European Competition Law Annual 2006: Enforcement of Prohibition of Cartels

(A) Ehlermann 07 Prelims 16/10/07 16:11 Page i EUROPEAN COMPETITION LAW ANNUAL 2006: Enforcement of Prohibition of C...

Author: Claus-Dieter Ehlermann | Isabela Atanasiu

66 downloads 5447 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

(A) Ehlermann 07 Prelims

16/10/07

16:11

Page i

EUROPEAN COMPETITION LAW ANNUAL 2006: Enforcement of Prohibition of Cartels


16/10/07

16:11

Page ii


16/10/07

16:11

Page iii

EUROPEAN COMPETITION LAW ANNUAL: 2006 Enforcement of Prohibition of Cartels

Edited by

Claus-Dieter Ehlermann and

Isabela Atanasiu

OXFORD AND PORTLAND OREGON 2007


16/10/07

16:11

Page iv

Published in North America (US and Canada) by Hart Publishing c/o International Specialized Book Services 920 NE 58th Avenue, Suite 300 Portland, OR 97213-3786 USA Tel: +1 503 287 3093 or toll-free: (1) 800 944 6190 Fax: +1 503 280 8832 E-mail: [email protected] Website: www.isbs.com

© The editors and contributors severally 2007 The editors and contributors have asserted their right under the Copyright, Designs and Patents Act 1988, to be identified as the authors of this work. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission of Hart Publishing, or as expressly permitted by law or under the terms agreed with the appropriate reprographic rights organisation. Enquiries concerning reproduction which may not be covered by the above should be addressed to Hart Publishing at the address below.

Hart Publishing, 16C Worcester Place, Oxford, OX1 2JW Telephone: +44 (0)1865 517530 Fax: +44 (0) 1865 510710 E-mail: [email protected] Website: http://www.hartpub.co.uk

British Library Cataloguing in Publication Data Data Available

ISBN: 978-1-84113-645-5 Typeset by Hope Services, Abingdon Printed and bound in Great Britain by Biddles Ltd, King’s Lynn, Norfolk


16/10/07

16:11

Page v

PERMANENT SPONSORS OF THE A N N UA L E U I C O M P E T I T I O N WORKSHOPS

Blake, Cassels & Graydon LLP Contact: Calvin S. Goldman, QC Commerce Court West 199 Bay Street Toronto, Ontario Canada M5L 1A9 Tel: 416 863 22 80 Fax: 416 863 26 53 E-mail: [email protected] Cleary Gottlieb Contact: Mario Siragusa Rome Office Piazza di Spagna 15 I-00187 Rome Tel: 39 06 6952 2210 Fax: 39 06 6920 0665 E-mail: [email protected] Hengeler Müller Contact: Jochen Burrichter Benrater Strasse 18-20 40213 Düsseldorf Germany Tel: 49 211 830 41 38 Fax: 49 211 830 41 70 E-mail: [email protected] Howrey Simon Arnold & White LLP Contact: James F. Rill, Esq. 1299 Pennsylvania Ave., NW Washington, DC 20004 Tel: (001 202) 383 65 62 Fax: (001 202) 383 66 10 E-mail: [email protected]


vi

16/10/07

16:11

Page vi

Permanent Sponsors of the Annual EUI Competition Workshops

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 Contact: Jim Venit 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 LLP Contact: Ian S. Forrester, QC 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] WilmerHale Contact: John Ratliff Bastion Tower Place du Champ de Mars/Marsveldplein 5 B-1050 Brussels, Belgium Tel: (32 2) 285 49 08 Fax: (32 2) 285 49 49 E-mail: [email protected]


16/10/07

16:11

Page vii

1 1 T H A N N UA L E U C O M P E T I T I O N L AW AND POLICY WORKSHOP: ENFORCEMENT OF PROHIBITION OF CARTELS

List of Participants

Allendesalazar Rafael, Martínez Lage & Asociados—Madrid, Spain. Barnett Thomas, Assistant Attorney General for Antitrust, Department of Justice—Washington DC, USA. Bellamy Christopher, President of the Competition Appeal Tribunal— London, UK. Bloom Margaret, King’s College, London, and Freshfields Bruckhaus Deringer—London, UK. Burrichter Jochen, Hengeler Müller—Düsseldorf, Germany. Calkins Stephen, Wayne State University School of Law—Detroit, USA. Calviño Nadia, Deputy Director General with special responsibility for mergers, DG Competition, European Commission—Brussels, Belgium. Collins Philip, Chairman, Office of Fair Trading—London, UK. Denis Waelbroeck, Ashurst—Brussels, Belgium. Donncadh Woods, Deputy Head of Unit A-1 (Antitrust policy and strategic support), DG Competition, European Commission—Brussels, Belgium. Ehlermann Claus-Dieter, WilmerHale—Brussels, Belgium. Fingleton John, Chief Executive, Office of Fair Trading—London, UK. Forrester Ian, White & Case LLP—Brussels, Belgium. Forwood Nicholas, Judge at the Court of First Instance of the EC— Luxembourg. Goldman Calvin, Blake, Cassels & Graydon LLP—Toronto, Canada. Grout Paul, Head of Economics Department, Centre for Market and Public Organisation, University of Bristol—Bristol, UK. Guersent Olivier, Deputy Head of Cabinet with Commissioner Neelie Kroes, European Commission—Brussels, Belgium. Harding Christopher, Department of Law, University of Wales— Aberystwyth, UK. Harrington Joseph, Jr., Department of Economics, The Johns Hopkins University—Baltimore, USA. Joshua Julian, Howrey LLP—Brussels, Belgium. Klawiter Donald, Morgan Lewis & Bockius LLP—Washington DC, USA. Komninos Assimakis, White & Case LLP—Brussels, Belgium.


16/10/07

16:11

Page viii

viii List of Participants at the 2006 Workshop Kovacic William, FTC Commissioner—Washington DC, USA. Lawrence Jon, Freshfields Bruckhaus Deringer—London, UK. Motta Massimo, Economics Department, European University Institute— Florence, Italy. Ratliff John, WilmerHale—Brussels, Belgium. Rey Patrick, Institut d’économie industrielle, Université des Sciences Sociales—Toulouse, France. Roth Peter, Monckton Chambers—London, UK. Siragusa Mario, Cleary Gottlieb—Rome, Italy. Ullrich Hanns, Universität der Bundeswehr München—Munich, Germany. Venit James, Skadden—Brussels, Belgium. Whish Richard, School of Law, King’s College—London, UK. Wils Wouter, Legal Service of the European Commission, Brussels, Belgium, and Visiting Professor at King’s College—London, UK.


16/10/07

16:11

Page ix

CONTENTS

List of Sponsors List of Participants Table of Cases

v vii xiii

INTRODUCTION

xxv

SESSION A: UNEARTHING CARTELS—THE EVIDENCE

1

PANEL I: OLIGOPOLY THEORY AND ECONOMIC EVIDENCE 3 Discussion 3 Working Papers 49 I Joseph E. Harrington, Jr: Behavioural Screening and the Detection of Cartels 51 II Patrick Rey: On the Use of Economic Analysis in Cartel Detection 69 III Paul A. Grout and Silvia Sonderegger: Structural Approaches to Cartel Detection 83 IV Richard Whish: Oligopoly Theory and Economic Evidence 105 V Rafael Allendesalazar, Paloma Martínez-Lage and Roberto Vallina: Oligopolies, Conscious Parallelism and Concertation 119 PANEL II: INVESTIGATING POWERS OF THE COMPETITION AUTHORITIES Working Papers I Thomas O. Barnett: Seven Steps to Better Cartel Enforcement II Philip Collins: Some Background Notes on the Investigative Powers of the Competition Authorities III Ian S. Forrester, QC: Searching Beneath the Cherry Tree in the Garden: European Thoughts on How to Enhance the Task of Uncovering And Thereby Deterring Cartels IV Calvin S. Goldman, Robert E. Kwinter and Angie L. Morris: The Investigative Powers of the Canadian Competition Bureau: Domestic and International Dimensions

139 139 141 153

161

191


x

16/10/07

16:11

Page x

Contents V

Olivier Guersent: The EU Model of Administrative Enforcement Against Global Cartels: Evolving to Meet Challenges

SESSION B: THE INSTITUTIONAL FRAMEWORK PANEL III: PUBLIC ENFORCEMENT (ADMINISTRATIVE AND CRIMINAL) Discussion Working Papers I Wouter P.J. Wils: Is Criminalization of EU Competition Law the Answer? II Nadia Calviño: Public Enforcement in the EU: Deterrent Effect and Proportionality of Fines III James S. Venit: Modernization and Enforcement—The Need for Convergence: On Procedure and Substance IV Stephen Calkins: Coming to Praise Criminal Antitrust Enforcement V Gregor Langus and Massimo Motta: On the Effect of EU Cartel Investigations and Fines: On the Infringing Firms’ Market Value SESSION C: TOOLS OF ENFORCEMENT PANEL IV: IMMUNITY/LENIENCY/FINANCIAL INCENTIVES/PLEA BARGAINING Discussion

213 221 223 223 265 267 317 337 343

363 377 379 379

PANEL V: PRIVATE ENFORCEMENT 403 Working Papers 423 I Denis Waelbroeck and Donald Slater: The Commission’s Green Paper on Private Enforcement: “Americanization” of EC Competition Law Enforcement? 425 II Eddy De Smijter and Donncadh Woods: The Commission Green Paper on Damages Actions for Breach of the EC Antitrust Rules 447 III Jon Lawrence: Seeking the Perfect Balance: Some Reflections on the Commission’s Green Paper on Damages Actions for Breach of EC Antitrust Rules 457 IV Mario Siragusa and Marco D’Ostuni: A Reflection on Some Private Antitrust Law Issues 477 PANEL VI: IMMUNITY/LENIENCY/FINANCIAL INCENTIVES/ PLEA BARGAINING Working Papers

487 487


16/10/07

16:11

Page xi

Contents I II III IV V VI

Donald C. Klawiter: US Corporate Leniency After the Blockbuster Cartels: Are We Entering a New Era? Julian M. Joshua: That Uncertain Feeling: The Commission’s 2002 Leniency Notice Margaret Bloom: Despite Its Great Success, the EC Lenience Programme Faces Great Challenges William E. Kovacic: Bounties as Inducements to Identify Cartels John Ratliff: Plea Bargaining in EC Anti-Cartel Enforcement—A System Change? Jochen Burrichter and Daniel J. Zimmer: Reflections on the Implementation of a “Plea Bargaining”/ “Direct Settlement” System In EC Competition Law

PANEL VII: ENSURING THAT EFFECTIVENESS OF ENFORCEMENT DOES NOT PREJUDICE LEGAL PROTECTION—RIGHTS OF DEFENCE Working Papers I Peter Roth QC: Ensuring that Effectiveness of Enforcement Does Not Prejudice Legal Protection. Rights of Defence. Fundamental Rights Concerns II Christopher Harding: Effectiveness of Enforcement and Legal Protection III Nicholas Forwood: Effective Enforcement and Legal Protection—Friends or Enemies BIBLIOGRAPHY

xi

489 511 543 571 597

611 625 625

627 643 661 665


16/10/07

16:11

Page xii


16/10/07

16:11

Page xiii

TABLE OF CASES I. Judgments of the EC Courts: Ahlström Osakeyhtiö et al. (Woodpulp II), Joined Cases C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85, [1993] ECR I-1307. ............................... 18, 23, 121, 123–6, 129, 396 Aalborg Portland v. Commission, Joined Cases C-204/00 P, C-205/00 P, C-211/00 P, C-213/00 P, C-217/00 P and C-219/00 P, [2004] ECR I-6453 ..................................................... 410, 656 Acerinox, Case C-57/02 P, [2001] ECR II-3859. ....................................... 654 AC-Treuhand v. Commission, Case T-99/04, judgment pending. .............. 182 Airtours v. Commission, Case T-342/99, [2002] ECR II-2585.................... 122 AKZO v. Commission [1991] ECR I-3359 ................................................ 437 Akzo-Nobel v. Commission, Case C-7/04 P(R), [2004] ECR I-8739........... 414 Akzo-Nobel v. Commission, Joined Cases T-125/03 R and T- 253/03 R, [2003] ECR II-4771.......................................... 406, 414, 419, 421, 636–7, 646 AM & S v. Commission, Case 155/79, [1982] ECR 1575.................... 406, 636 Amministrazione delle finanze dello Stato v. San Giorgio, Case 199/82 [1983] ECR 3595 ........................................................... 438–9 Archer Daniel Midland Co. v. Commission, Case T-224/00, [2003] ECR-II 2597. ............................................................................. 326 Archer Daniel Midland Co. and Archer Daniels Midland Ingredients Ltd v. Commission [2006] ECR I-4429 ................................ 418 Association of Convenience Stores v. OFT Case 1052/6/1/05 ..................... 117 AstraZeneca, Case 223/01, [2003] ECR I-11809 ....................................... 438 Automec v. Commission (Automec II), Case T-24/90 [1992] ECR II-2223......................................................................................... 606 BAI v. Bayer and Commission, Joined Cases C-2/01 P and C-3/01 P, [2004] ECR I-26.................................................................... 327 Banks, Case C-128/92, [1994] ECR I-1251. .............................................. 179 Boehringer Mannheim v. Commission, Case 45/69, [1970] ECR 769.......... 657 Brasserie du Pêcheur v. Bundesrepublik Deutschland and The Queen v. Secretary of State for Transport, ex parte: Factortame and Others, Joined Cases C-46/93 and C-48/93, [1996] ECR I-1029................................................................... 444, 479–80 British-American Tobacco Company Ltd (BAT) and R. J. Reynolds Industries Inc. v. Commission, Joined Cases 142 and 156/84, [1987] ECR 4487. .............................................. 128


xiv

16/10/07

16:11

Page xiv

Table of Cases

BRT v. SABAM (BRT I), Case 127/73, [1974] ECR 51. ........................... 447 Cascades v. Commission, Case C-279/98 P, [2000] ECR I-9709. ............... 277 Cimenteries CBR and Others v. Commission Joined Cases T-25/95 etc., [2000] ECR II-491. ................................................... 273, 656 Cipolla and Macrino, Joined Cases C-94/04 and C-202/04, [2006] ECR I-11421.............................................................................. 629 Commission v. SGL Carbon (Graphite Electrodes), Case C-301/04 P, [2006] ECR I-5915. ................................... 168, 175, 407, 629, 640, 642 Commission v. Anic Partecipazioni, Case C-49/82 P, [1999] ECR 4125. .... 655 Commission v. EU Council (Environmental Crimes), Case C-176/03, [2005] ECR I-2867. ........................................... 188–9, 233 Council v. Commission Case C-176/03, [2005] ECR I-7879 ................. 42, 314 Commission v. Italy, Case C-119/04, [2006] ECR I-6885. ......................... 629 Comateb and Others v. Directeur général des douanes and droits indirects, [1977] ECR I-165 ......................................................... 439 Compagnie Royale Asturienne des Mines SA (CRAM) and Rheinzink GmbH v. Commission, Joined Cases 29/83 and 30/83, [1984] ECR 1679................................................................. 125, 127 Consorzio Industrie Fiammiferi (CIF) v. Autoritá Garante della Concorrenza e del Mercato, Case C-198/01, [2003] ECR I-8055............ 329 Continental Can v. Commission, [1973] ECR 215...................................... 437 Courage Ltd v. Bernard Crehan, Case C-453/99, [2001] ECR I-6314.................................................. 180, 248, 263, 428, 444 Courage Ltd v. Bernard Crehan, Case C-453/99, [2001] ECR I-6297................................................................................... 448, 477 Danone v. Commission, Case T-38/02, [2005] ECR II-4407....................... 629 Dansk Rørindustri A/S and Others v. Commission, Joined Cases C-189/02 P, C-202/02 P, C-205/02 P, C-206/02 P, C-207/02 P, C-208/02 P and C-213/02 P, [2005] ECR I-5425. ...................... 178, 646–8, 652–3, 659 Edis, Case C-231/96, [1998] ECR I-4990. ................................................. 283 Elf Atochem v. Commission, Case T-9/97, [1997] ECR II-919................... 280 Enichem Anic v.Commission, Case T-6/89, [1991] ECR II-1695. ............... 277 Enso Española v. Commission, C-282/98 P, [2000] ECR I-9817. ............... 632 Enso Española v. Commission, [1998] ECR II-1875 .................................. 632 ERSA v. Ministero delle Politiche Agricole e Forestali, Case C-347/03, [2005] ECR I-03785. .................................................... 629 France v. Commission, Joined Cases C-68/94 and C-30/95, [1998] ECR I-1375................................................................................ 618 Gencor v. Commission, Case T-102/96, [1999] ECR II-753. ........................ 71 General Electric v. Commission, Case T-210/01, [2005] ECR II-5527........ 629 Gerhard Züchner v. Bayerische Vereinsbank AG, Case 172/80, [1981] ECR 2021. .......................................................................... 128, 132


16/10/07

16:11

Page xv

Table of Cases

xv

Germany v. Commission, Case 24/62, [1963] ECR 63................................ 619 Hercules Chemicals NV v. Commission, Case T-7/89, [1991] ECR II-1711. ............................................................................. 128 HFB and Others v. Commission, Case T-9/99, [2002] ECR II-1530. ................................................................................ 277, 631 Hoechst v. Commission, Joined Cases 46 and 227/88, [1989] ECR 2859. .................................................................................... 616, 633 Hoffmann-La Roche Ltd. v. Commission, [1979] ECR 461........................ 437 Hüls v. Commission, Case C-199/92 P, [1999] ECR I-4287................ 272, 632 Hydrotherm v. Compact, Case 170/83, [1985] ECR 3016. ......................... 276 Imperial Chemical Industries Ltd. (ICI) v. Commission, Case 48/69, [1972] ECR 619............................................... 122–4, 128, 132 JFE Engineering Corp. and Others v. Commission, Joined Cases T-68/00, T-71/00 and T-78/00, [2004] ECR II-2501...... 280, 529, 629 Kalfelis, Case 189/87, [1988] ECR 5565.................................................... 468 Krupp Thyssen Stainless and Acciai Speciali Terni v. Commission, Joined Cases T-45/98 and T-47/98, [2001] ECR II-3823. ...................... 279 Limburgse Vinyl Maatschappij and Others. v. Commission (PVC II), Joined Cases T-305/94, T-306/94, T-307/94, T-313/94, T-314/94, T-315/94, T-316/94, T-318/94, T-325/94, T-328/94, T-329/94 and T-335/94, [1999] ECR II-931. ...................... 127–9, 166, 277, 628–9, 640 Manfredi, Case C-295/04 [2006] ECR I-6619............................. 429, 440, 474 Mannesmannröhen-Werke v. Commission, Case T-112/98, [2001] ECR II-729. ................................................. 166, 407, 421, 621, 640 Martinelli v. Commission, Case T-150/89, [1995] ECR II-1165......................................................................................... 647 Masterfoods, Case C-344/98, [2000] ECR I-11369..................... 263, 442, 450 Matra-Hachette v. Commission, Case T-17/93, [1994] ECR II-595. ............................................................................... 524 Max-Mobil v. Commission, Case T-54/99, [2002] ECR II-313. ................. 629 Michelin v. Commission, Case T-203/01, [2003] ECR II-4071. .......... 279, 443 Microsoft v. Commission, Case T-201/04, not yet reported....................... 184 Mo Och Domsjo v. Commission, Case T- 352/94, [1998] ECR II-1989......................................................................................... 655 Montecatini v. Commission, Case C-235/92 P, [1999] ECR I-4539. ........... 632 National Pasnasonic v. Commission, [1980] ECR 2056 ............................. 280 Nederlandse Federatieve Vereniging voor de Groothandel op Elektrotechnisch Gebied v. Commission, Case C-105/04 P, [2006] ECR I-1185. ......................................................................... 629–30 Nintendo Co & Nintendo Europe v. Commission, Case T-13/03, judgment pending................................................................................. 178 Opel NL and GM NL v. Commission, Case T-368/00, [2003] ECR II-4491......................................................................................... 279


xvi

16/10/07

16:11

Page xvi

Table of Cases

Opinion 2/94 on Accession by the Community to the ECHR, [1996] ECR I-1759................................................................................ 628 Orkem SA v. Commission, Case 374/87, [1989] ECR3283................................................. 166-7, 406–7, 621, 638–41 Otto v. Postbank, Case C-60/92, [1993] ECR I-5683................................. 640 P&O European Ferries (Vizcaya) SA v. Commission, Case T-116/01, [2003] II-2957............................................................... 629 Rhône-Poulenc v. Commission, Case T-1/89, [1991] ECR II-869............... 276 Roche Nederland et al v. Frederick Primus et al, Case C-539/03, [2006] ECR I-6535. ....................................................................... 250, 468 Roquette Frères SA v. Directeur général de la concurrence, Case C-94/00, [2002] ECR I-9039. ...............................169, 405, 632, 634–5 Schöppenstedt v. Council, Case 5/71, [1971] ECR 983............................... 443 Sigma Tecnologie v. Commission, Case T-28/99, [2002] ECR II-1845......................................................................................... 526 Solvay v. Commission (Soda Ash), Case 27/88, [1989] ECR 3355. ............ 167 Solvay v. Commission (Soda Ash), Joined Cases T-30 to T- 36/91, [1995] ECR II-1175. .............................................................. 530 Stergios Delimitis v. Henninger Bräu, Case C-234/89, [1991] ECR I-935.................................................................................. 447 Suiker Unie v. Commission, Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114/73, [1975] ECR 1663. ........................ 120, 123–5, 128 Tate & Lyle v. Commission (British Sugar), Case T-202/98, [2001] ECR II-2035. ............................................................................. 125 Tatry, Case C-406/92, [1994] ECR I-5439. ............................................... 468 Thyssen Krupp and Thyssen Krupp Acciai Speciali Terni v. Commission, Joined Cases C-65/02 P and 73/02 P, [2001] ECR II-3757........................................................................... 654–5 Tokai Carbon v. Commission, Joined Cases T-236/01, T-239/01, T-244/01 to T-246/01, T-251/01 and T-252/01, [2004] ECR II-1181. ....................................... 175, 538, 640, 649 Tokai Carbon v. Commission, Joined Cases T-71/03, T-74/03, T-87/03, to T-91/03, [2005] ECR II-10.....................................166 Transocean Marine Paint Association v. Commission, Case 17/74, [1974] ECR 1063................................................................ 618 Van Gend & Loos, Case 26/62, [1963] ECR 1............................................ 451 Van Landewyck v. Commission, Joined Cases 209/78 etc., [1980] ECR 3125. ................................................................................. 632 Volkswagen AG v. Commission, Case C-338/00 P, [2003] ECR I-9189................................................................................ 619 Walt Wilhelm and Others v. Bundeskartellamt, Case 14/68, [1969] ECR 1. ............................................................................... 332, 657 Verein für Konsumenteninformation v. Commission, [2005] ECR II-1121 .............................................................................. 175


16/10/07

16:11

Page xvii

Table of Cases

xvii

II. Decisions of the European Commission: Acrylics, Decision of 31 May 2006, IP/06/698. ......................................... 533 Alloy Surcharge, OJ L 100 [1998]. ............................................................ 329 Austrian Banks – ‘Lombard Club’, OJ L 56 [2004]............................. 176, 327 Bank Charges for Exchanging Euro-Zone Currencies, OJ L 15 [2003]. ....................................................................................... 96 Belgian Architects Association, OJ l 4 [2005]. .................................... 329, 631 Bleaching Products, Decision of 3 May 2006, IP/06/560........................... 533 Carbon and Graphite, IP/03/1651.............................................................. 649 Cartonboard, OJ L 243 [1994]. ................................................................. 526 Choline Chloride, OJ L 190 [2005]..................................................... 256, 539 Citric Acid, OJ L 239 [2002]................................................................ 95, 175 Coca-Cola, OJ L 253 [2005]. ............................................................. 600, 614 De Beers, OJ L 205 [2006]. ....................................................................... 614 Ferry Operators — Currency Surcharges, OJ L 26 [1997]. .......................... 96 French Beef, OJ L 209 [2003]...................................................................... 96 French-West African Ship-Owners’ Committees, OJ L 134 [1992]. ..................................................................................... 97 German Bundesliga, OJ L 134 [2005]................................................. 600, 614 Graphite Electrodes, OJ L 100 [2002] .............................................40, 97, 649 Industrial Bags, IP/05/1508....................................................................... 180 Italian Cast Glass, OJ L 383 [1980]........................................................... 182 Microsoft, Case COMP/C-3/37.792, Decision of 23 April 2004, full text available at http://ec.europa.eu/comm/competition/ antitrust/cases/decisions/37792/en.pdf. ............................................. 183–4 Napier Brown - British Sugar, OJ L 284 [1988]. ........................................ 329 Needles, Decision of 26 October 2004, text available at http://ec.europa.eu/comm/competition/antitrust/cases/ decisions/38338/en.pdf. ........................................................................ 539 Optical Fibres, OJ L 236 [1986]............................................................. 128–9 Petrochemicals, OJ L 239 [1994]................................................................. 97 Pioneer, OJ L 60 [1980]. ........................................................................... 529 PO/Interbrew, OJ L 200 [2003]. ........................................................ 186, 329 PO-Michelin, OJ L 142 [2003]. ................................................................. 328 Pre-Insulated Pipes Cartel, OJ L 24 [1999].............. 277, 328–9, 526, 646, 651 Raw Tobacco Italy, OJ C 60 [2006]. ....................................... 533–4, 539, 549 Rubber Chemicals, Decision of 21 December 2005, IP/05/1656. ............................................................................ 180, 533, 549 Seamless Steel Tubes, OJ L 382 [2003]. ...................................................... 97 Vegetable Parchment, OJ L 70 [1978]. ...................................................... 128 Vitamins, OJ L 6 [2003]. ............................................................ 326, 328, 518 Zinc Producer Group, OJ L 220 [1984]. .................................................... 124


16/10/07

16:11

Page xviii

xviii Table of Cases

III. US Cases: Advanced Micro Devices, Inc. v. Intel Corp., No. C 01-7033, 2004 WL 2282320 (N.D. Cal., Oct. 4, 2004). ........................................ 564 American Column & Lumber Co. v. United States, 257 U.S. 377 (1921)..................................................................................... 354 Apperson v. Fleet Carrier Corp., 879 F.2d 1344, 1352 (6th Cir. 1089). ..................................................................................... 584 Associated Press v. United States, 326 U.S. 1 (1945)................................. 353 Blackledge v. Allison, 431 U.S. 63, 76 (1977). ........................................... 611 Blomkest Fertilizer, Inc. v. Potash Corp. of Saskatchewan, 203 F.3d 1028 (8th Cir. 2000). .............................................................. 357 Blue Shield of Virginia v. McCready, 457 U.S. 465, 472 (1982)................. 171 Braswell v US, 487 US 99, 108 SCt 2284 (1988). Broadcast Music, Inc. v. CBS, Inc., 441 U.S. 1 (1979). ............................. 355 Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993)....................................................................... 355, 360 Brown Shoe Co. v. United States, 370 U.S. 294 (1962).............................. 353 Brunswick Corp. v. Pueblo Bowl-O-Mat Inc., 429 U.S. 477 (1977)............ 354 Business Electonics Corp. v. Sharp Electronics Corp., 485 U.S. 717 (1988)....................................................................... 355, 359 California Dental Ass’n v. FTC, 526 U.S. 756 (1999)..................... 355–6, 358 Chicago Board of Trade v. United States, 246 U.S. 231 (1918).................. 354 City of Columbia v. Omni Outdoor Advertising, Inc., 499 U.S. 365 (1991). ............................................................................. 356 Confederated Tribes of Siletz Indians of Or. v. Weyerhaeuser Co., 411 F.3d 1030 (9th Cir. 2005). .............................................................. 357 Continental Airlines, Inc. v. United Airlines, Inc., 277 F.3d 499 (4th Cir. 2002). ..................................................................................... 358 Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36 (1977). ......... 354–5 Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752 (1984). ............................................................................. 355 Dell Computer Corp., 128 F.T.C. 151 (1999). ........................................... 357 Eastern States Retail Lumber Dealers’ Ass’n v. United States (1914). .................................................................................................. 353 F. Hoffman-La Roche Ltd. v. Empagran S.A., 542 U.S. 155 (2004).................................................................... xxiii, xxv, 171, 256–7, 355, 562 FTC v. Arch Coal, Inc., 329 F. Supp. 2d 109 (D.D.C. 2004). .............. 71, 357 FTC v. Ethyl Corporation, [1979-1983 Transfer Binder] T.R.R. (CCH) 22,003 (1983).................................................... 153 FTC v. Indiana Federation of Dentists, 476 U.S. 447 (1986). ............. 356, 358 FTC v. Mylan Labs. Inc., 62 F. Supp. 2d 25 (D.D.C. 1999) ..................... 275


16/10/07

16:11

Page xix

Table of Cases

xix

FTC v. Superior Court Trial Lawyers Ass’n, 493 U.S. 411 (1990). .................................................................. 355–6, 358 FTC v. Ticor Title Insurance Co., 504 U.S. 621 (1992). ............................ 356 Goldfarb v. Virginia State Bar, 421 U.S. 773 (1975) ................................. 259 Great Atlantic & Pac. Tea Co. v. FTC, 440 U.S. 69 (1979). ...................... 355 Hale v. Henkel 201 U.S. 43, (1906)........................................................... 631 Hartford Fire Ins. Co. v. California, 509 U.S. 764 (1993). ..................... 355–6 Hickman v. Taylor, 329 U.S. 495 (S. Ct. 1947). ........................................ 530 Illinois Brick Co v. Illinois, 431 U.S. 720 (1977) ................................ 439, 461 Illinois Tool Works Inc. v Independent Ink, Inc., 126 S. Ct. 1281 (2006). .................................................................. 355, 358 Image Technical Services v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir. 1997). ..................................................................................... 358 In re Application of Microsoft, Case N° 06-80038 JF (PVT), (ND,Ca). .............................................................................................. 531 In re Application of Microsoft, MBD N° 06-10061-MLW, (Mass.) 17 April 2006. .......................................................................... 531 In re Methionine Antitrust Litigation, 221 F.R.D. 1 (N.D.Cal. 2002). ...... 564 In Re Methionine Antitrust Litigation, No. C-99-3491-CRB N.D.Cal. .............................................................................................. 174 In re Sulfuric Acid Antitrust Litig. (N.D. Ill. July 7, 2005)........................ 505 In re Vitamins Antitrust Litigation, 217 F.R.D. 229 (D.D.C. 2002). .............................................................................. 174, 564 Intel Corp.v. Advanced Micro Devices, Inc., 542 U.S. 241 (2004). ............................................................................ xxv, 564 Intel Corp., Dkt. No. 9288, 1999 FTC LEXIS 38 (consent order 1999)............................................................................. 357 International Salt Co. v. United States, 332 U.S. 392 (1947). .................... 354 Interstate Circuit, Inc. v. United States, 306 U.S. 208 (1939). ................... 354 ISO Antitrust Litigation, 203 F.3d 1322 (Fed. Cir. 2000). ........................ 358 Joblove v. Barr Laboratories, Inc., 429 F.3d 370 (2d Cir. 2006). ............... 358 LePage’s, Inc. v. 3M, 324 F.3d 141 (3d Cir. 2003).................................... 357 Lorain Journal Co. v. United States, 342 U.S. 143 (1951). ........................ 354 McCormick & Co., 2000 FTC LEXIS 43 ................................................. 358 Maple Flooring Mfrs. Ass’n v. United States (1925).................................. 354 Matsushita Electric Indus. Co. v. Zenith Radio Corp., 475 U.S. 574 (1986)....................................................................... 355, 359 National Society of Professional Engineers v. United States, 435 U.S. 679. (1978).. .................................................................... 356, 358 NCAA v. Board of Regents of University of Georgia, 468 U.S. 85 (1984)......................................................................... 355, 358 North Texas Specialty Physicians, FTC Dkt. 9312................................... 358 Northwest Wholesale Stationers, Inc. v. Pacific Stationery and Printing Co., 472 U.S. 284 (1985). ........................................................ 355


xx

16/10/07

16:11

Page xx

Table of Cases

NYNEX Corp. v. Discon, Inc., 525 U.S. 128 (1998). ................................. 355 Polygram Holding, Inc. v. FTC, 416 F.3d 29 (D.C. Cir. 2005). ......... 351, 358 Professional Real Estate Investors, Inc. v. Columbia Pictures Industries, Inc., 508 U.S. 49 (1993).................................................... 355–6 Rambus Inc, Dkt 9302 (FTC Aug. 2, 2006) .............................................. 358 Re SGL Carbon Corporation, 3rd Cir. (1999). .......................................... 650 RX Care of Tenn., Inc., 121 F.T.C. 762 (1996). ........................................ 358 Santobello v. New York, 404, U.S. 257 (1971). .................................. 606, 611 Schering-Plough Corp. v. FTC, 402 F.3d 1056 (11th Cir. 2005). ........... 357–8 Schmitz v. Bernstein Liebhard & Lifshitz, LLP, 376 F.3d 79 (2d Cir. 2004). ...................................................................................... 564 Siegel v. Chicken Delight 488 F.2d 43 (9th Cir. 1971) ............................... 436 Southern Motor Carriers Rate Conference v. United States, 471 U.S. 48 (1985). ............................................................................... 356 Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447 (1993). ........................ 355 Spirit Airlines, Inc. v. Northwest Airlines, Inc., 431 F. 3d 917 (6th Cir. 2005). ..................................................................................... 357 State Oil Co. v. Khan, 522 U.S. 3 (1997). .......................................... 355, 358 Stone Container Corp., 125 F.T.C. 853 (1998).......................................... 357 Stop & Shop Supermarket Co. v. Blue Cross & Blue Shield, 373 F.3d 57 (1st Cir. 2004). .................................................................. 358 Sugar Institute, Inc. v. United States (1936).............................................. 354 Summit Health, Ltd. v. Pinhas, 500 U.S. 322 (1991). ................................ 355 Texaco Inc. v. Dagher, 126 S. Ct. 1276 (2006). ......................................... 355 Texaco v. Hasbrouck, 496 U.S. 543 (1990). .............................................. 355 Time Warner Inc, 2000 FTC LEXIS 40.................................................... 358 Timpkin Roller Bearing Co. v. United States (1951).................................. 354 Toys “R” Us, Inc. v. FTC, 221 F.3d 928 (7th Cir. 2000). ...................... 357–8 Bell Atlantic Comp. v. Twombly 1265 S. Ct. 2965 (2006)................... 355, 357 United States ex rel Herbert v. Nat’l Acad. of Sciences, 1992 WL 247587 (D.D.C. 1992). .......................................................... 575 United States v. Addyston Pipe & Steel Co., 85 F. 271 (6th Cir. 1898). ..................................................................................... 354 United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir. 1945). ...................................................................................... 354 United States v. American Linseed Oil Co. (1923)..................................... 354 United States v. American Tobacco Co., 328 U.S. 781 (1946). .................. 354 United States v. AMR Corp., 335 F.3d 1109 (10th Cir. 2003)............ 353, 356 United States v. Andreas 216 F.3d 645 ..................................................... 579 United States v. Arnold, Schwinn & Co., 388 U.S. 365 (1977)............ 354, 436 United States v. Arthur Anderson, LLP, 334 F.3d 281 (5th Cir. 2004), reversed, 544 U.S. 696 (2005). ...................................... 347 United States v. Booker, 543 U.S. 220 (2005)............................. 146, 186, 304 United States v. Brown University, 5 F. 3d 658 (3d Cir. 1993)................... 358


16/10/07

16:11

Page xxi

Table of Cases

xxi

United States v. Colgate & Co., 250 U.S. 300 (1919)................................. 354 United States v. Container Corp. of America, 393 U.S. 333 (1969). ..... 26, 354 United States v. Cuisinart, Inc 1981–1 Trade Cas. (CCH) ........................ 359 United States v. Dairy Farmers of Am., Inc., 426 F.2d 850 (2005). ........... 353 United States v. Dentsply International, Inc., 399 F.3d 181 (3d Cir. 2005)................................................................................ 353, 357 United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377 (1956)..................................................................................... 354 United States v. General Dynamics Corp., 415 U.S. 486 (1974)................. 353 United States v. Grinnell Corp., 384 U.S. 563 (1966). ............................... 354 United States v. LSL Biotechnologies, 379.3d 672 (9th Cir. 2004)............. 353 United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001)..................................................................... 353, 355, 357 United States v. Oracle Corp., 331 F. Supp. 2d 1098 (N.D. Cal. 2004)................................................................................... 357 United States v. Penn-Olin Chemical Co., 378 U.S. 158 (1964). ................ 353 United States v. Philadelphia National Bank, 374 U.S. 321 (1963). ........... 353 United States v. Rose, 2006 U.S. App. LEXIS 11679 (5th Cir. 2006)........ 357 United States v. Sealy, Inc. (1967). ........................................................... 354 United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940). ................ 354 United States v. Sungard Data Systems, Inc., 172 F. Supp. 2d 172 (D.D.C. 2001). ..................................................................................... 357 United States v. Therm-All, Inc., 373 F.3d 625 (5th Cir. 2004). ................ 357 United States v. Topco Associates, Inc., 414 U.S. 801 (1973). ................... 353 United States v. Trans-Missouri Freight Ass’n, 166 U.S. 290 (1897). ........ 354 United States v. United States Gypsum Co., 438 U.S. 422 (1978). ..... 356, 360 United States v. VISA USA, Inc., 344 F.3d 229 (2d Cir. 2003). ................ 353 Valassis Communications, Inc., Dkt. C-4160, 2006 FTC LEXIS 25 (2006). ................................................................................. 357 Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, 540 U.S. 398 (2004)...................................................................... 141, 238, 355, 357 Volvo Trucks North America, Inc. v. Reeder-Simco GMC, Inc., 126 S. Ct. 860 (2006). .................................................................... 355, 358 Weyerhauser Co v. Ross-Simmons Hardwood Lumgber Co., 126 S. Ct. 2965 (2006)........................................................................... 355 White v. US, 322 U.S. 694 (1911). ............................................................ 277

IV. Judgments of the European Court for Human Rights (ECHR): Albert and Le Compte v. Belgium [1983] ECHR 1. ................................... 273 Bendenoun v. France [1994] 18 ECHR 54. ......................................... 271, 273 Bosphorus v. Ireland, [2005] ECHR 30. .................................................... 629


16/10/07

16:11

Page xxii

xxii Table of Cases Bryan v. UK [1996] 21 ECHR 342. ........................................................... 632 Buck v. Germany, Application no. 41604/98, ECHR judgment of 28 April 2005.................................................................................... 636 Cantoni v. France [1996] ECHR 52. .................................................. 653, 659 De Cubber v. Belgium, [1984] 7 ECHR 236. ............................................. 273 Engel and Others v. the Netherlands, ECHR judgment of 8 June 1976, Series A no. 22. ................................................................ 271 Findlay v. United Kingdom, [1997] 8 ECHR 221, Reports 1997-I. ............ 273 Foley v. United Kingdom, Appl no. 33274/96, judgment of 20 June 2000......................................................................................... 637 Funke v. France, A/256-A [1993] 16 ECHR 297 ........................................639 Goodwin v. United Kingdom, Appl no. 25680/94, judgment of 11 July 2002 ......................................................................................... 629 JB v. Switzerland, [2001] 12 ECHR 281 ................................................... 639 OOO Neste St. Petersburg and Others v. Russia, ECHR judgment of 1 February 1983, text available at http://www.echr.coe. int/Eng/Judgments.htm................................................................. 272, 632 Özturk v. Germany, ECHR judgment of 21 February 1984, Series A no. 73. ............................................................................. 271, 273 Saunders v. United Kingdom, [1997] 23 ECHR 313. ............................ 639–40 Schmautzer v. Austria, [1995] 21 ECHR 511. ........................................... 633 Senator Lines GmbH v. the 15 Member States of the European Union, Appl no. 56672/00, text of the judgment available at http://www.echr.coe.int/Eng/Judgments.htm........................................ 272 Société Colas Est and Others v. France, Application no. 37971/97, ECHR 2002-III. .............................................. 405, 633, 635 Stenuit v. France, A/232-A, judgment of 11 July 1989. ............................. 272 Umlauft v. Austria [1996] 22 ECHR 76. ................................................... 633

V. National jurisdictions (miscellaneous): Canada British Columbia Securities Commission v. Branch [1995] 2 S.C.R. 3 ........ 196 Canada (Commissioner of Competition) v. Air Canada ............................ 194 Edwards [1999] O.J. No. 3819 (Sup. Ct.). ................................................. 200 Hudson’s Bay Co. v. Canada v. Canada (Director of Investigation and Research) (1992), 10 O.R. (3d) (C.A.). .......................................... 197 Irvine v. Restrictive Trade Practices Commission, [1987] 1 S.C.R. 181. ..... 196 Mutual Legal Assistance in Criminal Matters Act, R.S.C. 1985, c. 30 (4th Supp.) and SOR/90-704. .................................. 206 R. v. Grant, [1993] 3 S.C.R. 223 at 241-242. ............................................. 198 R. v. Jarvis, [2002] 3 S.C.R. 757. .............................................................. 196


16/10/07

16:11

Page xxiii

Table of Cases

xxiii

R. v. Ling [2002] 3 S.C.R. 814. ................................................................. 196 R v. Litchus [1997] 3 S.C.R. 320 (SCC) .................................................... 192 R. v. McClure, [2001] 1 S.C.R. 445........................................................... 202 Southam, [1984] 2 S.C.R. 145. .................................................................. 197 Germany BGH, Case GSSt 1/04, Entscheidungssammlung des Bundesgerichtshofes in Strafsachen (BGHSt) 50, 40 = Neue Juristische Wochenschrift (NJW) 2005, 1440, 1443 = Neue Zeitschrift für Strafrecht (NStZ) 2005, 389. ..................... 612 Bundesverfassungsgericht, Case 2 BvR 1133/86, Neue Juristische Wochenschrift (NJW) 1987, 2662 = Neue Zeitschrift für Strafrecht (NStZ) 1987, 419. ................................................................................. 616 Bundesgerichtshof (BGH), Case 4 StR 240/97, Entscheidungssammlung des Bundesgerichtshofes in Strafsachen (BGHSt) 45, 193 = Neue Juristische Wochenschrift (NJW) 1998, 86. ........................................... 617 BGH, Case 4 StR 240/97, Entscheidungssammlung des Bundesgerichtshofes in Strafsachen (BGHSt) 45, 193 = Neue Juristische Wochenschrift (NJW) 1998, 86. ........................................... 618 Spain Aceites, judgment of the TDC of 8 July 1992, Case 294/91. ..................... 135 Asturiana de Zinc, judgment of the TDC of 9 January 1987, Case 223/89. .......................................................................... 131, 134, 136 Autoescuelas Alcalá, judgment of the Spanish Audiencia Nacional of 11 April 2003..................................................................... 332 Autoescuelas Collado Villalba, judgment of the TDC of 18 December 1998, Case 421/97. .......................................................... 134 Autoescuelas Extremadura, judgment of the TDC of 16 February 2005, Case 582/04............................................................. 133 Autoescuelas, judgment of the TDC of 25 October 1988, Case 237/88. ......................................................................................... 135 Azúcar, judgment of the TDC of 15 April 1999, Case 426/98................... 134 BT/Telefónica, judgment of the Spanish Supreme Court of 23 March 2005. ............................................................................. 328, 329 Cider cartel, judgment of the Spanish Audiencia Nacional of 6 May 1999. ......................................................................................... 329 Detergentes concentrados, judgment of the TDC of 6 March 1992, Case 306/91. ............................................... 131, 133, 136–7 Distribuidores cine, judgment of the TDC of 10 May 2006, Case 588/05. ......................................................................................... 136 Driving schools of Tenerife, judgment of the Spanish Audiencia Nacional of 22 September 2003. ........................................................... 331


xxiv

16/10/07

16:11

Page xxiv

Table of Cases

Española de Zinc, judgment of the TDC of 18 July 1986, Case 223/86................................................................................... 131, 134 Estate Agents, judgment of the Spanish Audiencia Nacional of 16 October 2003. .................................................................................. 328 Hormigón Gerona, judgment of the TDC of 4 June 2001, Case 492/00....................................................................................... 134–5 Películas de vídeo, judgment of the TDC of 11 May 1998, Case 387/96. ......................................................................................... 134 Prensa del corazón, judgment of the TDC of 2 November 1988, Case 239/88. ......................................................................................... 135 Prensa dominical, judgment of the TDC of 4 July 1988, Case 235/87....................................................................................... 134–5 Servicios Agua Tenerife, judgment of the TDC of 23 September 2004, Case 567/03. ......................................................................................... 134 Suresa/Correos, judgment of the Spanish Audiencia Nacional of 26 September 2005. .............................................................................. 327 Tabacalera, judgment of the Spanish Audiencia Nacional of 13 December 2004. ............................................................................... 329 Tipos de Interés Bancario, judgment of the TDC of 5 July 1996, Case R 144/96. ..................................................................................... 136 Transmediterránea/Euroferrys/Buquebús, judgment of the TDC of 13 June 2003, Case 543/02. ................................................................... 134 Vacunas Antigripales, judgment of the TDC of 30 September 1998, Case 395/97. ......................................................................................... 134 United Kingdom Arkin v. Borchard Lines [2003] Comm EWHC 2844................................. 433 Arkin v. Borchard Lines [2005] EWCA Civ 655 ........................................ 433 Courage v. Crehan [2001] All ER (EC) 886................................ 180, 452, 480 Office of Fair Trading v. X, [2003] 2 All ER (Comm) 183. ................ 635, 639 Provimi Ltd v. Aventis Animal Nutrition SA [2003] 2 All ER (Comm) 683............................................................ 250, 258, 467 R (Morgan Grenfell) v. Special Commissioner [2003] 1 AC 563................ 637 R v. Derby Magistrates Court, Ex p B [1996] AC 487, 508. ...................... 637 Sevenoaks Survey, UK Office of Fair Trading Press Releases 88-06, 19 May 2006, and 214/05, 9 November 2005, available at www.oft.gov.uk................................................................. 602 Three Rivers v. Bank of England (No 6) [2005] 1 AC 610......................... 637

(B) Ehlermann 07 Intro

16/10/07

13:58

Page xxv

I N T RO D U C T I O N This volume contains the written contributions prepared for the 11th edition of the Annual EU Competition Law and Policy Workshop, held on 2–3 June 2006 at the Robert Schuman Centre for Advanced Studies of the European University Institute in Florence, as well as the (edited) transcripts of the debate that took place on this occasion. Background. The EU Competition Law and Policy Workshop is a multiannual programme exploring topical policy and enforcement issues in the area of EC competition law. Each year the programme brings together a group of top-level EU and international policy-makers, judges, legal practitioners, economic consultants and academics for a one-and-a-half-day intensive roundtable debate in an informal and non-commercial environment. The 11th edition of the Annual EU Competition Law and Policy Workshop was devoted to the enforcement of the prohibition of cartels.1 The choice of this topic is easily explained. The fight against cartels has long been the top priority for the competition authorities in the US and Canada. During the 1990s, this fight also became the top priority for the European Commission. The reform of the leniency programme in 2002 has given the Commission a rather efficient tool for the detection of cartels. The modernization of EU competition policy by means of Regulation 1/2003, which abolished both the Commission’s exemption monopoly under Article 81(3) EC and the corresponding notification requirement, has allowed the Commission to allocate the necessary human resources to the detection, prosecution and fining of cartels. To a certain extent, the Commission has even become the victim of its own success: nowadays there are more leniency applications than the Commission is able to “digest”, that is to say, more applications are received than can be transformed into binding decisions which sanction the wrong-doers.2 1 Detailed information about the previous editions of the Workshop, and about the European Competition Law Annuals series, is available on the EUI website (see www.eui.eu/RSCAS/ Research/Competition/). The written contributions prepared for the more recent editions of the Workshop can be downloaded freely from the same website. 2 This Introduction follows the structure of the Workshop as indicated in the original programme, and sketches the essential content of the written contributions and the oral discussions of the six sessions in the order in which the proceedings unfolded. However, there are a few subjects which were addressed in several sessions despite being formally assigned to only one of them. The most obvious of these recurring themes is leniency. Another is the policy and practice of the Commission with respect to the level of fines in general and in individual cases. Even the compensation of victims of cartels is referred to in several sessions. We have resisted the temptation to regroup all these references in separate, specifically dedicated sections of this Introduction, so as to maintain the “historical” truth of the Workshop. The reader who is particularly interested in these subjects will therefore have to look at the Introduction and the book as a whole.


xxvi

16/10/07

13:58

Page xxvi

Introduction

It is therefore not surprising that the conclusion of the introductory paper submitted by Neelie Kroes, the Member of the European Commission responsible for Competition, addresses improvements of the existing leniency programme and the introduction of direct settlements, comparable to the US system of plea bargaining. The improvements envisaged for leniency consist of: clarifications of the threshold for immunity to encourage more precise and substantiated leniency applications; further specification of the duty of cooperation; and the adoption of rules on an oral procedure for corporate statements. The objective of a direct settlement procedure will be to reduce the duration of the administrative procedure leading to a binding Commission decision, thus releasing enforcement resources for the efficient pursuit of other cartel cases. For the Commissioner, such procedural “savings” must take place either before or upon the issuing of the Statement of Objections. At the same time, she warns that the margin for the reduction of fines in direct settlement procedures will be limited. Structure of the Workshop. Discussions at the Workshop were divided into six sessions. In what follows, we provide an overview of the main points that surfaced from the presentation of the written contributions and the subsequent debate.

1. Session One: Oligopoly Theory and Economic Evidence a. Written contributions Under the broader heading “Unearthing Cartels—the Evidence”, the first session of the Workshop discussed “Oligopoly Theory and Economic Evidence”. The discussions were introduced by five papers. Paul A. Grout’s oral presentation at the Workshop was based on the results of a study undertaken in 2005 (together with Silvia Sonderegger, University of Bristol) for the Office of Fair Trading, entitled “Predicting Cartels”.3 The study examined the economic and structural factors that are typically considered to contribute to the formation, maintenance or detection of cartels. Based on economic theory, case studies and econometric analysis, the OFT study came to the conclusion that, to begin with, three fundamental characteristics facilitate the formation of a cartel, i.e. homogeneous products, stable output and market conditions, and large and stable market players. In 3 Grout P. and Sonderegger S. (2005): Predicting Cartels, OFT Economic Discussion Paper, March 2005. Following the Workshop, the authors prepared a written contribution—reproduced in this volume—which updates and expands upon the outcomes and conclusions of the 2005 OFT study.


16/10/07

13:58

Page xxvii

Introduction

xxvii

addition to these basic requirements, certain other factors facilitate collaboration, i.e. transparency, payroll effects (high payroll per employee), concentration, barriers to entry and excess capacity. Finally, it is important to ask the questions “why and when”. Cartel formation is more likely if there has been a long-run decline in demand and/or of prices affecting all or almost all companies in the market, and where there is a sudden market shock that affects all companies in the market. The Grout and Sonderegger study of 2005 shows not only the well-known concentration of cartels detected thus far in certain economic sectors but also the strong correlation between those cartels on both sides of the Atlantic. While this is explainable in terms of structural and behavioural factors present in the US as well as in the EU, the strong correlation could also be explained in terms of some sort of “hysteresis”: the detection of a cartel by one authority leads other competition authorities to concentrate their efforts on the same activities in their respective jurisdictions. More interesting as guidance for the future, however, is Grout and Sonderegger’s list of sensitive sectors in which, so far, no cartels have been found. While the Grout and Sonderegger study concentrates on structural characteristics, the paper of Joseph E. Harrington turns to behavioural factors. He identifies a series of collusive markers. Collusive price markers are: (i) increased price uniformity: a higher list (or regular) price and reduced variation in prices across customers; (ii) a series of steady price increases preceded by steep price declines; (iii) price rises and import declines; (iv) strongly correlated prices of different firms (e.g., in bidding auctions, where bids are clustered); (v) a high degree of uniformity across firms in product price, and other dimensions including the price of ancillary services; (vi) low price variance; and (vi) price regime switches (collusion phases followed by punishment phases). Collusive quantity markers are: (i) market shares that are highly stable over time; and (ii) the presence of a subset of firms in which each firm’s share of total supply is highly stable over time. Patrick Rey’s paper overlaps with those of Paul Grout and Joseph Harrington. He discusses both structural and behavioural factors which facilitate or might otherwise indicate collusion. He notes that, while it would be natural to look for supra-competitive prices or profits (generated by a cartel), such an approach would require a deep knowledge and expertise of the examined industries—which regulators are supposed to have but competition authorities are not. Rey also notes that courts have been reluctant to take suspect behavioural patterns as evidence of collusion: indeed, price parallelism, market share stability and price reactions to changing demand conditions can all be compatible with competitive market conditions. All three authors consider structural and behavioural factors to be useful complements to the existing leniency programmes. In particular, Joseph Harrington emphasizes that, the greater the risk of discovery and the higher/more painful the penalties, the more effective leniency will be.


16/10/07

13:58

Page xxviii

xxviii Introduction “Amnesty plus” and “penalty plus” (generated through the so-called “omnibus question”) contribute to this risk, but so does the danger of ex officio scrutiny by a competition authority. The screening of certain economic sectors in which cartels are more likely to exist than in others will therefore reinforce the efficiency of leniency programmes. Richard Whish’s paper describes two instruments made available by UK competition law in cases where competitors are behaving in a parallel manner but where there is insufficient evidence, if any, of a cartel. The Office of Fair Trading (OFT) might itself conduct a “market study”; or it might also (like various sectoral regulators) make a “market investigation reference” to the Competition Commission (CC). Market investigation references can lead directly to the imposition of remedies, whereas sectoral investigations under Article 17 of Regulation 1/2003 cannot. Market investigation references are therefore a useful complement to Articles 81 and 82 EC and the instruments available to enforce them. Rafael Allendesalazar’s paper 4 discusses EC and Spanish competition law, as well as the attitude of European Courts and the Tribunal de Defensa de la Competencia (TDC), with respect to concerted practices on the one hand and conscious parallel behaviour on the other. The paper criticizes the fact that, in spite of the different treatment of the two situations under EC and Spanish law, the TDC has not acknowledged the fundamental difference between them. Spanish competition law should not, by virtue of presumptions, be applied to behaviours that are unilateral, logical, rational and competitive answers to market conditions.

b. Discussion The discussion among participants during the first session concentrated on the three economics papers dealing essentially with the ex officio detection of cartels. The answer to the question whether economists can assist in such ex officio detection was a qualified “yes”. Everybody agreed that the identification and quantification of structural and behavioural factors facilitating or signalling the existence of a cartel is not an appropriate method to identify collusion per se. But screening is a useful tool for competition authorities to prioritize, either in determining which sectors should be looked at more closely in order to find a cartel, or in choosing among leniency applications in order to determine which “tip off” to pursue first. That position was clearly confirmed by at least one of the major cartel enforcers, while others showed some scepticism, in particular in view of the potential high costs of monitoring certain sectors (e.g. the gasoline industry in the US). 4

Co-authored by Paloma Martínez-Lage and Roberto Vallina.


16/10/07

13:58

Page xxix

Introduction

xxix

Participants agreed that screening is a useful complement to leniency. The efficiency of a leniency programme depends on the risk of detection and punishment. Ex officio screening contributes to this risk: it enhances the fear of discovery of a cartel. This conclusion was nicely encapsulated in the formula “you cannot catch a thief with an economist, but you can scare him”. There was also consensus that screening methods should be refined through further research based both on theoretical insights and on case studies. Sectoral investigations are a useful starting point and have led DG Competition to prioritize its efforts (e.g. in undertaking dawn raids). The relationship (complementarity) between ex officio screening and leniency programmes invited several participants, in particular practicing lawyers, to reflect on their experiences with the existing leniency programmes. It was recognized by all that leniency is an efficient tool for destabilizing cartels. Reference was made to the complementarity of corporate and individual leniency and to the usefulness of a system of bounties. At the same time, it was emphasized that the effectiveness of leniency depends on proper incentives, and that hurdles discouraging would-be applicants from applying (such as the “omnibus question” and “penalty plus” in the US) should therefore be avoided. Interveners agreed that leniency applications can be used as a strategic tool to harm competitors, notably in a legal system (such as in the EC) based exclusively on fines. Such harm is particularly dangerous for smaller, financially weaker firms because of the relatively stronger impact of fines on such companies. While practicing lawyers seemed to regret this type of strategic use, representatives of competition authorities stressed their indifference visà-vis the motives of the whistleblower. For competition law enforcers, everything that destabilizes a cartel is good; leniency applicants are therefore welcome. For the enforcer, what is more important is the kind of cartel the whistleblower reveals. A young, healthy and well-functioning cartel is a more interesting target for prosecution than an old, dying one. A certain number of leniency applicants might well be members of the second group, a consideration which underlines the usefulness of increasing the fear of detection through increased, efficient ex officio screening. One participant raised the question of the importance of corporate culture and the changes likely to affect it. The increased sensitivity towards corporate wrong-doing in other sectors of the law (such as accounting) could have a major impact, but the discussion about its effects, in particular in Europe, were (regrettably) not pursued in detail. However, in the US, the attitude notably towards document retention has changed discernably.


xxx

16/10/07

13:58

Page xxx

Introduction

2. Session Two: Investigative Powers of the Competition Authorities Written contributions The five papers presented in this Session reflect the situation in the US, Canada, the EU and the UK. Thomas O. Barnett’s contribution sets out seven steps to better cartel enforcement. These seven steps are: (i) a focus on the detection, prosecution and punishment of hard core cartels (as opposed to the investigation of rule of reason or monopolization cases—with consequences well illustrated in Stephen Calkins’ contribution to the third session of the Workshop); (ii) the treatment of cartels as serious crimes and of cartel members as criminals, facilitated by the substantial increase in maximum jail term and criminal fines in recent US legislation, putting antitrust crimes on a par with other economic crimes such as bribery, bank fraud, and mail and wire fraud; (iii) amnesty and “amnesty plus” programmes, the efficiency of which requires significant benefits, a genuine fear of detection, predictability and transparency; (iv) the vigorous prosecution of obstruction of justice; (v) the combination of cartel offences with other violations of the law; (vi) predictability and transparency of the enforcement policy; and (vii) wide publicity of anti-cartel enforcement efforts. Calvin S. Goldman’s contribution5 discusses cartel enforcement in Canada. The section dealing with investigative powers highlights constitutional problems such as: the protection against self-incrimination; the issue of extraterritoriality in computer searches (where Canada seems to be less restrained than the US!); the recent authorization of wire-tapping; and the main elements of amnesty and leniency programmes. The section on international cooperation in antitrust investigations sketches the origins and the evolution of cooperation with the US, which Canada intends to extend to other jurisdictions; it also refers briefly to the work under way in the framework of the OECD. The concluding section confirms that the broad array of investigative tools available, the dramatic rise in international cooperation and the efficiency of leniency programmes signal the ever-growing likelihood that antitrust violations will be detected. The conclusion also stresses the growing need for coordinated responses by the defence in order to match the increased coordination between enforcement agencies. For Olivier Guersent, the EU enforcement system ensures effective and timely investigation and decision making. Efficiency has increased since the 5

Co-authored by Robert E. Kwinter and Angie L. Morris.


16/10/07

13:58

Page xxxi

Introduction

xxxi

creation of a cartel directorate, but the Commission’s output should rise to 10 cartel decisions per year, reached within 24 to 30 months each, in spite of the fact that a substantial increase in resources is unlikely. The high number of leniency applications (20 per year since the establishment of the 2002 leniency programme) suggests a streamlining of procedures in the form of a “European plea bargaining” compatible with the existing legislative framework, and a “transfer of immunity” together with the transfer of cases within the ECN, so as to facilitate decentralization of cases which are retained today by the Commission due to lack of immunity at national level. The EC public enforcement system should be complemented by more effective private actions, as advocated by the Commission’s Green Paper on Private Enforcement. Criminal enforcement is excluded for the foreseeable future at EU level but it might progress in certain Member States, although it entails higher standards of proof, increased rights, and the need for efficient leniency programmes for individuals which interact with the leniency programme at EC level. Ian Forrester examines the investigative tools available to the Commission under Regulation 1/2003. His paper also examines the new Commission leniency programme, its adaptation in view of the discovery procedure in the US, and its limits with respect to cooperation in the framework of the ECN. Compared to the US amnesty programme, the paper finds the EC leniency programme less attractive because of a lack of confidence in the Commission’s fairness in setting fines (citing, inter alia, relatively high fines for smaller companies), the relative slowness of the procedures, their opacity and the lack of predictability of the steps following the confession. “Amnesty plus” and a plea bargaining procedure would be a useful addition. The paper reviews current Commission initiatives, such as: favouring the increase of private enforcement (which is considered to be a useful complement to public enforcement if it is limited to achieving justice and not punishment, and which could be facilitated by obliging cartel members to accept arbitration of claims made by victims); the fining of “facilitators”; and the use of trustees, as in the pending remedy procedure against Microsoft. Criminalization could be prescribed by the EC, but would require the Commission to change its investigation procedures, increasing the rights of the defence. An alternative to criminalization would be the disqualification of directors. The payment of bounties, following the Korean example, should be considered. Philip Collins’ paper concentrates on the situation in the UK. It is sceptical with respect to treating facilitating devices as being forbidden as such (as opposed to forbidding them only when ancillary to a cartel). The paper describes the civil and more onerous criminal investigatory powers available to the OFT. The OFT has reviewed its internal procedures so as to ensure that they conform with the higher standards of criminal procedures. For Collins, own-initiative investigations are useful but less necessary than at first sight, as leniency applications can be triggered by a series of factors other than fear


16/10/07

13:58

Page xxxii

xxxii Introduction of detection, such as corporate compliance and governance activities, due diligence in the case of mergers, management changes, etc. From the brief oral presentations of the papers, the following points are particularly worth noting. Convergence of the laws of different jurisdictions in substance and procedure is desirable so as to facilitate international (and intra-EC) cooperation. Companies might be more inclined to tell the truth when asked directly than is normally assumed, also in view of the changes in corporate governance that have occurred outside the antitrust area. Compliance programmes are important, in particular if they engage senior management. Direct settlements are useful in appropriate cases. Such settlements should provide for the compensation of victims, to be determined through mediation or arbitration. But direct settlements should not impede the deterrent effect of sanctions nor discourage leniency. Unfortunately, the recording of the discussion following the oral presentations of the papers failed. This volume therefore presents a regrettable gap (of 30 to 40 minutes) for the rest of the second session of the Workshop.

3. Session Three: Administrative (Civil) v. Criminal Sanctions a. Written contributions The broadest paper is undoubtedly the contribution by Wouter P. J. Wils. In view of its breadth and depth, it is impossible to summarize its contents in this Introduction. Wils himself delineates the scope of his paper by setting out five main questions, which he discusses in great detail, with respect to antitrust enforcement in the EU and its criminalization. First, what is meant by “criminalization”, as opposed to public enforcement of a civil or administrative nature? Second, is there a tendency in the EU Member States to criminalize antitrust enforcement, in comparison with US antitrust enforcement and with antitrust enforcement at the level of the EU? Third, is criminal antitrust enforcement, and more specifically imprisonment, desirable in general, i.e. independently of whether it takes place at the level of the Member States or of the EU, or whether it is harmonized at EU level? Fourth, is it problematic for antitrust enforcement to be criminalized at the level of individual Member States without parallel criminalization at the level of the EU institutions? Fifth, would it be legally possible to criminalize antitrust enforcement at the level of the EU, or to have EU harmonization of criminal antitrust enforcement in the Member States? The author gives positive answers to questions two to five. Particularly notable—in view of the ensuing discussion—is his affirmative response to the fourth question, i.e. that criminalization at the level of individual Member


16/10/07

13:58

Page xxxiii

Introduction

xxxiii

States would have negative consequences for the role of the Commission, which might well lose its central role in cartel enforcement in the EU—a high priority of overall EU antitrust policy. Stephen Calkins’ paper examines the issue of criminal sanctions essentially in light of the US experience. In the first part of his paper, he analyzes the reasons in favour of adding criminal to “civil” penalties, to punish individuals instead of (or in addition to) companies, and why incarceration has become the preferred sanction of the US Department of Justice (DOJ). The second part of the paper reviews the consequences of expanded antitrust remedies. Criminal enforcement has become dominant in the non-merger work of the DOJ. The concentration on criminal enforcement has had profound repercussions on the respective roles of the DOJ and the FTC, as well as on the development of competition law. Merger cases are still brought by both the DOJ and the FTC. But the exploration of novel issues has been taken over by the FTC or private parties, such as the conduct of dominant firms, the limits of intellectual property rights, vertical restraints and price discrimination. As sanctions are becoming more serious, courts have become more reluctant to find a violation. Finally, the author sketches some consequences for the EU. Because of the interactions between different elements of competition law, increasing penalties will lead to a demand for clearer legal standards. Individual responsibility has a greater deterrent effect than does holding a company responsible. Heavier penalties may come to dominate, both within the basket of available sanctions and in the relationship between agencies. In relation to national competition authorities (NCAs), the EU Commission could find itself in the role of the FTC vis-à-vis the DOJ. Nadia Calvino’s contribution concentrates on setting the right administrative fines. The paper examines first the question of whether there is an optimal fine, and concludes that fines should maximize deterrence while at the same time assuring proportionality from a double perspective: considering the profit derived by the individual offender, and also the harm caused to society as a whole. In discussing the usual two-step approach in setting the fine, the paper compares the traditional method, followed in the 1998 Commission Guidelines with the—preferred—more economic approach applied in the new (September 2006) Guidelines,6 which set the initial amount by using the volume of sales in the relevant market and the duration of the cartel. The paper discusses further the different aggravating and mitigating criteria which are used in setting the final amount of the fine, such as repeated violations, the presence of intention or negligence, the more active or passive role of the company concerned, its obstructive or cooperative attitude, the ne bis in idem principle, and the capacity to pay. The discussion 6 Guidelines on the method of setting fines imposed pursuant to Article 23 (2) (a) of Regulation No. 1/2003, OJ C 210 [2006]. The new Guidelines were published after the Workshop was held, and their contents were not yet known to most participants.


xxxiv

16/10/07

13:58

Page xxxiv

Introduction

is not a commentary on the new Guidelines, but a useful explanation of the reasoning which underlies them. Finally, the paper briefly compares the setting of fines in the EU with the situation in the US. Massimo Motta’s presentation set out the results of an empirical analysis (undertaken together with Gregor Langus) of the effect of EC antitrust fines and of the opening of EC investigations on the share prices of firms which have been caught infringing Article 81 and/or Article 82.7 The paper describes the research methodology (standard event study techniques), the events used (the dates of the dawn raid, the sending of the Statement of Objections, the Commission’s Decision and the first Court judgment), and the data considered (all decisions resulting in a fine from 1969 until 2005 concerning firms quoted in a stock exchange; in total, a number of 61 decisions involving 130 firms). The analysis shows significant effects on the share price (an average drop of 2.2%–2.4%) following dawn raids, but not following the other events. Even without taking into account the probability that the cartel will be uncovered, the study comes to the conclusion that the costs incurred if the investigation takes place (measured in the loss of capitalization of a magnitude of 2.2%–2.4%, of which the fine represents at most 70%) might not constitute a sufficient deterrent, in view of the estimated average duration of cartels and their average overcharge. The paper suggests as a starting point for the calculation of the fine the relevant market turnover and a much higher percentage than 10%—a suggestion reflected in the recently adopted Fining Guidelines.8 James Venit’s contribution pleads for greater convergence between the EU and the US antitrust enforcement systems. He discusses three practical examples of actual or potential difficulties arising from different rules. The first relates to the differences in the scope of legal privilege: the refusal of the EU to recognize legal privilege for outside lawyers not admitted to an EU Member State bar and for in-house counsel could lead to the loss of legal privilege in the US. The second example relates to leniency programmes. The EU programme not only lacks the useful instrument of a “marker” but also requires behaviour that might entail the loss of amnesty in the US. Finally, the existence of criminal sanctions for individuals leads to situations in which the interests of companies and those employed by them are no longer aligned. Individuals might need separate legal counsel and might be inclined to cooperate less in investigations if they expose themselves to personal risks. 7 Co-authored by Gregor Langus. Motta’s presentation at the Workshop was based on a preliminary version of the paper, where in particular neither the effect of the Commission’s Decision nor the effect of judgments of the Community Courts turned out to be statistically significant. The final version of the paper—published in this volume—is based on a larger dataset including only cartel cases. The final version reveals that the coefficient of the impact of the Decision and of annulments by the Courts are statistically significant. As a consequence, some of the comments made during the Workshop regarding the preliminary version would not apply to the final version. 8 Supra note no. 6, at para. 21.


16/10/07

13:58

Page xxxv

Introduction

xxxv

b. Discussion The participants agreed with the proposal that the evolution of relations between the DOJ and the FTC might well prefigure the future relations between those NCAs that have recourse to criminal sanctions and the EU Commission. The EU Commission might even be eclipsed in this respect by such NCAs. The participants also stressed the growing acceptance of individual and even criminal sanctions, including imprisonment, as proven by ongoing discussions in certain Member States. Although it was generally recognized that individual sanctions have “costs” for successful prosecution, such as higher standards of proof and even better rights of defence, it was also noted that the introduction of penalties for individuals might lead to more efficient instruments of detection. The lack of success of individual sanctions in certain situations, such as bid rigging, might be due to the lack of such instruments (e.g. leniency). The relationship between the successful prosecution of individuals and the applicable procedure is just one example of the interaction between different elements of the antitrust enforcement systems, highlighted in particular in Stephen Calkins’ paper. Advocates of more efficient private enforcement in the EU would of course regret a diminution in the rate of condemnations, as private follow-on actions would decrease correspondingly. The participants also noted the disequilibrium in the US between the growing importance of criminal sanctions imposed on individuals and the rarity with which such sanctions are imposed on companies. In addition, antitrust enforcement appears to be much stronger than the enforcement of violations of other rules of corporate governance. The participants showed a clear preference for individual sanctions, in particular director disqualification. It was noted that such a sanction can be formally provided for, but can also be the de facto consequence of a certain corporate culture (as in the US). In any case, the introduction and application of criminal sanctions should be carefully phased in and should be preceded by the effective application of administrative sanctions in order to create a positive climate for the acceptance of criminalization. With respect to administrative fines on companies, the participants noted the difficulty of striking the right balance between the need for deterrence and respect for the principles of proportionality and justice. Fines have risen substantially in the recent past. The move towards a more economic approach (considered to be preferable to the traditional approach of the 1998 Guidelines) entails the risk of losing sight of proportionality. In addition, if high fines effectively force a company out of business, this can produce undesirable economic consequences. The weakening and


xxxvi

16/10/07

13:58

Page xxxvi

Introduction

eventual failure of the firm might lead not only to undesirable social consequences but also to a higher concentration rate and higher market prices. On the other hand, the effect of dawn raids on share prices is astonishingly small when compared to the effect of share prices in the case of the announcement of mergers. Financial markets, which are usually well informed, seem to consider the future loss of the benefit of cartel-induced supra-competitive prices to be of relatively little importance. The economic effects of cartels might therefore be overestimated. Or are cartels which are discovered and which are the object of dawn raids in fact primarily “dying” cartels, a suspicion already voiced in the first session of the Workshop?9 In any case, more research into the economic effects of cartels would be useful.

4. Session Four: Private Enforcement of EC Competition Rules on Cartels a. Written contributions Donncadh Woods’ written contribution10 highlights a few points of the EU Commission’s Green Paper on damages actions.11 The contribution stresses the importance of addressing the problem of information asymmetry between the plaintiff and the defendant. This problem can easily be solved for followon actions. But it requires difficult procedural choices for stand-alone actions, either in terms of different modes of disclosure of evidence held by the defendant, or of shifting the burden of proof. The paper addresses further: the level of damages (single or double) and the attribution of costs in case of unsuccessful claims; the passing-on defence and the rights of indirect purchasers; collective actions of consumers; the need to protect the attractiveness of leniency programmes; the problem of fault; and issues of jurisdiction and applicable law. Denis Waelbroeck’s contribution12 comments on the Commission’s Green Paper on Damages, and benefits from Waelbroeck’s earlier experience in preparing the so-called “Ashurst study”.13 Like the Green Paper, the contribution is not limited to the strengthening of private enforcement of the 9

See supra point 1.b., penultimate paragraph. Co-authored by Eddy De Smijter. 11 Green Paper on damages actions for breach of the EC antitrust rules, COM (2005) 672. 12 Co-authored by Donald Slater. 13 Waelbroeck D., Slater D. and Even-Shoshan G. (2004): Study on the Conditions of Claims for Damages in Case of Infringement of EC Competition Rules, report prepared for DG COMP, European Commission, text available at http://ec.europa.eu/comm/competition/antitrust/others/ actions_for_damages/comparative_report_clean_en.pdf. 10


16/10/07

13:58

Page xxxvii

Introduction

xxxvii

prohibition of hard core cartels. The paper discusses a wide range of problems, starting with the objectives of strengthening private enforcement (compensation, deterrence, detection, bringing competition law closer to the citizen, cost recovery and competitiveness). The paper concludes that the main objective should remain compensatory; however, certain measures envisaged in the Green Paper are in clear conflict with the compensatory philosophy underlying the rules on damage awards in the Member States. In addition, bad decisions (“false positives”) resulting from the complexity of the law will have a chilling effect on the economy. The paper examines further the conditions of competition law-based damages claims (the impact of per se rules, the usefulness of specialized courts, the issue of fault, the passing-on defence, group actions, discovery, the effect of prior decisions of competition authorities, leniency and contingency fees). A recurrent concern is the justification for special rules for damages actions in competition cases. The author doubts that such actions are sufficiently different from similar actions in other areas of the law on damages to justify special treatment. Consequently, he warns against spill-over effects. Jon Lawrence’s contribution also stresses the need to find an equilibrium so that both claimants and defendants are assured of a fair and efficient system. However, his contribution appears to be less sceptical and reserved than the paper by Waelbroeck. According to Lawrence, the present situation is imbalanced because a claimant faces too many challenges. However, it is equally important to ensure that, in the future, the balance is not tilted too far in the other direction. A series of interests have to be balanced. The paper considers: (i) the interests of the claimants and of the defendants; (ii) the need to ensure compensation and deterrence (Lawrence is opposed to any kind of punitive damages); (iii) the optimal use of private and public enforcement (public enforcement is better suited for deterrence); (iv) fairness and efficiency (the Commission seems to privilege efficiency); (v) the interests of the EU and of the Member States (Lawrence doubts whether harmonization is achievable or even desirable; he favours encouragement of reforms and national “model laws” leading to “positive” forum shopping). A lex specialis for all antitrust cases would be wrong. One has to distinguish between: (i) follow-on and stand-alone actions; (ii) hard core cartels and less clear-cut areas of substantive law; (iii) actions brought by consumers and claims of indirect purchasers. All options should be considered together in order to evaluate their aggregate effect. The objective of compensation should be at the heart of private enforcement. Mario Siragusa’s contribution14 focuses on three issues discussed in the Green Paper. The paper considers that the fault requirement is not a significant impediment to private action, at least not with respect to actions for damages in hard core cartel cases. The introduction of a rebuttable presumption 14

Co-authored by Marco D’Ostuni.


16/10/07

13:58

Page xxxviii

xxxviii Introduction would be the best solution, but this would not represent very much of a change. The introduction of a harmonized discovery procedure inspired by procedures in Member States with a common law tradition would be useful, on condition that the requirements for fact-pleading and court intervention are carefully defined in order to minimize the risk of US-style “fishing expeditions”. In contrast to the foregoing proposals, Siragusa maintains that it is inappropriate to give binding effect to administrative decisions taken by NCAs. The paper concludes that the current framework for private antitrust enforcement in the Member States has not been tested fully, inter alia because of a lack of awareness about the rights conferred by competition law and the remedies available. National rules considered to be inadequate may prove to be suitable, once put to the test.

b. Discussion The participants stressed the importance of cultural factors. Sir Christopher Bellamy noted the existence of a sort of anti-litigation or “honour among thieves” culture. He advocated in favour of a dialogue between antitrust lawyers, special litigators, economists, judges and politicians/administrators. He stressed four areas of reform: first, the distribution of costs; second, rules on evidence (disclosure, backed up by heavy sanctions for the destruction of documents); third, the need for specialized judges; and fourth, some rules of thumb for the determination of damages. Sir Christopher cautioned against being over-ambitious. Special rules should be limited to hard core cartels and obvious abuses of dominant positions, so as to ensure that judges do not react defensively. The participants recognized the usefulness of a specialized private enforcement bar. However, it is unlikely that such a bar will emerge centrally (as it did in the course of the last decades as a result of dealings with DG Competition in Brussels), considering that private enforcement will have its centres of gravity in the Member States. With respect to the advisability of legislative action at EU level, participants were divided in their opinions. It is understandable that defenders of the Commission’s Green Paper stressed the logic of harmonized national rules concerning private remedies that correspond to the substantive rights and obligations that flow from Articles 81 and 82 EC. Others were rather sceptical—or even opposed—to the idea of obliging Member States to introduce special rules for the private enforcement of antitrust law which do not fit into the respective national systems. Similarly, there was disagreement as to whether violations of competition law were sufficiently different from violations of other forms of economic law to justify special rules limited to antitrust cases.


16/10/07

13:58

Page xxxix

Introduction

xxxix

The participants therefore gave strong support to the idea of individual national legislative initiatives which create a favourable environment for private enforcement and which attract private enforcement actions even from other Member States. Such initiatives could be a strong stimulus for the respective national legal systems, their domestic courts and their legal services industry. Such initiatives would create competition among national legal orders. They would allow the testing of different solutions, avoiding both the weaknesses of the present situation in the Member States and the danger of excessive solutions, as in the US. US lawyers would stand to gain much if the US became the privileged jurisdiction for the adjudication of private enforcement actions throughout the world. Fortunately, however, the decision of the US Supreme Court in the Empagran case15 has limited the risk of an excessive supranational reach of US law, thanks inter alia to the comity-conscious amicus position taken by the DOJ. One practical aspect of the efficient private enforcement of antitrust violations which was particularly stressed is the importance of consolidating claims of several victims located in different Member States in a single EU jurisdiction so as to avoid excessive costs. It was felt that this aspect was not sufficiently addressed in the Green Paper, and that it required further reflection and discussion. Reference was made to a recent German experience which shows that the consolidation of claims can be the result of private initiative, i.e., of a company which purchases the rights of victims in order to enforce them en bloc. Such an initiative might have been stimulated by German legislation which confers a binding effect on decisions taken not only by the Bundeskartellamt but also by other NCAs. Also relevant to the discussion was the Canadian experience, which shows that private enforcement can flourish in a system allowing class actions and contingency fees but not treble damages. The participants gave broad support to the idea of giving competition authorities the power to require compensation via mediation or arbitration, thus endorsing a suggestion which had been put forward already in the second session of the Workshop.16 This suggestion contrasts with the practice in the US and Canada, where agencies stay out of private enforcement. In any case, private enforcement influences the role of competition authorities. With active private enforcement, the authorities do not have a monopolistic gate-keeping function in deciding what will be pursued as an antitrust violation or how antitrust doctrine develops. Private enforcement helps competition authorities to pursue their own enforcement actions. The authorities have to monitor private enforcement and act as amicus curiae in order to influence the evolution of doctrine. 15 16

F. Hoffmann-La Roche Ltd. v. Empagran S.A., 542 U.S. 155, 164 (2004). See supra point 2, penultimate paragraph.


xl

16/10/07

13:58

Page xl

Introduction

5. Session Five: Leniency, Financial Incentives and Plea Bargaining a. Written contributions Donald C. Klawiter’s contribution examines leniency in the US. The paper sketches the origins and components of today’s leniency policy. In practice, the seven components of the US approach can be reduced to three principles: (i) first in and full cooperation; (ii) restitution to the injured party; and (iii) an appreciation of the fairness and propriety of granting leniency to a company. “Amnesty plus” and “penalty plus” are additional elements of the “carrotand-stick” approach. The paper describes how leniency has changed the investigative landscape; how to educate corporate decision-makers; how to react to the service of a subpoena or to a dawn raid; and how to proceed before the Antitrust Division of the DOJ. The author stresses the implications of the decision to apply for leniency: leniency will not work without the full cooperation of both the DOJ and the company in a true partnership. The paper warns that leniency does not come without costs and risks, in particular the quasi-certainty that the company will be sued for civil damages in the US. However, leniency applications provide advantages (including the de-trebling of damages and the lifting of joint responsibility for the first applicant), and the DOJ takes care to diminish the risks of discovery through the oral character of its proceedings. In addition, a leniency application in the US will have spill-over effects in other jurisdictions, although the DOJ does not share such information with other authorities. Klawiter regrets the absence of a “onestop shop” for leniency in Europe. His conclusion: the decision to apply for leniency is a decision of top management. Julian Joshua’s written contribution reviews the EU leniency programme in the light of the US rules and practice as well as the basic requirements of certainty, transparency and generosity. He criticizes the 2002 Leniency Notice for several reasons: the subjectivity of the criteria used in certain tests; the danger of using oral statements not in order to establish a road map but as proof, without any guarantees as to their truthfulness (the protection from US discovery is thus achieved at the expense of rights of defence and the integrity of the process); the lack of transparency; and the absence of a “marker” system and of an “amnesty plus” programme. For Joshua, while the incentives in the US are all geared toward getting in first, the EC system is less clear, and in particular the applicant must decide between the conflicting criteria of completeness and priority. In the EU, the situation is further complicated by the absence of a “one-stop shop” for leniency applications. Moreover, the success of the 2002 Leniency Notice is not matched by a


16/10/07

13:58

Page xli

Introduction

xli

notable increase in Commission decisions imposing fines. In the absence of a rapid sanction for discovered cartels, leniency applications are likely to decrease, as a successful leniency programme depends critically on, inter alia, the threat of severe sanctions for those who participate in a cartel and fail to self-report. Joshua recommends an integrated solution which: (i) reestablishes the Commission as the primary prosecutor of cartels with an EU dimension; (ii) drops the evidential threshold; (iii) introduces both a marker system and (iv) a “leniency plus” complement; (v) creates the possibility of negotiated settlements; and (vi) establishes predictable fining guidelines. Margaret Bloom’s paper shares many of Julian Joshua’s criticisms and suggestions. She too considers that certainty and transparency could be improved by clarifying certain elements which condition access to leniency or which determine the benefits, and that the issue of a one-stop shop for leniency applications in the EU has to be addressed. She suggests that references to a leniency application in the Statement of Objections should be minimized, so as to avoid the disadvantages that flow from such references in private actions, and that the position of individuals and that of companies should be as consistent as possible to ensure that there is no conflict of interests. The paper provides an overview of the number and effects of leniency applications since the Commission adopted the 1996 Leniency Notice, and discusses the differences of leniency programmes within the EU and the options for a “one-stop shop” system. The contribution examines in detail the interface between leniency and private enforcement (in particular the risk that increased exposure to private enforcement in the EU will diminish the incentive to apply for leniency), as well as the interface between leniency and criminal enforcement. Bloom’s analysis of the international implications of leniency programmes leads to a discussion of the effects of the Empagran17 and AMD/Intel decisions18 of the US Supreme Court. Her paper touches briefly on financial incentives (which are not likely to be introduced in the near future) and on plea bargaining, each of which should be investigated. Plea bargaining/direct settlement is the subject of Jochen Burrichter’s contribution.19 Starting with the German experience of a quasi-plea bargaining procedure before the Bundeskartellamt, the paper suggests that a direct settlement system before the Commission could consist of a short-form decision based on Article 7 of Regulation 1/2003, preceded by a simplified Statement of Objections. The direct settlement procedure could alleviate the Commission’s workload: the Commission would not have to fully investigate a case and to fully reason its fining decision; such a procedure would probably avoid subsequent Court actions. The procedure would also present advantages for the undertakings concerned: they would benefit from an expedited 17 18 19

See supra note 15. Intel Corp. v. Advanced Micro Devices, Inc., 542 U.S. 241 (2004). Co-authored by Daniel J. Zimmer.


