, EU Presidency Working Document of 14.03.2002. Draft Notice on Cooperation within the Network of Competition Authorities, European Commission Staff Working Paper of February 2002 (EU Restricted), submitted to the Council Working Group. European Commission (1992): Proposal for a Council Directive on the Mutual Recognition of Licences and Other National Authorizations to Operate Telecommunications Services, Including the Establishment of a Single Community Telecommunications License and the Setting-up of a Community Telecommunications Committee (CTC), (92/C248/05), OJ C 248/4 of 25.9.1992. European Commision (1993): Notice on Cooperation between National Courts and the Commission in applying Article 85 or 86 of the EC Treaty, OJ C 39/6 [1993]. European Commission (1997): Notice on cooperation between national competition authorities and the Commission in handling cases falling within the scope of Articles 85 and 86 EC, OJ C 313/3 of 13.2. 1993. European Commission (1997): Notice on the internal rules of procedure for processing requests for access to the file in cases under Articles 85 and 86 of the EC Treaty, Articles 65 and 66 of the ECSC Treaty and Council Regulation (EEC) No 4064/89, OJ C 23/3 [1997]. European Commission (1999): Fifth Report on the Implementation of the Telecommunications Regulatory Package, available at http://europa.eu.int/ISPO/infosoc/telecompolicy/ 5th-en.pdf . European Commission (1999): The 1999 Communications Review. Towards a New Framework for Electronic Communications Infrastructure and Associated Services, COM (1999)539 final. European Commission (1999): White Paper on Modernisation of the Rules Implementing Articles 85 and 86 (now 81 and 82) of the EC Treaty, OJ C 132/1 of 12.5.1999. European Commission (2000): Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, COM(2000) 582 final, 27.09.2000. European Commission (2001): Progress Report to the Council of the European Union on the Proposal for a Council Regulation on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty and amending Regulation (EEC) No 1017/68, (EEC) No 2988/74, (EEC) No 4056/86 and (EEC) No 3975/87 SN4467/01, 7 November 2001. European Commission (2002): Report to the Commission to the Council and the European Parliament on the application of the Agreement between the European Communities and the Government of the USA regarding the application of their competition laws — 1 January 2000 to 31 December 2000, COM (2002) 45 final. European Parliament and Council Directive 2002/19/EC on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive), OJ L 108 of 24.4.2002. European Parliament and Council Directive 2002/20/EC on the authorisation of electronic communications networks and services (Authorisation Directive), OJ L 108 of 24.4.2002.