16/10/07

13:58

Page xlii

xlii Introduction and predictable decision, from limited negative media publicity, reduced exposure to civil liability claims, and a reduction of the fine. Use of the procedure would be at the discretion of the Commission. It would have to respect the principles of legal certainty, in particular predictability and due process. With respect to leniency, direct settlements should be available to undertakings even if leniency is not at all available (either because the Commission is already in possession of ample evidence or because the applicant has coerced other undertakings to participate in the cartel) or if the applicant already enjoys partial leniency. Third parties should not be integrated into the settlement procedure, but they should be invited to comment in response to a notice to be published by the Commission before the settlement becomes final. The precise rules should be set out in a Commission Notice. John Ratliff’s contribution on plea bargaining/direct settlements compares the new instrument with the settlement procedure expressly provided for in Article 9 of Regulation 1/2003 (which cannot be used for cases involving a fine), and notes the differences between the two. The paper raises a host of substantive and procedural problems, such as: the fundamental question of whether the idea of plea bargaining is acceptable to the EU and its Member States; the minimum elements required in a simplified decision and the procedure leading to such a decision; what the concerned undertaking needs to know about the case which is made against it and about the fine which it may expect, in particular the discount which becomes available through plea bargaining; the repercussions of differences of position between several codefendants, some of whom intend to settle directly while others refuse to do so; and the rights of third parties, in particular complainants, compared to their rights in normal investigation procedures. Finally, the paper presents an overview of the essential elements of plea bargaining in the US, where plea bargaining is widely accepted but takes place in a fundamentally different legal and institutional context. Both the issues raised and the comparisons with the US system show the complexities of a new “hybrid” settlement procedure and explain why this useful complement to a more efficient cartel enforcement strategy is a latecomer in the tool kit of the Commission and the Member States. William Kovacic’s contribution examines bounties as inducements to identify cartels. A bounty system could provide a source of cartel instability even more acute and disruptive than existing leniency programmes. Such a system would raise the likelihood that government prosecutors would discover cartels; it would thus deter cartel formation and impede cartel implementation. The paper uses as an example and source of inspiration the US Civil False Claims Act (CFCA). The CFCA delegates the decision to prosecute wrongdoers to the DOJ and to private persons. Private informants are entitled to a share of the amounts recovered by the DOJ. Under certain conditions, the informant may also bring a private suit. Private monitoring is an antidote to concealment by cartel members; private prosecution is a remedy against lax or


16/10/07

13:58

Page xliii

Introduction

xliii

inefficient public enforcement. Protecting the informant only against economic or social retaliation and granting leniency may not be enough to motivate an individual to come forward: the bounty is a compensation for the risks and costs that the informant incurs. The group of possible beneficiaries and the modalities of the bounty system have to be carefully defined in order to avoid false positives and abuses. If this is not done by the legislator, the judges will find ways to restrict an over-inclusive system, as shown by their reactions to private antitrust actions in areas of the law less clear than hard core cartels. A bounty system should be a complement to leniency programmes. Such a system should be limited to hard core horizontal cartels and should be introduced as an experiment for a trial period of ten years.

b. Discussion As with the written contributions, the interventions in the discussion of the fifth session of the Workshop can be divided into three parts, i.e, those relating to leniency, bounties and direct settlements. The interventions relating to leniency complete the references to this topic which were made already during the first and second sessions of the Workshop. In addition to the points set out in the written contributions on leniency and highlighted in the corresponding short oral presentations, it is worth noting that participants stressed the overall importance of trust among all concerned, i.e., the applicant and its lawyers, the authority to whom the application is first made, other relevant authorities who have also received applications, and—in the US—the lawyers acting for private plaintiffs. Another point which was emphasized concerns the need for harmonization of leniency programmes, at least in the EU but hopefully even beyond. In addition, reference was (again) made to the problems of corporate governance: issues related to the question of whether or not to apply for leniency, and how to manage the consequences of such an application, raise extremely delicate questions for top management. In reply to the criticism that, compared to the US, the barrier for leniency applications in the EU is too high in terms of the type of evidence required, the observation was made that the quality of evidence counts, particularly when one considers—as participants did—that the efficiency of the leniency programme depends on the credibility of the system of sanctions, and thus on the number of successful prosecutions. The perspective of high evidential requirements for future leniency applications may have a useful by-product, as it provides an incentive for the preservation (instead of the destruction) of evidence. Another criticism concerned the lack of transparency with respect to the setting of the level of fines, which can give rise to the suspicion that an individual fine will first be increased so as to neutralize the later reduction incurred by an


16/10/07

13:58

Page xliv

xliv Introduction application of the leniency rule. In reply, it was noted that the introduction of a leniency system may well be the occasion for a general review of the existing system (and increase in the level) of fines,20 but that a manipulation of fines in individual cases in response to a leniency application would of course be illegal. From the point of view of The Workshop organizers and participants, it is rewarding to note that several of the points made during the Workshop are addressed in the Draft Commission Notice on Immunity from fines and reductions of fines in Cartel Cases, published by the Commission on 29 September 2006.21 We note in particular that the Notice defines more clearly the kind of evidence required, introduces flexibility with respect to the point in time at which applicants should terminate their participation in the alleged cartel activities, introduces a marker system and describes the procedure to protect corporate statements. In addition, on the same day, the Commission launched the ECN Model Leniency Programme,22 which aims at the introduction of a “one-stop leniency shop” in the EU. Compared with leniency, the discussion regarding bounties was relatively short. However, it is noteworthy that the introduction of a system of financial rewards was not rejected but considered to be a useful complement to the already-existing leniency programmes. Such a system contributes to destabilizing cartels because it gives an incentive to those informed persons who have no reason to apply for individual leniency, either due to the fact that they are not personally involved in the cartel and therefore cannot be liable for individual sanctions, or because they are not exposed to individual sanctions for the simple reason that such sanctions do not exist (as in the EU and the majority of EU Member States). On the other hand, it is obvious that the introduction of a system of financial rewards would add to the already-mentioned problems of corporate governance. With respect to direct settlements (a notion preferred to “plea bargaining”), participants agreed that Article 9 of Regulation 1/2003 could not be used as a legal base. But no-one expressed the view that direct settlements are forbidden under Regulation 1/2003. Views were divided as to whether the criteria for such settlements should be introduced by a Commission Notice or whether a formal Notice might be preceded by a period of trial and error in individual cases, so that the Commission Notice would only consolidate the results of practical experiences “on the ground”. It was recalled that the Commission’s Leniency Notice of 1996 already provided for a reduction of fines of 10% in case of non-contestation of facts, a

20 As explained earlier, the previous system of fines was in fact superseded after the Workshop was held. See supra note no. 6 and accompanying text. 21 OJ C 298 [2006]. 22 Available at: http://ec.europa.eu/comm/competition/ecn/model_leniency_en.pdf.


16/10/07

13:58

Page xlv

Introduction

xlv

provision which was not taken up in the Commission’s Leniency Notice of 2002. However, a direct settlement procedure would go beyond these (timid) beginnings, as it would require an admission of facts and the disposition to pay the “agreed” fine. In addition, a point already made in the second session was recalled: direct settlements should provide for redress of damages caused by the cartel to private parties. Participants agreed that, legally, the Commission would not be entitled to request a definitive waiver of the right to appeal a fining decision. However, it was noted that, in case of an admission of the facts, an appeal is not likely (or not likely to succeed), unless arguments are made which relate to the level of the fine, such as discriminatory treatment. It was also noted that the institutional system of the EU makes negotiations about the level of the fines difficult. Fines are decided formally by the Commission as a collegiate body, and not by the Member of the Commission responsible for competition, and even less so by officials of DG Competition. In addition, this process is traditionally non-transparent; decisions on the level of fines are taken at the last minute. The situation in Brussels is therefore totally different from the situation in the US and the Member States. Under these circumstances, will negotiations with officials of DG Competition be sufficient? Finally, it was recognized that not all cases will lend themselves to direct settlements. There will remain cases which are very controversial and which will need the respect of the full administrative procedure and all its safeguards in view of the likelihood of proceedings before the Community Courts.

6. Session Six: Effectiveness of Enforcement and Legal Protection v. Rights of the Defence a. Written contributions Nicholas Forwood’s contribution starts with the proposition that the primary measure of effectiveness of the cartel prohibition and its enforcement mechanism is its degree of deterrence with respect to entering into a cartel, as (ex post) punishment and compensation of victims will usually remain suboptimal. Cartel members act with a high degree of rationality. Their inclination to “cheat” through whistle-blowing will increase with the importance and the certitude of the benefits they can expect in doing so. Judge Forwood is in favour of a “hard law” approach to the application of a leniency programme, i.e., a close control by the Community Courts as to whether the


16/10/07

13:58

Page xlvi

xlvi Introduction programme has been correctly applied. Conversely, with respect to fines, he advocates guidelines which leave a certain degree of flexibility; an over-rigid or formalistic application of fining guidelines would be counterproductive. However, the guidelines might themselves be successfully challenged if in breach of Community law principles, such as proportionality or nondiscrimination. In addition, Judge Forwood advocates the maintenance of the existing administrative system, in which decisions on fines are taken by the Commission under the full control of the Community Courts. The Courts would neither be well positioned—nor perhaps willing—to adopt guidelines for the future fining policy or to formulate and apply a leniency programme. Both are matters for the Commission, which is entrusted by the EC Treaty to formulate and apply EU competition policy. Christopher Harding’s contribution examines issues of legal protection in the emerging quasi-criminal EC law, in particular in relation to the application of fines, the temporal dimension of cartel proceedings, and the interfaces resulting from the co-existence of enforcement procedures in third countries, the EU and its Member States. Legal disputes arising out of cartel enforcement have increasingly become a matter for criminal and human rights lawyers, as the emphasis has shifted from aspects of investigation to issues of penalty. The call for greater transparency and predictability of fines is understandable, but it is also contrary to the desired deterrent effect, insofar as predictable fines may simply become part of the estimation of business risk when evaluating whether to participate in a cartel. It is difficult to deal with the argument that an otherwise justified large fine threatens the financial viability of a company; the practice of the Commission is not consistent. Fining guidelines limit the discretion of the Commission, but such guidelines may of course be changed. A change which is reasonably foreseeable does not raise problems of (forbidden) retroactivity. Changes of corporate identity which have occurred since the prohibited cartel activity took place have to be evaluated. The principle of ne bis in idem does not apply where sanctions have already been imposed by foreign jurisdictions, but such sanctions must nevertheless be taken into account in assessing the proportionality of a fine. In conclusion, Harding warns that the Commission’s concern for efficiency should not affect the objectives of deterrence and retributive justice. Peter Roth’s paper concentrates on cartel enforcement and the protection of fundamental rights as defined in the European Convention of Human Rights (ECHR) and the EU Charter of Fundamental Rights, the latter of which is expressly referred to in the preamble of Regulation 1/2003. The paper notes the particularities of competition law, which is addressed to undertakings, i.e., mostly legal persons, concerned with commercial confidentiality, and applied by administrative authorities, although the fines which are imposed by them constitute “criminal” penalties. The conditions for dawn raids distinguish between business premises and private homes; with respect


16/10/07

13:58

Page xlvii

Introduction

xlvii

to the former, they should even distinguish between premises of persons legal and those of natural persons. Proportionality is a fundamental principle, but can hardly be evaluated by a national court, which does not have the right to see any of the underlying evidence before authorizing the dawn raid. The ECJ jurisprudence concerning the privilege against self-incrimination may be adequate for documentary requests, but is unsatisfactory with respect to questions addressed to individuals—in particular when individuals are exposed to individual sanctions, be they custodial or financial. The problems are enhanced by the increasing cooperation among competition authorities, which apply different procedural and substantive legal rules. The ECJ jurisprudence on legal privilege should be reviewed: it should recognize the public interest in ensuring that corporations and individuals can obtain effective legal advice in an increasingly globalizing and complex world.

b. Discussion The following discussion showed unanimous support for Roth’s plea regarding legal privilege. It was pointed out, in this regard, that a recent Spanish case is in line with the trend for a more generous recognition of privilege. Participants noted the value of an in-house counsel who has the full confidence of his colleagues, and the need for lawyers from different jurisdictions to communicate freely in cases of international cartels. However, there remained differences of opinion with respect to the conditions which limit the scope of legal privilege, as regards both the persons and the documents that should benefit from it. In information exchanges among competition authorities, the higher standard should prevail, as recommended by the OECD. The criticism of the ECJ jurisprudence in relation to the privilege against self-incrimination triggered a reflection about corporations: as the corporate existence is a privilege granted by law, could this privilege not be conditioned by limitations attached to the recognition of certain fundamental rights which are fully justified for natural persons, but not necessarily for legal persons? Could the same not be true for persons employed by the corporation, if they are not exposed to the risk of personal sanctions? Finally, the discussion returned once more to leniency: should the first applicant be entitled to full amnesty even if the applicant was the founding member (or one of the founding members) of the cartel? While deterrence might plead against such a solution, the interest in detecting cartels pleads in favour of generosity: the incentive for any member of the cartel to ask for leniency is a destabilizing factor. While transparency and predictability of the level of fines might not be good for deterrence, they are advisable for an efficient leniency programme: also in Europe, administrative discretion has progressively been restricted in the light of experience, particularly in the US.


16/10/07

13:58

Page xlviii

xlviii Introduction It is remarkable that, at no point in the discussions was the current institutional regime for cartel enforcement in the EU questioned. The difference between the US court-based system and the European administrative agency model was apparently taken for granted. On the contrary, the only explicit evaluation of the merits of the two systems (in Nicholas Forwood’s written contribution, which expresses a clear preference for the European model) remained untouched. The situation might have been different if the new fining Guidelines of the Commission23 had already been known to the participants. The new Guidelines provide for a dramatic increase in the level of fines, in particular for long-lasting cartels. Fines can easily reach the legislative limit of 10% of total turnover. They can also easily reach or exceed 1 billion Euros, if the company is sufficiently big and successful. While the Workshop had in fact demonstrated that the existing EU sanctions against hard core cartels did not have a sufficiently deterrent effect, the papers and proceedings also show that fines on undertakings are not the most efficient instrument by which to achieve such an effect. Sanctions against individuals might be a useful complement. However, the Commission does not dispose of any such sanctions; it has only one option, i.e., to increase the level of fines on companies. The fines resulting from the new Guidelines may not be proportionate in a vertical sense, i.e. with respect to the wrong they are supposed to address from the narrow point of view of cartel (or antitrust) law enforcement. Furthermore, the new levels of fines are probably not proportionate when compared with sanctions imposed to punish wrongdoing in other areas of economic law which are comparable to violations of the prohibition of cartels. We do not know of any other area of economic law where similarly high fines are imposed on companies. Whatever one may think about proportionality, it is highly problematic that fines of the size which undertakings will face from now on have been decided by the EU’s executive, instead of being decided by the judiciary. Such fines were certainly not envisaged when the European model of administrative competition agencies was conceived and enacted. Instead, it would be salutary for the Commission to recognize that the only type of sanctions at the disposal of the EU, i.e., fines for companies, is no longer by itself sufficient. The Commission should propose to broaden the spectrum of choice by opening up the possibility of imposing sanctions against individuals. Such a proposal would of course logically entail the need to transfer the power to impose such sanctions to the CFI. The Commission would therefore have to limit its role to that of a public prosecutor, like the competition agencies on the other side of the Atlantic.

23

See supra note no. 6.


16/10/07

13:58

Page xlix

Introduction

xlix

As unrealistic as this proposal may seem today, it is nothing but the logical result of the West-East drift in modern antitrust law, which we have all experienced so forcefully during the last years. Who would have thought in June 2002, when the discussions regarding criminal sanctions against individuals emerged at a similar Workshop at the EUI, that they would so quickly become a reality in several Member States? Claus-Dieter Ehlermann December 2006, Brussels


16/10/07

13:58

Page l

(C) Ehlermann 07 Session A Pt 1

12/10/07

12:16

Page 1

SESSION A

UNEARTHING CARTELS—THE EVIDENCE

1 PANEL DISCUSSION

Participants: Allendesalazar, Rafael Barnett, Thomas Bloom, Margaret Calkins, Stephen Collins, Philip Ehlermann, Claus-Dieter Fingleton, John Forrester, Ian Goldman, Calvin Grout, Paul Guersent, Olivier Harrington, Joe

Joshua, Julian Kovacic, William Motta, Massimo Ratliff, John Rey, Patrick Schinkel, Mark Peter Siragusa, Mario Ullrich, Hanns Venit, James Waelbroeck, Denis Whish, Richard Wils, Wouter


12/10/07

12:16

Page 2


12/10/07

12:16

Page 3

Panel Discussion

3

Session A: Unearthing Cartels—The Evidence Panel I: Oligopoly Theory and Economic Evidence CLAUS-DIETER EHLERMANN—Welcome to the eleventh edition of the EU Competition Law and Policy Workshop held at the EUI in Florence. I must start by announcing that, unfortunately, Commissioner Neelie Kroes cannot join us today, due to unforeseen professional obligations. Her opening speech will be presented by Mr. Olivier Guersent, representing the Commissioner’s cabinet. The first session will be chaired by Prof. Massimo Motta.

䉴

Ladies and gentlemen,1 I am delighted to be here in Florence for the Eleventh Annual EU Competition Law and Policy Workshop. I would like to thank Claus-Dieter Ehlermann in particular for his kind invitation to open this event. The purpose of this workshop is to allow key policy makers, judges, academics and practicing lawyers to discuss critical issues in the field of EC competition law. The subject of our reflections over the next two days, the “Enforcement of Prohibition of Cartels”, is at the very top of the priorities of the Commission and of the European Competition Network. Cartel conduct strikes at the very heart of the principal virtue of economic activity—competition between companies for markets and for customers. This competition, this rivalry, stimulates innovation and increases efficiency and consumer welfare. Of course, Adam Smith famously remarked on the strong propensity of companies to avoid this rivalry, and instead to cartelize and maximize industry profits. The economic benefits to companies are clear. Yet collusion between independent economic entities, whether successful or not, serves no purpose for society as a whole. So we must have a framework that imposes penalties heavy enough to outweigh the benefits that companies expect to receive from cartelization. We must both punish the cartels that we find, and deter the rest. For the framework to be effective, we need four main elements. First, fines that are sufficiently high to act as a genuine punishment and deterrent. Second, a working leniency regime to provide appropriate incentives to cartel participants to defect and report cartels. Third, adequate resources. Fourth, co-operation between enforcement agencies. In what follows, I shall outline my views on each of these elements. I shall conclude with some considerations of future policy developments. Let me begin with where we stand today.

1 The text in italics reproduces the speech prepared by Neelie Kroes, Member of the European Commission responsible for competition, for the event.


4

12/10/07

12:16

Page 4

Session A – Unearthing Cartels—The Evidence

A. State of Play

Commission action against hard core cartels increased significantly over the period 2000–2005. The Commission adopted thirty-eight decisions over the period, representing an average of six decisions per year. This is already an improvement compared to the record for the thirty years from 1970 to 2000 of only three decisions every two years, but I want to see that number increase yet further. These thirty-eight decisions were addressed to about three hundred companies, who were fined a total of nearly four and a half billion euros. During this recent period, the Commission focused its actions on significant hard core cartels of worldwide or European scope, and each case involved a number of economic entities. The average fine per addressee also increased significantly, in line with the Commission’s announced intention to impose fines that have the necessary deterrent effect. Most of the cartel decisions are systematically appealed, which thus provides—something absent in some other jurisdictions—a rich and illuminating body of case law. At the same time, dealing with the court appeals requires significant resources. The Commission’s record in front of the European Courts is excellent. During this period (2000–2005), the Court has never overruled us on the substance of a case. As for the fines, companies fined in cartel cases since the 1998 Fining Guidelines were introduced have on average ultimately had to pay nearly 90% of the fines imposed by the Commission. (This is for cartel cases where either there was no appeal or where the Court of First Instance has ruled.) Of course, as companies don’t go to court if they think the fine is too low, a 100% record is effectively impossible. In parallel with the Commission’s anti-cartel activities there have been an important number of cartel decisions adopted by national competition authorities. The entry into force of Council Regulation No. 1/2003, and the impetus given by the European Competition Network’s mechanisms of information and coordination, have greatly contributed to the fight against cartels. If the recent experience of anti-cartel enforcement is anything to go by, then I think there are three conclusions to draw: first, the cartel culture in the European economy extends far beyond traditional manufacturing sectors to all sectors, including agriculture and a number of service industries. Second, the cartel detection rate is improving. Finally, the deterrent effect of fines merits a re-examination.


12/10/07

12:16

Page 5

Panel Discussion

5

B. Fining Policy The increased level of fines levied under the Fining Guidelines adopted in 1998 2 has certainly drawn many comments, not least from those who have had to pay them. No one denies that it has been an important element in combating cartels— not only because of the increased deterrent effect but also because the increase in fines provides greater incentives to race to file for leniency. The incentives are particularly high for addressees of past infringement decisions—but we still continue to see repeated infringements. Just a few weeks ago, we increased two fines by 50% in the hydrogen peroxide cartel because of such recidivism. This week we had to do it again in the Metachrylate cartel. While I welcome the trend in the adoption of compliance programmes by the corporate sector, it’s clear that in many companies, compliance has not gone far enough. Companies and competition authorities must both step up their efforts to eradicate this cartel culture. Having acquired more than eight years of experience under the current Guidelines, I intend to propose new Guidelines for adoption by the Commission. As the European Court of Justice has just made clear in the Archer Daniels Midlands ruling, the Commission has discretion as to its fining policy. I intend to exercise that discretion to increase the deterrent effect of our fines, to set a better balance of fines between large companies and SMEs in the same a cartel, and to give companies greater predictability as to the level of fines. Another way I intend to attack the cartel culture is to strengthen the rights of companies and individuals who are the victims of illegal anticompetitive conduct. It is absurd that, nearly fifty years after the EC Treaty was signed, rights under the Treaty are still not effectively protected. Enforcement by competition authorities must be complemented by private actions for damages by victims of illegal anticompetitive conduct. As well as being about justice, significant private enforcement would provide a further cost and therefore a disincentive to cartels. The Green Paper issued by the Commission at the end of 2005 identifies obstacles to an effective system of damage actions for breach of the EC antitrust rules. Various options designed to address each obstacle were presented for discussion, and I’m extremely happy at the response so far. We are currently analyzing all the contributions received in the public consultations. These and the ongoing discussions within the ECN will be carefully taken into account before deciding on further actions.

2

OJ C 9 [1998].


6

12/10/07

12:16

Page 6


C. Leniency Policy The 2002 Leniency Notice has proved to be a major improvement over the 1996 Notice upon which it was based. The increased transparency and predictability it gives applicants is evident in the statistics on applications, although, given their sensitive nature, I shall limit myself to overall totals. Under the 2002 Notice, the Commission had received eighty-seven applications for immunity by the end of 2005. Over the same period, the Commission granted fifty-one decisions for conditional immunity. It either rejected or decided not to deal any further with twenty-three applications, and it was scrutinizing thirteen more recent applications. In one case, the Commission withdrew conditional immunity. In the same period, seventeen applications for reductions in fines were provisionally granted (of which eight have been confirmed), and thirty-four applications have so far been rejected. The reason that fewer applications for reductions are granted compared to applications for immunity is because they are decided upon much later in the procedure—at the latest before the Statement of Objections—compared to immunity applications, which are decided at the outset. The Commission’s case against the cartel in question is therefore better developed, and the ability of a third party to add value is reduced. This total of one hundred and thirty-four applications up to end of 2005 under the 2002 Notice was made by companies headquartered in twenty different countries, primarily in EU-15 Member States. These figures suggest that the leniency policy is fulfilling its objectives of: —destabilizing cartels; —encouraging their members not only to defect but also to report the cartels to the competition authorities; and —creating a race amongst participants to report. By the end of 2005, twice as many applications had been submitted under the new Notice as compared with its predecessor. As a result of the 1996 Notice, the Commission has adopted 36 cartel decisions so far. I have little doubt that this ratio will be maintained, if not improved upon. What is also important and not often mentioned is that, since 2002, three cartel decisions were adopted resulting from ex officio investigations based on complaints and market monitoring. The Commission’s ex officio investigations almost always trigger immunity and leniency applications in related geographic or product markets. In applying the Leniency Notice, the Commission takes great care to protect corporate statements of the applicants in order to guarantee that companies that co-operate with the Commission are protected from discovery requests from jurisdictions not applying EC competition rules. When necessary, the Commission has intervened as amicus curiae in other jurisdictions to make this point clear.


12/10/07

12:16

Page 7

Panel Discussion

7

Some commentators say that the need for multiple filings of leniency applications in the ECN deters reporting of cartel infringements. The extent of this problem is hard to judge: third parties will always ask for a system with more guarantees and greater benefits for leniency applicants. However, the problem may be limited in practice due to the case allocation and coordination mechanisms of the ECN. At the same time, over the last few months, intensive discussions have been underway in the ECN in order to reduce, to the extent possible, the discrepancies between the different leniency programmes in force or under consideration. It should be possible to clarify the essential features of the treatment which an applicant can expect to be given in any ECN jurisdiction. I am glad to say that much progress has been achieved through this cooperation between the national competition authorities. The result, I think, will be generally regarded as leading to a de facto one-stop shop for leniency within the ECN. We therefore have considerable experience with the application of the 2002 Notice. The general conclusion is that it works well, but we are continually looking for ways to improve the system. The various suggestions for changes made by different commentators are being looked at and carefully evaluated by DG Competition. The main conclusion that I draw from the ongoing internal review is that the processing of applications could be improved and made quicker. We need to provide guidance to companies and their lawyers so that the quality of leniency applications improves to match their quantity.

D. International Cooperation I have already touched upon the very close cooperation within the ECN, which enables it to operate as an effective enforcement network. The prosecution of international cartels has also benefited considerably from the intensification of the links developed by the Commission over the years with a number of enforcement agencies. I’m thinking here in particular not only of the US Department of Justice and the FTC, the Canadian Competition Bureau, and the JFTC, but also of colleagues in Australia, Korea and other jurisdictions. Notwithstanding the strict limitation on what information can be exchanged due to confidentiality obligations, international cooperation nevertheless can have a considerable impact on cartel prosecution. Where a company has filed for immunity in several jurisdictions, the applicant usually provides waivers to enable authorities to freely discuss and share the evidence that the applicant has provided. This allows agencies to discuss and coordinate their planned enforcement actions. (as in the case of Aluminium Fluoride, for example). I’m sure that you are all aware that we have been able to coordinate our dawn raids with other jurisdictions—if not through the press, then through, shall we say, more direct experience.


8

12/10/07

12:16

Page 8


Similarly, where initial information emanates from market monitoring and not from leniency applications, the cooperating agencies have been able to provide one another with useful leads at the early stages of investigations. Information from complainants can also, with their agreement, be shared among agencies. Beyond case cooperation, various fora such as the OECD and the ICN provide frameworks for policy discussions and exchange of best practices in relation to, for instance, leniency rules, issues related to discovery, effective inspection techniques such as forensic IT and on investigative techniques in general. Such cooperation, I believe, is essential in the fight against international cartels, if only to counter the rather easy cooperation that would appear to prevail amongst the international cartel members themselves.

E. Enforcement Resources This leads naturally to a consideration of the importance of resources in the fight against cartels. The substantial and sustained retreat of the prevalent cartel culture needs serious allocation of resources. This is why I directed that DG Competition create a specialist cartels directorate specifically charged to deal with all hard core cartels uncovered by leniency applications. Deterrence gains in effectiveness if infringements are rigorously and speedily pursued with decisions that stand up to judicial review. This in turn gives a necessary signal to the corporate sector that the detection rate is ever increasing and that important efforts need to be deployed to make irreversible changes in the way they conduct business. I am pleased to note that this approach is also embraced by national competition authorities in the ECN.

F. Conclusion I would like to conclude with some reflections regarding further policy developments that I intend to implement in respect of the Commission’s cartel enforcement. I have already indicated that the threshold for immunity in our Leniency Notice needs to be clarified to guide more precise and substantiated leniency applications, which can be dealt with more quickly. Similarly, both the efficiency of the Leniency Notice and legal certainty for applicants would improve if the content of the duty of cooperation were further clarified and extended to all leniency applications. A number of other changes need to be made, such as the insertion of rules on oral procedure for corporate statements. The main elements of the Notice will not materially change.


12/10/07

12:16

Page 9

Panel Discussion

9

At my request, DG Competition has also been examining the introduction of direct settlements. Such settlements would be comparable to some extent to the US system of plea bargaining, but made appropriate to the decisional process of the Commission. Experience shows that companies are eager to have incentives for acknowledging their involvement in a cartel and accepting liability for it. At the same time, the current leniency programme does not reduce the procedural steps of our administrative procedure, or the time taken to complete them. Interestingly, though, while our procedures may be considered heavy and long, they are in general more rapid than the plea-bargaining process in the US. Before I will consider direct settlements, certain preconditions must be fulfilled. Any system must reduce the duration of the procedure and release enforcement resources for other cases. Any system must also maintain the deterrent effect of enforcement action. In view of these constraints, it would seem that, under the existing regulatory framework of Council Regulation No. 1/2003, any savings in procedures must take place either before or upon the issuing of a Statement of Objections. If we are to maintain deterrence, companies must not view direct settlements as a speedy and cost-effective way to end one cartel episode only to pursue another. Hence, reductions of fines for leniency applicants in the existing system are limited. At this stage I would limit myself to outlining these few elements of our internal reflections in order to provide some food for the debate that we are to have in this workshop. I look forward with interest to your discussions. 䉴 MASSIMO MOTTA—The first session deals with oligopoly theory and economic evidence. I propose to change slightly the order of the presentations so as to take into account their subject matter. Paul Grout will talk first, about the main characteristics of cartels. He will be followed by Patrick Rey, who will speak about facilitating factors and how to detect collusion. Next, Joe Harrington will talk about how to screen for cartels. He will be followed by Rafael Allendesalazar, talking about standards of proof in the area of concerted practices. We will conclude the round of presentations with Richard Whish, who will address the question of what happens when we do not have evidence of collusion: are there other instruments available?

䉴 PAUL GROUT—I usually tend not to regret things in my life, but in recent years there are at least two things that I do regret: one of them is wasting two hours to see “The Da Vinci Code”, and the other is the title we have given to the OFT Report on which I shall base my presentation3—and which was 3 Grout P. and Sonderreger S. (2005): Predicting Cartels, OFT Economic Discussion Paper No. 773/March 2005, text available at http://www.oft.gov.uk/NR/rdonlyres/4E9FFDC4-31F44F7D-896E-FFE635A136F3/0/OFT773.pdf.


10

12/10/07

12:16

Page 10


chosen essentially out of laziness. . . . The original title of the project was something like “The Economics of Cartels”, and it eventually became “Predicting Cartels” because we realized that the scope of what we were doing was much more narrow. Yet what we actually do in this report is to look at case studies and industry data in order to see if that may help us in any way to see where cartels might be. That is a long way short of predicting where cartels are; we just offer some piece of guiding information that is useful. Usually, when I speak before an audience of non-economists, I get the impression that the audience may have two takes on it: one is that “all this technical stuff ” allows us to predict very accurately where cartels are—and then I spend a lot of time trying to persuade people that this is not the case; and the other is that, so little is explained that there is little or no use at all in undertaking this sort of exercise. The truth is somewhere in between. In the next few minutes I would like to talk about some of the constraints in using this analytical approach, and some of the issues that come out of it that are worthwhile to remember. I shall not go through the theoretical literature, but just summarize the core messages. The theory tends to be a bit uncertain about various factors, but at the end of the day, what it says is that, if there is a cartel, the price is abnormally high and anyone in the cartel group that decides to deviate would have short-term gains but would lose in the future. All of the theory is, one way or another, about looking at the factors that make deviation more or less likely. So it is about issues like the reading of signals. This is actually one of the critical themes where theory and data line up: the easier it is to read the signals on whether people [are likely to] deviate from what they have agreed, the more likely you are going to be able to run a successful cartel. The more differences there are within the firms in an industry, the harder it is to police an eventual cartel. If one looks at case studies and industry evidence, what are the kinds of problems and issues that are relevant? The first is the so-called “postselection” issue: you are not actually looking at a set of cartels in the world, but only at a set of cartels that have been found and proven. So, when we look at the case studies and industry evidence, are we looking at the factors that led to cartelization, or are we actually looking at the factors that made cartels fail? There is also a particular issue with entry barriers, because the reason why we found that some of these cartels existed is not that there are entry barriers in the industry, but that the cartels have actually tried to erect the barriers to entry. So there is a sort of mismatch between the theory and the empirical evidence. Second, and related to that, are we looking at cartels, or are we looking at the cartels that we can actually convict? These two categories are not necessarily the same. Looking at past histories is a much better predictor of cartels that we are able to convict, and we may not be able to uncover all the cartels. There is an interesting feature coming out of the data, namely the similarity between jurisdictions.


12/10/07

12:16

Page 11

Panel Discussion

11

Third, are we looking at industry data or cartels in current markets? Those two things need not be the same, and this is an important point that I mention just in passing. I will not talk now about the various factors that jump out of cartel case studies, which are covered in our written contribution to this volume. One of the conclusions that come out of the analysis of the data is that industries with larger firms are more likely to have cartels. But perhaps the most important single conclusion coming out of the data analysis is that most of the industries where cartels have been discovered in the EU actually have very homogeneous outputs. One of the tests for the extent to which the industry has a homogeneous output is how much R&D is going into the industry. If there is lots of R&D going in, firms are discovering things, and products are differing over time. If you look at the amount of R&D going on in all the sectors where cartels were found in the EU, you can see how small these numbers are: they are almost all industries where there is literally no R&D. So, one of our observations is that these industries have relatively homogeneous products. This coincides with economic theory, because in the case of homogeneous products we are much more likely to read the signals of what each firm is doing—they are all going to be hit by the same shocks, and then you can pick up individual responses. Next, we took all horizontal price-fixing cases from the US over a 15-year period, and compared them with all the cartel data in the EU over a roughly similar period. We then allocated them to their 3-digit industry classification, and what jumps out of their analysis is that there is a huge similarity in how cartels occur. The group correlation between the EU and US is 0.68, which is huge, and if we take shipping out, it becomes almost 0.8. Furthermore, that correlation totally underestimates the similarity, because they seem to be correlated if they appear in exactly the same 3-digit classification, and in lots of instances these cartels appear in the very next classification in the US to EU—which makes it look as if there is almost no correlation, whereas in fact they are very correlated, because these 3-digit industries are not randomly allocated. The question is: why is this happening? Only about 20% of these cases are common. One of the possibilities is that there are factors in these industries that really drag firms into cartels. The completely opposite explanation is that people in one jurisdiction look at what has been successfully done in another jurisdiction—and we know lots of that goes on. Therefore, there is a kind of hysteresis in the process: once you hit upon firms in one jurisdiction, everybody in the other jurisdictions dives in. So, there is the issue of whether the received truth is really the truth on where the cartels are, or whether we are looking at a sort of self-reinforcing process—and I suspect there is quite a bit of the latter going on. But our study also suggests that there are some industry-specific factors to be taken into account as well. To sum up the messages that come out of our exercise: first, this is only an input—it is not a mechanism for predicting cartels, but just a way of looking into things that helps. Second, variability in growth is an issue—if there is lots


12

12/10/07

12:16

Page 12


and lots of variability in the market, you are less likely to see cartels. We are likely to discover a cartel when there is a long period of slow decline in the market, and then a sudden hit that seems to lead people to get together. Traditional entry barriers do not seem to have a lot of effect in this area. Scale matters; regressions typically explain somewhere between 10 and 22%. So here is a set of instances where pure econometrics may predict where there may be cartels though they have not been found. One of the interesting issues here is private secondary education. This came from a data set where almost all of the information came out of the industries, and it predicts that secondary education will have a very high level of probability in the top 30 of these industries. Simultaneously—though without knowing it at the time— secondary education was subject to investigation, and it is very interesting that this turned out a few months after our report was published. I did not mention profitability, but it is a big issue. Profitability at the industry level does not work, you have to do it very closely—and actually in a recent paper I have discussed this: you can use profitability in individual cases, but not in this kind of more general approach. 䉴 PATRICK REY—My presentation is about the two main ways to identify cartels. The first one is the so-called “police patrol” approach, which consists of sending the law enforcers out to look for trouble. This is difficult to do. In the context of antitrust enforcement against cartels, this means looking for signs that something fishy is going on in a market. This may be useful for what I would call “pre-detection”: typically it would be difficult to be sure that a cartel is in fact operating; what you usually get are indications that there may be a cartel. The second is what I would call the “fire alarm” approach: you install fire alarms in houses, so that when a house burns you get the signal. In our context, this means that you want to encourage those who have the information to come forward and hand it over to the antitrust agencies. I would argue that this way may be less costly and easier to pursue from the standpoint of prosecution, and that is why indirectly it may also be useful for deterrence purposes. Having said this, I think these two ways are complements that work in the “stick-and-carrot” manner. In particular, the fire alarm, or carrot, will work better if you have an appropriate stick, and it may be useful to rely on the other approach to show that antitrust agencies look closely at industries where something fishy may be going on. Economic analysis is useful to both approaches, and I would like to go into a bit more detail on this. First, regarding the “police patrol” approach, there are several ways to look for evidence of trouble. You can look for structural factors—and this I think is what Paul Grout just covered, so I will be very brief on this—you can look for performance, like supra-competitive profitability of prices, or for behavioural patterns—and that will be covered in more detail by Joe Harrington.


12/10/07

12:16

Page 13

Panel Discussion

13

On structural factors: there is some consensus on the variables that we should look for, and I would agree with the variables that Paul Grout has just discussed. In the report for DG COMP on the economics of tacit collusion4 that I co-authored with some colleagues at IDEI, we have the same kind of list of “usual suspects”. Those factors can be useful to determine in what industries, markets and so on, there could be collusion, but they cannot be used to determine whether there is indeed collusion. So, they are useful only at the prediction stage—to identify the candidates for collusion. In his written contribution for this Workshop, Joe Harrington made the point that those factors do not distinguish between explicit and tacit collusion, which may also be an issue for our purposes. Here I would like to argue that this may be useful per se as an input, or as a piece in a more global framework. But it may also be useful for deterrence purposes to know that the agencies are closely watching those industries. I have in mind a paper by Sylvestre Frézal5 stressing that, if you have the choice between randomly selecting the industries to look at, or doing this in an pre-announced sequence, the second approach is much more effective. Of course, the above-mentioned factors could be used to modulate the intensity of scrutiny across various industries; this would make the policy more effective, by looking more frequently at those industries where collusion is more likely. I will now refer briefly to performance indicators, where the idea would be to look at prices or profitability expected from normal competition. The issue here is to assess what normal competition is; this is a challenging task, requiring a lot of information on costs, demand, etc. On the profitability dimension there is another caveat, which is that, even if firms do not compete, they may dissipate their rents in other ways, so that their profitability may not be that great—I could talk about French supermarkets as an example. Generally speaking, this is a task better left to regulators than to competition agencies. So I do not see that avenue as very useful for our purposes. Then there is the possibility to look at behavioural patterns that are naturally associated with collusive behaviour, and probably less often associated with competitive behaviour. This gets closer to the question of whether there is collusion taking place at the moment, as opposed to “could there be collusion”. We will talk more about this tomorrow, but I think this is a good area for further research, both conceptual and empirical, although at the moment the theory is probably not sufficiently developed to be useful for enforcement purposes. One of the issues here is that, again, it is difficult to rule out normal

4 Ivaldi M., Julien B., Rey P., Seabright P. and Tirole J. (2003): The Economics of Tacit Collusion, IDEI Working Papers 186, Institut d’Économie Industrielle (IDEI), Toulouse, text available at http://ec.europa.eu/comm/competition/mergers/review/the_economics_of_tacit_ collusion_en.pdf. 5 Frézal S., (2006): “On Optimal Cartel Deterrence Policies”, 24 International Journal of Industrial Organization |23|.


14

12/10/07

12:16

Page 14


competition. Indeed, the standard of proof is: “are you sure that this cannot be associated with normal competition”? But then, you will find lots of imaginative people coming forth with one particular scenario in which the behaviour would be compatible with normal competition. So, this may be useful for screening, but it would be difficult to rely on this approach to be sure that there is collusion going on. In my written contribution I discussed a couple of examples showing how easy it is to make standard “usual suspects” like parallel prices or stable market shares compatible with normal competition. I even made the point that, if you look at these indicators from a particular angle, they would be more associated with normal competition than with collusion. There is also a bit of a cat-and-mouse problem if you rely on this approach: you look for specific patterns, and if you find them, you think there is collusion; but of course, then the cartels see that the agencies are looking for this particular pattern and they change their strategy, so you have to revise your approach, and then again the cartels find out, and this can go on for some time. Joe Harrington’s point about how to distinguish between explicit and implicit cartels also illustrates this point. Turning to the “fire alarm” approach, the idea here is to encourage those who have the information to come forward—or, in other words, to destabilize cartels. This approach relies on the idea that cartels are illegal, and therefore parties have to rely on self-enforcement of their agreement—which, in turn, depends on the trade-off between the short-term gains that one would have from cheating on the cartel and the long-term benefits from continuing to collude. The idea is to exacerbate those tensions. Here you can play both: with the stick, and reduce the benefits of continued collusion by increasing sanctions both for firms and individuals; and with the carrot, by enhancing the net gain from reporting through rewarding informants and protecting them from retaliation and sanctions (including civil sanctions—sometimes you have to be careful about the links with criminal sanctions). I think that this approach has several advantages: it is less costly, good for detection and prosecution, and more directly linked to the deterrence purpose. In addition, this may also be a way to encourage the parties to actually keep evidence, because they know that they can apply for leniency when the time comes. There is an emerging body of literature on leniency programmes (and several people in this room have contributed to this literature: Massimo Motta, Joe Harrington, William Kovacic and myself). There are a couple of messages that come out pretty clearly from this literature. For example, you do not want to send the signal that one can engage in a cartel and then simply apply for leniency if the cartel is exposed. So, it is good to restrict leniency to a limited number of informants, probably the first one; this way, you get all the benefits of encouraging reporting while avoiding the negative effects of lenient treatment. You should also grant leniency even when a cartel investigation is already on the way (this is a point that Massimo Motta made in a


12/10/07

12:16

Page 15

Panel Discussion

15

joint paper with Michele Polo6). There is complementarity between corporate and individual leniency programmes, so it is good to have both, in order to exacerbate the tensions between companies and the employees that may become informants. In a paper co-authored with William Kovacic7 we argued this point, and we also suggested that it may be a good idea to offer bounties (monetary rewards to informants, particularly to individual informants), e.g. in the form of a share of the fines imposed on the cartel members. Another paper that I am completing at the moment8 shows that the carrot is more effective if there is a stick. We do not want to send the wrong signal with leniency and, for this purpose, it is important to have a strong stick, or more precisely to maintain an appropriate balance between the carrot and the stick. Here as well, economic analysis can contribute to improving the effectiveness of leniency programmes. JOE HARRINGTON—The fight against cartels is carried on in three stages: detection, prosecution and penalization. Historically, the role of antitrust authorities has been very important in terms of prosecution, and in the EU it was obviously critical also for penalization. In recent years it is also becoming important in terms of penalization in the US, but with very few exceptions it has not been very important when it comes to detection. Today I will talk about an increased role for the antitrust authorities in detecting cartels. Paul Grout and Patrick Rey already touched upon this, and I want to add to their comments by focusing on the activity of screening, i.e. trying to identify industries that are relevant for investigation. As mentioned by previous speakers, there are two principal methods to do that: the structural method, where you look at industry traits—as Paul Grout did—and see if these are conducive to collusion; and there is a behavioural method, where we look at prices and quantities or any other patterns that might be influenced by collusion. I will only talk briefly about the structural method, and will focus more on the behavioural method. As mentioned by previous speakers, there is a list of industry traits for which there is consensus that they are conducive to collusion. I do not want to be very pessimistic about the structural method. Several people presented it as complementary to other methods: I think that this is the right approach. I am pessimistic about the structural method of detection being used in isolation, as there is a risk of having too many false positives. When we think about “the perfect store” for collusion, we have in mind two firms, identical products, etc., and we expect a cartel to come out of it, whereas in most industries with those

䉴

6 Polo M. and Motta M. (1999): Leniency Programs and Cartel Prosecution, IGIER Working Paper No. 150, text available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=165688. 7 Aubert C., Kovacic W. and Rey P. (2005): “The Impact of Leniency and Whistleblowing Programs on Cartels”, 21 International Journal of Industrial Organization 347. 8 Chen Z. and P. Rey (2006): “On the Design of Leniency Programs”, mimeo.


16

12/10/07

12:16

Page 16


characteristics there is no cartelization. Anyway, that statement is based on what we know about cartels that were discovered. Otherwise, I think that this is a real concern, and theory says this is a problem of admitted variables. Economic theory says these factors are conducive to collusion, but there may be many outcomes, and collusion is just one of the possibilities. Economic theory cannot really say when collusion is chosen over a non-collusive solution when both are possible. So, the problem is that there are several factors indicating when collusion is possible, and those things are non-quantifiable. The structural approach is based on industry data indicating where it is more likely that a cartel will be formed. The behavioural approach is, I think, more promising, because it is based on evidence of actual collusive behaviour. There are two broad approaches to implementing a screening approach. One is to identify collusive markers—this is based very much on economic theory and the empirical evidence we have, and I will talk next about this in more detail. And the other is looking for a structural break in the market datagenerating process, which is something that could be done without much use of theory. In terms of collusive markers, we are looking for several characteristics of collusive behaviour, as reflected in prices and quantity, etc. We look into these and say: while they are not inconsistent with competition, they are probably more consistent with collusion. This is why it is important to keep in mind that this would be the best method for identifying the industries to investigate, and it is not a method for identifying collusion per se. For example, a lot of cartels have engaged in buy-back schemes, and that will certainly lead to a negative correlation with market share over time. A feature that has been verified theoretically and empirically over time is that market shares can be highly stable where prices are strongly correlated across firms, because that is very much why cartels can emerge under competition. Another interesting point is that a number of cartels have pursued the “home market” principle, that is, the idea that a cartel member should largely supply its domestic market. The implication is that, upon collusion, prices rise and imports fall. That would be sort of a “hot property” coming out of the competitive model. A third feature noticed in a number of cartels is that they move towards having a higher regular price and fewer discounts, and that is again something that is not so easily coming out of competition. A fourth feature, for which a number of theories have been produced, is that the price variance tends to be lower with collusion. This is reflected in a recent FTC study9 examining collusion at auctions of frozen perch (a type of fish). What we have here10 is the price series in three phases. The first phase, 9 Abrantes-Metz R., Froeb L, Geweke J. and Taylor C. (2005): A Variance Screen for Collusion, FTC Bureau of Economics Working Paper No. 275. 10 The speaker refers to Figure 3 in his written contribution to this volume.


12/10/07

12:16

Page 17

Panel Discussion

17

on the left, is during collusion (the cartel had not yet broken down), then the second phase is a transition phase towards the end of the cartel, and on the right is the ex post cartel situation. What we see is that, indeed, prices were much more stable when firms were colluding. The idea here is that you want to look at the variance of price for a given market and compare it to some sort of benchmark. That benchmark could be from the past experience of the industry, or other industries—but I would be of the idea of using a collusive marker. It is important that this screening be based on easily accessible data, but sometimes data on prices and quantities may be hard to get, and at any rate, a sophisticated econometric analysis would not be feasible. The second approach is to look for a structural break in the market datagenerating process. Collusion has to change that process, so it should be detectible. If you want to apply economic theory, here you would look for some sort of change in the process. For example, change of prices, or the price-cost relationship has changed, or the relationship among firms’ prices has changed (this is particularly relevant to auctions, and the relationship between bids), or there has been some variance of price versus market share. I will try to illustrate this by looking at the recent case of collusion in some NASDAQ securities markets.11 This was a clearly debatable case of collusion (although they did settle for a large amount of damages), but it is a case where the market (for example, the market for Intel and Microsoft securities) forms the market-players. What we have on the horizontal axis shows when a facilitating practice was adopted, and the vertical axis shows the spread for the stock, which is the price-cost margin. There is a very clear break in the pricecost margin. This is something that scholars have stumbled across, but I can easily imagine that, if these markets had been monitored (and in this case the data was easily available) they would have identified the break in the pricegenerating process. So, the point here is that if there are data available, the screening is easy to conduct. I will conclude by mentioning that I have made this presentation at similar venues before, including at the US Department of Justice, and the reaction was, “well, we are busy enough as it is, we do not need more cases”. This sounds to me like, if I were able to come up with a method to distinguish between accidental and intentional homicides, and the police would tell me “no, we do not need it, we already have too many homicide cases as it is”. Their argument is, we do have leniency programmes, so why do we need more? The point is that leniency programmes really seem to work best when there is some suspicion about collusion, i.e. firms get the sense that they may be detected. Screening can potentially play that role. Last Fall, Scott 11 See Figure 4 in the speaker’s written contribution to this volume. See also Christie W. and Schultz P. (1999): “The Initiation and Withdrawal of Odd-Eighth Quotes Among Nasdaq Stocks: An Empirical Analysis”, 71 Journal of Financial Economics 317.


18

12/10/07

12:16

Page 18


Hammond of the US Department of Justice gave a talk12 where he said that you cannot catch a thief with an economist. He may be right, but my point is, it may be sufficient to scare the thief with an economist, and screening may have a role to play in that. 䉴 RAFAEL ALLENDESALAZAR—In this presentation I will attempt to summarize how EC and Spanish courts have tried to distinguish between parallel conduct that responds to the normal structure of the market and parallel conduct that is an implicit cartel. Cartels are the worst cases for a competition lawyer. Nowadays it is extremely difficult to come across an explicit cartel; most of the cartels are implicit. Many of our clients act in oligopolistic markets, where competitors tend to be interdependent and adapt their own strategy to the expected reaction of their competitors. The issue is how to tell when this kind of conduct corresponds to some kind of collusion. As you know, Woodpulp II is the main EU case relevant to this context.13 It is an extremely important case because it involved parallelism but also what would be considered a facilitating practice. The fact that the companies announced the prices in advance, and there was an almost simultaneous announcement by the companies of the prices they were going to apply in the future months, is usually considered to be a facilitating practice. The competition authority considered this to be a practice caught by Article 81 EC, whereas the ECJ disagreed, and established that facilitating practices are not by themselves forbidden, as sometimes they may respond to pro-competitive conditions. In this case, it was found that the facilitating practice could have an alternative explanation that was not caught by Article 81 EC. Essentially, the Court’s decision established that parallel conduct cannot be regarded as proof of collusion unless collusion is the only plausible explanation for such conduct. This is a very high standard of legality. When counselling companies, we have to be aware that competition authorities do not always take into account this standard. Once they find facilitating practices, they tend to presume that there is collusion, and thus in practice the burden of proof shifts to the companies involved. This kind of proof is often extremely difficult to produce, because the result may be exactly the same whether they have colluded or simply adapted their practice to market conditions. I think it is important for the competition authorities to look at the actual conditions of the market. In

12 Scott Hammond (2005): “Ten Strategies for Winning the Fight Against Hard-Core Cartels”, speech delivered at the OECD Competition Committee Working Party No. 3, Prosecutors Program, Paris, 18 October 2005, text available at http://www.usdoj.gov/atr/ public/speeches/212270.htm. 13 Joined Cases C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85, A. Ahlström Osakeyhtiö and Others v. Commission (Woodpulp II) [1993] ECR I-1307.


12/10/07

12:16

Page 19

Panel Discussion

19

some cases they have looked at the nominal price lists, and not at the real conditions applied to the customers. This is not only a logical mistake, but it can also be interpreted by companies as an incentive to stick to nominal prices, a practice that would make the market more transparent, thereby making collusion easier. The difficulty of distinguishing between collusion and non-concerted practices is even greater in jurisdictions like Spain, where the law contemplates conscious parallelism as something different from collusion. It is true that this distinction (borrowed from the US) has seldom been applied, but when it was applied, it was done in a confusing way, because the Spanish court established that conscious parallelism is different from a concerted practice, and an autonomous action by a company that consciously adapts its strategy to that of a competitor may be an anticompetitive practice, and sometimes it is synonymous with a concerted practice. In any case, what this shows is that when you water down the standard of legality it becomes difficult for companies to defend themselves. To sum up, I think that, on the one hand, it is essential to fight cartels, and of course we cannot always hope to find a “smoking gun”, i.e. all the documents and correspondence. On the other hand, there is a real danger that, if competition authorities chase parallel conduct which is not really associated with collusion, and condemn spontaneous parallelism, they will prevent companies from adapting themselves intelligently to market conditions. Furthermore, the remedies are unworkable: what can companies do if parallelism is really natural? I was happy to read in Patrick Rey’s paper that economists will develop tools to distinguish supra-competitive markets from normal imperfect competition. This is the basic problem: to know when imperfect competition is normal. But it is essential to avoid false positives, because companies in these cases do not have the means to react in a rational economic way, so they find themselves obliged to collude so as to apparently diverge and thus avoid having competition authorities go after them.

RICHARD WHISH—In my presentation I want to address the question of what a competition authority should do if it discovers parallel behaviour in an industry, but there is no “smoking gun” and there has been no immunity applicant to provide the authority with evidence of collusion. Maybe there have been dawn raids, and not sufficient evidence was found to start infringement proceedings under Article 81 EC. Is there anything else that the competition authority can, or should do in those circumstances? Or, to put the question otherwise: is Article 81 EC the only tool in the tool kit? Of course, it is perfectly possible to have parallel behaviour in a competitive industry. If the explanation is that firms charge at their marginal cost, there is no antitrust problem. But there are also a variety of ways in which there can be a market failure, and there are different ways of addressing this

䉴


20

12/10/07

12:16

Page 20


failure other than through proceedings under Article 81 EC. Why might a market be uncompetitive? Perhaps consumers lack proper information, or indeed, they might be given misleading information. Switching costs may be very high, there may be technical problems which inhibit switching, and there may simply be switching inertia on the part of consumers. There may be very high barriers to entry, which often are attributable to governmental restrictions of competition. What can a competition authority do in these circumstances? I wanted to share with you two of the devices used by the Office of Fair Trading to deal with these problems. The first is the initiation of what we call “market studies”, and the second is the making of market investigation references by the OFT to the Competition Commission. As to market studies: under the Enterprise Act of 2002, the OFT has the general responsibility to keep markets under review. In pursuing this function, the OFT asks the question of whether markets are working well for consumers. It takes a pro-active stance in trying to identify examples of market failure and to think of ways in which they can be addressed. It is important to note that the OFT is not only a competition authority, but also a consumer protection authority, and it enjoys a wide variety of powers not only under Articles 81 and 82 EC and their domestic equivalents, but also in relation to unfair contract terms, misleading information, the provision of information to consumers, etc. The OFT initiates about 7–8 studies of this kind in any given year. Sometimes these are fairly short, taking 6 to 8 months, and they are essentially of a fact-finding nature. Other times they can take a year or longer, in which case the study is much more penetrating. In terms of the outcome of a market study, there are a variety of actions that can be taken, and I have listed those in my written contribution. The list is taken from the OFT Guidance on conducting market studies.14 The possible outcomes include: deciding that the industry should be given a clean bill of health (as there are no market failures); assisting consumers by providing them with information, or better information; encouraging an industry to take voluntary action in order to cure market failures; or sponsoring the adoption of codes of practice within the industry. The OFT also has an advocacy role vis-à-vis the government in cases where it identifies public restrictions of competition. The outcome of a market study may be the initiation of proceedings under the relevant provisions of competition or consumer protection law. In some circumstances, market studies are followed by a market investigation reference to the Competition Commission. In my written contribution [in this volume] you will find a table of all market studies that 14 OFT (2004): Market Sudies: Guidance on the OFT Approach, London, text available at http://www.oft.gov.uk/NR/rdonlyres/03AD31EC-3533-40BE-AE79-2A83424E9C9B/0/ OFT519.pdf.


12/10/07

12:16

Page 21

Panel Discussion

21

the OFT has commissioned to date. They cover a wide variety of types of behaviours. Such studies do not in any sense address specifically the oligopoly problem, but my point is that the oligopoly problem could be addressed through the initiation of a market study. The other tool that I referred to before is the initiation of market investigation references to the Competition Commission. This instrument has a specific legislative basis in the Enterprise Act of 2002. It is explicitly a competition law mechanism, a supplement to the enforcement of Articles 81 and 82 EC and their domestic equivalents. It is, in a sense, a safety net: certain problems that cannot be addressed through conventional antitrust investigations may be addressed through the market investigation system. The Enterprise Act of 2002 establishes that the OFT can make a reference to the Competition Commission where any feature, or combination of features, of the market may prevent, restrict or distort competition in the UK or part of it. When the case reaches the Competition Commission, its function is to address that question and determine whether there are any adverse effects on competition. If this is so, it has a wide range of remedies available to it, including for example the discontinuation by order of abusive conduct or anticompetitive agreements, or a requirement that information be made available to customers and to consumers. It is also possible to order somebody to supply customers downstream in the market, and in very serious cases it is even possible to order structural remedies, e.g. compulsory divestitures in order to make the market more competitive than it currently is. In a Guidance document published by the OFT and the Competition Commission it is made explicit that the oligopoly problem is one that the two institutions consider could and should be addressed under this particular mechanism. Another point that I wanted to make is that there is a system whereby, instead of the OFT making a reference to the Competition Commission— which might then last a considerable amount of time, i.e. up to two years—the OFT can negotiate with the undertakings concerned commitments in lieu of a reference to the Competition Commission. The system is still fairly new: we only had one case that has gone all the way through to a final determination, and there are four others in the pipeline, one of which is in the retail banking sector in Northern Ireland, which looks like it could be an example of the oligopoly problem in circumstances where there is no actual evidence of collusion. To conclude, there are obvious similarities between the market study system and the market investigation reference system, on the one hand, and sectoral investigations under Article 17 of Regulation 1/2003 on the other (as we know, there are two industries being scrutinized under this procedure at the moment).15 But the notable thing about market investigation references is that, in the end, there are remedies that can be imposed as part of that 15 Energy and financial services—see DG Competition’s Report on Energy Sector Inquiry, SEC(2006) 1724 (10 January 2007).


22

12/10/07

12:16

Page 22


mechanism, whereas sectoral investigations under Article 17 of Regulation 1/2003 do not lead in themselves to the adoption of remedies. MARGARET BLOOM—I wanted to ask Olivier Guersent: how much has economic analysis been used in relation to ex officio investigations undertaken by the Commission? I would also ask Thomas Barnett: what is the share of cartel cases started ex officio by the US Department of Justice, and how effective was the use of economic analysis in such investigations?

䉴

OLIVIER GUERSENT—In DG COMP we established a working group to discuss the relevant factors in the scrutiny of individual cases, and the economic tools that could be used when scrutinizing sectors that are the best candidates for our cartel investigations. We did use economic analysis in the past, but the rate of success tells me that we might as well not have used it at all. In about half of the cases it proved to be successful, and in the other half it proved that, although we had the same economic indicators, we did not find any evidence of collusion.

䉴

䉴 THOMAS BARNETT—The short answer is that we use a wide range of tools to identify cartels—and here I would insert a comment in response to Prof. Harrington, who seems to believe that we use only amnesty programmes. His remark reflects some ignorance about how we go about cartel investigations. I am not saying that we engage in a wide range of econometric studies to detect cartels; a lot of what we do is based on cost-benefit analysis. Olivier Guersent suggested that, with market studies, we might end up using all our resources in the Antitrust Division, that there are issues about false positives to be considered in that regard, and the question is how much we are going to learn from such exercises—as we may see changes in price structure that are explained by reasons other than collusive activity, etc. In response to that, there is one example, which is fairly public, where the FTC and the US Department of Justice work together (to some extent it is more the FTC than us), namely that of the gasoline industry. The FTC monitors pricing in the gasoline retail market on a regular basis, looking for anomalies and determining whether those uncover some sign of collusive activity; then the information is referred to the Department of Justice. The FTC spends enormous resources doing this, and it is not really feasible to try to do this across a larger range of industries. On other occasions we have employed other tools and resources, such as phone calls, document requests, or other methods. So we are somewhat selective about where to direct our efforts.

JULIAN JOSHUA—The cases that Olivier Guersent talked about were ex officio cases that started to give results only in the last five years—I can speak

䉴


12/10/07

12:16

Page 23

Panel Discussion

23

with a certain personal knowledge about this, because most of them were cases from my earlier life, and they are only now filtering through. As far as the use of economic analysis is concerned, all I can do is praise another economist, Prof. Elzinga from the University of Virginia, who expressed himself in terms similar to Scott Hammond,16 saying that, if he had a limited budget, rather than employing economists to track down cartels, he would hire Hercule Poirot, Father Brown and Sherlock Holmes . . . But it is true that the role of economists is very important in connection with screening devices. Many of those cases that came up on an ex officio basis at the Commission arose in all sorts of circumstances, for example following tips from informers, but very few came up as a result of actual “detective work” done by economists, and that happens because of the false positives. I remember that the DG COMP library, which was a very dusty place in the old days, contained— besides very good old Trade Case reports and lots of US case law—a very useful publication called The European Chemical News. It was very interesting reading, as it publicly announced things like industry “price initiatives” for such-and-such product. So I do not think that economic analysis as such ever managed to really point the way. In the US, I do not think that cases would ever be handled without the help of economists, whereas the Commission sort of seems to sometimes manage without. With regard to Woodpulp II,17 this was one of those cases where, at the time, the Director General was very proud that the Commission was going to prove collusive behaviour by way of economic evidence. The case involved not only parallel pricing and a facilitating device in the sense of a base-point system, but there was actually also a lot of actual collusive evidence: eight of the core members had established something called “the Bristol Club”, named after the Bristol Hotel in Paris, where they met to fix prices. The trouble with that case as it was presented was that, you had on the one hand these hard core carrel members, and on the other hand several other participants who were adopting a base-point system. The Commission did not really distinguish between those different forms of conduct, and so there was a failure to distinguish between hard-core collusion and mere adherence to a facilitating device, which nowadays would be considered a concerted practice. JOHN RATLIFF—I wanted to pick up on one of the points made by Prof. Harrington in his paper, where he explains that the omnibus question in the US is extremely useful in bringing forward cartel revelations. I am one of those who would advocate not using the omnibus question. The reason is that, I 䉴

16 17

See supra note no. 12. Supra note no. 13.


24

12/10/07

12:16

Page 24


think, if you overload the burden on the company there is a risk that this will operate as a disincentive, discouraging the company from coming in at all. You want them to come in with a big cartel case that they worry about, but you also want to force companies to reveal other cartels that they are involved in, so that they start, wondering if their balance sheets will actually survive the treble damage suits. With the omnibus question the stakes change greatly, and you may find they do not come in at all. In general, the way this should be done is through incentives, and not by imposing further burdens and conditions on going in. An amnesty programme is fine, because it leaves in place the incentives to go in. Incentives to tell more are also fine, but I do not see why the companies should be penalized for not going in with all the other possible cartels found. They may have good reasons for not doing so. Paul Grout also suggested in his paper that we should not rush to judge the company that does not come in and use amnesty plus, because there may be other reasons which explain their behaviour. The core point is that we should not add more disincentives to going in through having conditions that are too elaborate and sanctions for revealing other cartels. We should proceed case by case. It is already a big victory if one cartel is revealed, even if others are not. In practice companies may well anyway decide to stop their other collusive activities after the first case. This is a bit like many other types of regulatory compliance: a company does not necessarily confess it, but will often rectify it as best it can. That should be a good enough result for the antitrust enforcer in many cases. 䉴 JOE HARRINGTON—I find it hard to figure out how cartels are really discovered, because things can be misleading at first glance. For example, in some EU cases of application for leniency, even if the leniency did not produce any kind of information per se, the cases were connected with US cases—but, then again, US cases are really hard to see through. I think there really is a lack of understanding about what triggers initial discovery. There are thousands of industries out there: what was the first event leading someone to think there is a cartel in a particular industry? This question is relevant to a lot of the policy issues we are discussing today, because it gets to the point of what leniency applicants are responsive to. If leniency applications come in because of a heightened sense of prosecution, then they may come forward regardless of other concerns. If, instead, it is truly spontaneous reporting— the company CEO wakes up one day and decides he wants to lead a new life— then the issues about overburdening become relevant. The problem is, we are at a point of ignorance—or, at least, I am—about this process. Probably the world outside competition agencies, and lawyers in particular, have better information in this sense. We need to understand what leniency applications are responsive to, and that makes it necessary to identify “time zero” when it comes to the detection of cartels.


12/10/07

12:16

Page 25

Panel Discussion

25

WILLIAM KOVACIC—One of the provocative questions that comes out of the presentations made by Joe Harrington, Patrick Rey and Paul Grout is: where to look for cartel activity? Suppose we think of designing a general monitoring programme: you want to look for a combination of structural and behavioural characteristics. How do you decide where to look? What investments would you make in order to begin formulating the structural and behavioural profile that might lead you in more specific directions? I want to offer a couple of thoughts about what those screens might be. One is, I would look for markets characterized by some form of public barrier to entry—that is, a public policy that in some form restricts or limits entry. Several of the cases mentioned here so far reinforce this idea. Frozen perch18 was about a defence procurement market, and critical to the operation of the perch cartel was a domestic content restriction that forbade the US purchasers from buying perch offshore. Had that restriction been eliminated, the US would have been buried up to 100 feet in perch, but because the perch supply was limited to 4 US firms which had been supplying all the perch to the Department of Defence over several decades, that facilitated their option to cooperate. One possible lesson is that when one notices systematic public efforts to restrict entry, that may be the place to begin hunting for certain behavioural clues consistent with possible cartel behaviour. A second example, which I think illustrates Richard Whish’s combination of competition and consumer protection perspectives, is the eyeglasses market. In the US, one of the leading FTC initiatives in the 1970s and through the 1980s was to look for restrictions on the provision of optometric services. Two things came out of this exercise: one, it was possible to identify, by way of careful economic studies and a side-by-side comparison of jurisdictions, states which imposed restrictions on who could make and sell eyeglasses, and states that did not. It was thus possible to document in a convincing way that, in jurisdictions without barriers to entry, the prices for eyeglasses were much lower. This led the FTC to initiate a rule-making process that struck down some of these restrictions. Second, one of the solutions to stimulate rivalry that came out of this exercise was an information-based solution—specifically, to take the simple step of mandating the physician to give the patient a copy of the prescription, so as to facilitate the patient’s choice of going to other suppliers to get the eyeglasses. So, the second thought is about where to start looking when deciding where to hunt: are there public policies that stimulate the accomplishment of the Stiglerian task which is central to making collusion work? Public barriers to entry strike me as a good example along those lines. A thought on the discussion that Joe Harrington, Margaret Bloom and Thomas Barnett had about specific industry monitoring: from 1992 to 2006, 䉴

18

See the paper cited supra note no. 9, at pages 5–7.


26

12/10/07

12:16

Page 26


the FTC has published fairly substantial reports on the US gasoline sector. I can confirm Thomas Barnett’s comments: I doubt that there was a single petroleum industry document the FTC has not looked at over this time, in search for evidence that would suggest parallel action. Had we found it, I would now be ordering champagne for all of us, but so far we have not found it. But it is interesting to note that, though there is a massive use of resources to do such market inquiries, they probably offer some insights as to what are the practices that facilitate cooperation. A merger study published in the 1980s quotes— without disclosing the name of the firm in question—a document that discusses the implications of a US Supreme Court decision from the late 1960s, Container Corp.,19 which prohibited firms from calling each other to verify prices on a routine basis. The document in question originated from a senior marketing executive in a major oil refining company, and was addressed to his colleagues in the industry. The letter was more or less along the following lines: “I want to tell you about a distressing development, the recent Container decision of the Supreme Court. I will tell you what I think it means. The Supreme Court said that we cannot in real time share information about actual prices. Before this decision, we were able to call up competitors in the industry and confirm whether there were or not actual deviations of price, and if the deviation was a mistake or intentional. This allowed stability in the retail market price.” This sort of evidence confirms the intuition in the economic literature, that this kind of real-time price monitoring and information exchange arrangements have serious possibilities for facilitating cooperation. Finally, based on the presentations made up to this point, I was thinking that the emphasis on structural and behavioural traits has enormous implications for merger enforcement. This is an area where the possibility of tacit coordination counts a great deal for the evaluation of coordinated effects. Suppose we try to apply this to the hunt for express cartels—the refinement of these techniques has enormous implications for the proof of likely anticompetitive effects in merger review. 䉴 MARIO SIRAGUSA—I would like to go back to Prof. Harrington’s question as to what may motivate a company to apply for leniency. I wonder whether, at least in some circumstances, the decision to apply for leniency could also be a competitive move, or even an exclusionary move—in other words, an attempt to cause damage to competitors. Very often the companies that are the core organizers of the cartel are also the ones who go for leniency. In a system like the European one, where we only have pecuniary sanctions for cartel activities, this is a perfect tool for bringing damage to competitors. If you get leniency, you do not pay, but the competitors are going to pay. Prices

19

United States v. Container Corp. of America, 393 U.S. 333 (1969).


12/10/07

12:16

Page 27

Panel Discussion

27

will rise afterwards, and you recover the sanctions you have paid through the increased prices. Very often the marginal competitors are forced out of the market. I had a case where this clearly happened as a result of the leniency application: the cartel leader succeeded in buying up all the other market participants, thus monopolizing the market. This is an abominable effect of the leniency programme. So, I would ask you to consider this unorthodox view of leniency programmes as well. I did not hear any reference in the Commissioner’s speech to the rights of defence in leniency applications, but only an almost triumphant description of the effects of leniency. As a lawyer who defends companies in this situation, I see very often leniency cases that rely on indirect evidence—for example, statements by lawyers without any right of crossexamination for the defendants. I think this is a significant weakness of the EU leniency system, and I hope that the Court will strengthen the rights of defence in this area. I seriously believe that some companies are motivated in the sense that I mentioned: leniency is a very good way to cover up years of cartel activity, inducing everybody else to share the cost of the cartel. I am very worried because I think that the current system lacks equilibrium. I am not against leniency programmes, of course, but I think that there is something missing. Maybe it is just that for now there are only pecuniary sanctions, which make leniency a very important anticompetitive and exclusionary tool. This damages particularly the small companies, which are often coerced to participate in cartels. They do not have the knowledge and skill to process a leniency programme, so they become victims of the leniency programme and sometimes are even expelled from the market. JAMES VENIT—I wanted to reply to John Ratliff’s comment about the omnibus application. There are means to get around the treble damage disincentive. The US has done this by abolishing treble damages for leniency applicants. In my personal experience, the greatest barrier to leniency applications, at least in Europe, is insisting on the kind of evidence that is very hard to come up with. Fortunately in the case I was involved in we had a “smoking gun”, but if we had not had one, I would have been really concerned about the barrier to entering the leniency programme. I would think that nobody in their right mind would make a leniency application unless there is something behind it, because there is nothing to be gained by coming in if there is nothing to be proven. Another comment about detection and motivation—I am not going to comment on Mario Siragusa’s points, which I found fascinating. I would think that less protection is a motivation for coming in a leniency programme. One of the questions I always had in mind was whether you could use corporate culture to secure these laws, to change the incentives for engaging in cartel behaviour and for reporting it. If the corporate culture regarded cartels

䉴


28

12/10/07

12:16

Page 28


as an investment that is just too risky to undertake, then compliance standards and behaviour would change rather radically. This is something that we need to consider. 䉴 STEPHEN CALKINS—Richard Whish talked about the benefits of having a variety of tools. Joe Harrington asked how they are applied. I recall that last week I was reading in US newspapers that the price of chickens has been plummeting at the wholesale level and there has been no change in the price at the retail level. Anytime that you see this kind of dramatic change at one end and none at the other, the US Department of Justice ought to consider making a phone call to see if it can stimulate leniency applications. I pick Paul Grout’s presentation to both praise and criticize. I think it is highly commendable to look at the facts and hold them up against the theory. The criticism is that he is not doing quite enough of the latter. Indeed, he puts up a list of standard characteristics of when a cartel is likely to be present, and he does such a good job of it that other people will then refer back to his list, but his list is disproved in part by the rest of his paper. For example, he says that cartels are likely to occur where we have a concentrated industry. And yet, his own paper shows that this is incorrect. Indeed, many duopolists do not need to fix prices in a smoke-filled room because they can find more subtle ways to coordinate prices. I wish we would stop thinking that concentrated industries are where you can find cartels. In his paper, Grout also says that barriers to entry do not seem to be a relevant factor, so one should remove them from the list. I am not sure I agree with that, unless I knew what his definition of entry barriers was. I agree that the absence of what some may have called entry barriers does not seem to prevent carrels. So perhaps those asserted “barriers” should not be on the list. Otherwise, I can only applaud the application of reality to theory, and wish that economists would take a step further in refining the list of cartel indicators, thus aligning theory better with discovered realities.

MARK PETER SCHINKEL—As a relatively young economist, I am totally excited about these possibilities of detecting cartels, and it seems to me that we should explore that route. The base-point price theory is difficult to understand, and law enforcers should have a hold of it before going to actually look for those points. It is interesting that in similar law conferences I got the same type of reaction from young people interested in developing all sorts of economic tests, etc., but the enforcement agencies seem to be quite reserved with respect to this avenue, and are more enthusiastic about leniency programmes, whereas I personally have doubts that this factor is strategically considered by cartels. My question is addressed to some of the people in this room who were present at the beginning of the leniency programme. I wonder whether this is normal institutional inertia, to be resistant to the

䉴


12/10/07

12:16

Page 29

Panel Discussion

29

development of new detection tools, or whether there is another explanation. So, was this scepticism present at the beginning of the development of the leniency programmes? JOHN FINGLETON—I am an economist who also works in an enforcement agency, so I am perhaps more likely to see these issues from both sides. Screening can be very useful, and I also find the retrospective analysis of cases very useful. William Kolasky’s paper also looked at the features of previous cartels uncovered by the US Department of Justice for interesting insights. In the area of cartels, competition agencies are increasingly using a combination of screening and monitoring tools, because we operate in a target-rich environment, where there are a lot of cases to take up, and we need to prioritize our resources. Economic analysis is probably less valuable for cartel investigations than in other antitrust areas, because a lot of the evidence we want is not that visible, and that is why leniency is more important. On Joe Harrington’s point, the complementarity between leniency and screening is even more important from the competition agency’s point of view, because it works in the other direction: when you have a lot of potential cases, and you have to choose among them, you probably want to choose the ones where the economic and structural features of the market indicate that there are likely to be X-inefficiency, higher prices, etc., so that the impact of bringing the case will be greater. Some knowledge about how markets function can be useful. Oil prices are a good example in this sense: governments tend to want competition authorities to look at oil prices when crude oil prices are high, while this is when the domestic distribution agents have the least possibility to operate a cartel. It is instead when prices are falling on the international market that competition authorities should be particularly vigilant. Little economic knowledge, i.e. “high prices are equivalent to a cartel”, is harmful. Some more detailed knowledge about how markets operate is good. Having an economist sitting at the side of the antitrust investigator is an essential ingredient. I cannot resist responding to Mario Siragusa’s gentle provoking: I think that his point ignores the ways of proper compliance. Many cartels might not exist if we have successful leniency programmes. We cannot just look ex post at particular cases without considering the ex ante conditions, i.e. the kind of economic behaviour induced by these programmes. In Ireland, there is now before the jury a very interesting cartel case involving home heating oil.20 One of the defendants in this case, a small firm, argued it was coerced into participating in the cartel, but then another firm from the industry declared that they had refused to participate in the cartel, and still remained in business—

䉴

20 For details on this case, see e.g. http://ec.europa.eu/ireland/general_information/key_issues/ competition/ index_en.htm.


30

12/10/07

12:16

Page 30


it was tough, but it was the right thing to do. The judge told to the first defendant that such an argument was a self-serving criminal act, more or less like thievery. So, the message to the firms is that they should not get involved in the cartel. Coercion is not a defence for cartel activity in a criminal law system, and it should not be so in a civil law system either. HANNS ULLRICH—I have a remark and a question. The remark is that Germany does not yet have too much experience with leniency programmes, but it does have some with the so-called “legalized cartel” situations, e.g. licensing contracts for distribution systems. In this area, under German law there is a requirement that the contract be concluded in written form. When this condition is not fulfilled, there are problems when one of the partners wants to get out of the contract. We have a similar situation in the case of cartels and leniency programmes: one of the parties to the agreement benefits from the opportunity to get out, and at the same time, it has the advantage of placing an extra penalty on its competitors. So, leniency programmes do have a drawback. You have to think that there is not always a clear-cut case of fixed prices. In many instances the parties are on the fringe of what is legal, and use this kind of programme for strategic purposes. My question is certainly a naïve one: the economists have told us under which conditions, and in which markets, we may or may not find cartels. Much of it is fairly conventional wisdom—we always knew that homogeneous products and the lack of transparency of the market favour cartels. Now, hard core cartels are not exclusively price-fixing cartels, and I think that bid-rigging is detected in other ways. I take it that most output restrictions would lead one way or another to price fixing. But what about territorial customer restrictions? Do we have similar economic theory to detect this kind of competition distortion? The reason why I am asking this is that, if you come up with a ‘perfect’ economic theory about how to detect price-fixing cartels, then company managers will find other forms of achieving their purpose.

䉴

OLIVIER GUERSENT—When we look at the relationship between deterrence and leniency programmes, there are a number of relevant factors to take into account, that we have not touched upon yet. John Fingleton explained that the first reason why we (i.e. the competition agencies) perform sectoral economic analysis is in order to prioritize our interventions. This is why the Commission has launched sectoral inquiries, for instance in the energy markets. In response to Richard Whish: sector inquiries are not some sort of academic exercise run by the Commission; they do have consequences, as the several dawn raids we performed following such sectoral inquiries indicate. I think that, from the point of view of an enforcement agency, one way of approaching the issue of the usefulness and complementarity between deterrence and leniency programmes is to do a qualitative analysis of leniency 䉴


12/10/07

12:16

Page 31

Panel Discussion

31

cases. If we did that, we would probably find that leniency cases do not display the same profile as ex officio cases. First of all, the former include a morethan-average proportion of near-to-die cartels. I mean to say that people do not usually rush to report young and well-working cartels; they come in when things are starting to go less well. I think it would be good for the Commission to do a retrospective analysis, now that we have had enough cases, precisely on this point. These are not the cartels we really want to catch. Then, to a certain degree, I would agree with Mario Siragusa: leniency is sometimes used not only as a safety device, but also as an offensive tool. But I do not mind. This is certainly not a concern for me as law enforcer. Usually, if there is continuity in the management of a firm, meaning that the same managers run a successful cartel for about 10 years, they will decide to report once the gains from reporting are higher than the ones from continuing the cartel. The cases where there is a change of management are different: the quality of the reporting is, in my view, higher. The same goes for what I call the “trouble shooters”. Here I have a very specific example in mind: that of a US company that has tried for a long time to penetrate the EU market, but was kept out by a EU cartel. Finally it joined the cartel in order to be allowed on the market, and then reported it. This is a perfect “offensive tool”, if you wish, but it is perfectly acceptable to me. The only way to ensure a reasonable proportion of good leniency applications is to also have effective detection. In its absence, you will only get reporting of dying cartels. In response to Mr Schinkel’s question: in my view, the reservation on the part of the enforcement agencies comes from the fact that, if you go about how you find out that a cartel is in place in a given market, then you get very close to opening the debate on the impact of the cartel, or its actual effects on the market, which is a debate that the Commission would like to avoid. 䉴 RICHARD WHISH—I can see how sectoral inquiries can be a useful exercise, which then allows the Commission to decide on the course of action. This is entirely consistent with my written contribution. The analogy was with market studies in the UK. My point was that, in the UK system, at the end of the market investigation remedies become immediately available to the OFT, whereas they do not become immediately available to the Commission at the end of a sectoral investigation. 䉴 IAN FORRESTER—I would like to echo Mario Siragusa’s point, and restate it from a different angle. The classic European cartel—rather old-fashioned— is a gathering of national champions who hold an industry meeting in some convenient hotel. At that meeting, the leaders of the cartel reach an understanding, which is fairly private, while other parties attend for interest, curiosity or legitimate trade activities. Then the cartel breaks down, and the leaders are heavily penalized. One of them, or a number of them, gets


32

12/10/07

12:16

Page 32


substantial or total immunity. The small players suffer in two ways: one is that they may not know all that was going on, although they suspected it, but it is virtually impossible for them to take the risk of complaining, because they would likely be eliminated from the market. More specifically, the way in which the fine is calculated for the small players, as a lump sum, is unreasonably burdensome upon them. Of course, we all do want to see cartels eliminated, but I do echo Mario Siragusa’s concern: the prosecution of national champion cartels along with a few smaller players, in a sense, brings about a certain amount of injustice for the latter. PAUL GROUT—Two short comments. One is on how to interpret the OFT report on which I based my presentation.21 That report should be viewed as an early look into what could be achieved if we think that particular approach is worth pursuing. The answer that comes out from the paper is clearly “yes”, we can learn something from applying this approach. I think that the issue of entry barriers is a very good example of the problem at stake here: different approaches give you very different answers. If you look at the aggregate data across industries, you do not get a clean message indicating that strong entry barriers means there is cartel activity going on. But when you get into the case studies, you see exactly why this is the case: it is not that entry barriers do not matter, but they matter in different ways for different cases. In some cases entry barriers help to form the cartels: if you cannot keep competitors out of the market, you actually cannot make any money. So in some sense the entry barriers are facilitating the cartel. But if you look at the shipping industry, this actually is a very easy market to enter—there are all these tramp vessels around the world that you can rent very easily. The reason why these cartels get discovered are two: one is that there is actually legalized sharing of information, and companies slip very easily from what is legal to do to what is not. Second, in the absence of entry barriers, companies already on the market share information and suddenly decide they are going to sell very cheaply when a tramp vessel tries to enter the market—and this is exactly how you discover the cartel. Therefore it is actually quite difficult to be able to read into the available data. Second, in terms of actually using any of this evidence, I would not advise you to use it for screening. The way that I would use it is when the authority gets some kind of flaky information from three different industries, all of which looks equally plausible, and the question is where to orient the agency’s efforts. In this sense, our material would be helpful to indicate that in one of these industries there is more likely to be a storm blowing. It is nothing more than that, so we should not try to oversell it (though on the other hand, when 䉴

21

Supra note no. 3.


12/10/07

12:16

Page 33

Panel Discussion

33

compared to lawyers’ fees, it is actually a cheap way of doing things—you just ask a few economists to work for a few weeks generating the data . . .). Another point: one of the things I find terrifying, and I wanted to look to in order to see if it is truly an international phenomenon, or only typical of the EU, is how all these cartels have been found exactly in the same industries. If you look at our US and EU data, in most of these industries no cartels have been found. The cartel cases appear in less than one third of the observed industries. The correlation between US and EU is enormous: if you take shipping out—and in Europe, clearly, shipping is different than in the US, so there is every reason to leave it out of the analysis—the pure statistical correlation is 80%, meaning that actually less than 20% of these cases are not common. On top of that, because a lot of these cases are actually in differently classified but actually very close industries, the true correlation is getting close to 100%. There are two observations to draw from this: one is that industry characteristics totally determine cartelization, and the other is, agencies find cartels in one jurisdiction and then other jurisdictions go after the same industry. There is an enormous reinforcing hysteresis effect, a process that has been going on totally informally. So, if the economic analysis mechanisms manage to randomize the process, getting different authorities to look in different places, I think that would be a good thing. One of the things we intended to do in our study was to go inside these groups of cases, identify the authorities involved and actually go to them and do a true case study. I seriously worry about the results of this data, because it indicates that in a sense there is almost too much pre-screening going on inside the authorities in a totally informal manner. 䉴 THOMAS BARNETT—On the point about strategic reporting of cartels in order to damage competitors, that might concern me if I thought that people entered cartels with the specific purpose of then reporting them in order to damage competitors. There is uncertainty about who gets through the door first, and in our jurisdiction especially that is a concern, but once a cartel is reported, I tend to agree with Olivier Guersent that, whatever the motivation, the fact of reporting is desirable. Second, trying to answer the question on what are people responding to when they decide to come in and report, our perception is that there is some sort of precipitating event that destabilizes a cartel. To give you a practical flavour of that: it can be random discovery by senior management; it could be a change in management; I remember a situation where through a document related to a merger the lawyers discovered the cartel; it could be the withdrawal or non-attendance of a participant in cartel meetings that creates uncertainty, or a phone call from the Department of Justice to serve a grand jury subpoena—and indeed, if there is anything there, typically this can lead to cartel immunity applications, customer complaints, or a change in the law. We (the DoJ) believe that the elimination of several liability in 2004 has


34

12/10/07

12:16

Page 34


increased the willingness of people to apply for leniency, to the point where sometimes we have had to turn applications down, because what was reported was not actually a cartel—which is an interesting position to be in for an antitrust agency. A final comment on Jim Venit’s point about corporate culture and securities law: I have come across some anecdotal evidence suggesting that the criminal sanctions have created a cultural environment within corporations, not only on compliance generally, but specifically with respect to document retention, to the point that the prospect of individuals destroying evidence, even after there is suspicion that the government may be after them, has probably dropped. Nowadays there is a sense that destroying evidence is such a bad thing to do that you simply do not engage in it. This may be an example of securities law driving a change in corporate culture. MASSIMO MOTTA—Also related to Mario Siragusa’s point, I think that, to the extent that the small firms are also the less efficient, and that may often be the case, then driving them out of the market is precisely what we want to obtain through a competitive process anyway. The cartel allows firms to maintain high prices, so the less competitive firms can also remain in the market. Once the cartel breaks down, prices are falling, and the less competitive firms are driven out of the market. In such cases the effect is desirable. Nevertheless, I think that there is something to the general points raised by Mario Siragusa and Ian Forrester, which is that, at some point, we should perform an ex post evaluation of leniency programmes, because not necessarily all goes in the right direction. For instance, I was reading recently a brief by an economic consultancy showing that, with the current leniency programme in place, the average of the fines imposed by the Commission has gone down. If this is going to give firms the impression that the risk of being caught colluding has decreased, it may prompt some firms to engage in cartel activity hoping for a more lenient treatment.

䉴

䉴 MARIO SIRAGUSA—I was referring to cases where the small firms go out of the market not because of the lower prices after cartel disclosure, but because of the fines imposed on them. Sanctions are bringing companies to bankruptcy, and then prices are not going down after cartel disclosure as expected.

PATRICK REY—I only wanted to emphasize again that we want cartels to be deterred in the first place. I take the point that we do not want to encourage firms to set up cartels and then denounce them, but at the same time, the lack of leniency can also be counter-productive. Also, the companies approached to join the cartel might be suspicious of the reasons why they were approached, and this serves deterrence purposes. Our objective is to destabilize cartels, to induce suspicion among the participants.

䉴


12/10/07

12:16

Page 35

Panel Discussion

35

A quick point on entry barriers: it was pointed out that entry barriers are measured by capital expenditures, and at the same time, we know that in R&D innovation-driven markets collusion is not an issue. But I think that if you look at other entry barriers, especially publically erected entry barriers, you would get a very clear positive correlation. So I think that we should not discard entry barriers too quickly. Finally, I wanted to join Joe Harrington’s claim for ex post investigation. Olivier Guersent has said that in the EU there are not enough cases for a meaningful ex post evaluation. I think that it would be useful to look not only at what triggered the case, but also to see what exactly makes the cartel work—in other words, what buttons we could push in order to destabilize cartels. 䉴 WOUTER WILS—A quick reaction on your point regarding the level of fines: I think it is very important to look at the overall evolution of the level of fines in the EU before the introduction of leniency. Of course leniency leads to a reduction of the fines (if I remember well, the study you referred to showed a reduction of 37%), but if before the introduction of leniency the level of the fines had actually increased, then the reduction is not a problem. Historically speaking, there is a very substantial increase in the level of fines before leniency. So if the level of the fines had doubled, a reduction of 37% means that the final level is at any rate higher than before the introduction of the leniency programme.


36

12/10/07

12:16

Page 36


Panel II: Investigation Powers of the Competition Authorities CALVIN GOLDMAN—I had the privilege of being the Head of the Canadian Competition Bureau at the time when our competition law was modernized and strengthened in two material areas: the review of mergers and cartel enforcement. With that experience, and the one acquired afterwards as a legal practitioner representing companies during cartel investigations, I will try to offer some thoughts not only about the Canadian experience with cartel enforcement, but also on some international aspects and issues that are currently of some significance to legal practitioners involved in this area. I will highlight three of them in my opening remarks. First, in the Canadian context, cartels are treated as indictable criminal offences, carrying severe sanctions, including fines and a potential for individual imprisonment. My first point, reflecting on the issues raised in the debate, is that the fact that in the Canadian system cartels are prosecuted criminally, and treated like any other crime, or in particular like any other fraud, creates the greatest deterrent effect possible. My personal belief, after thirty years of experience in this field, is that you do not actually have to send people to prison for an extended period of time. This issue was recently the subject of debate in a similar conference in Toronto, and previously we saw the same kind of debate in Australia, at the ABA and elsewhere. In Australia we had a long debate on this, as the Attorney General is moving forward with criminal sanctions. My belief is that it is the threat of criminal investigation, the threat to company executives that their names will appear in the newspapers as being under criminal investigation, that they may actually stand up in a criminal court and have handcuffs removed, that their entire life’s work may be tarnished by that kind of publicity, those are the maximum deterring factors. In fact, I remember one case where the CEO who was under investigation in the US before a grand jury and in parallel proceedings in Canada needed serious treatment by a psychologist—eventually he was not convicted, but the fact that he had been under criminal investigation threw him into a kind of despair that he had never experienced before. I am not suggesting that imprisonment terms are unwarrented, but the deterrent effect comes from the fact that individuals will be imprisoned. It works, and it should continue to be applied, in my view. Second, from the experience in Canada, the coming into force of the MLAC—the Canada-US mutual legal assistance treaty that came into force in 199022—and therefore the ability to have parallel criminal proceedings and effective sharing of information between the US and Canadian enforcement agencies, coupled with parallel immunity or amnesty programmes, enables the DoJ and the Canadian Competition Bureau to function most effectively, 䉴

22 Treaty between the Government of Canada and the Government of the United States of America on Mutual Legal Assistance in Criminal Matters, in force from 24 June 1990, full text available at http://www.oas.org/Juridico/mla/en/traites/en_traites-mla-can-usa2.html.


12/10/07

12:16

Page 37

Panel Discussion

37

independently, but in a truly coordinated manner. It is those two ingredients, i.e. the fact that both US and Canadian proceedings are criminal and the existence of a treaty allowing for the sharing of information, that puts the US and Canadian antitrust enforcement agencies in a better position. It is also necessary, since so many cases cross the border between Canada and the US, and often to other parts of the world, including Europe, that defence counsel engage in similar coordination. In my written contribution I underlined the critical importance of defence counsel coordinating to the same extent as the enforcement agencies, especially in cases where you are approached by a possible amnesty applicant, and you must move quickly and effectively across many jurisdictions, and then continue to ensure that what you do in one jurisdiction does not undermine the defence in another. Last point: the ability to work effectively in a cooperative and coordinated manner, both at the government level and in terms of defence, requires that the laws converge both in terms of process and substance. If the process or the substance is significantly different, amnesty programmes cannot work together. For example, if one jurisdiction makes it clear that as you get amnesty you cannot apply for leniency in another jurisdiction, this would distort the process tremendously across jurisdictions. My written contribution touches upon a very important subject in cartel investigations today, namely computer searches. In our technology era, for example, the Canadian Competition Bureau has a forensic electronic department with the ability to download instantly all kinds of information from computers anywhere. But do they have the right to enter an office in Toronto or Vancouver and download information from a computer in an office located in another jurisdiction? The Canadian Bureau’s position is that they have the right to do so, because the Canadian Competition Act permits them to access anything available. The US has taken a more cautious position on this, occasionally stating that it will not knowingly download information that it knows is not located in their jurisdiction. And what happens if the people in the offices located in another jurisdiction decide to pull the plug on the computers in Toronto? Is that obstructing justice? To conclude, in my written contribution I also commented upon recent initiatives at the OECD on information exchange. I commend the OECD for trying to develop a template for information exchange, building on the OECD recommendation of some years ago on the effective treatment of international cartels. In my written contribution I pointed out that the template has accomplished a good normative standard in several areas, from privilege protection to confidentiality protection. This kind of framework ought to become the foundation for more bilateral agreements. But there is the issue of notice, where the position of various representatives of the ABA and IBA, to some extent, runs to the other side of the table from that of the governments. The position of private practitioners is that notice should be given to those whose information is sent across borders, unless it jeopardizes the investigation. The


38

12/10/07

12:16

Page 38


position of the governments and the OECD is that such notice need not be given if the other safeguards are provided in the bilateral agreements. It remains to be seen how the OECD template works in practice. IAN FORRESTER—I will begin with a personal anecdote: last year, coincidentally, through a family member, I was contacted by the victim of a cartel. In the late 1980s this man ran a quite successful business, wholesaling consumer products to shops and end users. One day he received a phone call from an employee of one of his suppliers, saying that several suppliers had been discussing the problems presented to them by the man’s activities and he would need to change his behaviour; otherwise there would be unpleasant consequences. The man’s company could not change its behaviour, for a variety of reasons, and in 4 months’ time it had been eliminated from the market. It sold the stock, and the two owners of the company went into other trades. In 1989, this man wrote to DG IV and described the story in a long and detailed letter, and in a few months he received a negative response, saying: “Sorry, we have other priorities, we cannot help.” Fifteen years later, in 2004, the Commission announced it was investigating and fining the very companies about whom he had complained in 1989. The person is now in correspondence with one of those companies, endeavouring to get an indemnification from it for having ruined his business 15 years ago. He obviously encounters great difficulties, because of the obvious issues of proof of causality for the damages incurred, etc. I therefore have every reason to be sympathetic to the proposition that we should facilitate indemnification for those who were victims of cartels. However, I am able to recognize that there are considerable difficulties in establishing that as a principle. It is not evident how we could distinguish the rights of cartel victims as being separate from and superior to the rights of other victims of personal injury, personal abuse and other kinds of sufferings. Each of those face considerable difficulties in finding compensation. I am not sure how we could reduce obstacles for one category of victims and leave the others as they stand. Having said this, I do have some suggestions which might assist in that direction—and which I will pass on “free of charge” to the European Commission . . . I have a number of ideas which seem to me worth pursuing, and which are not inconsistent with European legal traditions. First—and I do not think anyone has said it so far, but let me be the first to state the obvious—the nature of cartels is such that proof of their existence has to be obtained from the inside. That being so, we could give consideration in Europe to the Korean practice of paying rewards to those who denounce cartels. Rewards in Korea can be high, up to 850,000 Euros, and the amount depends on the quality of the evidence and the value of the goods affected by the cartel. I think Korea is the first to have such an elaborated system, in place since 2–3 years ago—a very intriguing idea, and of course it is self-financing, and if well administered, it could become a source of huge profit. 䉴


12/10/07

12:16

Page 39

Panel II: Investigation Powers of the Competition Authorities

39

A second path, which the Commission has pursued in Organic Peroxides,23 is to penalize the facilitator. Law firms are of course accustomed to sitting with clients during sensitive meetings with competitors and clients. This is part of the lawyers’ job, to advise clients on when the legal limits of discussion were reached and it can go no further. I do not think it would be good to prosecute lawyers on the basis of the kind of evidence available against them. Rather, you might want to ask hotels to request that those who book conference facilities for industry associations might certify the legality of the meetings. My suggestion for dealing with the problem of cartel victims is to include in the text of decisions concerning cartels a promise by the infringing party to settle claims through arbitration. In arbitration, many of the genuine difficulties of proof of damages and causality would be eased, and the result might be more just. I recommend to the Commission not to pursue the kind of retaliation associated with the industry of treble damages in the US. That is, in my view, not justice, but a kind of privateering (privateers were licensed pirates in the 18th and 19th centuries, who were empowered by the crown to rob under a flag). Like many Europeans, I have some problems with the aesthetics of treble damage practice. One of the very intriguing things that the European Commission is doing at the moment is exploring the possibility of admitting oral confessions by those seeking leniency—confessions which are not, I believe, admissible in US proceedings. And a couple of extra thoughts: first, we have now a system whereby the target company has a duty to deliver evidence. I suspect that most companies are willing and accustomed to telling the truth, and less likely to conceal evidence than investigators probably believe. I suspect that the dramas associated with dawn raids probably are not really necessary—although it is a wonderful technique for lawyers to get the attention of their clients. To be straightforward, I think that companies do not have an inclination to shred documents or to lie when confronted with precise questions. This leads me to a further observation, a point which James Venit raised earlier. Nowadays, companies—especially the large ones listed in the US— are under massive duties of disclosure, which have increased over the last 5 years: environment, sexual harassment, treatment of employees, corrupt practices, taxation, all kinds of topics are subject to heavy disclosure obligations. For example, the penalties applicable for not disclosing the fact of not having in place good accounting mechanisms can be savage. I think that competition enforcers have not exploited sufficiently this possible line of action. Supposing you were to identify some potentially recidivist industries—as identified by the economists this morning, where you notice low R&D, stable products, etc.—and invite them to certify their compliance with competition 23

Commission decision of 10 December 2003, OJ L 110 [2005].


40

12/10/07

12:16

Page 40


rules. I am not sure how people would react to an inquiry coming out of the blue and saying “please describe to us the mechanisms in place for detecting and discouraging infringements of competition rules, please state . . .” etc. I am not sure how helpfully they would answer, but I am sure they would give it very close attention. In my experience, companies do not generally tell lies, they give their understanding of the truth. The documents under the cherry tree referred to in the title of my paper were taken home by a German employee of a worldwide company who thought it would be quite wrong to destroy documents, and therefore took them home and planted them under the cherry tree in his garden. As far as the legend tells us, they are still there . . . I think you may be overestimating the effectiveness of force and the likelihood of lying in enforcement matters. To conclude, a note of caution: there is no doubt that criminal sanctions are one of the most important features making the US a leader in successful cartel enforcement. But the more we rely on penalizing individuals in Europe, the more doubts will arise as to whether this is compatible with an administrative law regime where the companies are under a duty to cooperate with the investigation. That means we may have to choose between the desirability of terminating the undesired activity, and punishing it. It may be easier in Europe to terminate than punish. OLIVIER GUERSENT—I would like to comment on a few points. First, facilitating practices, which I did not treat in my written contribution. I think one should give a definition for this. I would submit that the concept comprises practices which bring about an infringement of competition law or conceal such infringement, but which are not constitutive elements of the infringement itself. From the standpoint of the European Commission, it all depends on who facilitates the practice: if it is one of the cartel members, then it is rather easy—we have had that situation a number of times in the past, when one of the cartel members is doing hotel reservations and arranging for the meetings, etc. Sometimes it goes as far as qualifying this member as the cartel leader, therefore triggering an aggravating circumstance—this is what happened in the Graphite Electrodes case,24 where clearly Mr Koehler was providing the facilitating devices, a situation that led us to qualify the firm as leader of the cartel. This role can also be played by an association of undertakings— another classic situation. The problem is when the facilitator is not a market participant. This is the case that Ian Forrester referred to: there are companies in Europe, mainly in Switzerland, whose raison d’être is to facilitate cartels. There were situations where the companies kept on their premises all the documents related to their 䉴

24

OJ L 100 [2002].


12/10/07

12:16

Page 41


41

agreements. One of the reasons why we manage to get these documents is because under the leniency programme the applicants withdraw the documents from the company’s safes and bring them to the Commission. The Decision [against the non-market participant] was of course challenged in court. The Commission had imposed a symbolic fine, which I strongly opposed at the time, because we had had this same case before, in the 1980s, and second-time fines should have been higher. My second point is on the “administrative versus civil system” issue. The purpose of my written contribution is to demonstrate that the administrative enforcement system that we have in place at the EU level is legitimate and efficient, but that it could become even more efficient through a number of measures that we can take internally—for example, by improving leniency programmes and better internal organization, which can certainly be complemented by other tools. First, the existing instruments can be improved: I would certainly agree with the argument in Julian Joshua’s written contribution, in the sense that leniency standards should be clarified. Then Julian Joshua and I may disagree on what clarification means. Personally, I would argue for toughening the standards, because, if anything, as the statistics in Commissioner Kroes’ speech also showed, quite a number of leniency applications have been accepted but eventually they did not lead to anything. I would personally drop Article 8(a) in the Leniency Notice, and maintain only Article 8(b). Second, I would like to make a few remarks on new instruments that can be developed, especially on direct settlement. The Directorate in DG Competition that deals with cartel enforcement handles about 40 to 50 cases at a time. Even if a case were to take no longer than 30 months, this means that we cannot expect it to deliver more than about 10 decisions per year. So, we believe that the way to solve this problem is to introduce a system of direct settlement. The next issue is: what should this direct settlement system be like? Should we, or should we not, change the legislative framework? If we change it, we can process more cases, but the change may take 5–6 years. So, I would rather submit that we have to find a solution that does not require a change of the legislative framework. Under the current system, you cannot impose a fine without a Statement of Objections. Next, the direct settlement system should be shaped so as to allow the Commission to save time and resources. We should have negotiations before the Statement of Objections, and then we could have a short Statement of Objections—which is in the interest of everybody, both the Commission and the companies, especially in the context of exposure to civil damages actions in the US. I very much hope that we will be able to come up with a preliminary design of a direct settlement system, after discussion within the ECN, by the end of 2006. Last remarks, on the issue of complementary tools: I have very little doubt that criminalization of EC proceedings is not something that I will see in my


42

12/10/07

12:16

Page 42


lifetime. But that does not mean we should stop considering it, because there is a trend towards criminalization in the Member States, and I do believe that, depending on the design of criminal regimes in the Member States and the way in which they interact with the administrative system at the EU level, the two tracks can either become a very powerful instrument in the hands of the ECN, or they can mutually destroy each other. Nobody in his right mind will come to Brussels to apply for immunity, to then go back home and be put in prison. So we have to find a way to solve this problem. There are a number of interesting tracks, both within the ECN, and in the recent ruling of the ECJ in the Council v. Commission case,25 in the area of the environment, where there are some interesting considerations as to the degree of interference that the First Pillar can have in the criminal law system of the Member States in order to ensure compliance with Community legislation. A very last remark, on international cooperation: I think it is quite clear that it is very difficult to cooperate if we have different systems. This created some difficulties in our cooperation with the US—we could not always do what we would have liked to do together. But still, we can achieve quite a lot, and I sincerely hope this difference will not be an impediment to a secondgeneration cooperation agreement. PHILIP COLLINS—If I could start with a few observations, coming from someone who has not only moved from Brussels to London, but also from the public to the private sector and back. On the point that John Ratliff raised, I think this is critical for the functioning of leniency in the European system. If you look at the motivations for applying for leniency, a lot of them arise spontaneously, as a result of things like businesses being bought and sold, and managements being changed. If you impose too heavy a burden on groups of companies which apply for leniency, or invite them to come out on a whole range of other things, that raises some practical problems for the leniency programme. Ian Forrester raised important issues for the legal profession and for enforcement agencies related to corporate governance. I believe that one of the most effective ways in which lawyers and agencies can encourage compliance is by engaging with management at the very senior level, and in particular in the case of public companies, not exactly directors. There is an interesting case in the UK, not in the competition field, but in the water industry, where a distinguished gentleman who is not exactly the chairman of the company has been forced to stand aside by the chairman of the UK industry association as a result of criminal investigations opened for giving misleading information to a regulator. My experience is that legal compliance activities carried out by law firms with their clients are really effective. I realize that

䉴

25

Case C-176/03, [2005] ECR I-7879.


12/10/07

12:16

Page 43


43

in the long term defending cartels is more remunerative for a lawyer than advocating compliance, but I do think that the legal profession should effectively engage with companies, and help encourage compliance. A few remarks on facilitating devices: I do not think that the public enforcer should spend a lot of time and resources on facilitating device cases, because the burden of proof is very high, and the risks of failure are high in terms of the evidential proof for the court. The UK experience is that, for addressing issues in the areas where there are apparent market failures, a detailed investigation is the way to go. In relation to administrative procedures, I would express our position in relation to settlements: this morning Paul Grout mentioned the private secondary schools cartel in the UK, and there is some interest in the settlement that we have achieved there. The whole case came about because two people turned on a school computer, found evidence of cartel activity, and then delivered it to the press. There were many distinguished personalities sitting on the boards of governors of those schools. There are more than 12,000 independent schools in the UK, and the evidence collected covered not more than about 100 of them, so the question was how to use it in a way that would produce an effective decision and outcome. If someone would have told me before John Fingleton’s arrival at the OFT that this case would be settled and finished in six months, I would not have believed him. But the fact of the matter is that the case was prone to settlement, because there was a clear interest on the part of the schools to put this situation behind them. There was also agreement that there should be some sort of modest contribution for the compensation of those who suffered damage. Do not forget that all these bodies were charities, not profit organizations. There are several conclusions that come out from this experience: in appropriate cases there must be some point where the opportunity for settlement arises, and in such cases some form of abbreviated Statement of Objections is necessary. Then in such cases you may also be overwhelmed by the amount of evidence coming at you, so you need to identify the core infringement and the facts—it may be near what you have to prove in a normal case—and you have to come to some form of settlement before you arrive at the decision. Then a key point in the settlement has to be the provision of some form of compensation to those who suffered loss. I certainly endorse Ian Forrester’s proposal that the compensation could be settled through some form of mediation; that is certainly an interesting suggestion. But we also have to be careful that, by going down the settlement route, we do not impede the deterrent effect or discourage leniency. It is also important that we make sure we do not take this road with cases that, at the end of the day, cannot be settled, because the parties are taking a very tough position related to evidence. That is, frankly, a waste of time. Two more observations: the UK introduced criminalization only shortly after having introduced the regime on the application of Articles 81 and 82


44

12/10/07

12:16

Page 44


EC. At the time I thought that was undoubtedly a complicating factor, but there have been relatively few cases where the OFT considered using criminal powers. The important point about this is the strong deterrent effect that criminal proceedings have. If you can pursue in parallel civil and criminal proceedings, then you have the option at some stage down the road to drop the criminal charges, but in the meantime it does help people to focus their minds on the issues. There was also the point on whether we should introduce individual rewards, as in the South Korean model: I do not think that it is realistic for the UK to follow that model, but there are certainly precedents in UK law of more modest amounts being offered for information, typically about £ 10,000, and this is something we are certainly considering at the moment in the context of the debate on initiatives for encouraging parties to come forward. In the written contribution I also mentioned some interesting civil and criminal investigative powers that the OFT enjoys—the latter we have not used yet, but they might increasingly be used in the future. One final point about the remarks made by Mario Siragusa and Ian Forrester, about companies being in danger of being driven out of business through cartel proceedings. I find it interesting to hear about small companies that have been driven out of business by fines, but interestingly, 2 or 3 years later they appear again on the market, perhaps under different ownership. THOMAS BARNETT—I want to talk first about what I call “seven steps to better cartel enforcement”; second, about a few developments in the US that reflect these concepts; and third, I’ll conclude from a more international perspective. The number one step, in our perspective, is to focus prosecutors on “hard core” collusive activity, and not get engaged in matters that are not naked price fixing or customer allocation. Focusing on these cases makes proof easier, makes it easier to develop a culture that this type of conduct is wrong, makes it easier to obtain substantial sanctions once you have proved the violation, and gives your prosecution team a more focused target. Second, it is important to treat cartels as serious crimes, not only in terms of deterrence, but also in terms of compliance and developing a culture in the business community. Third, provide amnesty programmes. This continues to be our most effective tool not only for detecting cartels, but also for building a case for prosecuting them. It is very often the case that, once we begin to investigate a particular market for cartel activity, a lot of applicants start coming in, and we get to prioritize leniency applications not only in the order of their importance, but also—and this gets into Olivier Guersent’s point—we start from one application to gradually build a stronger and stronger case. In the US, it is very difficult to take a case before the jury and obtain a conviction unless you have someone who can testify he was in the room, there was no ambiguity about what they were doing, they were fixing prices. 䉴


12/10/07

12:16

Page 45


45

Fourth, vigorously prosecute obstruction of justice. It is particularly important to protect the integrity of your investigation. Sometimes we end up with companies who are prosecuted just for this, and not for the cartel activity, which we have a more difficult time in proving, but we place that as a high priority. Fifth, charge cartels in conjunction with other offences. Obstruction of justice is an example of that: sometimes we may have proof of false statements to procurement agencies without actual proof of cartels, and it is important to be aware of the other criminal charges that can be brought to bear. Sixth, provide transparency and predictability. We talk a lot about that in terms of the effectiveness of an amnesty programme, but it is also important on a broader scale, to give a clear signal to the business community about where the dividing line is, what is a cartel and what is not, and to give them an incentive to comply with the law. Seventh, publicize these enforcement efforts—that is very much selfexplanatory. Turning to some of the steps that we have taken in the US that reflect these principles, I underline that price fixing, bid rigging, the allocation of markets and output is what we focus on. We make it our number one priority, and we publicize that fact—we reflect that in having a separate criminal enforcement group since 1994—and I think in that respect that Neelie Kroes has made an excellent move in creating a cartel directorate at DG COMP. We have a welldefined focus for our criminal enforcement efforts: we have clear boundaries for private firms; they know when they are crossing the line; and this helps us when we go to court in the sense of meeting our burden of proof. It is essentially about whether they did it or not, and not about the competitive effects of the activity that has taken place. We have said for years now that cartel activity is the equivalent of fraud—it is stealing—and we have been conveying that to courts, so that courts are more willing to impose sanctions when a violation is found. In 2004, we had the antitrust criminal penalty enhancement reform, which increases the maximum prison terms available under the Sherman Act for individuals to 10 years, and increases the maximum criminal fine to $ 100 million. To avoid confusion, if we can prove ‘double the amount gained, double the amount lost’ we can get a higher fine, but we do not have to prove that in order to get a fine up to $ 100 million. This is a reflection of the fact that the US Congress takes seriously our statements that cartels are serious fraud activities. The US Sentencing Guidelines were revised last year26 to reflect this. I will note that, while our Supreme Court has made these Guidelines advisory, they must nevertheless be considered by the courts, and indeed we have not seen any significant deviations from the Guidelines in the sentences that have been imposed. This again underlines the results of our efforts to make judges think of cartels as fraudulent activities. 26

See http://www.ussc.gov/2005guid/2005cong.pdf.


46

12/10/07

12:16

Page 46


Finally, the US Congress this year [2006] gave us wire-tap authority in criminal antitrust investigations. The extent to which we will use that remains to be seen, but it is another serious signal to the business community about how cartels are perceived. To give you a numerical perspective, we have continued to have success in our criminal enforcement efforts. Up to this year we have collected up to more than $ 500 million in fines, and the trend in incarceration—expressed as average months of imprisonment—has risen significantly since 1999. The lowering trend in 2006 is due to the fact that this year we had a number of foreign citizens (South Koreans) who have agreed to serve jail time in the US, and the foreign nationals have received lower sanctions than US nationals. You can also see a general upward trend in the total number of months of imprisonment. Regarding an international perspective, from our standpoint, one of the most encouraging developments has been the international expansion of antitrust enforcement regimes and of criminal antitrust enforcement regimes in other jurisdictions. There are now 16 countries with criminal antitrust enforcement system, and 35 countries with some form of amnesty programme. This provides a number of benefits for the US: it helps expand within the international business community the culture that cartel activity is wrong, fraudulent, and should not be engaged in. It thereby helps increase deterrence, and it helps our ability to gather evidence: instead of active opposition, we get active cooperation. The expansion of leniency programmes can help precipitate our leniency applications. Indeed, when I gave the list of what precipitates applications, I probably should have included the expansion of leniency programmes abroad, because when somebody reports a cartel that may be focused in Japan or in Europe, they are likely to report it in the US as well. It also helps us in terms of getting access to individuals. This is one of the reasons why we had an increasing number of foreign nationals agree to serve time in US prisons, and it helped us so that we have the option of seeking forced extradition, as we are seeking with Mr Norris in the UK. Finally, we are getting benefits in terms of sharing information and best practices with other competition authorities. That makes our efforts more effective and it makes the world less safe for antitrust offenders. DENIS WAELBROECK—Two brief comments on leniency. First, my feeling is that there is over-reliance on leniency on the part of competition authorities. There are also cultural differences between jurisdictions in the way we perceive leniency programmes. My perception is that when you talk to AngloSaxon clients, it is easier to make them accept leniency systems, but it is more difficult to convince continental clients. There is a sort of “code of ethics” in continental Europe, covering “things that one should not do”. Therefore I think one should be somewhat careful before over-relying on leniency programmes as a method of discovering cartels. People are happy to go into leniency once a cartel has been discovered.

䉴


12/10/07

12:16

Page 47


47

Second, Thomas Barnett said that leniency is important in order to establish the facts. My problem is: what is the evidential value of the statement made by a person who has an interest in having the other parties be prosecuted and fined? At least in Europe, I have the feeling that we are too keen on accepting the story as it has been told by an interested party. 䉴 EDITORIAL NOTE—Due to technical problems, the last part of this recording is unfortunately not available.


12/10/07

12:16

Page 48

(D) Ehlermann 07 Session A Pt 2 Harrington

12/10/07

12:16

Page 49

SESSION A


2 PANEL I: OLIGOPOLY THEORY AND ECONOMIC EVIDENCE

WORKING PAPERS


12/10/07

12:16

Page 50


12/10/07

12:16

Page 51

I Joseph E. Harrington, Jr.1 Behavioral Screening and the Detection of Cartels*

A. Introduction Effectively fighting cartels requires that cartels be discovered, discovered cartels be successfully prosecuted, and successfully prosecuted cartels be penalized. Operating effectively in all three stages—detection, prosecution, and penalization—is crucial to disrupting existing cartels and deterring new ones from forming. Let us consider the role of economic and econometric analysis in each of these three stages.2 Historically, the biggest role of economic analysis lies in the penalization stage. To determine the damage inflicted by a cartel, one must estimate what the price would have been had collusion not occurred. Economic analysis has been essential in estimating the but for price and using it to measure damages. This is directly relevant to penalization in the US—as customer damages are significant—and is (or ought to be) relevant in determining the appropriate government fine, especially if fines are intended to deter cartel formation.3 Economic analysis plays a significantly weaker role in the prosecution stage.4 In the US, economic evidence is generally insufficient to prove a violation of Section 1 of the Sherman Act. There needs to be evidence of explicit coordination (or, as Werden (2004) prefers, “of a spoken agreement”) among the suspected cartel members. Economic analysis can play a supportive role, however. For example, in the case of the Ohio school milk case, economic analysis showed how prices were lower when (transportation) costs were higher which is inconsistent with competition but quite consistent with a particular collusive scheme (Porter and Zona, 1999). Nevertheless, since 1 Department of Economics, Johns Hopkins University, Baltimore, MD 21218 410-5167615,-7600 (Fax), * This paper was written for the 11th EU Competition Law and Policy Workshop (Florence, June 2–3, 2006). It draws on Harrington (2006a,b). This research is supported by the National Science Foundation under grants SES-0209486 and SES-0516943. e-mail: [email protected], site: www.econ.jhu.edu/People/Harrington. 2 From hereon, when I refer to “economic analysis”, I will be referring to the use of either economic theory and/or econometric methods. 3 For a recent review of economic methods for calculating damages, see Clark, Hughes, and Wirth (2004). 4 For a review of this issue for the US, see Werden (2004).


52

12/10/07

12:16

Page 52


economic theory generally doesn’t distinguish between tacit and explicit collusion—and it is explicit collusion that is illegal in most countries—it is difficult for economic analysis to deliver definitive evidence that observed behavior is consistent only with the presence of a hard-core cartel. And finally we come to the issue of the detection of collusion. It is here that economic analysis has been least significant. Cartels are discovered in many ways but, with some rare exceptions, economic analysis is not one of them. In this paper, I intend to argue that economic analysis can play a more significant role in the detection of cartels, with the intent of spurring policy discussions and scholarly research. Screening refers to a process whereby industries are identified for which the existence of a cartel is likely. An industry that is picked up by a screen is one that warrants not prosecution but rather a more intense investigation which directly contrasts collusion and competition as competing explanations of market behavior.5 Screening is then the first phase of a multi-stage process which may or may not end with prosecution. Before moving on, I do want to admit that screening for cartels may be difficult, given our current methods. But my intent is largely to promote research into methods of screening by economists in academia, think tanks, consulting firms, and government. That there is some reason to hope that this process may ultimately result in a workable and effective screening program is suggested by success in analogous programs for other forms of illegal activity. Screening programs are in place to identify credit card fraud, insider trading, and tax evasion. Admittedly, those applications are rich in data but then there are some industries in which data is available. We should not forsake the possibility that screening for cartels can achieve some of the success that has occurred in screening for other illegal activities.

B. Screening approaches Broadly speaking, methods of cartel detection using economic analysis can be partitioned into those that are structural and those that are behavioral. A structural approach identifies markets with traits conducive to cartel formation. For example, it has been shown that collusion is more likely with fewer firms, more homogeneous products, and more stable demand.6 Grout and Sonderegger (2005, p. 15) is representative of this approach:

5 For a review of some of the methods for comparing competition and collusion as alternative explanations of the data, see Harrington (2006a). I would also recommend reading Porter (2005). 6 For some industry traits associated with collusion, see Symeonedis (2003), Motta (2004), and Grout and Sonderegger (2005).


12/10/07

12:16

Page 53

Behavioural Screening and the Detection of Cartels 53 “[. . .] the fundamental background reduces to three core issues—product, volatility, and company criteria. The first core question is whether the industry has a homogeneous product or not. Cartels are far more likely if the product is fairly homogeneous between companies in the market. . . . Second, does the industry display volatile turnover over a sustained period of time? Cartels are more likely if output and market conditions are normally stable. [. . .] Finally, are the leading players in the market large and relatively constant? If there are significant changes in the market shares or regular exits and entrants then cartels are less likely.”

One could imagine investigating industries that score high on these relevant traits with the hope of finding evidence of a cartel. I have been told that the Dutch competition authority has deployed structural methods with some success including, for example, uncovering collusion in the shrimp industry.7 In spite of that success, I am pessimistic as to the efficacy of structural methods. Their weakness lies in that, in most economies, there is a high chance of false positives—the indicators suggest collusion is likely but in fact there is no cartel. To see why this could be the case, imagine the “ideal” market for collusion: two firms, homogeneous products, stable demand, no large buyers, excess capacity, and so forth. Even though such a market would surely be flagged by a structural investigative tool, my own prior belief is that a very high fraction of those markets are not cartelized. Based on what we know (which, admittedly, is only discovered cartels), the frequency of collusion in most economies is rather low. Hence, given a low prior probability of collusion, the posterior probability—conditional on all those structural variables taking values conducive to collusion—is still probably quite low. At the heart of this problem are omitted variables. There are multiple equilibria— some involving collusion, some not—and non-observed variables can influence whether firms settle upon a collusive equilibrium. A structural approach is based on data about an industry which makes it more likely that a cartel will form. This is to be contrasted with a behavioral approach which uses data that is itself evidence that a cartel has formed. A behavioral approach focuses on the market impact of that coordination; suspicions may emanate from the pattern of firms’ prices or quantities or some other aspect of market behavior. Buyers could become suspicious because of a parallel movement in prices or an inexplicable increase in prices. A sales representative for a colluding firm may become suspicious because she is instructed not to bid for the business of certain potential customers (as part of a customer allocation scheme) or not to offer reasonable price concessions when business might be lost to competitors. For example, the European Commission investigated the stainless steel industry (and found collusion) because buyers complained to the Commission about a sharp increase in

7 This information is based on private communication with Peter A. G. van Bergeijk of the Nederlandse Mededingingsautoriteit (December 16, 2004.)


54

12/10/07

12:16

Page 54


prices.8 Though we do not know the specifics of their complaints, reportedly the cartels in graphite electrodes9 and thermal facsimile paper10 were also begun because of buyer complaints. The focus of this paper is on a behavioral approach to screening. In trying to detect a cartel, it is useful to think about two phases. To illustrate them, the price path for the citric acid cartel is reproduced from Connor (1998). The path labeled “contract” is a more accurate reporting of actual transaction prices. As one can see, the cartel significantly though gradually raised price starting in mid 1991 until late 1992 at which point price became somewhat flat over time until it started to fall with the cartel’s collapse. The only point I want to make here is that one could imagine screening methods designed to pick up the transition from non-collusion to collusion— looking for a radical change in firm behavior—or the stationary collusive phase—finding differences in behavior from when firms compete. Finding evidence of collusion is apt to be easiest during the transition and the method to do so is to find evidence of a structural break in the data-generating process, whether it is a firm’s price, quantity, allocation of market shares, or some other relevant variables. With either approach, it is important to know what to look for; an issue I turn to in the next section.

List

85 80 75 70

•

•

• •

• • •

•

• 85 60 55

••

• • •• •• • •• • • •• • • •• Contract

••

•• • • Conspiracy

50 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Figure 1: Citric Acid Cartel Price Path (Connor, 1998) 8 “The industry was investigated for cartel activity after buyers complained to the European Commission about the rapid increase in prices.” Graham Hind, “English Cutlers `Hit by Cartel’,” The Times (London), August 21, 1994; cited in Levenstein, Suslow, and Oswald (2004). 9 Ferromin International Trade Corporation, et. al. vs. UCAR, et. al., In the United States District Court for the Eastern District of Pennsylvania, Second amended complaint, filed May 1, 1999, at paragraph 50. 10 John Clifford and Bill Rowley, “Tackling Cross-Border Conspiracy,” International Corporate Law, May 1995.


12/10/07

12:16

Page 55

Behavioural Screening and the Detection of Cartels 55

C. Collusive markers To provide economic evidence of collusion, one needs to know what behavioral patterns are indicative of collusion. An important line of work is that which provides markers that serve to distinguish collusion from competition. In this section, I will review some of what theory and actual cartels have to offer. The discussion will focus on patterns in prices and quantities because a screening approach should be based on easily available data. While the theories might generate predictions consistent with some of the noted empirical evidence on cartels, I do not claim that those theories explain that evidence, nor that my discussion is comprehensive. My objective is simply to catalog a few tentative collusive markers based on theory and evidence. Before embarking on this review, an important disclaimer is that evidence supporting collusion need not imply evidence against competition. When we find evidence of collusion, there is always the possibility that there actually is no collusion and the problem is we’ve misspecified the non-collusive model. Similarly, failure to find evidence of collusion may be due to misspecifying the collusive model; for example, we’ve focused on the wrong collusive equilibrium. At best, collusive markers can serve to screen industries to determine whether they are worthy of more intense investigation.

1. Theory of collusion The basic logic whereby collusion is sustainable as an equilibrium in a repeated game model is predicated upon rewards and punishments. A Nash equilibrium for the static game is one in which each firm’s behavior (which is typically a price, a bid, or a quantity) is optimal given the (correctly anticipated) behavior of other firms. Firms collude for the purpose of raising price above the static Nash equilibrium (or non-collusive) level so as to yield higher profits. This necessarily means that a firm’s behavior doesn’t maximize current profit; a firm’s collusive quantity is below that which maximizes current profit or its collusive price exceeds that which maximizes current profit. As cheating on the collusive outcome raises current profit, firms can only be deterred from doing so if they experience some punishment. This punishment is typically in the form of future foregone profits due to an intensification of competition. More specifically, if firms set the collusive price then they continue colluding but if a firm cheats then firms revert to some low-profit punishment path. This may mean going to the static Nash equilibrium for some length of time or an outcome with even lower profits (perhaps pricing below cost) or an asymmetric equilibrium which is particularly detrimental to the firm that deviated (perhaps requiring that the firm produce very little).


56

12/10/07

12:16

Page 56


It follows from this argument that a firm which considers deviating from a collusive outcome realizes it would raise current profit but lower its future profit stream. The equilibrium condition requires that the foregone future profit stream is at least as great as the gain in current profit from deviating. When the punishment is reversion to the non-collusive outcome for periods, the condition is of the form: c nc 2 c nc … T c nc d c, where c, nc, and d is the collusive profit, non-collusive profit, and the (maximal) profit from deviating, respectively. (0, 1) is the common discount factor across firms. In this simple case, the model is stationary and the solution is symmetric. Deviation yields higher current profit of d c but lower future profit of c nc over the next T periods (with firms returning to the collusive outcome thereafter). Equilibrium requires this condition to hold so that abiding by the collusive agreement is optimal for all firms. Since c nc, this will hold when T is sufficiently high and is sufficiently close to one so that firms sufficiently value future profits.

2. Price markers A defining feature of collusion is obviously higher prices and, therefore, the most obvious marker is price increases. However, it is not necessarily a very effective one since price can rise for many reasons unrelated to collusion. What a number of European Commission cases suggest is that price increases occur simultaneously with more uniformity in price across customers.11 Cartels often reduce the number of distinct prices charged to customers. For example, it is common to move from a regime characterized by a list price and a series of ad hoc negotiated discounts to either a non-negotiable price list (such as with the fine arts auction houses cartel) or an agreed-upon set of discounts to the list price (as with the citric acid cartel). This increased uniformity in price across customers could have been done for several reasons. It might have been done for simplicity; remember that collusion requires firms to discuss and agree to every price and this is a recurring situation. A second reason is that fewer prices may have made monitoring easier. With more customer-specific prices, it may become difficult to determine whether the price charged was appropriate given the market or customer type. While in a competitive environment one could imagine a firm moving to an “everyday low price” by eliminating discounts, this seems to be done along with reducing, not raising, the list or regular price. This makes the combination of

11

Unless otherwise noted, references to actual cartels are based on Harrington (2006c).


12/10/07

12:16

Page 57

Behavioural Screening and the Detection of Cartels 57 fewer discounts off of a list price along with a higher list price suggestive that firms are coordinating their behavior. Collusive Marker: A higher list (or regular) price and reduced variation in prices across customers. It has also been documented that cartel formation is often preceded by large price declines which, in some cases, were due to weakening demand and/or excess capacity. Notable examples include carbonless paper, citric acid, commercial explosives, fine arts auction houses, graphite electrodes, plasterboard, seamless steel tubes, soda ash, and vitamins. Unless the market is characterized by some cyclical factors such as seasonal demand or supply movements, it would be unlikely for competition to result in significant price declines and then a steadily rising price over the span of a few years. Collusive Marker: A series of steady price increases is preceded by steep price declines. A common feature to implementing a collusive allocation is the “homemarket principle” whereby cartel members reduce supply in each other’s home markets. The objective is to move to an arrangement where, ideally, each cartel member takes control of their home market and then share the global demand that was not part of any cartel member’s home market. The home-market principle was associated with collusion in the markets for chorine chloride, copper plumbing tubes, isostatic graphite, lysine, methionine, and seamless steel tubes. In a competitive market, one would expect a rise in a firm’s price to result in more imports, ceteris paribus. However, an allocation scheme based on the home-market principle would result in the suspicious combination of a higher price and fewer imports. Collusive Marker: Price rises and imports decline. The preceding markers pertain to movement in a firm’s price. There are also markers that focus on the relationship among firms’ prices. That parallel pricing behavior is a possible feature of collusion has long been noted. One problem with it as a marker is that non-colluding firms’ prices will also tend to move together if they are responding to common demand shocks or highly correlated cost shocks (such as a change in a common input price). Still, it is possible that there is a stronger correlation in colluding firms’ prices. It has also been pointed out that some bid rigging cartels have engaged in identical or near-identical bidding. This leads to a second problem which is that it would appear easy for colluding firms to avoid the appearance of collusion by simply having the members of a bidding ring scale up competitive bids and thus make the relationship between their bids consistent with competition (and firms having different valuations or costs.). This would suggest that parallel pricing behavior may not be effective as a marker, at least in the auction setting.


58

12/10/07

12:16

Page 58


Interestingly, it does not appear that bidding rings use such an easy detection-avoidance procedure (Bajari and Summers, 2002). One possible reason is due to constraints imposed by maintaining cartel stability. This possibility has been shown in Marshall and Marx (2004). The setting is a first-price sealed bid auction in which no ex post information is provided regarding the identity of the winning bidder and their bid (an assumption, however, that is not typically applicable). It is also assumed that the cartel is less than allinclusive—some bidders are not members of the cartel. Suppose the cartel meets beforehand to decide on what each bidder should bid (and who should be the high bidder). The situation is one of imperfect monitoring; the cartel member who was designated to submit the highest bid cannot distinguish failing to win the item because a non-cartel member outbid it or another cartel member cheated and outbid it. In order to ensure that cartel members want to bid as prescribed, two cartel members’ bids may need to be clustered. Using the (truthful) reports of their valuations at the pre-auction meeting, the cartel selects the bidder with the highest report—let us refer to him as the cartel representative (at the auction)—to bid at a certain level with all other cartel members told to bid less. Absent concerns about cheating, the cartel representative would bid optimally in light of its valuation and the distribution on non-cartel members’ bids. The problem, however, is that if it bids too low, another cartel member may cheat by outbidding the cartel representative’s bid. To destroy that incentive to cheat, the cartel representative sets a higher bid so that the other cartel members don’t want to outbid it and are content to set a lesser bid. But now suppose these other cartel members all set very low bids. The problem that emerges is that the cartel representative would have an incentive to bid lower since the only reason to bid so high was to discourage cheating. In order to keep that from happening, one of the other cartel members must set a bid a little below the cartel representative’s bid; it’ll keep the cartel representative from cheating without affecting whether the cartel representative wins. That bids are clustered in this way is unique to when bidders collude. Collusive Marker: Firms’ prices are strongly positively correlated. Going beyond the auction setting, it is indeed a common feature of cartels that firms charge the same price. As mentioned, this could be quite consistent with competition. What makes one more suspicious that there is collusion is when uniformity extends to many other dimensions such as the prices of ancillary services. For example, in the industrial and medical gases cartel, the cartel not only set a common price for gas but also a common transport charge and a common rate for renting a gas cylinder. Failure to agree on any of these other charges could have undermined the cartel as firms could then price gas at the agreed-upon level but cheat by offering, say, free transport. Controlling other dimensions upon which firms can compete was also strikingly clear in the fine arts auction houses cartel. Christie’s and Sotheby’s


12/10/07

12:16

Page 59

Behavioural Screening and the Detection of Cartels 59 didn’t just agree to identical commission schedules but also with respect to at least six other contractual features. Collusive Marker: A high degree of uniformity across firms in product price and other dimensions including the prices for ancillary services. Properties relating to the variability of a firm’s price may also indicate collusion. Two papers taking very different approaches show that prices can be less responsive to cost when firms collude. Athey, Bagwell, and Sanchirico (2004) argue that fixed prices may be part of an optimal collusive scheme. Consider a setting in which firms choose price and each firm’s cost varies over time and across firms and is private information. In each period, colluding firms exchange messages about their costs and then choose price. In characterizing an optimal collusive mechanism, there is a tension between efficiency and the amount of collusion. Given firms have homogeneous products, the unconstrained joint profit-maximizing scheme is to have the firm with the lowest cost produce all output at its monopoly price. The problem is inducing firms to truthfully reveal their cost since a firm with high cost may want to signal it has low cost in order to be able to have a positive sales quota. When firms are sufficiently patient, the incentive to mislead is sufficiently great that the best collusive scheme is to have price and (equal) market shares fixed over time and thus messages between firms are made irrelevant. When firms are moderately patient, there is partially rigid pricing so the price function is a step function of cost in which case price is often unchanged but then experiences a large change. This also serves to distinguish collusion from competition. The argument for stable prices in Harrington and Chen (2006) comes from the cartel wanting to avoid buyers detecting collusion because of suspicious price changes.12 It is not assumed that buyers know how a cartel prices, nor are consciously looking for collusion, but rather that buyers become suspicious when the observed price series is sufficiently anomalous or inexplicable when compared to the history of prices. More specifically, buyers have beliefs on the current price change based on the history of price changes. When the realized price change is sufficiently unlikely in light of their beliefs, detection is said to occur. The cartel is aware of how its price path affects beliefs and thereby the probability of detection. Upon cartel formation, firms inherit the non-collusive price and buyers’ beliefs which are predicated upon price changes when firms were not colluding. In a sense, detection occurs when buyers pick up the “break” in the pricing function associated with cartel formation. Ideally, a cartel would like to raise price fast and have it adjust quickly to cost shocks but it must temper any such price movements in light of the prospect of detection. The optimal cartel price path is found to have a transition phase—in which price rises largely irrespective of cost—and a stationary phase—in which price 12

In most price-fixing cases, these are industrial buyers.


60

12/10/07

12:16

Page 60


70 Collusive Non-collusive

65

pt

60

55

50

45

40

0

20

40

60 80 100 Time Period Figure 2: Simulated Cartel Price Path (Harrington and Chen, 2006)

120

is responsive to cost. A simulated price path is shown in Figure 2 where the dashed path is the non-collusive price path, which moves around because (constant) marginal cost follows a random walk (that is, unit cost in period t equals unit cost in period t-1 plus a random shock that is independently and identically distributed). Starting with cartel formation in period 40, the solid line is the cartel price path. (The non-collusive price path is provided after period 40 for purposes of comparison.) While price is sensitive to cost in the stationary phase, note that it is much less variable than the non-collusive price. Intuitively, though the cartel might want to raise price a lot in response to a series of large positive cost shocks, such a price series may be perceived as unlikely by buyers and thus induce suspicions. Therefore, the cartel doesn’t respond commensurately to large cost shocks. Relative to non-collusive pricing, the impact of cost shocks on price is muted and takes a longer time to pass through. Thus, the variance of price is lower with collusion. There is some limited evidence showing that price variance is indeed lower under collusion. Examining collusion at auctions of frozen perch, AbrantesMetz, Froeb, Geweke, and Taylor (2005) find that the price variance during collusion is indeed distinctly lower then what is observed after the cartel was discovered (excluding the transition from collusion to non-collusion); see Figure 3 below. Mixed evidence is provided by Bolotova, Connor, and Miller


12/10/07

12:16

Page 61

Behavioural Screening and the Detection of Cartels 61 (2005) who find a lower price variance under collusion for lysine but a higher price variance for citric acid. Collusive Marker: Low price variance. Another marker pertaining to price variability also serves as a caveat to the previous marker. Consider a setting in which collusion is imperfectly monitored (Green and Porter, 1984). In each period, firms choose quantities and then observe price. Price depends on firms’ quantities and an unobserved demand shock. As a firm’s quantity is never observed by other firms, a deviation cannot be directly observed. However, price is observed and, in expectation, a higher quantity results in a lower price. Of course, since price depends on an unobserved demand shock, a low price could be due to a low demand shock rather than some firm cheating by producing above their collusive quota. In this setting, a mechanism that will work to sustain collusion is one in which if the realized price is ever too low then firms switch to a punishment phase which is competitive behavior for periods, after which they return to the collusive phase.13 Equilibrium then entails stochastic regime switches where a one-time low demand shock triggers a movement from the collusive phase to the punishment phase—associated with it is a fall in the average price—and, after some length of time, there is a regime switch back to the collusive phase—with a rise in the average price. One then observes abrupt changes in average price which cannot be explained by contemporaneous demand and cost movements. Collusive Marker: Price is subject to regime switches.

3. Quantity markers The markers relating to quantity show how collusion imposes intertermporal structure on market share. To establish this point, let us return to the price game when firms’ costs are stochastic and private information. Cartel members can convey messages about their costs prior to setting price and quantity. As mentioned earlier, an optimal equilibrium can have firms keeping prices and market shares fixed so there are indeed stable market shares. This was mentioned for when cost shocks are independent across firms and over time but it also holds when firms’ costs are persistent over time (Athey and Bagwell, 2004). More generally, firms settle on a collusive outcome with stable market shares when cost persistence is sufficiently high relative to firms’ patience. This result is due to the following logic. When cost persistence increases, it becomes more valuable to a firm to convince its fellow cartel members that it 13 Alternatively, the punishment phase could be an intense price war for some stochastic length of time (Abreu, Pearce, and Stachetti, 1986).


62

12/10/07

12:16

Page 62


has low cost as it influences not only current beliefs (and potentially the current collusive output quotas) but also future beliefs on cost and thus can enhance a firm’s future quota. Given this augmented incentive for a firm to report its cost is low (even when its cost is actually high), inducing truthful revelation either requires firms to be more patient—so they are content to wait for higher market share in the more distant future when they may truly have low cost—or to set lower prices (thus reducing the gain in current profit to a high cost firm from reporting it is low cost). When firms are not very patient, the preference is to forego efficiency in order to support higher collusive prices, which means market shares do not respond to cost shocks. A number of cartels chose to freeze market shares at their pre-cartel levels which can produce stability in actual market shares. The previous year’s market shares were used in the cartels for copper plumbing tubes, organic peroxides, and several vitamins (A, E, and folic acid, in particular), while the average of several years were used in the cartels for citric acid, sorbates, and zinc phosphate. Collusive Marker: Market shares are highly stable over time. Market share stability may only be observed when one uses the right measure. If the cartel is not all-inclusive then a cartel member’s market share may fluctuate because non-cartel supply is changing (often increasing due to the high prices of cartel members) but, at the same time, it’s share of cartel supply is stable. Then there are cases like the sorbates cartel. Collusion occurred in the European market and was comprised of one European company, Hoechst, and four Japanese producers: Chisso, Daicel, Nippon, and Ueno. At the general cartel meeting, sales were allocated between Hoechst and the Japanese producers as a group. The four Japanese producers would then meet and allocate their collective quota amongst themselves. From 1979 to 1994, the allocated market share of Hoechst had little variation as it ranged from 51 to 53%. Even more striking, each Japanese company’s allocated share of Japanese supply never varied over those 15 years.14 Collusive Marker: There is a subset of firms for which each firm’s share of total supply for that subset of firms is highly stable over time. Returning to the previous theoretical model, when instead patience is high relative to persistence, the best collusive equilibrium may have market shares moving over time as firms achieve a more efficient mechanism in which a firm with lower cost has a higher market share (Athey and Bagwell, 2001, 2004). This is shown in a simple situation where cost is high or low. The way the mechanism works is to engage in intertemporal market share favors. A firm that announces low cost and receives a high market share in the current period can expect a lower market share in the next period. This induces the 14

Data was not available to determine whether actual market shares had the intended stability.


12/10/07

12:16

Page 63

Behavioural Screening and the Detection of Cartels 63 firm to truthfully reveal. For if it is high cost and announces low cost, it sacrifices future market share when indeed it might truly be low cost. (Note that market share is especially valuable to a firm when it has low cost.) Thus, market shares are predicted to change over time (with firms’ costs) and, furthermore, a firm’s market share is negatively correlated over time. In the absence of collusion, market share would vary over time but would not be negatively correlated. Similar results of intertemporal market sharing arise in models of repeated auctions which, contrary to the preceding model, do not allow messages to be sent and assume prices are private information. The solution is a historydependent bid rotation scheme; the probability of winning is decreasing in the frequency with which a bidder has won in the past. Thus, firms are favored who have tended to lose recent auctions (Blume and Heidhues, 2003; Skrzypacz and Hopenhayn, 2004). In a non-auction setting, this may take the form of cartel members engaging in a buy-back scheme whereby firms that sold over its quota have to buy output from firms that sold under their quota. For example, in the lysine cartel, Haarman & Reimer was required to buy 7,000 tons from Archer Daniels Midland for exactly this reason. The attractiveness to punishing with a buy-back scheme rather than a price war is established in Harrington and Skrzypacz (2005) and was used in several recent price-fixing cartels including citric acid and vitamins A and E as well as lysine. Collusive Marker: A firm’s market share is negatively correlated over time.

D. Screening in action In this section I would like to illustrate how screening might operate. Let me first consider applying screening to the stationary phase. In examining collusion at auctions of frozen perch, Abrantes-Metz et al (2005) find that the price variance was much lower during the cartel regime compared to the post-cartel phase (excluding the transition from collusion to non-collusion). This is shown in Figure 3 which is reproduced from their paper. The cartel price path (which occurs over 1/6/87–7/9/88) is less sensitive to cost than the (posttransition) non-collusive price path (which occurs over 9/20/88–9/26/89). In practice, a screen could monitor the price variance and assess whether it is low relative to some benchmark. In fact, this research was done with the objective of developing a screen in terms of the price variance. Furthermore, one of the co-authors—Luke Froeb—was, at the time, Director of the Bureau of Economics at the U. S. Federal Trade Commission. Dr. Froeb is not the only FTC official to explore the prospect of screening. Former Director Jonathan Baker used price increases after an industry-specific trough in demand to identify the


64

12/10/07

12:16

Page 64


4 3.5 Price 3 2.5 2 1.5 Cost 1 0.5 0

1/6/87

7/9/88

9/20/88

9/26/89

Figure 3: Collusion and the Variance of Price (Abrantes-Metz et al, 2005) exercise of market power and did so with some success (FTC History, 2003, pp. 108–110). An alternative approach is to identify the transition from non-collusion to collusion or from collusion to non-collusion (as in the case in Figure 3). Cartel formation necessarily entails a change in the price-generating process—and possibly in the market share-generating process—or otherwise it wouldn’t be profitable. Furthermore, the change could be abrupt and thereby detectable. Has average price changed? Has the relationship between a firm’s price and cost changed? Has the relationship among firms’ prices changed? Has the variance of price and market share changed? To argue the implementability of a screen based on finding a structural break, consider the recent episode of collusion in some Nasdaq security markets. In those markets, collusion took the form of avoiding the quoting of odd-eighths so that a market maker would post a bid price of, say, $10 or 10 1/4 but not 10 1/8. This resulted in a minimum bid-ask spread of 1/4. Figure 4 (which is based on a figure from Christie and Schultz, 1999) shows that there is a sharp change in the dollar spread—which effectively measures the pricecost margin—surrounding the switch from quoting all eighths to largely avoiding odd-eighths. Screening for a structural change in the spreadgenerating process would have probably picked up the advent of collusion.


12/10/07

12:16

Page 65

Behavioural Screening and the Detection of Cartels 65

* ** * * * * * ****** * * * 0.45 ******** ** * * * * ** * * ***** * * ** * * * * * ***** *** * *** 0.4 Dollar spreads * 0.35

*** ****

* ****** * * * * ** ***** * * * * ***** *** * *** * ** ********* * * * * *

–60 –50 –40 –30 –20 –10 0 10 20 Day relative to switch

Dollar inside spreads

0.5

0.3

0.25 30

40

50

60

Figure 4: Collusion and Bid-Ask Spreads (Christie and Schultz, 1999)

E. Screening and leniency programs In light of the impact of leniency programs in the US, E.U. and elsewhere, one possible reaction to a proposal to actively screen for cartels is that it isn’t really necessary given that they are being discovered through leniency programs. In response, I want to make two points. First, it is an open question— at least to me—as to how effective leniency programs have been in discovering cartels. I am convinced by their role in prosecution as the evidence is much stronger when it is provided by one of the cartel members. And there clearly are cases, some which I mention below, in which a leniency program was responsible for the discovery of a heretofore unknown cartel. Still, the extent to which leniency programs adequately discover cartels remains an open question. The second point I intend to make is that the presence of an active leniency program makes the case for screening more, not less, compelling because they are complements. In assessing the impact of leniency programs, the evidence is rather solid that they make a difference. The redesign of the US program in 1993 and the redesign of the EU’s program in 2002 were both followed by a very large


66

12/10/07

12:16

Page 66


increase in leniency applications. Of course, one cannot draw inferences about discovery from leniency applications. Even if leniency was awarded prior to an investigation, this is not to say that there was not some knowledge or suspicion about a cartel operating. Indeed, a cursory review of the evidence can be quite misleading, as evidenced by the sorbates cartel case. Chisso received amnesty (under Section A) from the European Commission and paid no fines. In its decision, the European Commission stated: “On 29 September 1998, Chisso informed the Commission about the sorbates cartel in which it participated. At that time, the Commission had not undertaken an investigation, ordered by decision, of the enterprises involved, and did not have any information to establish the existence of the alleged cartel.” [Commission of the European Communities, Commission Decision of 1.10.2003 (Case COMP/E1/37.370—Sorbates)—paragraph 426]

Before one concludes that the leniency program can be credited with detection, let us travel across the Atlantic. Literally the day after Chisso came forward to the European Commission, the U. S. Department of Justice announced that one of the manufacturers of sorbates had agreed to plead guilty to engaging in a global cartel.15 The E.U. leniency program was not responsible for the discovery of the European sorbates cartel. The point of this example is not to argue that leniency programs do not discover cartels but only that there is some misinformation about the role of leniency programs in the discovery process. To pursue this matter a bit more rigorously, let us consider the incentives to use the leniency program. Consider a cartel that formed after the advent of a leniency program and has been operating effectively, such as the dRAM cartel. The question is: What will now induce one of the cartel members to come forward and apply for leniency? What has caused it to forsake collusive profit and knowingly suffer at least some penalties?16 There must have been some change somewhere to cause a firm to stop colluding and apply for leniency. Let us consider what some of those changes might have been. The most obvious change is that information about collusion has leaked out and this has raised the prospects of prosecution. However, the focus here is on conditions that induce spontaneous reporting—a firm applies for leniency when there is no knowledge about collusion. One possibility is that there is a change in management and the new management disagrees with the corporate strategy of price-fixing.17 An example is the monochloroacetic acid cartel. 15 U. S. Department of Justice, “Eastman Chemical Company Agrees to Plead Guilty to Price Fixing Charges and Pay $11 Million Criminal Fine.” Press Release. September 30, 1998. 16 Until there are rewards, leniency cannot ever truly be full. This is obviously the case in the US as a firm is still liable for single customer damages, but it is true more broadly as the employees involved in the conspiracy are likely to suffer some punishment by the company and those penalties are not waived. 17 A career concerns perspective is that the new management finds this hidden liability of the previous management—the penalties that would have to be paid if caught—and wants to ensure that this liability is not attributed to them.


12/10/07

12:16

Page 67

Behavioural Screening and the Detection of Cartels 67 Upon acquiring the Hoechst chemicals business, the management of Clariant discovered that the business was engaging in collusion. The new management chose to inform the European and American authorities and apply for leniency.18 A second source of change that might induce a firm to come forward is the steady accumulation of penalties. It is possible that penalties might reach a level that a firm prefers to come forward rather than risk having to pay them. That choice becomes even more compelling if collusion has actually collapsed since applying for leniency does not reduce future profits; non-collusive profits will be earned whether or not a firm applies for leniency. The US has capitalized on an accentuated chance of discovery to induce spontaneous reporting. The context is one in which a firm has been caught colluding in one market and the issue is whether they choose to reveal information about collusion in other markets for which there is currently no evidence. Cartels have been discovered this way through the Omnibus Question and, in conjunction, the Amnesty Plus and Penalty Plus programs. A firm that is currently being prosecuted can, under the Amnesty Plus program, receive full amnesty by reporting collusion in separate markets and receive reduced penalties in the markets for which it is already being prosecuted. Furthermore, if they do not report this other collusion and it is ultimately found, the Penalty Plus program enhances the severity of penalties. The incentives to apply for leniency are stronger because: i) given that the firm is currently being prosecuted for collusion, it is reasonable to attach higher probability that its collusion in these other markets will ultimately be discovered; and ii) the expected penalty is larger for not coming forward. Information about collusion in other markets can come out during the Omnibus Question phase. At the conclusion of a witness interview under the leniency program, the U. S. Department of Justice asks: “Do you have any information whatsoever, direct or indirect, relating to price-fixing, bid rigging, etc. with respect to other products in this industry or in any other industry?”. Failure to answer the question truthfully means a loss of amnesty and possible prosecution for perjury. I believe the power of the Omnibus Question rests largely with the incentives it creates for the individual manager rather than the corporation. Though revealing the company is colluding in some other markets may cause additional financial harm to the company (though, as argued above, the Amnesty Plus and Penalty Plus may actually mean the firm is benefited), this employee is apt to be more concerned with jeopardizing the amnesty they’ve received from prison sentences and individual fines. Having already revealed their part in collusion, I suspect there is a strong incentive to avoid perjury and tell all to the authorities. The Omnibus Question can be a meaningful tool for cartel detection. There have clearly been cases discovered through these programs and the Omnibus Question. (A clear documentation of how many would be useful.) 18

“The Week: June 12, 2002,” Chemical Week, 6/12/2002, Vol. 164, Issue 24.


68

12/10/07

12:16

Page 68


That a leniency program can use one cartel prosecution to spur the discovery of other cartels is exemplified with this sequence of events:19 “[. . .] the investigation and prosecution of the international citric acid cartel led to the investigation and prosecution of the international sodium gluconate cartel, which resulted in the investigation and prosecution of the international sodium erythorbate cartel, which, in turn, led to the investigation and prosecution of the maltol cartel.”

As just argued, there are cases in which it is reasonable to conclude that a leniency program discovered previously unknown cartels. Nevertheless, it is an open question regarding to what extent leniency programs take advantage of some knowledge or suspicion about collusion—inducing cartel members to come forward due to the heightened concern about prosecution because of that knowledge or suspicion—and to what extent they cause spontaneous reporting and thereby truly detect cartels. I am unconvinced that leniency programs in and of themselves are sufficiently effective in discovering cartels that we ought to forsake other avenues. That statement is sufficiently weak that most people will agree with it. My proposal is to take it seriously by suggesting we pursue screening. The most compelling reason for working towards the creation of a screening program at this time is that its efficacy has been enhanced by the presence of an effective leniency program. A firm is more apt to stop colluding and come forward and apply for leniency when the chances of prosecution are greater. An antitrust authority that identifies an industry for further scrutiny through some form of screening and conveys these suspicions to the suspected firms could well induce one or more cartel members to come forward and apply leniency. Hence, the development of a leniency program makes the case for screening more compelling. In a recent speech, Deputy Assistant Attorney General Scott Hammond claimed:20 “You can’t catch a thief with an economist.” That may well be true. But with an active screening program, you may be able to scare a thief with an economist and that could be enough to induce them to report their crimes.

19 Deputy Assistant Attorney General Scott Hammond (Criminal Enforcement, Antitrust Division, US Department of Justice), “Cornerstones of an Effective Leniency Program”, ICN Workshop on Leniency Programs, Sydney, Australia, November 22–23, 2004. 20 “ Ten Strategies for Winning the Fight Against Hardcore Cartels”, OECD Competition Committee, Paris, France, October 18, 2005.

(E) Ehlermann 07 Rey

12/10/07

12:16

Page 69

II Patrick Rey1 On the Use of Economic Analysis in Cartel Detection

A. Introduction In principle, various economic approaches are available to detect cartels. For example, following the ‘structure-conduct-performance’ paradigm, a structural approach could seek to identify the industry characteristics that make the industry prone to collusion; a conduct approach could look at behavioural patterns that are more likely to be associated with collusion than with competition; and a performance approach could, for example, look for supracompetitive profitability. These approaches can be useful, particularly in contexts such as merger control, where the spotlight is already placed on a given industry and competition agencies must figure out whether collusion is likely to be there. However, these methods are costly to adopt in a routine way to detect collusion, in the absence of warnings already attracting attention to a particular case. Moreover, at this stage, these approaches do not provide robust, clear-cut answers. Overall, while further theoretical and empirical research is probably warranted in this area, it is still difficult to use these approaches in a systematic way. Yet economic analysis can also be used in a different way. By nature, cartels cannot depend on binding, legally enforceable contracts; they must therefore rely on some form of self-enforcement. This can make cartels fragile, and economic analysis can help to identify the critical factors that may exacerbate this fragility and induce cartel participants to come forward and report their activity. To put it another way, playing firms against each other might make it possible to move from a ‘police patrol’ to a ‘fire alarm’ mode of operation, to use the terminology of McCubbins and Schwartz:2 rather than sending out police patrols in the street and looking for evidence of trouble, it may be better to try to induce those engaged in illegal activity—at least some of them— to denounce it. This approach can also contribute to deterring the formation of new cartels.

1 Professor of economics, University of Toulouse, Ecole Polytechnique and Institut Universitaire de France. 2 See McCubbins M. and Schwartz T. (1984): “Congressional Oversight Overlooked: Police Patrols vs Fire Alarms”, 28 American Journal of Political Science 165.


70

12/10/07

12:16

Page 70


This paper is organized as follows. First, I briefly review the structural factors that may facilitate collusion, which may be used to identify candidates for further investigation. Then, I discuss the factual evidence that might be further indicative of collusion, before presenting some of the conclusions from recent developments in the economics literature on leniency and whistleblowing programmes.

B. Structural factors3 The economic literature on collusion4 has identified several factors that may facilitate collusion. First, there are some basic structural variables,5 such as: —Low number of competitors: the larger the number of parties involved, the more difficult coordination becomes;6 and as the number of firms increases, each firm gets a smaller share of the pie generated by collusion, which makes deviation more tempting (short-run gains from deviation increase, while the long-run benefit of maintained collusion is reduced).7 3 This discussion borrows from the IDEI report for DG Comp, “The economics of tacit collusion”, available at http://ec.europa.eu/comm/competition/mergers/review/the_economics_of_ tacit_collusion_en.pdf. For a general overview of the economics literature on collusion, see Motta M. (2004): Competition Policy: Theory and Practice, Cambridge University Press. 4 The economic literature focuses mainly on tacit collusion, which, in contrast to typical unlawful cartels, lacks explicit coordination mechanisms or information exchanges. However, explicit cartels and tacit collusion share many common features, since in both cases firms must have proper incentives to implement the collusive agreement. Therefore, the insights of the literature on tacit collusion, which rely on the analysis of self-enforcement, apply as well to explicit collusion. 5 While the literature often relies on conceptual analyses, there are also some empirical studies on tacit collusion, following the initial work of Posner R. (1970): “A Statistical Study of Antitrust Enforcement”, 13 Journal of Law and Economics 365, and of Hay G. and Kelley D. (1974): “An Empirical Survey of Price Fixing Conspiracies”, 17 Journal of Law and Economics 13. These studies are surveyed by Levenstein M. and Suslow V. (2004): What Determines Cartel Success?, Working Paper, University of Michigan, who emphasize their disparity. However, they provide support for the importance of some of the structural factors discussed here (such as the number of firms, entry barriers, and market stability). Symeonidis G. (“In Which Industries Is Collusion More Likely? Evidence from the UK”, 51 Journal of Industrial Economics 45 (2003)) uses a natural experiment (the adoption of the 1956 Restrictive Trade Practices Act) to identify the industries subject to price-fixing agreements in the 1950s. His analysis suggests that collusion is more likely the higher the degree of capital intensity, and that it is less likely in advertisingintensive industries than in low-advertising industries. He also finds a non-monotonic relationship between market growth and the likelihood of collusion (moderate growth makes collusion more likely, while rapid growth makes it less likely). However, he finds no clear link between concentration and the incidence of collusion. 6 See for example Compte O. and Jehiel P. (2002): “Multi-Party Negotiations,” mimeo, CERAS. 7 This insight is valid when holding all other factors constant. However, the number of firms is endogenous and reflects other structural factors, such as barriers to entry and product differentiation.


12/10/07

12:16

Page 71

On the Use of Economic Analysis in Cartel Detection 71 —High entry barriers: in the absence of entry barriers, any attempt to maintain supra-competitive prices would trigger market entry (e.g., short-term or ‘hit-and-run’ entry strategies), and this would erode the profitability of collusion;8 moreover, the prospect of future entry tends to reduce the scope for retaliation. —Frequent interaction between the firms: firms can more easily sustain collusion when they interact more frequently, since they can then react more quickly to deviations. —Market transparency: quick retaliation when one firm undercuts the others requires such deviation to be identified by the other participants; as a result, collusion is more difficult to sustain when individual behaviour is not readily observable and cannot easily be inferred from available market data.9 Second, there are characteristics of the demand side: —Market growth: collusion is easier to sustain when short-term gains from a deviation are small compared with the cost of future retaliation; thus, for a fixed number of market participants, collusion is easier to sustain in growing markets, where today’s profits are small compared with those of tomorrow.10 —Absence of significant fluctuations or business cycles: collusion is less sustainable in markets that are subject to demand fluctuations (see FTC v. Arch Coal),11 since peak periods increase short-term gains from a deviation relative to the potential cost of later retaliation.12 Hence, demand fluctuations hinder collusion, particularly when fluctuations are deterministic (as in the case of seasonal cycles, where it is anticipated that the future will be less rosy) as opposed to random. 8 However, John Sutton has stressed the possibility of durable situations where the profits from high prices would simply be dissipated through inefficient duplication of fixed costs and an excessive number of market participants. 9 This issue was first highlighted by Stigler’s classic paper, “A Theory of Oligopoly”, 72 Journal of Political Economy 44 (1964). Since then, it has been formally analyzed by Green E. and Porter R. (1984): “Non-Cooperative Collusion under Imperfect Price Information”, 52 Econometrica 87, and Abreu D., Pearce D. and Stachetti E. (1985): “Optimal Cartel Equilibria With Imperfect Monitoring”, 39 Journal of Economic Theory 251. It may be noted that market stability can contribute to enhancing market transparency, since inferring deviations from collusive conduct is easier and requires less detailed information when the market is stable. See Kühn K. (2001): “Fighting Collusion by Regulating Communication between Firms”, 32 Economic Policy 169, discussing the role of information exchanges in this context. 10 This conclusion appears somewhat at odds with some court cases and opinions. Indeed, demand growth is in practice often interpreted as a factor hindering collusion. One possible reason for this apparent discrepancy is that the above reasoning assumes that the number of market participants remains fixed despite market growth (due for example to regulations or limited access to key inputs or technologies), while in practice, entry may be easier in growing markets. See, e.g., Case T-102/96 Gencor v Commission [1999] ECR II-753. 11 329 F. Supp. 2d 109 (D.D.C. 2004), appeal dismissed per curiam. 12 See Rotenberg J. and Saloner G. (1986): “A Supergame-Theoretic Model of Business Cycles and Price Wars during Booms”, 76 American Economic Review 390; Haltiwanger J. and Harrington J. (1991): “The Impact of Cyclical Demand Movements on Collusive Behavior”, 22 RAND Journal of Economics 89.


72

12/10/07

12:16

Page 72


—Demand elasticity: elasticity of demand has no clear impact on the sustainability of collusive prices, but collusion is more profitable when demand elasticity is low, which in turn can influence the firms’ willingness to establish a cartel and facilitate its implementation despite the risk of being caught by antitrust authorities. Similarly, collusion is a greater concern for consumers when demand is inelastic, both because there is more scope for large price increases and because consumers are hurt more by a given price increase when they have few alternatives.13 —Buying power: even a perfect cartel may find it difficult to impose high prices on powerful buyers, which makes cartel activity less profitable and, by the same token, more fragile.14 —The absence of club and network effects: club effects (that is, when consumers benefit from being in the same ‘club’: using the same software, typing on the same type of keyboard, subscribing to the same operator, and so forth) tilt the market in favour of a single participant, thereby creating a ‘winner take all’ type of competition which is not prone to collusion. In addition, club effects create lock-in effects that reinforce the position of the market leader and thus increase the benefits derived from such a position. Finally, there are characteristics related to the supply side: —Mature industries with stabilized technologies: by allowing one firm to gain a significant advantage over its rivals, innovation limits the scope for collusion; the prospect of innovation reduces the value of future collusion as well as the cost from possible retaliation from these rivals. —Symmetric costs: it is easier to collude with equals.15 First, cost asymmetry,16 for example, may make it difficult to agree on a common pricing policy, since firms with a lower marginal cost will insist on lower prices than what high-cost firms would wish to sustain; more generally, different cost structures may rule out ‘focal points’ and may exacerbate coordination problems. Second, technical efficiency would require allocating market share to low-cost firms, which might be difficult to achieve in the absence

13 Thus, the less elastic the demand the greater the potential harm to consumers. However, the impact on total welfare is more ambiguous, since price increases generate less distortion when demand is inelastic. See, e.g., Tirole J. (1988): The Theory of Industrial Organization, MIT Press, Cambridge Massachusets, for a discussion of this issue. 14 See Snyder C. (1996): “A Dynamic Theory of Countervailing Power”, 27 RAND Journal of Economics 747, noting that large buyers can design procurement schemes that reduce the scope for collusion. 15 This intuition may also explain the informal discussions about the role of so-called ‘mavericks’. A maverick firm can be interpreted as a firm with a drastically different cost structure (or a drastic difference in any other relevant dimension), which is thus unwilling to participate in or stick to collusive action. 16 See Bain J. (1948): “Output Quotas in Imperfect Cartels”, 62 Quarterly Journal of Economics 617 for an early discussion.


12/10/07

12:16

Page 73

On the Use of Economic Analysis in Cartel Detection 73 of side-transfers.17 Third, low-cost firms may be more tempted to deviate from a cartel agreement, both because they may gain more from undercutting their rivals and because they have less to fear from a possible retaliation by high-cost firms.18 —Symmetric capacities: an asymmetric distribution of production capacities hinders collusion; the firm with the largest capacity has more incentive to undercut its rivals, particularly if their production capacities limit their retaliatory power.19 —Product homogeneity: a firm with a better product is in a situation similar to that of a firm with lower costs, i.e., it has more to gain from cheating on the cartel (or, put another way, it may favour a different ‘net’ price, adjusted for the quality differential), and has less to fear from retaliation;20 in addition, product differentiation may contribute to reducing market transparency.21 —Multi-market contact: firms can more easily sustain collusion when they are present on several markets.22 First, multi-market contact may make firms interact more frequently and allow them to maintain a cartel in markets where industry characteristics would otherwise prevent them from doing so. Second, multi-market contact may soften asymmetries arising in individual markets—for example, one firm may have a competitive advantage in one market and its rival may have its own competitive advantage in another market; while a (single) market-level analysis may then suggest 17 However, side-transfers need not be monetary. For a discussion of these issues, see Osborne M. and Pitchik C. (1983): “Price Competition in a Capacity-Constrained Duopoly”, 38 Journal of Economic Theory 238; Schmalensee R. (1987): “Competitive Advantage and Collusive Optima”, 5 International Journal of Industrial Organization 351. 18 See Harrington J. (1989): “Collusion Among Asymmetric Firms: The Case of Different Discount Factors”, 7 International Journal of Industrial Organization 289; Thalm J. (2005): “Optimal Collusion under Cost Asymmetry,” mimeo, University of Toulouse. See also Mason C., Phillips O. and Nowell C. (1992): “Duopoly Behavior in Asymmetric Markets: An Experimental Evaluation”, 74 Review of Economics and Statistics 662, noting that, in experimental duopoly games, cooperation is more likely when players face symmetric production costs. 19 See Lambson V. (1987): “Optimal Penal Codes in Price-Setting Supergames with Capacity Constraints”, 54 Review of Economic Studies 385; Lambson V. (1996): “Some Results on Optimal Penal Codes in Asymmetric Bertrand Supergames”, 62 Journal of Economic Theory 444; Davidson C. and Deneckere R. (1984): “Horizontal Mergers and Collusive Behavior”, 2 International Journal of Industrial Organization 117; Davidson C. and Deneckere R. (1990): “Excess Capacity and Collusion”, 31 International Economic Review 521. 20 See, e.g., Ross T. (1992): “Cartel Stability and Product Differentiation”, 10 International Journal of Industrial Organization 1; Martin S. (1993): “Endogenous Firm Efficiency in a Cournot Principal-Agent Model”, 59 Journal of Economic Theory 445. 21 See Raith M. (1996): “Product Differentiation, Uncertainty and the Stability of Collusion,” London School of Economics-STICERD Discussion Paper Series EI/16:49. 22 The classic reference is Bernheim B. and Whinston M. (1990): “Is Everything Neutral?”, 96 Journal of Political Economy 308. See also Parker W. and Röller L.-H. (1997) “Collusive Conduct in Duopolies: Multimarket Contact and Cross-Ownership in the Mobile Telephone Industry,” 28 RAND Journal of Economics 304, and Evans W. and Kessides N. (1994) “Living by the ‘Golden Rule’: Multimarket Contact in the U.S. Airline Industry”, 109 Quarterly Journal of Economics 341.


74

12/10/07

12:16

Page 74


that collusion is difficult to sustain, multi-market contact in such a case restores an overall symmetry that facilitates collusion. International cartels such as the vitamins cartel provide good examples of such multi-market contacts. —Structural links: structural links can facilitate collusion among firms. For example, cross-shareholdings reduce the gains derived from undercutting the other firm. Joint venture agreements may further enlarge the scope for retaliation because a participating firm can punish a deviating partner by, e.g., investing less in the venture.23 For these reasons, collusion is more likely to emerge in markets where competitors are tied through structural links. —Cooperative and other contractual agreements: even in the absence of structural links, simple cooperative agreements can contribute to foster collusion. As in the case of joint ventures, these cooperative agreements can for example enlarge the scope for retaliation, thereby enhancing the ability to punish deviating partners. This may be particularly relevant for industries such as the telecommunications industry, where competitors need to reach interconnection agreements in order to offer widespread connectivity. These agreements not only enlarge the scope for retaliation, they also have a direct impact on the operators’ pricing strategies.24 More generally, firms may alter their contractual agreements, either between themselves or with third parties, so as to facilitate collusion.25 This brief overview suffices to show that many factors may affect the scope for collusion. Most often, a given market will have some characteristics that facilitate collusion, and others that tend to hinder collusion. Predicting the likelihood of collusion on this basis alone is thus a complex question.26 In addition, while these factors may help to identify those markets in which collusion might be attractive or feasible, they cannot be relied upon to deter-

23 For a detailed analysis, see Martin S. (1995): “R&D Joint Ventures and Tacit Product Market Collusion”, 24 European Journal of Political Economy 357. 24 For example, telecoms operators that compete in linear prices could give each other incentives to maintain high prices, even in the absence of repeated interaction, by agreeing on a high reciprocal access charge—see, e.g., Armstrong M. (1998): “Network Interconnection in Telecommunications”, 108 The Economic Journal 545; Laffont J., Rey P. and Tirole J. (1998): “Network Competition: I. Overview and Nondiscriminatory Pricing”, 29 RAND Journal of Economics 1. 25 Marketing agreements can be employed to that effect. Jullien B. and Rey P. (“Resale Price Maintenance and Collusion,” working paper, University of Toulouse, 2002) show, for example, that producers of consumer goods can resort to Resale Price Maintenance to impose more uniform prices across local retail markets, thereby making it easier to detect deviations from a collusive price. Record companies have been accused of marketing their offerings according to simple pricing grids (with only a few categories, instead of personalized prices for each author or title) for a similar purpose. 26 Moreover, the interplay of the factors may be important, as illustrated by the effects of market growth and entry barriers. If entry barriers are so large that entry is highly unlikely to occur, market growth fosters collusion; if instead entry barriers are moderate, market growth may stimulate entry, which in turn impedes collusion.


12/10/07

12:16

Page 75

On the Use of Economic Analysis in Cartel Detection 75 mine whether firms are actually colluding.27 In the context of merger control, where the merger control office must evaluate ex ante the future evolution of the industry, short of determining whether collusion will indeed occur, a highly difficult if not impossible task, the merger control office can review these factors to assess whether collusion will become easier to sustain.28 In an ex post antitrust context, these factors may be looked at to determine whether collusion is a plausible concern, but their analysis will not suffice to establish the existence of actual cartel activity. An analysis of the history of the industry might help to answer that question.

C. Factual evidence and behavioural patterns Two types of factual evidence might be relevant to identify collusion: first, since the object or effect of collusion is to impede competition, antitrust authorities could look at signs of supra-competitive prices or profits; alternatively, they can look for behavioural patterns that are more likely to be associated with collusion than with competition. I briefly discuss these two types of evidence.

1. Supra-competitive performance This approach relies on a comparison between the actual performance of the industry and the performance that one would expect from ‘normal’ competition in a comparable industry. This requires reasonably good access to reliable price or accounting data, but it is also necessary to have an idea of what the comparable competitive levels would be. The latter issue is clearly a tough one, and it requires detailed data and analysis of costs and demand conditions (not only in the current situation but in the supposedly competitive one as well). This exercise is all the more difficult to realize in concentrated industries, which 27 Formally speaking, the same market situation can give rise to different equilibria. In particular, firms may well compete in each period as if it were the last one, even if another equilibrium exists in which they maintain monopoly prices in each and every period. 28 While many factors appear relevant when evaluating the impact of a merger on collusion, some may be more important than others. First, whenever there are no entry barriers, or where firms interact rather infrequently, or the industry is primarily driven by innovation, collusion is unlikely to constitute a significant concern. A second series of factors, such as the number of market participants, the degree of symmetry among those participants, and the existence of contractual or other types of arrangements, are all relevant and likely to be directly affected by mergers. The other factors (market transparency, product homogeneity, demand trend and fluctuations, multi-market contact, and so forth) may be relevant when the previous factors do not suffice to send a clear signal.


76

12/10/07

12:16

Page 76


will often be subject to ‘imperfect competition’ anyway (that is, even a purely static, non-cooperative form of ‘normal’ competition would still yield significant price-cost margins and profits). More generally, this type of in-depth study requires deep knowledge and expertise about the industry, and it is more naturally associated with regulatory supervision than with the antitrust oversight. One can also rely upon econometric methods when sufficiently rich data are available. This approach can, for example, provide an estimate of the likelihood of collusion. However, it is sensitive to various modelling assumptions and specifications with respect to the demand and supply side.29 Recently, however, the development of structural models of oligopolistic markets using panel data have made predictions possible even in the absence of precise information on costs: in essence, sufficiently detailed data on prices and quantities enable econometricians to estimate both demand and cost parameters. This opens up new scope for application. Furthermore, this approach allows one to distinguish supra-competitive mark-ups from “normal but imperfectly competitive” ones.30 This approach can thus be useful in determining whether collusion is likely or not in a given industry. However, it is time- and resource-consuming (data collection, to begin with, can be a real issue). In addition, this approach cannot yield certainty: by their nature, econometric methods rely on probabilities; they can provide an estimate of the likelihood of collusion, as well as confidence intervals, but they cannot produce definite yes/no answers.

2. Collusive patterns Alternatively, one can look for behavioural patterns that are more naturally linked to collusion than to competition. The usual suspects include price parallelism and stable market shares. Concerns have also been raised by a lack of price response to changing demand conditions. 29 For example, using the same dataset, Porter R. (“On the Incidence and Duration of Price Wars”, 23 Journal of Industrial Economics 415 (1985)) and Ellison G. (“Theories of Cartel Stability and the Joint Executive Committee”, 25 RAND Journal of Economics 37 (1994)) reach different conclusions about the likelihood of different collusive schemes adopted by an 1880s US railroad cartel. 30 For example, Nevo A.’s analysis of the US ready-to-eat cereal industry (see “Mergers with Differentiated Products: The Case of the Ready-to-Eat Cereal Industry”, 31 RAND Journal of Economics 395 (2000); “Measuring Market Power in the Ready-to-Eat Cereal Industry”, 69 Econometrica 307 (2001)) suggests that mark-ups as high as 45% are more likely to be due more to product differentiation (making consumers willing to pay for their favourite brand) and multibranding (giving firms an incentive to maintain high prices on each brand, to avoid cannibalizing their other brands) rather than to collusion. Similarly, the analysis of Pinkse J. and Slade M. (“Mergers, Brand Competition, and the Price of a Pint”, 48 European Economic Review 617 (2004)) suggests that the margins observed in the UK brewing industry result from intrinsically imperfect competition (due for the main part to product differentiation, and for the remaining part to local concentration).


12/10/07

12:16

Page 77

On the Use of Economic Analysis in Cartel Detection 77 Courts have been reluctant to take such patterns as evidence of collusion. In my view, this reluctance is justified and may even extend to interpreting these patterns as signals, if not proof of collusion, since such patterns can be consistent with competitive as well as collusive behaviour. Therefore, in the absence of any further empirical investigation, one cannot conclude that the industry has been, in fact, colluding. Three examples can illustrate that point.

2.1. Price parallelism Consider first a benchmark situation where all competitors offer the same, homogeneous product. In those circumstances, by nature and whatever the effectiveness of competition, the prices of all active firms will necessarily be always equal, since any price above the others would attract no customer. All these prices are therefore deemed to move together: one should expect a perfect parallelism in the evolution of prices, irrespective of whether firms are colluding or competing head to head against each other. This case is admittedly an extreme one, but by continuity the reasoning extends to situations where firms have similar offerings. The more similar the firms’ offerings, the narrower the price range that one should expect, regardless of whether firms collude or compete. But parallelism can also obtain in industries with significant product differentiation. To see this, let us use the simplest model of oligopolistic competition, the so-called ‘linear model’, where both costs and demand are supposed to have a simple, linear shape.31 In this simple model, whatever the nature of the strategic interaction (in prices à la Bertrand or in quantities à la Cournot) and whatever the degree of product differentiation, prices are the same across firms, whether they compete or perfectly collude. Thus, here again one should expect as much parallelism with competition as with collusion. To explore the issue further, one can introduce some asymmetry on either the cost or the demand side. Doing so breaks down the equality of the various firms’ collusive prices as well as of their competitive prices, but it does not eliminate the scope for price parallelism. To see this, suppose for example that there are two firms 1 and 2 with different unit costs, c1 < c2, and that the demand for each product i is of the form qi = d – (pi – spj)/(1 – s), where pi and pj denote respectively the price of that firm and the rival’s price; d is a demand scale parameter; and s is a parameter reflecting the degree of substitution between the products.32 It can be checked that, whereas, under competitive conditions, the prices of the two firms differ from each other (the firm 31 That is, firms have a constant unit cost and face a demand of the form q = d – a p + ⌺ a i i i i j ij pj, where d i reflects the size of the demand and the a’s parameters reflect the sensitivity of the demand with respect to the various prices. 32 That is, the products are perfect substitutes when s = 1 and completely differentiated when s = 0.


78

12/10/07

12:16

Page 78


with a lower cost charging a lower price), they both respond in the same way to a change in the demand parameter d.33 And while collusive price would also evolve in a parallel fashion, the price range is actually larger for collusive prices than for competitive prices.34 Thus, in the context of this simple example, a narrow range would be more a signal of competition than of collusion. A similar analysis could be conducted for alternative forms of asymmetries, or for a larger number of firms, at the cost of notational (and computational) complexity.35 The bottom line is that, while the level of prices might be used to determine whether firms are competing or colluding (provided that the observer has enough information on cost and demand conditions), the fact that they evolve in parallel does not, by itself, allow for such identification.

2.2. Market share stability Here again, this feature does not by itself make it possible to distinguish collusion from competition, since market share stability can be consistent with competition as well as collusion. To see this in a simple way, it suffices to consider any situation where firms face symmetric supply and demand conditions. Firms would then have the same market share, both in the most profitable collusive outcome and in the competitive one, whatever the scale of demand. A shift in demand thus has no impact on the firms’ relative market shares. This example is of course extreme, and in practice asymmetry in the firms’ ability to compete will generally yield asymmetric market shares. Yet this does not imply that market shares would fluctuate when there is a general shift in demand, nor does it imply that market shares should be fluctuating more in the case of competition than in the case of collusion. While asymmetry in the firms’ technological ability, for example, leads in general to asymmetric market shares, one would expect this to be the case with collusion as well as with competition. And it may even be the case that market shares are more stable or diverge less with competition than with collusion. To illustrate these points, we can turn again to the linear model presented above, where the demand depends on a scale parameter d and on the degree of substitutability between the products, measured by s, and where two firms 33 A decrease in d by 1 unit induces the same decrease in both firms’ prices, equal to (1-s)/ (2-s). In the polar case of perfect substitutes (product homogeneity, s = 1), prices do not adjust to demand changes and are instead entirely driven by costs. 34 The price range is (c – c )/2 in the case of collusion and is smaller and equal to (c – c )/ 2 1 2 1 (2 + s) in the case of competition. 35 For example, if the demand for one product exhibits a constant elasticity with respect to the price of that product, competitive prices are driven solely by (cost levels and) the value of the elasticity. Thus, as long as demand fluctuations affect the market size but not the demand elasticity, competitive prices would remain constant—an extreme form of ‘parallelism’. In contrast, collusive prices would respond to changes in market sizes. Thus, in that case competitive prices would actually be ‘more parallel’ than collusive ones.


12/10/07

12:16

Page 79

On the Use of Economic Analysis in Cartel Detection 79 have different unit costs, c1 < c2. In this very simple model, the difference in output levels is entirely driven by the price differential; and whether firms compete or perfectly collude, this price differential depends on the cost differential c2 – c1 and on the degree of substitutability s, but it is unaffected by a shift in the demand parameter d. Thus, while output levels are different (the firm with the lower cost having a higher output and thus a higher market share than the other), the difference between the output levels remains constant over time in spite of changes in the demand parameter d, whether firms compete or collude.36 This in turn implies that the difference in market shares is smaller under competitive conditions than in the case of collusion, since the output differential is the same and total output is larger with competition. In addition, a declining shift in demand increases the market share differential under both competition and collusion, but the adjustment is smaller in the case of competition.37 Thus, in the light of this simple example, a ‘small’ adjustment of market shares when demand is declining should be interpreted as a sign of competition rather than collusion.

2.3. Price reaction to changing demand conditions Collusion concerns are sometimes raised when prices do not appear to react sufficiently to changes in the environment, such as shocks to demand. However, even colluding firms would have an incentive to revise their coordinated behaviour in reaction to changing demand conditions. Furthermore, there is no clear reason suggesting that the optimal adaptation of prices would systematically involve smaller price reductions under collusion than under competition. It could actually be argued that, as competitive prices are driven more by cost considerations, they might respond even less than competitive ones to changes in demand. Accordingly, a ‘small’ change in prices should be interpreted as a sign of competition rather than collusion. On the magnitude of competitive price responses to changes in demand. Generally speaking, the outcome of non-collusive competition among a limited number of competitors is primarily driven by the level of their (marginal) costs of production and by the cross-elasticity of demand for each firm’s product. For example, consider the benchmark situation in which firms produce the same homogeneous product at the same constant unit cost. The 36 The output differential is given by (1+s)(c2 – c1)/(2+s)(1 – s) in the case of competition and by (1+s)(c2 – c1)/2(1 – s) in the case of perfect collusion. In both cases, an increase in, say, the degree of substitutability s has a qualitatively similar impact on (namely here, it induces an increase in) both output differentials. 37 The market share differential is given by (1+s)(2 – s)(c2 – c1)/(2+s)(1 – s)2(d – c1 +c2)/2) in the case of competition and by (1+s)(c2 – c1)/(1 – s)2(d – c1 +c2)/2) in the case of perfect collusion. The impact of a reduction in the demand scale parameter d on the market share differential is thus (2 – s)/(2+s) (< 1) smaller with competition than with collusion.


80

12/10/07

12:16

Page 80


competitive price then simply reflects the level of the unit cost, and is entirely insensitive to demand shocks. Thus, in this most competitive situation, prices would not change when demand falls. Consider now another situation, where firms still have constant unit costs but offer differentiated products. Competition is admittedly less intense in that case, so that competitive prices reflect both cost and demand conditions. Suppose further that, for each firm, the demand for its products exhibits a constant elasticity with respect to that firm’s prices.38 In that case, the normal competition among the firms yields prices that depend on this constant elasticity and not, for example, on the scale of the demand. Thus, as long as shifts in demand have no impact on the own-elasticity of demand for one firm’s products with respect to its own price, the competitive price levels remain unchanged. Do collusive prices adjust less than competitive prices to changes in demand? As already hinted, competitive prices tend to be driven more by cost considerations, and thus less by demand considerations, than collusive or monopolistic prices. To see this, let us again use the simple linear model of oligopolistic competition already presented above. In this simple model, whatever the nature of competition (in prices à la Bertrand or in quantities à la Cournot) and whatever the degree of product homogeneity or differentiation: —both the competitive and the collusive price levels can be expressed as a weighted average of the demand size parameter and of the unit costs; and —the weight put on the demand parameter is larger, not smaller, in the case of perfect collusion.39 In the light of this simple illustration, a supposedly small change in prices should be interpreted as a sign of competition rather than of collusion. Furthermore, as already discussed in the previous section, downwards shifts in demand or demand fluctuations tend to make collusion more difficult to sustain, particularly when demand is low or falling.40 It follows that collusive 38 That is, the demand for firm i’s products is of the form q = Q (p )p ⫺␧, where p represents i i -i i -i the other firms’ prices. 39 In the simplest case, where two firms produce homogeneous goods and the market demand then takes the simple form Q = d – p, the perfectly collusive price (i.e., the monopoly level) puts equal weights on the demand and cost parameters ( pm = (d + c )/2), whereas the competitive price just reflects the unit cost in the case of price competition à la Bertrand ( pB = c), and puts on the demand parameter a weight that is half of that put on cost in the case of quantity competition à la Cournot ( pC = (d + 2c )/3). In the latter case, the difference between the two weights furthermore increases with the number of firms. For example, in the case of five firms, the ratio between the two weights becomes 1 to 5: pC = (d + 5c)/6. 40 See, e.g., Rotemberg J. and Saloner G. (1986): “A Supergame-Theoretic Model of Business Cycles and Price Wars during Booms”, 76 American Economic Review 390, or Haltiwanger J. and Harrington J. (1991): “The Impact of Cyclical Demand Movements on Collusive Behavior”, 22 RAND Journal of Economics 89.


12/10/07

12:16

Page 81

On the Use of Economic Analysis in Cartel Detection 81 prices then exhibit even larger adjustments to changes in demand conditions, since firms cannot sustain as much collusion in bad times as they can in good times. However, as emphasized above, none of these features (i.e., price parallelism, stable market shares, lack of price response to changing demand conditions) allows an outside observer to distinguish collusion from competition. They are all consistent with competition as well as with collusion. Furthermore, while this paper has addressed each feature in turn for the sake of exposition, it should be clear at this point that a combination of the three features is also consistent with competition as well as with collusion. More recently, some work has been done to explore new and promising ideas, in order to identify specific behavioural patterns that are more likely to be indicative of collusion than competition. For example, Joseph Harrington has stressed the role of sudden changes in behaviour. On the one hand, the establishment of a cartel would be expected to lead to rapid price increases, while occasional collusion break-downs might lead to temporary price wars; cartels will seek to avoid sending signals that are too obvious, which might lead them to adapt prices more slowly.41 One might expect a cat-and-mouse situation, where colluding firms try to avoid adopting precisely the type of pattern that one would normally associate with cartels. This would lead to revised expectations about cartel behaviour, and so on and so forth. The issue is further complicated by the fact that many cartels probably operate in a less than fully effective way, which multiplies the number of possible patterns. Overall, while further research in this area is probably warranted, at the moment it appears difficult to rely solely on this approach to detect cartels.

D. Leniency and whistleblowing programmes Economic analysis, both conceptual and empirical, can also serve to better understand what makes cartels work or not work, in order to exacerbate cartel fragility and, by the same token, contribute to cartel deterrence. Alternatively, one could say that competition agencies are confronted with an information acquisition problem: firms know whether they collude; the agencies do not. The agencies could therefore try and devise so-called ‘revelation 41 See Harrington J. and Chen J. (2006): “Cartel Pricing Dynamics with Cost Variability and Endogenous Buyer Detection”, 24 International Journal of Industrial Organization 1185. In the same vein, it has been argued that, once a cartel has been successfully prosecuted by antitrust agencies and is facing civil suits for damages, its former members might still maintain (unilaterally, this time) supra-competitive prices in order to minimize the evaluation of the damages.


82

12/10/07

12:16

Page 82


mechanisms’ in order to induce the firms to report this information.42 This is the purpose of the leniency programmes that grant favourable treatment to cartel members who come forward with information that helps competition authorities to dismantle a cartel. Economic analysis can be used to study the design of these leniency programmes so as to improve their effectiveness; and indeed, there has recently been a body of work in this area, starting with the works of M. Motta and M. Polo,43 and of G. Spagnolo.44 Several messages stem from this emerging literature,45 for example: —it is good to restrict leniency to the first informant (provided that this informant brings sufficient evidence or information leading to successful prosecution); —it may be a good idea to keep offering leniency (to a first informant) even once an investigation is underway; —there is some complementarity between corporate and individual leniency programmes; —it may be desirable to go further and offer bounties (monetary rewards, such as a share in the fines imposed on the cartel members) to individual informants; —the stronger the ‘stick’ (criminal sanctions such as fines and jail sentences) the more effective the ‘carrot’ of leniency; a weaker stick calls for larger amnesty rates (or larger rewards). While this body of work is still very much in progress, there lies a strong hope for enhancing cartel detection and deterrence.

42 Building on the pioneering work of Maskin E. (“Nash Implementation and Welfare Optimality”, MIT 1977, published in 66 Review of Economic Studies 23 (1999)), the literature on Nash and Subgame Perfect implementation has confirmed the intuition that “a secret is no longer a secret when it is shared by several agents”. See Moore’s very nice survey, “Implementation, contracts, and renegotiation in environments with complete information”, in Laffont J., ed., Advances in Economic Theory, Sixth World Congress, Econometric Society Monographs, Vol. 1, 182 et seq (1992). 43 “Leniency Programs and Cartel Prosecution”, 2 International Journal of Industrial Organization 347 (2003). 44 “Optimal Leniency Programs”, Discussion Paper, Fondazione Enrico Mattei (2000). 45 Besides the papers already mentioned, see Aubert C., Kovacic W. and Rey P. (2006): “The Impact of Leniency and Whistleblowing Programs on Cartels”, 24 International Journal of Industrial Organization 1241; Harrington J., “Optimal Corporate Leniency Programs,” Johns Hopkins University (2005); Frezal S. (2006): “On Optimal Cartel Deterrence Policies”, 24 International Journal of Industrial Organization 1231; Rey P. (2003): “Toward a Theory of Competition Policy”, in Dewatripont M., Hansen L. and Turnovsky L., eds., Advances in Economics and Econometrics: Theory and Applications, Eighth World Congress, Cambridge University Press; Chen Z. and Rey P., “Optimal Design of Leniency Mechanisms”, mimeo, University of Toulouse.

(F) Ehlermann 07 Grout&Sonderegger

12/10/07

12:17

Page 83

III Paul A. Grout and Silvia Sonderegger 1 Structural Approaches to Cartel Detection

I. Introduction Cartels are agreements between firms with the aim of reducing the level of competition and increasing prices and profitability. In modern times these agreements are frequently verbal, and may therefore be very difficult to detect. Although past evidence of cartel detection provides some indication of where cartels may exist, it would be helpful to understand more fully the factors that lead to cartel formation, in order to identify the location of the cartels that have not yet been uncovered. Although many of the critical factors will depend on the specifics of the individuals involved and unique features of the case, there may be common economic factors that can help inform the process of deciding where to allocate effort in the detection process. This paper is concerned with this question. There are several ways to tackle the problem of cartel detection, and each has a role to play. It has become common to distinguish between structural and behavioural approaches, although the distinction is less clear than it initially may appear.2 While the former considers what characteristics of an industry increase the probability that cartels will arise or be sustainable, the latter asks what behaviour of the participants is likely to be associated with cartels. Competition authorities cannot investigate every possible market equally intensively and will therefore typically use both types of information. The aim of this paper is to highlight some of the main lessons that arise from structural approaches to cartel detection. The paper is divided into two sections. The first section provides a detailed non-technical overview of the theory literature; the second section discusses the lessons that can be drawn from case studies and statistical approaches. The primary reason for providing a detailed account of the existing theoretical literature is that this literature is able to sidestep a major problem that plagues both the behavioural and structural approaches, namely that of sample selection, and so is particularly helpful in this context.3 1 Department of Economics, University of Bristol, and CPMO. E-mails: P.A.Grout@ bristol.ac.uk; [email protected]. 2 See also Harrington J. (2006): “Detecting Cartels”, in Buccirossi P., ed., Handbook of Antitrust Economics, MIT Press, Massachusetts, for a discussion of this distinction. 3 Another reason is that non-technical surveys of the implications of this literature are not readily available in the law literature.


84

12/10/07

12:17

Page 84


The sample selection problem arises because of the limited evidence that is available to researchers. Using case studies and statistical approaches to identify structural or behavioural characteristics that are associated with cartels requires separation of cases of cartelization from cases where there are no cartels. But there is scant evidence on cartels that have not been discovered. What we have is a series of successful (and some unsuccessful) legal cases from which to draw our conclusions. Hence, we are almost certainly faced with a sample selection bias. To use a labour force analogy, the process is similar to one where a researcher tries to understand the causes of unemployment from a data set where most of the unemployed individuals have been wrongly classified as employed. To assess the problems that the researcher faces, it clearly matters a great deal whether the misclassification is random or not, as it will be much easier to draw reasonable conclusions from the evidence if misclassification is random. In the case of cartels, random discovery is almost certainly unrealistic, so when inferring evidence from case studies and statistical approaches we face a problem that is not present in theoretical analysis.

II. Theoretical insights 1. Introduction The theoretical literature on cartels is primarily concerned with the identification of the conditions that allow independent firms to implement jointly determined outputs and prices without resorting to explicit agreements. For this purpose, economists have borrowed from the theory of repeated games. The analogy with repeated games arises because cooperation requires single firms to depart from their myopically optimal actions. In the absence of legally binding agreements, cooperation in the present may therefore only be sustained by the threat of future punishment in the event of a deviation. Hence, in economic theory firms can maintain collusive prices only if they can credibly threaten to punish deviations by aggressive competition in the future. Clearly, this presupposes repeated interaction. At the most general level, the sustainability of collusion depends on the interplay between the size of the (short-lived) gain that can be obtained by deviating from the collusive agreement, and the (long-run) opportunity cost of the resulting reversion to non-cooperative interaction (punishment).4 In particular, collusion is sustainable only when the potential short-term gain 4 Friedman J. (1971): “A Noncooperative Equilibrium for Supergames”, 28 Review of Economic Studies 1, was the first to study cooperative equilibria sustained by the threat of future punishment (trigger strategy equilibria).


12/10/07

12:17

Page 85

Structural Approaches to Cartel Detection

85

from a deviation is outweighed by the long-run losses from a price war. This is more likely to be the case whenever the firms’ preference for present as opposed to future profits is not too high and the likelihood and speed of retaliation after a price cut are high. In what follows, we review the industry characteristics that the literature has singled out as being relevant when assessing the sustainability of tacit collusion within a market.

2. Number of firms A small number of market participants is generally thought to increase the likelihood of collusion5, for several reasons. First, having a small number of participants makes it easier to find a consensus regarding the modalities of the collusive agreement.6 Second, a small number of participants makes deviations from the collusive agreement easier to monitor.7 Third, as the number of firms increases, each firm gets a lower share of the market. This has the effect of both increasing the appeal of deviations and of decreasing the fear of punishments, hence making collusion harder to sustain.

3. Entry barriers If entry barriers are low, high prices in the present will attract new competitors in the future. This decreases the future gains from collusion, and therefore makes punishment less costly.

4. Capacity constraints and excess capacity The role played by capacity constraints on the feasibility of cartels is ambiguous. On one side, a capacity constrained firm has less to gain from undercutting its rivals, as it is able to accommodate only a fraction of the extra demand this would generate. On the other side, however, capacity constraints limit firms’ retaliatory power. This is because the strongest penalty that firms 5 See for instance Tirole J. (1988): The Theory of Industrial Organisation, MIT Press, Massachusetts, p. 242, and Bain J. (1956): Barriers to New Competition, Harvard University Press, Cambridge, Massachusetts. 6 The underlying idea is that a higher number of firms might increase the likelihood of asymmetries (e.g. of costs) among the participants. See also the section on cost asymmetries/quality differences below. 7 That is, market transparency may be greater when there are only a few firms operating in the market. See also the section on market transparency/private information below.


86

12/10/07

12:17

Page 86


can envisage is to produce at full capacity. So we have two effects that go in opposite directions. Brock and Scheinkman (1985)8 provide a good illustration of this ambiguous effect. Within a symmetric setting—where all the firms in a market have the same capacity constraint—these authors show that a non-monotonic relationship exists between cartel sustainability and the size of the capacity held by each firm. When industry capacity is sufficiently small, the deviationinhibiting effect of capacity constraints dominates. This implies that adding additional capacity makes collusion harder to sustain. When industry capacity is sufficiently large, on the other hand, the punishment-reducing effect of capacity constraints dominates. Adding additional capacity, therefore, makes collusion easier to sustain. Compte et al. (2002)9 analyze the effect that asymmetries in capacity constraints have upon the sustainability of cartels, focussing on the joint impact of asymmetry and aggregate capacity. Their findings share some similarities with those of Brock and Scheinkman (1985): the introduction of asymmetric capacities hinders collusion when aggregate capacity is limited, but favours it when aggregate capacity is large. This has interesting implications for assessing the impact that horizontal mergers may have upon the likelihood of collusion within a market. While the effect of capacity constraints upon the sustainability of collusion is ambiguous, the literature is less ambiguous on the need to hold excess capacity in order to support collusion in the presence of capacity constraints. Davidson and Deneckere (1990)10 explicitly model the firms’ choice of scale of operation, and find that excess capacity is carried in all collusive equilibria. This is because, unless excess capacity is held, deviations from the collusive scheme cannot be punished, and collusion cannot be sustained. In this setting, therefore, large capacities are a necessary condition for collusion to occur. In practice, the effect of capacity constraints may be difficult to assess. This is because periods of unexpectedly low demand typically result in depressed production, and therefore in lower capacity utilization. As a consequence, it 8 Brock W. and Scheinkman J. (1985): “Price Setting Supergames with Capacity Constraints”, 52 Review of Economic Studies 371. See also Abreu D. (1986): “Extremal Equilibria of Oligopolistic Supergames”, 39 Journal of Economic Theory 191, and Lambson V. (1987): “Optimal Penal Codes in Price-Setting Supergames with Capacity Constraints”, 54 Review of Economic Studies 385. 9 Compte O., Jenny F. and Rey P. (2002): “Capacity Constraints, Mergers and Collusion”, 46 European Economic Review 1. See also Lambson V. (1995): “Optimal Penal Codes in Nearly Symmetric Bertrand Supergames with Capacity Constraints”, 14 Journal of Mathematical Economics 1. 10 Davidson C. and Deneckere R. (1990): “Excess Capacity and Collusion”, 31 International Economic Review 521. A similar idea is also explored in Rotemberg J. and Saloner G. (1989): “The Cyclical Behaviour of Strategic Inventories”, 104 Quarterly Journal of Economics 73, Benoît J.-P. and Krishna V. (1987): “Dynamic Duopoly: Prices and Quantities”, 54 Review of Economic Studies 23 and Osborne M. and Pitchik C. (1987): “Cartels, Profits and Excess Capacity”, 28 International Economic Review 413.


12/10/07

12:17

Page 87


87

may be hard to disentangle the effect of demand decline from that of excess capacity in favouring cartel formation.11

5. Demand/demand stability In a classic article, Rotemberg and Saloner (1986)12 explore the response of oligopolies to fluctuations in the demand for their product. Taking a simple but illustrative view of the business cycle, they assume that the level of market demand is determined in an independent fashion each period, so that the expected level of future demand is independent of the current demand level. This ensures that the expected cost of punishment is independent of the present state of demand. Today’s demand level, however, does affect the short-term incentive to cheat, since a price-cut is more attractive when demand is high. The implication is that, when demand is at a peak, collusion is harder to sustain. Haltiwanger and Harrington (1991)13 relax the assumption of demand independence across periods, and introduce a more realistic model of the business cycle, where demand movements are cyclical over time. They show that the most difficult point of the cycle for firms to collude is when demand is at its highest. The intuition is that, when demand is at a peak, short-term gains from deviation are maximal, while the potential cost of punishment is at a minimum. Bagwell and Staiger (1997)14 extend the analysis to persistent business cycles—situations where demand movements are stochastic and persistent. They find that collusion can be more easily sustained if the expected duration of booms is sufficiently long, and that of recessions is sufficiently short. Moreover, transitory shocks make collusion harder to sustain, independently of whether the market is in a boom or a recession phase. Staiger and Wolak (1992)15 explore the response of collusive prices to changing demand conditions when firms operate under capacity constraints. In their setting, periods of unexpectedly low demand result in the emergence of excess capacity; as discussed above, this may destabilize the collusive agreement. Based on the above results, we draw the following conclusions. First, demand instability hinders collusion. This is because, with respect to the case 11 See Staiger R. and Wolak F. (1992): “Collusive Pricing with Capacity Constraints in the Presence of Demand Uncertainty”, 23 RAND Journal of Economics 203. 12 Rotemberg J. and Saloner G. (1986): “A Supergame-Theoretic Model of Business Cycles and Price Wars During Booms”, 76 American Economic Review 390. 13 Haltiwanger J. and Harrington J. (1991): “The Impact of Cyclical Demand Movements on Collusive Behaviour”, 22 RAND Journal of Economics 89. 14 Bagwell K. and Staiger R. (1997): “Collusion over the Business Cycle”, 28 RAND Journal of Economics 82. 15 Staiger R. and Wolak F. (1992): “Collusive Pricing with Capacity Constraints in the Presence of Demand Uncertainty”, 23 RAND Journal of Economics 203.


88

12/10/07

12:17

Page 88


where demand is stable at an average level, demand fluctuations tighten the requirements that ensure the sustainability of collusion. Second, demand growth facilitates collusion. This is because high future profits (generated by high demand) increase the opportunity cost of cheating in the present and therefore make firms more likely to collude. However, it should be noted that this last prediction is conditional on entry barriers being sufficiently high. In the absence of barriers to entry, demand growth will trigger new entry in the market. This will, at least partially, counteract the collusion-enhancing effect of demand growth.

6. Frequency of interaction and price adjustments Frequent interactions/price adjustments facilitate collusion by shortening the time of reaction to deviations from the collusive agreement. If the amount of time that must elapse before any deviation can be punished is significant, the gains from deviation can be enjoyed for longer. This makes deviation more attractive. Moreover, a longer reaction time pushes retaliation further in the future. When firms discount the future, thus makes punishment less frightening. Snyder (1996)16 argues that the impact of frequency of interaction upon firms’ ability to collude may be particularly important in the presence of large buyers.17 This is because large buyers can strategically concentrate their orders across time in order to make firms’ interactions less frequent and therefore to make collusion harder to sustain. Methods to make deviations from a collusive agreement more readily observable and more costly to implement include the imposition by manufacturers of resale price maintenance on wholesale or retail resellers of their products18 or the use of “most favoured customer” clauses19 (requiring that the seller charge a buyer a price no higher than what it charges any other buyer).

7. Market transparency/private information The importance of market transparency in determining the sustainability of tacit collusion was first highlighted by Stigler (1964)20 and subsequently 16 Snyder C. (1996): “A Dynamic Theory of Countervailing Power”, 27 RAND Journal of Economics 747. 17 See also the section on buyer power, below. 18 See e.g. Telser L. (1960): “Why Should Manufacturers Want Fair Trade?”, 3 Journal of Law and Economics 86. 19 See Cooper T. (1986): “Most Favoured Customer Pricing and Tacit Collusion”, 17 RAND Journal of Economics 377. 20 Stigler G. (1964): “A Theory of Oligopoly”, 72 The Journal of Political Economy 44.


12/10/07

12:17

Page 89


89

formally established by Green and Porter (1984)21 and Abreu, Pierce and Stacchetti (1986).22 The logic of the argument can be seen as follows. Suppose that firms are unable to observe demand conditions, and to monitor their rivals’ sales and prices. Moreover, assume that demand randomly fluctuates from period to period. In this setting, firms who register low sales are unable to establish whether this is the consequence of adverse overall demand conditions, or of secret undercutting by rival firms. How should they react? Suppose that, after a period of low sales, firms do not modify their behaviour, and stick to the collusive price. This would eliminate any punishment for undercutting, and would therefore encourage firms to deviate from the collusive agreement. It follows that collusion can be sustained only if firms react to low sales by launching a price war. In this setting, this implies that collusion can only be sustained in environments where price wars erupt whenever demand is low. With respect to the case where demand is observable, this reduces the expected future benefits from collusion, and makes undercutting more tempting. Thus, the firms’ inability to distinguish between favourable and unfavourable demand conditions undermines their ability to collude. Importantly, the argument advanced by Green and Porter (1984) relies on the unpredictability of demand, and would not hold in markets where demand is deterministic. Similar to demand unpredictability and lack of transparency, the existence of private information also makes collusion harder to sustain. Roberts (1985),23 Cramton and Palfrey (1990),24 Kihlstrom and Vives (1992),25 Athey and Bagwell (2001)26 and Athey, Bagwell and Sanchirico (2004) 27 concentrate on the case where firms possess private information about their costs. The idea is that, at any given point in time, some firms may have high costs, while others enjoy low costs, due for instance to variations in local conditions, labour relations, inventory management and so on. In this setting, joint profit-maximization prescribes that market shares should be inversely related to firms’ costs. Because this requires them to 21 Green E. and Porter R. (1984): “Noncooperative Collusion under Imperfect Price Information”, 52 Econometrica 87. 22 Abreu D., Pearce D. and Stacchetti E. (1986): “Optimal Cartel Equilibria with Imperfect Monitoring”, 39 Journal of Economic Theory 251. See also Clarke R. (1983): “Collusion and The Incentives for Risk-Sharing”, 14 Bell Journal of Economics 383, for an early discussion of the issues. 23 Roberts K. (1985): “Cartel Behaviour and Adverse Selection”, 33 Journal of Industrial Economics 401. 24 Cramton P. and Palfrey T. (1990): “Cartel Enforcement with Uncertainty About Costs”, 31 International Economic Review 17. 25 Kihlstrom R. and Vives X. (1992): “Collusion by Asymmetrically Informed Firms”, 1 Journal of Economics and Management Strategy 271. 26 Athey S. and Bagwell K. (2001): “Optimal Collusion with Private Information”, 32 RAND Journal of Economics 428. 27 Athey S., Bagwell K. and Sanchirico C. (2004): “Collusion and Price Rigidity”, 71 Review of Economic Studies 317.


90

12/10/07

12:17

Page 90


relinquish market share, high-cost firms have strong incentives to misrepresent their true cost realizations. Truthful communication may therefore be possible only if higher-cost firms are assured of side-payments, or of some other form of future benefits, by lower-cost fellow cartel-members. In turn, this decreases the low-cost firms’ incentive to abide by the collusive agreement. Thus, a trade-off exists between joint profit maximization and cartel enforceability. This reduces the profitability of cartels. Compte (1998),28 Kandori and Matsushima (1998)29 and Verboven (1998)30 study the informational asymmetries that arise when firms hold diverse and imperfect information about each other’s actions. This may for instance be the case when competition is localized, and different firms compete directly with different sets of rivals. In such markets, cheating by one firm only affects a subset of the other firms in the industry. If punishing a deviating firm is costly (for instance because it involves a price war that penalizes all firms in the market), this creates the following problem: a firm that suspects cheating by one of its competitors may consider concealing this information, in order to avoid having to enforce the costly punishment. With respect to a setting where all information is public, the presence of private information therefore imposes an additional constraint, arising from the trade-off that exists between the severity of the punishment that can be imposed after a deviation and the likelihood that deviations will be discovered. This makes collusion harder to sustain. Compte (1998) and KandoriMatsushima (1998) show that under these conditions communication may alleviate the problem. Thus, communication-enhancing devices such as trade unions may promote collusion. This is in contrast with McCutcheon, (1997),31 who emphasizes how communication may facilitate renegotiation away from punishments that are costly to implement. When everybody is hurt by the punishment, firms may decide to let bygones be bygones, and to restart the collusive agreement. In turn, this makes collusion infeasible: firms know that they will not be punished for cheating, and will therefore find it optimal to cheat. In that case, communication may be detrimental to collusion, by making renegotiation more likely.

28 Compte O. (1998): “Communication in Repeated Games with Imperfect Private Monitoring”, 66 Econometrica 597. 29 Kandori M. and Matsushima H. (1998): “Private Observation, Communication and Collusion”, 66 Econometrica 627. 30 Verboven F. (1998): “Localized Competition, Multimarket Operation and Collusive Behaviour”, 39 International Economic Review 371. 31 McCutcheon B. (1997): “Do Meetings in Smoke-Filled Rooms Facilitate Collusion?”, 105 Journal of Political Economy 330.


12/10/07

12:17

Page 91


91

8. Cost asymmetries/quality differences Cost asymmetries and differences in the quality of the products sold by different firms make collusion harder to sustain.32 In asymmetric settings, lowcost (respectively, high quality) firms are hard to discipline. This is because the harm that their high-cost (respectively, low quality) rivals can inflict upon them by triggering a price war is limited. Moreover, low-cost/high quality firms have stronger incentives to deviate than their high-cost/low quality counterparts. These stronger incentives arise from higher price-cost margins in the case of cost differentials, and from lower elasticity of demand in the case of quality differences. To compensate, collusive agreements should assign larger market shares to low-cost/high quality firms, in order to decrease the (relative) gains that such firms can obtain from deviation. This however adversely affects the highcost/low quality firms’ incentives to participate in the collusive agreement, creating a trade-off between incentive provision to low-cost/high quality firms and to their high-cost/low quality rivals. This limits the range of market share allocations that are consistent with collusion, and therefore makes collusion harder to achieve.33 An interesting application of this general point concerns the impact of innovation upon collusion. Because innovation fosters asymmetries in costs and/or qualities, we expect collusion to be less likely in innovative markets.34

9. Horizontal product differentiation The effect of horizontal product differentiation upon the sustainability of collusion is ambiguous. When products are horizontally differentiated, different consumers rank equally priced products differently; this is because goods possess characteristics that appeal to different types of consumers. With respect to the situation where all goods are perfect substitutes, this decreases the elasticity of demand faced by any single firm; for any given price differential, consumers are more reluctant to abandon their favourite variety. On the 32 See Hackner J. (1994): “Collusive Pricing in Markets for Vertically Differentiated Products,” 12 International Journal of Industrial Organization 255, for an analysis of the sustainability of collusive prices in markets where goods are vertically differentiated. 33 An exception to this prediction can be found in Eswaran M. (1997): “Cartel Unity over the Business Cycle”, 30 Canadian Journal of Economics 644, where (within the context of a model of business cycles) some members of the cartel may go bankrupt during periods of slack demand. In this case, a low cost producer may help collusion by curtailing its own output in periods of slack demand, thereby ensuring the viability of the more inefficient cartel members. 34 This point is made more formally in Ivaldi M. et al. (2003): The Economics of Tacit Collusion, Final Report for DG Competition, European Commission.


92

12/10/07

12:17

Page 92


one hand, this results in firms finding deviations less attractive; this is because the extra demand that firms can capture by slightly undercutting their rivals is limited. By the same token, however, a lower elasticity of demand also restricts the firms’ retaliatory power, by limiting the harm that can be inflicted upon other firms via a price war. The net result of these opposing forces depends upon the exact nature of the case under scrutiny.35 Raith (1996)36 studies the impact of horizontal product differentiation upon the sustainability of collusion when firms are unable to observe the actions of their rivals, but can nonetheless draw inferences from their own demand. In this setting, differentiation decreases the correlation between the demands faced by different firms. This makes deviations harder to detect, and makes collusion harder to sustain. Hence, when markets are not transparent, horizontal product differentiation makes collusion harder to sustain. Symeonidis (2002)37 considers the case where firms can sell more than one variety, and finds that collusion becomes harder as the number of varieties sold by each firm increases. This is because the increase in profit associated with a rise in the number of varieties sold by a firm is higher when the firm is a monopolist than when it has a limited market share. It follows that selling a higher number of varieties increases the benefit from deviation more than it increases the (opportunity) cost of punishment.

10. Multi-market contact In the words of Edwards (1955):38 “When firms encounter each other in several markets, the multiplicity of their contact may blunt the edge of their competition”. Bernheim and Whinston (1990)39 formalize this intuition by precisely identifying the conditions under which this holds. Intuitively, multi-market contact acts as an expansion of the market where the firms compete. This has an ambiguous effect: on one hand it magnifies the gains of deviation, while on the other it increases the strength of the resulting punishments. In a perfectly

35 See for instance Chang M. (1991): “The Effects of Product Differentiation on Collusive Pricing”, 9 International Journal of Industrial Organization 453, and Ross T. (1992): “Cartel Stability and Product Differentiation”, 10 International Journal of Industrial Organization 1. 36 Raith M. (1996): Product Differentiation, Uncertainty and the Stability of Collusion, London School of Economics-STICERD Discussion Paper. 37 Symeonidis G. (2002): “Cartel Stability and Multiproduct Firms”, 20 International Journal of Industrial Organization 339. 38 Edwards C. (1955): “Conglomerate Bigness as a Source of Power”, Business Concentration and Price Policy 331, National Bureau of Economic Research Report, Princeton, NJ. 39 Bernheim D. and Whinston M. (1990): “Multimarket Contact and Collusive Behaviour”, 21 RAND Journal of Economics 1.


12/10/07

12:17

Page 93


93

symmetric setting—where identical firms with identical, constant-returns-to scale technologies meet in identical markets—these opposing forces exactly nullify each other. In that case, multi-market contact has no impact upon the sustainability of collusion. In contrast, when markets and/or firms differ, or when technology exhibits decreasing returns to scale, multi-market contact may foster collusion. This is the case because the presence of several markets softens/offsets the anti-collusive incentives that may be generated within single markets. To illustrate this point, consider a situation where firms have different costs of production, which are negatively correlated across markets. In that case, multi-market contact facilitates collusion, by allowing each firm to hold higher market shares in those markets where it is more efficient. This allocation ensures that the greatest fraction of the extra demand that firms would generate by undercutting would come from the market where they have higher costs, and at the same time the greatest fraction of the profits lost in the event of a price war would come from the market where firms have lower costs. The first effect decreases the firms’ gains from deviation, while the second effect increases the costs of punishment. Overall, therefore, multi-market contact relaxes the incentives that are necessary to sustain collusion.

11. Cross-ownership There are many cases in which firms acquire their rivals’ stocks as passive investments that give them a share in the rivals’ profits but not in the rivals’ decision-making. Malueg (1992) and D. Gilo et al. (2006)40 study these passive investment/cross-ownership patterns, and show that they may have two opposing effects upon the firms’ inclinations to collude. On the one hand, cross ownership weakens the incentive of firms to deviate; this is because firms internalize part of the losses that they inflict on rivals when they deviate. On the other hand, cross ownership may also soften the firms’ incentive to impose a harsh punishment, following a breakdown of the collusive scheme. This makes deviation more tempting. We conclude that cross-ownership agreements have an ambiguous effect upon the sustainability of collusion.

12. Buyer power Large, powerful buyers make collusion harder to sustain. Because buyers benefit from price wars between sellers, large buyers will utilize their power 40 Malueg D. (1992): “Collusive Behaviour and Partial Ownership of Rivals”, 10 International Journal of Industrial Organization 27; Gilo D. et al (2006): “Partial Cross Ownership and Tacit Collusion”, 37 RAND Journal of Economics 81.


94

12/10/07

12:17

Page 94


strategically, in an attempt to break any collusive agreement that may exist among the sellers. This may for instance be achieved by concentrating orders across time, in order to make the sellers’ interactions less frequent (as in Snyder (1996)41) or by shading current demand, in the hope that this will lead sellers to believe that a secret price cut has occurred (as in Compte (2000) 42).

III. Case studies and statistical analysis 1. Some primary messages from case studies In this section we identify some of the primary lessons that emerge for EU case studies and statistical analysis on European data. We also provide a comparison between the cartel cases that have recently been identified in the EU and the US. This may allow us to uncover any pattern that could be useful in identifying the industries where cartels are most likely to arise. We address this question using EC cases from 1990 to the present and US Department of Justice cases of horizontal price fixing from 1994 to the present. For each jurisdiction, we construct an index that gives the number of cartel cases discovered in each industry during the periods noted above. We define industries by the Standard Industrial Classification (SIC) three-digit level. There are 204 three-digit industries. Having allocated the cartels to SIC three-digit classification we calculate the correlation between US and EU cartels. The correlation using the raw numbers per industry is 0.678. This is surprisingly high. Hence, despite the fact that less than 20% of the cases are common between these two jurisdictions, we find there is a strong positive correlation between the data sets. This high correlation identified between the EU and the US cases is interesting, but it is not obvious what it implies. On the one hand, this correlation may be a sign that similar industry characteristics in the EU and US are associated with cartelization. In this case, the evidence would imply that industry characteristics are important determinants of the likelihood of collusion. On the other hand, one should also take into account the possibility that antitrust authorities in the EU and the US may not be acting independently. If one country/jurisdiction is successful in finding cartels in a particular market then others may focus extra effort in that area. This will increase the correlation between the cases identified in different jurisdictions. Similarly, even though the cases are not common, discovery of one cartel may provide information 41 Snyder C. (1996): “A Dynamic Theory of Countervailing Power”, 27 RAND Journal of Economics 747. 42 Compte O. (2000): Secret Price Cutting and Strategic Buyers: An Illustration of the Divide and Conquer Tactic, CERAS, Ecole Nationale des Ponts et Chausées, Paris, mimeo.


12/10/07

12:17

Page 95


95

that leads to discovery of others in the same market (but in other countries). In this case, the common location of cartels across the industry is not only a sign of specific factors in that industry but also the result of luck that spills over to other jurisdictions. Also, even if there are common, industry-based factors that affect cartelization in the EU and the US, these factors may not necessarily be economic. However, we believe that this last possibility may however be quite unlikely. One of the most prominent features that emerges from the case studies is that most cases relate to markets with relatively homogeneous products. Intuitively, the scope for brand differentiation in industries where cartels arise—such as shipping, basic chemicals, currency exchange, tubes, carbonless paper, petrochemicals, and electrodes—appears to be limited. This confirms the theoretical view that horizontal product differentiation renders collusion harder to sustain by limiting the cartel members’ retaliatory power and by making it harder to reach a consensus over the modalities of the collusive agreement. The scope for vertical product differentiation (quality differences) and/or cost asymmetries also appears to be rather limited. The finding that, within the industries where cartels have been discovered in the EU, the mean value of R&D expenditure as a proportion of total turnover is much lower than the mean value across all industries corroborates this observation. For UK industries this data is given in Table 1. Indeed, the low level of R&D is quite striking. So EU evidence appears to be consistent with the theoretical view that cost/quality asymmetries hinder collusion. Table 1 Mean R&D expenditure as a proportion of total turnover Industry Mean of all industries Mean of case study industries Manufacture of basic chemicals Sea and coastal water transport Manufacture of other non-metallic mineral products Manufacture of tubes Manufacture of articles of concrete, plaster and cement Manufacture of pulp, paper and paperboard Production, processing and preserving of meat and meat products Manufacture of electrical equipment not elsewhere classified Manuf. of pharmaceuticals, medicinal chemicals and botanical products Manuf. of non-refractory ceramic goods other than for construction

0.2905 0.0516 0.0359 0.0029 0.0462 0.0247 0.0124 0.0116 0.0233 0.1153 0.2151 0.0295

A second common feature of the case studies is that the markets where cartels have operated are very concentrated. For instance, in the Citric Acid 43 43

Citric Acid, Case COMP 2002/747/EC, Commission decision published in OJ L 239 [2002].


96

12/10/07

12:17

Page 96


case, the cartel members (five firms in total) controlled 56–61% of the world market, and 60–70% of the European market. In the area of shipping, the cartels accounted for the majority—indeed sometimes 100%—of the market. This common observation is corroborated when we look at C3, C5 and the mean number of firms in the case study industries—see Table 2 (again using UK data).44 The aggregate mean value for the cartel case study group is significantly larger than the mean for all industries. So on the whole, the case studies evidence confirms the theoretical prediction that the likelihood of collusion is inversely correlated with the number of firms and the degree of concentration in the market. However, there needs to be a note of caution here in that the EU cases almost by definition are dealing with large firms and so, although this does not imply that the companies have high market shares, it would be surprising if they did not. Hence one has to be careful in using the case studies to indicate that cartels arise where one has large firms with high market shares. Table 2 Number of firms Industry

No. firms

Mean of all industries Mean of case study industries Manufacture of basic chemicals Sea and coastal water transport Manufacture of other non-metallic mineral products Manufacture of tubes Manufacture of articles of concrete, plaster and cement Manufacture of pulp, paper and paperboard Production, processing and preserving of meat and meat products Manufacture of electrical equipment not elsewhere classified Manuf. of pharmaceuticals, medicinal chemicals and botanical products Manuf. of non-refractory ceramic goods other than for construction

C3

C5

7544 843 1262 1033 399 379 1086 476

39.86 48.05 57.90 53.51 42.37 48.45 62.52 36.75

46.55 56.82 68.04 62.73 49.82 58.07 69.59 41.51

1173 2405

36.62 31.09

47.08 40.36

476

59.26

71.69

844

52.08

59.28

A third feature of many case studies is that the role of declining demand is surprisingly common. Cartels appear to be associated both with sudden adverse demand shocks and long, slower declines. Examples of the former include the Ferry Operators 45 case, French Beef 46 and German Banks.47 44 C3 and C5 stand for 3-firm and 5-firm concentration ratios, respectively. In general, the N-firm concentration ratio is the percentage of market output generated by the N largest firms in an industry. 45 Ferry Operators—Currency Surcharges, OJ L 26 [1997]. 46 French Beef, OJ L 209 [2003]. 47 Bank Charges for Exchanging Euro-Zone Currencies, OJ L 15 [2003].


12/10/07

12:17

Page 97


97

Examples of the impact of prolonged adverse demand conditions are found in Petrochemicals ,48 Seamless Steel Tubes,49 Graphite Electrodes,50 and French-West African Ship-Owners Committees.51 In all these cases, the protracted nature of the adverse demand conditions resulted in the emergence of excess capacity, industry restructuring and exit, which in turn favoured cartel formation. The notion that sudden negative shocks and prolonged adverse demand conditions are connected with cartel formation is certainly intuitive as well as being grounded in theory. Clearly, if a shock hits all firms in an industry then each firm has to make a quick decision as to how to respond in a situation when each firm is aware that others are in the same position. The drive to find out how others are responding and to co-ordinate may be extremely strong, particularly if the industry is normally relatively stable and the large players in the market are well known to each other. Furthermore, intuitively, when demand is temporarily low the gains from aggressive behaviour in the present are minimized, while the benefits from cooperation in the future are maximized. Therefore, it is not surprising that sudden negative shocks may precipitate collusion. Similarly, prolonged adverse demand conditions may encourage firms to seek to limit the damage to their businesses. On the surface, this may appear to contradict the theoretical prediction that falling demand should hinder, rather than facilitate collusion. This apparent dichotomy can nevertheless be reconciled by noticing that steadily falling demand generates excess capacity; in turn, excess capacity might be conducive of collusion. Moreover, prolonged crises often result in exit; this decreases the number of market participants, and makes cooperation easier to achieve.

2. Statistical analysis A third way to identify which industry characteristics are favourable to cartelisation is to test whether specific characteristics are good statistical predictors of the presence of cartels. To undertake an exhaustive study of this type is a major exercise, which the authors have not conducted to date. However, Grout and Sonderegger (2005)52 provide the results of a preliminary investigation of this type. This investigation was sufficiently fruitful to be informative in its own right and to indicate that such approaches are effective. In this 48

Petrochemicals, OJ L 239 [1994]. Seamless Steel Tubes, OJ L 382 [2003]. 50 Graphite Electrodes, OJ L 100 [2001]. 51 French-West African Ship-Owners’ Committees, OJ L 134 [1992]. 52 Grout P. and Sonderegger S. (2005): Predicting Cartels, Office of Fair Trading, Economics Discussion Paper, March 2005. To download, see the web address given at page 9 of this volume, footnote 3. 49


98

12/10/07

12:17

Page 98


section we consider some of the difficulties with this type of analysis, and provide a brief summary of the investigation and its conclusions. The approach has two stages. First, the statistical model uses the economic independent variables to explain the level of cartelization in any industry (using successful cases as the measure of where cartels are located across industries). Second, given the specific values of the independent economic values in any industry, the model can then be used to predict the probability that a cartel exists in that industry. One is then able to rank the industries that do not have cartels according to the likelihood that there is a cartel that has yet to be discovered. There are some clear difficulties with this approach. One problem that arises was discussed in the introduction, namely that the evidence we have of where cartels are located is based on ‘failed’ cartels—implying that there may be a sample selection problem. This, of course, is an issue for all empirical analysis in this field. Another problem is that cartels arise in markets. Yet, almost universally, governments collect economic data according to industries. So there has to be a leap of faith to use industry data to infer what is happening in markets. Furthermore, the process has to be conducted at a disaggregated level. This is partly done to provide sufficient cross-section data points, but it is also necessary if predictions are to bring significant additional insight to that which can be gleaned directly from case studies and simple ‘eye-balling’ of the data. For example, identifying, say, six specific industries where cartels may be particularly prevalent is more informative if these six are drawn from a classification with over two hundred categories than if they are drawn from, say, ten categories. As indicated, industries are defined for the purposes here by the Standard Industrial Classification (SIC) three-digit level. There are 204 three-digit industries, although, because of data restrictions, slightly less have been used in the study. A high level of disaggregation introduces more measurement error and random industry factors. Moreover, there is an even more fundamental problem, namely that it is not obvious that, even with perfect data, economic factors ‘work’ at this level of disaggregation. That is, the ‘economic’ differences between, say, manufacturing and transport may be significant in determining cartel formation but it is not obvious whether the differences in manufacturing at a very disaggregated level (e.g., separating manufacture of knitted and crocheted fabrics from manufacture of knitted and crocheted articles, as we do in the analysis) will be relevant. This may be because, at this level of disaggregation, any mismatch between markets and industry classification is marked, and the differences between industries may become too fine to significantly influence cartel formation, and hence identification. There are several clear messages that arise from the series of models investigated in Sonderegger and Grout (2005). The main message from the analysis is that the approach does work at this level. That is, despite the data problems and the level of disaggregation, we are able to find strong, significant patterns


12/10/07

12:17

Page 99


99

in the data. There were four factors that appeared to be particularly important in the statistical results.53 First, variability in growth was shown to have a negative impact on cartels. This was an extremely robust result holding in all the models considered. Second, growth in turnover had a positive impact on cartel formation, and was again extremely robust. Third, the cost disadvantage ratio (a measure of economies of scale) was the sole entry barrier that was significant in explaining cartel presence, implying that more significant economies of scale increase the likelihood of cartels. Last, industry concentration—in the form of C3 (the share of the three largest firms in the industry)—was also positively associated with cartels. The regressions typically explained somewhere between 14% to 24% of variability in the data on the location of cartels. That is, the economic variables explained about one fifth to one sixth of the variability in the cartel data. Given the measurement error problems in the data, this is not an unreasonable figure. As indicated above, the model can be used not only to identify critical factors, but also to rank industries according to likelihood of cartelization. Tables 3 and 4 provide some of these rankings. The specific model that has been used for these tables is as follows. Each industry was given a value of 1 if the European Commission has discovered cartels within that industry classification, and a value of 0 if no cartels have been found. Thirty-six industry classifications had cartels. This was used as the incidence of cartelization. A (logit) model was then used on disaggregated UK data to predict a probability of cartelization for each industry. The top thirty industries, ranked by probability, are given in Table 3. Table 4 gives the top thirty industries where no cartels have been discovered. The benefits of even such a preliminary approach become apparent when one compares the predictions with the raw cartel data. The allocation of EC and US cartels to industry classification displays the well-known concentration on manufacturing and transport. Using this evidence as the basis for future investigation would imply a focus on manufacturing and transport sectors. The econometric approach allows a more detailed assessment. The statistical predictions of the most likely locations for cartels does indeed include several manufacturing industries where cartels have not yet been found (e.g., manufacture of aircraft and spacecraft, manufacture of paper and paperboard, manufacture of weapons) but, significantly, it also throws up other industries as being prime candidates that would not have arisen from a simple non-econometric interpretation of the data. For example, secondary schools, auditing and tax consultancy, architecture and related consultancy, and telecommunications all come out extremely high on the list. 53 Formally, these factors were particularly important, i.e., they were significant at 1% or 5% in all approaches.


100

12/10/07

12:17

Page 100


There are two particularly interesting predictions delivered by this preliminary investigation that suggest that statistical structural approaches can be informative. The first example concerns secondary schooling. Because no cartels had been discovered to date in this sector, this industry sector was entered as a 0 in the analysis. Moreover, looking at past case studies, this sector appeared an unlikely candidate for cartels. However, since the analysis was conducted, a large number of the leading private secondary schools in the UK were found guilty of collusion. The fact that schools were not present in the cartel data and yet were thrown out as highly likely to contain cartels by the statistical analysis (i.e., in the top 20% of industries where cartels have not been discovered) highlights how useful these disaggregated econometric approaches can be. The second illustration concerns shipping and basic chemicals. These are the industry classifications with most cartels in the EU. In our data set, there were 10 in shipping and 11 in basic chemicals, while no other industry classification had more than 4 (with only two having 4). The data that were used for the predictions given in Table 3 did not use the fact that these two industries had far more cartels than other industries (the analysis only distinguished between industries in which at least one cartel had been found, and industries where no cartels had been found). In spite of this, these industries appear high on the list, and are therefore indicated as being among the most likely to contain cartels. Overall, the general message of this preliminary study is that structural industry features do help explain the location of cartels. Clearly, if there were no structural relationships then the results would be less robust. Table 3 All Industries 1 2 3 4 5 6 7 8 9 10 11 12

13

Probability of at least one cartel.

Building of complete constructions or parts thereof; civil engineering Telecommunications Activities of other transport agencies Manufacture of cement, lime and plaster Scheduled air transport Manufacture of basic chemicals Manufacture of pharmaceuticals, medicinal chemicals and botanical products Manufacture of motor vehicles Software consultancy and supply Manufacture of aircraft and spacecraft Manufacture of grain mill products, starches and starch products Legal, accounting, bookkeeping and auditing activities; tax consultancy; market research and public opinion polling; business and management consultancy Manufacture of other food products

0.89 0.84 0.80 0.77 0.73 0.72 0.71 0.68 0.68 0.65 0.61

0.55 0.52


14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

12/10/07

12:17

Page 101


101

Cargo handling and storage Activities of travel agencies and tour operators; tourist assistance activities Publishing Manufacture of railway and tramway locomotives and rolling stock Other land transport Manufacture of tubes Recycling of metal waste and scrap Manufacture of articles of paper and paperboard Manufacture of basic iron and steel and of ferroalloys Manufacture of weapons and ammunition Manufacture of beverages Processing and preserving of fruit and vegetables Manufacture of motorcycles and bicycles Quarrying of sand and clay Building installation Sea and coastal water transport Non-scheduled air transport

0.50 0.46 0.44 0.44 0.43 0.41 0.40 0.40 0.39 0.39 0.39 0.38 0.38 0.37 0.36 0.35 0.34

Table 4 Ranking of industries where no cartel(s) have yet been discovered 1 2 3 4

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23


Telecommunications Manufacture of aircraft and spacecraft Manufacture of grain mill products, starches and starch products Legal, accounting, bookkeeping and auditing activities; tax consultancy; market research and public opinion polling; business and management consultancy Cargo handling and storage Activities of travel agencies and tour operators; tourist assistance activities Publishing Manufacture of railway and tramway locomotives and rolling stock Other land transport Recycling of metal waste and scrap Manufacture of articles of paper and paperboard Manufacture of weapons and ammunition Radio and television activities Processing and preserving of fruit and vegetables Manufacture of motorcycles and bicycles Quarrying of sand and clay Manufacture of prepared animal feeds Non-scheduled air transport Manufacture of domestic appliances not elsewhere classified Mining and agglomeration of hard coal Manufacture of glass and glass products Other computer related activities Architectural and engineering activities and related technical consultancy

0.84 0.65 0.61

0.55 0.50 0.46 0.44 0.44 0.43 0.40 0.40 0.39 0.39 0.38 0.38 0.37 0.35 0.34 0.34 0.34 0.33 0.32 0.30


102

12/10/07

12:17

Page 102


Table 4 (cont.) Ranking of industries where no cartel(s) have yet been discovered 24 25 26 27 28 29 30


Manufacture of cutlery, tools and general hardware Other recreational activities Manufacture of other chemical products Printing and service activities related to printing Secondary education Manufacture of bricks, tiles and construction products, in baked clay Postal and courier activities

0.29 0.28 0.28 0.27 0.27 0.27 0.25

IV. Concluding remarks In this paper we have considered whether structural approaches can contribute to cartel detection. The approach suggests that industry characteristics may be an important factor in encouraging cartels, and that, by identifying those factors that are particularly likely to lead to cartelization, antitrust authorities may be aided in the task of cartel detection. We have provided a detailed, non-technical overview of the theoretical literature on the subject. Although this literature delivers ambiguous predictions at times, it is nonetheless useful for discriminating between different industry characteristics, and distinguishing between those that are unlikely to affect cartelization and those that are likely to have an impact on the ability of firms to collude. We have also reviewed a series of case studies on cartels, drawn from EC Commission cases from 1990 to the present. A number of patterns emerge from these cases. For instance, most of the cartels appeared in industries where product differentiation and/or cost differences between producers are rather limited. A second element of similarity between the cases concerned the high concentration of the markets in which cartels were found to operate. Finally, declining demand (both in terms of sudden shocks and protracted adverse demand conditions), coupled with the emergence of excess capacity, appears to have been at the origin of several of the cartels in our data. These findings are in line with theoretical predictions, and provide corroborative evidence that industry characteristics may indeed play an important role in detecting cartels. An intriguing discovery was that EC and US cases have a rather high degree of correlation, in spite of the fact that less than 20% of the cases were common between the two jurisdictions. Although there may be more than one explanation for this finding, this may also provide evidence in favour of a structural approach to cartel detection.


12/10/07

12:17

Page 103


103

Finally, we have discussed some of the statistical evidence on the relationship between industry characteristics and cartelization that was obtained in a previous study by the authors (Grout and Sonderegger (20005)). In spite of several potential problems, the analysis delivered a number of robust predictions, which appeared consistent with theoretical findings. The analysis also allowed us to deliver a series of predictions aimed at identifying those industries where no cartels have yet been found, but where they are likely to exist. Overall, therefore, it is clear that economics has a role to play in cartel detection, and the paper shows that structural approaches are useful. However, as discussed in the paper, these types of approaches also carry some problems, and should therefore be seen as only one component in the broader use of economics in cartel detection.


12/10/07

12:17

Page 104

(G) Ehlermann 07 Whish

12/10/07

12:17

Page 105

IV Richard Whish* Oligopoly Theory and Economic Evidence

A competition authority that has evidence that competitors are behaving in a parallel manner, for example by charging the same prices or by refraining from poaching on each other’s territories, will naturally consider whether there is a sufficiently strong case to bring proceedings under Article 81 EC or the domestic equivalent thereof; and, in some jurisdictions, whether a criminal prosecution may be possible, including against individuals. However, there will be some cases where the evidence is insufficient to bring cartel proceedings. A question then arises as to whether the authority has other tools available to address any market failure, and as to remedies that might be appropriate to address that failure. One possibility might be to bring a case under Article 82 EC for collective abuse of a collective dominant position, although it is difficult to see how parallel pricing in itself could be found to be abusive; it may be possible, theoretically, to argue that excessive parallel prices are an exploitative abuse, although I am not aware of any such case having been brought.1 In UK competition law there are two other, to some extent inter-related, ways in which a problem of market failure of the kind described above may be tackled in circumstances where Articles 81 and 82 EC (or the so-called Chapter I and Chapter II prohibitions in the Competition Act 1998) are inapplicable for want of evidence of an agreement or abuse. First the Office of Fair Trading (‘the OFT’) may itself conduct a ‘market study’; and secondly the OFT (and various sectoral regulators such as the Office of Communications (‘OFCOM’)) may make a ‘market investigation reference’ to the Competition Commission (‘the CC’). Each of these devices will be described in this paper.

A. Market studies Section 5 of the Enterprise Act 2002 provides that one of the general functions of the OFT is to obtain, compile and keep under review information about matters relating to the carrying out of its functions. One of the ways in which the OFT carries out this general function is by conducting ‘market studies’ of * 1

Professor of Law, King’s College London. See Whish R. (2003): Competition Law, 5th edition, LexisNexis UK, pp. 527–528.


106

12/10/07

12:17

Page 106


markets which appear not to be working well for consumers but where enforcement action under competition or consumer law does not, at first sight, appear to be the most appropriate response. Market studies are carried out by the Markets and Policy Initiatives Division of the OFT (‘MPID’), and are intended to enable the OFT to understand as well as possible how markets are working and whether the needs of consumers are being met. The OFT has published “Market Studies: Guidance on the OFT approach”2 on the procedure that it normally follows when selecting markets for studies and in performing such studies. Market studies may be triggered in various ways: for example the enforcement divisions of the OFT or trading standards officers might bring problems to the attention of MPID, or they might be prompted by a ‘super-complaint’ from a designated consumer body.3 MPID also reviews existing undertakings given following Competition Commission investigations to determine whether they remain necessary or should be changed or removed. Market studies are distinct from market investigation references under Part 4 of the Enterprise Act, although it is possible that a market study might lead to a market investigation reference as occurred, for example, in the case of Store Cards.4 An account of the MPID’s work in studying markets can be found in the OFT’s Annual Report and Resource Accounts,5 and a section of the OFT’s website provides details of its completed and current market studies.6 Sometimes the OFT conducts a short market study, of three to six months or so, often limited to fact-finding; a decision to launch such a study is usually taken by the Director of MPID. A full market study, which may last a year or even more, is decided on by the OFT Board. Commencement of a market study is announced by a Press Release; and the OFT publishes a report of its findings. The OFT can request information under section 5(1) of the Enterprise Act 2002, but it does not have any powers of compulsion; it does have formal powers to obtain information under section 174 of the Act where the possibility of making a market investigation reference arises. Various outcomes may follow a market study by MPID. The following possibilities are set out in paragraph 3.18 of the OFT’s guidance: —giving the market a clean bill of health; —publishing information to help consumers; —encouraging firms to take voluntary action; —encouraging a consumer code of practice; —making recommendations to the Government or sector regulators; —investigation and enforcement action against companies suspected of breaching consumer law or competition law; —a market investigation reference to the Competition Commission. 2 3 4 5 6

OFT 519; the OFT intends to publish revised guidance by March 2007. On super-complaints under section 11 Enterprise Act 2002, see Whish, supra note 1, pp. 415–416. See below. See e.g. HC 906 (2003), pp. 18–32; HC 739 (2004), pp. 68–76; HC 171 (2005), pp. 62–69. http://www.oft.gov.uk.


12/10/07

12:17

Page 107

Oligopoly Theory and Economic Evidence 107

B. Table of market studies Completed market studies

Date of OFT report

Extended warranties for electrical goods

July 2002

Consumer IT December goods and services 2002 Pharmacies

Outcome of study Reference made to the CC under the (now-repealed) Fair Trading Act 1973 Generally the consumer IT market works well for consumers; OFT will continue to monitor the market

January 2003 Recommendation of Government action to liberalise entry to the community pharmacy market

Private Dentistry March 2003 Better information on prices and treatments (nb this followed a should be given to consumers, and there should super-complaint by be improvements to self-regulation of the market the Consumers’ Association) Payment systems

May 2003

OFT to examine the effectiveness of the commitment within the Banking Code to inform consumers about the length of clearing cycles; to complete the investigation of the MasterCard notification; to consider whether action is required on access to merchant acquiring and on debit card networks; and to monitor undertakings following the CC’s report under the Fair Trading Act 1973 on banking services

Liability insurance June 2003 market

Changes recommended to certain practices and the OFT to keep the market under review

Taxi services

Recommendation that elements of the regulatory framework for taxi services should be improved

November 2003

New car warranties December 2003

Manufacturers and dealers should improve their advice to consumers on their servicing options; also they should remove servicing restrictions from their new car warranties: failure to do so could lead to a formal investigation by the OFT under Article 81 EC. Subsequently servicing ties were removed from new car warranties

Debt consolidation March 2004 Better financial awareness among consumers and provision of clear, accurate and relevant information by credit providers needed to make the use of debt consolidation fairer and clearer


108

12/10/07

12:17

Page 108


Table Cont. Completed market studies

Date of OFT report

Store cards

March 2004 Reference made to the CC under s 131 Enterprise Act 2002: see market investigations below

Estate agents

March 2004 Recommendation that the Estate Agents Act 1979 should be amended, that self-regulation should be improved, and that consumers should take further action to protect their own position

Doorstep selling (nb this followed a super complaint by the NACAB)

May 2004

Outcome of study

Consumers require better information as to their rights in relation to doorstep selling, and existing legislation requires amendment to provide consumers with greater protection

Financial Services December and Markets Act 2004

No adverse effects on competition found to flow from the Act

Public sector procurement

September 2004

Preliminary research published; further discussions to take place with OGC

Public subsidies

November 2004

First stage of research published; further work to be done on the effect of subsidies in practice

Ticket agents

January 2005 Tickets should bear their face value and indicate when further charges may apply; Society of Ticket Agents and Retailers to develop model terms for its members to use in consumer contracts

Care homes (nb May 2005 this followed a super-complaint by the Consumers’ Association)

The Government should establish a central information point or ‘one-stop shop’ where people can get clear information about care for older people; better access to complaints procedures should be achieved; there is a need for fairer contracts

Classified directory April 2005 advertising services

Reference made to the CC under s 131 Enterprise Act 2002: see market investigations below

Property searches

Local authorities are required to give access to the information needed to complete a home information pack; clearer guidance to local authorities on how they should set prices for providing property information; agreement about revised targets with local authorities to ensure that this information is made available quickly and on the same timescale that they apply to themselves

September 2005


12/10/07

12:17

Page 109

Oligopoly Theory and Economic Evidence 109 Liability insurance June 2005 follow-up review

Lower increases in premiums, better communication between insurers and policyholders and a reduction in the number of businesses denied cover are among the key improvements identified by OFT in its follow-up review

Public subsidies stage two

November 2005

OFT presented proposals to the European Commission for reforming state aid controls so that the impact on competition is taken into account

January 2006

OFT recommendation that guidance be given to providers of subsidies to assess their potential effect on competition

Commercial use of public information Pharmaceutical Price Regulation Scheme Payment protection insurance Internet shopping

C. Market investigation references 1. Introduction Part 4 of the Enterprise Act 2002 provides for the reference of market investigations by the OFT and various sectoral regulators to the CC. These provisions replace the law on scale and complex monopoly situations contained in the Fair Trading Act 1973. In July 2001 the Government signalled its intention to overhaul the monopoly provisions in its White Paper Productivity and Enterprise—A World Class Competition Regime.7 In particular the Government proposed that decisions in the new regime for investigating markets should be taken primarily by the OFT and the CC, and that the Secretary of State should become involved only in cases which raise exceptional public interest issues; the decisions to be adopted by the OFT and CC would be made against a new test based on the prevention, restriction or distortion of 7

Cm 5233 (2001) chapter 6; see also the Government’s Response, December 2001.


110

12/10/07

12:17

Page 110


competition, as opposed to the public interest test set out in section 84 of the Fair Trading Act. Part 4 of the Enterprise Act 2002 entered into force on 20 June 2003. The OFT8 has power to make a reference to the CC where it has reasonable grounds for suspecting that any ‘feature or combination of features’ of a market prevent, restrict or distort competition in the UK or a part of it. Provision is made for the OFT to accept legally binding undertakings in lieu of a reference.9 The CC must determine whether there is an adverse effect on competition; if so it must decide on suitable remedies, bearing in mind the need to achieve as comprehensive a solution as is reasonable and practicable to any adverse effects identified. In taking remedial action the CC may also take into account any ‘relevant customer benefits’, as defined by the Act. A wide array of powers is available to the CC following its investigation including, where appropriate, the power to impose a structural remedy. References by the OFT and reports of the CC are required to be published.

2. Guidance The OFT has published Market investigation references: Guidance about the making of references under Part 4 of the Enterprise Act10 which explains how it intends to apply the Act. The CC has adopted the Competition Commission Rules of Procedure,11 which entered into force on 20 June 2003; they are available on its website.12 The CC has also published various guidelines of relevance to market investigations, including: —market investigation references: Competition Commission Guidelines13 —general advice and information14 —statement of policy on penalties.15

3. Making references The OFT (concurrently with the sectoral regulators) may make a market investigation reference to the CC when it has reasonable grounds for sus8 The sectoral regulators have the same powers as the OFT in relation to market investigation references. 9 See below. 10 OFT 511. 11 CC 1. 12 http://www.competition-commission.org.uk. 13 CC 3. 14 CC 4. 15 CC 5.


12/10/07

12:17

Page 111

Oligopoly Theory and Economic Evidence 111 pecting that one or more ‘features’ of a market prevent, restrict or distort competition in the supply or acquisition of goods or services in the UK or in a part thereof. Features of a market include the structure of the market concerned or any aspect thereof; the conduct of persons supplying or acquiring goods or services who operate in that market, whether that conduct occurs in the same market or not; and the conduct of those persons’ customers upstream and downstream in the market. Conduct for these purposes includes a failure to act and need not be intentional. The OFT, when making a reference, is not required to specify whether particular features of a market are a matter of structure on the one hand or of conduct on the other. Section 133 of the Enterprise Act specifies what the OFT must include in a market investigation reference; in particular it must provide a description of the goods or services to which the feature or combination of features that are restrictive of competition relate. A reference may be framed so as to require the CC to confine its investigation to goods or services supplied or acquired in a particular place or to or from particular persons. Provision is made for the variation of references. Section 169 of the Act requires the OFT to consult before making a reference and section 172 requires it to give reasons for its decision.

4. The discretion of the OFT whether to make a reference The OFT has a discretion, as opposed to a duty, to make a market investigation reference when the statutory criteria appear to be met. The OFT Guidance says that it will make a reference only when the following criteria, in addition to the statutory ones, are met: —it would not be more appropriate to deal with any competition issues under Articles 81 or 82 EC, under the Competition Act 1998 or by other means, for example the powers of the sectoral regulators; —it would not be more appropriate to accept undertakings in lieu of a reference; —the scale of the suspected problem, in terms of the adverse effect on competition, is such that a reference would be appropriate; —there is a reasonable chance that appropriate remedies will be available. The OFT Guidance provides further insights into the way in which its discretion will be exercised.

5. Relationship between competition act and market investigations The OFT’s policy is always to consider first whether a suspected problem can be addressed under Articles 81 or 82 EC or under the Competition Act 1998;


112

12/10/07

12:17

Page 112


it would only consider a market investigation reference where it has reasonable grounds to believe that market features restrict competition, but not to establish a breach of EC competition law or the Chapter I and/or Chapter II prohibitions, or when action under these provisions has been or is likely to be ineffective for dealing with any adverse effect on competition identified. The OFT Guidance goes on to explain that a market investigation reference might be appropriate for dealing with tacit coordination in oligopolistic markets or with problems arising from networks of vertical agreements. It adds that the majority of references are likely to involve industry-wide market features or multi-firm conduct, of which tacit coordination and parallel vertical agreements are examples. The OFT will review these criteria in the light of emerging case-law on the Chapter II prohibition should it appear that it is inadequate to deal with conduct by a single firm which has an adverse effect on competition; also it may make a reference where there has been an abuse of a dominant position under Article 82 EC or the Chapter II prohibition and it seems that a structural remedy going beyond what could be achieved under these provisions is necessary.

6. Relationship with the EC Modernisation Regulation The OFT Guidance discusses the relationship between the EC Modernisation Regulation and the market investigation provisions. That Regulation requires national competition authorities to apply Articles 81 and 82 EC where agreements or abusive conduct have an effect on inter-state trade; as a general proposition it is not possible to apply stricter national law than Article 81, but this is possible in the case of Article 82 EC. The OFT Guidance points out that this does not prevent investigations of agreements and conduct that infringe Articles 81 and 82 EC, but that it does affect the remedies that can be imposed. Where Article 81 EC is applicable to an agreement or agreements, it is unlikely that the OFT would make a reference. However the OFT notes the possibility that in certain circumstances the benefit of a block exemption can be withdrawn from vertical agreements, and that this could be a recommendation of the CC after a market investigation reference. The OFT Guidance also notes that the CC could achieve a stricter outcome in relation to abusive conduct than could be achieved under Article 82 EC. In the event that the CC were, during the course of a market investigation reference, to discover an infringement of Article 81 or 82 EC, it would not itself have the power to apply those provisions to the behaviour in question, since it has not been designated as a national competition authority for the purpose of applying the EC competition rules: only the OFT and the sectoral regulators have been so designated.


12/10/07

12:17

Page 113


7. Scale of the problem The OFT Guidance discusses the proposition that a reference would be made only where the scale of a suspected problem, in terms of its effect on competition, is such that a reference would be an appropriate response. It will consider whether the adverse effects on competition of features of a market are likely to have a significant detrimental effect on customers through higher prices, lower quality, less choice or less innovation; where the effect is insignificant the OFT would consider that the burden on business and the cost of a reference to the CC would be disproportionate. The OFT Guidance also says that, generally speaking, the OFT would not refer a very small market; a market only a small proportion of which is affected by the features having an adverse effect on competition; or a market where the adverse effects are expected to be short-lived. It adds that a reference might not be made where any adverse effect on competition appears to be offset by customer benefits.

8. Availability of remedies The OFT would not refer a market if it appeared that there were unlikely to be any available remedies to deal with an adverse effect on competition, for example where a market is global and a remedy under UK law would be unlikely to have any discernible effect.

9. Restrictions on the OFT’s ability to make a reference The OFT cannot make a reference if it has accepted undertakings in lieu of a reference within the preceding 12 months. This limitation does not apply where an undertaking has been breached; nor where it was accepted on the basis of false or misleading information.

10. The OFT’s application of the reference test The discussion in this section so far has concerned the statutory criteria for making a market investigation reference, the way in which the OFT will exercise its discretion whether to refer, and the restriction on it making a reference where it has accepted undertakings in lieu. This leaves one further important issue, which is the OFT’s interpretation of the reference test set out in section 131 of the Act. Part II of the OFT Guidance contains a helpful discussion of this. In chapter 4 it discusses the meaning of ‘prevention, restriction or


114

12/10/07

12:17

Page 114


distortion of competition’. In chapter 5 it considers structural features of markets, including the concentration level within a market, vertical integration, conditions of entry, exit and expansion, regulations and government policies, information asymmetries, switching costs and countervailing power. Chapter 6 deals with firms’ conduct, in particular the conduct of oligopolies, facilitating practices, custom and practice and networks of vertical agreements. Chapter 7 considers the conduct of customers, which section 131(2)(c) of the Act considers to be a feature of a market, and specifically considers the issue of search costs, that is to say the cost that customers may have to incur in order to make an informed choice.

11. The determination of references Once a reference has been made to the CC it must decide whether any feature, or combination of features, prevents, restricts or distorts competition in the referred market(s). If the CC considers that there is an adverse effect on competition, it must decide three additional questions: first, whether it should take action to remedy the adverse effect on competition or any detrimental effect on customers it has identified: detrimental effects are defined as higher prices, lower quality, less choice of goods or services and less innovation; secondly, whether it should recommend that anyone else should take remedial action; and thirdly, if remedial action should be taken, what that action should be. When considering remedial action, the CC must have regard to the need to achieve as comprehensive a solution as is reasonable and practical to the adverse effect on competition and any detrimental effects on customers, and may in particular have regard to the effect of any action on any relevant customer benefits. If the CC finds that there is no anti-competitive outcome, the question of remedial action does not arise.

12. Investigations and reports Section 136(1) of the Act requires the CC to prepare and publish a report; this must be done within two years. The report must contain the decisions of the CC on the questions to be decided under section 134, its reasons for those decisions and such information as the CC considers appropriate for facilitating a proper understanding of those questions and its reasons for its decisions. The time limit of two years may not be extended, but the Secretary State may reduce it by order.


12/10/07

12:17

Page 115


13. Duty to remedy adverse effects When the CC has prepared or published a report under section 136 and concluded that there is an adverse effect on competition, section 138(2) requires it to take such action as it considers to be reasonable and practicable to remedy, mitigate or prevent the adverse effect on competition and any detrimental effects on customers. In doing so the CC must be consistent with the decisions in its report on the questions it is required to answer, unless there has been a material change of circumstances since the preparation of the report or the CC has a special reason for deciding differently. In making its decision under section 138(2) the CC shall have regard to the need to achieve as comprehensive a solution as is reasonable and practicable to any adverse effects on competition or detrimental effects on customers, having regard to any relevant customer benefits of the market features concerned.

14. Procedure before the CC The procedures of the CC during market investigation references are set out in the CC’s Rules of Procedure and the Chairman’s Guidance to Groups.

15. The CC’s Guidance The CC has published Market Investigation References: Competition Commission Guidelines (CC 3) which explains its approach to market investigation references. It deals in turn with issues of market definition, the assessment of competition and remedial action, and contains an extremely useful guide to the competitive process and various factors that adversely affect competition. Chapter 3 of Part 4 of the Enterprise Act deals with the powers of the OFT and the CC to accept undertakings or to impose orders to ensure compliance with the Act and to monitor and enforce them. It begins with the powers of the OFT to accept undertakings in lieu of a reference to the CC; it then sets out the interim and final powers of the CC. Undertakings and orders are legally binding and enforceable in the courts. The OFT is required to maintain a register of undertakings and orders made under the market investigation provisions in the Enterprise Act; it is accessible on the OFT’s website.


116

12/10/07

12:17

Page 116


16. Undertakings and orders 16.1. Undertakings in lieu of a reference to the CC Section 154(2) of the Act gives power to the OFT to accept an undertaking in lieu of a reference to the CC. It can do this only where it considers that it has the power to make a reference to the CC and otherwise intends to make such a reference. In proceeding under section 154(2), the OFT must have regard to the need to achieve as comprehensive a solution as is reasonable and practicable to the adverse effect on competition concerned and any detrimental effects on customers, taking into account any relevant customer benefits. The OFT considers that undertakings in lieu are unlikely to be common. Before accepting undertakings in lieu the OFT is obliged to publish details of the proposed undertaking, to allow a period of consultation and to consider any representations received; a further period of consultation is required should the OFT intend to modify the undertakings. If an undertaking in lieu has been accepted it is not possible to make a market investigation reference to the CC within the following 12 months, unless there is a breach of the undertaking or it was accepted on the basis of false or misleading information. The OFT has used this procedure on one occasion, in the case of Postal franking machines;16 OFCOM has also used the procedure in the case of BT.17

16.2. Interim undertakings and orders to prevent pre-emptive action The CC can accept interim undertakings and make interim orders to prevent pre-emptive action which might prejudice any final remedy adopted by the CC. The Secretary of State may exercise these powers in public interest cases.

16.3. Final powers Sections 159 to 161 of the Enterprise Act deal with the final undertakings and orders that are available to the CC after it has completed its investigation and reached its conclusion on the questions contained in section 134. Section 159 provides for the acceptance of undertakings and section 160 for the making of an order where an undertaking is not being fulfilled or where false or misleading information was given to the OFT or CC prior to the acceptance of the undertaking. Section 161 empowers the CC to make a final order. The orders that can be made are set out in Schedule 8 of the Act.

16 17

The undertakings of 17 June 2005 are available at http://www.oft.gov.uk . The undertakings of 22 September 2005 are available at http://www.ofcom.org.uk.


12/10/07

12:17

Page 117


17. Enforcement functions of the OFT Section 162 of the Act requires the OFT to keep enforcement undertakings and enforcement orders under review and to ensure that they are complied with; it is also required to consider whether, by reason of a change of circumstances, there is a case for release, variation or supersession. The CC may ask the OFT to assist in the negotiation of undertakings. Section 167 provides that there is a duty to comply with orders and undertakings, which is owed to anyone who may be affected by a breach of that duty; any breach of the duty is actionable if such a person sustains loss or damage, unless the subject of the undertaking or order took all reasonable steps and exercised all due diligence to avoid a breach of the order or undertaking. Compliance with an order or undertaking is also enforceable by civil proceedings brought by the OFT or the CC for an injunction.

18. Review of decisions under Part 4 of the Enterprise Act Section 179 of the Act makes provision for review of decisions under Part 4 of the Act. Section 179(1) provides that any person aggrieved by a decision of the OFT, the Secretary of State or the CC may apply to the CAT for a review of that decision: the aggrieved person could be a third party with sufficient interest. When dealing with appeals under section 179(1) the CAT will apply the same principles as would be applied by a court on an application for judicial review. The CAT may dismiss the application or quash the whole or part of the decision to which it relates; and, in the latter situation, it may refer the matter back to the original decision maker for further consideration. An appeal may be brought, with permission, against the CAT’s decision on a point of law. Part 3 of the Competition Appeal Tribunal Rules makes provision for appeals under section 179 of the Act.18 There has so far been one appeal to the CAT under these provisions, in the case of Association of Convenience Stores v OFT.19 This was an appeal against the refusal of the OFT to make a market investigation reference to the CC of the grocery market. On 1 November 2005 the CAT handed gave a ruling setting aside the OFT‘s decision; subsequently the OFT did make a reference of this market to the CC.20

18

SI 2003/1372. Case No 1052/6/1/05, CAT [2005] 36. 20 See OFT Press Release 84/06, 9 May 2006, and Competition Commission News Release 27/06, 9 May 2006. 19


118

12/10/07

12:17

Page 118


D. Table of market investigation references There have so far been five market investigation references to the CC, as set out in the following table. Title of report

Date of reference

Date of report Outcome

Store Card Credit Services

18 March 2004

7 March 2006 Adverse effect on competition; various remedies imposed to increase transparency on the market for consumers and to separate payment protection from the basic credit service

Domestic bulk liquefied petroleum gas (LPG)

5 July 2004

Home Credit Market (nb this followed a super-complaint from the NCC)

20 December 2004

Classified Directory Advertising Services

5 April 2005

Northern Ireland 26 May 2005 Banking (nb this followed a super complaint from Which? in conjunction with the General Consumer Council for Northern Ireland)

(H) Ehlermann 07 Allendesalazar

12/10/07

12:17

Page 119

V Rafael Allendesalazar, Paloma Martínez-Lage and Roberto Vallina 1 Oligopolies, Conscious Parallelism and Concertation

A. Introduction The extreme market structures of perfect competition and absolute monopoly are scarce in the real world, where oligopoly is a much more prevalent situation. In many oligopolistic markets, undertakings tend to be competitively interdependent, and each firm consciously adapts its own strategy to the expected reactions of its competitors. This situation may result in supracompetitive oligopoly pricing. But firms that charge supra-competitive prices do not, for that reason alone (and in the absence of some sort of explicit or tacit agreement), infringe competition law. It has been rightly pointed out that treating such behaviour as illegal would require competitors to act irrationally by ignoring their perceived interdependencies, and would also call for unworkable remedies.2 The difficulty for competition law, competition authorities, and also for the undertakings themselves, is to distinguish between situations in which strategic coordination implies some sort of illicit collusion and when it merely corresponds to spontaneous coordination resulting from the rational response of each member of the oligopoly to perceived interdependencies. This paper has two main parts. In Section B we outline the general principles governing the application of Article 81 EC to strategic coordination in oligopolies, as defined by the case law. In the light of such principles, we see in Section C how the notions of concerted practices and conscious parallelism are defined under Spanish competition law, especially in view of the practice of the Spanish competition authorities. Finally, in Section D we draw our conclusions.

1 Rafael Allendesalazar is a partner at Martínez Lage & Asociados. Paloma Martínez-Lage and Roberto Vallina are associates at Martínez Lage & Asociados. 2 US Department of Justice and FTC (1999): Note for the Mini-Roundtable on Oligopoly, OECD Paris available at: www.ftc.gov/bc/international/docs/compcomm/1999—miniroundtable%20on%20Oligopoly.pdf


120

12/10/07

12:17

Page 120


B. Article 81 EC, concerted practices and oligopolies Concertation within the meaning of Article 81 EC implies, as such, parallel conduct or at least some degree of coordination. However, parallel conduct does not always amount to concertation within the meaning of Article 81 EC. According to the ECJ, the notion of concertation under Article 81 EC must be understood in light of the concept inherent in the provisions of the EC Treaty relating to competition: each economic operator must determine independently the policy which he intends to adopt on the market. The Treaty thus lays down a requirement of independence.3 However, this requirement does not prevent economic operators from adapting themselves intelligently to the conditions of the market; rather, it prohibits any practice that may influence competitors’ conduct on the market.4 The problem lies in distinguishing between legitimate conscious parallelism and illicit collusion.5 Oligopoly is the place where the boundaries between these concepts can become elusive.

1. Oligopolies Oligopoly has been traditionally defined (at least by lawyers) as a market structure in which there are only a few suppliers (at least two).6 However, it 3 Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114/73 Suiker Unie v Commission [1975] ECR 1663, at paragraphs 173–174. 4 Ibid. 5 The essence of this problem has been very clearly stated in the OECD definition of conscious parallelism contained in its Glossary of Industrial Organisation Economics and Competition Law:

“Under conditions of oligopoly, the pricing and output actions of one firm have a significant impact upon [those] of its rivals. Firms may after some period of repeated actions become conscious or aware of this fact and without an explicit agreement coordinate their behaviour as if they were engaged in collusive behaviour or a cartel to fix prices and restrict output. The fear that departure from such behaviour may lead to costly price cutting, lower profits and market share instability may further create incentives for firms to maintain such an implicit arrangement amongst themselves. This form of conscious parallel behaviour or tacit collusion generally has the same economic effect as a combination, conspiracy or price fixing agreement. However, whether or not conscious parallel behaviour constitutes an illegal action which is restrictive of competition is [a] subject of controversy in both competition law and economics. Price uniformity may be a normal outcome of rational economic behaviour in markets with few sellers and homogenous products. Arguments have been advanced that the burden proof must be higher than circumstantial evidence of concerted or parallel behaviour and uniform pricing and output policies. In other words, conscious parallelism in and of itself should not necessarily be construed as evidence of collusion. The problem arises more from the nature of the market or industry structure in which firms operate than from their respective behaviour.” 6 Faull J. and Nikpay A. (1999): The EC Law of Competition, Oxford University Press, Oxford, p. 24.


12/10/07

12:17

Page 121

Oligopolies, Conscious Parallelism and Concertation 121 has been rightly pointed out that what matters is not really the number of firms. From an economic point of view, what matters is market power, that is, whether undertakings can, either individually or collectively, reduce output or raise prices to the detriment of consumers.7 The relevant feature of oligopolies from the point of view of Article 81 EC is that, in such markets, competitors are interdependent. According to the “oligopolistic interdependence” theory, in a genuine oligopolistic market the structure of the market enables competitors to closely monitor each other’s movements, as a result of which they are bound to match one another’s marketing strategy.8 Thus, if the necessary conditions are met (transparency, etc.), market players in a homogeneous oligopolistic market tend towards a stable, noncompetitive equilibrium. If a firm in such a market cuts its prices, it will increase its market share for a short period of time. However, the competitors will probably react by matching the price cut, thus neutralizing the gains of the price cut in the long term. The initial gain obtained by the firm could thus be outweighed by these long-term sales at this new reduced price. In short, if the conditions are met, undertakings in oligopolistic markets may have little incentive to compete, and they must closely follow and analyze the other operators’ conduct. If this is the case, the undertakings reach a non-competitive equilibrium. Game theory (and the prisoner’s dilemma) supports this conclusion. According to game theory, if in a game the worst solution for each player is to ignore the behaviour of the other players, since such behaviour will affect the outcome of the game, the players will end up cooperating. They will either decide from the very first moment that reaching an agreement is the most advantageous solution, or they will reach the conclusion that it is necessary to cooperate by trial and error.9 In such a case, the players will cooperate “as if” they had reached an agreement. For game theory it does not matter whether such cooperation is reached through an express agreement or by tacit cooperation (acting “as if” there was an agreement). What matters is whether the conditions of the game create the incentives for the players to cooperate. It can thus be seen that the structural conditions of the market in which oligopolists operate can be such that they have little or no incentive to compete. In these circumstances, parallel conduct would not be the result of the will or intention of market players but a natural consequence of the market structure, or the “natural operation” of the market, in the wording of the ECJ.10 Nevertheless, the very existence of an oligopoly does not entail a less com7

Whish R. (2003): Competition Law, 5th edition, Lexis Nexis, pp. 504–505. Ibid., at pp. 506–507. 9 Franzosi M. (1988): “Oligopoly and the Prisoner’s Dilemma: Concerted Practices and ‘As If ’ Behaviour”, 9 European Competition Law Review 385. 10 Joined Cases C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85 A. Ahlström Osakeyhtiö e.a. (Woodpulp II) [1993] ECR I-1307, at paras. 101–102. 8


122

12/10/07

12:17

Page 122


petitive industry or market. Day-to-day business experience proves that, in some oligopolistic markets, competition is intense and fierce.11 This shows that not every oligopoly creates interdependence between competitors: for a non-competitive outcome, certain conditions must be met, e.g., transparency (a vital element), possibilities of retaliatory measures, product homogeneity, etc.12 Oligopolies thus pose a major challenge for competition law. On the one hand, some oligopolies could affect the very essence of the objectives that competition law aims to protect: they could lead to a restriction of output and may increase prices well above the competitive level.13 On the other hand, Article 81 EC only applies to agreements and coordinated practices. That is, Article 81 EC requires at least a concurrence of wills or a meeting of minds between the undertakings concerned. If the market conduct of an enterprise is not the result of either an agreement or at least a conscious common intention of coordination among competitors, Article 81 EC is not applicable. The question therefore arises as to what extent, if any, parallel behaviour in an oligopolistic market is the result of a meeting of minds or whether, on the contrary, it is the result of the autonomous will of each undertaking? The answer to this question is the key element in applying Article 81 EC in most cases.

2. Oligopolies, concertation and Article 81 EC The ECJ stated in ICI (and since then has repeatedly affirmed) that concertation within the meaning of Article 81 EC must be defined as:14 “[. . .] a form of coordination between undertakings which, without having reached the stage where an agreement properly so-called has been concluded, knowingly substitutes practical cooperation between them for the risks of competition”.

The ECJ further added that: “[b]y its very nature, then, a concerted practice does not have all the elements of a contract but may inter alia arise out of coordination which becomes apparent from the behaviour of the participants”.

Therefore, concertation implies a form of cooperation that: (i) protects the 11

See comments on this by Whish R., supra note no. 7, pp. 510–511. See, from a merger control perspective, Case T-342/99 Airtours v Commission [2002] ECR II-2585, at paras. 58, 60 et seq. 13 Of course, this begs the question of what is the competitive price level? And how can it be determined? These questions are hard if not impossible to answer. See the reference made to the normal conditions of competition in Case 48/69 Imperial Chemical Industries Ltd. (ICI) v. Commission [1972] ECR 619, at para. 66. 14 See ibid. 12


12/10/07

12:17

Page 123

Oligopolies, Conscious Parallelism and Concertation 123 undertakings concerned from the risks of competition; and (ii) usually becomes apparent from the behaviour of the participants (for example, from parallel conduct with regard to their pricing policy). The problem now lies in how to apply and interpret the notion of concertation in the context of an oligopoly. As we have seen, market players in an oligopolistic market may assess the potential actions of competitors and take account of the interdependence of their strategies.15 To what extent, if any, does the oligopolists’ analysis of the market and the strategies of their competitors amount to a concerted practice within the meaning of Article 81 EC? The ECJ has acknowledged that, although each economic operator must determine its own commercial policy independently, it is legitimate for the economic operators to adapt themselves intelligently to the existing and anticipated conduct of their competitors.16 However, if adapting intelligently to the existing and anticipated conduct of their competitors is legitimate, it is foreseeable that all the existing operators will adapt similar conduct, i.e. parallel conduct. Such conduct could be easily read as evidence of “coordination which becomes apparent from the behaviour of the participants”. The ECJ rightly acknowledged this, and ruled that parallel conduct as such is not caught by Article 81 EC and creates no presumption of collusion between the undertakings concerned. Nevertheless, the Court’s case law seemed to provide for two exceptions to this principle: —First, in ICI the ECJ stated that, although parallel behaviour may not by itself be identified with a concerted practice, parallel conduct may amount to strong evidence of concertation if it leads to conditions of competition which do not correspond to the normal conditions of the market, having regard to the nature of the products, the size and number of the undertakings, and the volume of the said market.17 —Second, in Woodpulp II the ECJ ruled that parallel conduct cannot be regarded as proof of concertation unless concertation constitutes the only plausible explanation for such conduct.18

In spite of the fact that the language varies from one case to another, these two “exceptions” can be read as a single one: parallel conduct is not proof of concertation if the conduct of the enterprises can be explained by market conditions. In ICI, the ECJ said that parallel conduct may amount to strong evidence of concertation if it leads to conditions of competition which do not correspond to “the normal conditions of the market”. However, in practice it is very diffi15

Faull J. and Nikpay A. (1999), supra note no. 6, p. 24. Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114/73 Suiker Unie v Commission [1975] ECR 1663, at paras. 173–174. 17 Case 48/69 Imperial Chemical Industries Ltd. (ICI) v Commission [1972] ECR 619, at para. 66. 18 Joined Cases C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85, A. Ahlström Osakeyhtiö e.a. (Woodpulp II) [1993] ECR I-1307, at para. 71. 16


124

12/10/07

12:17

Page 124


cult to ascertain what such “normal” conditions are, and this reference to the “the normal conditions of the market” was never repeated.19 The position of the ECJ in ICI must be read in light of the subsequent cases, and mainly in light of Woodpulp II. In Woodpulp II, the ECJ reformulated its previous statement in ICI. The ECJ again said that account must be taken “of the nature of the products, the size and the number of the undertakings and the volume of the market in question”, but with an important difference. The goal of such an analysis is no longer to establish what “the normal conditions of the market” are but to ascertain whether or not the parallel conduct can be explained otherwise than by concertation. Thus, the case law must be seen as an evolution whose outcome is clear. Parallel conduct is not prohibited, and it does not create a presumption iuris tantum of collusion. Nevertheless, parallel conduct can be considered sufficient proof of collusion if concertation is the only plausible explanation.

3. Plausible explanations The list of plausible explanations is not exhaustive, but to date, two main circumstances have been considered to be plausible explanations for parallel behaviour: (i) price leadership; and (ii) market structure. As we have said, it is legitimate for economic operators to adapt themselves intelligently to the existing and anticipated conduct of their competitors.20 Consequently, if there is a price leader in the market, competing undertakings could try to adapt themselves to the leader’s commercial policy. Such parallel pricing behaviour in an oligopoly producing homogeneous goods would not in itself amount to concertation within the meaning of Article 81 EC.21 A price leader could be characterized as an economic operator which, due for example to its market share, is able to act independently of its competitors, knowing that they would almost certainly follow suit (“dominant price leadership”).22 Another possibility is so-called “barometric price leadership”, where the firm taking the lead is not dominant but is widely accepted as the best performing operator in the sense of meeting demand and adapting to evolving market conditions (cost increases, etc.). The conclusion could be quite different if additional evidence is adduced,

19 This may be because the notion of “normal market conditions” was heavily criticized by commentators. 20 Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114/73 Suiker Unie v Commission [1975] ECR 1663, at paras. 173–174. 21 Commission Decision EEC/84/405 of 6 August 1984, Case IV/30.350—zinc producer group, OJ L 220 [1984], at paras. 75–76. 22 Waelbroeck M. (1997): Commentaire Megret. Vol. IV: Concurrence, ULB Brussels, p. 138–139. See also supra note no. 21.


12/10/07

12:17

Page 125

Oligopolies, Conscious Parallelism and Concertation 125 such as evidence of contacts between undertakings on desirable price changes prior to the adoption of a new price, or of an exchange of information that reinforces such contacts.23 In any event, an agreement between competitors to follow or to choose a price leader will be regarded as concertation within the meaning of Article 81 EC. Another possibility is that the very structure of the market leads to parallel conduct. The ECJ has implicitly acknowledged that, in some oligopolistic markets, competitors are interdependent. Accordingly, in Woodpulp II the ECJ appointed a group of economic experts to examine the characteristics of the affected market during the period covered by the Commission’s contested decision. After examining the market in the relevant period, the experts concluded that the normal operation of the market was a more plausible explanation for the uniformity of prices than concertation.24 The ECJ accepted this as a plausible explanation for the parallel conduct and upheld the applicants’ argument that the Commission had not sufficiently proved concertation within the meaning of Article 81 EC. The ECJ thus implicitly accepted the practical consequences of the “oligopolistic interdependence” theory. It is worth noting that, in Woodpulp II the experts did not say that the normal operation of the market was the sole plausible explanation for the uniformity of prices. For the ECJ, it was enough that there were other explanations apart from collusion. Concertation thus cannot be inferred from the mere existence of parallel conduct.

4. Burden of proof According to the case law of the ECJ, in order to establish that parallel behaviour is the result of concerted action, the evidence must be “sufficiently precise and coherent”.25 In light of the foregoing, in Woodpulp II Advocate General Darmon considered that this statement of the ECJ implied that it is necessary to establish a degree of certainty that goes beyond any reasonable doubt. He further added that, in “accordance with the principles governing the burden of proof, it is for the Commission to demonstrate that; the burden of proof cannot be shifted simply by a finding of parallel conduct”.26 Still further, the Advocate 23 Ibid. See also Case T-202/98 Tate & Lyle v Commission (British Sugar) [2001] ECR II-2035, at paras. 34–46. 24 Joined C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85 A. Ahlström Osakeyhtiö e.a. (Woodpulp II) [1993] ECR I-1307, at paras. 75 et seq. 25 Joined Cases 29/83 and 30/83 Compagnie Royale Asturienne des Mines SA (CRAM) and Rheinzink GmbH v Commission [1984] ECR 1679, at para. 20.


126

12/10/07

12:17

Page 126


General considered that, in any event, if a plausible alternative explanation is put forward by the parties, then concertation cannot be deemed to be established.27 The position of Advocate General Darmon in Woodpulp II was quite clear: the burden of proof is always borne by the competition authority. Parallel conduct neither creates a presumption of concertation nor shifts the burden of proof. In view of the conclusions of Advocate General Darmon, in Woodpulp II the ECJ ruled on this point as follows.28 “[70] Since the Commission has no documents which directly establish the existence of concertation between the producers concerned, it is necessary to ascertain whether the system of quarterly price announcements, the simultaneity or nearsimultaneity of the price announcements and the parallelism of price announcements as found during the period from 1975 to 1981 constitute a firm, precise and consistent body of evidence of prior concertation. [71] In determining the probative value of those different factors, it must be noted that parallel conduct cannot be regarded as furnishing proof of concertation unless concertation constitutes the only plausible explanation for such conduct. It is necessary to bear in mind that, although [Article 81] of the Treaty prohibits any form of collusion which distorts competition, it does not deprive economic operators of the right to adapt themselves intelligently to the existing and anticipated conduct of their competitors (see the judgment in Suiker Unie, cited above, paragraph 174). [72] Accordingly, it is necessary in this case to ascertain whether the parallel conduct alleged by the Commission cannot, taking account of the nature of the products, the size and the number of the undertakings and the volume of the market in question, be explained otherwise than by concertation.”

The position of the ECJ seems, at first glance, to follow Advocate General Darmon. However, this may not be a realistic interpretation of the existing case law. First, competition authorities reason (at least in the officials’ minds) in a quite different way: parallel conduct is either proof of concertation, or at least creates a presumption of collusion. No matter how the case law defines the principles governing the burden of the proof in these cases, a competition authority will presume the existence of concertation if there is parallel behaviour. It will then be a hard task for the defendants to find “sufficiently precise and coherent” evidence that concertation is not the only plausible explanation. It is neither practical nor realistic to expect competition authority officials to presume that there is no concertation in the face of parallel behaviour. Second, Woodpulp II must be read in the light of the previous case law. 26 Joined C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85 A. Ahlström Osakeyhtiö e.a. (Woodpulp II) [1993] ECR I-1307. 27 Ibid. 28 See Opinion of Advocate General Darmon in Joined C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85, A. Ahlström Osakeyhtiö e.a. (Woodpulp II) [1993] ECR I-1307, at paras. 96 and 196.


12/10/07

12:17

Page 127

Oligopolies, Conscious Parallelism and Concertation 127 Indeed, in CRAM and Rheinzink, the ECJ stated: “The Commission’s reasoning is based on the supposition that the facts established cannot be explained other than by concerted action by the two undertakings. Faced with such an argument, it is sufficient for the applicants to prove circumstances which cast the facts established by the Commission in a different light and which thus allow another explanation of the facts to be substituted for the one adopted by the contested Decision.” (emphasis added)

This seems to be a much more practical approach. Competition authorities tend to presume that parallel conduct is the result of concertation if they see no clear evidence supporting an alternative explanation. It is then for the defendants to provide an alternative plausible explanation. This does not mean that a competition authority must never look for alternative explanations. As an initial matter, a competition authority must also examine on its own initiative whether such an alternative explanation exists or not. However, if after a preliminary analysis the competition authority does not find a plausible alternative explanation, the burden of proof will shift to the defendants (at least, in practice). It is then for the defendant undertakings to provide evidence which casts the facts established by the competition authority in a different light and which thus allows another explanation for those facts.29

5. Additional evidence of concertation Nevertheless, the problem of the plausible explanation, the burden of proof or the structure of the market only arises if evidentiary support for the case is insufficient. Parallel conduct could be considered as proof of collusion if it can be connected with other evidence of collusion, in particular with facilitating practices or “plus factors”. These facilitating practices have been defined as activities that promote interdependent behaviour among competitors by reducing their uncertainty as to each other’s future action, or by diminishing their incentive to deviate from a coordinated strategy.30 For instance, parallel conduct can be deemed to be proof of concertation if it is accompanied by evidence of any of the following:31

29 See Joined Cases T-305/94, T-306/94, T-307/94, T-313/94 to T-316/94, T-318/94, T-325/94, T-328/94, T-329/94 and T-335/94 Limburgse Vinyl Maatschappij NV, Elf Atochem SA, BASF AG, Shell International Chemical Company Ltd, DSM NV, DSM Kunststoffen BV, Wacker-Chemie GmbH, Hoechst AG, Société artésienne de vinyle, Montedison SpA, Imperial Chemical Industries plc, Hüls AG and Enichem SpA v Commission (PVC II) [1999] ECR II-931, at para. 728. 30 See Areeda P. (1986): Antitrust Law, Little Brown, Boston, para. 1407 b. 31 See Ritter L. and Braun W. D. (2005): European Competition Law: A Practitioner’s Guide, Kluwer Law International, The Hague, 3rd edition, at pp. 108 et seq.


128

12/10/07

12:17

Page 128


—Documents. Taken together, a finding of parallel market conduct and documents which show that the practices were the result of concerted action are sufficient proof of concertation32. —Contacts between competitors. In its landmark Sugar case, the ECJ held that Article 81 EC strictly precludes any direct or indirect contact between operators that can influence the conduct on the market of an actual or potential competitor.33 In such a case, the defendants cannot explain the existence of the parallel conduct by the fact that they have adapted intelligently to the existing and anticipated conduct of their competitors. Such contacts remove in advance the uncertainty as to the future conduct of the competitors, and thus protect the undertakings from the risks of competition. These contacts could consist of, for example, meetings between competitors.34 —Disclosing to competitors the course of conduct that each undertaking has decided to adopt (or contemplates adopting) on the market. For example, announcements of price increases, together with parallel conduct, can be regarded as proof of concertation.35 —Exchange of Information. Exchange of sensitive data can be considered proof of concertation, especially if the data are closely linked to the competitive conditions in respect of which the conduct of competitors is parallel.36 —Reciprocal supply agreements between competitors. Reciprocal supply agreements between competitors could be regarded as proof of concertation, especially if, at the same time, the concerned undertakings refrain from supplying competitors’ clients.37 —Common board members. Having representatives on the board of directors or any other management body of a competitor could be considered a device facilitating collusion.38 32 See Joined Cases T-305/94, T-306/94, T-307/94, T-313/94 to T-316/94, T-318/94, T-325/94, T-328/94, T-329/94 and T-335/94 Limburgse Vinyl Maatschappij NV, Elf Atochem SA, BASF AG, Shell International Chemical Company Ltd, DSM NV, DSM Kunststoffen BV, Wacker-Chemie GmbH, Hoechst AG, Société artésienne de vinyle, Montedison SpA, Imperial Chemical Industries plc, Hüls AG and Enichem SpA v Commission (PVC II) [1999] ECR II-931, at paras. 724–728. 33 Ibid. 34 See, inter alia, Case T-7/89 Hercules Chemicals NV v Commission [1991] ECR II-1711 at paras. 259–261. 35 Case 48/69 Imperial Chemical Industries Ltd. (ICI) v Commission [1972] ECR 619, at paras. 83 et seq. 36 Case 172/80 Gerhard Züchner v Bayerische Vereinsbank AG. [1981] ECR 2021, at para. 21. 37 Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114/73, Suiker Unie v Commission [1975] ECR 1663, at paras. 173–174, and Commission Decision of 23 of December 1977, Case IV/29176—Vegetable Parchment, OJ L 70 [1978]. 38 Joined Cases 142 and 156/84 British-American Tobacco Company Ltd (BAT) and R. J. Reynolds Industries Inc. v Commission [1987] ECR 4487, at para. 46 inter alia. See also Commission Decision EEC/86/405 of 14 July 1986, Case IV/30.320—optical fibres, OJ L 236 [1986].


12/10/07

12:17

Page 129

Oligopolies, Conscious Parallelism and Concertation 129 —Associations of enterprises. If all (or a vast majority) of the undertakings investigated are members of an association of enterprises, this can be an element supporting an accusation of concertation by a competition authority. —Network of joint ventures coordinated by a parent company. In the Optical fibres case, the Commission found that concertation in an oligopolistic market could result of a network of interrelated joint ventures with a common technology provider and a common parent company.39 The above non-exhaustive list of examples shows that, in the presence of parallel conduct, any device, practice or framework facilitating collusion can be considered sufficient supporting proof of concertation. In such cases, it is not enough for the defence merely to provide a plausible alternative explanation of the parallel conduct. The additional proof (exchange of information, documents, etc.) must also be rebutted.40 Furthermore, it should be recalled that most of the facilitating practices can serve pro-competitive as well as anticompetitive purposes. Therefore, the mere fact that some sort of facilitating practice exists should not alone lead to a conclusion of illicit collusion among the defendants.

6. Prices and parallel conduct We would like to end this section with a little reflection on parallel conduct. In a number of markets there is a significant gap between the ‘official’ or the ‘catalogue prices’ and the actual transaction prices charged to such customers. In those markets, the ‘official prices’ of the market players may be similar, but the effective transaction prices could vary significantly from one market player to another.41 For a competition authority, the similarity of the ‘official prices’ would be clear evidence of collusion (at least in the officials’ minds). The point is that, in such markets, competition could have shifted to the rebates and conditions offered to each buyer. Secret (or not so secret) discounts could spread all over the industry, and the effective prices charged to customers could therefore be very different from those on which the competition authority’s attention is focused. In our experience, the competition authorities tend to pay too much 39 Commission Decision EEC/86/405 of 14 July 1986, Case IV/30.320—optical fibres, OJ L 236 [1986]. 40 See Joined cases T-305/94, T-306/94, T-307/94, T-313/94 to T-316/94, T-318/94, T-325/94, T-328/94, T-329/94 and T-335/94 Limburgse Vinyl Maatschappij NV, Elf Atochem SA, BASF AG, Shell International Chemical Company Ltd, DSM NV, DSM Kunststoffen BV, Wacker-Chemie GmbH, Hoechst AG, Société artésienne de vinyle, Montedison SpA, Imperial Chemical Industries plc, Hüls AG and Enichem SpA v Commission (PVC II) [1999] ECR II-931, at paras. 724–728. 41 In this regard, see the commentary on the Woodpulp II case in Motta M. (2004): Competition Policy: Theory and Practice, Cambridge University Press, at p. 219.


130

12/10/07

12:17

Page 130


attention to the announced or ‘official prices’ rather than to effective transaction prices (maybe because this is a much easier task). In our opinion, this is wrong. If important rebates are not only widespread but are almost the rule in the market, the ‘official prices’ are only a reference point, and do not reflect the reality of the market. In such cases, whether there is concertation in setting up the ‘official prices’ is not so relevant. When there is a big gap between ‘official prices’ and the real market prices, free riding is not only possible, it will be the rule, and retaliatory measures against such free riding will be hard to implement. Competition authorities should essentially focus on proving that parallel conduct exists as to the effective prices charged to customers, which could only be possible if prices are also similar and there exists some degree of transparency at that level. By focusing on ‘official’ prices and ignoring the effect of individual discounts leading to different real prices, competition authorities could be unwillingly introducing an element of artificial transparency in the market, thereby reducing the uncertainty as to other competitors’ pricing practices and facilitating collusion.

C. Concerted and consciously parallel practices in the case law of the Tribunal de Defensa de la Competencia The equivalent in Spanish law of Article 81 EC, Article 1 of the Ley 16/1989 de Defensa de la Competencia (“LDC”), refers to “any collective agreement, decision or recommendation or any concerted or consciously parallel practice aimed at producing or enabling the effect of impeding, restricting or distorting competition in all or any part of the domestic market”. (emphasis added) Article 1 LDC thus mentions consciously parallel practices among the prohibited conduct, unlike Article 81 EC. The question then arises as to whether, in the LDC, concerted practices and consciously parallel practices are equivalent concepts or whether on the contrary they refer to two different types of conduct and thereby introduce a disparity between Spanish and EC competition law. We will analyze the judgments of the Tribunal de Defensa de la Competencia (“TDC”), which support each of these two possibilities (compare sections 1 and 2 below), in order to try and draw a conclusion from the existing case law.

1. Concerted and consciously parallel practices as different categories of infringements of Article 1 LDC The LDC has inherited the expression of consciously parallel practices from the law that regulated competition in Spain until 1989, when the LDC was


12/10/07

12:17

Page 131

Oligopolies, Conscious Parallelism and Concertation 131 enacted. In fact, Act 110/1963 (Ley 110/1963 de Represión de Prácticas Restrictivas de la Competencia, or the “LRPR”) already mentioned consciously parallel practices. The precedent for this Spanish term, first included in the LRPR, may be found in the theory of conscious parallelism developed by the North American courts in applying Section 1 of the Sherman Act. In fact, in the 1950s, just before the enactment of the LRPR, US courts had expressed the inability of Section 1 of the Sherman Act to deal with cases in which it was difficult to demonstrate the existence of an agreement or in which there was a restriction of competition but no agreement between the parties involved. In applying the LRPR, the TDC defined consciously parallel practices as those practices where the involved parties, without concluding any kind of agreement, and acting “unilaterally but harmonically”, remove from free competition one of its characteristic components.42 Thus, according to the TDC’s judgments in Asturiana de Zinc and Española de Zinc, in a consciously parallel practice there is not necessarily an agreement between the companies involved: it is enough that they act unilaterally but harmonically, and thus avoid competing with each other. In this respect, in Asturiana de Zinc, the TDC stated that “the supposition that behind a consciously parallel practice there has to be, necessarily for it to be prohibited, an agreement, is not imposed by law; on the contrary, it should be accepted that the prohibition embraces such conduct, independently of the agreement, insofar as the involved parties, deliberately, and acting unilaterally but harmonically, are removing from competition one of its characteristic components”. Further to these cases under the LRPR, the key judgment on consciously parallel practices after the enactment of the LDC was adopted by the TDC on 6 March 1992 in a case known as Detergentes.43 The facts were the following: in September 1989, new four-kilo concentrated detergent containers of different companies appeared simultaneously on the Spanish market, indicating that they were equivalent to the usual five-kilo containers of nonconcentrated detergent. The companies involved—Henkel, Camp, Procter & Gamble and Lever—denied the existence of any agreement or decision on the reduction of size of the containers or on the production and sale of concentrated detergent. They claimed that the change from the five-kilo container to the four-kilo one was due exclusively to the reduction of sulphates in the detergents, which was beneficial for the environment. The relevant companies acknowledged having discussed this issue in an industry association (ADTA) but they claimed that it had simply been an exchange of opinions and that there had been no agreement whatsoever between them. There had 42 Judgment of the TDC of 18 July 1986 in Case 223/86, Española de Zinc; judgment of the TDC of 9 January 1987 in Case 223/89, Asturiana de Zinc. 43 Judgment of the TDC of 6 March 1992 in Case 306/91, Detergentes concentrados.


132

12/10/07

12:17

Page 132


also been contacts with the Spanish Administration in order to exchange opinions on the effects on the environment of sulphates and other detergent components. However, the Administration had only pointed out the disadvantages of dumping in water chemical components of detergents, without imposing any standardization or normalization of the products. As there was no evidence of a collective recommendation or decision by ADTA or of a formal agreement between the companies, the TDC explored the possibility of a “concerted or consciously parallel practice”, as expressed in Article 1 LDC. It started by defining concerted practices in accordance with the judgments of the ECJ in ICI 44 and Zuechner.45 As the ECJ stated in ICI, Article 81 EC draws a distinction between the concept of “concerted practices” and the concepts of “agreements between undertakings” or “decisions by associations of undertakings”. As explained by the ECJ, the reference to concerted practices in Article 81 is intended to bring within the scope of the prohibition a form of coordination between undertakings which, without having reached the stage where an agreement properly so called has been concluded, knowingly substitutes practical cooperation between them for the risks of competition. Starting from there, the TDC concluded that “concerted practices” are those practices which, restricting competition in the market, imply a conscious parallelism of conduct between economic operators, usually competitors, and a certain cooperation and coordination between them. By contrast, “consciously parallel practices”, according to the TDC in its judgment, do not require the consensual element, i.e., the element of coordination. However, the TDC specified that, if liability is to be established, such consciously parallel behaviour “should not be explained by normal or foreseeable reactions of the operators in the market”. In that particular case, the TDC concluded that the four companies involved had engaged in a concerted practice through their parallel conduct, and in particular through their mutual dissemination of strategic commercial information. According to the TDC, all the elements of concerted practices were present: (i) there was a parallelism of conduct (the simultaneous launching of the four-kilo concentrated container); (ii) there was conscious behaviour (the conduct of the companies in ADTA meetings and the discussion of the effects of sulphates on the environment and the launching of concentrated detergents, thereby expressing a common will); (iii) there was cooperation and coordination (the exchange of information, the solicitation of third parties to adhere to the common behaviour, the waiting period of some companies already set to launch the new product format, etc.); and (iv) there was a distortion of competition (while the object was not to raise prices in this case, the coordination was intended to avoid the risks of an individual product launch). 44 45

Case 48/69 Imperial Chemical Industries Ltd. v Commission [1972] ECR 619. Case 172/80 Gerhard Zuechner v. Bayerische Vereinsbank [1981] ECR 2021.


12/10/07

12:17

Page 133

Oligopolies, Conscious Parallelism and Concertation 133 Accordingly, the TDC did not need to resort to the concept of consciously parallel practices in the Detergentes judgment because there was evidence of coordination between the companies (the meetings at ADTA). Nevertheless, the existence of such a concept was plainly affirmed by the court. In fact, it is quite clear from the Detergentes judgment that consciously parallel practices are an independent type of infringement of Article 1 LDC. The TDC recapitulates both EC and national case law and tries to set the boundaries between the different types of conduct that fall under Article 1 LDC. According to this judgment, there are four types of conduct that fall foul of Article 1 LDC: (i) decisions or recommendations by associations of undertakings; (ii) formal or proper agreements between undertakings; (iii) concerted practices: they are not formal agreements but they require a coordination or cooperation (i.e., concertation) between the undertakings involved; (iv) consciously parallel practices: coordination is absent from this concept; the undertakings act “unilaterally but harmonically” to restrict competition. In light of these categories, and in light of the precedents cited by the TDC, a consciously parallel practice within the meaning of Article 1 LDC would be prohibited conduct in respect of which there is neither an agreement nor any coordination or cooperation between the companies involved.

2. Concerted and consciously parallel practices as equivalent expressions and means of proof of such conduct Notwithstanding the above judgments, there are several cases where the TDC has not acknowledged a difference between concerted practices and consciously parallel practices, and has instead used both expressions as if they were equivalent. For example, in a recent judgment, dated 16 February 2005,46 the TDC found that the defendant companies involved had engaged in a “concerted practice or consciously parallel conduct in order to fix the prices” of their services. In this judgment the TDC uses both terms together every time, as a sort of redundancy, as if they were equivalent. The court does not refer, in any part of the judgment, to the distinction made in the Detergentes judgment. Furthermore, we cannot leave out the fact that the TDC has always pointed to some sort of “plus factor” when condemning companies for having pursued a consciously parallel practice. In fact, as explained above, in the 46

Judgment of the TDC of 16 February 2005 in Case 582/04, Autoescuelas Extremadura.


134

12/10/07

12:17

Page 134


Detergentes judgment, where it drew the clearest distinction between the different forms of conduct prohibited under Article 1 LDC, the TDC found in the end that the companies had been involved in a concerted practice. Similarly, in the two above-mentioned judgments issued under the LRPR (Asturiana de Zinc and Española de Zinc), in which the TDC established that the consensual element was not necessary for the existence of a consciously parallel practice, the court concluded that there was concerted or cooperative behaviour among the defendant companies. Moreover, the means of proof used by the TDC to establish the existence of consciously parallel practices have also contributed to a blurring of the boundaries between that concept and the concept of concerted practices. In particular, in a judgment issued on 11 May 1998,47 the TDC set forth the three elements that are necessary for a consciously parallel practice to exist: (i) the facts have to be sufficiently proven; (ii) there must be a causal link between these facts and the allegedly infringing conduct; and (iii) there cannot be another rational interpretation for the observed behaviour. Reading this in accordance with other TDC judgments on Article 1 LDC cases, it may be seen that these features are in fact the three steps necessary to demonstrate the existence of a consciously parallel practice. In the above-mentioned judgment of 11 May 1998, the TDC found that a consciously parallel practice could not be established. Although the first requirement was fulfilled, the second one was uncertain and the last one was not satisfied because there was an economic report that provided an alternative explanation of the parallel behaviour. The report established that the characteristics of the relevant market (i.e., the market for sales and rental of videotapes) favoured the existence of similar recommended prices but that in fact there was great dispersion between the prices actually applied. This showed that the apparent probative value of the recommended prices was deceptive. The TDC’s three-step rule is also used to establish the existence of concerted practices, since it is applied whenever there is no direct evidence of the infringement. A quite recent judgment may be referred to in this respect. In a judgment dated 23 September 2004,48 the TDC used the three elements described above to rule out the existence of a concerted practice on prices. The court stated that it is frequent in competition cases to rely on presumptions in order to demonstrate an infringement,49 mostly when the case deals with collusive 47

Judgment of the TDC of 11 May 1998 in Case 387/96, Películas de Vídeo. Judgment of the TDC of 23 September 2004 in Case 567/03, Servicios Agua Tenerife. 49 There are in fact several cases where the TDC has found the existence of a concerted practice consisting of price fixing by relying on presumptions: judgment of 13 June 2003 in Case 543/02, Transmediterránea/Euroferrys/Buquebús; judgment of 30 September 1998 in Case 395/97, Vacunas Antigripales; judgment of 18 December 1998 in Case 421/97, Autoescuelas Collado Villalba; judgment of 15 April 1999 in Case 426/98, Azúcar; judgment of 4 June 2001 in Case 492/00, Hormigón Gerona. Even under the LRPR, the TDC had already used this mechanism on many occasions in order to prove concerted practices relating to prices using the coincidence and 48


12/10/07

12:17

Page 135

Oligopolies, Conscious Parallelism and Concertation 135 practices. In this respect the TDC’s approach follows the case law of the Spanish Constitutional Court, which has stated that, under certain conditions, proof by presumptions may be used to establish certain facts that constitute an infringement. According to the case law of the Constitutional Court, the acceptance of presumptions as valid evidence in punitive law requires the three elements described above: (i) the indications or signs must be completely proven; (ii) the relation between those indications and the presumed conduct (the concerted or consciously parallel practice) must be sufficiently argued; and (iii) if there are other explanations for the signs or indications, they must be thoroughly analyzed and ruled out. In the above-mentioned judgment of 23 September 2004, the TDC stated that, if there was no direct evidence of an infringement of Article 1 LDC, it was necessary to examine the case under the principles of proof by presumptions in order to establish whether, on the basis of a total coincidence of prices among the companies involved, the conclusion could be drawn that there had been a concerted or consciously parallel practice—once again used as equivalent terms by the TDC. Regarding the first requirement of proof by presumptions, the TDC found the required factual indications in the form of documents contained in the file: it was established that the three defendant companies had applied identical prices (and variations) over a seven-year period. Nevertheless, the TDC pointed out that this was common in markets with homogeneous products, including the market for water in the case before it. Therefore, the TDC found that the alternative explanations provided by the companies had to be carefully analyzed and ruled out (third requirement) in order to establish an infringement, since the second requisite (the causal link) was also fulfilled. Finally, the TDC found that the explanations provided by the involved companies were highly credible and had not been appropriately ruled out during the investigation: the existence of a concerted practice could not be established. The explanations put forward by the parties, and accepted by the TDC, were based fundamentally on price leadership behaviour. By contrast, in an earlier case,50 the TDC used this proof by presumptions to establish the existence of a concerted practice relating to prices. In its judgment, the court stated that “concerted practices prohibited by Article 1 LDC are tacit agreements or forms of coordination between economic operators that cannot be expressly proven”. Therefore, it said, “it is necessary to rely on concrete signs or indications by which the existence of a tacit concertation between independent operators can be inferred, taking into account that tacit concertation, unlike written agreements, lacks express evidence”. simultaneity of their variations: judgment of 4 July 1988 in Case 235/87, Prensa dominical; judgment of 25 October 1988 in Case 237/88, Autoescuelas; judgment of 2 November 1988 in Case 239/88, Prensa del corazón and judgment of 8 July 1992 in Case 294/91, Aceites. 50 Judgment of the TDC of 4 June 2001 in Case 492/00, Hormigón Gerona.


136

12/10/07

12:17

Page 136


Thus, the TDC has used proof by presumptions in several cases to prove or rule out the existence of both concerted and consciously parallel practices. Nevertheless, proof by presumptions may have the undesirable effect for companies of obliging them to prove their innocence when their behaviour is, in some respect, “parallel” to that of their competitors. This would amount to reversing the burden of proof, which should always rest with the competition authority.51 In fact, it has been said52 that there is no other reason for the inclusion of consciously parallel practices in Article 1 LDC but to facilitate the establishment of an infringement by reversing the burden of proof. In other words, on this view it is not intended to introduce a new type of prohibited conduct. The TDC has endorsed this position, stating that the function of the concept of consciously parallel practices is to facilitate the proof of an agreement between the parties.53 This would support the interpretation that consciously parallel practices require concertation or coordination to fall foul of Article 1 LDC. The difference with other concepts embraced by this Article is that the consensual element exists but cannot be proved by direct evidence. In any event, Article 1 LDC should not, by virtue of proof by presumptions, be applied to conduct that amounts to a unilateral, logical, rational and competitive response to market conditions. It is important in this respect, in order for the TDC to establish that there has been an infringement of Article 1, to rule out the possibility that the parallel behaviour results from anything other than collusion. In fact, in the Detergentes judgment, the TDC stated that, for consciously parallel conduct to be an infringement of Article 1 LDC, it “should not be explained by normal or foreseeable reactions of the operators in the market”. That rule was actually applied in a judgment dated 5 July 1996,54 in which the TDC concluded that, under the circumstances of the case, the conduct of the banks involved was not a concerted practice because the parallelism of their conduct could be explained by the ”normal reactions of economic operators in the market”. By contrast, in a very recent judgment,55 the TDC did not give an adequate answer to the alternative explanations given by the parties for their allegedly infringing conduct. In fact, in its judgment, the TDC did not reach the third step of the proof by presumptions described above. It used presumptions as valid evidence based only on the first two requirements of the constitutional doctrine of proof by presumptions. The TDC considered that the indications 51 According to Article 2 of Regulation 1/2003, the burden of proving an infringement rests with the party or the authority alleging the infringement. 52 Pascual y Vicente J. (2000): “Las conductas prohibidas en la reforma de la Ley de Defensa de la Competencia”, 205 Gaceta Jurídica de la Unión Europea y de la Competencia 13. 53 Judgment of the TDC of 9 January 1987 in Case 223/89, Asturiana de Zinc. 54 Judgment of the TDC of 5 July 1996 in Case R 144/96, Tipos de Interés Bancario. 55 Judgment of the TDC of 10 May 2006 in Case 588/05, Distribuidores Cine.


12/10/07

12:17

Page 137

Oligopolies, Conscious Parallelism and Concertation 137 or signs of concertation were sufficiently proven and that there was a causal link between those indications and the alleged infringement. However, it did not proceed to rule out the alternative explanations offered by the companies involved. We believe that this step cannot be overlooked if an uncontrolled or exaggerated application of Article 1 LDC is to be avoided.

3. Conclusions that can be drawn from the case law of the TDC on concerted and consciously parallel practices Taking into account the discrepancy introduced by the expression “consciously parallel practices” in the relationship between Article 1 LDC and Article 81 EC, consciously parallel practices seem to be a rather awkward category of Spanish competition law. In fact, the definition of consciously parallel conduct given in the Detergentes judgment has not gone without criticism. The criticism stems from the fact that if there is no consensual element (coordination or cooperation) at all in the concept of consciously parallel practices, a problem arises as to where the frontier is, under Article 1 LDC, between a normal reaction to market circumstances—that is, a legal parallelism of conduct—and a “consciously parallel practice” prohibited by Article 1 LDC. Furthermore, the LDC’s stated purpose says that it has the aim of prohibiting restrictions to competition that arise from collusion. Therefore, in a coherent interpretation of the LDC, it cannot be sustained that Article 1 proscribes unilateral conduct. The proposal for a new LDC maintains in Article 1 the expression “consciously parallel practices”.56 However, while the title of Article 1 LDC as it stands is “Prohibited conduct”, in the proposed text Article 1 is entitled “Collusive conduct”. Thus, if collusion means coordination, cooperation or concertation, how can conduct that is unilateral be caught by a provision with such a title? Moreover, the apparent aim of the TDC in Detergentes to go further than Article 81 and to catch all restrictive practices that do not fall under Article 6 LDC (i.e., the equivalent of Article 82 EC) even if they are unilateral, seems to be incompatible with Regulation 1/2003. According to Article 3(2) of Regulation 1/2003, the application of national competition law may not lead to the prohibition of agreements, decisions by associations of undertakings or concerted practices which do not restrict competition within the meaning of Article 81(1) EC. Consequently, consciously parallel practices do not seem to have a place in Spanish competition law as a different category of prohibited 56 This inclusion was criticized by some of the commentators to the White Paper on the Reform of the Spanish System of Competition Defence.


138

12/10/07

12:17

Page 138


conduct (i.e., to the extent they are not equivalent to concerted practices), and we believe that the expression should be understood as a synonym for concerted practices, despite the fact that this interpretation essentially renders the expression redundant.

D. Conclusions Dealing with parallel conduct in oligopolistic markets is one of the most difficult tasks for competition authorities. Economic theory assumes that firms in an oligopoly can often perceive interdependencies facilitating strategic coordination. Although this strategic coordination may result in supracompetitive prices, this should not by itself lead to the conclusion that there has been illicit collusion. Since direct proof of illicit collusion is often difficult, when parallel behaviour is accompanied by facilitating practices, as a matter of practice competition authorities tend to shift the burden of proof to defendant undertakings. Although this attitude may seem logical, competition authorities must also remain open-minded to logical explanations of parallelism and to countervailing justifications of facilitating practices. Furthermore, competition authorities should not forget that parallel conduct may be a rational response by firms competing in an oligopolistic market, and that most facilitating practices can have both pro- and anticompetitive purposes. While competition authorities should use all legitimate means to prosecute illegal cartels, whether based on explicit or implicit collusion, condemning spontaneous parallelism prevents undertakings from adapting themselves intelligently to market conditions and implies unworkable remedies. An example of the doubts with which competition authorities address parallel conduct can be found in the different approaches applied by the Spanish competition authorities to conscious parallel practices, considered in some cases to include unilateral practices while in other cases they are treated as synonymous with concerted practices.

(I) Ehlermann 07 Barnett

12/10/07

12:17

Page 139

SESSION A


2 PANEL II: INVESTIGATION POWERS OF THE COMPETITION AUTHORITIES

WORKING PAPERS


12/10/07

12:17

Page 140


12/10/07

12:17

Page 141

I Thomas O. Barnett1 Seven Steps to Better Cartel Enforcement

A. Introduction Cartels are “the supreme evil of antitrust.”2 The fixing of prices, bids, output, and markets by cartels has no plausible efficiency justification; therefore, antitrust authorities properly regard cartel behavior as per se illegal or a “hard core” violation of the competition laws. While recognition of the dangers posed by cartels has reached something close to an international consensus, it is not enough merely to label cartels as harmful; cartelists have proven resistant to powers of persuasion or shame, perhaps because the anticompetitive rents available through cartel behavior can be so large. Cartels can only be deterred through vigorous prosecution. The United States has long experience prosecuting cartels and its efforts have yielded solid results. Some of those results are quantifiable: in fiscal years 2004 and 2005, and so far in 2006, the Antitrust Division of the Department of Justice has obtained fines of $350 million, $338 million, and $467.5 million, respectively, and has brought criminal cases against 63 firms. Some of the results are less tangible, but no less real: the Division has uncovered evidence that some cartelists choose to compete in the United States, even while continuing cartel behavior in other nations, due to the fear of US prosecution. The success of the United States model leads me to offer the following set of useful practices for the detection and prosecution of cartels.

1 Thomas O. Barnett is the Assistant Attorney General for Antitrust at the United States Department of Justice. Prior to his confirmation to that position in February 2006, he served the Department as Deputy Assistant Attorney General or Acting Assistant Attorney General since April 2004. Before joining the Department, Mr. Barnett was a partner at the law firm of Covington & Burling in Washington, D.C., where he served as vice chair of the firm’s Antitrust and Consumer Protection practice group. 2 Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 408 (2004).


142

12/10/07

12:17

Page 142


B. Seven steps for detecting, prosecuting, and ultimately deterring cartels The Antitrust Division’s anti-cartel enforcement program has been built over many years of dedicated effort. Based on our experience, each of the following seven practices has contributed to the success of the program: (1) focus prosecutors on “hard core” collusive activity; (2) treat cartels as serious crimes; (3) provide an amnesty program and “amnesty plus”; (4) vigorously prosecute obstruction of justice; (5) charge cartels in conjunction with other offenses; (6) provide transparency and predictability; and (7) publicize these enforcement efforts.

1. Focus prosecutors on “hard core” cartel activity The most important step in prosecuting cartels, and particularly in deterring them, is to make clear to all that anti-cartel enforcement is a priority. While this can be done in myriad ways, the United States chiefly relies on two strategies. First, we have separated criminal from civil enforcement, permitting a group of our attorneys to focus solely on cartels. Second, we have declared anti-cartel enforcement to be the highest priority in an explicit enforcement hierarchy.3 As to the first point, the separation of criminal and civil antitrust cases has a long history at the Department of Justice but the separation of attorneys into criminal and civil sections is a relatively recent development. During the first hundred years of US antitrust enforcement, the same attorneys often would investigate both a criminal and a civil theory in the early stages of a case. Even in the early 1990s—and even after the Antitrust Division’s attorneys were organized into sections based on industry segments or legal topics—most sections had the ability to bring both criminal and civil cases. This changed in 1994, when the Antitrust Division reorganized to create a separate criminal group. After further refinement, the current criminal enforcement organization of the Antitrust Division is: a front office Deputy for Criminal Enforcement (Scott Hammond); his second in command, the Director for Criminal Enforcement (Marc Siegel), and eight offices dedicated to anticartel work and located in Atlanta, Chicago, Cleveland, Dallas, New York, 3 E.g., Pate R. H. (2004): “Securing the Benefits of Global Competition”, address before the Tokyo America Center 4–6 (September 10, 2004), at http://www.usdoj.gov/atr/public/speeches/ 205389.pdf; see also Barnett T. O. (2004): “Antitrust Enforcement Priorities: A Year in Review”, address before the Fall Forum of the ABA Section of Antitrust Law 2–3 (November 19, 2004), at http://www.usdoj.gov/atr/public/speeches/206455.pdf.


12/10/07

12:17

Page 143

Seven Steps to Better Cartel Enforcement

143

Philadelphia, San Francisco, and Washington, DC. All told, we devote roughly 40% of the Antitrust Division’s attorneys to criminal enforcement. Does this organization make a difference? We believe that it does. People matter, and the creation of a specialized criminal enforcement team allows us to attract, train, and retain people whose passion is criminal enforcement. From the leadership to the line attorneys, our criminal enforcers are immersed in criminal law issues: they share practical advice with other prosecution branches of the government; they work closely with the United States Attorneys responsible for federal criminal enforcement in each US judicial district; they spend time on “details” (temporary duty assignments) as prosecutors in other Department components; they work with and learn investigatory skills from such entities as the Federal Bureau of Investigation; and in short, they live and breathe criminal law. There is a great benefit to the criminal side’s targeted focus.4 At the same time, the Division carefully delimits its criminal enforcement to focus only on hard core violations. The higher burden of proof in criminal cases (requiring proof “beyond a reasonable doubt,” as opposed to the “preponderance of the evidence” standard used in United States civil law) and the narrowness of what criminal enforcement condemns (the fixing of prices, bids, output, and markets, as opposed to the “rule of reason” or monopolization analyses used in civil antitrust law) establish clear, predictable boundaries for business. In addition, this narrow focus helps conserve prosecution and judicial resources by reducing the number of potential cases and also by reducing the complexity of proof: the focus on collusion largely removes the need for the detailed economic testimony common in civil antitrust actions. When criminal cases focus on conduct that has no plausible business justification and that usually occurs in secret, accompanied by preemptive cover-ups and misrepresentation, defendants cannot reasonably argue that they failed to grasp the illegality of their actions.5 All of these features—high burdens of proof, well-defined coverage, clear boundaries—allay the potential fears of law-abiding businesspersons, who can easily determine whether their own conduct will form the basis of a criminal case.6 4 And there is a corresponding benefit to having our civil enforcers specialize in their areas of greatest interest and expertise—we have a Networks & Technology civil section, a Telecommunications & Media civil section, and an Economic Analysis Group, to name just three of our other enforcement components. A list and brief description of the Antitrust Division’s sections and offices can be found at http://www.usdoj.gov/atr/sections.htm. 5 Cartels often use extreme measures to conceal their existence. The Division has uncovered cover-ups ranging from the creation of bogus trade associations, the use of code names, and sophisticated ruses to keep general counsel in the dark, to hiding incriminating evidence in the attic of a cartel member’s grandparent’s home, wholesale document destruction, and witness tampering after an investigation begins. Most of these were features of the lysine cartel, which the Division prosecuted in the 1990s. See generally http://www.usdoj.gov/atr/cases/archer0.htm; http://www.usdoj.gov/atr/public/press_releases/1996/0988.pdf. 6 These features also help to persuade courts to impose significant sanctions on cartelists once they have been convicted.


144

12/10/07

12:17

Page 144


My second point, our decision to make anti-cartel enforcement the highest priority in our antitrust hierarchy, is also a useful signal to industry. To be sure, we continue vigorously to protect competition in the areas of mergers and non-merger civil conduct, but we give special emphasis to cartel enforcement. The antitrust hierarchy also means that we have aligned enforcement priorities with our level of certainty about consumer harm: cartels are always harmful to consumers, whereas mergers and non-merger civil conduct are sometimes harmful but other times will lead to greater efficiencies that enhance consumer welfare. In the areas of mergers and non-merger civil conduct, antitrust enforcers need to account for the possibility of false positives, meaning the assumption of anticompetitive effects where, in fact, none exist or where the potential anticompetitive effects are outweighed by proconsumer benefits. The antitrust hierarchy helps enforcers to avoid deterring businesses from good, hard competition. I have heard the view that anti-cartel enforcement is too difficult for agencies that are relatively new to antitrust enforcement and that such agencies should concentrate only upon civil enforcement until they develop experience. I respectfully disagree. It is preferable to begin the anti-cartel process— even if one must learn from a few missteps along the way—than to let cartels function unimpeded, harming consumers. And as for the view that an agency in its early years should concentrate on civil conduct because such conduct is easier to find, this view reminds me of the man who, after a night of drinking, had lost his keys. A policeman finds the man searching unsuccessfully for the keys under a street lamp. The policeman, who also does not find the keys, asks “Are you sure you lost them here?” The man replies “Oh no, I lost them in that dark alley, but the light is much better for searching right here.” While conduct subject to civil investigations might be more open and therefore easier to see, the most consumer harm is to be found in the dark alleys of criminal cartels. Moreover, to take the analogy one step further, even though conduct in civil investigations might be easier to spot, discerning whether such conduct is beneficial or harmful to consumer welfare can be difficult, so things in the “light” might not be so clear after all.

2. Treat cartels as serious crimes, and cartel members as criminals The penalty for cartel behavior should fit the crime. Penalties should reflect the fact that cartels inflict enormous consumer harm7 with no likelihood of 7 See generally OECD (2002): Report on the Nature and Impact of Hard Core Cartels and Sanctions Against Cartels Under National Competition Laws, at http://www.oecd.org/dataoecd/ 16/20/2081831.pdf, Annex A (collecting estimates of consumer harm from numerous cartels). For example, the consumer harm from the citric acid cartel in the United States alone has been estimated at $100 million.


12/10/07

12:17

Page 145


145

corresponding efficiency gains. Penalties should also take into account the facts that cartelists’ motivation is strictly financial and that cartelists are quite capable of making a cost/benefit decision that factors in an occasional discovery and fine as a cost of doing their illegal business. Therefore, cartel penalties not only should be large enough to remove financial incentives, but also should look past financial incentives to affect cartelists directly: cartel penalties should include substantial jail time for responsible individuals. Our investigators have found that nothing in our enforcement arsenal has as great an effect as the threat of substantial incarceration in a United States prison— nothing is a greater deterrent and nothing is a greater incentive for a cartelist, once exposed, to cooperate in the investigation of his co-conspirators. Penalties are vitally important but so too are other elements of the criminal enforcement arsenal. Cartelists should be subjected to the same investigatory and apprehension practices that we apply to other suspected financial felons. This includes the use of the full range of criminal investigatory techniques (from search warrants to the use of informers wearing recording devices), the use of Interpol “Red Notices” and border watches, and extradition. Our criminal enforcement attorneys tell us that it is particularly important to use investigators trained specifically for criminal analysis, who on a weekly basis practice such techniques as criminal interrogation, forensics, and detection of corporate fraud, and who are comfortable operating in an environment in which they must assume that defendants might engage in evasion and the destruction of evidence. Criminal investigations are qualitatively different from most civil investigations, so it is wise to approach them from the outset with the assistance of dedicated criminal investigatory staff. Finally, coordination between international enforcement agencies is a particularly important priority in criminal investigations, due to the difficulty of detection, likelihood of evidence destruction, and risk of flight. The United States recently strengthened its laws against criminal cartels to reflect some of the principles mentioned here. The Antitrust Criminal Penalty Enhancement and Reform Act of 2004 (ACPERA)8 more than tripled the maximum jail term available under the Sherman Act, from three years to ten years, and increased the Division’s ability to seek substantial criminal fines. In November 2005, the United States Sentencing Commission increased the penalties allowed under the antitrust guideline, United States Sentencing Guidelines § 2R1.1, taking into account both the revised Sherman Act maximum jail sentence and recent experience with the enormous volumes of commerce affected by international cartels.9 The maximum jail term under § 2R1.1 is now nine years and a longer term is possible via adjustments from other guidelines sections, such as the “role in the offense” enhancement. 8

Pub. L. No. 108-237, Title II § 215(a), 118 Stat. 668 (2004), codified as 15 U.S.C. § 1. The new § 2R1.1 adds enhancements for affected volumes of commerce more than $250 million, $500 million, $1 billion, and $1.5 billion. 9


146

12/10/07

12:17

Page 146


While § 2R1.1 is not mandatory, courts are required to consider it when rendering judgment against criminal antitrust defendants.10 More recently, the United States also amended its investigatory statutes to provide wiretap authority in criminal antitrust investigations.11 Congress’s decision to grant that power is a signal that the United States places antitrust crimes on par with such other significant economic crimes as bribery, bank fraud, and mail and wire fraud. The United States has for some years made it a priority to obtain prison sentences for cartel participants wherever they reside. In our international cartel investigations, we have negotiated plea agreements with 27 foreign nationals to serve time in U.S. prisons. Also on this front, the United States has for the first time moved to extradite a criminal antitrust defendant. In June 2005, in The Government of the United States of America v. Ian P. Norris, a United Kingdom magistrates’ court found a British national to be extraditable on a U.S. antitrust charge. On February 24, 2006, the High Court rejected one appeal by the defendant. At the time of this writing, additional appeals are still pending. The United States’s efforts in the Norris case should send a powerful signal that cartelists will not be allowed to hide behind borders.

3. Provide amnesty and amnesty plus Although there are many reasons to create amnesty programs, by far the greatest is their value to detection. It is notoriously difficult to discover cartel behavior or, once discovered, to compile sufficient evidence to successfully prosecute cartel members in court. As one past Justice Department official noted, “[t]he most startling characteristic of the multinational cartels we have prosecuted is how cold blooded and bold they are. [T]hey [go] to great lengths to cover up their actions—such as using code names with one another, meeting in secret venues around the world, creating false ‘covers’—i.e., facially legal justifications—for their meetings, using home phone numbers to contact one another, and giving explicit instructions to destroy any evidence of the conspiracy.”12 To penetrate such elaborate concealment strategies, prosecutors must have a tool to convert cartel members into cooperative witnesses, so that prosecutors can gain access to background information, testimony, and the documents that otherwise might be destroyed. Amnesty programs are such a tool. 10

United States v. Booker, 543 U.S. 220, 245–46 (2005). 18 U.S.C. § 2516(1)(r). 12 Kolasky W. (2002): “DOJ Antitrust Compliance Programs: The Government Perspective”, speech before the PLI Corporate Compliance 2002 Conference 4–5 (July 12, 2002), at http://www.usdoj.gov/atr/public/speeches/11534.pdf. 11


12/10/07

12:17

Page 147


147

Amnesty programs work by changing the cost/benefit calculation of cartelization.13 To do this properly, an amnesty program requires three elements: First, there must be severe sanctions for firms and individuals that do not obtain amnesty—not only severe as a general matter, but also severe as a relative matter, in terms of what is lost by not cooperating. An amnesty program must provide significant benefits as compared with the alternative strategies of staying in a cartel or withdrawing but remaining silent. In the United States, the benefits are quite substantial; for example, successful amnesty applicants receive immunity from federal criminal antitrust prosecution and, as the result of ACPERA, are required to pay only single damages in any related private civil lawsuit, as opposed to the treble damages otherwise available under U.S. law.14 To make this element meaningful, the amnesty program must require significant, material cooperation and must allow prosecutors to terminate amnesty for non-cooperation or for continued participation in the cartel. Second, there must be a genuine fear of detection. Antitrust authorities must have a credible threat to discover cartel behavior even without amnesty applications by cartel members. If firms perceive that the risk of being caught by antitrust authorities is very small, stiff maximum penalties will not be sufficient to deter cartel activity or to cause firms to report their wrongdoing to authorities in exchange for amnesty. Once that credible threat exists, the threat of being turned in by one’s fellow corporate cartelists will increase. It is also very helpful to have an individual leniency policy, which creates the potential for an amnesty race as between a corporation and its own culpable employees. The real value and measure of the United States Individual Leniency Program is not in the number of individual applications we receive, but in the number of corporate applications it generates—when corporate management learns that an individual employee has committed cartel conduct, management takes an uncomfortable risk if it does not apply for amnesty immediately. And third, there must be predictability and transparency to the program. The United States program is set forth in two documents—the Corporate Leniency Policy15 and the Individual Leniency Policy16—that allow the program to function almost automatically, which gives potential applicants a high degree of assurance that, if they take the risk of coming forward, they will get the reward. Transparency in the amnesty program has the same benefits as transparency in the entire enforcement program, as I discuss further below. 13 See generally Hammond S. (2004): “Cornerstones of an Effective Leniency Program”, address before the ICN Workshop on Leniency Programs (November 23, 2004), at http://www.usdoj.gov/atr/public/speeches/206611.pdf. 14 15 U.S.C. § 15. 15 At http://www.usdoj.gov/atr/public/guidelines/0091.pdf (1993). 16 At http://www.usdoj.gov/atr/public/guidelines/0092.pdf (1994).


148

12/10/07

12:17

Page 148


An amnesty program or other enforcement practices must not give so much leniency that the deterrent effect of cartel penalties is diminished. This is an important principle to keep in mind when evaluating a request for a reduced fine from a “second-in” cooperator, meaning a cartel member who is not the first to expose or cooperate against a cartel but who may still provide valuable additional evidence and assistance. The Antitrust Division’s Deputy for Criminal Enforcement, Scott Hammond, recently published a presentation entitled Measuring the Value of Second-In Cooperation,17 in which he discussed the factors that go into assigning value to such cooperation. The United States does not provide absolute fixed discounts for second-in cooperation, as has been adopted, for example, in Japan; while there are benefits to the transparency of a fixed-discount approach, Mr. Hammond argues that those benefits are outweighed by the need for proportionality. Prosecutors are better served where they have the ability to adjust sentencing benefits to the actual amount of cooperation offered by second-in cooperators, which varies widely from case to case and may, in some cases, have no value at all. The discussion of second-in cooperation leads me naturally to a discussion of “Amnesty Plus,” which refers to benefits that prosecutors can offer to a cartel member who discloses previously undetected antitrust offenses, involving a cartel different from the one that first brought that cartelist to the prosecutors’ attention. Put more simply, Amnesty Plus induces firms that are already under investigation to clean house and report violations in other markets. Again, Mr. Hammond makes a strong case for such a program: Here is a remarkable statistic: roughly half of the Division’s current international cartel investigations were initiated by evidence obtained as a result of an investigation of a completely separate market. Most of the corporate defendants in international cartel cases are multinational companies selling hundreds of different products. It will come as no surprise then to learn that the Division’s experience is that if a company is fixing prices in one market, the chances are good that it is doing so in other markets as well. If an executive readily meets with competitors to allocate customers, then he or she has likely done it before in his or her career. And, if you go back further in time, you will likely find a mentor who taught the colluding executive the tricks of the trade. Armed with this experience, the Division has had great success engaging in a strategy of “cartel profiling” techniques aimed at ferreting out violations that sprout “cartel trees”—where one investigation will eventually give root to prosecutions in a half-dozen or more different markets.18

Our practice of rolling one investigation into another is well known in the antitrust community and should strike fear in the heart of cartelists. Through

17 Hammond S. (2006): “Measuring the Value of Second-in Cooperation in Corporate Plea Negotiations”, address before the 54th Annual Spring Meeting of the American Bar Association (March 29, 2006), at http://www.usdoj.gov/atr/public/speeches/215514.pdf. 18 Id. at 9–10.


12/10/07

12:17

Page 149


149

Amnesty Plus, exposure of a single member of a single cartel has the potential to bring a series of cartels tumbling down like a house of cards. As implemented by the Department of Justice, Amnesty Plus has its own set of rewards and penalties. As to rewards, the cooperating company not only receives the benefits of full amnesty for the newly revealed offense, but also receives a substantial additional discount in its fine for participation in the first conspiracy. The size of the additional discount depends on a number of factors, including: (1) the strength of the evidence provided by the cooperating company in the amnesty product; (2) the potential significance of the uncovered case, measured in such terms as the volume of commerce involved, the geographic scope, and the number of co-conspirator companies and individuals; and (3) the likelihood that the Division would have uncovered the cartel absent the self reporting—for example, if there is little overlap in the corporate participants involved in the original cartel and the Amnesty Plus matter, then the credit for the disclosure likely will be greater. And as to penalties, the Department applies “penalty plus”: if a firm under investigation fails to discover and report a second offense, then later finds itself negotiating a plea after the second conduct is discovered by the Division, it should expect significant consequences. If the Division learns that the company discovered the second offense and simply decided not to report it when it had a chance to qualify for Amnesty Plus credit, then the sentencing consequences will be even more severe. Therefore, companies understand that they cannot afford to remain willfully ignorant by limiting the scope of their internal investigations. The risks to the companies and their executives are too great.

4. Vigorously prosecute obstruction of justice As I have mentioned, cartelists are criminals and frequently are adept at concealment. True to type, many react to an investigation by actively obstructing the investigations of antitrust prosecutors. The cost/benefit mindset of the cartelist does not evaporate upon learning of an antitrust investigation; instead, the cartelist is likely to evaluate the potential costs of obstruction versus the “benefit” of impeding at least the scope of any resulting lawsuit. Accordingly, antitrust authorities should ensure that the cost side of the equation is severe. In May 2006, the Cartels Working Group of the International Competition Network released a paper entitled Obstruction of Justice in Cartel Investigations.19 Among its recommendations is that the punishment imposed for impeding a cartel investigation should be on par with punishment for the 19 At http://www.internationalcompetitionnetwork.org/capetown2006/ObstructionPaper-withcover.pdf.


150

12/10/07

12:17

Page 150


original offense. I endorse that view. Rather than discuss the problem of obstruction in greater detail here, I commend the report as important reading.

5. Charge cartels in conjunction with other offenses Prosecutors do not hesitate to combine charges in most areas of the criminal law; for example, they combine embezzlement charges with claims for illegal wire transfers and evidence destruction, among others. The same approach is appropriate for cartel offenses. The Division has uncovered cartel activity in conjunction with violations of laws against fraud, kickbacks, unfair government contracting, breach of fiduciary duty, and tax evasion, to name just a few. Antitrust prosecutors should have the power and the inclination to pursue a cartelist for each and every criminal violation, both to vindicate such proscriptions on their own merits and to induce cooperation against other members of the cartel. Leads developed through pursuing a collateral offense often lead to the exposure of a cartel offense, or vice versa. An effective enforcer will not ignore offenses simply because those offenses do not have the word “antitrust” in the title.

6. Provide transparency Antitrust enforcers should maximize the transparency and predictability of their enforcement policies. Informing market participants of the rules and likely consequences provides a critical foundation for effective deterrence. Business need to know where the line has been drawn so that they have an opportunity to conform their behavior to the law. Further, as discussed above, the success of an amnesty program depends on potential applicants having confidence that they will obtain the benefits of amnesty if they comply with a set of clear requirements. Similar considerations apply when negotiating plea agreements with defendants outside an amnesty program. The benefits of transparency reduce to a simple formula: in the United States’s experience, cooperating parties come forward in direct proportion to the predictability and certainty of their treatment following cooperation. It is generally advisable to establish transparent standards for all aspects of the criminal enforcement process, including the opening of hard core cartel investigations, the decision whether to bring charges, and the calculating and imposing of sanctions. Providing such a roadmap gives clear signals to industry. It also makes prosecutors’ jobs easier: not only do they have a clear


12/10/07

12:17

Page 151


151

procedure to follow internally, they also have something to point to when defendants seek unduly favorable settlements or accuse prosecutors of unfairness.

7. Publicize enforcement efforts Finally, it is important for antitrust enforcers to publicize their anti-cartel efforts. Deterrence is preferable to prosecution, whether as a matter of marketplace effects or of enforcement resources. Deterrence requires that violators learn the penalties they face and the rewards available if they confess. The goal is to destabilize cartels through the fear of harsh penalties, the incentive to cooperate and expose co-conspirators, and the recognition that enforcers are predictable and relentless in their approach. Over time, publicizing enforcement efforts can even change the norm of what is acceptable or tolerated in the marketplace. A chief difficulty faced by any enforcement program—antitrust or otherwise—is the tendency of deviant behavior, left unchecked, to become the norm. If criminal activity is presented as inevitable, some will simply accept it as a cost of doing business and a few will even seek to participate, seeing an opportunity for a share of anticompetitive gains. Aggressive enforcement combined with appropriate publicity helps break this cycle, reminding market participants that they do not have to tolerate the criminals in their midst. If publicity is particularly effective, it can lead both to more complaints and to complaints that are more actionable, as cartel victims and amnesty applicants learn to give enforcers the specific information necessary to make a case. And finally, let us not forget the reality that criminal anti-cartel enforcement is a government function, and government resources are always scarce. Antitrust enforcers should not be shy about publicizing the fact that anti-cartel enforcement is, on a dollarfor-dollar basis, among the best uses of law enforcement resources from the standpoint of consumer welfare.

C. Conclusion The ultimate goal of cartel enforcement is deterrence, and deterrence only works when consequences are real. To effectively deter cartels, antitrust enforcers must aggressively and predictably prosecute cartelists and use the full range of weapons in the enforcement arsenal, from fines to jail time to restrictions on international movement. All of these consequences affect the cost/benefit analysis of cartels, whether as a matter of the corporate bottom


152

12/10/07

12:17

Page 152


line or of the individuals who know they may serve time in jail. It is gratifying to us that some cartels avoid violating the law in the United States specifically because of our enforcement policies. This phenomenon illustrates that aggressive cartel enforcement can effectively deter such collusion. Further, as the number of countries with aggressive cartel enforcement programs increases, the effectiveness of each individual program should increase as well. The cartels will have fewer easy targets, a lower expected profit, a greater likelihood of detection, and a higher expected sanction. Accordingly, I encourage the international antitrust community to continue its efforts to expand and improve global cartel enforcement.

(J) Ehlermann 07 Collins

12/10/07

12:18

Page 153

II Philip Collins* Some Background Notes on the Investigative Powers of the Competition Authorities

A. How to treat facilitating practices? Facilitating practices can potentially be an important feature in the factual, legal and economic analysis of a cartel case where there is evidence of an express agreement or a concerted practice. But in the absence of such evidence of an express agreement or a concerted practice, facilitating practices are difficult to characterize legally as direct infringements of competition law. There are situations where such practices adopted unilaterally may be pro-competitive, leading to more transparent and efficient markets and to lower prices to consumers. US law shows that such practices cannot be attacked under the Sherman Act; rather, they have to be regarded as ancillary to a principal, collusive agreement (FTC v. Ethyl Corporation 1979). There are some examples of national authorities apparently finding facilitating practices to be direct infringements of competition law in themselves (see Italian Competition Authority Decisions no. 8553, Gasoline, concerning retail petrol prices and 8346 RC Auto, concerning car insurance information sharing). However, both cases also appear to be capable of being characterized as cases where the suppliers had adopted, in concert, a mechanism designed to produce the desired collusive outcome. Similarly, the EC Commission’s decision in UK Agricultural Tractor Registration Exchange (1992) may be regarded not as a case of “facilitating practices”, but rather as a case of an effects-based analysis of an information exchange agreement on competition in the specific market concerned. In the UK, there have not been any infringement cases to date where facilitating practices have been found in a cartel case. However, the OFT has attempted in the past to use available market intelligence to assess whether there may be cartel activity in certain sectors. In March 2005, we published a discussion paper (OFT733) commissioned from Professor Paul Grout of the University of Bristol.1 This was an empirical * Chairman, Office of Fair Trading, London, UK. 1 Grout P. and Sonderreger S. (2005): Predicting Cartels, OFT Discussion Paper of March 2005, available at: www.ft.gov.uk/NR/rdonlyres/4E9FFDC4-31F4-4F7D-896E-FFE635A136F3/0/OFT7 73. pdf.


154

12/10/07

12:18

Page 154


study into the economic and structural factors that are typically considered to contribute to the formation, maintenance or detection of cartels. The study was based principally on a survey of the economic literature and on market data and an analysis of past-published EC and US cartel decisions (since the early 1990s) and was designed to identify common economic factors that can help inform the process of deciding where to allocate effort in the process of detecting cartels. The report makes interesting observations on facilitating structures and behaviours that may be indicative of the existence of cartels. In the OFT’s cartel work, the most obvious type of publicly available information we see which may suggest facilitating practices and collusion are near-simultaneous headline price increases by the same or a very similar amount in concentrated markets. In one case, which did not proceed to a decision, we had received inconclusive complaints of price-fixing. However, an investigation of the sector using publicly available market data and further information requested from the complainants established that the following facilitating structures and behaviours were present in that market: —Highly concentrated market (4 to 5 players) —Cross-supply arrangements —Joint-venture arrangements —Actual and aborted M&A activities between the players —Significant price differentials between supplies in the UK and supplies in another country despite the fact that firms in both countries faced very similar input costs —Advance announcements of headline price increases —Parallel near-simultaneous subsequent price movements —Active trade associations However, having used our statutory powers, we found no evidence of actual coordination or collusion between the parties and we did not consider that the existence of the above ‘facilitating factors’ alone was in that case sufficient to establish an infringement. Our experience is therefore that, while such information on facilitating structures and conduct is undoubtedly highly relevant and informative, it cannot, by its very nature, be decisive. In a broader context, it should be noted that the use of a facilitating practice or practices can be a feature or part of a combination of features of a market in the United Kingdom which prevents, restricts or distorts competition and may be the subject of action as a result of a market investigation reference made by the OFT to the Competition Commission under the UK’s Enterprise Act 2002.2 2 See, in particular, OFT 511, Market Investigation Guidelines, Section 6, Firms’ Conduct, especially paras. 6.10 and 6.11 on facilitating practices.


12/10/07

12:18

Page 155

Notes on Investigative Powers of the Competition Authorities

155

B. The current administrative system of investigation: does it suffice? The current administrative system that has been broadly adopted and refined in Europe has been supplemented in more recent years by enhanced powers of inspection and, in some jurisdictions, such as the UK, by the introduction of parallel criminal procedures for (broadly) hard core offences involving dishonesty.

1. Civil administrative investigation procedures generally In Europe there is now a well-established system of civil administrative investigation procedures which normally relies on the following basic powers: —Formal written requests to undertakings to provide documents and explanations; and —Inspections of business premises with or without notice (these can be conducted, depending on the applicable legislative framework, under court warrants or simply with the authorization of the competition authority). In some countries, such as the UK, it is now also possible to inspect private homes if it is believed that relevant material is kept there. In fact, in the UK we have two types of powers to choose from when we need to attend premises to obtain information: —We can seek court warrants from the High Court to search business and residential premises; or —We can internally authorize OFT officers to attend business premises with or without notice and ask for the production of relevant documents; but this does not include a power to search. Increasingly, competition authorities across Europe, as in the UK, now have the power to ask individuals questions under compulsion. In the UK: —we can ask individuals to provide ‘specified information which the [OFT] considers relates to any matter relevant to the investigation’; —we can also always ask individuals, in writing or during an inspection of business or residential premises, to provide explanations of documents and to tell us where specified documents are located.


156

12/10/07

12:18

Page 156


2. The UK’s civil investigatory powers In the UK, the general investigatory procedures are supplemented by specific investigatory powers.

2.1. Directed surveillance Under the Regulation of Investigatory Powers Act 2000 (RIPA), the OFT has since 2003 had the power to authorize ‘directed surveillance’ in civil cartel cases. This covers any surveillance of individuals in public spaces for the purposes of a specific investigation that is: —covert (i.e. unknown to the target); and —likely to lead to the OFT obtaining ‘private information’ about an individual. Before authorizing such measures, we need to be satisfied that the recommended surveillance measure is necessary and proportionate and that so-called ‘collateral intrusion’ into the private affairs of third parties is kept to an absolute minimum. This power may for example be used to observe suspected cartelists meeting each other at a hotel. While we would be allowed to observe the individuals entering the hotel and even in the common areas, such as the lobby, we could not use the power of directed surveillance to follow suspects into a conference room, for example, and listen to any conversations. There is one other potentially interesting use of this power. Where one of two individuals participating in a telephone call consents to us taping that conversation, the exercise would be classified as ‘directed surveillance’ and the more stringent authorization requirements for interceptions of communications by telecommunication systems would not be triggered.

3. Informants: covert human intelligence sources or “CHIS” In the UK there is a specific authorization regime which allows the police and other designated agencies, including the OFT, to ‘task’ individuals to gather relevant intelligence or information. In the UK law enforcement community, such individuals are referred to as a ‘covert human intelligence sources’ or “CHIS”. In most cases, a CHIS will wish to remain anonymous for the duration of a case—or indeed indefinitely. As a result, we cannot use their information as evidence in our cartel cases, as we would be unable to disclose the source of that information. The information provided to us by a CHIS would therefore


12/10/07

12:18

Page 157


157

normally be used only as intelligence to direct the use of the investigatory procedures described earlier. To date, the number of approaches by informants to the OFT has been low. It may be therefore that we should consider the use of additional incentives to motivate people to come forward—in particular, financial incentives. Currently, we rely on people to blow the whistle on cartel conduct without any financial reward. We rely, instead, on their sense of public duty—or in some cases their willingness to satisfy a grudge against their employer! (Of course, in a potentially criminal case, a whistleblower may himself be potentially exposed to prosecution for the cartel offence, in which case the desire to gain immunity under our criminal immunity (or ‘no-action’) regime may provide the necessary incentive.) In the UK it is not unusual in serious crime cases (such as drug cases) for informants to be paid a sum of money for their cooperation. In most cases these sums would not ordinarily exceed £10,000. The OFT is currently considering whether to allow for small sums to be paid to informants in appropriate cases.3

4. Additional powers in criminal cases The only additional powers we have in the UK in criminal cases are the following: —The power to conduct compulsory interviews (s193(1) EA02) —The power to use intrusive surveillance (used to allow surveillance in private spaces) —The power to interfere with property (used to allow surveillance devices to be planted in private spaces). The intrusive surveillance power (coupled with the power to interfere with property) is rightly subject to very onerous external authorization requirements. While we welcome these additional powers and are prepared to make use of them in appropriate cases, the complexities associated with employing them should not be underestimated. 3 We understand that other countries have introduced such systems, but on a much larger scale. For instance, South Korea has allocated a sum of several million US$ for remunerating informants who provide the authority with useful information in cartel cases. Under this programme an individual can be paid up to US$ 1 million. However, payments of that scale (or even smaller scale payments) are likely to give rise, at least in common law systems, to tricky issues of motivation and ethics. Where a person who has been given a financial reward for information gives that information as evidence in court, it is likely to be attacked as unreliable and the prosecution may only be able to overcome that attack if there is evidence available which corroborates the informant’s account of events.


158

12/10/07

12:18

Page 158


5. Some implications of parallel civil and criminal regimes In the UK we now have the prospect of being able to conduct parallel civil and criminal investigations for hard core cartel cases. The question often arises whether introducing a criminal regime for cartel conduct—be it for undertakings and/or individuals—gives rise to a need significantly to strengthen an authority’s range of investigatory powers. In the UK, the powers available to the OFT for civil investigations of undertakings and criminal investigations of individuals are fairly similar. Indeed, to allow us to switch between the two regimes after an investigation has started, we are now making sure that all internal procedures are identical and conform with the higher criminal standards (e.g. note taking, document handling, records of chain of evidence). This had the positive effect that, as an authority, we strengthened all our investigation procedures as a result of criminalization in the UK.

6. Leniency and the importance of own-initiative investigations Within the current administrative system of investigation, there are various tools for making the system more effective and which enable cartels to be detected and administrative and other powers to be used. Corporate leniency is the most obvious and most well-known example. Leniency programmes are fast developing into the primary generator of cartel cases for competition authorities. However, one issue is how dependant a competition authority should become on its leniency programme. There is an argument that it is vital to ensure that competition authorities maintain a certain number of owninitiative investigations not triggered by leniency applications. If the authority has no track record of detecting cartels in the absence of leniency applications, then it could be argued that it would be irrational for anyone to apply for leniency, as the probability of independent detection (i.e. in the absence of leniency applications) would be zero. This argument has some force and is intuitively attractive. However, a number of considerations blunt the force of the argument: —Human beings do not always behave rationally. —The probability of detection is only one consideration in whether to apply for leniency. Others include: —Corporate compliance and ethics policy, which may have zero tolerance for illegal corporate conduct;


12/10/07

12:18

Page 159


159

—Increased corporate governance activity and systems (e.g. the role of non-executives on boards and the requirement for corporate governance reviews, procedures and statements, such as under Sarbanes-Oxley); —Identification of unlawful activity during the preparation or conduct of due diligence, on the vendor’s or the purchaser’s side, as a result of the business being put up for sale, with a resultant desire to quantify exposure and provide for allocation of potential liabilities; management changes, with new managers not wishing to inherit practices and liabilities and jeopardize their future careers. —The existence of individual whistle-blowers, typically disgruntled or sacked employees. —The risk of complaints from customers or leaks to, or speculation in, the trade or businesses press about a company or market. These factors will ensure that leniency-generated cases will remain a feature of antitrust enforcement no matter what in the future. However, despite this, it is the OFT’s view that no competition authority should ever give up its attempts proactively to detect cartel conduct, especially in sectors it has identified as ‘priority areas’ or where there are strong indications that markets are not working as well as they could and where there are suspicions of cartel activities. Thus, broadening the intelligence base, and encouraging complaints from customers and others, remains critical, as does ongoing market monitoring. For instance, the OFT is now obtaining substantial data which may be valuable in a competition law context from consumer complaints to its Consumer Direct telephone advice service.4 Market monitoring is carried out not only to inform our competition enforcement work but also to identify markets which are not working well for consumers and which may then be the subject of market studies carried out by the OFT or market investigations carried out on a reference by the OFT to the Competition Commission.

C. International cooperation against cartels Businesses are increasingly global and regionally managed; cartels are frequently international. Effective cooperation between competition authorities is vital to address and deter cartel activities.

4 The OFT took over responsibility for Consumer Direct on 1 April 2006. The service currently takes around 1.5 million calls per annum from consumers.


160

12/10/07

12:18

Page 160


1. Information exchanges The need for better and less burdensome procedures for international cooperation in cartel cases has been considered by many international bodies dealing with cartel issues, such as: —OECD, Working Party No. 3 on Co-operation and Enforcement of the Competition Committee of the OECD Directorate for Financial, Fiscal and Enterprise Affairs (DAFFE/COMP/WP3): ‘Recommended Practices for Formal Information Exchanges in International Cartel Investigations’, October 2005. —ICN, Cartels Working Group, Subgroup 1 ‘General Framework’: ‘Cooperation between competition agencies in cartel investigations’, May 2006. The necessary statutory framework is in place in the UK to allow the OFT to be an active partner in international anti-cartel enforcement. The UK framework for information exchanges is regulated by a statute under which the OFT has the power to disclose ‘specified information’ to an ‘overseas public authority’ for certain purposes. These include the investigation and bringing of proceedings in civil cases as well as the investigation and prosecution of criminal cases. Information exchanges across the ECN are regulated by the Network Notice (which includes special safeguards for the exchange of leniency and leniency-generated information between EC competition authorities). In the broader international sphere, many jurisdictions still have restricted gateways for disclosure of information derived from their cartel investigations, and this is an area on which further work is required, encouraged by work done within the OECD and the ICN.

2. Assistance with investigatory measures The OFT is relatively frequently asked to assist the Commission with investigations in the UK under Art 22(2) of Regulation 1/2003. The OFT has also carried out investigatory measures in the UK on behalf of other EC competition authorities under Art 22(1) of Regulation 1/2003. In addition, the OFT has on occasion assisted overseas authorities more informally by carrying out certain voluntary fact-finding measures (e.g. taking a voluntary statement from a UK-based individual).

(K) Ehlermann 07 Forrester

12/10/07

12:18

Page 161

III Ian S. Forrester, QC1 Searches Beneath the Cherry Tree in the Garden: European Thoughts on How to Enhance the Task of Uncovering and Thereby Deterring Cartels

A. Introduction These annual gatherings in the beautiful setting of the European University Institute have significantly shaped how competition law in Europe has been interpreted and structured. After an invigorating debate from 1996 to 1999 on the purpose and scope of the Treaty’s core provisions on unlawful agreements and the European Commission’s idiosyncratic interpretation of Article 81(1) and (3) EC, we moved to discussing how the Commission could achieve better and more consistent enforcement by harnessing the resources of other competition authorities and other agencies, without making unacceptable sacrifices in terms of quality. The debate about the wisdom of Regulation 1/2003 is now concluded, and we may presume that the regulatory basis for the new regime is not going to change significantly in the professional lifetimes of most of us.2 The topics selected for discussion at previous gatherings have been very challenging. However, a number of people have said to me this year that cartels are a rather routine topic, a matter of proof and evidence, not intellectually difficult. In one sense, it is true that the prosecution of cartel-like activity may be “merely” an issue of gathering evidence and demonstrating to the requisite standard of proof that punishable wrongdoing has occurred. But genuinely difficult choices are presented. Do we wish to maximise the number of cartels unearthed in the expectation that participants will discontinue? Do we wish to punish a maximum number of those who engage in cartel behaviour? Do we wish to maximize the pain of those who are discovered to have engaged in cartel activity? Is punishment a more or less important goal than cessation? Are the enforcement targets equally those companies (“undertakings” in the cumbersome language of the Treaty) and the men or women 1 Queen’s Counsel at the Scots Bar, Visiting Professor, University of Glasgow; White & Case, Brussels. Warm thanks are expressed to Anthony Dawes, Krzysztof Kuik, Kai Struckmann and others for their contribution to this paper. The opinions expressed are wholly personal. 2 Claus-Dieter Ehlermann presciently observed: “Regulation No. 17 contains all the basic procedural rules for Community competition policy. If and when it is revised, it will only be revised once for the foreseeable future.” See “Developments in Community Competition Law Procedures”, EC Competition Policy Newsletter 2 (1994).


162

12/10/07

12:18

Page 162


who are employed by those companies? Do we wish to maximise the chances of “genuine victims” (as opposed to those putative victims whose true situation gives their lucky counsel a ticket to financial rewards) to obtain indemnification? Should the penalties for such behaviour be savagely economically painful, more painful than negligently causing the death of a worker? These choices cannot be pursued simultaneously. (Indeed, I submit that we are not usually made aware that choices lie before us.) The current processes by which the European Commission investigates possible offences and decides what to do about them are inadequate to the preparation of criminal prosecutions. We cannot expect DG Competition officials to be trusted with powers to perform electronic surveillance; nor to be expected to prove their accusations to the level of near certainty demanded by criminal procedure; nor do they prove their case in a court; nor are we (yet) culturally disposed to regard cartels as they are regarded in the United States. However, I think most of us would agree on two propositions. First, the United States is the world leader in successfully identifying and prosecuting those who engage in cartels. Secondly, while “Europe” fully endorses the general principle that cartel activity is socially and economically undesirable, the repression of cartels is handicapped by a lack of sympathy for a number of aspects of US antitrust practice. It is interesting to reflect on a number of these emotional factors before making suggestions about how the Commission’s powers and priorities might change or be reinforced. This may help identify some weaknesses and opportunities in the current enforcement environment.

1. Cartels exist I begin with a personal anecdote. Cartels in Europe exist. I have by chance personal knowledge through a member of my family of one victim. He was the owner of a small British company in a wholesaling business in a consumer product. He was informed by an employee of one of the company’s suppliers that several competing suppliers of his wholesale business had decided that the company’s business was disrupting suppliers’ activities too much, and that the company must therefore be eliminated from the marketplace if it would not change its behaviour. The interlocutor was not insulting, impolite; indeed, he was rather sympathetic. But since the company could not change its behaviour, its business was effectively extinguished within four months. The stock was liquidated and the owners sought new jobs. The boss of the business wrote to DG IV describing what had happened. Some weeks later, a reply was received, courteously stating that because of shortage of resources and the need to set priorities the Commission would not be pursuing the complaint. That was in 1989. About fifteen years later, the Commission


12/10/07

12:18

Page 163

Enhancing the Task of Uncovering and Thereby Deterring Cartels 163 condemned a number of companies, including the former suppliers of the defunct business, for having engaged in cartel-like behaviour along the lines of the conduct complained of by my acquaintance. He is now endeavouring to extract some modicum of compensation for having seen his business ruined, but because of the applicable law on evidence of proof he faces formidable difficulty in obtaining any reparation. The cartel member is strenuously denying any duty to indemnify him. The relevance of this story is that it illustrates the difficulty of redress, and the virtual dependence of European victims on enforcement by the public authorities. Even when a complainant is armed with specifics, it is not easy to obtain redress. Although the availability of damages for breaches of the competition rules has been a theoretical possibility for thirty years and more, awards by courts have been rare. By far the best way of getting details of an infringement is from an insider. How to induce insiders to volunteer information involves exploiting a mixture of fear and self-interest. The Commission has an interest in encouraging those who are participants in industry-wide agreements to tell the public authorities; in encouraging non-participants to be aware of the illegality of engaging in such behaviour; in recalibrating the perceived benefit of coming forward with evidence by making penalties more predictable; and treating more fairly those who have chosen to volunteer information. However, I submit that the European Commission has been less successful than the US authorities in developing a regime to induce confessions. The goal of any cartel regime should be to induce a wise general counsel to rush to confess the very day that the counsel hears of a problem. To that end, a well-working leniency programme is crucial. Among the inhibiting factors which have made European leniency programmes less attractive are lack of confidence in the Commission’s fairness in setting fines; the dangers of creating huge civil liabilities in the United States; the relative slowness of the procedures and their non-transparency; and the lack of predictability of the steps following the making of a confession. It is extraordinarily interesting that the Commission is adopting creative means to palliate the problems occasioned by the distortions of analysis due to US civil litigation, notably the rules on treble damages. I shall be submitting that the Commission might consider additional targets, powers or practices that could assist its enforcement activity. Some of these suggestions are not very bold, and some are likely to be dismissed as unthinkable. In order for a leniency programme to be truly successful, it must offer a huge incentive to take the risk of volunteering the truth as opposed to waiting and hoping in silence. This implies total immunity from penalty if the public authority hears about it for the first time from the confessing party, a continuing receptiveness to receiving confessions from a participant even after the public authority knows of the cartel, as well as a readiness to extend concessions to late-comers in order to complete the public authority’s files


164

12/10/07

12:18

Page 164


revealing the nature of the infringement (Julian Joshua elegantly calls this the principle of only one first prize with less attractive prizes for runners-up). How should this review fit within the modernization package and the reshaping of the Commission’s administrative enforcement powers?3 The decentralisation aspect of modernisation increased the resources of the European Commission by enlisting the support of national competition authorities (NCAs) and national courts in the enforcement of EC competition law, and the procedural reforms actually provided the Commission with new tools. A review of Chapters V. to VIII of the new Regulation4 and of the accompanying Regulation 773/20045 and Notices readily demonstrates that in the new system the Commission enjoys a much stronger procedural framework for an effective enforcement of the competition rules.6 However, the Commission has sought to develop an array of new methods, outside and beyond the confines of Regulation 1/2003, to complement its new investigative arsenal. These initiatives have ranged from increasing reliance on NCAs and private enforcement, to the more novel possibilities of fining undertakings that facilitate a cartel and the use of trustees to monitor the implementation of its decisions. While these initiatives may be to ensure the “optimal enforcement” of the competition rules, it is open to debate whether they are appropriate. I invite us to reflect on the penalties visited upon those found to have engaged in cartel activities. The sums of money are immense: they have totalled €99 million (December 1999, steel tubes); €110 million (June 2000, lysine); €218 million (July 2001, graphite electrodes); €855 million (November 2001, vitamins); €313 million (December 2001, carbonless paper); €478 million (November 2002, plasterboards); €222 million (September 2004, copper plumbing tubes); €290 million (November 2005, industrial bags); €388 million (May 2006, hydrogen peroxide); €266 million (September 2006, road bitumen). Comparing the wages of different sins is always unsatisfactory, but the effort is not foolish as it reveals whether our society officially considers it worse to put out rubbish on the wrong day, to drive at an excessive speed, to assault another supporter at a football match, or to use hateful language towards an ethnic minority. By that imperfect standard it would seem that 3 See 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 173; Riley A. (2003): “EC Antitrust Modernisation: The Commission Does Very Nicely—Thank You! Part One: Regulation 1 and the Notification Burden”, 24 European Competition Law Review 657, at p. 604; Venit J. (2003): ‘‘Brave New World: The Modernization and Decentralization of Enforcement Under Arts 81 and 82 of the EC Treaty’’, 40 Common Market Law Review 545. 4 Regulation 1/2003 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ L 1 [2003]. 5 Regulation 773/2004 relating to the conduct of proceedings by the Commission pursuant to Articles 81 and 82 of the EC Treaty, OJ L 123 [2004]. 6 See Forrester I. (2004): “Modernisation: An Extension of the Powers of the Commission?”, in Geradin D., ed., Modernisation and Enlargement: Two Major Challenges for EC Competition Law, Intersentia, Brussels, pp. 83–97.


12/10/07

12:18

Page 165

Enhancing the Task of Uncovering and Thereby Deterring Cartels 165 attending industry-wide meetings at which prices and customers are agreed is an offence graver than causing death by negligent maintenance of a railway track, selling meat contaminated by fatal E-coli bacteria, or evading taxes or excise duties. It may be the case that the current administrative system of investigation has reached its limit. At this stage of the legal process, the more Europe moves to the criminalization of competition law infringements, the more it will be necessary to give better protection and elaborate more rigorous procedural rules. These other, probably unwelcome, reforms might sharply slow matters down, and would trigger calls for fundamental changes in how the Commission investigates and reaches decisions in quasi-criminal cases. This article will review the current investigative powers of the Commission in relation to its fight against cartels and whether they are sufficient. It will first compare the new powers given to the Commission by Regulation 1/2003 with those under Regulation 17/627 in the fight against cartels. It will then review the Commission’s recent initiatives to extend its investigative arsenal beyond the strict scope of Regulation 1/2003 through the use of novel tools, and then make some suggestions.

B. Current Commission Tools Under Regulation 17/62, the Commission had two main investigatory powers at its disposal: the power to request information in a two-stage process (Article 11 of Regulation 17) and the power to carry out inspections (Article 14 of Regulation 17). However, in the words of Regulation 1/2003, “The detection of infringements of the competition rules is growing ever more difficult, and, in order to protect competition effectively, the Commission’s powers of investigation need to be supplemented.”8 Consequently, Regulation 1/2003, while maintaining and enhancing these existing powers, granted the Commission a series of new powers. These include: —The power immediately to request information from undertakings (Article 18(1)).

7

OJ 13 [1962]. Recital 25 of Regulation 1/2003. See also Monti M. (2000): “Why and How? Why Should We Be Concerned With Cartels and Collusive Behaviour?”, speech delivered at the 3rd Nordic Competition Policy Conference Stockholm, 11–12 September 2000: “[. . .] in a global economy where infringements become more and more sophisticated, it is of a paramount importance that the Commission is properly equipped with investigative powers that allow it to effectively detect infringements of the Community competition rules. To do so, we must make the Commission’s inspection powers more biting.” 8


166

12/10/07

12:18

Page 166


—The power to interview any natural or legal person who consents to be interviewed for the purpose of collecting information relating to the subject matter of an investigation (Article 19). —The power to seal any business premises and books or records for the period under inquiry and to the extent necessary for the inspectors to ask any representative or member of staff for explanations on facts or documents relating to the subject matter and purpose of the inspection, and to record the answers (Article 20(1)(d) and (e)). —The power to search private homes, where the Commission has a reasonable suspicion that books or other records related to the business and to the subject matter of the inspection, which may be relevant to prove a serious violation of Article 81 or Article 82 EC, are being kept in any other premises. These increased powers are complemented by the significant increase in fines which the Commission may impose on undertakings for non-compliance: fines of up to 1% of global turnover under Article 23 to periodic penalty payments of up to 5% of global turnover under Article 24. The tariff prior to Regulation 1/2003 was in tens of thousands of Euros; a daily fine of €2 million per day is being threatened in June 2006 for alleged non-compliance against at least one company.9

1. Information requests Information requests (“Article 18 requests”) are widely used by the Commission as a means of obtaining information from companies and third parties. The Commission can now issue simple non-obligatory information requests to undertakings. Although not obligatory, these are unlikely to be treated as trivial or as capable of being disregarded. Alternatively, the Commission can adopt a formal decision, compelling the delivery of certain information. (At the same time, Article 18(4) now recognizes that such information can be supplied by a company’s legal representatives, although the company still remains fully liable if the information supplied is incomplete, incorrect or misleading.) At the same time, the case law10 has also recognized a limited right to undertakings against self-incrimination. Two early judgments, Solvay v. 9 Microsoft: a fine of €280,500,000 was imposed under Article 24 by a Decision of 12 July 2006. 10 Case 374/87 Orkem SA v. Commission [1989] ECR 3283; Case T-112/98 Mannesmannröhren-Werke [2001] ECR II-729; Joined Cases C-238/99 P, C-244/99 P, C-245/99 P, C-247/99 P, C-250/99 P to C-252/99 P and C-254/99 P Limburgse Vinyl Maatschappij and Others v. Commission (“PVC II”) [2002] ECR I-8375; Joined Cases T-71/03, T-74/03, T-87/03 and T91/03 Tokai Carbon Co. Ltd and others v. Commission of the European Communities (“Graphite Electrodes”) [2005] ECR II-10.


12/10/07

12:18

Page 167

Enhancing the Task of Uncovering and Thereby Deterring Cartels 167 Commission11 and Orkem v. Commission12 raise the question of whether there is something akin to the privilege against self-incrimination in responding to decisions requesting information under Article 11 of Regulation 17/62. The Court noted that Regulation 17/62 by its terms recognizes no such right, and indeed instead calls for active cooperation from the party investigated. Examining the question from the point of view of fundamental rights, the Court found that while the legal systems of the Member States generally recognise a right for a physical person in a penal procedure not to testify against himself, they do not recognise such a right for legal persons (companies) accused of infringing economic law, and competition law in particular. It reached similar conclusions with regard to the European Convention on Human Rights, and the International Convention on Civil and Political Rights. However, a somewhat different result was reached in examining the status of the rights of the defence: “While . . . the Commission is entitled to oblige the company to furnish all necessary information relating to facts of which it is aware and, if necessary to communicate to it the relevant documents in its possession even if these may be used to prove anti-competitive behaviour by it or by another company, it may not, by a request for information, compromise the rights of defence of a company. Thus the Commission may not oblige a company to furnish replies by which it would be induced to admit the existence of the infringement which it is for the Commission to prove.”13

Based on these criteria, the Court approved requests for factual information concerning meetings, the identity of the participants, and the documents relating thereto. Questions relating to initiatives on prices were approved to the extent that they related to the subject and modalities of these initiatives, but disapproved to the extent that they related to their objectives. This has been codified by Recital 23 of Regulation 1/2003.14 On the one hand, undertakings must actively cooperate with the Commission’s investigation. On the other, the Commission may not compel an undertaking to provide it with answers that may involve an admission on the undertaking’s part of the existence of an infringement. As to the so-called “gap” between the 11

Case 27/88 Solvay v. Commission [1989] ECR 3355. Cited above. See more generally, Forrester I. and Norall C. (1990), in Barav. A. and Wyatt D., eds., 1989 Yearbook of European Law, Clarendon Press, Oxford, p. 305. 13 Orkem, paras. 34 and 35. 14 “The Commission should be empowered throughout the Community to require such information to be supplied as is necessary to detect any agreement, decision or concerted practice prohibited by Article 81 of the EC Treaty or any abuse of a dominant position prohibited by Article 82 of the EC Treaty. When complying with a decision of the Commission, undertakings cannot be forced to admit that they have committed an infringement, but they are in any event obliged to answer factual questions and to provide documents, even if this information may be used to establish against them or against another undertaking the existence of an infringement.” 12


168

12/10/07

12:18

Page 168


Luxembourg Courts’ case law and Article 6 of the Strasbourg ECHR, Advocate General Geelhoed in the Graphite Electrodes appeal15 dismissed the notion that a request for documents is contrary to the right to remain silent, guaranteed by Article 6. I am not completely convinced that there is no problem of fundamental rights. In my view, it remains sensitive to compel a company to reply truthfully and helpfully to detailed questions and then to punish the company for what its answers reveal.

2. Power to take statements The power to take statements is not to be confused with the Commission’s power to ask for on-the-spot oral explanations during a dawn raid. Article 19 of Regulation 1/2003 grants the Commission the power to take statements from any natural or legal person. However, the extent of this power is fairly limited. The Commission must obtain the consent of the person to be interviewed. The information sought and collected from the statements must relate to the subject matter of the investigation. And Regulation 1/2003 does not provide for any sanctions if false or misleading information is given in these statements. Article 19 statements should not be underestimated: the discussion may be a bit tense, but the exchange can still be useful. I think experienced practitioners would generally say that a company is more likely to be cooperative and truthful than Commission officials tend to expect. Asking a question is likely to elicit information.

3. Dawn raids The Commission’s powers in relation to dawn raids under Article 20 of Regulation 1/2003 are largely similar to those previously granted under Article 14 of Regulation 17/62. Most adjustments and changes have been in favour of the Commission. For example, Article 20(2)(b) now reads “officials and other accompanying persons authorised by the Commission to conduct an inspection are empowered to examine the books and other records related to the business, irrespective of the medium on which they are stored,” confirming the Commission’s previous contention that it could inspect all

15 See the opinion of AG Geelhoed of 19 January 2006 in Case C-301/04 P Commission of the European Communities v. SGL Carbon (Graphite Electrodes appeal), paras. 65–67.


12/10/07

12:18

Page 169

Enhancing the Task of Uncovering and Thereby Deterring Cartels 169 records, even in electronic format. Similarly, Article 20(2)(c) now allows the Commission “to take or obtain in any form copies of or extracts from such books or records”, there being no need for the Commission to get permission to use the undertaking’s photocopying facilities. Article 20(2)(d) of Regulation 1/2003 is entirely new and allows the Commission to “seal any business premises and books and records”. The seal can be placed “for the period and to the extent necessary for the inspection” and Recital 25 confirms that “seals should normally not be affixed for more than 72 hours”. Article 20(2)(e) allows the Commission “to ask any representative or member of staff of the undertaking or association of undertakings for explanations on facts or documents relating to the subject-matter and purpose of the inspection and to record the answers”. Inspectors can therefore target the people they believe are the best placed to give answers to their questions. A reliable urban myth tells us that there is a cherry tree in the garden of the director of a German company beneath which snooze embarrassing documents related to a serious infringement of the competition rules. And so last but not least, “in order to safeguard the effectiveness of inspections” Article 21 extends the Commission’s power of investigation to the search of the private homes of individuals. According to Recital 26, this was because “[e]xperience has shown that there are cases where business records are kept in the homes of directors or other people working for an undertaking”. Many practitioners would not disagree. The decision to conduct such a search cannot be executed without prior authorisation from the national judicial authority of the Member State concerned. The anxiety of the drafters is clear enough—to avoid a sympathetic court’s annulling the Commission’s investigative initiative. A national court requested to issue such a warrant cannot review either the legality of the Commission’s decision or the necessity of the inspection. Nor may it request access to the content of the Commission’s file. All it is free to consider is whether the Commission’s decision is authentic and whether the envisaged coercive measures are neither arbitrary nor excessive with regard to the subject matter of the investigation.16 Fundamental rights of defence during a dawn raid include the right not to be subject to an unauthorized investigation, the right to obtain legal advice, the right to withhold the provision of legally privileged documents (that category is the subject of debate) and the right to protection against selfincrimination. However, the penalties for non-cooperation with a dawn raid are fierce: an undertaking may be liable for fines up to 5% of its average daily turnover in the preceding business year.

16 This article is a codification of the judgment in Case C-90/00 Roquette Frères [2002] ECR I-9011.


170

12/10/07

12:18

Page 170


4. Are companies rightly feared to be rogues? I suspect that the Commission underestimates the degree of scrupulousness with which corporate candidates for investigations, especially very large ones, conduct their affairs. More specifically, any large company is under multiple duties to report to various agencies, to have systems in place that will facilitate the detection of improper or corrupt payments, and to ensure compliance with environmental, regulatory, stock exchange, fiscal and safety rules. From my own experience I submit that Commission officials who conduct inspection visits are likely to suspect, or even to expect, that destruction of evidence will occur unless the inspectors rapidly take precautions to secure the premises. Though such fears are understandable, and in certain past cases well founded, I believe they are mistaken in the vast majority of cases. I further submit that there have been occasions when raids were conducted not to discover vulnerable evidence but to demonstrate the Commission’s toughness or energy, or perhaps to suggest that the target company may be guilty of something.17

5. Leniency Confessions are the commonest method through which the Commission first learns about the existence of a cartel that it is subsequently able to investigate. Leniency is thus an upstream asset, the attractiveness of which is essential for Commission policy. The US was the first jurisdiction to realise the potential of such a system in 1978. The Department of Justice’s 1993 amnesty programme for whistleblowers has had great success in revealing the existence of many cartels.18 The EU’s experiments with leniency have had more uneven success. In the first version of the Leniency Notice, adopted in 1996,19 there was no guarantee for the first company that came forward with “decisive evidence” of the existence of a cartel that it would obtain 100% immunity. Instead, the 1996 Notice provided that the undertaking could obtain either complete immunity from fines or a reduction in the level of the fine of at least 75%. Since the level of fines was not revealed till the moment when the

17 I name as an example the now ancient case of football, when national football associations were “raided” for information about the charges made to suppliers of footballs for affixing a tournament logo, information which was neither secret nor sensitive, and could have been obtained by a telephone call or a letter. 18 For example, both the Citric Acid cartel and the Vitamins cartel were brought to light because of whistleblowing. 19 Commission Notice on the non-imposition or reduction of fines in cartel cases, OJ C 207 [1996].


12/10/07

12:18

Page 171

Enhancing the Task of Uncovering and Thereby Deterring Cartels 171 Commission actually issued its final decision,20 the promise of a discount from an unknown and unpredictable figure was not very meaningful. The Commission chose to adopt a new Notice in 2002 21, a more sophisticated approach to its treatment of whistleblowing. The 2002 Notice now guarantees 100% immunity to the initial whistleblower, provided the whistleblower complies with several other conditions.22 The 2002 Notice has created incentives for undertakings to confess.23 But the increasing use of leniency by companies confronting cartel problems has raised a number of questions.

5.1. The discoverability of EU leniency statements in US civil proceedings Potential leniency applicants are increasingly reluctant to seek leniency in Brussels because of the risk that the information provided to the Commission may be discoverable in actions brought by third parties in the US,24 as this will make more likely civil exposure to treble damages claims in the US.25 Tactical choices about what to do, when and how are greatly influenced by the US situation. Mandatory treble damages awards to successful civil antitrust plaintiffs have been a cornerstone of US civil antitrust enforcement since the enactment of the Sherman Act in 1890. The Clayton Act incorporates the treble damages provision. Congress created treble damages awards in order to advance four goals: (1) compensation; (2) deterrence; (3) disgorgement of illegal profits; and (4) punishment.26 From 1890 to 1940, plaintiffs brought just 175 private actions. But from 1941 to 1985, plaintiffs brought nearly 30,000 private actions.27 The 20 See Joshua J. (2004): “EC Fining Policy Against Cartels After the Lysine Rulings: The Subtle Secrets of X”, Global Competition Review 5. 21 Commission Notice on Immunity from fines and reduction of fines in cartel cases, OJ C 45 [2002]. 22 Provided the whistleblower fully cooperates on a continuous basis throughout the procedure, immediately puts to an end to its participation in the cartel and has not coerced other undertakings to participate in the cartel. 23 See Jephcott M. (2002): “The European Commission’s New Leniency Notice—Whistling the Right Tune?”, 23 European Competition Law Review 378, at p. 380; Arp J. and Swaak C. (2003): “A Tempting Offer: Immunity From Fines for Cartel Conduct Under the European Commission’s New Leniency Notice”, 23 European Competition Law Review 9. 24 Although the potential worldwide exposure of undertakings regarding the ability of foreign purchasers to assert treble damage claims under the US antitrust laws for injuries sustained in foreign commerce where the underlying anti-competitive “conduct” also caused effects in the US marketplace has been limited by the US Supreme Court in Hoffmann-La Roche, Ltd. v. Empagran S.A, 2004 WL 1300131 (2004). Foreign purchasers are now precluded from bringing suit in US courts where their foreign injuries are “independent of any adverse domestic effect.” 25 See Nordlander: K. (2004): “Discovering Discovery—US discovery of EC Leniency Statements”, 24 European Competition Law Review 646. 26 Blue Shield of Virginia v. McCready, 457 U.S. 465, 472 (1982). 27 Jones C. (2004): “Exporting Antitrust Courtrooms to the World: Private Enforcement in a Global Market”, 16 Loyola Consumer Law Review 409, at pp. 410–11.


172

12/10/07

12:18

Page 172


wisdom of treble damages is by no means generally endorsed.28 In criminal cases, the US government may seek fines equal to double the conspirators’ gain or the victims’ loss. Additional civil treble damages awards could in theory result in penalties five times greater than the injury.) The prospect of huge recoveries provides a lucrative incentive for plaintiffs’ lawyers to bring claims. Indeed, I am told by fellow practitioners in the US that “normal” business tort claims have been re-characterized as antitrust claims in order to increase the pressure and the potential gains. I suggest that the concept of damages as a remedy for victims of antitrust law breaches cannot credibly be challenged. US arrangements can credibly be challenged.

5.2. The litigator as privateer A distinctive characteristic of the US court system is the use of litigation as a vehicle for the personal enrichment of lawyers acting for plaintiffs whose relationship with the lawyer is that of asset rather than client. Entrepreneurial counsel in personal injury cases may hire cameramen dressed as firemen to photograph smoking debris (I knw personally of one such episode!); they find persons susceptible of being called experts on various disciplines and have on at least one matter procured from these experts written statements confirming the existence of damage, sometimes scores or hundreds of such statements being issued in a very short time; and in many cases they seek discovery of immense volumes of documents from the defendant, partly to inflict administrative pain and partly (secondarily maybe) to seek for damaging evidence. The vaster the class of plaintiffs, the greater the burden on the defendant. These are very broad generalizations, and of course meritorious claims exist alongside weak claims, and of course the plaintiffs’ bar consists of talented lawyers zealously advancing their clients’ interests. Nevertheless, it is reasonable to note a widely spread sentiment in Europe and elsewhere that Things Have Gone Too Far. One vivid example of litigious excess that still drags on are the claims brought against manufacturers of asbestos-related products. 850,000 claims were brought, most of the later ones on behalf of claimants without symptoms, and indeed without awareness of any lung impairment. A number of large enterprises have been rendered insolvent by the claims, and many employees lost their jobs as a result.29 The benefits to victims of asbestosis, in 28 See, for example, Cavanagh E. (1987): “Detrebling Antitrust Damages: An Idea Whose Time Has Come?”, 61 Tulane Law Review 777, at p. 789; Easterbrook F. (1985): “Detrebling Antitrust Damages”, 28 Journal of Law and Economics 445. 29 Brickman L. (2005): “An Analysis of the Financial Impact of S. 852: The Fairness in Asbestos Injury Resolution Act of 2005”, 27 Cardozo Law Review 991, at p. 992. Asbestos litigation has been described as “the gift that just keeps on giving to trial lawyers.” Remarks of Bob Nardelli, White House Economy Conference Discussing Lawsuit Abuse (December 15, 2004), available at http://www.whitehouse.gov/news/releases/2004/12/20041215-11.html.


12/10/07

12:18

Page 173

Enhancing the Task of Uncovering and Thereby Deterring Cartels 173 the sense of having contracted a disease caused by workplace exposure to asbestos have been uneven and disputed. The economic damage to many employees of otherwise prospering companies is significant and undisputed. The Antitrust Criminal Penalty Enhancement and Reform Act was enacted in June 2004.30 The legislation limited the civil liability of a cooperating leniency applicant by “detrebling” damage awards to actual damages. This legislation was driven by concerns that the imposition of civil treble damages on top of criminal fines discouraged potential applicants from cooperating with the government. Senator Orrin G. Hatch remarked that despite the success of the DOJ’s Leniency Program, “a major disincentive to self reporting still exists[:] the threat of exposure to a possible treble damage lawsuit.”31 Senator Herb Kohl—maybe optimistically—also stated that the detrebling provision “will result in more antitrust wrongdoers coming forward to reveal antitrust conspiracies, and thus the detection and ending of more illegal cartels.”32 The Standards Development Organization Advancement Act33 provided that standards development organizations (“SDOs”) whose activities have been disclosed to, and approved by, the DOJ and FTC would be subject to only single damages. During debate in the Senate, Senator Hatch observed that “the threat of treble damages deters SDOs from their pro-competitive standard-setting activities.”34 Similarly, in the House of Representatives, William Delahunt argued that the detrebling provision of the legislation would remedy the fact that SDOs were frequent targets of frivolous lawsuits. He stated that despite their “pro-competitive effects, these SDOs can find themselves named as defendants . . . Once they are sued, these organizations are forced to expend considerable resources on protracted discovery proceedings before they . . . finally . . . prevail on motions for summary judgment which occurs in 100 percent of the cases, from my information.”35 So the wisdom of treble damage claims is doubted by those with real expertise.

5.3. Current practice in handling leniency requests in Washington and Brussels The Antitrust Division has a policy of granting leniency in certain conditions to corporations which report their unlawful conduct on certain conditions. It grants automatic leniency from criminal prosecution to the first corporation that reports it automatic leniency from criminal prosecution, and a limitation 30 Pub. L. No. 108–237, §§ 201–21, 118. Stat. 661 (2004) (codified in scattered sections of 15 U.S.C.). 31 150 Cong. Rec. S3610-02, S3614, 2004 WL 714783 (April 2, 2004). 32 Id. at S3616. 33 15 U.S.C. §§ 4301–05 (2006). 34 150 Cong. Rec. at S3614. 35 149 Cong. Rec. H5104-01, H5106, 2003 WL 21339488 (June 10, 2003).


174

12/10/07

12:18

Page 174


of liability to victims of actual damages. It will grant leniency, before the launching of its own investigation, if the corporation took prompt and effective action to terminate its part in the activity and if it reports its conduct with full cooperation. After an investigation has begun, a corporation may also receive leniency if the DOJ does not have evidence against the corporation likely to result in a sustainable conviction. There are other limiting conditions. Most importantly, directors, officers, and employees also receive leniency if they cooperate.36 Since amnesty is granted to the first-comer, there is a race to confess. Counsel, who is trying to digest the story, clarify the facts and get management to reach a decision, may telephone the DOJ to get a “marker” to secure the corporation’s place in the queue. A successful amnesty claim of course eliminates threat of jail for corporate executives.37 Thus, crucial to the success of the US leniency regime is the invocation of criminal sanctions against company executives; their natural desire to avoid jail is a powerful driver of leniency applications; The company’s interests are exposed to a serious threat in exchange for the removal of a different, but even more acute, threat to its executives. This phenomenon, at the heart of the Washington regime, has no counterpart in Brussels. Civil damages claims are accelerated by the seeking of leniency, which purchases no immunity in their regard. Now, there is a deterrent to seeking leniency in Brussels: will it accelerate liability in the United States? The European Commission has confirmed its intention to reinforce the protection of leniency applicants from disclosure. It has reasserted that any written statement made under the 2002 Leniency Notice forms part of the Commission file and, as such, may not be disclosed or used for any other purpose than the enforcement of Article 81 of the EC Treaty. Secondly, the Commission will continue to intervene as amicus curiae in civil proceedings in the US where discovery of leniency corporate statements is at issue.38 These may be welcome but are not likely to be effective. Finally, and most interestingly, undertakings can make corporate statements in oral form in order to avoid the threat of discovery of written statements in civil litigation. Under the 2002 Notice, a corporate statement can take the form of a written document signed by or on behalf of the undertaking or of a statement made in oral form, of which the Commission makes a recording

36 Dep’t of Justice, Antitrust Div., Corporate Leniency Policy (1993), available at http://www.usdoj.gov/atr/public/guidelines/0091.pdf. 37 Hammond S. (2001): “When Calculating the Costs and Benefits of Applying for Corporate Amnesty, How Do You Put a Price Tag on an Individual’s Freedom?”, presentation at National Institute on White Collar Crime 3–8 (Mar. 8, 2001), available at http://www.usdoj.gov/atr/ public/speeches/7647.htm. 38 The Commission has intervened before the US Courts as amicus curiae in the course of litigation in both the Vitamins (In Re Vitamins Antitrust Litigation, Misc. No. 99-197 D.D.C. Jan. 23, 2002 and Sept. 30, 2002) and Methionine cartel cases (In Re Methionine Antitrust Litigation, No. C-99-3491-CRB N.D.Cal.).


12/10/07

12:18

Page 175

Enhancing the Task of Uncovering and Thereby Deterring Cartels 175 and a transcript. For example, in its Citric Acid 39 Decision, it granted a leniency applicant a 90% reduction on the level of its fine on the basis of the oral evidence that the applicant had disclosed during a meeting with the Commission. The legitimacy of such a practice has now also been confirmed by the CFI in the Graphite Electrodes case where it stated that “not only ‘documents’ but also ‘information’ may serve as ‘evidence’ which materially contributes to establishing the existence of the infringement”.40 To formalize this practice, the Commission has recently proposed an amended version of the Leniency Notice.41 “The main purpose of the procedure for corporate statements is to provide, in a clear, transparent and legally secure manner means by which undertakings that want to confess their participation in a cartel to the Commission can do so without undue fear of such corporate statements being used against them in procedures that are not in application of the European competition rules.”42

Under the new procedure, companies would make oral confessions to the Commission, which will record and prepare its own transcript. Applicants will then be required to confirm that the transcript is a correct rendering of the oral statement. This confirmation may be given orally and recorded. The transcript will serve as evidence, supplemented if necessary before the Community Courts by the original recordings. The draft amendment also makes clear that there would be very limited access to the file and that no other mechanical copies of the oral statements can be made. However, whether such an amendment will be sufficient to avoid disclosure in all circumstances is open to question. As has been highlighted by Kerse & Kahn,43 the fact that a document is created by the Commission and that a leniency applicant retains no copy of the document does not necessarily mean that discovery of the corporate statement will be impossible in civil proceedings in Europe or the United States. The Commission has stated that under normal circumstances, it will not accept requests under Regulation 1049/200144 for access to documents that it receives under the 2002 Leniency Notice. This policy has been the subject of a challenge in Case T-2/03 Verein für Konsumenteninformation v. Commission45 where a national consumer association requested access to the

39

OJ L 239 [2001]. Joined Cases T-236/01, T-239/01, T-244/01 to T-246/01, T-251/01 and T-252/ Tokai Carbon v. Commission, [2004] ECR II-1181, at para. 431. 41 Proposed amended 2002 Commission Notice on Immunity from fines and reduction of fines in cartel cases, http://ec.europa.eu/comm/competition/antitrust/legislation/leniency2_en.pdf. 42 Draft amendment of the 2002 Commission Notice on Immunity from fines and reduction of fines in cartel cases, http://ec.europa.eu/comm/competition/antitrust/legislation/leniency_en.pdf. 43 EC Antitrust Procedure, 5th edition, Thomson (2005), at p. 422. 44 Regulation 1049/2001 regarding public access to European Parliament, Council and Commission documents, OJ L 145 [2001]. 45 [2005] ECR II-1121. 40


176

12/10/07

12:18

Page 176


Commission’s file in the context of the Austrian Banks 46 cartel case. The Commission refused access to parts of the file, including in relation to documents submitted under the 1996 Notice because it argued that “allowing third parties access to those documents would deter undertakings from cooperating with the Commission and would be detrimental to inspections and investigations in future cases. The same reasoning applies to documents drawn up by third parties.”47 However, the CFI, while not ordering access to specific documents, stated that in principle, the Commission must carry out a concrete assessment of the content of each individual document in order to determine whether access to a document can be denied under the specific exceptions of Regulation 1049/2001 and that “[. . .] it is only in exceptional cases and only where the administrative burden entailed by a concrete, individual examination of the documents proves to be particularly heavy, thereby exceeding the limits of what may reasonably be required, that a derogation from that obligation to examine the documents may be permissible.”48

The Commission must therefore make an individual analysis of a request for access to each individual document and may be forced to allow discovery of certain documents in cases of private litigation where a leniency applicant has been found to have breached Article 81 EC.

5.4. Parallel leniency applications in the European Competition Network There is said to be some uncertainty about problems associated with the possible exchange between the other members of the European Competition Network of the information contained in a leniency application made to one EU jurisdiction.49 Although the Notice on cooperation within the Network of Competition Authorities50 provides answers to a certain number of concerns, it remains silent on others. On the one hand, the Notice clearly states that: —An application for leniency to one authority will not be considered as an application for leniency to any other authority. If this is taken at face value, it would be in the interest of an applicant to apply for leniency to all competition authorities that may be affected by the infringement.51 —Under Articles 11(2) and 11(3) of Regulation 1/2003, the Commission and the NCAs are under a reciprocal obligation to inform each other of cases 46 Commission decision of 11 June 2002 relating to a proceeding under Article 81 of the EC Treaty (Case COMP/36.571/D-1: Austrian banks—“Lombard Club”), OJ L 56 [2004]. 47 Para. 81 of the judgment. 48 Para. 112 of the judgment. 49 Swaak C. and Mollica M. (2005): “Leniency Applicants to Face Modernisation to EC Competition Law”, 26 European Competition Law Review 507. 50 OJ C 101 [2004]. 51 Para. 38 of the Notice.


12/10/07

12:18

Page 177

Enhancing the Task of Uncovering and Thereby Deterring Cartels 177 which have been initiated on the basis of leniency applications. However, information relating to leniency applications submitted to the network by one member of the network cannot be used by other members to open their own investigations.52 —However, this information may be transmitted to another NCA if the leniency applicant gives its consent to such a transfer. NCAs are under a duty(!) to encourage leniency applicants to give this consent.53 —The undertaking’s consent is not required if another member of the network has (i) also received a leniency application from the undertaking relating to the same infringement or (ii) given a written commitment that neither the information transmitted or any other information it may obtain, may be used to impose sanctions on the leniency applicant, or on any employee or former employees.54 The Notice does not decide how leniency applications in one jurisdiction may be affected by the potential reallocation of cases between members of the Network: —Reallocation from Commission to NCA: where an undertaking, having approached the Commission and made a leniency application solely to it in the belief that the Commission would have sole jurisdiction, then sees the case being subsequently reallocated within the ECN to an authority where the leniency applicant had not or could not have applied for leniency. —Reallocation from NCA to Commission: conversely, after a leniency applicant has made a leniency application to one NCA, the case is reallocated to the Commission to which no leniency application had been made. —Reallocation between NCAs: a leniency applicant, has sought leniency before NCA’s with a leniency programme, but then sees the case is reallocated to an NCA which either does not operate a leniency programme or operates a programme for which the company is ineligible. I suspect these concerns are more theoretical than practical but it may be too early for confidence to be possible. It would be strange if a corporate group had to make applications to all 25 Member States, each with their own distinct features and procedures, and of course to the Commission! Certain Member States still do not even have a leniency programme. Consequently, it has been suggested that either a new Notice is required or that a “one-stop leniency shop” should be developed. This question can be revisited in a few years: it may prove to be a non-problem.

52 53 54

Para. 39 of the Notice. Para. 40 of the Notice. Para. 41 of the Notice.


178

12/10/07

12:18

Page 178


5.5. The unfair treatment of smaller undertakings Finally, it has also been said that the Leniency Notice, while increasing the incentives for larger undertakings to come forward and blow the whistle, has had the perverse effect of reducing these same incentives for smaller undertakings. Trade association meetings have frequently occasioned accusations about illicit cooperation. Often the meetings have been attended by a range of small and large companies and by national associations. The industry leaders, commonly four or five national champions, can arrange their understandings privately on the fringes of plenary discussions on safe topics; and smaller players, who attend the official gathering but not the private ones, get an indication of the way the wind is blowing. Their goal in attending the gathering may often be to gather market intelligence, which good intentions do not immunise them from being accused of being a willing participant in the infringing conduct. They will be reluctant to come forward to the Commission because they have more to lose from the discovery of a cartel than their larger rivals for several reasons, including the impact of the potential fine. When calculating the level of a fine, the Commission takes as the “basic amount” a lump sum determined without reference to the turnover of the undertakings involved.55 As a result, this sum often reaches a level exceeding 10% of a smaller firm’s turnover (and not just its profits), enough to cripple it. For a larger firm, the same lump sum represents a proportionally smaller amount of its turnover,56 a serious inconvenience (and one increased by the possible use of multipliers for deterrence) but less grave a threat. The application of the Leniency Notice aggravates this phenomenon. The realities of the business world are such that a smaller company runs more risk of being crushed in reprisal than a larger company which will be better able to weather the storm following its decision to incriminate others (and may indeed experience some benefit vis-à-vis customers when competitors are floundering in response to its accusations). As a result, not only will the level of the fine be 55 For criticism of the manner in which the Commission sets fines, see Van Bael I. (1995): “The Lottery of EU Competition Law”, 4 European Competition Law Review 237, referring to practice under former rules. Case T-13/03 Nintendo Co & Nintendo Europe v. Commission is in my respectful opinion a recent example of an aberrantly severe penalty. (I am representing Nintendo in its appeal.) 56 The unfairness of this lump sum calculation was considered by Advocate General Tizzano in his Opinion in Joined Cases Dansk Rørindustri A/S (Case C-189/02 P), Isoplus Fernwärmetechnik Vertriebsgesellschaft mbH and Others (Case C-202/02 P), KE KELIT Kunststoffwerk GmbH (Case C-205/02 P), LR af 1998 A/S (Case C-206/02 P), Brugg Rohrsysteme GmbH (Case C-207/02 P), LR af 1998 (Deutschland) GmbH (Case C-208/02 P) and ABB Asea Brown Boveri Ltd (Case C-213/02 P) v. Commission of the European Communities [2005] ECR I-5425, paras 91–133. Whilst he concluded that the pleas in law of the parties could not be supported, he did note that the “intensification, deriving as it does from a calculation method based on flat-rate amounts, is liable for the most part to hit small and medium-sized undertakings” (para. 131) and that such a result is not “fully consistent with the requirements of individualisation and progressiveness” of penalty setting (para. 130).


12/10/07

12:18

Page 179

Enhancing the Task of Uncovering and Thereby Deterring Cartels 179 in relative terms more severe for a smaller company than for a larger one, but a smaller company is also less likely, in practice, to be granted either immunity or a reduction in the level of its fine. Either it keeps its head down and prays that a larger member of the cartel does not choose to blow the whistle, or it must take it upon itself actively to report to the authorities each activity in which it participates. If it bets wrong, it risks potentially crippling fines. Neither option is particularly appealing. The Commission should reconsider its fining policy vis-à-vis these companies, in order to ensure that they are not overly punished for the same sins (or lesser sins) as the leaders of the illicit behaviour.

C. Current Commission initiatives Alongside the tools currently provided in Regulation 1/2003 and the Leniency Notice, the Commission has recently also sought to broaden the scope and efficacy of its investigative powers through the use of a number of more novel devices, which we will now consider.

1. Private enforcement Article 81 EC has often been invoked by private litigants as a defence against actions for breach or performance of certain contractual obligations.57 Under Regulation 17, agreements which were not notified to the Commission and which did not benefit from the provisions of a Block Exemption Regulation, were automatically void under Article 81(2) EC, even if it was later found that the agreement would have fulfilled the four conditions of Article 81(3) EC for individual exemption. This, in the eyes of some, pernicious use of Article 81 EC by persons who sought to escape their contractual duties has now been eliminated. It remains the case of course that a person who has signed an anticompetitive contract may still claim its invalidity, but not on the formal ground of its non-notification. As to the possibility of using Article 81 EC as a “sword”, both Regulations 17/62 and 1/2003 are silent. Whilst both the European Parliament58 and the Commission59 had at various times expressed the view that Articles 81 and 82 EC could and should be interpreted as allowing the recovery of damages and that such actions would be a useful complement to the Commission’s own 57 See in this regard, Wils W. P. J. (2003): “Should Private Antitrust Enforcement Be Encouraged in Europe?”, 26 World Competition 473. 58 See its position on the Commission’s proposal for what later became Regulation 17, OJ 1410 [1961], para. 11. 59 See the list of Commission statements referred to by Advocate General Van Gerven in his Opinion in Case C-128/92 Banks [1994] ECR I-1251, ftn. 112.


180

12/10/07

12:18

Page 180


administrative enforcement actions, no explicit provisions to this effect were included in Regulation 17. In practice, there had been relatively few cases where Articles 81 and 82 EC have been invoked in damage actions before national courts.60 In 2001, the European Court of Justice confirmed, for the first time, the possibility of seeking compensation for loss caused by a contract or by conduct that was found to be in breach of EC competition law.61 The marathon Crehan cases in the English courts62 show how easy it is for skilled lawyers to find many reasons why damages for breach of the competition rules ought not to be awarded in any given case and how difficult it can be to seperate the various causes of damage. Were the loses incurred due to bad business choices, bad luck or bad market conditions? Or were they caused by breaches of the competition rules? Did the operator of the pub pay too much for his beer because he unfortunately accepted worse terms than he could have obtained from another brewer, or were the terms caught by the prohibition Article 1(1), affecting trade between Member States? Before the European Court of Justice, Crehan confirms the need for an adequate remedy to be available in the national courts. The Commission sees a number of advantages to increased private enforcement of EC competition law.63 In recent press releases in cartel cases, it has invited any person or firm affected by anti-competitive behaviour to bring the matter before the courts of the Member States and seek damages, submitting elements of the published decision as evidence that the behaviour did occur and was illegal.64 It has also recently released a Green Paper65 on how to facilitate actions for damages caused by violations of European competition law, with a specific focus on price fixing cartels. Whether increased private enforcement is the best tool for the effective and efficient application of EC competition law is open to debate.66 On the one

60 Whish R. (2003): Competition Law, 5th edition, Butterworths, London, refers to some of the known examples of damages cases up until 2003. It is difficult to estimate the precise number of damages claims and awards as many cases are either not reported or are confidentially settled. In an attempt to address this issue, the Commission has created a National Court Cases Database, containing non-confidential versions of National Court judgments that are forwarded to it under Article 15(2) of Regulation 1/2003. http://europa.eu.int/comm/competition/antitrust/ national_courts/index_en.html 61 Case C-453/99 Courage v. Crehan [2001] ECR I-6314. 62 Courage v. Crehan [2001] All ER (EC) 886. 63 It argues that increased private enforcement will ensure that businesses and consumers harmed by anticompetitive activity are compensated for their losses; will enhance the overall level of respect for the EC competition rules by discouraging companies from engaging in anticompetitive activity; and will bring the benefits of Community law closer to the citizen. 64 See the Press Releases in the two most recent cartel decisions: Industrial bags (IP/05/1508 of 30 November 2005) and Rubber Chemicals (IP/05/1656 of 21 December 2005). 65 Damages actions for breach of the EC antitrust rules, Commission Green Paper, SEC(2005) 1732. For comment, see Diemer C. (2006): “The Green Paper on Damages Actions for Breach of the EC Antitrust Rules”, 27 European Competition Law Review 309. 66 On the basis of the assumption that the goal of competition law enforcement is to maximize the likelihood that the law is respected.


12/10/07

12:18

Page 181

Enhancing the Task of Uncovering and Thereby Deterring Cartels 181 hand, authors such as Wouter Wils67 dispute the need for increased private enforcement, arguing that there are several factors (the wider investigative powers of competition authorities, the divergence between the private and general interest, and the increased costs of private enforcement) that show the inherent superiority of public over private enforcement. Mr. Wils doubts whether private enforcement in the EU could play any useful supplementary role, either in terms of providing additional sanctions or in bringing additional cases. Granting enforcement agencies increased resources or allowing the imposition of individual penalties would be more effective, claim such sceptics, than private enforcement. However, the academic literature generally favours an increased role for private enforcement within the EU.68 Private enforcement has a valuable role to play as a supplement to public enforcement and should become the primary means of compensating the victims of competition law infringements before the national courts. The theoretical economic arguments developed against private enforcement “are weak, insufficient, and lack an observable basis in the real world.”69 As the anecdote with which I started would suggest, I consider that the obtaining of indemnification should not be impossibly hard. At the end of this paper, I suggest a mechanism at the moment of the taking of the Commission decision that could facilitate compensation. However, I think it would be neither effective nor desirable for Europe to create a malignant legal industry in order to add another element of deterrence to the existing reasons to refrain from cartel activity. Civil private enforcement should be available as a vehicle to achieve justice, but not to deliver punishment. And litigation by endless discovery, by attrition, by abuse of process, should be no part of European law.

2. Fining the facilitators Can a supplier of services to undertakings or associations of undertakings which facilitated their engagement in anti-competitive agreements or practices be held liable under Article 81(1), alongside the other cartel members? The Commission answered this question with a qualified “yes” in its recent Organic Peroxides70 Decision. In addition to fining the participating companies, the 67 The Optimal Enforcement of EC Antitrust Law (2003) Kluwer Law International, chapter 8, a very thoughtful piece. 68 See among others Komninos A. (2002): “New Prospects for Private Enforcement of EC Competition Law: Courage v. Crehan and the Community Right to Damages”, 39 Common Market Law Review 447, a perceptive and thorough study; Ehlermann C.-D. and Atanasiu I., eds. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland, Oregon. 69 Jones C. (2004): “Private Antitrust Enforcement in Europe: A Policy Analysis and Reality Check”, 24 World Competition 13. 70 Commission decision of 10 December 2003 (COMP/37.857—Organic Peroxides); http://ec.europa.eu/comm/competition/antitrust/cases/index/by_nr_75.html#i37_857.


182

12/10/07

12:18

Page 182


Commission also fined AC Treuhand, a consultancy firm, for acting as secretary and essential organizer of the cartel.71 The Commission, referring to its prior Decision in Italian Cast Glass,72 concluded that Treuhand could also be held liable for the activity of the cartel as it met all of the requirements for the application of Article 81(1) EC. Treuhand is an undertaking within the meaning of Article 81(1) EC since it is a company that carries on an economic activity. It had participated in the agreement as an undertaking or took decisions as an association of undertakings. A sudden departure of AC Treuhand, would have, at least temporarily, disrupted the functioning of the agreement. Interestingly, the Commission stated that it did not consider it “[. . .] necessary to prove the exact role of hybrid entities such as AC Treuhand, which clearly have infringed Article 81 of the Treaty. AC Treuhand participated in the infringement directly for the purpose of restricting competition in the sector of OP, even it does not produce OP itself, and/or it took decisions with that purpose. Hence AC Treuhand infringed Article 81 of the Treaty.”73

Due to the novelty of the matter, the Commission imposed only a fine of €1,000 on Treuhand. The Decision is noteworthy: the Commission may target independent companies that act as facilitators or organizers of cartel arrangements. However, the full implications of the Decision are unsure. Treuhand has appealed the Decision to the Court of First Instance,74 alleging that as it was not a party to the cartel, it was unable to comment on the Commission’s allegations during the investigation, thereby infringing its rights of defence and acting in breach of the fundamental right to due process. It also submits that the Commission acted in breach of the principle of nullum crimen, nulla poena sine lege since Treuhand was neither a party, as an undertaking, to the agreement restricting competition nor was it a group of undertakings. On the other hand, the ambit of the Decision itself is unclear in so far as it seems to suggest that its scope of application is limited to cases where undertakings provide services that exceed that of a “purely secretarial function”. On an extensive reading of the Decision, it could even be argued that hotelowners should be under an obligation to report anticompetitive behaviour that occurs in the murky corners of their business centres or conference rooms. One might argue that a knowing host should be blameworthy for allowing such meetings to take place (some famous hotels are indeed regularly cited as cartel venues in Commission decisions!). It is to be hoped that the CFI’s judgment will shed some further light on this interesting and controversial issue. 71 For comment, see Forrester I., MacLennan J. and Komninos A. (2005): “EC Competition Law 2003–2004”, in 2005 Oxford Yearbook of European Law at p. 557. 72 Commission decision of 17 December 1980 relating to Italian cast glass, OJ L 383 [1980]. 73 Paras. 345 and 346 of the decision. 74 Case T-99/04 AC-Treuhand v. Commission, judgment pending.


12/10/07

12:18

Page 183

Enhancing the Task of Uncovering and Thereby Deterring Cartels 183 Perhaps the reasoning in the Organic Peroxides Decision could also be applied to law firms in light of the lawyer’s duty to encourage respect for the law.75 Consider two examples. First, the law firm knowingly facilitates the organisation of the cartel, allowing it to hold meetings at the firm’s premises and giving legal advice to the cartel on the best manner in which to organize the cartel in order to escape detection. That (improbable) law firm could also be held to be breach in Article 81(l) EC through its active participation and facilitation of the actions of the cartel. By contrast, consider a law firm which is asked by its client whether the client could use the law firm’s premises to hold a meeting with several competitors, the aim of the meeting being the drafting of a common response to a proposed piece of legislation which the industry feels will hamper business conditions. Suppose the lawyer discovers after the meeting that one of the clients has forgotten a folder labelled “European Prices Task Force: BURN if DG Competition asks for a copy”, revealing that the meeting was organized to work towards the setting up of a cartel? This is a nice question, not likely to arise in practice. If ever European competition law modules contain an ethics module, one can imagine various twists to either scenario. It commonly happens that lawyers attend meetings between competitors, precisely to enhance the participants’ understanding of the legal principles, and to keep a note of the proceedings for the avoidance of doubt and for the protection of participants. So it would be undesirable to try to convert the lawyer from counsellor of the client to zealous smiter of the unlawful.

3. The use of trustees Under Regulation 1/2003, investigatory and enforcement powers in relation to Articles 81 and 82 EC are conferred exclusively upon the Commission (or national competition authorities).76 Nowhere does the Regulation authorise the Commission to delegate its investigatory or enforcement powers to third parties, in particular private persons. In its Decision in Microsoft,77 the Commission took the novel step of requiring Microsoft, as part of the remedy imposed upon it, to: “[. . .] submit a proposal to the Commission for the establishment of a suitable mechanism assisting the Commission in monitoring Microsoft’s compliance with this Decision. That mechanism shall include a monitoring trustee who shall be independent from Microsoft Corporation.”78

75 See Riley A. (1991): “European Competition Law: The Client v. the Public Interest”, 12 European Competition Law Review 39. 76 Pursuant to Article 5 of Regulation 1/2003. 77 Commission decision of 24 March 2004 relating to a proceeding under Article 82 of the EC Treaty (Case COMP/C-3/37.792 Microsoft). 78 Article 7 of the Microsoft Decision.


184

12/10/07

12:18

Page 184


It is submitted that such a delegation of public enforcement powers is illegal. While the role of the Trustee may seem simply to assist the Commission in ensuring Microsoft’s compliance with its obligations under the Decision, in reality the powers given to the Trustee in order to fulfil that task go far beyond those of a third party expert, and are more akin to those of a public authority. Articles 3.1, 3.3, 5.5 and 5.6 of the Microsoft Trustee Decision79 indeed allow for the possibility that the Trustee may carry out investigations on its own initiative. Article 3.1 provides generally that the Trustee shall monitor Microsoft’s compliance, and adds detailed provisions about investigations that the Trustee would be required to carry out, apparently on its own initiative (see Article 3.1(c) and (d)). Article 3.3 makes it clear that the Trustee must accept third party complaints and investigate them on at least a preliminary basis before notifying the Commission. Article 5.5 provides that the Trustee will carry out tasks specified in a Trustee Mandate in addition to other tasks after having been specifically instructed by the Commission. Article 5.6 provides that the Trustee is to be “pro-active”, and shall carry out tasks covered by a Work Plan as well as, in the absence of approval from the Commission, additional work that is not “significant”. Furthermore, as icing on the innovation cake, the Trustee is not subject to the material constraints under which the Commission operates. Not only does the Trustee have “the possibility to hire experts to carry out certain precisely defined tasks on his behalf (. . .) access to Microsoft’s assistance, information, documents, premises and employees to the extent that he may reasonably require such access in carrying out his mandate”,80 but his entire costs, “including a fair remuneration for the Monitoring Trustee’s activities” are also to be borne by Microsoft. The Decision therefore allows the Commission to outsource its investigative powers to an independent party, to whom the Commission delegates some of its investigative powers. This is a novel approach to inadequate resources. As the costs are paid by the defendant, this could have strong attractions. The enforcement implications of using outsiders to police compliance in big cases are enormous. It is now for the CFI81 to decide whether such an initiative is legitimate and whether the phenomenon of outsourcing, so popular in current and political discourse, will also find favour in EC competition law.

79

See http://ec.europa.eu/comm/competition/antitrust/cases/index/by_nr_75.html. Recital 1048 of the Microsoft Trustee Decision. 81 Case T-201/04 Microsoft v. Commission, judgment pending. The author is co-counsel for Microsoft in its appeals. 80


12/10/07

12:18

Page 185

Enhancing the Task of Uncovering and Thereby Deterring Cartels 185

D. The current administrative system: can it ever be sufficient? The current system of administrative enforcement of EC competition law is focused on setting and imposing sufficiently high corporate sanctions on undertakings to ensure that the undertaking is both sufficiently punished for its actions and that other undertakings are sufficiently deterred by the threat of such sanctions so as to comply with the competition rules. However, there is a growing realization that the limits of this administrative system are now being reached and that there may be a need to consider incorporating elements of a different nature. This section will consider different options, both theoretical and in place in other jurisdictions, and question how and whether these could be applied to the current Community regime.

1. Methods that involve further development of the existing system 1.1. “Amnesty Plus” The idea behind “Amnesty Plus” programmes is to reward companies if they inform competition authorities about collusive activity in a market not yet being investigated. This encourages cartel participants caught in an ongoing investigation to come forward for amnesty for another cartel. The US Department of Justice has been using such a system since 1993. It provides that a company which does not qualify for leniency for the first cartel but discloses a second cartel (for which leniency conditions are satisfied) can receive amnesty for the second one and a substantial reduction in penalty for the first one, even though it was not first in the queue to confess that one. In 2001, more than half of the DOJ’s 30 global cartel investigations were the result of Amnesty Plus leads. A similar programme has also been implemented in the UK in paragraphs 3.10 to 3.11 of the Director General of Fair Trading’s Guidance as to the appropriate amount of a penalty.82 If a company which is the subject of an investigation by the OFT in relation to its participation in a suspected cartel, also provides evidence to the OFT of the existence of a separate cartel in another market, it will obtain not only complete immunity from fines in relation to its activities on the second market but will also receive a reduction in financial penalties which would have been imposed on it in the first market, in addition to the reduction which it would have received for its co-operation in the first market alone.83 82

(OFT 423). For more detail, see Henry D. (2005): “Leniency Programmes: An Anaemic Carrot for Cartels in France, Germany and the UK?”, 26 European Competition Law Review 13. 83


186

12/10/07

12:18

Page 186


In the EU, the benefits that could arise from an “Amnesty Plus” programme can be seen from the two Commission Decisions in the Belgian 84 and Luxemburg 85 Breweries cartel cases. The Commission’s investigation into the Luxemburg cartel was only made possible by Interbrew’s choosing to disclose the existence of a separate cartel on the Luxemburg market in the course of the Commission’s investigation into Interbrew’s participation in a cartel on the Belgian market. Whilst Interbrew was granted 100% immunity from fines under the 1996 Notice in the Luxemburg Breweries case, it received no further reduction in fines in the Belgian Breweries case in recognition of this extra information. It can therefore be argued that Interbrew’s decision to inform the Commission of the existence of a second cartel and its participation therein represents an incongruity. Undertakings will be unlikely to inform the Commission of the existence and their participation in other cartels unless they are given incentive, in the form of “Amnesty Plus” programmes, to do so.86

1.2. Plea bargaining It has also been suggested that a form of plea bargaining should also be introduced. Plea bargaining has been successfully used by the US Department of Justice in its prosecution of cartels.87 In essence, the parties to the cartel and the enforcement agent together agree on the nature and scope of the illegal cartel activity and the appropriate penalty to be imposed. A modified form of plea bargaining has also already been adopted in certain Member States. For example, the new French system of leniency88 provides for a form of plea bargaining under what is called the “transaction procedure”. During proceedings before the Competition Council and after the notification to the parties of the statement of objections, a party can benefit from a reduction in fines (not full immunity) if it decides not to contest the existence of the alleged practices and offers sufficient commitments to modify its behaviour in the future, including a compliance programme. There are signs that the Commission is also considering whether to create its own form of plea bargaining. A past disincentive to seeking leniency in Brussels has been the uncertainty over what fine the Commission might 84 Commission Decision COMP IV/37.614/F3 PO/Interbrew and Alken Maes, 5 December 2001. 85 Commission Decision COMP/37.800/F3, Luxemburg Brewers, 5 December 2001. 86 McElwee D. (2004): “Should the European Commission Adopt “Amnesty Plus” in its Fight Against Hard-Core Cartels”, 25 European Competition Law Review 558. 87 Although the success of the DOJ may now be called into question for the Supreme Court’s recent decision in United States v. Booker 543 220 (2005), which severed the previously mandatory Federal Sentencing Guidelines from the Sentencing Reform Act, thereby relegating the Guidelines to “advisory” status and reducing the amount of leverage in antitrust cases that these mandatory guidelines had given the DOJ in fine negotiations. 88 Article L.464–2 II, III and IV of the Commercial Code and in Article 44 of the implementing decree of 30 April 2002.


12/10/07

12:18

Page 187

Enhancing the Task of Uncovering and Thereby Deterring Cartels 187 choose to impose. Since fines in different cases could fluctuate to an extraordinary degree, the doubting company was unsure. It doubted the meaning of a “50% reduction” of an immense but incalculable amount. It has been widely believed that the Commission was non-transparent and unpredictable in its setting of fines, so to adopt plea bargaining would be a major change. In a speech on 7 April 2005,89 Commissioner Kroes suggested that the Commission may look into developing some form of plea bargaining power with a view to negotiating direct settlement agreements with alleged cartel members if the measures adopted in the 2002 Notice are not sufficient “to deliver swift enforcement with timely punishment”. The introduction of some form of plea bargaining would remove the need for the Commission to investigate each and every case. It could lead to faster enforcement decisions and also free additional Commission resources to focus on key areas, in line with the rationale behind the modernization regime. The Commission has at the moment a considerable queue of liniency applications awaiting attention. The Commission would doubtless welcome a respite from the need to record in a formal, detailed, published and judiciallyreviewable act, the comprehensive details of how a cartel functioned. The need for such throroughness would be eliminated if the infringer had conceded its guilt and accepted it penalty.

1.3. Paying whistleblowers The use of paid “moles” or agents provocateurs would seem to be the natural extension of the criminalization of cartel enforcement. If it is decided that cartels are so serious as to merit criminal penalties, then equally it can be argued that certain methods used to detect other types of criminalized behaviour such as fraud or tax evasion or theft, should also be allowed to facilitate the discovery of cartels. In 2002, Korea became (and still is) the first jurisdiction in the world to create an informant reward programme to encourage insiders and market participants in upstream and downstream markets to provide crucial information to the Competition Authority relating to the existence of cartel behaviour. Such a system is seen as being cost effective as it will save investigative costs and induce voluntary compliance by the businesses. Under article 64(2) of the Korean Monopoly Regulation and Fair Trade Act, the authorities “may pay the compensation within the limit of the budget for the person who informs the Korean Fair Trade Commission of the violation of this Act and provides the evidence that proves this violation.” The level of reward is based on the level of sanction and the quality of evidence 89 “The First Hundred Days”, speech delivered at the 40th Anniversary of the Studienvereinigung Kartellrecht 1965–2005, International Forum on European Competition Law, 7 April 2005.


188

12/10/07

12:18

Page 188


provided.90 First, a standard amount is derived according to the level of sanction. Where the surcharge in the price of the products due to the cartel is less than KRW 500 million, the standard amount is set at 5% of the surcharge, with a minimum of KRW 5 million; where the surcharge is between KRW 500 million and 50 billion, the standard amount is set at 1%; finally where the surcharge is over KRW 50 billion, the standard amount is set at 0.5%. Once the standard amount is calculated, the final amount will then depend on the quality of evidence provided to the Authority. “Top grade evidence” such as a signed agreement will bring the reward to 80%–100% of the standard amount. “Medium grade evidence” such as the minutes of the meeting is worth 60%–80% of the standard amount. Finally, “low grade evidence” includes supporting information of cartel agreement and coordinated action and is worth 40%–60% of the standard amount. Initially the maximum award was set at KRW 20 million (€17 000), but because there were not many reports, the maximum reward was raised to KRW 100 million (€85 000 ) in November 2003 and to KRW 1 billion (€850 000) in 2005.

2. If we consider a radical change is necessary 2.1. Criminal sanctions The desirability of the criminalisation of the enforcement of EC competition law is much debated.91 Whilst Article 5 of Regulation 1/2003 allows Member States to provide for the imposition of criminal sanctions for violations of Articles 81 and 82 EC,92 such a power is lacking from the enforcement arsenal of the European Commission. It is unclear whether the European Community has the competence to criminalize certain violations of European competition law. In its recent judgment in Case C-176/03,93 the ECJ ruled 90 See Annual Report on Competition Policy Developments in Korea 2004, OECD Directorate for Financial And Enterprise Affairs, Competition Committee, DAF/COMP(2005)32/29. 91 See the discussion in Part IV of Ehlermann C.-D. and Atanasiu I., eds. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland, Oregon, at pp. 453–462; Kolasky W. (2004): “Criminalizing Cartel Activity: Lessons from the U.S. Experience”, 5th lunch talk of the Global Competition Law Centre (Brussels, 29 September 2004), accessible at http://www.gclc.coleurop.be/ lunchtalk20040929.htm, at 11–12; Rosochowicz P. (2004): “The Appropriateness of Criminal Sanctions in the Enforcement of Competition Law”, 25 European Competition Law Review 752; Wils W. P. J. (2005): “Is Criminalization of EU Competition Law the Answer?”, 28 World Competition 117, a very thoughtful review of the major issues. 92 “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 (. . .) decisions (. . .) imposing fines, periodic penalty payments or any other penalty provided for in their national law.” 93 Case C-176/03 Commission v. Council, [2005] ECR I-7879, at para. 48. For comment, see White S. (2005): “Harmonisation of Criminal Law under the First Pillar”, 31 European Law Review 81.


12/10/07

12:18

Page 189

Enhancing the Task of Uncovering and Thereby Deterring Cartels 189 that the lack of Community competence in the field of criminal law and criminal procedure “[. . .] does not prevent the Community legislature, when the application of effective, proportionate and dissuasive criminal penalties by the competent national authorities is an essential measure for combating serious (environmental) offences, from taking measures which relate to the criminal law of the Member States which it considers necessary in order to ensure that the rules which it lays down on environmental protection are fully effective.”

The Commission has since issued a Communication setting out how it intends to interpret and implement the judgment.94 It proposes that “appropriate measures of criminal law can be adopted on a Community basis only at sectoral level and only on condition that there is a clear need to combat serious shortcomings in the implementation of the Community’s objectives and to provide for criminal law measures to ensure the full effectiveness of a Community policy or the proper functioning of a freedom.” This therefore implies the possibility that a similar line of reasoning may be applied to the field of competition. The need to ensure the effective enforcement of EC competition rules at national level can be interpreted as permitting the adoption of legislation under the Community pillar, requiring that Member States adopt “effective, proportionate and dissuasive criminal penalties” in the field of competition law. Criminalization at EU level would ensure uniformity of available sanctions across Member States for similar violations of EC competition law. This might be thought to be an advantage. However, the criminalization of competition law also brings with it additional problems. In a recent interview,95 Judge Vesterdorf highlighted that while the criminalization of EC competition law may have strong deterrent effect and lead to a reduction in the number of infringements, it would also have the potential to change the manner in which the Commission currently conducts its investigations. Because of the need to ensure a greater degree of protection of the rights of individual managers and directors, the nature of the cooperation between the Commission and investigated undertakings would necessarily change. Documents which under an administrative system would be handed over to the investigating body might now be withheld as they could subsequently be used in criminal proceedings. The criminalization of EC competition law will also require the Commission to cooperate with national judges during the criminal phase of its investigation as it cannot, as an administrative body, have the power to impose criminal sanctions. Consequently, while we must 94 Communication on the implications of the Court’s judgment of 13 September 2005 (Case C-176/03 Commission v. Council ), COM(2005) 583 final. 95 Vesterdorf B. (2006): Interview given to the Revue Lamy de la Concurrence: “pouvoir donner une réponse utile au juge national peut demander une connaissance approfondie du marché”, RLC 2006/7 no. 493.


190

12/10/07

12:18

Page 190


agree that the criminalization of EC competition law has the potential to reduce the number of infringements, the debate cannot stop there. I seriously doubt whether current Commission procedures are adequate to the challenge of satisfying national criminal standards. Decisions are not immune to political pressure, they are taken by a college of Commissioners, not by a court of judges, and the evidential rigour deployed in their drafting is not always adequate. This does not mean any disrespect for those who have the important job of enforcing Community policy against cartels, which is a vital and laudable activity. Their limited resources and the necessity for their duty do not eliminate certain features of the current regime: undertakings must cooperate with the investigation; no trial occurs; administrative law governs and not criminal law; and the penalties are enormous.

2.2. Job losses for executives infringing competition law An alternative to the potential criminalisation of EC competition law could be the professional disqualification of directors who are found to have participated in a cartel. Such a system is currently under consideration in several Member States as either a complement or as an alternative to criminal sanctions. In the UK, the disqualification of directors is used as a complement to the imposition of criminal sanctions. Section 9 of the Company Directors Disqualification Act 1986, as amended by Article 204 of the Enterprise Act 2002, provides that, on application from the OFT, a competent court must make a Competition Disqualification Order (CDO) against a director of a company if the undertaking which is a company of which he is a director commits a breach of competition law, and if the court considers that such conduct has made the director unfit to manage a company. By contrast, in the Netherlands, the Dutch Parliament, supported by the Dutch Competition Authority (the NMA) has recently proposed to amend its Competition Act in order to introduce the possibility to remove from their jobs the executives of companies that are found to have participated in cartels. This is seen as an alternative solution to the introduction of criminal penalties, which many national judges may be reluctant, in practice, to enforce. The knowledge by directors that they risk losing their jobs would serve as a threat, whilst being easier to enforce than criminal sanctions.

2.3. Settlements Consideration should be given to including in decisions terminating cartel infringements a provision whereby the infringing company promises to accept arbitration of claims from victims of the infringing conduct. This would have assisted the victim whose plight I described at the beginning.

(L) Ehlermann 07 Goldman et al.

12/10/07

13:09

Page 191

IV Calvin S. Goldman, Robert E. Kwinter and Angie L. Morris* The Investigative Powers of the Canadian Competition Bureau: Domestic and International Dimensions

Introduction As described more fully below, the Commissioner of Competition (the “Commissioner”), who is the head of the Canadian Competition Bureau, has a wide variety of tools available to her in investigating cartels and gathering evidence under the Competition Act (the “Act”).1 However, even without resort to formal investigative powers such as search and seizure and section 11 orders, targets of an investigation often voluntarily cooperate with the Commissioner and the Competition Bureau (the “Bureau”).2 With the dramatic growth in multi-jurisdictional cooperation among antitrust authorities around the world, the Commissioner’s investigations also have taken on an increasingly international dimension. In recent years the Commissioner’s investigative powers have been applied in a greater number of cases pursuant to agreements with foreign antitrust authorities in investigations of offences that have attracted both domestic and international interest. Cartels are a serious criminal offence under the Act, punishable by corporate and individual fines up to $10 million per count and imprisonment for up to 5 years. Criminal process and sanctions likely carry the greatest deterrent effect and have also facilitated the Bureau’s exchange of information with other jurisdictions, particularly the US, further promoting enforcement efforts. From the perspective of deterrence, the possibility of penal sentences is likely as significant a factor as the length of available sentences. For corporate executives, the prospect of being publicly identified as a criminal offender in respect of an offence to which penal consequences attach can be particularly sobering. This paper reviews the formal and informal investigative means by which the Commissioner can gather information under the Act; it also discusses * Cal Goldman and Rob Kwinter are partners at Blake Cassels & Graydon LLP in Toronto, Canada. Mr. Goldman is a former Director of the Canadian Competition Bureau (a position now known as the “Commissioner of Competition”). Angie Morris is an associate at Blake Cassels & Graydon LLP in Toronto, Canada. 1 Competition Act, R.S.C., c. C-34, as amended. 2 See John F. Clifford and Jeffrey R. Roode, “Practical Issues in Responding to Bureau Investigations and Ensuring Competition Law Compliance, presented at the Insight Competition Law Practices for Canadian Companies Conference in December 1999.


192

12/10/07

13:09

Page 192


some of the recent developments and issues arising from the Bureau’s cooperative efforts in international antitrust enforcement.

B. Overview 1. The Commissioner The Commissioner, Canada’s chief competition law enforcement official, has exclusive authority to administer and enforce the Act and is responsible for the investigation of violations of Canadian competition laws.3 The Commissioner must commence a formal inquiry whenever she has reason to believe a criminal offence has been, or is about to be committed or that grounds exist for the Competition Tribunal to make an order regarding a reviewable practice.4 While most inquiries begin at the Commissioner’s instance, an inquiry must also be commenced when the Minister of Industry so directs, or on the sworn application of six Canadian residents.5

2. The Bureau In Canada, the managers and officers of the Bureau actually investigate alleged violations of the Act pursuant to legislative authority given to the Commissioner and her staff under the Act. If the Bureau finds evidence of criminal conduct, then the Commissioner or one of her deputies can refer the matter to the Canadian Department of Justice (referred to generally as the “Attorney General” or the “Crown”) for prosecution.6 Prosecutions are brought before the criminal court and elements of the offence charged must be proved beyond a reasonable doubt.7 For civil reviewable matters, it is the Commissioner who decides whether or not to file an application with the Competition Tribunal to seek to prohibit the conduct. The Attorney General provides counsel to the Commissioner in both criminal and civil matters, but in practice tends to follow the Commissioner’s recommendations. Bureau officers will also work closely with counsel for the Attorney General throughout the prosecution process. 3

Section 7(1)(a) of the Competition Act. Section 10(1)(b) of the Competition Act. 5 Section 10(1)(a) of the Competition Act 6 Section 23 of the Competition Act. 7 This is in contrast to the standard of proof used in civil proceedings, which is the lower threshold of “on the balance of probabilities.” For a recent discussion of what is meant in practical terms by proof “beyond a reasonable doubt,” see the Supreme Court of Canada’s decision in R. v. Litchus, [1997] 3 S.C.R. 320 (SCC). 4


12/10/07

13:09

Page 193

The Investigative Powers of the Canadian Competition Bureau 193

3. The investigative powers The Commissioner and the Bureau have considerable investigative powers at their disposal to deal with alleged violations of antitrust laws. These powers become available to the Bureau once the Commissioner initiates a formal inquiry into a possible violation of the Act. Once an inquiry is underway, the Commissioner may make formal or informal requests for information from targets of the investigation and third party participants in the relevant market. Formal investigative tools available to the Commissioner include: search and seizure, examinations under oath, the production of records, computer searches, wire taps, whistle blowing and immunity provisions. Despite the difficulties in obtaining evidence from foreign jurisdictions, foreign entities should not be lulled into the mistaken belief that they are completely beyond the Bureau’s reach. The Commissioner may be able to access relevant documents or personnel either by search warrant or section 11 order if the foreign entity has a Canadian subsidiary. And more importantly, the Bureau also has both formal and informal avenues for securing the assistance of foreign antitrust authorities in its investigations outside of Canada.

C. Investigative powers

1. Informal information requests Informal investigations (sometimes called “preliminary examinations”) may take a variety of forms, ranging from telephone calls from Bureau case officers to written requests for documents or other information (“Information Requests”). Information Requests usually include a series of specific questions drafted by the Bureau with a view to gathering evidence relevant to their investigation. These informal requests may be directed at the targets of the investigation and/or their suppliers, customers and competitors. Increasingly, the Bureau requests that responses to Information Requests be accompanied by a sworn affidavit of full compliance. For this reason, it is important that the Information Request is fully understood and that compliance can be determined. Where the Bureau issues an overly broad Information Request or one that requests information that is clearly irrelevant and/or which requires a response within an unreasonably short time frame, negotiation should be undertaken with the Bureau to pare down the request so as to reduce the administrative burden on the respondent and to limit the request to the most relevant documents and information.


194

12/10/07

13:09

Page 194


Requiring targets to attest to full compliance with an Information Request adds a degree of formality to the investigation. However, it is important to note that informal Information Requests are not mandated by the Act and responses are not mandatory. That said, failure to respond to an Information Request in a full and timely manner could cause the Commissioner to use her formal investigative powers to compel the production of the information sought.8

2. Subpoenas for oral testimony, records and written returns 2.1. Introduction An order under section 11 of the Act, commonly referred to as a “section 11 order” is an information-gathering tool available to the Commissioner once a formal inquiry has been commenced. Section 11 orders are available in respect of investigations into criminal and civil reviewable matters to compel testimony, require the production of documents or a written return (i.e. a written response to questions posed by the Bureau).9 The Commissioner typically applies for a section 11 order on an ex parte basis, meaning that neither the person subject to the order nor the person to whom the inquiry pertains will be notified beforehand. The legal threshold for the issuance of a section 11 order is relatively low.10 A section 11 order can be obtained by the Commissioner if a judge is satisfied that an inquiry is being made under section 10 of the Act and that a person has, or is likely to have, information that is relevant to the inquiry.11 In Canada (Commissioner of Competition) v. Air Canada, Reed J. held that the Commissioner must also provide sufficient information with respect to these elements for the exercise of judicial discretion in favour of granting an order:12 “Section 11 provides that a judge may, not shall, issue an order. Residual discretion exists. Also, I cannot conclude that section 11 authorizes the issuing of an order to produce information if the Commissioner were acting on a “whim”. I cannot envisage a court granting a section 11 order on the basis of a bald assertion by the Commissioner than an inquiry has been commenced. It seems to me that any judge would require more than that. He or she is likely to require some description of the

8

See ibid. at p. 3. Sections 11(1)(a), 11(1)(b) and 11(1)(c) of the Competition Act respectively. 10 As a result, and since the threshold is lower than the “reasonable grounds” required for a search warrant (which is used for “dawn raids”), this section has recently been the subject of a constitutional challenge which is still pending. 11 Section 11 of the Competition Act. 12 [2001] 1 F.C. 219 (T.D.). 9


12/10/07

13:09

Page 195

The Investigative Powers of the Canadian Competition Bureau 195 nature of the alleged conduct that is the subject of the inquiry, the basis of the Commissioner’s decision to commence an inquiry and his reason for believing that conduct to which the inquiry is addressed has occurred. Also, the judge must be satisfied that the person against whom the order is sought is likely to have relevant information. This does not mean that the Court second guesses the Commissioner’s decision that he has reason to believe that the conduct that is the subject of the inquiry in question occurred, but it does allow the Court to refuse to grant an order when there is insufficient evidence to support a conclusion that a bona fide inquiry has been commenced.”13

Where a judge is satisfied with the Commissioner’s application, the judge may order a person: (i) be examined under oath; (ii) produce records or other evidence; and/or (iii) deliver a written statement. Significantly, section 11(3) provides that a person subject to such an order cannot refuse to comply on the basis that their evidence may incriminate themselves, or subject them to further examinations.14 However, evidence given by an individual pursuant to section 11(1) cannot be used or received against that individual in any criminal proceeding thereafter instituted against her (other than a prosecution for perjury under section 132 or for giving contradictory evidence in a subsequent judicial proceeding under section 136 of the Criminal Code).15 Section 11 orders are generally used against the targets of an inquiry but may also be used against any third parties who may have information relevant to the Commissioner’s inquiry. Following complaints about the overly broad scope of many such orders, the Competition Bureau now assigns a senior officer to review drafts of all section 11 applications in order to increase the focus and relevance of questions. Although there is some latitude for discussion with the Bureau as to what information will be considered responsive to questions in a section 11 order, any actual change to the questions typically requires an application for variation of the order before the court.

2.2. Examination under oath A person being examined pursuant to a section 11 order is entitled to be represented by counsel16 although the role of counsel is limited to objecting to 13

Ibid. at para. 31. Section 11(3) of the Competition Act. There is a constitutional challenge pending, because of the risk of compulsory self-incrimination, as this section also applies to a possible “target” of a criminal inquiry. 15 Evidence tendered under oath may not be used in a subsequent proceeding, except in a prosecution for perjury or to the extent that contradictory evidence is given. See Kay K. (2001): “Fundamentals of Criminal Anti-Competitive Conduct”, in Houston D., ed., Papers of the Canadian Bar Association Annual Fall Conference on Competition Law—2000, Juris Publishing, Inc., New York, 231, at page 240. Note that the compelled testimony rules apply to persons, while the exclusion of evidence in future proceedings only applies with respect to individuals. 16 Section 12(3) of the Competition Act. 14


196

12/10/07

13:09

Page 196


improper questioning and clarifying a client’s statements.17 Section 12(4) also entitles the person(s) whose conduct is being investigated to attend the examination of third parties unless the presiding officer decides to exclude that person(s) on the grounds that their presence would be prejudicial to the effective conduct of the examination or inquiry, or, for reasons of confidentiality.18 The ability of the Commissioner to compel a person to be examined under oath initially raised concerns about self-incrimination under section 7 of the Charter of the Rights and Freedoms, which guarantees everyone, “the right to life, liberty and security of the person and the right not to be deprived thereof except in accordance with the principles of fundamental justice.”19 However, the Supreme Court of Canada considered this issue under similar legislation in Branch20 and determined that section 7 rights were not offended. Nevertheless, the recent cases of R. v. Jarvis and R. v. Ling appear to have raised the bar in terms of the use of investigative powers in criminal investigations.21 Although these cases related to tax audits that evolved into investigations for penal liability, the Supreme Court unanimously held that the Charter applies with full force and effect to persons under investigation for “regulatory crimes” from the point at which officials have as their predominant purpose the determination of penal liability.22 Although the decision in Branch upheld the right of the Commissioner to use section 11 orders in criminal investigations, it is now arguable that there is scope to revisit this issue, in part, based on the findings in Jarvis and Ling.23

2.3. Production of records A section 11 order may compel persons named in the order to produce documents in their possession, or, in the case of corporations, the possession of any of the corporation’s domestic or international affiliates. Documents that 17

Irvine v. Restrictive Trade Practices Commission, [1987] 1 S.C.R. 181 at 231–135. Section 12(4) of the Competition Act. 19 Canadian Charter of Rights and Freedoms, Part 1 of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c.11, section 7 [Charter]. 20 British Columbia Securities Commission v. Branch, [1995] 2 S.C.R. 3 [Branch]. 21 R. v. Jarvis, [2002] 3 S.C.R. 757, R. v. Ling [2002] 3 S.C.R. 814 [Jarvis and Ling]. In Jarvis, the Supreme Court ruled that the federal Department of Finance could not exercise its compulsory audit powers at the same time it was conducting a parallel criminal investigation into a taxpayer’s finances. 22 Jarvis at para 97. 23 See Assaf D. and Lefebvre E. (2003): “The Investigatory Tools under Sections 11 and 15 of the Competition Act: Selected Issues of Law and Practice”, Canadian Bar Association Competition Section, Annual Fall Conference, October 2, 2003. The basis for this argument is threefold: there was no clear majority ruling on the issue as the court in Thomson was divided on the issue of the constitutionality of the provision under section 7 of the Charter; the modern view of Competition Act offences are that they are serious criminal violations and not merely “regulatory crimes” as described in that decision; and the current Supreme Court in Jarvis and Ling have held that there “when the predominant purpose of a question or inquiry is the determination of penal liability, the ‘full panoply’ of Charter rights are engaged”. Jarvis, at 96. 18


12/10/07

13:09

Page 197

The Investigative Powers of the Canadian Competition Bureau 197 are produced in response to a section 11 order (unlike oral testimony or written returns by an individual) are not subject to any immunity and may be used against the recipient in current or future investigations.

3. Search and seizure 3.1. Introduction The Bureau’s powers of search and seizure are set out in sections 15 and 16 of the Act. Section 15 allows the Commissioner to make an ex parte application to the court to obtain a search warrant. This provision was revised in 1986 after the decision in Hunter v. Southam Inc.,24 where the Supreme Court of Canada struck down the search provisions under subsections 10(1) and (3) of the Combines Investigation Act 25 as being inconsistent with the right against unreasonable search and seizure as set out in section 8 of the Charter. Because of the greater potential impact on Charter rights of a search or seizure, the legal threshold for the granting of a section 15 order is higher than that for a section 11 order. The Commissioner must demonstrate that there are reasonable grounds to believe that (i) a criminal offence has been committed, is about to be committed, or grounds exist for the Competition Tribunal to make an order respecting a reviewable matter, and (ii) there are records or other things that will afford evidence with respect to the commission of the offence or matter at the premises to be searched.26 The Commissioner’s application for a search warrant is accompanied by a sworn information which outlines the basis for the request including the “reason to believe” required for the judge to grant the order. Once granted, an order is enforceable anywhere in Canada regardless of the court issuing the order.27 There is no right of appeal once an order is granted.28 An order may still be subject to judicial review under the inherent right of the superior court to provide prerogative writs; however, the court in 603022 Ontario v. Canada (Director of Investigation and Research) held that this right should not be

24

[1984] 2 S.C.R. 145 [Southam]. R.S.C. 1970, c. C-23. 26 In Southam, the Supreme Court of Canada held that a search and seizure is valid under section 8 of the Charter if: (i) there is a system of prior authorization usually in the form of a valid warrant; (ii) the warrant is issued by a person capable of acting judicially; and (iii) there are reasonable and probable grounds to believe that an offence has been committed or that there are evidence to be found at the place of the search. 27 Section 15(4) of the Competition Act. 28 Hudson’s Bay Co. v. Canada (Director of Investigation and Research) (1992), 10 O.R. (3d) (C.A.). 25


198

12/10/07

13:09

Page 198


exercised to review warrants or the return of seized documents as the admissibility of evidence should be dealt with by the trial court.29 The Commissioner must be in possession of a warrant to search the premises,30 unless exigent circumstances exist, in which case a warrantless search may be conducted provided there are reasonable grounds to believe that an offence has, or is about to be committed, or that grounds exist for the Competition Tribunal to make an order but it is impractical to obtain a warrant due to the circumstances.31 Exigent circumstances might exist where the delay in obtaining a warrant would result in the destruction or loss of evidence.32

3.2. Contents of a search warrant The search warrant must identify the matter in respect of which it is issued, the specific individuals authorized to conduct the search, the premises to be searched and a description of the records or other things that are sought.33 If counsel is present at the time of the search, he or she ought to be careful to observe that only those persons named in the warrant are present and that the records taken comply with the description of what is sought in the warrant. Persons who control the premises searched are bound to allow access to the Commissioner’s representatives named in the warrant and to allow the search for, and seizure of, evidence.34 If the Commissioner’s representatives have been refused access or expect to be refused access, the Commissioner can apply under section 15(6) for an order to have a peace officer provide access.35 While the Commissioner’s powers of search and seizure are used more prevalently in criminal investigations, the Bureau has used these powers occasionally to gather evidence in non-criminal inquiries.36

4. Computer searches 4.1. Introduction Section 16 provides for the search of computer systems and the seizure of electronic records during a search being conducted under section 15. This provision allows computer systems to be searched for any data “contained in 29

(1988), 21 C.P.R. (3d) 575 (Ont. H.C.). Section 15(7) of the Competition Act. 31 Section 15(8) of the Competition Act. 32 R. v. Grant, [1993] 3 S.C.R. 223 at 241–242. 33 Section 15(2) of the Competition Act. 34 Section 15(5) of the Competition Act. 35 Section 15(6) of the Competition Act. 36 Clifford J. and Rowley J. (1996): “Search and Seizure: Canada Gets Tough”, 5 (no. 6) Corporate Counsel’s International Advisor 83. 30


12/10/07

13:09

Page 199

The Investigative Powers of the Canadian Competition Bureau 199 or available to the computer system” wherever located.37 Once retrieved, data covered by the warrant may be copied or reproduced in the “form of a printout or any other intelligible output” and then be seized.38 To make full use of its section 16 powers, the Commissioner has established a specialized Electronic Evidence Unit to conduct forensic searches of computer systems and to handle other electronic evidence obtained through efforts such as Internet sweeps and Web site captures.39 The term “computer system” includes portable computers, pocket and hand-held electronic diaries, and person digital assistants containing a calendar, telephone list and other information.40 If a login name or password is required to access a computer system, the Commissioner’s position is that an individual on the premises is required to assist the person conducting the search to gain access to the computer. While there is no specific case law to support this position, the issue may be moot because in most instances the technology available to search officials provides them with access to all data available to a computer regardless of whether passwords are obtained.

4.2. Extraterritoriality Multinational corporations are increasingly using computers to link their businesses in various jurisdictions, and as a result, data stored in one country is often available to employees in another. The Bureau takes the position that the wording of the statutory provision that says “contained in or available to” permits an extraterritorial search of foreign databases which may be accessible from a computer located on a Canadian premises being searched.41 It is arguable that accessing computer data from outside the country represents an unwarranted extraterritorial application of Canadian law. Similarly, there are grounds for asserting that an antitrust authority’s seizure of computer 37

Section 61(1) of the Competition Act. “Criminal Investigations: Process and Procedure” Remarks of Harry Chandler, Deputy Director of the Investigation and Research, September 24, 1998. Bureau investigators have seized data stored in the United States and in other countries, retrieved from computers located in Canada. 39 For a more expansive discussion of the issues concerning computer searches, see Goldman C., Witterick C. and Kissack M. (1997): “Cross-Border Computer Searches”, Grocery Manufacturers of America 1997 Annual Legal Conference, Washington, D.C., October 7, 1997. 40 Chandler H. (2004): “Criminal Investigations: Process and Procedure”, Canadian Bar Association Competition Law Annual Conference, 24 September 1998. The Act’s definition of the term “computer system” refers to the definition found in ss. 342.1(2) of the Criminal Code. 41 See Sullivan M. and Filion J. (2004): “The Basics of International Cartel Enforcement in Canada”, presented at the Osgoode Hall Continuing Legal Education Program: Canada’s Competition Regime: Thinking Strategically A Practical Guide for Business, January 14, 2004, and Sutton J. (1996): “Investigations under the Competition Act: Recent Issues”, in Important Changes in Competition Law and Competitive Business Practices, the Canadian Institute, 10 May 1996, and Goldman C. and Facey B. “Cross-Border Cooperation and Information Sharing in International Antitrust Cases: The Need for Balance”, presented at a conference on Advising Transnational Entities in the New Millenium, Dallas, 19 June 2001, at page 26. 38


200

12/10/07

13:09

Page 200


records stored in a foreign jurisdiction constitutes an infringement of national sovereignty or a breach of national privacy laws.42 In contrast to the Canadian enforcement position, we understand that the US Department of Justice’s historic position is that it will not knowingly seize documents through a computer search which are located outside the US Analogous computer search and seizure provisions exist in the Criminal Code. Although judicial consideration of these provisions is limited, at least one case seems to support the Bureau’s position that it may validly search and seize data accessible to a computer present in a location named in a warrant. In R v. Edwards 43, the Court upheld as constitutionally reasonable a search of the defendant’s cellular phone service provider’s computer records pursuant to a warrant. The defendant argued that the search was unreasonable and violated section 8 of the Charter since the information the police sought was not located at the address named in the warrant, but rather was stored in computers located elsewhere. The court rejected this argument, stating: “In my view, in this day and age, with the development of computer technology and the so called ‘information highway’, information exists where it is capable of being accessed, translated and recorded. Indeed, it would appear that s. 487 (2.2) of the Criminal Code recognizes this aspect of police seizure capabilities.”44

*** “[. . .] the information properly existed at the head office of the defendant’s cellular phone service provider where it could be accessed, translated and converted into hard copy for reference.”45

Some individuals or entities may be tempted to preclude the Bureau from gaining access to offshore databases during a search by simply “pulling the plug” and severing the computer links between the Canadian subsidiary and its foreign affiliates. This would be an inherently risky course of action, as such action could raise issues of criminal obstruction under the Act.46 Canadian counsel should be consulted before taking such action. Bureau investigators have seized data stored in other countries, from computers located in Canada47 and until issues relating to the permissible scope of computer searches are resolved, the Bureau is likely to continue to take an expansive view of its rights under section 16. 42 See Goldman, Witterick and Kissack, supra note no. 39; Goldman C. and Kissack M. (1998): “US/Canada Antitrust Cooperation and Cross-Border Corporate Searches”, ABA 1998 Annual Meeting, Toronto, Canada, August 3, 1998. 43 [1999] O.J. No. 3819 (Sup. Ct.) [Edwards]. 44 Edwards at paragraph 89. 45 Edwards at paragraph 90. 46 Section 64 of the Act provides that no person shall in any manner impede or prevent or attempt to impede or prevent any inquiry or examination under the Act. Every person who contravenes this provision is liable to a fine of up to $5,000 (Cdn.) and/or imprisonment for a term up to two years. 47 Search and Seizure at 87.


12/10/07

13:09

Page 201

The Investigative Powers of the Canadian Competition Bureau 201

5. Penalties for obstruction and destruction of documents under sections 11, 15 and 16 Obstruction of an investigation or examination is a criminal offence under section 64 of the Act punishable by a fine of up to $5,000 and/or a prison term of up to two years. Failure to comply with a section 11 order without good and sufficient cause can also result in a fine of up to $5,000 and/or imprisonments for up to two years.48 Anyone who destroys or alters documents covered by a section 11 order is guilty of an offence and liable on conviction to a maximum fine of $50,000 and/or maximum imprisonment for five years.49 It is therefore, important that counsel advise clients under an investigation that they ought to ensure that all documents remain in the state in which they existed at the time the order was issued. It is also a criminal offence for any person in possession or control of a premises or computer system subject to a search warrant to fail to provide access to the premises or computer system without good and sufficient cause.50 This offence is punishable by a maximum fine of $5,000 and/or imprisonment for up to two years. Destruction or alteration of records or other things sought under a search warrant is punishable by fine of up to $50,000 and/or imprisonment for up to five years.51 When a corporation fails to comply with an order, any officer, director or agent of the corporation who directed or otherwise participated in the decision to not comply or to destroy/alter records may be charged and, if found guilty, be subject to a fine and/or imprisonment.52

6. Privilege in the context of sections 11, 15 and 16 The Commissioner is prohibited from examining, copying or seizing documents without providing a reasonable opportunity for claims of privilege to be made.53 Any person in authority at the premises may make a privilege claim. If a claim of privilege is made, the record in question is sealed and placed in the custody of an officer of the court or third person agreed upon by both parties.54 The party alleging the privilege must then make an application to 48 49 50 51 52 53 54

Section 65(2) of the Competition Act. Section 65(3) of the Competition Act. Section 65(1) of the Competition Act. Section 65(3) of the Competition Act. Section 65(4) of the Competition Act. Section 19(2) of the Competition Act. Section 19(3) of the Competition Act.


202

12/10/07

13:09

Page 202


the court within thirty days for a determination as to whether the document is in fact the subject of a proper privilege claim. If such an application is not brought within thirty days, the Commissioner may bring an ex parte application for an order that the documents be delivered to her.55 The test for the application of privilege is whether the document or record in question was prepared for the purposes of obtaining legal advice or in anticipation of litigation.56 Reviewing responsive documents required under section 11, 15 and 16 for possible privilege claims can be costly and time consuming and can require the attention of senior personnel within the target’s business. To reduce the compliance burden, counsel should read the order carefully, speak to the Bureau to clarify any ambiguous provisions and attempt to reduce the scope of the order if full compliance could result in an enormous volume of documents. It is important to note, that the seizure of electronic data may result in an inadvertent breach of solicitor/client or litigation privilege, because the information seized may not include the contextual information needed to determine whether it is privileged.

7. Wiretaps Wiretaps are a relatively new addition to the Commissioner’s arsenal of investigative tools. In 1999, Section 183 of the Criminal Code was amended to add three antitrust offences to the list of offences for which judicial authorization to intercept private communications may be obtained.57 The Commissioner is permitted to obtain and use wiretaps in investigations of deceptive telemarketing, conspiracy to fix prices or share markets, and bid rigging.58 Interception of private communications may be permitted with the consent of either the originator or recipient under section 184.2 or without consent under section 184.3 of the Criminal Code.59 To obtain an order permitting a wiretap, the Commissioner must submit a sworn affidavit setting out:

55

Section 19(5) of the Competition Act. R. v. McClure, [2001] 1 S.C.R. 445 at 460. 57 Bill C-20, 1999 c.2, assented to March 11,1990, section 47. 58 Criminal Code, R.S.C. 1985, c. C-46, section 183(c). This provision includes the following criminal offences under the Competition Act as an offence for which a wiretap may be used: (c) any of the following provisions of the Competition Act, namely, (i) section 45 (conspiracy) in relation to any of the matters referred to in paragraphs 45(4)(a) to (d) of that Act, (ii) section 47 (bid-rigging), or (iii) subsection 52.1(3) (deceptive telemarketing). 59 The ex parte application for a wiretap must be signed by the Solicitor General of Canada or a designated agent, and the accompanying affidavit from a representative of the Commissioner must include an explanation as to what other types of investigative procedures have been tried and failed or why they are unlikely to succeed, or why the matter is so urgent that it would be impractical to carry out the investigation using only other investigative procedures. 56


12/10/07

13:09

Page 203

The Investigative Powers of the Canadian Competition Bureau 203 a. the facts upon which the application is based; b. the type of communication to be intercepted; c. the names, addresses and occupations, if known, of the person whose communications are to be intercepted; d. the nature and location of the place, if known, at which communications are to be intercepted; e. the manner of interception to be used; f. the length of time for which the wiretap is requested; and g. whether other investigative procedures have been tried and failed and why it appears they are unlikely to succeed.60 Wiretaps are only used by the Bureau in exceptional circumstances and interception is discontinued as soon as it becomes clear that privileged communications are being overheard.61

8. Whistle blowing and immunity Another investigative tool used by the Commissioner involves encouraging people to come forward and inform her of competition law violations. In criminal matters, there are two mechanisms that facilitate the provision of evidence by persons having information about an offence. First, a person who has knowledge that a criminal offence has been or is about to be committed but who has not participated in that offence may notify the Commissioner and be eligible for protection from reprisals from his employer.62 In particular, an employer shall not dismiss or take employment actions against an employee who, acting in good faith and on the basis of a reasonable belief discloses a past or future offence by the employer or any other person, and refuses to take part in an offence or takes action to prevent the commission of an offence.63 Secondly, corporate and individual applicants having taken part in a criminal offence may seek to avail themselves of the Bureau’s Immunity Program and seek immunity or leniency from the Attorney General. This process may involve full immunity, partial immunity or a negotiated plea settlement. A request for full immunity may be granted in cases where the Bureau was unaware of the offence and the party is the first to disclose it, or where the Bureau was already aware of the offence but the party is the first to provide sufficient evidence to warrant a referral to the Attorney General for

60 61 62 63

Criminal Code, s.185(1). Competition Bureau, “Interception of Private Communications” 22 February 1999. Sections 66.1 and 66.2 are the “whistleblower” provisions of the Competition Act. Section 66.2 of the Competition Act.


204

12/10/07

13:09

Page 204


prosecution.64 The requirements for full immunity are discussed in detail in the Bureau’s Immunity Bulletin; the main provisions provide that: —the party must terminate its participation in the illegal activity; —the party must not have been the instigator or leader, nor the sole beneficiary of the activity in Canada (co-leaders and co-instigators are considered eligible);65 —the party must provide complete and timely cooperation by: revealing any and all offences in which it may have been involved; providing full, frank and truthful disclosure of all evidence and information available to it or under its control; and cooperating fully on a continuing basis at its own expense; and —the party make restitution for the illegal activity, where possible.66 Failure to abide by all of the requirements may vitiate an immunity agreement and allow a subsequent party to be recommended for immunity. Also, although full immunity is only granted to one party, some form of favourable treatment may still be available to those parties wishing to subsequently cooperate with the Bureau.67 It is important to emphasize that, even if amnesty or immunity has been granted in one jurisdiction, a separate application must be made in Canada for immunity to also be granted. There is also an additional program available for immunity plus applications, a detailed discussion of which is beyond the scope of this paper.

D. International cooperation in antitrust investigations 1. Introduction The Bureau’s enforcement of the Act and its investigations into alleged violations of Canadian antitrust laws have taken on an increasingly international dimension. International investigations often require the Bureau to cooperate 64 Competition Bureau Information Bulletin, Immunity Program Under the Competition Act. See also Conformity Continuum Information Bulletin. Criminal prosecutions under the Competition Act are the responsibility of the Attorney General, and the Commissioner may refer a matter to the Attorney General for consideration when there is evidence of an offence, for whatever action the Attorney General may wish to take. The Attorney General has the sole authority to grant immunity to a party implicated in an offence under the Competition Act. The Bureau investigates the matter and makes a recommendation to the Attorney General. The Attorney General then independently considers if the interests of the public are best served by granting immunity. 65 Competition Bureau, Immunity Program—Frequently Asked Questions. 66 Competition Bureau Information Bulletin, Immunity Program Under the Competition Act. 67 See Conformity Continuum Information Bulletin.


12/10/07

13:09

Page 205

The Investigative Powers of the Canadian Competition Bureau 205 with antitrust authorities in various jurisdictions to ensure that anticompetitive behaviour is discovered and that the evidence required for domestic prosecution can be obtained from foreign jurisdictions (or vice versa). International cooperation in antitrust investigations occurs in a variety of ways. It may be informal on an ad hoc basis, or based on formal cooperation agreements which provide a framework setting out when and how cooperation should take place. It may occur via bilateral or trilateral cooperation arrangements that tend to be fairly specific in outlining the framework for cooperation, or via multilateral arrangements which tend to be more general. In addition, antitrust authorities also participate in efforts to co-ordinate investigations and enforcement activities with other jurisdictions through multilateral organizations such as the Organization for Cooperation and Development,68 the International Competition Network, the World Trade Organization and the United Nations Conference on Trade and Development.

2. Canada-US Cooperation Extraordinarily, the Bureau’s cooperation with US antitrust authorities began as an exercise in dispute management due to tensions arising from US investigations into several Canadian industries and Canadian opposition to the US approach to extraterritoriality.69 The first initiative was the 1959 Fulton-Rogers Agreement,70 an informal arrangement that established a communication channel to manage disagreements on competition matters. The Mitchell-Basford Understanding of 1969 and a Memorandum of Understanding (MOU) in 1984 also focused on conflict avoidance. Canada-US cooperation transformed from a defensive mechanism into a successful proactive relationship with the following three initiatives: —the signing of the Treaty on Mutual Legal Assistance in Criminal Matters between the Government of Canada and the Government of the United 68 In the field of antitrust, one of the OECD’s most practical contributions has been the nonbinding 1998 OECD Recommendations for Effective Action Against Hard Core Cartels. 69 In Canada and elsewhere, initial reactions to the extraterritorial application of antitrust laws resulted in a variety of measures including blocking statutes to prevent domestic implementation of foreign judgments.70 Sections 82 and 83 of the Competition Act specifically provide that the Commissioner may apply to the Competition Tribunal to prevent the implementation of any foreign judgment, law and directive which may have adverse effects on competition, trade or industry in Canada. However, most nations now seek to avoid enforcement conflict via international enforcement cooperation efforts. 70 The arrangement was named after the Canadian Minister of Justice and the US Attorney General of that time. It was established in the middle of a US investigation of a patent pool among a group of Canadian manufacturers. See von Finckenstein K. (2001): “International Antitrust Cooperation: Bilateralism or Multilateralism”, Presentation to a joint meeting of the American Bar Association Section of Antitrust Law and the Canadian Bar Association National Competition Law Section, May 31, 2001, Vancouver, Canada.


206

12/10/07

13:09

Page 206


States of America (the “MLAT”), which came into force on January 14, 1990;71 —the extension of the 1991 Extradition Treaty between Canada and the United States to offences punishable by the laws of both countries by imprisonment for a term exceeding one year or any greater punishment (which includes antitrust offences);72 and —the signing of the 1995 Canada-US Agreement regarding the application of their competition and deceptive marketing practices laws.73 The MLAT, in particular, has provided an unique basis for cooperation between Canada and the United States. Other cooperative agreements do not permit the same degree of information exchange and investigative cooperation as the MLAT does. For example, US Grand Jury Transcripts can be sent to Canadian antitrust officials even though these cannot be sent to state Attorneys General.

3. The MLAT The MLAT between Canada and the US is implemented in Canada by the Mutual Legal Assistance in Criminal Matters Act (“MLACMA”)74. The MLAT provides that Canada and the US will assist each other in “all matters relating to the investigation, prosecution and suppression of criminal offences.” The primary purpose of the MLAT is to seek evidence located in another jurisdiction of activity that is considered criminal under the laws of the requesting jurisdiction.75 Canada has entered into similar treaties with Australia, Bahamas, Mexico and the Netherlands.76 71 The Treaty between the Government of Canada and the Government of the United States on Mutual Assistance in Criminal Matters, Canada Gazette, Part I, 953(1990) [MLAT]. 72 Treaty of Extradition, March 22, 1976, United States-Canada, 1976 Can. T.S., No. 3 (as amended by an exchange of Notes on June 28 and July 29, 1974 and a Protocol dated January 11, 1988). This extradition process has been used in at least one case under the Act where Canada sought, and was granted, extradition from the US of an executive charged with misleading advertising. 73 Agreement between the Government of Canada and the Government of the United States of America Regarding the Application of their Competition and Deceptive Marketing Practices Laws, August 3, 1995. In a recent speech, the Commissioner stated that the Bureau makes use of the 1995 Agreement “almost on a daily basis.” von Finckenstein K. (2000): “Opening Remarks”, American Bar Association, Section of Antitrust Law Panel on Global Warming: International Reaction to the ICPAC Report, New York City, July 11, 2000. 74 Mutual Legal Assistance in Criminal Matters Act, R.S.C. 1985, c. 30 (4th Supp.) and SOR/90-704. 75 See Sullivan M. and Filion J. (2004): The Basics of International Cartel Enforcement in Canada, presented at the Osgoode Hall Continuing Legal Education Program: Canada’s Competition Regime: Thinking Strategically A Practical Guide for Business, January 14, 2004. 76 Australia: CTS 1990/2, Ottawa, June 19, 1989, entered into force March 14, 1990; Bahamas: CTS 1990/3, Nassau, March 13, 1990, entered into force July 10, 1990; Mexico: CTS 1990/34, Mexico, March 16, 1990, entered into force October 21, 1990; and the Netherlands: CTS 1992/9,


12/10/07

13:09

Page 207

The Investigative Powers of the Canadian Competition Bureau 207 Investigative assistance is available under the MLAT following the initiation of an investigation or prosecution by the requesting state.77 Examples of investigative assistance provided under the MLAT include: exchanges of information; the provision of documents, records and articles of evidence; execution of requests for search and seizure; and obtaining sworn testimony.78 Such assistance is provided pursuant to the MLAT even if the conduct under investigation is not the subject of a simultaneous investigation in the other jurisdiction. The conduct also need not be considered illegal in both jurisdictions.79 The MLAT is a supplement to other forms of cooperation. It expressly provides that the two countries may provide assistance pursuant to other “agreements, arrangements or practices.”80 An MLAT request itself may involve or lead to reciprocal assistance. In making a request under the MLAT the requesting state will provide information to the other because the MLAT requires the country seeking assistance to provide “such information as the Requested State requires to execute the request.”81 An MLAT request can therefore lead to the requested state initiating its own investigation if such information reveals conduct that may also constitute an offence under its laws. This has happened in at least one antitrust matter.82 The US and Canadian antitrust authorities do not necessarily publicly announce utilizations of the MLAT. However, in 1994 the MLAT was used in the context of price fixing conspiracies in the plastic cutlery and thermal fax paper industries.83 The thermal fax paper case was the first joint investigation The Hague, May 1, 1991, entered into force May 1, 1992. Treaties with the United Kingdom and Hong Kong are limited to narcotics offences. 77 In Canada, an investigation can be initiated by a crown prosecutor, an agency with criminal law enforcement responsibility or an agency with regulatory responsibilities. 78 In one case, a US district court, relying on its authority to respond to letters rogatory and its inherent jurisdiction, issued an order granting an Assistant United States Attorney the power to take whatever steps were necessary to respond to a request from Canada under the MLAT. The Assistant Attorney issued subpoenas to compel testimony: Rill J. and Goldman C. (1995): “A US and Canadian Perspective on International Antitrust Enforcement Post IAEAA, the 1995 International Guidelines and the Canada-US MLAT”, remarks to the Trade Policy Forum VIII, PECC, April 20–21, 1995, Taiwan. 79 MLAT, Article II(3). 80 MLAT, Article III (1). This provision is further reflected in subsection 3(2) of the MLAA, which states that: “[. . .] Nothing in this Act or a treaty [such as the MLAT] shall be construed so as to abrogate or derogate from an agreement, arrangement or practice respecting cooperation between a Canadian competent authority and a foreign or international authority or organization.” The MLAA defines a “competent authority” as the Attorney General of Canada, the Attorney General of a Province or “any person or authority with responsibility in Canada for the investigation of prosecution of offences,” which would likely include the Bureau. 81 Article VI(3). In particular, the information must include a description of the evidence, information or assistance sought and the purpose for which it is sought. 82 Rill and Goldman, supra note 78. 83 See 66 Antitrust & Trade Reg. Report 661 (1994) and 67 Antitrust & Trade Reg. Report 108 (1994).


208

12/10/07

13:09

Page 208


by the Bureau and US Department of Justice (the “DOJ”). As a result of information which was brought to the attention of the Bureau, both the Bureau and the DOJ commenced investigations that culminated in criminal charges in both countries.84 This cooperative investigation entailed exchanges of information, sharing documents and jointly interviewing witnesses. In the disposable plastic dinnerware case, pursuant to a US request for assistance under the MLAT, Canadian authorities raided the Canadian offices of one of the co-conspirators. As a result of this raid, a number of executives, including one Canadian, went to jail in connection with the resulting US prosecutions.85 The US and Canada have cooperated in a wide range of criminal investigations, including graphite electrodes, ductile pipe and bulk vitamins investigations.86 The success of the MLAT as an investigative tool in criminal matters was a major reason for recent amendments to the Act to include provisions allowing for MLATs in civil matters with qualified jurisdictions that enter a treaty.87 In 2003, the Ontario Court of Appeal upheld the Commissioner’s use of the MLAT to provide the US DOJ with assistance in investigating a criminal conspiracy in the United States.88 The Commissioner had made ex parte applications for search warrants and information gathering orders (section 11 orders). The Applicants brought an unsuccessful application to set aside the 84 On July 12, 1994, the Director announced that a Cdn. $950,000 fine had been levied against Kanzaki Specialty Papers Inc. for violation of the conspiracy provisions of the Act. See “Fine Levied under the Competition Act in Joint Canada-US Investigation,” Bureau of Competition Policy News Release NR-11526/94-15 (July 12, 1994). Two days later, the US Department of Justice issued a release announcing that charges were laid against Kanzaki, a Japanese officer of a related firm, Mitsubishi Corporation of Tokyo and Mitsubishi International Corporation of New York, and that the defendants had agreed to plead guilty and pay fines of more than US $6 million. See “Antitrust Division Breaks International Price Fixing Conspiracy in Fax Paper Industry”, Department of Justice Release (July 14, 1994). On the day the US charges were announced, an unprecedented joint press conference was held in the United States, attended by US Attorney General Janet Reno, US Assistant Attorney General Anne Bingaman and the Canadian Director of Investigation and Research. The coordination between the two governments included joint interviews of witnesses and, according to the comments of Assistant Attorney General Anne Bingaman at the joint press conference, the US sharing document summaries prepared by the Bureau. The authors are aware of other situations where the two governments have jointly interviewed witnesses in the context of a joint criminal investigation. 85 “Antitrust Division Breaks Price Fixing Conspiracy in Disposable Plastic Dinnerware Industry,” Department of Justice Release (June 9, 1994). 86 Valentine D. (2000): “Cross-Border/US Cooperation in Investigations and Enforcement Actions”, Canada/United States Law Institute, Case Western Reserve University School of Law, April 15, 2000; Stark C. (2000): “Improving Bilateral Antitrust Cooperation,” Competition Policy in the Global Trading System, Washington, D.C. June 23, 2000; Goldman C, Katz M. and Facey B. (2001): “Cross-Border Cooperation and Information Sharing in International Antitrust Cases: The Need for Balance,” Southwest Legal Foundation Annual Symposium on Private Investments Abroad, Dallas, Texas, June 19–20, 2001. 87 See sections 30.01 to 30.29 of the Competition Act as enacted by amendments contained in S.C. 2002, c. 16. 88 Ontario (Commissioner of Competition) v. Falconbridge Ltd., [2003] 24 C.P.R. (4th) 289, 24 C.P.R. (4th) 289.


12/10/07

13:09

Page 209

The Investigative Powers of the Canadian Competition Bureau 209 warrants and orders. The court concluded that the criminal conspiracy provisions of the Sherman Act were covered under the MLACMA, that the evidence gathering orders under the MLACMA did not attract the same protections as a search warrant under the Criminal Code, and that the MLACMA only requires that the relevant treaty provided for legal assistance with respect to the subject matter of the request. Given the success of the Canada-US relationship, efforts are now underway in Canada to extend this model of bilateral cooperation to the Bureau’s relations with other foreign antitrust authorities.

4. Investigative cooperation with other jurisdictions Canada has entered into cooperative agreements with Japan, Mexico, the European Community, Chile, Australia and New Zealand, and the United Kingdom regarding coordination of and cooperation in enforcement activities, recognition of each others’ interests, and notification of investigative activities. It should be noted that MLAT and other agreements do not override domestic confidentiality protections. The agreements between Canada and the European Community, and Canada and Mexico are very comprehensive (the other agreements referred to are not as detailed, but cover most of the same topics). They cover, amongst other things: (1) notification—any enforcement activities that may affect the interests of the other party including mergers, information seeking, and any judicial or administrative activity and shall be in advance with sufficient time for consultation; (2) consultation—either party may request consultation regarding any matter in the agreement; (3) coordination—obligations of both parties to assist the other’s competition authority and coordination of enforcement activities; (4) positive comity—where activities in one party’s jurisdiction has anti-competitive effects in the other, that party may request the other to initiate enforcement proceedings; (5) avoidance of conflict—where a conflict occurs between enforcement activities in each territory, the parties will work to arrive at an appropriate accommodation of competing interests; (6) exchange of information—both parties agree to exchange information upon request by the other that is relevant to the enforcement being contemplated or conducted; (7) semi-annual meetings; (8) confidentiality—each party shall, to the fullest extent possible, maintain the confidentiality of any information communicated under the agreement.


210

12/10/07

13:09

Page 210


5. International co-ordination The combination of a) immunity programs in Canada, the US and elsewhere; b) waivers of confidentiality (most often given by an immunity/amnesty applicant in favour of a foreign enforcement authority) and c) MLAT and other agreement provisions—all have led to far greater enforcement co-operation. In recent years we have seen an increasing number of co-ordinated cartel investigations, including the use of simultaneous dawn raids. Such coordination leads to separate, but parallel investigations which must then be responded to on a co-ordinated, multi-jurisdictional basis on the defence side.

E. OECD developments The issue of information exchange between and among enforcement agencies in international cartel investigations has long been a topic of interest to the OECD’s Working Party 3 on International Co-operation.89 While there is a general consensus that information exchange is important to the laudable goal of anti-cartel enforcement, the business and legal communities in particular have stressed the need for adequate protection of confidential business information, due process and legal privileges. In October 2005, the OECD released its Best Practices for the Formal Exchange of Information Between Competition Authorities in Hard Core Cartel Investigations (“OECD Best Practices”). The OECD received and considered submissions from the Business and Industry Advisory Council to the OECD (BIAC)90, the American Bar Association, the International Chamber of Commerce and the International Bar Association. The OECD Best Practices address suggested safeguards in four principle areas: authority to exchange information; provisions concerning confidentiality, use and disclosure in the requesting jurisdiction; protection of legal professional privilege; and notice. A key area of some controversy was the issue of notice. Many in the business community expressed the view that affected parties should receive advance notice before their information was provided to a requesting agency so long as it would not prejudice the investigation. The OECD adopted the position that advance notice should not be given where the other safeguards have been complied with; however, where 89 See for example the OECD’s January 24, 2002 paper on “Information Sharing in Cartel Cases—Suggested Issues for Discussion and Background Material”, although OECD’s consideration of these issues certainly predates this paper. 90 Cal Goldman served as Chair of the BIAC Working Group on Information Exchange and Rob Kwinter served as the Working Group’s principal drafter.


12/10/07

13:09

Page 211

The Investigative Powers of the Canadian Competition Bureau 211 an agency does provide notice (which can be before or after the fact) the agency should ensure that such notice will not violate a court order, domestic law, or an obligation under a treaty or other international agreement, or jeopardize the integrity of an investigation in either the requesting or requested jurisdiction. It remains to be seen how any future co-operation instruments treat the issue of notice.

F. Conclusion The movement toward greater international cooperation in antitrust investigations is unlikely to abate. Indeed, the increase in the number of countries that have cartel laws, the proliferation of amnesty programs and the growing number of international agreements between and among jurisdictions, all point to greater international cartel enforcement. Advances in technology are also likely to lead to increased investigative tools and cooperation in antitrust enforcement as multinational businesses increasingly use computers to link their international business operations and share information across jurisdictional boundaries. This also suggests that parties that are the subject of a multi-jurisdictional competition law investigation and their counsel should carefully co-ordinate their responses to any investigation to ensure consistency of approach and to ensure that steps taken in one jurisdiction do not prejudice the party’s position in another. The many jurisdictions that have “first in” rules for immunity eligibility also means that time may be of the essence, often on a simultaneous basis. It is imperative, for example, that a firm contemplating an immunity application in one jurisdiction quickly ascertain which other jurisdictions might be involved, and immediately obtain expert legal advice in respect of the application of both the substantive and procedural cartel laws in every such jurisdiction, and then implement steps to ensure such protection as may be available for the firm in each of those jurisdictions. Greater co-operation across jurisdictions will likely be seen both in the public and private sectors arising from cartel cases that have international dimensions. The dramatic rise in international agency cooperation, the broad array of investigative tools available to antitrust agencies in Canada and elsewhere, and the prevalence of leniency programs all signal that the likelihood of detection of antitrust violations is growing on a higher tangent. Accordingly, companies will be well served by the implementation of, and strict adherence to, antitrust compliance programs on a worldwide basis.


12/10/07

13:09

Page 212

(M) Ehlermann 07 Guersent

12/10/07

13:09

Page 213

V The EU Model of Administrative Enforcement Against Global Cartels: Evolving to Meet Challenges

Olivier Guersent 1 I would like to articulate my remarks around three themes. First, I should like to emphasize that the combination of investigation and decision-making in competition cases in one authority under the control of reviewing courts is a legitimate institutional choice, whether it be on grounds of efficiency, transparency or due process. Second, I would like to underline that this does not mean that the system cannot be bettered by organizational improvement and/or by adding new tools (e.g., direct settlement for cartels). Thirdly, I will stress that this does not mean that the system cannot be complemented by, and smoothly interact with, both civil and criminal enforcement.

A. A legitimate system that has proved its efficiency Firstly, it is the firm conviction of the European Commission that the Community administrative system of investigation ensures effective and timely investigation and decision making by an authority with a high level of expertise. We believe that this has been amply demonstrated in the past, not least over the last 10 years since the entry into force in 1996 of the first Community Leniency programme and the creation in 1998 of the first cartel unit. Secondly, due process and the procedural rights of undertakings are ensured. The administrative procedures ensure a high degree of transparency (in the form of Statements of Objections, hearings, reasoned decisions, etc.), which the Commission constantly seeks to improve (see, e.g., the recent updated Notice on access to the file). The fairness of proceedings and quality of decisions are supported by in-house checks and balances, notably due to the scrutiny of the Commission’s Hearing Officers.

1 Olivier Guersent is the Deputy Head of Cabinet of European Competition Commissioner Neelie Kroes. The views expressed are those of the author and do not necessarily reflect those of his employers. The author would like to thank Kirti Mehta, Flavio Laina and Frederique Wenner for the work on which this contribution is based. Thanks are also due to Rebecca Ellis for her kind help in drafting this contribution.


214

12/10/07

13:09

Page 214


Finally, the decisions of the Commission in the competition field are subject to effective judicial review by an impartial tribunal: the European Courts exercise strict control over the decisions of the Commission. All parties concerned by Commission antitrust decisions can bring actions before the Court of First Instance (CFI) seeking the annulment of the contested decision or the reduction of the fine imposed on them. The case law clearly shows that the CFI’s scrutiny is intensive and that the court does not refrain from meticulous and stringent review of the Commission’s decisions. This scrutiny extends to verifying the accuracy of the facts the Commission has relied upon. It can even be argued that the experience gained by the CFI in antitrust cases may represent an advantage of the EU’s administrative procedures compared to the US adversarial process, where the courts called to decide on the US agencies’ allegations do not necessarily have any specific knowledge of competition issues.

B. Improving and evolving So yes, the current system of administrative enforcement, both at EU and at national level in Europe, is clearly legitimate and efficient. Does this mean that it cannot be bettered or complemented? Certainly not, and the Commission is actually taking steps in that regard.

1. Improving the organizational framework The fight against cartels is one of the key areas in which DG Competition is and will be concentrating its efforts. Consequently, one of the very first decisions that Commissioner Kroes took was to create a Cartel Directorate in DG Competition. This decision became effective on 1 June 2005 and the cartel work in DG Competition was reorganized by grouping together some 60 specialized staff members2 in a single Directorate specifically dedicated to the enforcement and development of competition policy in relation to cartels. On 1 December 2005, a fourth Operational Unit was created and added to the three existing ones in the Directorate. The Directorate brings economies of scale and pools the Commission’s cartel knowledge. The aim is to bring together a body of specialized case handlers that have primary responsibility for the handling of cartel cases and sole responsibility for dealing with immunity applications under the Commission’s leniency programme. Sectoral 2

These comprise around 45 case handlers and 15 support staff.


12/10/07

13:09

Page 215

The EU Model of Administrative Enforcement Against Global Cartels 215 Directorates will continue to have an ongoing role in the detection and prosecution of cartels. We expect several benefits from this restructuring. Firstly, cartel work has a strong legal focus, both in the investigation and development of a case, and in the need to apply correct procedures and ensure that the rights of defence are respected. Consequently, a certain degree of specialization and expertise is needed for effective action against cartels. We expect the reorganization to reinvigorate the Commission’s cartel work and lead to more intensive involvement in policy development, ultimately resulting in a more effective fight against cartels and more decisions against cartels in future. Let me be clear on this: our aim is to reach—on a sustainable basis—a credible level of enforcement so as to provide for an adequate level of deterrence. In practical terms, this means a level of enforcement in the range of the 10 cartel decisions per year that we used to have in 2001–2002. We expect to be there already in 2006, for the first full year of operation of the Cartel Directorate. This represents around twice as many decisions when compared to the final year before the creation of the Directorate, i.e., 2004. Secondly, the new Directorate is to be the focal point for the outside world on cartel issues. This includes not only immunity and leniency applicants but also operational cooperation with other competition authorities both within the EU and with our main partners outside the EU. There are also sources of further efficiency gains in this domain. For example, we are working hard within the ECN to reach a satisfactory solution on the issue of the transfer of immunity when transferring cases within the network. Once this is done, it should allow the Commission to better prioritize its cartel enforcement. Currently, there are a number of cases that the Commission has to retain just to safeguard the position of an immunity applicant who risks not being the “first through the door” in the national leniency regime. In the future, the Commission will no longer have to retain these cases. Another example is bilateral cooperation agreements with non-EU jurisdictions. Competition authorities must work together if they are to match the increasing ability of cartels to operate on a global scale. This calls for a new generation of antitrust cooperation agreements that will allow authorities to share evidence. This is something we are reflecting on with our main partners and it should translate into further effectiveness and efficiency gains in our fight against global cartels. Finally, the Cartel Directorate also maintains and develops further relations of confidence and trust with the business and legal community that are essential to an effective leniency framework.


216

12/10/07

13:09

Page 216


2. Getting the administrative enforcement system to adapt and evolve The Cartel Directorate also has a key role in bringing together operational expertise and policy making to ensure that new policy formulation in the cartel area further improves the overall efficiency and effectiveness of the Commission’s fight against cartels. Again, I want to be clear: we have streamlined our procedures so as to deal with an average cartel case (based on an immunity application) within 24 to 30 months. With 45 case handlers working in case teams of two people, this means that our production capacity is at best around 10 decisions per year—a level that we hope to reach already this year. Although cartels are high on the Commissioner’s agenda, internal staffing decisions have to take account of competing priorities, notably in state aid and mergers, as well as the need to enforce non-cartel Article 81 and 82 work and to perform a number of basic functions of market surveillance, including the detection of cartels not reported through leniency or sector inquiries. This clearly means that there is little prospect in the short term of a further sharp increase in the resources dedicated to anti-cartel enforcement. However, this should not inspire a sense of security amongst cartelists, since the Commission is adept at flexibility. In this context, and while the 2002 leniency programme generates over 20 immunity applications per year on average, it is evident that improving the organizational framework within the Commission and refining our cooperation with sister authorities within and outside the EU cannot be our only answer to this challenge. It is therefore clear that, in order to avoid the creation of a structural backlog, the Commission has to find alternative ways to deal with cartel cases by fine tuning the existing enforcement tools and forging new tools which complement those we have today. I shall focus on what we consider to be the most promising policy evolution in this respect. In the spring of 2005, Commissioner Kroes announced a string of new measures aimed at addressing the challenges I have just described. The centrepiece is without doubt what she referred to as a “European plea bargain”, which can accurately be described as a direct settlement system for cartels. My colleagues in DG Competition and in the Legal Service are currently working on the possible design of such a direct settlement system. It is too early to be more specific for the time being—I hope we will be able to present our first views toward the end of the year. But in my view, any system must meet a minimal cahier des charges: it must be attractive, clear and transparent for business; it must bring about administrative efficiencies for the Commission; and it has to minimize the necessary legislative changes, if any.


12/10/07

13:09

Page 217

The EU Model of Administrative Enforcement Against Global Cartels 217

C. A system that could and should be complemented It is fair to say that the current EC administrative enforcement system forces the Commission to play on one string what our colleagues in other major jurisdictions—and notably in the US—can play on several strings (administrative enforcement; criminal enforcement; and civil enforcement). However, in the specific institutional context of the EU, both civil and criminal enforcement trigger particular complexities. We are firmly convinced that more effective private enforcement is an essential complement to public enforcement within the EU. The role of the public authorities—and by implication the administrative system—will continue to be of critical importance in detecting and deterring antitrust practices such as cartels. Private action cannot and should not replace public enforcement. This is particularly so when one considers the lack of pre-trial disclosure in most Member States. Why are antitrust damages actions such a vital counterpart to enforcement actions by competition authorities? First, because the victims of illegal anticompetitive behaviour should be compensated. Second, because more private actions will increase deterrence and increase compliance with the law. Thirdly, because private actions will further develop a culture of competition amongst market participants, including consumers in Europe, and will raise awareness of the competition rules. The Green Paper published at the end of last year on damages claims for compensation for antitrust violations identifies obstacles to a more effective system of damages actions for breach of the EC antitrust rules. Various options designed to address each obstacle are presented for discussion. We are currently analyzing in detail all the contributions we received in the public consultation, which we will take into careful account before deciding on any further course of action. Deterring undertakings and individuals from entering or re-entering cartel arrangements is one of our main objectives. We therefore welcome all sanctions, including criminal sanctions, which contribute to the deterrent effect on cartels. As regards criminal enforcement, you all know that the Commission does not impose criminal sanctions, and it is very unlikely that it will be in a situation to do so in the foreseeable future. However, a certain number of Member States do provide for criminal sanctions for breaches of EC competition law. This is the case, for example, in Ireland (where the first individual was very recently sentenced in connection with a price-fixing cartel) and in Estonia (where imprisonment for up to 3 years is possible). In other Member States, such as France, there are distinct but connected criminal infringements for those individuals that are personally involved in a cartel (also with jail terms


218

12/10/07

13:09

Page 218


of up to 3 years). In some other Member States, criminal sanctions are attached to certain types of infringement of competition rules (e.g., bidrigging in Austria and Germany; dishonest behaviour in the UK). There are also currently debates in Sweden concerning the possible criminalization of competition infringements. However, experience in our Member States shows that introducing criminal sanctions for individuals will not necessarily improve the efficient enforcement of competition law as such. These sanctions imply a higher standard for the rights of defence for the individuals, but often also for the undertakings. Criminal proceedings also trigger a higher standard of proof for the authority/prosecutor. It follows that, for criminal sanctions to be an effective complement to administrative enforcement, they must be accompanied by very effective investigative powers, and the judicature should be ready to impose to their full extent the sanctions provided for by law. In cases where criminal enforcement at national level would coexist with administrative enforcement at EU level, a third condition must in my view be fulfilled. There should be an efficient criminal leniency programme for individuals at national level that smoothly interacts with the EC administrative leniency regime for undertakings. Indeed, as regards cartels falling within the scope of Article 81 and covering several Member States, it is unlikely that anyone in their right mind would report the administrative infringement and obtain immunity for the company in Brussels if they run the slightest chance of facing criminal prosecution and a jail sentence for the same facts in one or several Member States. Therefore, much depends upon the coexistence of national criminal leniency programmes with efficient and transparent coordination mechanisms between EC-wide administrative corporate immunity and Member State-wide personal criminal immunity. Depending on how the two interact, criminal enforcement by Member States within the EU could either be an impediment or an asset in the global fight against cartels in Europe.

D. Conclusion No enforcement systems should or could be evaluated in abstracto. In other words, there are good and bad enforcement systems. I rather believe that enforcement systems can only be properly evaluated within the context of the legal culture that has produced them. As regards the EU, it is therefore my conviction that the current system of administrative enforcement, with its advantages and its drawbacks, is probably the best possible system for the EU. Other systems, like the American one, have other advantages but they also have other drawbacks. As Don Klawiter said recently during the ABA


12/10/07

13:09

Page 219

The EU Model of Administrative Enforcement Against Global Cartels 219 Spring 2006 meeting (referring to the consultation on private enforcement), “do as we say, do not do as we do”. This of course means that each of us—as public enforcers—should evaluate our own system on a regular basis. And we should look closely at other systems abroad, with a view to adapting our enforcement organizations, methods and tools so as to better meet our basic joint goal: eradicating cartels.


12/10/07

13:09

Page 220

(N) Ehlermann 07 Session B Pt 1

12/10/07

13:09

Page 221

SESSION B

THE INSTITUTIONAL FRAMEWORK

1 PANEL DISCUSSION

Participants: Allendesalazar, Rafael Barnett, Thomas Bellamy, Christopher Bloom, Margaret Burrichter, Jochen Calkins, Stephen Calviño, Nadia Collins, Philip Ehlermann, Claus-Dieter Fingleton, John Forwood, Nicholas Goldman, Calvin Grout, Paul Harding, Christopher

Harrington, Joseph Joshua, Julian Klawiter, Donald Komninos, Assimakis Kovacic, William Lawrence, Jon Motta, Massimo Rey, Patrick Roth, Peter Siragusa, Mario Venit, James Waelbroeck, Denis Wils, Wouter Woods, Donncadh


12/10/07

13:09

Page 222


12/10/07

13:09

Page 223

Panel Discussion

223

Session B: The Institutional Framework Panel III: Public Enforcement (Administrative and Criminal) JOHN FINGLETON—The discussion this morning already introduced the subject that we will discuss this afternoon, namely public and private enforcement. We have already seen that there are contrasting views on the two sides of the Atlantic on the subject of criminalization. For example, our colleagues from the US who have worked either in the public or private sector tend to see this differently from the EU colleagues who have worked in an administrative system. I found it interesting that many of the negative comments made this morning about leniency programmes were actually related to civil/administrative law systems, and in a criminal enforcement system many of those problems seem to dissolve. For example, if criminal sanctions are available, and one can reduce fines and increase personal sanctions, this may reduce some of the industry-wide harmful effects of cartelization. (The “prize” for this morning went to Mario Siragusa for his comment on the “dismissal of staff”, which I thought to be a wonderful proposal for increasing the supply of efficient cartel organizers . . .) 䉴

JAMES VENIT—In my presentation I will focus on the conflict that arises from the same case being pursued in two different enforcement systems. The result is essentially one of leaning into the rights—or legal protection—in one of the two systems. One set of rules prevails over the other, conceivably to the prejudice of the other enforcement system and of the rights of the individuals concerned. The bottom line is that there is a need for some degree of harmonization, although this is very difficult to achieve, because the kind of conflict that arises is difficult to reconcile institutionally. I have in mind three examples, one dealing with differences in the leniency programmes —which I am not going to discuss in detail now. The second deals with differences in legal privilege—which I will go into only briefly, because I do not want to sound like I’m pleading for some approach to legal privilege. One of the consequences of the European approach, where in-house counsel privilege is not yet recognized, can be far-reaching. If, for instance, a Commission inspection seizes documents considered to be privileged in the US, then in US proceedings the question arises as to what are the consequences of those documents having been used in Europe. Can the US judge decide that the privilege has been waived because the documents have been used in a non-privileged context? Or can he still continue to treat the documents as privileged? This is a question to which I am not sure there is a clear answer in US law, but I know that companies I have personally advised have been affected by this problem. The consequences of this situation may be

䉴


224

12/10/07

13:09

Page 224

Session B – The Institutional Framework

quite draconian, because in the US there is a doctrine called “subject-matter waiver”, whereby, when you lose the privilege, it can be not only for the specific document, but for the whole subject matter. So, that is one example of how the lack of harmonization in approaches can lean into individual rights. Another interesting example is the converse, namely, how a criminal jurisdiction can lean into the efficiency of an administrative jurisdiction and its investigative procedures. Just to give you a concrete example, in the EU the Commission has the power to inspect, seize documents and also to question employees, so long as it does not go over the line by phrasing the questions in a way that will incriminate the company. But in the EU system the individual is not liable if he provides untruthful answers. The company may be liable. The company is also under an obligation to cooperate during the inspection, which means probably that the employees should produce forthcoming answers. Suppose you have a criminal investigation on the same cartel in the US, and one of the people that the Commission wants to talk to is a potential target for criminal prosecution by the US Department of Justice. Is that individual going to be willing to answer the questions posed by the Commission inspector, or is he going to be concerned that his answers could cause him personal liability in the US? Again, it is a question to which I do not know the complete answer, but it certainly seems to be a potential risk. Another interesting situation, when you are the lawyer of the company that is being inspected, is: do you have the ethical obligation to alert the client to the fact that he may subject personally to the risk of criminal prosecution in another jurisdiction, and that he should be getting his own counsel? These are issues that never came up until we started to have criminal sanctions in the EU. I think that 10 or 15 years ago, nobody would have thought of telling the executive of a company that he had better get his own legal counsel because he may be prosecuted in another jurisdiction. But now that is much more of a reality. This can impact the effectiveness of the Commission’s proceedings, because now we need to protect the rights of the individuals. Solutions? There are solutions if the other legal system protects the rights of individuals—for instance, a US court deciding not to use a particular testimony because it was taken in a system where the level of protection of individual rights is not as high as in the US. But this is not necessarily the outcome you will always get. Another solution would be an agreement not to use instruments that hamper each other’s inspection procedures, but that is highly unlikely to happen. This is one of the reasons why I think that criminal enforcement will overtake, and eventually replace, civil/administrative enforcement. Everything you do in a case is driven by the criminal aspects. If the case is pursued in both a criminal and a non-criminal law system, this will affect the functioning of the latter. People will start to realize this, and it will become one of the strong arguments in favour of moving towards criminalization. Just one comment on the discussion this morning: I thought it was a fascinating study in cultural differences. European colleagues talk about the rights


12/10/07

13:09

Page 225

Panel Discussion

225

of defence being disturbed by the use of oral testimony, whereas in the US everything is done basically in an oral form—initial affidavits, plea bargaining statements. . . . This huge cultural difference is fascinating to observe, but also a bit disturbing at some point. NADIA CALVIÑO—Before I was invited to this Workshop I never had the chance, or I was never forced to think about sanctions. This invitation is particularly timely because it coincided with the fact that, in Spain, we are in the middle of a drastic revision of the competition statute, and the setting of fines will be one of the issues tackled in the process. Thus, in my written contribution I tried to look for practical answers to the Spanish questions. Since we are not contemplating criminalization of antitrust enforcement for the moment, I focused on fines because it is a more practical approach. When I started to look into the issues of fining, I realized that there are two main principles that may clash: effectiveness and deterrence versus justice and proportionality. I also noticed that the idea of deterrence is very attractive to economists and enforcement agencies, whereas, perhaps not surprisingly so, the idea of proportionality is more often heard from lawyers and firms. There are different theories on the relationship between deterrence and proportionality in the setting of fines. One of the main issues is how to set the initial amounts, and then reduce or increase them according to the particular circumstances of each infringement. There are mainly two approaches to setting the basic amount of the fine: a more ex ante, structural approach—which was until now the approach of the EC Commission—whereby we graduate fines according to the seriousness of the various types of infringement. The second approach will probably play a greater role in the EC in the future, and it is related to the economic effects of the conduct—the volume of sales in the relevant market, the market share of the offenders, etc. There is nowadays a certain tension between these two approaches, at least at the EC level, and we are currently discussing the drafting of the new Guidelines on the methods of setting fines.1 This tension— between an approach of ex ante, formal rules—which I believe lawyers prefer, at least until their clients get very high fines—and the more economic approach—which is more demanding on the competition authorities and also on firms, because it requires more accurate and detailed market data which is not readily available—will be discussed in the next months, and it certainly was also given some thought in Spain while the new competition law was being drafted. I think we should try to get the best of both approaches. In my written contribution I reviewed the pros and cons of each approach, and I discussed additional criteria that could be taken into account.

䉴

1 OJ C 210 [2006], text available at http://eur-lex.europa.eu/LexUriServ/site/en/oj/2006/c_210/ c_21020060901en00020005.pdf.

(Y) Ehlermann 07 Klawiter

12/10/07

13:13

Page 489

I Donald C. Klawiter* US Corporate Leniency After the Blockbuster Cartels: Are We Entering a New Era?

The Antitrust Division’s Corporate Leniency Policy1 is, without any doubt, the single most effective tool in the detection and prosecution of cartels ever devised by the enforcement community. As the Division approaches the thirteenth anniversary of the Corporate Leniency Program this summer, there is no dispute that this policy deserves the credit for the stunning success of the Division’s international cartel enforcement during the last ten years and, as a result, aggressive enforcement around the globe. If imitation is the sincerest form of flattery, the Division must be enormously flattered that many jurisdictions, including Canada,2 the European Commission3 and many of its member states, Brazil, Australia, Japan and Korea, among many others, have enthusiastically jumped on the leniency bandwagon. This paper will describe and evaluate the American leniency policy and consider the benefits and drawbacks of the policy from the prospective of the corporate applicant.4

A. The US leniency policy 1. The origins and components of the policy The Antitrust Division announced its first leniency policy on October 4, 1978, when Assistant Attorney General John H. Shenefield startled the business * Donald C. Klawiter is a partner in the Antitrust Practice Group at Morgan, Lewis & Bockius LLP in Washington, DC and is Chair of the ABA Section of Antitrust Law. His practice focuses primarily on international cartel investigations and cases, criminal and civil, internal investigations, compliance and corporate governance advice in these matters. 1 US Department of Justice, Antitrust Division, Corporate Leniency Policy (August 10, 1993), reprinted in 4 Trade Reg. Rep. (CCH) para. 13,113. 2 The Competition Bureau (Canada), Draft Immunity Program Under the Competition Act (February 17, 2000), available at http://strategis.ic.gc.ca/SSG/ct01705e.html. 3 Commission Notice on Immunity From Fines and Reduction of Fines In Cartel Cases, February 19, 2002, OJ C 45 [2002]. 4 This article is an updated version of an earlier article, entitled “Corporate Leniency in the Age of International Cartels: The American Experience”, published in 14 Antitrust Magazine (Summer 2000).


226

12/10/07

13:09

Page 226


I would like to conclude with three points. First, I agree with those who have pointed out that we should be sceptical about the effectiveness of fines, and that criminalization is probably more of a deterrent than fines. Fines set too high lead to less competition on the market—the point of Mr Siragusa. If they are too low, they are not deterrent at all, and anyway, ultimately it might be the consumer paying the fine in 2 years’ time. But we should also be realistic and move step by step: in some jurisdictions, including Spain and the EC, criminalization is not on the agenda for the time being. Second, there is nothing but advantages in laying down very detailed rules on fining. This is not so obvious to some people, but I believe that the more detailed we can get in our laws, the better for the competition system as a whole—first of all, because market operators, which are rational animals, take decisions based on a cost-benefit analysis, so the more they know about how they will be punished if they infringe the law, the better the information they have for taking the right decision, i.e., not to infringe the law. Second, this guarantees legal certainty, or at least enhances it. Third, it makes the work of enforcers much easier: if we have clear criteria for setting the final fine, we have an easier task in making our decisions, and judicial review will also be easier. Finally, striking the right balance implies taking into account the economic approach, but at least keeping in mind an ex ante criterion, namely, that more serious infringements will draw a maximum penalty, or at least a penalty greater than that for less serious infringements. MASSIMO MOTTA—This presentation draws on a paper that I am currently writing together with Gregor Langus, an expert in econometrics who is at the EUI.2 The purpose of this research, which is still very preliminary,3 is to promote a debate on the level of antitrust fines, and in general, on antitrust deterrence. The project starts from anecdotal evidence which seems to point to the fact that there are lots of repeat offenders of antitrust law and the fines do not seem to trigger any management changes. The question that we want to address is: are the fines set at the right level? Are firms seriously affected by the fines or not? The idea is to explore the effect of antitrust fines and antitrust investigations in general in the EU on the share prices of the firms that were fined. To the best of our knowledge, this is the first research of the kind in Europe. There are several reasons why a firm might see a reduction in its market value in relation to an (expected) antitrust fine, or more generally, when an antitrust investigation is opened. First, there is a direct effect, in that the pay䉴

2 The final version of this research paper was published in November 2006. See Langus G. and Motta M. (2007): The effect of EU antitrust investigations and fines on a firm’s valuation, http://www.iue.it/Personal/Motta/Papers/FinesFebruary2007REV.pdf. 3 See previous footnote for updated results.


12/10/07

13:09

Page 227

Panel Discussion

227

ment of the fine will diminish the firm’s cash, and this directly affects (in a negative sense) the value of the firm itself. We estimated this direct effect to be between 30 and 70%. The lowering of share prices is also due to other indirect effects, for instance: the fact that, due to the diminution of assets available to the firm, there are some profitable projects that will not be undertaken. There could also be legal costs; and probably even more important, the closure of an antitrust investigation, sometimes even without a negative decision, signals to the market that a profitable activity—for instance a cartel—will be ended. For all these reasons, once an antitrust investigation is opened, or an infringement decision is taken, we should observe a reduction in the value of the firms. We isolate four events typical of an antitrust investigation in Europe, i.e., the dawn raid, the Statement of Objections, the Commission Decision and the judgment of the ECJ, and we observe how they reflect on the market value of the firms. We explored a database composed of all the antitrust decisions taken by the Commission between 1985 and 2005—which is a huge pool of observation, but since we look at the effect of the fine on firms that are quoted on the stock markets, obviously the pool of cases is reduced. In some cases, since we have to isolate the precise moment in which an event takes place, the number of cases is reduced even further—for instance, in the case of dawn raids, in several instances we do not have the exact date when the dawn raid took place. The idea behind this methodology—which is standard in the finance literature—is that, if markets are efficient, stock prices will correctly reflect the value of the firm in the eyes of investors. In particular, all the relevant available information about a certain firm will have an impact on share prices. If there is any news about a change in the fundamentals of the firm, it will be immediately reflected in the share prices. This implies that, first, if we know the moment at which a certain piece of news about a fundamental change arrives on the market—for instance, if we know the moment in which an antitrust proceeding is started—and second, if we are able to estimate the share price which would prevail if this news had not arrived to the market, then we would be able to compute the value of the news. What we do, basically, is compute the difference between the counterfactual price and the real price. Obviously, the main difficulty here is to calculate the counterfactual price, and there are techniques that we use for this purpose that I will not describe here. The first interesting result is that the dawn raid, which is the really unexpected news breaking into the market, seems to have a very strong statistical effect on the firm’s share price. In the case in which a dawn raid takes place, in our sample the return of the firms will be 1.2% lower than the counterfactual return. This drop occurs on the very same day of the news. But also in subsequent days there is an overall drop in share prices of about 2.2%. The latter number is the strongest effect we found in our sample, and this does not


228

12/10/07

13:09

Page 228


change depending on whether you look at Article 81 EC or Article 82 EC cases. What is more controversial is whether the Commission’s Decision has an effect on the firm’s valuation. Here, if you look at the whole sample—that is, Article 81 and 82 EC cases together—there is a further drop on the order of 3%, but this is not so if you only look at Article 81 EC cases. Second, and not surprisingly, we find that the Statement of Objections does not have any effect whatsoever on share prices. Perhaps more surprisingly, ECJ judgments that lower or significantly reduce the fine do not have any impact whatsoever. To sum up, the research tells us that the antitrust investigation and decision will have an impact of between 2.2% and 5.2% on the share prices. Is that a lot, or is that too little? I would like to conclude by referring to research conducted by Prof. John Connor, who has surveyed a lot of empirical evidence about the impact that cartels have on market prices.4 Prof. Connor estimated that a new cartel’s price over-charge is on the order of 25% per year. This should give you an idea about the total harm to consumers: you have to multiply that for the number of years for which the cartel lasts, and of course, we should also add the deadweight loss. So, on the one hand, we have this drop of 2.2%, or at maximum, 5.2% of share prices; and on the other hand we have fines on the order of about 75% of the excess profit extracted, if we want to have some correcting effect of the fine applied. Of course, if we take into account that there should also be a deterrence effect, the level of the fines should be even higher, because fines are applied only to cartels that are discovered. If we want to measure the deterrence effect, we should also take into account how frequent antitrust fining is. 䉴 JOHN FINGLETON—It seems to me that your research brings up the question of whether it is actually the shareholders, or the consumers, who eventually pay the fines. Earlier we heard the implicit assumption that it is always the consumer who pays, but this is evidence that at least some shareholders do pay some of the fines.

䉴 STEPHEN CALKINS—In my presentation I will address three main points. First, what does it mean to have criminal antitrust enforcement? Second, what are the consequences of adopting criminal antitrust enforcement. Third, I will put forward a couple of reflections on the European situation. In the US, when we talk about criminal enforcement, we may be referring to three different things. One is: do we punish the individuals or the entity?

4 For a list of his recent work in this area, see http://www.agecon.purdue.edu/directory/ details.asp?username=jconnor.


12/10/07

13:09

Page 229

Panel Discussion

229

Second, the proceedings criminal or civil? Third, do sanctions involve prison, or something else? On the question of punishing individuals, as opposed to punishing the entity, I am firmly in favour of the former, for reasons of deterrence. Some years back, while I was counselling a company, their internal counsel said to me, while pointing to an FTC decision whereby the individual executives of a company were obliged to pay large amounts of money: “Thank goodness that the FTC made the president and chief executive pay personally. That is the only way I can get my president and chief executive to listen to me.” For purposes of deterrence, personal punishment is much more effective. The other argument for individual punishment—which sometimes is not mentioned—is one of fairness: if a company makes a lot of money by running a cartel, it is quite unfortunate if some future stock purchaser should pay an inflated price for that over-valued company and then suffer from having to pay a share of the fine. The Antitrust Division of the US Department of Justice has in the last decade indicted more individuals than corporations. They are firmly— and, I think, wisely—in the business of going after the individuals. On the question of criminal versus civil proceedings, I support criminal proceedings, for a variety of reasons, including the question of stigma. Only last week the state of Connecticut won a consent order against an insurance company that had apparently engaged in massive wrongdoing. The insurance company issued its own press release claiming that what they did was “entirely lawful”. Companies do not release press declarations like that, if they are pleading guilty to crimes—and I think that is one of the reasons why criminal sanctions make a difference. On the other hand, it is only fair to mention that the US Justice Department is in fact of two very different minds when it comes to crime and corporations. On the one hand, the Antitrust Division is in the business of indicting corporations for the crime of price-fixing; on the other hand, the Department of justice goes after major corporate fraud by indicting individuals and not corporations. If you go to the fraud website of the Department of Justice, you will find a list of 92 companies that have allegedly engaged in fraud (including Enron), and 381 individuals who have been indicted, but only one company that was ever indicted—Arthur Andersen (and anyway, the US Supreme Court tossed that out later). So, the fraud section of the Department of Justice does not indict corporations; it rather enters into deferred prosecution agreements, saying: “Try to get your house in order, and we will worry later about whether to indict you.” The Antitrust Division proceeds very differently. With respect to prison, clearly it is the greatest deterrent, and you already heard from Thomas Barnett that prison time is up in the US, both in terms of aggregate days and in terms of the individual sentences. As to the consequences of moving to a world of criminal penalties, I will talk first about consequences for the Department of Justice, and second, about the consequences for substantive antitrust doctrine.


230

12/10/07

13:09

Page 230


The Department of Justice, which of course is one of the two major antitrust enforcement bureaus in the United States, now overwhelmingly brings criminal cases. If one looks at the number of cases brought—there is a graph in my written contribution—you see that the criminal cases dominate over civil cases. The Department of Justice is clearly in the business of criminal law enforcement. If you look at the appellate briefs that the Antitrust Division files, in the last 10 years 26 cases were in criminal matters, while there were amicus briefs in 42 private matters, and appellate briefs in only 14 civil matters, because the Division does not do that much civil enforcement anymore. Indeed, if you look at the history of US antitrust enforcement, before 1978, US antitrust law was made largely by the cases brought by the Department of Justice, whereas nowadays much of what we in the US know as “antitrust law” is made of cases brought by private individuals and by the FTC—which does not handle criminal law enforcement. The Department of Justice is playing a different role. It is so attracted (and understandably so) by the criminal world that it does not do much civil litigation. How much of that is because of the penalty system, or because of policy preferences, is hard to tell. But the fact that the FTC operates quite differently than the Department of Justice surely suggests that, if you get into the business of wielding a criminal stick, you will emphasize the use of that stick. The other point is about criminal penalties and substantive antitrust doctrine. I have written about this several times, so I will talk about it only briefly. One can see in US law something that I call an “equilibrating tendency”. Once the consequence of finding a violation is severe, the courts are less likely to find a violation. If it is per se illegal to get into a vertical price agreement, the courts will make it very hard to prove an agreement. If treble damages result from predatory pricing, the courts will make it very hard to prove predatory pricing. So there are equilibrating tendencies in antitrust. This has made a big difference in the development of US law, and I am confident that it would make a big difference in the development of European law with the introduction of criminal sanctions. Just some quick comments on the European situation: First, I have no doubt that, to the extent that criminal penalties are spreading in Europe, it will become correspondingly more difficult to tolerate relatively vague standards, and it will become more difficult to prove violations. To the extent that there is private enforcement, the legal standards will become clearer. You cannot have a system whereby private persons can enforce substantive law unless that law is relatively clear, because you can no longer entrust the wise people in Brussels to take the right decision. Second, one really ought to think separately about each of those questions: individual versus company, criminal versus civil, prison versus something else. Third, the more severe penalties will tend to dominate. In the US, the FTC—where I was general counsel—must refer the case to the Department of Justice if criminal proceedings are appropriate. All the FTC can do is get


12/10/07

13:09

Page 231

Panel Discussion

231

out of the way, because criminal enforcement trumps civil enforcement. We shall see what will happen in Europe as criminal powers are increasingly exercised by the national competition authorities. Someone asked me during our lunch today: “Will the European Commission become a sort of FTC?” I do not think he meant that as a compliment. . . . In terms of equilibrating tendencies, one has to wonder what will happen if the national competition authorities end up being able to wield quite draconian punishments, while the Commission has only milder remedies available to it. How will this affect the substantive law? The milder penalties at EU level would invite relative ease in finding violations, while the more serious national penalties would work in the opposite direction—with the result being hard to predict. 䉴 WOUTER WILS—I will start with the usual disclaimer, namely that I will express only my personal views, and not necessarily those of the Commission. When we talk about cartels in Europe, it is useful to distinguish between two types of cartels: first, the global, or transatlantic cartels—which account roughly for about half of the cases that the Commission deals with. Invariably, these cartels are detected in the US by the Department of Justice, so that, in fact, what the Commission does is a kind of follow-on. As Margaret Bloom has observed on several occasions, that is in fact very good evidence of the usefulness of criminal enforcement. For these cases, in reality it does not matter so much whether we have the right penalties in Europe. We benefit anyway from the positive spillover from US enforcement. What matters is that we should not create negative spillover by having incompatible systems. The move from the 1996 to the 2000 Leniency Notice can be understood, at least in part, in that light. A second type of cartels comprises the local, or national ones—including, if I am not mistaken, bid-rigging cases, although the Commission has never dealt with such cases, for the understandable reason that public procurement takes place at the national level. It is interesting to distinguish these cases because, in quite a few Member States, for bid-rigging there are criminal penalties—there are such penalties in Germany; and since last year, Hungary has followed this original pattern; and in France the old criminal system tends to be applied nowadays mostly in the field of public procurement. If we look at the European cartels, how has the situation changed in the last five years, and how will it evolve in the future? Five years ago, I wrote a paper on exactly the same subject for a previous edition of this Workshop,5 in which I was pleading for introducing penalties for individuals. At that time, in Europe the only penalty for cartels consisted of fines imposed on

5 Wils W. P. J. (2003): “Does the Effective Enforcement of Articles 81 and 82 EC Require Not Only Fines on Undertakings, But Also Individual Penalties, and in Particular Imprisonment?”, in Ehlermann C.-D. and Atanasiu I., eds., European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, pp. 411 et seq.


232

12/10/07

13:09

Page 232


undertakings by the Commission. I am not going to repeat the arguments that I invoked then. The idea was radical then, but in the meantime the situation has changed a lot. I think that, today, there is a real perceived threat of criminal enforcement in Ireland, the UK and Estonia. Here, again, there is a positive spillover to the rest of the EU and the Commission from the point of view of deterrence. Again, there is also the risk of negative spillover in one direction or the other, but I think that the negative spillover is fairly well managed at the moment. The important points are Article 12 in Regulation 1/2003 (the rules on exchange of information) and the provisions in the Commission Notice on the functioning of the ECN,6 respectively the provisions on information exchange in the area of leniency programmes. What about the future? Maybe it is because I have less immediate retirement plans, but I would not speculate about what is going to happen in my administrative lifetime. I would just go back five or ten years ago and ask: how many people would have thought then that the UK would introduce criminal offences for cartels? Ten years ago, the UK was one of the pro-cartel, weak competition enforcement systems, so I expected them least of all to make the jump they have made from the bottom to the front of the ranking. Who would have expected five or ten years ago, in general, that we would have a European arrest warrant? There are lot of changes in the European antitrust enforcement system, and more generally, in the EU criminal law enforcement sphere. So, I would leave this discussion open. For instance, in the Netherlands a couple of years ago the head of the national competition authority proposed to introduce sanctions for individuals. Criminal antitrust enforcement would be a big move in this country, because it was only recently that it moved to an administrative enforcement system. Notwithstanding this, he opened the debate on sanctions for individuals, and proposed to introduce director disqualification, as in the UK. The Dutch government did not follow this proposal, and proposed instead the introduction of fines for individuals. At present, a heated debate is taking place in the Dutch parliament—where I have been invited twice as an expert. I was very impressed by the level of knowledge of the parliamentarians involved in the debate. At the end of March 2006, two leading Dutch parliamentarians announced that they will not follow the government’s proposal, and supported director disqualification (actually, one of them even supported imprisonment, but since the other one did not follow, they compromised on supporting director disqualification). In Sweden, another country where reforms are currently discussed, the debate goes in the direction of not introducing criminal sanctions. The debate was structured from the beginning by excluding the possibility of individual leniency, and I think this really goes to the crux of the matter, because I do not think that many of the people who believe that criminal sanctions would 6

OJ C 101 [2004].


12/10/07

13:09

Page 233

Panel Discussion

233

be much more effective than what we have now have seriously considered that the current system is inefficient. There is a lot of evidence in this sense. Olivier Guersent said this morning that most of the leniency cases are end-of-life cartels, but the big problem is related to leniency, because, on the one hand, the possibility of having individual leniency is one of the good things about having prosecution of individuals, as it multiplies the possibilities of exploiting leniency, and on the other hand, if you do not have it, the whole system becomes counterproductive. At the moment leniency was introduced, the justification for it was not obvious either, for several reasons. This is not to doubt the effectiveness of the leniency system—a subject on which the economists write papers. It is rather a discussion on the moral and legal acceptability of leniency, and on concerns about the moral element, which are partially reflections of the old cartel culture. Denis Waelbroeck commented about some European business people who still have their “code of ethics”: they think they cannot break away from the cartel, which is just another way of saying that they have a cartel culture. The moral concern, when you agree that something is a crime, is whether it is acceptable not to punish the person who committed it. On the point of legal treatment, I think the obvious answer is that, since companies get an equal chance to apply for leniency, this is equal treatment ex ante. On the moral element, it may depend on how you structure it. This morning, when I heard criticism of the omnibus question in the US, my reflection was that, if there is only this question, then it is actually a good thing: it shows that you want from the person a real element of conversion, and when there is real conversion, and not just playing around the system, it may be morally acceptable to have leniency. Will we have criminalization at the European level? Olivier Guersent said he would not expect it in his lifetime. I would not be so sure about it. It is certain that the political and institutional obstacles are substantial, but I do not think it is a legal problem—already in my paper of five years ago I argued it is legally possible to do it, and Prof. Zuleeg, a former judge at the ECJ, was of the same opinion.7 I think that, after the ECJ judgment of last September on criminal sanctions in the environmental sphere,8 this position is not so farfetched anymore, especially if one transposes the reasoning in that judgment to the competition area. So, it is legally possible, although it would of course demand substantial institutional change: we would have to separate prosecution and judgment. The question that Stephen Calkins raised in his paper is whether we really need that, or whether we could just have criminal enforcement at the level of 7 Zuleeg M. (2003): “Criminal Sanctions to Be Imposed on Individuals as Enforcement Instruments in European Competition Law”, in Ehlermann C.-D. and Atanasiu I., eds., European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, pp. 453 et seq. 8 Case C-176/03 Council v. Commission [2005] ECR I-07879.


234

12/10/07

13:09

Page 234


the Member States, and have the Commission lead separate procedures at EC level. I do not think that this is necessarily a bad approach. I just leave the question open as to the most efficient division of work, and observe, however, that in the US there is a federal cartel enforcer, so I suppose that in the end this is the line we want to follow—in the end, we want the Commission to continue to be in the business. Last remark: there is an interesting issue of debate at the moment, namely the impact of possibly enhanced private enforcement on the debate on criminalization. Many of those commenting on the Green Paper have pointed out—and the Green Paper includes a section on this—that development of private enforcement may have a negative effect on the leniency programme. One possibility would be that the pressure to criminalize increases. 䉴 NADIA CALVIÑO—Just a short clarification: in the new Spanish competition law we will double the fines on individuals, and we are discussing a leniency programme not only for companies, but also for individuals.

JOHN FINGLETON—I identify three main themes of debate from the presentations we have heard so far. One is about having an EU system whereby the Commission does administrative enforcement, and an increasing number of Member States do criminal enforcement. This links with Stephen Calkins’ point about the natural tendency of institutions. If people see successful criminal enforcement in some Member States, is that going to create a natural snowball effect, whereby other countries are pressured into having criminal enforcement? And where does that lead us? The second theme is fines for companies versus sanctions for individuals, like imprisonment, etc. The third theme is about consistency in enforcement at the international level. The core arguments in favour of and against criminalization are probably well known, so we should try to focus the debate on the “next generation” issues—or where we should go to from here.

䉴

RAFAEL ALLENDESALAZAR—A comment on the level of antitrust fines in Spain, in relation to Nadia Calviño’s presentation: I think that, besides proportionality and deterrence, there is a third principle to be taken into account, and that is the signalling effect. Fines should clearly signal which practices are more harmful for competition than others. This is somehow linked to proportionality. One would tend to believe that hard core cartels would be fined much more heavily than exclusionary practices, but nevertheless, the highest fine applied so far by the Commission refers to an exclusionary practice.9 I do not know if that is the proper signal to send to companies. 䉴

9

Commission Decision of 24 May 2004, OJ L 23 [2007].


12/10/07

13:09

Page 235

Panel Discussion

235

JOE HARRINGTON—Two comments. The first relates to Massimo Motta’s empirical study, which I find to be a very worthwhile exercise. To me, the interesting issue is to try and disentangle what part of the study’s effect is due to the loss of profits, and what part is due to antitrust penalties. To see the balance between the effect of penalties and the profit that firms earn from cartel activity is important from the standpoint of the deterrence objective. I wonder whether the fact that the ECJ judgments do not seem to have an impact means that the loss of profit is a more significant component—and where does this leave us in relation to the size of penalties? The second comment is more general: we have seen a number of tables and figures showing that leniency applications, fines and prison sentences are up. That is very good, but it has also been noted that the primary objective is to deter cartel formation, and it is important for all parties involved to try to get a sense about what is actually happening with the rate of collusion. This is a major challenge: collusion is one of those domains, like tax evasion and extortion, where it is difficult to observe the true rate of infringement. I am saying the obvious, but I think it is important that it be stated, because if at some point the success of antitrust enforcement will presumably be measured by lack of leniency applications, fines falling, etc., is this an indication that the number of cartels is falling? It is a challenge for the economists to try to find ways of measuring the real rate of cartel activity.

䉴

MARGARET BLOOM—Wouter Wils pointed out the importance of having a leniency programme in a criminal enforcement system. I think that the issue is who actually does the prosecution. In the US, the Department of Justice is both the prosecutor and the leniency grantor. In Israel it is the same. In Ireland there is a separate prosecutor from the antitrust agency, and in the UK it is slightly more complicated. In England and Wales, the OFT can prosecute, but the main prosecutor is the Serious Fraud Office. The OFT gives leniency in consultation with the Serious Fraud Office. In Scotland, Lord Advocate prosecutes; and there is no formal leniency scheme. In all those countries where the prosecutor is different from the competition authority, it is more difficult to persuade the prosecutor that there should be a leniency regime.

䉴

JOHN FINGLETON—Indeed, originally there were great difficulties in Ireland due to this split, but we discovered that the prosecutors eventually got very keen on pursuing cartel cases.

䉴

䉴 CHRISTOPHER HARDING—A couple of observations about the process of criminalization, and what we know about it. In this context we have to ask ourselves to what extent we can transplant a model from North America to the European legal culture. Perhaps we could distinguish between two


236

12/10/07

13:09

Page 236


separate possible impacts of criminalization in the European context. One would be the impact on leniency programmes and how they operate—and we already heard about the possible effect of criminal proceedings on the way in which leniency programmes are working. But that may be different from the impact of criminalization on long-term deterrence—which relates to Joe Harrington’s earlier point. It may be that criminalization has quite a considerable impact on leniency, on cartels being busted and on clear-up rate, but that does not mean that in the longer term the eradication of cartels will be more certain. I think that is a much more difficult question to answer. 䉴 PAUL GROUT—Just a couple of observations. Massive research linked to what we heard this morning about corporate governance points towards the need for most public companies to be quoted, to force a situation where directors see themselves obliged to offer transparency on the market, because of the impact of investigations upon their companies. We should also note the potential contingency liabilities—I think this is a very good discipline. We have seen a number of cases in the UK where, for one reason or another, public companies have felt obliged to make an announcement, and make an appropriate provision of their accounts. Second, more research is needed on the impact of individual cartels on the economy and on consumers. In the absence of that kind of information it is very difficult to justify some of the levels of fines that have been proposed or actually imposed. We would also benefit enormously from more specific data that rely on specific cases, meaning more data that run from the front of the case throughout its life and thereafter. Third, a personal thought about criminalization: in the UK and elsewhere there have been very serious cases, for instance cases of railway crashes, where directors were prosecuted criminally, and even if people have been killed, many were injured, and hundreds of thousands were potentially injured by the negligence of the individuals concerned and their companies, the level of the fines imposed was relatively modest: around £ 10 million. So a lot more research is needed in order to justify the high level of antitrust fines, be they civil or criminal, to demonstrate the damage to the economy and to consumers.

䉴 DENIS WAELBROECK—There seems to be quite a choir around this table in favour of criminalization, and I wanted to add two personal comments on this aspect. In my mind, to criminalize the sanctions would exacerbate a number of problems that I perceive with respect to Regulation 1/2003. National authorities in all 25 Member States apply Articles 81 and 82 EC, they meet in a network to share information and cooperate, and thus a case born in the UK may end up in Luxembourg. I have a problem with applying the same provisions where there are different sanctions, especially when you do not


12/10/07

13:09

Page 237

Panel Discussion

237

know if you are sanctioned for behaviour in one Member State or in the whole of the Community. This raises interesting questions. The second point is one of fairness. I think that cartels are terribly boring cases for lawyers; the only interesting aspect is of a psychological nature, namely: why do people still continue to engage in cartels even though they know they should not do it? The answer I seem to get is that there are clear instructions in companies that, from one year to another, the profit should be increased by 50%. If you do not produce such results, you lose your job. But there is no way to do it without infringing the law. So it is a bit of risk-taking: either you do not increase the profit, and then you lose your job, that is a certainty; or you do increase profits through illicit means, and maybe you will be caught, or maybe not. I just wonder where is the real liability for this: is it with the person who engaged in the cartel, or is it with the management of the company, which probably knowingly imposed a policy that cannot be achieved without breaching the law? JULIAN JOSHUA—A “next generation” issue: how does the idea of the Commission emerging as the next FTC and the Member States’ authorities as the next 27 “Departments of Justice” fit into international cooperation? The Member States, especially those that have criminalization, will be able to provide judicial assistance to the US on the transatlantic level, especially on extradition. Obviously there will be far more attractive propositions for the DoJ for cooperation at the transatlantic level than at present, where assistance relationships are maintained with the Commission. The tension between treating cartels as a sort of economic phenomenon and treating them as bad behaviour results in visceral opposition to exchange of information even inside the network whenever you get anywhere near criminal enforcement. It seems that Ireland, which is allowed to exchange information for purposes of criminal enforcement of Article 81 can only exchange information with itself. So there could be implications for global cooperation, whereby national systems that have introduced criminalization can engage in far greater cooperation than those where enforcement is administrative. On corporate governance: there again appears to be a big gap between those Member States which see cartels as an economic phenomenon and those which treat them as serious delinquency and put the responsible parties in jail. There is an effect in the corporate governance area: if you were the CEO or chairman of a company in the US that was prosecuted for cartel activity, you would not remain at the helm of that company, and you certainly would not be seen a hero of industry and be allowed to sit on the supervisory boards of other companies. So there appears to be a cultural difference that is very visceral, and although we can talk about criminalization, we first have to come to terms with that huge gulf of incomprehension between the two approaches.

䉴

WILLIAM KOVACIC—I wanted to draw your attention to an article that Stephen Calkins has cited in footnote 6 of his written contribution for this

䉴


238

12/10/07

13:09

Page 238


Workshop,10 which I think is one of the five best papers written in the past half century. It is very important because it identifies very clearly that changes in any one element of a competition policy system ripple through its other features. If one element is severely out of balance, other developments will take place to correct it, often in ways that introduce other distortions into the process. Stephen Calkins introduces this caution in talking about criminalization of antitrust enforcement, but he has applied it also to treble damages introduced into private enforcement. I think this is a point that deserves attention, as jurisdictions can employ the mix of private enforcement and damage collection, and class actions, and discovery techniques, that they want to use. It is unmistakeable that the US doctrine on abuse of dominance acquired a decidedly non-interventionist cast since the 1970s because the formative cases that came before the appellate courts and the Supreme Court were private treble damage cases, in a number of which the Court found an infringement by a very narrow margin. The concern was that, if you turn on the liability switch, you get automatic trebling, and in amounts which in many instances could dwarf the net worth of the companies in question. So, faced with the question of applying a liability standard and setting in motion what they thought would be an undue torrent of damage collections, the courts decided to fix this: if the statute freezes the damage mechanism in place, then the courts would change the liability standard. That is exactly what I think they did. Indeed, when you look at the modern abuse of dominance cases, such as Trinko,11 I do not think it is possible to interpret them without noticing the over-hanging risk of massive treble damages that the courts took into account. Justice Stephen Breyer once even mentioned in a decision that to find liability would be to convert a business tort into a treble damage claim. This opinion probably would not have been written in the same way if the issue were converting a business tort into an actual damage claim. This is the status quo; the difference comes from the trebling, and I think you can find the same analysis in cases defining horizontal agreements. So, the principle that Stephen Calkins mentions is enormously informative in contemplating any adjustment in the system. These are interdependent variables, and in the US system the courts have fixed the distortions induced by turning on the volume on one element, but I argue they have done it in ways that in some instances are excessive. This denies non-treble damage litigants the opportunity, for example, to prove an agreement in several nondamage cases, or by circumstantial evidence in instances that I think would be useful. I am also thinking about the question that Joe Harrington raised, also touched on in Massimo Motta’s presentation, namely: do industries reform 10 Calkins S. (1986): “Summary Judgment, Motions to Dismiss, and Other Examples of Equilibrating Tendencies In the Antitrust System”, 74 Georgia Law Journal 1065. 11 Verizon v. Trinko, 540 U.S. 398 (2004).


12/10/07

13:09

Page 239

Panel Discussion

239

themselves over time, or are they inherently recidivist? I get at least a pattern of recovery within individual sectors, so it is almost like the model of the Terminator, who keeps coming at you after having been just slowed down for a bit. This is again an area where historical assessment of past experiences is enormously valuable, especially ways of systematically looking at sectors over time. With all the qualifications and problems associated with comparative study, I strongly endorse a reading of the US experience with criminal enforcement from the 1960s to the present. In many instances we faced questions of political legitimacy, and part of the success of the DoJ strategy is holding the prosecution in these cases. In at least three cases in the 1960s, the DoJ indicted companies, not individuals, for abuse of dominance. It got out of that business, I think, for fear that continuing to do that would make application of the statute murky in cases where enforcement was of high value, and second, so as to establish political legitimacy for criminal enforcement. After the statute was made more severe, in the 1960s and 1980s, the DoJ brought lots of cases of public procurement. To some extent, commentators ridiculed this strategy of bringing the same case again and again against public procurement authorities. But it happened to be against what is characterized by the DoJ as theft against the Treasury, and that was a good way to get juries to convict individuals involved. What you see is actually a very careful process of sequencing the introduction of the more powerful criminal sanctions over time. This history is enormously informative, not only about how to technically increase reliance on this tool, but also how to do it in a way that appears, in the eyes not only of the competition community but of society as a whole, to be a legitimate exercise of power. JOCHEN BURRICHTER—Just a few remarks on what may happen in the Member States. For my country, Germany, I can tell you that I do not expect, in my lifetime, criminalization of antitrust enforcement. Why not? Because the experience with the criminalization of bid-rigging has not been very convincing. First of all, the courts have not made use of it. People have not been sent to jail, although there is a theoretical threat. But even the threat that people responsible for bid-rigging could go to jail has not prevented cases of the kind. So, in parliamentary discussions about the possibility of criminalizing antitrust enforcement, this argument was raised in opposition. On the other hand, let me remind you that, in Germany already since 1958, there has been personal responsibility for companies violating competition law. The fines were recently increased to 1 million Euros. In the past, in most cases the companies paid the fines for their managers, but this is no longer possible, because German corporate law has been changed. If the Board of Directors decides in favour of one manager to do that, 1% of the shareholders can sue the Board of Directors and force them to claim damages even from the manager in question. I think that is much more efficient than a system of criminalization that is not practiced by the courts. Third, on the sanction of

䉴


240

12/10/07

13:09

Page 240


professional disqualification, which is actually an issue in a case where companies have allegedly violated the law, the insurance regulatory body inquires who was responsible for that, and there is a legal possibility to disqualify the people who have been involved in cartels, at least for a certain period of time, by preventing them from taking over similar jobs in such companies. I think this is a sanction we should consider more in the fight against cartels. 䉴 DONNCADH WOODS—Just a few words on what I would see—and I am not attacking in any way the criminal system—as some advantages of the administrative system. It is less resource-intensive and you have a lower standard of proof. The rights of defence are not going to be as great a barrier as in a criminal case. Do not underestimate the effects of financial fines in the administrative system. If you have a correct corporate structure within a company, the impact on its management of the type of financial losses incurred through administrative fines can be really significant. But we need to ensure that the losses are reflected in a proper corporate governance structure, and that they really have an impact on management. From my experience, I think those fines can have very serious consequences for the directors concerned. Another issue to bear in mind is private enforcement. Wouter Wils presented the opposite scenario, of private enforcement taking off and then having to bring in criminal sanctions to balance for leniency. Bear in mind that in Europe we do not have any system of discovery. If you were to switch to a criminal system, there would certainly be fewer cases than what you have at the moment. As we know, in the area of cartels, most private enforcement cases are follow-on cases. In the absence of a discovery system, that would mean less private enforcement.

䉴 PATRICK REY—Jochen Burrichter has already covered a couple of points that I wanted to raise on the role of sanctions. Again, do not forget the stick: we can play with the carrot as much as we want, but if there is no stick it will not work. I was going to question the role of fines because of the manipulation problem. Jail is a serious deterrent, but it is so difficult to achieve at the EU level. I was surprised to learn this morning that disqualification is actually easier to achieve or implement than criminal sanctions in the UK, and I wanted to ask whether there are similar experiences in other countries, because this appears to be a very promising avenue. The other point refers to the other side of the coin, which is leniency for the informants. I think this is the most effective way of supporting the claim I was making this morning, namely that leniency can be good in order to ensure that the evidence will be there, and it is really leniency for individuals that will achieve this. In other words, maybe some managers will bury evidence under their cherry tree, but at least it will be there. In that respect, I take the moral point against offering rewards to individuals who were involved in the


12/10/07

13:09

Page 241

Panel Discussion

241

infringement of the law, but at least we could offer rewards to individuals who just disclose the cartel without having been involved in its organization. I heard about a case in which the evidence was brought forward by the wife of a company manager. NICHOLAS FORWOOD—The most impressive thing for me in Massimo Motta’s presentation was the lack of response from the stock markets to decisions of my court. . . . Then I started to reflect on this. One possible explanation would be that the court is “unknown”—I read this comment in the Herald Tribune the other day. The other possibility is that perhaps markets are cleverer than we think, and are able to identify the cases where the initial allegations (in the form of dawn raids, etc.) did not match with what was actually happening in the marketplace. I wanted to ask Massimo Motta if the correlation that he found between price movements and dawn raids and convictions tends to be weaker in cases where the conviction is subsequently reversed.

䉴

PAUL GROUT—My point follows exactly upon this. I am happily impressed that Massimo Motta found any impact at all on share prices. I do quite a lot of financial economics research, and I know that a variation of 1.2% is actually nothing. We are looking here at companies where the shareholders suddenly find that what was actually generating lots of money for their company has disappeared. The analogue of this in the other direction is when shareholders suddenly discover that somebody out there thinks their company is worth a hell of a lot more than they think it is, and their management should be sacked and replaced. The typical bid premium for a takeover is 25 to 30% of the share price, almost all of which goes to the target, because of the efficiency of the market. So basically an inefficient company that is suddenly raided raises its share price by 25 to 30%, of which almost all goes to the target. So, 1.2% is really just nothing. To give it another perspective, I read an article in the Journal of Financial Economics about a month ago where the authors looked at the impact of the Labour Party playing around with the regulated industries in the UK in the late 1990s. It comes forward quite clearly that the impact of the interventions in 1997 and 1998 had an enormous effect. The interesting question is, why? Three possibilities come to mind. One is that cartels are actually part of the business, and this is true of a lot of the cases before the UK Monopolies and Mergers Commission. But in a lot of the cases that I studied, that is not the case at all. A second possibility is that the demand curve is such that when you raise the price by 25% you actually lose an enormous part of your sales. If that is so, the social welfare loss is actually quite small, because the demand curve is very flat. But I do not actually believe this is plausible either, because the people who set up the cartel are not going to take a risk for something that is not profitable. So

䉴


242

12/10/07

13:09

Page 242


actually we are left with Olivier Guersent’s point, namely that if the markets have got this right, what we are actually picking up is the end of the cartel’s life. If John Connor is right in that these people are raising the price by 25%, that is total profit, and unless it does enormous damage to your sales, the value is huge. It cannot possibly be just 1.2% or 2% of the value. So clearly these cartels must not have had much life left in them, or the market just does not understand. But my experience is that the market knows a lot, actually . . . THOMAS BARNETT—First, I want to underscore the point that Stephen Calkins and William Kovacic so eloquently made about equilibration. When I was talking about the importance of a very narrow definition of cartels that will enable you to impose more severe sanctions, it is essentially the converse of the equilibration argument. Second, on the cultural point, I do think that is one of our greatest challenges: to develop a culture of respect for competition law enforcement. Aa an example: the ADM tapes.12 These tapes allowed judges and business people to see how these people sat in a hotel room, fully aware of what they were doing, and joked about the FBI and FTC. This really educated people about how this was a criminal offence. Those tapes are still in use: recently, the Korean Fair Trade Commission arranged to play the ADM tapes in Seoul for a group of businessmen, and the reports that I received were that at the end of the session the room was completely sombre, and you can only imagine what was going through their minds. I cannot resist responding to Denis Waelbroeck about individual liability. Competition creates incredible pressures, there is no doubt about that, but I personally believe it is a serious mistake to let those pressures be an excuse for engaging in anything other than lawful competition on the merits. With respect to cartel activity, we can debate things on the civil side, but criminal conduct needs to be treated as such, notwithstanding the pressures that may come from management. 䉴

䉴 DONALD KLAWITER—Following up on Thomas Barnett’s comments, one of the points that I heard around this table is that we should worry about the effects of enforcement on the corporation—Mario Siragusa’s point about bankruptcy—as opposed to the consequences for the individuals who were actors in the case. We talked about how, in many jurisdictions, criminal sanctions are, at least for now, not likely. I think there is agreement around the table that personal responsibility does matter here, that this is the critical element that we ought to be dealing with. The first thing mentioned in this 12 For an account of the Lysine cartel case, see Hammond S. (2005): “Caught in the Act: Inside an International Cartel”, speech delivered at the OECD Competition Committee on 18 October 2005, available at http://www.usdoj.gov/atr/public/speeches/212266.htm, with direct hyperlink to the transcripts of the tapes in question.


12/10/07

13:09

Page 243

Panel Discussion

243

sense is criminal sanctions; the second is what I call “the corporate governance system”—basically disqualifying people from the job if they are involved in this kind of conduct; and the third would be individual fines (with a note of warning: individual fines that are paid by the individuals, and in fact it is a criminal penalty in the United States for the corporation to pay the fine in whatever form). If you take those three elements as manifestations of personal responsibility, I think that is where deterrence is. The US has two of them in effect and one of them applies de facto. The criminal fine cannot be paid by the corporation, the person can go to jail. The corporate governance decision is not made automatically, but I think that in virtually every case that I have seen in the last five years, the CEO involved lost this position by the end of the investigation, and so did his colleagues in the company who were involved. So, three elements is best, two is better, one is OK. That is how enforcers are going to address the deterrence issue in a very meaningful way. JAMES VENIT—Just a short comment related to those made Donald Klawiter and Thomas Barnett. If you look at the issue from the perspective of the firms involved, society is doing them a huge disservice by not sending a clear message. One of the things that surfaced in our discussion today is that, in Europe at least, we do not have a clear message about cartels. In some ways it is probably a way of doing business—people achieve a certain positive notoriety in certain countries for having been involved in cartels, while we hesitate about criminalization, and so on. From a compliance perspective, this is a disastrous message to be sent to companies. They need to be given very clear guidelines. Either we get out of the business of opposing cartels, or we bring the full force of criminal law, corporate governance and personal responsibility to bear on individuals who are in a position to react to that. When a company realizes it will lose leniency if it continues the illegal behaviour, this is a very powerful weapon: all of a sudden you have no difficulty in bringing about a compliance attitude within that company. However long or short-lived that may be, that is a separate issue. My point is that, ultimately what we should try to induce is a compliance culture.

䉴

WOUTER WILS—Three short points. First, a reaction on the predictions for Germany. I would think that the German experience with bid-rigging confirms what I said, in the sense that there is no leniency for bid-rigging, and that explains the problems in building cases. Of course, judges must be willing to convict, which is part of the general need for a cultural change, but I would think that is not impossible. If a change can happen in the UK, it can also happen in Germany . . . On the point that the administrative enforcement would be less resourceintensive, I think that is an unsubstantiated empirical point. I think that it would be interesting for someone to carry out a study on that, to look at the

䉴


244

12/10/07

13:09

Page 244


number of officials that work on cartel cases at the Commission and in the national competition authorities, to look at the number of cartels detected (not counting follow-on cases from the DoJ), and then to look at the same two figures for the DoJ, and compare them. My guess is we may well find that the criminal system is less resource-intensive. How can that be? Because they have stronger investigative powers, and you can only get that in a criminal enforcement system. Last point, on director disqualification, which I described in my paper as a defensible second-best: the effect depends on the type of person involved, and basically whether the person is close to retirement or not. But we should not have illusions of avoiding the bigger administrative burden or the higher level of proof or the institutional change. I cannot imagine, in any of the legal systems that I know, that you could impose director disqualification through administrative procedures. In the end, the question becomes one of whether you put the label of “criminal” on it or not. As I described in my paper, there is no uniform definition of that, so it is difficult to cast criminal sanctions as being in contrast to director disqualification. JOHN FINGLETON—The job of the chairman is to bring about an agreement, but I do not think we can have that in this room. But I do think there is agreement on the fact that the cost of cartels in Europe is going up, and statutory limits for fines are going up. The discussion about individual sanctions, both administrative and criminal, is more intense—the debate is more advanced in the Netherlands and Spain—and cartels are being taken more seriously by national governments, in the sense that more resources are being put into enforcement, both at the national and EU level, so there has been an increase in enforcement. All of that goes to the question of how is this relating to compliance: is it raising the costs of getting involved in cartels? There is no doubt that cartels were a cheaper corporate strategy in the past than they are today. Having said this, the issue of criminalization does not go away, because the fundamental argument for criminalization is that none of the rest makes cartel activity costly enough. There is a related point which has not come up yet, namely that, to the extent that we proceed with raising fines we begin to push the administrative system towards criminal law and towards human rights issues. Can we actually avoid being seen as a criminal system dressed up in an administrative background? If the European Court of Human Rights starts to trump the European Court of Justice on those issues, we will see a very dramatic effect. My final comment is about culture. If someone told me a few years ago that the biggest obstacle to successful cartel enforcement in Europe was cultural, I would have accepted it, with great fear about what the Irish courts would do. But the Irish competition authority recently initiated the first criminal trial before a jury, and it turned out to be very interesting when the case got

䉴


12/10/07

13:09

Page 245

Panel Discussion

245

to court, because the jury found the evidence to be incredibly compelling. You need a witness, a person to stand up and say that something happened in this way, in order to satisfy the burden of proof. The jury needs to see how the cartel operated. Perhaps the most compelling piece of evidence in that case was from a person who was asked to participate in the cartel and refused to. His testimony stood out because he showed coercion, and the jury was sympathetic to him. The judge said, in summing up the evidence to the jury, that they did not have to worry about the sanction, but only had to decide whether there was cartel activity or not, and the case went through much easier than we had thought. William Kovacic spoke before about the link between responsibility and liability. On the other hand, what I would not have anticipated was the high cost of cartel cases for the competition law enforcer. I do not disagree with Wouter Wils, but I think that one should consider the cost for an agency to bring a criminal case at a time when they do not have a record of civil prosecution. The Irish competition authority has taken ten years to bring this case, and its image was hugely damaged over that period, because it was seen as lacking success. My one piece of advice to countries introducing criminal law is to make sure everything else is working first. In Europe I do not think we can avoid anymore the debate about moving towards criminalization.


246

12/10/07

13:09

Page 246


Session IV: Private Enforcement JOHN FINGLETON—The subject of this session is private enforcement. It has been touched upon in the previous sessions—for example, William Kovacic made a significant observation about treble damages, and there were other issues that came up this morning, such as Ian Forrester’s point about justice and the fact that public enforcement cannot cover everything, the issue of under-enforcement versus over-enforcement, the remark about privateering regarding treble damages in the US, etc. We already have several ingredients for the discussion.

䉴

䉴 DONNCADH WOODS—Let me start with the usual caveat that the views I express here are my personal views. They do not necessarily reflect those of the Commission. The PowerPoint presentation that I prepared basically summarizes the six key issues where the Commission thinks we should intervene in order to facilitate actions for damages in Europe. At this point I want to address two of them: one is the purpose of this exercise, and the second is what my choice would be. To stimulate the debate, I would ask everybody to think: if they were to choose two actions to facilitate damage actions, what would those be? Coming back to the issue of purpose, the short-term objective of the Green Paper on Damage Actions13 is to open the debate—and so, I am very happy to be here today, as I am sure we will have a very interesting debate on damage actions. The long-term objective is to increase the effectiveness of actions for damages in Europe. Why do we want to do this? We want to ensure compensation for parties who have been injured by anticompetitive acts, and we also want to ensure, or enhance, deterrence. We want to do this while avoiding unmeritorious litigation. In order to stimulate the debate, and in the interest of brevity, I would simply ask the question of what are the steps you think should be taken in order to achieve this objective. My two choices are rather simple. The first would be to give the injured parties the toolbox to bring damage actions. By “toolbox” I mean basically that competition cases are fact-intensive and complex. If you do not have the facts, you cannot bring a case. So, you have to have access to evidence. This to me is one of the key issues. As explained in the Green Paper, this can be achieved by allowing some kind of disclosure of the relevant documents, or alternatively, you could shift the burden of proof in case of information asymmetry. For me, the key issue is access to evidence. The second choice is addressing this risk-reward balance, i.e., to incentivize litigation while at the same time minimizing unmeritorious claims. Here, we

13 Available at http://ec.europa.eu/comm/competition/antitrust/actionsdamages/index.html. By the time this volume is published, a corresponding White Paper should also be available at the same website.


12/10/07

13:09

Page 247

Panel Discussion

247

could look first at the issue of costs. Lowering this risk is a delicate exercise, and we cannot discuss contingency fees. But we do have an option according to which unsuccessful claimants would be liable only for legal costs on the defence side if their action failed. That would be a certain way of inciting meritorious claims. Another way would be to increase the rewards available to a successful claimant. In that respect, the Green Paper puts forward the option of double damages for horizontal cartels. I am aware that treble damages is a rather sensitive issue, but one could argue that double damages is not punitive. One argument in favour of this would be that, if you protect the litigant on the cost side, there is a greater risk that somebody could come forward with a weaker case. If you only had double damages and no protection on the cost side, you would still face a situation whereby, if you had a weak case and lost, you had to pay costs to the other side. There are many other issues out there that need to be addressed—for example, I am not raising the question of whether these two issues should be tackled at the EU or national level, and I am not going into the issue of legal instruments either. The fundamental issue that we have to decide first in Europe is what we want to do. 䉴 DENIS WAELBROECK—The purpose of my presentation is to discuss whether something needs to be done, and if so, what. I will start with two comments. The first is that it is easy to criticize everything and say no to everything, and I have seen that happening a lot. Personally, I find—as Ian Forrester was saying before—that there must be a system of compensation if damage is caused by cartelists. I do not think that this would be against our legal system, to have the victim compensated instead of allowing the cartelists to keep the profit in their pockets. So, I do not think compensation is a problem. I do have a problem with punitive damages, as we got a very clear message from 22 of the 25 rapporteurs in the Member States in our study14—exceptions being the common law Member States, which have some sort of exemplary damage system that is seldom applied. But in all the rest of the Member States it was perceived as contrary to public policy to have a form of double damages. In fact this goes back very far in history, to the days of the old Romans, where the public enforcement system was not very well developed, and litigants were awarded multiple damages in order to provide a sort of private sanction. As the system matured, all these multiple damages were abandoned. The compensation objective, as I said, is justified, and then we need to ask ourselves whether the current system is working properly. You know the results of our study on this. We may have overlooked one or two cases, but still, we found about 12 or 13 successful damage actions in 50 years of 14 Ashurst (2004): “Study on the conditions for claims for damages in case of infringement of EC competition rules”, available at: http://ec.europa.eu/comm/competition/antitrust/others/ actions_for_damages/comparative_report_clean_en.pdf.


248

12/10/07

13:09

Page 248


Community law for 25 Member States. There have been some cases more recently, and there is a trend towards more, and there are many settlements that probably we do not know of, but still, it simply does not work. If you look at the Crehan case,15 you understand why this is so: this gentleman started an action in 1993, and the case is still not over. He went up from the High Court to the Court of Appeals, then to the ECJ, back to the High Court, and then again to the Court of Appeal (where he got £ 130,000 of damages), and the case is now before the House of Lords. So there are not many incentives to start this type of action. If compensation is legitimate, what can we do to improve the system? I am not in favour of punitive damages, or for US-style class actions, but there are other proposals that may be more reasonable. I classify them in three categories: one step would be to clarify the rules; the second would be to assist parties in establishing the facts; and the third step would be to carefully rebalance part of the system. On clarifying the rules, in our study we found that, basically, in every European jurisdiction, the law is unclear. In Italy you do not know whether you should start the action before a normal court or the Court of Appeal, and this is a huge disincentive. You may start an action and then, after days of argument, you may be thrown out of the court. I think there is scope for clarification of the rules that would eliminate at least one element of uncertainty, and that goes not only for the procedural rules, but there is also a fundamental question that goes to the substance, and I recently read an interesting study in the US showing how private enforcement increases with per se rules, so I think we need clearer rules of substance, though not necessarily per se ones. One point I wanted to make here, taking the UK example, is that it may be worthwhile to consider the establishment of courts specialized in competition matters, because, given the complexity of the law, there is a high risk of bad decisions and false positives. Take Article 82 EC and the mess its application is in. If you go before a Belgian court, you may easily be condemned for something that is not an abuse, and I think there is a big potential for a chilling effect. If we want to have adequate application of the law, it is probably time to consider the establishment of specialized courts. On helping the parties to establish the facts, here there is a big hurdle: in most Member States there is no pre-trial discovery except in the UK, Ireland and Cyprus. In most cases you need to identify specific documents, which I think is a major hurdle. So without going into the excesses of the US model, there is room to change the rules. I agree with Donncadh Woods on this. The most difficult point is on the re-balancing of the system. Here there are many issues, and I will briefly comment on two of them. The passing-on defence: in my experience, you have weeks of arguments in court between economists, and nobody knows the truth, so I tend to agree with AGs 15

Case C-453/99 Courage v. Crehan [2001] ECR I-6297.


12/10/07

13:09

Page 249

Panel Discussion

249

Mancini and Tesauro that it is a difficult defence. Maybe the only person that ends up being rich is the cartelist. The other point is contingency fees, because I do not think you will ever have private enforcement unless there is some sort of success element, and I have a problem with bar rules excluding contingency fees. To sum up, there is room to improve the system. The Commission’s initiative was welcomed because it created a momentum. There is also room for improvement in the area of injunctive relief. 䉴 JON LAWRENCE—I am going to take a completely different approach, and be extremely narrow. I will approach the subject from the perspective of a commercial litigator, which is what I actually am, and I’ll ask myself the question: how can I advise commercial clients litigating in the European context? I will give you an example, which I will develop in some length, to illustrate the problem that I see. I see the problem as more than what we do, but how on Earth do we do it? The problem I will talk about does not have the sex appeal of double damages or the pass-through defence against purchaser claims, but it is nonetheless one of the most important questions that my clients ask about every time they are trying to bring an antitrust claim in a cross-border situation. It is about jurisdiction, and in particular, consolidation of claims. I will focus on the ability in Europe for groups of companies to bring claims in a single forum against groups of defendants. By that I mean all the cartelists and all the members of their groups being sued by all the claimants, that is, all the companies who bought the cartel product. Why does it matter? As Mario Siragusa knows from our experience in an ongoing case, if you tell a client that he has to bring proceedings through his local subsidiaries in potentially 25 different jurisdictions, he immediately thinks of the cost, and that is actually what he will ask first: how long will it take, how much will it cost, and where is it going to take place? All of that hinges on jurisdiction and consolidation. If you have to sue in a number of jurisdictions through local subsidiaries, you need multiple experts, multiple overseas lawyers, enormous linguistic, translation as well as evidential issues, different limitation problems, coordination difficulties, and of course the risk of diverse outcomes. The litigation risk multiplies exponentially once you cannot advise a client that he can bring all of his claims for all losses within the group within one country. What does the Green Paper do about this problem? Paragraph 2.8 of the Green Paper says nothing other than that Regulation 44/01, the Brussels Regulation, governs the situation. It puts forward no option for change in relation to consolidation or joined actions or jurisdiction questions. In the Staff Working Paper, at paragraphs 237 to 240, there is no mention of any proposed changes, it simply relies on the Shevill judgment to say that “clients have multiple options”. That may be true for consumers, who can go to their


250

12/10/07

13:09

Page 250


home court if they can demonstrate that the harmful event happened in that jurisdiction, but for a group of companies, and for commercial litigators working for a group of companies, the position is really very different. I shall illustrate this with two brief references to cases. The first one is the Provimi case.16 This was a Vitamins Cartel litigation case in the UK courts. I am going to simplify this massively, because there were a number of issues in the case, but essentially there was a group of claimants—let us say, for example, an English and a German company—who wanted to sue a cartelist, a Swiss company, through its subsidiaries around Europe. The German company had bought products from a German seller, and the English company had bought from an English seller. What they tried to do was to get all the buyers into court in England and sue all of the defendant subsidiaries in the UK courts. They essentially met two problems: they had to show that the English subsidiary defendant was a tortfeasor who committed a breach of Article 81 EC, and then that the German company had a claim against the English subsidiary. And through that mechanism they then tried to argue that the Brussels Regulation enabled them to bring in the German subsidiary defendant. How did the English court deal with that? This was at interlocutory stage, and the English court decided that, even though it could not be shown or it was not pleaded that the English subsidiary selling the products knew about the cartel, it had committed a breach of Article 81 EC by reason of selling the goods (implementation of the cartel). So they had a cause of action. The German company, on the claimants’ side, then had to run a convoluted argument according to which that company would have been able to buy from the English company if the latter had not been in the cartel, because it would have been able to arbitrage products across the border, and they produced some evidence for that. At a very preliminary stage in the English proceedings, the court decided that there was a novel point of EC law there, as to whether the English subsidiary had even committed a breach of Article 81 EC, and so permission was given to appeal and it would have gone through the appellate courts if the case had not been settled. Secondly, at the preliminary stage, the English court accepted the convoluted arbitrage argument, so it seemed to give some licence to bringing in multiple claims against multiple defendants. But the position remains very unclear, and I suspect that an appellate court may well have reached a different decision. The second example is the Primus case,17 which has not yet reached ECJ judgment stage. This is a patent infringement enforcement case, where a question has been referred from the Dutch Hoge Raad to the ECJ, asking: if a group of defendant companies carry out an infringement and they do so 16

Provimi Ltd v Aventis Animal Nutrition SA [2003] 2 All ER (Comm) 683. See Advocate General (AG) Léger’s Opinion of 8 December 2005, in Case C-539/03 Roche Nederland et al v. Frederick Primus et a.l 17


12/10/07

13:09

Page 251

Panel Discussion

251

under a common group policy, can you sue them together in the Dutch courts, and can all the relevant claimants come to The Netherlands and sue in a single court? Last December, AG Léger took the view that this is not possible. He concluded that the cases were not sufficiently connected, in Brussels Regulation terms, in order for there be a risk of irreconcilable judgments. The reason why I labour these points is that it is still considerably difficult to get any clarity around the jurisdiction rules that would permit whole groups of companies to bring an action against whole groups of defendants, and in a cartel situation that is something that would really facilitate commercial claims in court. I believe this issue has been relatively scantily dealt with in the Green Paper, and this should be revisited. This leads me to my broader point: why are these issues a frustration, and do we deal with them, in legislative terms, and in terms of culture if necessary? At this stage, I have to express some pessimism about where the Green Paper takes us. In Europe we do not have the same culture of litigation that I, as a litigator, see with great envy on the other side of the Atlantic. To try to reregulate the jurisdiction rules would be impossible under the Brussels Regulation. Dealing with these issues in general will be a massive problem in the absence of political will on behalf of the Member States and the Commission, and I am interested to hear what Mario Siragusa and other colleagues have to say on the prospects of the Green Paper coming to fruition at all. 䉴 MARIO SIRAGUSA—Jon Lawrence has given us some very good examples of the practical obstacles to be met, and Denis Waelbroeck has told us interesting things about what should be done. Reflecting on the effects of the Green Paper, I ask myself: are difficulties in procedural and substantive law in the Member States the reason why there is little private enforcement, or is it more a cultural thing, like Jon Lawrence suggested? I really believe that culture plays an important role in what we are talking about. I tried to go through a number of the possible changes suggested, but I think that in most of the cases our legal systems do offer a solution. Maybe the solution is not perfect, but it is not as bad as we think it is. I gave a few examples in my written contribution. I think, for instance: is it really necessary to change this very old requirement in our tort law about fault? In cartel cases, fault is very difficult to deny. Discovery is a real problem. Cartel cases are generally going to be follow-on cases, and there are already enough decisions of the Commission and national authorities on the leniency programme, so if a company jumps on this there will be an increase in litigation. There are measures in our civil law codes allowing the judges to forward all the evidence they have. It is true that the Italian civil code, for example, requires you to indicate documents, but I think that judges, if pressed, and if the culture of antitrust is more diffused, will interpret the law in a more positive way. So I think the idea is that we have to keep on trying, we have to bring cases, and only if we do so will we


252

12/10/07

13:09

Page 252


have an influence on how the judges apply those rules. If we sit doing nothing and wait for changes in the civil procedure, we might wait for another 20 years. And Italy is a country where litigation is very inexpensive. We have a lot of injunctive relief actions under Article 82 EC, this is very common, and those are almost always preliminary proceedings—in a few weeks they are decided, and then you start the action for damages, and this takes longer, but it is already a big improvement. Also, for instance, this idea that we have to make the decisions of the national authorities and the Commission compulsory on judges, this is completely against the constitutional order of several Member States. Do we really need to change that? I think everybody knows that a civil judge is going to be very impressed by a one-hundred-page decision, either of the Commission or the national authority, and is going to accept most of the facts of that decision even if we do not have in our system a principle that obliges them to do so. Sometimes, in big cartel cases, the Commission or the national authorities establish the facts, but do not really closely examine the facts of a specific defendant, and a judge can do that, so I think it is good that a judge can somehow re-establish the position of a single defendant in a cartel case. For me, the real reason why private litigation has not been so forthcoming is the expense of the process—in many Member States, litigation is very expensive and therefore quite an elite matter—and the second reason is the culture of the consumer. After all, even the Commission has put the consumer at the centre of antitrust enforcement only very recently. Until only a few years ago it was the competitive structure, etc. Therefore, the more we bring the consumer to the centre of the system the more our judges will be able to recognize when damage has been done and the more they will be inclined to compensate the injured. CHRISTOPHER BELLAMY—Jon Lawrence has told us about the difficulties of even follow-on actions bringing cross-border claims. . . . I would tend to share Mario Siragusa’s optimism in a cautious way on these topics, and I would ask the question: what are some of the conditions that need to be met before private enforcement can become effective in Europe? It seems to me, first of all, that we will need considerable dialogue between five different worlds. First, there is the specialist competition world—but not all competition lawyers are litigators, and in fact, the gentle world of merger filings is a very long way away from the normal world of litigation. Then in the second world there are litigation lawyers, not many of whom are also competition lawyers. So I think we need to reach a stage where a competition lawyer is also a litigator, and a litigator is also, at least to some extent, a competition lawyer. The litigation world is very special. It is a world of civil procedure, of tactics, of evidence, sometimes of trial by ambush . . . and sometimes even in a medieval way, of trial by ordeal. But it is certainly a very particular world. We are getting better at having these two worlds talk to each other, and looking around this room I see at least two or three of us who have at some stage seen the inside

䉴


12/10/07

13:09

Page 253

Panel Discussion

253

of a prison cell—of course, in the context of advising clients. Then thirdly there is the economics world, and it is encouraging in a forum like this to have such great input from our economic colleagues, but there is still some way to go: the economic world does not always feel entirely comfortable in legal contexts, and vice versa, from the lawyers’ point of view, competition cases are fact-intensive, so you need to see the facts and what pattern emerges, rather than starting with theory and seeing if the facts fit. Fourthly there is the judicial world, but if, as was said earlier, an economist might be able to scare a thief, he certainly scares most judges. Again, this is a matter of a learning curve, but we still have a situation in which many judges would try to avoid or circumvent the economic issues and retreat into procedural matters, or others where they feel they are on firmer ground. The fifth world, without which nothing would happen, is the political world, and within this world there are the ministries of justice, finance and commerce, all of whom need to be thoroughly involved in a discussion at Member State level. One of the difficulties, I think, is that if competition authorities are short of resources, the ministries of justice are even shorter, and that is going to be a major problem. There is an additional fact we need to acknowledge: we have not yet got a full competition culture, and to some extent we have, I think, an anti-litigation culture. We have a sort of “honour among thieves” culture in the business community. Those are important difficulties to overcome. One of the arguments that is not often formulated but I think is an important political argument is that, very often, the victim in the world of cartels is in fact the taxpayer, because it is very often the municipalities, the hospitals, or government agencies of one sort or another, to be ripped off by cartels. It is sometimes said that we do not want too much litigation, what we really want is a system in which cases can be settled and mediated, etc. But we have to bear in mind that settlements happen, or mediation occurs, only in cases where there is an effective litigation system in place; otherwise, there is no incentive to settle. In more detail, what do we need to make a system work, and can this be done without being perceived as being too Americanized? Again, I am not talking about follow-on actions. Can you go down a litigation route that does not have the full panoply of disadvantages of litigation? The first thing to emphasize is that litigation in itself is extremely risky, no matter how good the system. We have all lost winning cases and won cases on our seventeenth and worst point. For that reason, one cannot entirely rely on private litigation if we want a successful competition enforcement system. Not to mention the commercial pressures that potential claimants come under, in terms of possible commercial reprisals in the marketplace, of accusations of disloyalty to the club, in terms of not being invited for the next trade association meeting, etc. Of course, the firms in the most difficult position of all are the SMEs, which are very often too small to be able to be really helped by the public enforcement authorities but which cannot afford the expenses of litigation.


254

12/10/07

13:09

Page 254


There are four areas we need to think about, and one way or another they have already been mentioned, but the ones that I identify are, first of all, costs. Legal costs cannot really be dealt with by the system, and you are going to need some entrepreneurial lawyers prepared to take on the costs of these cases. As far as the costs on the other side are concerned, the rule that costs should follow the event is not necessarily suitable for these cases—in my personal opinion it is not suitable for any sort of case, but this is the rule, certainly in common law jurisdictions, and this rule will need some sort of clarification to re-balance the field in this area, and there are other fields, like employment law, matrimonial law, etc. Second, evidence: nobody wants to go down the road of extensive pre-trial depositions and discovery. On the other hand, we do need some provisions for disclosure, backed by very heavy criminal sanctions for destroying evidence. Question: what sort of threshold should be reached before a court can order disclosure? No one wants fishing expeditions; claimants have to produce something, but I do not know what that is. I do not think you can legislate in advance for it, but a claimant probably should not have to show very much before getting a disclosure order. Of course there is a great richness out there nowadays—cartels are easier to prove now than they have ever been, because of e-mails, deleted e-mails on computer systems, and mobile phone records. So the evidence is all there, it is just a question of getting it. The “Enron effect” is of course very positive here, because now there are doubts about destroying documents, and that should be reinforced. Thirdly, you need qualified judges: we have all seen cases that take too long because the judge is not strong enough to stop them. We need judges who are able to weed out vexatious cases. Patent judges are not a bad analogy: these are also complicated cases; we would not give them to non-specialists. In the whole damages area, very quickly: in the common law, a deliberate act which you calculate is worth doing because you will never have to pay damages can give rise to something called exemplary damages, which basically evens up the loss. So you cannot profit from your own wrong, even if there is a slight element of over-compensation. We are going to need a rule of thumb on how to calculate damages, with an extra charge of 10% or something like that, so you do not have to prove every single invoice. Presumptions will deal with the passing-on case, but that needs to be addressed, and Jon Lawrence is quite right, you have to establish a procedure whereby all the claimants can go into one court at the same time. This procedural issue can be solved, and if you solve that, you also solve the passing-on defence, because you avoid the risk of paying more than once. And finally, do not be very ambitious because it will backfire: keep these cases down to hardcore cartels and obvious cases of monopoly; otherwise, judges will be resistant to finding in favour of the claimants if they think the case is weak, and the law of unintended consequences will strike again.


12/10/07

13:09

Page 255

Panel Discussion

255

JOHN FINGLETON—We are dangerously close to consensus, so I encourage those of you who disagree with what has been said to say so. I will also remind you that we are discussing cartels here, and there are issues of private enforcement that run across all areas of antitrust, but I think we should try to focus on the specific issues related to cartels. Nobody has mentioned the role of public authorities in private enforcement, either in terms of handing over evidence or of amicus briefs. Building on Mario Siragusa’s optimism and Donncadh Woods’ plea for focus on the choice of two issues, we could take advantage of the diversity of this audience to bring out perspectives from different Member States.

䉴

CLAUS-DIETER EHLERMANN—I have a question of process, not procedure. The Green Paper is extremely rich with respect to solutions; one can only hope that questions help you formulate a reply. Even having gotten replies to your different issues, then there is the big question mark on how to implement this. What will DG COMP do with this material? Now I leave that question for a moment and go back to 2001, when in this room we discussed private enforcement, with an “appendix” on criminalization. Since 2001, criminalization has taken extraordinary leaps forward, totally unexpectedly. The Commission had advocated private enforcement for decades, and it had published a Notice to help judges deal better with EC competition cases. The process has its own dynamism, and we heard today that criminalization might make bad money chase the good money, so that private enforcement might take over and probably even marginalize DG COMP and the national authorities. . . . What we heard today is really an example of competition among legal systems. If you apply that model to private enforcement—this is not my observation, I think Jon Lawrence raised it in his written contribution—if one of the Member States implemented a sort of model regime for private enforcement, what would be the effect? Lawyers would of course be more than happy, because it is an industry, and the city of London, in that respect, it is true that it is expensive, but as Ian Forrester has told me several times, it is also a luxury—you pay for the quality of the service. My process question is, why is arbitration in London? Why are certain jurisdictions, and even on the civil side, very open to extraterritoriality? Extraterritoriality is an element of attracting business. My point is, instead of relying too much on DG COMP to produce wonderful draft regulations and recommendations, why not pick up the results of the discussion on the Green Paper and implement them at national level, to see what comes out of it.

䉴

CALVIN GOLDMAN—In Canada we had a very interesting experience with private actions on the doorstep of the US, and there is a template there that can provide a bridge to Europe. Our business community was greatly reluctant to adopt US-style treble damages, and they have not made their way into

䉴


256

12/10/07

13:09

Page 256


Canada, but after a series of constitutional challenges fighting off civil actions to supplement public enforcement, the Supreme Court of Canada finally broke through and settled the law in 1986. But it took one ingredient that we have not talked about here yet, in order to open the door to civil actions, and this addresses the cost issue that several participants talked about here. Also in Canada you did not see private actions until there was a cottage industry of lawyers (as you know we have all kinds of lawyers, including specialists in ambulance chasing) who became antitrust specialists because of one phenomenon: the introduction in the 1990s of class actions with contingency fees. It is that mixture, without treble damages, the ability to take thousands of plaintiffs and put them into one action by lawyers who can obtain a percentage of that proceeding. In ten years we have developed a cottage industry that now mirrors the US. Every time there is a US investigation, the Canadian case follows without treble damages. Second, the role of the government. The Canadian government and courts have been largely successful in staying about as neutral as they can, letting the attorneys of plaintiffs see the documents of the defendants and then letting their counsel fight over it. The Canadian Bureau of Competition stays out of it, and that is working. What is throwing the monkey wrench in the Canadian system is not our courts but the US courts. In the Choline Chloride case, we are still in the middle of it, but because of the decision of the Washington DC federal district court, all kinds of production orders were made of plaintiffs in Canada. The Canadian Attorney General had intervened in the US proceedings, but was largely ignored, and the US court ordered production of something that is protected under Canadian law. So you really have to watch out for the US courts, because it is really different from anything we have seen before, and of course, European disclosure rules and amnesty programmes also affect Canadian enforcement. Third, the protection on the defence side comes from having a valid and very stringent certification process. All these actions are commenced, and until their claims get through they have to get certification as to whether there is prima facie merit before the claim is allowed, and that is the check. THOMAS BARNETT—My main comment on the extraterritorial reach of the US courts in the cartel context is to talk about the Empagran case,18 where, as you know, we weighed in quite aggressively on the US Supreme Court to limit the territorial reach of the decision, in large part because of what we saw as a deleterious effect on leniency programmes in other countries. We recognized that, and also pointed to the sovereign right of different countries to determine what remedies should be available to their citizens for transactions that take place in their countries. There was an issue left open, regarding world markets, where the two were arguably inextricably linked, where you might

䉴

18

E. Hoffmann-La Roche Ltd. v. Empagran S.A., 542 US 155 (2004).


12/10/07

13:09

Page 257

Panel Discussion

257

have wanted claimants coming in to US courts. We have weighed in before courts numerous times with amicus briefs, trying to assert the same position. So far, at least, the tendency of the lower US courts is to accept the logic of the Supreme Court and not allow these types of claims. I cannot resist briefly commenting on public law and private actions. We do lawyers a tremendous service, in alerting them to the existence of a cartel, but beyond that we make every effort to basically stay out. We do not cooperate with the plaintiffs’ lawyers or feed them information or leads; they are left to themselves. I have to respond to Sir Christopher Bellamy’s perhaps off-hand comment that suggested that the evidence for cartels is out there, all we have to do is find it. I wish it were that simple. In our experience, cartelists go through an evolution, and they are very sophisticated. I love the story of a cartel in an Eastern European country where officials of a recently established competition authority went to a company they heard was leading a cartel activity, asking: “We heard that you may be fixing prices; is that true?” And the individual opened a file, pulled out a piece of paper with a written agreement confirming the existence of a cartel, signed by the members, with a stamp from the post office which he thought would make this all lawful. That is on one end of it. But in time we find that there is an increased sophistication: things are not written down, most of it is done through phone calls, and phone records can be helpful in proving that some people talked to each other, but what they said is another matter. They put things in code: for example, instead of talking about prices, they talk about golf outings. One of my favourite examples that was recently discussed at the ICN was a computer code that, when you looked at it, seemed to be complete gibberish, but once you sorted it out it was a code to generate a pricing and output report for the company, and it was embedded in it to send those reports to the other companies. So, although I agree that there are certain advantages to e-mail tracking, cartels are increasingly sophisticated and we need to keep up with them. 䉴 CALVIN GOLDMAN—Just to specify, the Canadian plaintiffs bar is tremendously appreciative to the DoJ for the position they took in Empagran.

ASSIMAKIS KOMNINOS—I would start with a personal note, I have done academic research in the area of private enforcement for several years and now I am practicing in Brussels. I have understood now that it is not due to rules, but to human beings, that we do not have a developed system of private enforcement in Europe. I am sure there are problematic rules, like the rule of the English Court of Appeal with regard to the pari delicto situation, or the fact that the Italian Corte di Cassazione some years ago was considering that Article 81 EC does not cover consumers in its protective scope. But I believe the problem is with humans. Starting from the judges, I think that if, in the future, we do not have Community courts of general jurisdiction, and if we

䉴


258

12/10/07

13:09

Page 258


do not follow the US model of federal circuit courts which try these cases from the starting point up to their end, I do not think we will ever reach the US level of private enforcement. With regard to lawyers, again there is a problem, because the competition legal milieu is mainly Brussels-based, so I can understand a certain reserve of the Brussels legal profession with regard to private enforcement, because professionally this might mean less business. Looking at the Member States, I really see no possibility of having a developed bar to deal with competition law claims locally. I would strike a note of caution here: for a number of years since 1999 we have taken for granted that decentralization and effective private enforcement are the same thing, or that they are compatible. But I think that what we actually need in private enforcement, rather, is centralization. We need centralized courts, ideally at the Community level, or certainly at the national level, and only when we attain a certain degree of development at the central level, only then can we start thinking about spillover and more decentralization. This is what happened with public enforcement: we first created the Brussels-based system and then we started thinking about decentralizing. JULIAN JOSHUA—In the Green Paper the emphasis has been on facilitating remedies for small consumers, but in the real world, I think that if you want to deter cartels the availability of civil remedies will not be a deterrent unless there is a possibility for a large group of purchasers to gather together, who will not be deterred by cost rules. I am thinking of small companies who bring in a civil claim in London: there is an application to strike out, the defendant with deep pockets gets a very expensive barrister, who talks for five days and then suggests dropping the case on the basis that each side pays its own costs. I agree we need some sort of centralization rather than decentralization. Under the modernization project we federalized EC law without creating a system of federal courts. We need to beef up the Brussels Regulation and make it even easier for companies to bring group actions in a single forum. There will obviously be legal obstacles. Although Provimi took the law forward by a series of steps, it still involves quite a bit of bootstrapping. As Jon Lawrence said, the case may not even survive appeal. I wonder if the principle could be extended to allow you to sue, for instance, not only other companies in the group that supplied you but also the other members of the conspiracy, a system like in the US where you end up with joint and several liability for the co-conspirators. Other obstacles: business culture—people do not like suing the people with whom they have ongoing business relationships, whereas in the US this does not seem to matter because everybody sues everybody anyway. I think that encouraging civil actions on a pan-European basis in one form by major groups is a road to go down if you want to have civil litigation that really has a deterrent effect on cartels.

䉴


12/10/07

13:09

Page 259

Panel Discussion

259

䉴 WILLIAM KOVACIC—When you create private rights that go beyond follow-on suits, I think that, in a certain way, you do transform the key role of the competition authorities. Without private rights, the competition authorities have a monopoly-like gate-keeping function in setting the agenda of what gets litigated, and in many ways influencing how doctrine develops. Private rights divest the agencies of that gate-keeping role. So if we think of what the agencies find over time, even in the area of cartel prosecutions or things that characterize these cartel matters, they do end up investing considerably more effort in an amicus curiae role. They have to establish a cost-watching function by which they observe cases that have an influence on doctrine, or they have to openly invite private parties to play that cost-watching function. So, to answer John’s earlier question, private enforcement unmistakably changes the role of competition agencies: it means that, instead of directly shaping the agenda, they have to spend a lot of resources on influencing the doctrine. Another thing to keep in mind is that, in hard core horizontal restraint matters, plaintiffs may not always be seeking damages; they may be seeking injunctions, and this is an area where the system does want to preserve the right incentives for people to bring cases. What kind of case do I have in mind? To go back to the theme pursued two years ago in this forum, the treatment of the liberal professions, the topic of our discussion there was how often in the US a private litigant, after the public agency has decided to stand aside, drops a formative case challenging state restrictions on matters that were serving state-based exemptions from competition policy. Going back to the formative Goldfarb case in the 1970s,19 and several others, it was a private litigant who, against conventional wisdom, and indeed against the advice of the competition agencies, decided to bring the case. That case had a crucial role in starting to topple many of the restrictions on liberal professions that had persisted in the US for decades. So, it may not only be the private litigant as a damage seeker, but the private litigant seeking a permanent injunction, and in that case there must be a general system of calculating the fees of the lawyer representing that person.

JOHN FINGLETON—I remember that, when this point was raised two years ago, Giuliano Amato commented that if we rely on individual heroes we are lost.

䉴

JOCHEN BURRICHTER—I concur with Claus-Dieter Ehlermann in that we need a certain process encouraging private enforcement, but I see this process already in place. Just as an example, the German legislator, a month before the publication of the Green Paper, changed the law to encourage private enforcement by, for example, introducing a binding effect not only for

䉴

19

Goldfarb v. Virginia State Bar, 421 U.S. 773 (1975).


260

12/10/07

13:09

Page 260


German decisions but also for Commission decisions and decisions of other Member States. I will not comment on whether this is wise, I just want to give you an example that there is something going on, and I also see that the culture of such complaints is changing. Another example: before this new German law entered into force, a complaint against the German cement industry was submitted to a local court in Düsseldorf concerning the amount of 120 million Euros, a case where the decision of the Bundeskartellamt was not yet binding. This brings me back to what Mario Siragusa has said, that in the existing system we practically have all the means in order to initiate private enforcement against hard core cartels. The argument that we need some sort of class action is especially unconvincing to me, because again, going back to that case, a German lawyer founded a company which he calls Cartel Damage Claims Company, which has bought the claims from smaller companies which bought cement from the cartelists and submitted this claim. The company has invested money into the investigation and paid for experts. This is a kind of class action, if you want. PETER ROTH—I just wanted to go back for a moment to Donncadh Woods’ question about what would be the two suggestions. His first one was the toolbox of giving access to evidence. In England we have this tool: the UK and Ireland probably have the most extensive disclosure provisions in Europe. Regarding conditional fees—we do not have Canadian or US percentage contingency fees, but we do have conditional fees in the sense of “no win, no fee”, and 100% uplift on the fee if successful. But still we do not have a lot of cartel damage claims. These claims are hugely expensive, and I think one has to recognize that extensive disclosure obligations may be necessary, but they certainly make the cases more expensive, and with disclosure of electronic evidence still more so. As for conditional fees, one of the very few cases that we have had involved a shipping liner conference. The lawyers acted on a conditional fee, the trial lasted ten weeks, the claim completely failed, and I do not think that any of the lawyers would be very enthusiastic about bringing on such a conditional fee a case again. I would support very strongly what ClausDieter Ehlermann indicated because it seems to me one cannot think of this problem within the confines only of competition law, and still less, cartel damages. These are problems of access to justice which arise across quite a spectrum of cases. I recently had the very unusual (for me) experience of having to assist the victims of an allegedly defective drug seeking to bring a claim against the pharmaceutical manufacturer. There many of the same issues and difficulties arose—the need for disclosure of a lot of evidence, dependence on expert evidence, great problems of causation, and as a result of all of that, huge problems of cost in trying to bring a claim at all. I think each legal system has to find a solution to these problems within its legal culture, and within its particular procedures. It may be very difficult to impose that from the Community level. If there is one particularity of cartel claims, it is the involvement of pub-

䉴


12/10/07

13:09

Page 261

Panel Discussion

261

lic authorities, which you do not have in other areas of damages claims, and that does introduce the potential for the authority, where it has investigated and found a cartel, to require a remedy for affected victims as part of its intervention. Phillip Collins mentioned what happened in the schools case in England—it could only happen because the schools were willing to offer that kind of arrangement. But might there be scope for introducing a power on the part of competition authorities to impose a requirement of some kind of compensation? We have all kinds of other imaginative remedies that are imposed in merger cases; why should it simply be a financial penalty on the wrongdoers in a cartel case—could there not also be the introduction of a mandatory compensation scheme, with mediation to settle individual claims? PHILIP COLLINS—I am also quite optimistic, and I think that a Europeanwide solution is simply not realistic because of too many conflicting interests. In particular, ministers of justice will find it very difficult to agree that there should be some special carved-out regime for competition cases. I agree here with Peter Roth: there are other areas where there are similar issues of access to justice, and I think that if we go down that route we would actually put ourselves in a much more difficult position than we need to. Gradual pressure and evolution in individual cases, and indeed the development of competition between different jurisdictions, will actually produce more evolved civil models that bring about effective private enforcement. I also agree that if we go down that route it will change the role of public authorities, whereas if we do take the second route, private enforcement will not inhibit public enforcement through things like the discovery of documents at an inappropriate stage. But there are certainly things that public authorities can do, like putting things into their decisions that will help private enforcement. It is very important, in trying to encourage private enforcement, to emphasize the harm that cartels do to the economy. It is not just a question of trying to equalize compensation between people who have suffered loss—these are cases where serious economic harm has been done to the country.

䉴

JON LAWRENCE—I still think we are a very long way from where we want to be in the Community. Consider the decision issued by a Nantes court on 11 May 2006, again in a vitamins cartel litigation where the Commission Decision had specified in its decision that there was harm to consumers. The courts denied the claim of one of the direct purchasers on the basis that the harm had been caused to consumers, and therefore implicitly accepted that there was a pass-through defence, and also put the burden of proof on the purchaser to disprove pass-through. This is why I fully endorse the model jurisdiction approach in my written contribution, but with a word of caution on that as well. When I discussed this idea with my US colleagues, they said, “Well, make sure it is a model system and not the kind of system that we

䉴


262

12/10/07

13:09

Page 262


ended up with in the US, where claimants rush to Alabama because of the particular features of that state.” This is why in my paper I advocated balance, rather than tipping the balance the other way. DONNCADH WOODS—I have not heard new arguments, but the flavour is different, certainly on the idea of developing a national model. I think I was very succinct in my presentation, and there have been a number of issues raised that I would like to go through. First of all, I commend the German example. Remember Mario Siragusa’s point about follow-on actions. This is not just about follow-on actions; we want to extend the scope of application and increase the deterrence effect. Therefore, we have to facilitate stand-alone actions. Jon Lawrence’s point about Regulation 44/2001 is very interesting. We did write a bit more, but the Paper was so long that we had to cut it down. The situation is much more complicated than you are probably aware of. Jon Lawrence described a situation of consolidation in terms of jurisdictions. You may succeed in doing that, but then you have the issue of the applicable law. If the effects are in a few Member States, then you would have to apply the laws of those Member States, and for issues such as calculating damages you have the “salami” effect, so it can get even more complicated. In the Green Paper we are not proposing US-style options: there are no juries in Europe for this kind of trials; the cost rules in the US are different; disclosure of evidence is much wider in the US; the issue of joint and several liability is quite important; and then you have opt-out class actions and contingency fees. I wanted to remind you that in Ian Forrester’s example, the man did not succeed. Mario Siragusa was presenting a more optimistic view, according to which the situation is improving in many Member States . . . but to my recollection there are many Italian consumers in an ongoing insurance cartel damages case, and they are waiting for an Article 234 EC reference. On the issue of access to justice, yes, you could argue that this should be addressed at the European level, and it should not be specific to antitrust—but bear in mind that there is an IP Enforcement Directive20 which provides for access to evidence, very similar to our proposal in the Green Paper. And then, “Nothing special about antitrust cases”: well, there is something special that make antitrust cases different from most tort cases. In tort cases it is usual to go to court and say, “I was injured by the act of the defendant”. That is not sufficient in an antitrust case. The plaintiff can go to court and argue that the defendant threatened to run him out of business. He may even have been run out of business and suffered a loss, but that is not sufficient: the plaintiff must prove an antitrust infringement, i.e., that there has been restriction of competition—reduction of output, increase in price, loss of innovation—and it is

䉴

20 Directive 2004/48/EC of the European Parliament and of the Council of 29 April 2004 on the enforcement of intellectual property rights, OJ L 157 [2004].


12/10/07

13:09

Page 263

Panel Discussion

263

a public interest test, of general harm to consumer welfare. One of the major points raised by Prof. Ehlermann was the level of action we were talking about. Paragraph 51 of the Commission Staf Working Paper annexed to the Green Paper is quite clear in the sense that the discussion is about all possible levels of intervention, from Community to national ones. So it is clearly very appropriate to discuss this idea of a model jurisdiction. But we are talking here about a common standard of infringement of Article 81 EC; it is common across the EU, so the right to an action for damages is a common EU right. In the Crehan judgment,21 the ECJ said the remedies to those rights are to be applied through national law. So, imagine that we have an EU cartel case that has effects in Italy, Germany and the UK. If we apply the laws of those jurisdictions, we know that in Italy there is a passing-on defence, and in Germany it is very difficult to use the passing-on defence—the concept of the calculation of damages is different. So there is a common EU standard as to what is an infringement but there are differences in remedies at the national level. Another issue that could be rather tricky: if we went down the route of the US system on standing, there you have standing for direct purchasers at the federal level and not for indirect purchasers, but the latter do have standing at the state level. So my point is, we have a common EU standard for the infringement, but be very careful about what happens at the national level. Divergences at national level can create obstacles to private enforcement. 䉴 JOHN FINGLETON—One point to take home is what Sir Christopher Bellamy said about not ignoring the political dimension to this, and the importance of advocacy. Ultimately it is clear what improvements are needed at Member State level: most would generally point to greater consistency across jurisdictions and dealing with the model point, but I think the importance of advocacy at national level is critical. On Mario Siragusa’s point that in the past we had a system of competition that was more about protecting competitors, whereas we can genuinely argue to national governments that this type of private litigation might have been harmful to business and productivity in the past, there is a much greater prospect that now it would be beneficial to business. Adding Claus-Dieter Ehlermann’s point, we can also argue to national governments that if they get this right, first the country gets a competitive advantage and an international services marketplace. If we take these two points together, it makes a very powerful basis for advocacy at the national level. Finally, when people talk about legal costs they should be careful. I have learned in time to try to get more competition into the legal system in Ireland, and will probably continue to do so in the UK, but at the end of the day the cost that we are talking about in Masterfoods 22 or Crehan 23 are

21 22 23

See supra note no. 15. Case 244/98 Masterfoods v. H.B. Icecream [2000] ECR I-11369. See supra note no. 15.


264

12/10/07

13:09

Page 264


the costs of cases that go on for 12 years or so, and this is not just about legal cost, but also about how the legal system treats a case. Of course, everything gets picked up as a cost at the end of the day, but part of it is about getting the legal system right, and to some extent we can wait for the courts to sort that out, but it goes back to the point about advocacy with respect to primary and secondary legislation.

(O) Ehlermann 07 Wils

12/10/07

13:10

Page 265

SESSION B


1 PANEL III: PUBLIC ENFORCEMENT (ADMINISTRATIVE AND CRIMINAL)

WORKING PAPERS


12/10/07

13:10

Page 266


12/10/07

13:10

Page 267

I Wouter P.J. Wils* Is Criminalization of EU Competition Law the Answer?

Abstract: This paper addresses five questions concerning criminalization of EU antitrust enforcement: First, what do we mean by ‘criminalization’, or ‘criminal’ enforcement (as opposed to public enforcement of a ‘civil’ or ‘administrative’ nature)? Second, is there a tendency in the EU Member States to criminalize antitrust enforcement (in comparison with US antitrust enforcement and with antitrust enforcement at the level of the EU institutions)? Third, is criminal antitrust enforcement, and more specifically imprisonment, desirable (in general, irrespective of whether it takes place at the level of the Member States or of the EU institutions, or whether it is harmonized at EU level)? Fourth, is it problematic that antitrust enforcement is criminalized at the level of individual EU Member States without parallel criminalization at the level of the EU institutions or without EU harmonization? Fifth, would it be legally possible to criminalize antitrust enforcement at the level of the EU institutions, or to have EU harmonization of criminal antitrust enforcement in the Member States?

* Member of the Legal Service of the European Commission, Visiting Professor at King’s College, London. The author gratefully acknowledges comments and contributions received from Margaret Bloom, Leo Flynn, Céline Gauer, Nicholas Khan, Clemens Ladenburger, Kirti Mehta, Walter Mölls, Andreas Reindl and Geert Wils. All views expressed in this paper are strictly personal, and should not be construed as reflecting the opinion of the European Commission, its Legal Service or any of the above-mentioned persons. The author can be reached at [email protected]. This contribution is based on a paper with the same title first presented at the Amsterdam Center for Law and Economics (ACLE) Conference on ‘Remedies and Sanctions in Competition Policy: Economic and Legal Implications of the Tendency to Criminalize Antitrust Enforcement in the EU Member States’ (Amsterdam, 17–18 February 2005), forthcoming in Cseres K., Schinkel M. P. and Vogelaar F., eds., Remedies and Sanctions in Competition Policy: Economic and Legal Implications of the Tendency to Criminalize Antitrust Enforcement In the EU Member States (Edward Elgar, 2006), and also published in 26 World Competition 117 (June 2005) and (with French translation) in 4 Revue Lamy de la Concurrence 139 (August/October 2005).


268

12/10/07

13:10

Page 268


A. What do we mean by ‘criminalization’? 1. No definition of ‘criminal’ in EC or EU law Article 23(5) of Regulation No 1/2003,1 the main regulation governing the enforcement of Articles 81 and 82 EC, provides that decisions by which the European Commission imposes fines on undertakings for violations of Articles 81 and 82 EC, or for obstruction of investigations into possible violations, “shall not be of a criminal law nature”.2 The last sentence of recital 8 of the same Regulation says that “this Regulation does not apply to national laws which impose criminal sanctions on natural persons except to the extent that such sanctions are the means whereby competition rules applicable to undertakings are enforced”.3 The Treaty on European Union has a Title VI concerning ‘Provisions on police and judicial cooperation in criminal matters’. Article 30(a) TEU provides that common action in the field of police cooperation shall include operational cooperation in relation to the prevention, detection and investigation of ‘criminal offences’. Article 31 TEU provides for common action on judicial cooperation ‘in criminal matters’. EC and EU law do not, however, contain any definition of what is criminal.

2. Distinguishing characteristics of criminal law Looking across legal systems (of the EU Member States, and of third countries, in particular English-speaking third countries), criminal law, as opposed to public law enforcement of a civil or administrative nature, appears to have the following six distinguishing characteristics:4

1 Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules of competition laid down in Articles 81 and 82 of the Treaty, OJ L 1 [2003]. This regulation has replaced, as of 1 May 2004, Council Regulation No 17, OJ 13 [1962] (Special English Edition 1959–62, p 87). 2 On the interpretation of this provision, see text accompanying notes 45 to 50 below, and further Wils W. P. J. (2005): Principles of European Antitrust Enforcement, Hart Publishing, Oxford and Portland, Oregon, section 1.1.4.3.3 (footnote 88). 3 On the meaning of this sentence, see text accompanying notes 74 to 77 below; see also text accompanying note 189 below as to the use of the word ‘criminal’ in the text of the EC Treaty. 4 See Wils W. P. J. (2002): The Optimal Enforcement of EC Antitrust Law, Kluwer Law International, London, section 8.7.2.1.


12/10/07

13:10

Page 269

Is Criminalization of EU Competition Law the Answer?

269

2.1. Criminal penalties Firstly, it appears that criminal law has a monopoly on the use of imprisonment. Whereas fines can be either criminal or civil or administrative, imprisonment appears to be essentially a criminal sanction. The possibility of a prison sanction does not seem to be a necessary condition for a prohibited act or for an enforcement procedure to be criminal, but it is certainly a sufficient condition. Compared with civil or administrative enforcement, criminal enforcement tends to make relatively more use of penalties imposed on individuals (natural persons), and relatively less of penalties imposed on corporate entities or legal persons. Individual penalties are however neither a necessary nor a sufficient condition for enforcement to be criminal. In Denmark, for instance, violations of Articles 81 and 82 EC and of the parallel prohibitions under national competition law are punished with fines which are of a criminal law nature but which are in practice only imposed on undertakings, not on natural persons. In Germany, on the other hand, leaving aside the specific criminal law provisions on bid rigging, violations of Articles 81 and 82 EC and of the parallel national prohibitions are punished through fines which are imposed on natural persons, and only derivatively also on undertakings, but which are of an administrative law nature.5

2.2. Criminal intent Secondly, the commission of a criminal offence usually requires that the prohibited act be committed with a guilty state of mind (criminal intent), and not by mere negligence.

2.3. Moral condemnation Thirdly, the imposition of criminal sanctions carries, and is designed to carry, a stigma effect. Criminal enforcement has a stronger message-sending role or expressive function than civil or administrative enforcement.6 5 For an overview of antitrust enforcement in the 25 Member States, see Cahill D. and Cooke J., eds (2004): The Modernisation of EU Competition Law Enforcement—FIDE 2004 National Reports, Cambridge University Press, and Holmes M. and Davey L., eds. (2004): A Practical Guide to National Competition Rules Across Europe, Kluwer Law International, London. 6 See Adenaes J. (1971): “The Moral or Educative Influence of Criminal Law”, 27 Journal of Social Issues 17, and Adenaes J. (1975): “General prevention revisited: research and policy implications”, 66 Journal of Criminal Law & Criminology 338, at pp. 341–343, as well as Dau-Schmidt K. (1990): “An Economic Analysis of the Criminal Law as a Preference-Shaping Policy”, Duke Law Journal 1, Sunstein C. (1996): “On the Expressive Function of the Law”, 144 University of Pennsylvania Law Review 2021, Kahan D. (1997): “Social Influence, Social Meaning, and Deterrence”, 83 Virginia Law Review 349, Katyal N. (1997): “Deterrence’s Difficulty”, 95


270

12/10/07

13:10

Page 270


2.4. Less strict relationship between penalty and harm Fourthly, in criminal law there appears to be a less strict relationship, if any, between the size of the penalty and the size of the harm caused than in the setting of civil sanctions. This appears to reflect the idea that criminal law does not seek to price certain behaviour (by making the actor bear the external costs of his behaviour) but rather to prohibit it (unconditionally, i.e., irrespective of the actual size of the external costs).7

2.5. Criminal powers of investigation Fifthly, under criminal law, enforcement authorities tend to have stronger investigative powers.8

2.6. Criminal rights of defence Finally, criminal procedures tend to have stronger procedural protections, designed to avoid false convictions. In particular, in criminal enforcement systems the adjudicative or decision-making function is always separated from the investigative and prosecutorial function or functions.9 Whereas it appears to be a necessary condition of criminal enforcement that the body which imposes the criminal penalties is separate and independent from the investigator and prosecutor, it is however not a sufficient condition. In Finland, for instance, violations of Articles 81 and 82 EC and of the parallel national prohibitions are punished with fines for undertakings which are of an administrative law nature but which are imposed by the Market Court, a specialized court entirely separate and independent from the Competition Authority, which acts as investigator and prosecutor. In criminal enforcement systems, the standard of proof Michigan Law Review 2385, Lynch G. (1997): “The Role of Criminal Law in Policing Corporate Misconduct”, 60 Law and Contemporary Problems 23, Kahan D. (1998): “Social Meaning and the Economic Analysis of Crime”, 27 Journal of Legal Studies 609, and Dau-Schmidt K. (1998): “Preference Shaping by the Law”, in Newman P., ed., The New Palgrave Dictionary of Economics and the Law, Macmillan, London, at p. 84. 7 See Cooter R. (1984): “Prices and Sanctions”, 84 Columbia Law Review 1523, and Coffee J. (1992): “Paradigms Lost: The Blurring of the Criminal and Civil Law Models—And What Can Be Done About It”, 101 The Yale Law Journal 1875. 8 For an overview of the law and economics of the collection of intelligence and evidence by the competition authorities from antitrust offenders, see Wils W. P. J. (2003): “Self-incrimination in EC Antitrust Enforcement: A Legal and Economic Analysis”, 26 World Competition 567, and chapter 5 of Principles of European Antitrust Enforcement, supra note 2. 9 See The Optimal Enforcement of EC Antitrust Law, supra note 4, section 8.7.4.5; Wils W. P. J. (2004): “The Combination of the Investigative and Prosecutorial Function and the Adjudicative Function in EC Antitrust Enforcement: A Legal and Economic Analysis”, 27 World Competition 201, and chapter 6 of Principles of European Antitrust Enforcement, supra note 2.


12/10/07

13:10

Page 271


271

for establishing violations also tends to be higher. In the English-speaking countries, criminal convictions normally require proof ‘beyond reasonable doubt’, whereas civil or administrative findings can normally be made by a ‘preponderance of the evidence’ or a ‘balance of probabilities’.

2.7. Links between the six characteristics The six characteristics listed above are clearly interrelated. The use of imprisonment and the stigma effect create the stronger need for procedural protections to avoid false convictions. These stronger procedural protections create, in their turn, the need for stronger investigative powers so that enforcement does not become ineffective. On the other hand, the message-sending role of criminal enforcement explains the requirement of a guilty state of mind (as this shows the need for wrong valuations to be repudiated) as well as the absence of a strict relationship between punishment and actual harm, reflecting the desire to deter the criminal behaviour unconditionally rather than merely to price it.

3. The wider notion of ‘criminal’ in the European Convention on Human Rights The notion of ‘criminal’ enforcement as described above and as used throughout most of this paper should be distinguished from the wider notion of ‘criminal’ as conceived under Article 6 of the European Convention on Human Rights.10 The European Court of Human Rights has developed, within the framework of Article 6 ECHR, an autonomous notion of ‘criminal’ which covers proceedings of an administrative law nature that fulfil the following conditions:11 the offence is defined by a general rule, applicable to all citizens and not only to some of them; the rule is linked to penalties in the event of noncompliance; the sanctions are intended not as a pecuniary compensation for damage but essentially as a punishment to deter repeated offences; and the sanctions are severe.12 10 See Dekeyser K. and Gauer C. (2004): “The New Enforcement System for Articles 81 and 82 EC and the Rights of Defence”, paper presented at the 31st Annual Fordham Corporate Law Institute Conference on International Antitrust Law & Policy, New York, 8 October 2004, published in Hawk B., ed. (2005): 2004 Annual Proceedings of the Fordham Corporate Law Institute—International Antitrust Law and Policy, Juris Publishing, New York, 537–573, at pp. 540–542; The Optimal Enforcement of EC Antitrust Law, as note 4 above, section 8.7.2.3; and Principles of European Antitrust Enforcement, supra note 2, sections 1.1.4.3.3 (footnote 88) and 6.1.4.1. 11 Restated in the words of Dekeyser K. and Gauer C, supra note 10, at 541. 12 Engel and Others v the Netherlands, judgment of 8 June 1976, Series A no. 22, paragraph 82; Özturk v Germany, judgment of 21 February 1984, Series A no. 73, paragraph 50; Bendenoun v France, judgment of 24 February 1994, Series A no. 284, paragraph 47.


272

12/10/07

13:10

Page 272


The European Court of Human Rights has never had the occasion to rule on an antitrust enforcement procedure at the level of the EU institutions or in one of the EU Member States.13 In 1989, the European Commission of Human Rights (a body which has since been abolished) decided on a case involving competition proceedings before the French competition authority.14 It decided that the sanctions imposed by that authority were of a criminal nature within the meaning of Article 6 ECHR. However, the case was not brought before the European Court of Human Rights. In 2004, the European Court of Human Rights decided on a case concerning proceedings under the Russian Law on Competition and the Restriction of Monopolies in the Commodity Markets, which had led to the confiscation of profits obtained by a number of companies violating that law. The European Court of Human Rights declared the application brought by the companies concerned for alleged violation of Article 6 ECHR inadmissible on the ground that the proceedings at issue were not criminal within the meaning of Article 6 ECHR.15 The EC Court of Justice appears to have acknowledged in its Hüls judgment of 1999 that the fines imposed by the European Commission under then-Regulation No. 17 (now Regulation No. 1/2003) may be qualified as ‘criminal’ within the meaning of Article 6 ECHR.16 Antitrust proceedings at the level of the EU institutions or in the EU Member States which are, according to EU law or to the applicable national law, of an administrative or civil nature, may thus fall under the wider notion of ‘criminal’ within the meaning of Article 6 ECHR. However, it should be pointed out that, in its case law on the compatibility of specific proceedings with the substantive requirements of Article 6 ECHR, the European Court of Human Rights appears to distinguish between, on the one hand, enforcement proceedings which are criminal within a more traditional, narrower sense, or which are considered as criminal under the applicable domestic law, and, on the other hand, proceedings which are only criminal within the wider meaning of Article 6 ECHR. 13 The case of Senator Lines GmbH v the 15 Member States of the European Union (application no. 56672/00) would have furnished such an occasion had it been heard on the merits. The European Commission had imposed a fine of 273 million Euros on Senator Lines, and the EC Courts rejected the request of the company to suspend the payment of the fine. Senator Lines GmbH alleged a violation of Article 6 ECHR (right of access to court), in that it was required to pay the fine before a decision was taken in the substantive proceedings before the EC Courts. It claimed that this would have resulted in the insolvency and liquidation of the company before the substantive issues were determined. However, the application was declared inadmissible by the European Court of Human Rights after the EC Court of First Instance decided to set aside the fine imposed by the European Commission; see Grand Chamber Decision of 10 March 2004, available at http://www.echr.coe.int/Eng/Judgments.htm. 14 European Commission of Human Rights, Stenuit v France, A/232-A decision of 11 July 1989. Stenuit withdrew its application in 1991. 15 Decision of 3 June 2004 as to the admissibility of applications by OOO Neste St. Petersburg and Others v Russia, available at http://www.echr.coe.int/Eng/Judgments.htm. 16 Case C-199/92 P Hüls v European Commission [1999] ECR 4287, paragraph 150.


12/10/07

13:10

Page 273


273

This distinction appears most clearly in the case law of the European Court of Human Rights on the separation of the decision-making function from the investigative and prosecutorial function or functions. Indeed, the European Court of Human Rights has ruled that, for reasons of efficiency, the determination of civil rights and obligations or the prosecution and punishment of offences which are ‘criminal’ within the wider meaning of Article 6 ECHR can be entrusted to administrative authorities, provided that the persons concerned are able to challenge any decision thus made before a judicial body which has full jurisdiction and which provides the full guarantees of Article 6(1) ECHR.17 This means that antitrust proceedings such as those at the level of the EU institutions are not incompatible with Article 6 ECHR, even if the European Commission combines the investigative and prosecutorial function with the decision-making function, given that the addressees of European Commission decisions imposing fines can bring an action for annulment before the EC Court of First Instance, which manifestly provides the full guarantees of Article 6(1) ECHR and which undertakes a comprehensive review of the Commission’s decisions.18 However, this alternative means of satisfying the requirements of Article 6(1) ECHR does not appear to be available in more traditional areas of criminal law or in areas considered criminal under domestic law. In those areas, the case law of the European Court of Human Rights appears to require separation of the adjudicative function from the investigative and prosecutorial function already at first instance.19

B. Is there a tendency to criminalize antitrust enforcement in the EU Member States? 1. US antitrust enforcement To the extent that there is currently a tendency in the EU Member States, or even at the level of the EU institutions,20 to criminalize antitrust enforcement, 17 Judgments of 23 June 1981, Le Compte, Van Leuven and De Meyere v Belgium, A/43 paragraph 51, of 1 February 1983, Albert and Le Compte v Belgium, A/58 paragraph 29, of 21 February 1984, Öztürk v Germany, A/73 paragraph 56, and of 24 February 1994, Bendenoun v France, A/284, paragraph 46. 18 See the judgment of the Court of First Instance of 15 March 2000 in Joined Cases T-25/95 a.o. Cimenteries CBR and Others v Commission [2000] ECR II-700, paragraph 719 (“When the Court of First Instance reviews the legality of a decision finding an infringement of [Article 81(1) and/or Article 82 EC], the applicants may call upon it to undertake an exhaustive review of both the Commission’s substantive findings of fact and its legal appraisal of those facts.”). 19 See the judgments of 26 October 1984, De Cubber v Belgium, A/86 paragraphs 31–32, and of 25 February 1997, Findlay v United Kingdom, Reports 1997-I paragraph 79. 20 See text accompanying notes 84 to 99 below as to EU Member States and text accompanying notes 53 to 55 below as to the EU institutions.


274

12/10/07

13:10

Page 274


this is undoubtedly inspired by US antitrust enforcement, even if criminal antitrust legislation in the US may historically in turn have been inspired by older European precedents.21 Since the original 1890 version of the Sherman Act, violations of Sections 1 and 2 of the Sherman Act have been punishable under US federal law by criminal fines for both enterprises and individuals, and by imprisonment for individuals.22 Initially, the maximum jail term was 1 year.23 In 1974, Congress raised the maximum jail term to three years,24 and raised it again in 2004 to ten years.25 The maximum criminal fines for enterprises were similarly increased to $1 million in 1974,26 to $10 million in 1990,27 and to $100 million in 2004.28 More than half of the US states also have criminal antitrust statutes that provide for jail sentences, and many of these statutes pre-date the federal Sherman Act.29 It was not until 1959, however, that prison sentences were imposed against businessmen for price fixing without acts or threats of violence.30 Since then, enforcement has increased very strongly. During the 1990s, the US Department of Justice successfully prosecuted an average of more than 35 individuals each year. During 1990-1999, 132 individuals were sentenced to actual prison terms averaging almost 9 months, and 154 were sentenced to periods of confinement in a halfway house or community treatment centre averaging 4 months.31

21 In particular, the US appears to have been influenced by English legislation of the 16th Century, which had however fallen into disuse by the 19th Century. See Green N. (2004): “The Road to Conviction—The Criminalisation of Cartel Law”, in Hawk B., ed., 2003 Annual Proceedings of the Fordham Corporate Law Institute International Antitrust Law & Policy, Juris Publishing, New York, 13, at 13–22. In France, following the Loi Le Chapelier of 2–17 March 1791, a criminal cartel prohibition was also contained in Article 419 of the Napoleonic Penal Code. 22 For an overview of the history of US criminal antitrust enforcement from 1890 to 2001, see Baker D. (2001): “The Use of Criminal Law Remedies to Deter and Punish Cartels and BidRigging”, 69 George Washington Law Review 693, at pp. 694–696. 23 Sherman Act, ch. 647, 26 Stat. 209 (1890). 24 Antitrust Procedures and Penalties Act, Pub. L. No. 93-528, § 3, 88 Stat. 1706, 1708 (1974). 25 Antitrust Criminal Penalty Enhancement and Reform Act of 2004, H.R. 1086. 26 Supra note 24. 27 Antitrust Amendments Act of 1990, Pub. L. No. 101-588, § 4, 104 Stat. 2879, 2880 (1990). 28 Supra note 25. 29 Dau-Schmidt K. (1993): “Sentencing Antitrust Offenders: Reconciling Economic Theory with Legal Theory”, 9 William Mitchell Law Review 75, at pp. 75–76. 30 Gallo J., Dau-Schmidt K., Craycraft J. and Parker C. (1994): “Criminal Penalties Under the Sherman Act: A Study of Law and Economics”, 16 Research in Law and Economics 25, at p. 40. 31 “Antitrust Deterrence in the United States and Japan”, remarks by S.M. Chemtob at a Conference on Competition Policy in the Global Trading System (Washington DC, 23 June 2000), available at http://www.usdoj.gov/atr/public/speeches/5076.pdf, at 7–8. For more statistics, see Gallo J., Dau-Schmidt K., Craycraft J. and Parker C. (2000): “Department of Justice Antitrust Enforcement, 1955–1997: An Empirical Study”, 17 Review of Industrial Organization 75.


12/10/07

13:10

Page 275


275

In his 2004 overview of US antitrust enforcement, Joel Davidow reported: “In the last fiscal year, defendants in [Department of Justice Antitrust] Division cases were sentenced to more than 10,000 jail days—a record—with an average sentence of more than 18 months. In the last four years, a total of over 75 years of imprisonment have been imposed on antitrust defendants, with more than 30 defendants receiving prison sentences of one year or longer. It is not just US executives who are facing prison sentences, but foreign executives as well. Business people from Canada, France, Germany, Sweden, and Switzerland have now served time in US prisons for violating US antitrust laws. Extradition of cartel leaders is now frequently sought. The Justice Department is conducting over 40 criminal investigations of international cartel activity.”32 Although the text of the criminal provisions in the Sherman Act covers all types of restrictive agreements as well as unilateral monopolistic behaviour, by long tradition the US Department of Justice limits criminal prosecution to hard core violations of Section 1 of the Sherman Act such as price fixing, bid rigging, market division, and customer allocation schemes among horizontal competitors.33 As the US Department of Justice has no powers to impose or to seek the imposition of civil or administrative fines, it only seeks prospective, injunctive relief against other violations of Section 1 and all violations of Section 2 of the Sherman Act. The Federal Trade Commission has relatively recently in competition cases applied its power, which it had traditionally used in consumer protection cases, to seek disgorgement of unlawful profits.34 Otherwise, deterrence and punishment are left to private treble-damage suits.35 The imposition of criminal penalties for violation of Section 1 of the Sherman Act requires a finding of criminal intent. It also carries a rather strong stamp of moral condemnation. Indeed, in the US, antitrust values 32 Davidow J. (2004): “Recent US Antitrust Developments of International Relevance”, 27 World Competition 407, at 409. See also most recently Hammond S. (2005): “An Overview Of Recent Developments In The Antitrust Division’s Criminal Enforcement Program”, address before the American Bar Association Midwinter Leadership Meeting, Kona, Hawaii, 10 January 2005, available at http://www.justice.gov/atr/public/speeches/207226.htm. 33 Waller S. (2003): “The Incoherence of Punishment in Antitrust”, 78 Chicago-Kent Law Review 207, at 216, referring in footnote 48 to US v. Dunham Concrete Prods., Crim. No. 1842 (E.D. La. 1969) as the last criminal monopolization indictment discovered by that author; Connor J. (2005): Price-Fixing Overcharges: Legal and Economic Evidence, American Antitrust Institute (AAI) Working Paper No. 04–05, available at http://www.antitrustinstitute.org/ recent2/355.pdf, at 7. 34 Section 13 (G) of the FTC Act; see FTC v. Mylan Labs. Inc., 62 F. Supp. 2d 25, 36 (D.D.C. 1999) and Waller S., supra note 33, at 218. 35 See Waller S., supra note 33, at 210 and 230–232; Calkins S. (1997): “Corporate Compliance and the Antitrust Agencies’ Bi-Modal Penalties”, 60 Law and Contemporary Problems 127, at 136–139 and 150; and letter dated 5 January 2005 from R.H. Pate, Assistant Attorney General, Department of Justice Antitrust Division, to D.A. Garza, Chair, Antitrust Modernization Commission, “Suggested Topics for Antitrust Modernization Commission Study”, available at http://www.usdoj.gov/atr/public/comments/207122.pdf, at 2.


276

12/10/07

13:10

Page 276


enjoy a very wide consensus,36 and price fixing is considered to be immoral, like theft.37 To investigate suspected criminal antitrust violations, the US Department of Justice can make use of a full arsenal of criminal investigatory powers.38 The US Department of Justice also operates a very successful Corporate Leniency Policy and an Individual Leniency Policy, which allow corporations and individuals to avoid criminal prosecution if they are the first to cooperate with the Department.39 Defendants in turn benefit from a full range of criminal rights of defence— which are, however, less extensive for corporations than for individuals40— and violations have to be proven ‘beyond reasonable doubt’.

2. Antitrust enforcement at the level of the EU institutions As to the enforcement of Articles 81 and 82 EC at the level of the EU institutions, Article 23(2) of Regulation No. 1/2003 only empowers the European Commission to impose fines for violations of Articles 81 and 82 on ‘undertakings’. Neither the Treaty nor the Regulation contains any definition of the term ‘undertaking’. The Court of Justice has held that, “in competition law, the term ‘undertaking’ must be understood as designating an economic unit [. . .] even if in law that economic unit consists of several persons, natural or legal”.41 In practice, the Commission does not simply impose fines on undertakings thus defined as economic units. For reasons of enforceability, it rather addresses its fining decisions to entities possessing legal personality, typically companies, to which the violation committed by the undertaking is imputed.42 As to the choice of the person or persons to whom a violation is to 36 In the words of Robert Reich, “antitrust is popular with the public” and “has a compelling ideological attraction”. He refers to “the ideal of competition continuing to receive broad support from a wide spectrum of American public opinion”. Reich R. (1980): “The Antitrust Industry”, 68 Georgetown Law Journal 1035, at p. 1054 and footnote 1. 37 See The Optimal Enforcement of EC Antitrust Law, supra note 4, section 8.6.7, and D.I. Baker, supra note 22, at 714. 38 See text accompanying notes 127 to 128 and 156 to 161 below. 39 US Department of Justice Corporate Leniency Policy (August 10, 1993), available at http://www.usdoj.gov/atr/public/guidelines/0091.pdf, and US Department of Justice Leniency Policy for Individuals (August 10, 1994), available at http://www.usdoj.gov/atr/public/guidelines/0092.pdf. See further text accompanying notes 158 to 160 below. 40 In particular, corporations cannot invoke the privilege against self-incrimination enshrined in the 5th Amendment to the US Constitution. See White v. US, 322 U.S. 694 (1911). 41 Judgment of 12 July 1985 in Case 170/83 Hydrotherm v Compact [1985] ECR 3016, paragraph 11. For a detailed analysis of the notion of ‘undertaking’ and the relationship between ‘undertaking’ and legal or natural persons, see The Optimal Enforcement of EC Antitrust Law, supra note 4, chapter 7. 42 See Opinion of Judge Vesterdorf, acting as Advocate General, in Case T-1/89 RhônePoulenc v Commission [1991] ECR II-869 at 916; judgment of the Court of First Instance of 20 April 1999 in Joined Cases T-305/94, T-306/94, T-307/94, T-313/94, T-314/94, T-315/94,


12/10/07

13:10

Page 277


277

be imputed, the general rule formulated by the Community Courts is that, “when [. . .] a violation is found to have been committed, it is necessary to identify the natural or legal person who was responsible for the operation of the undertaking at the time when the violation was committed, so that it can answer for it”.43 If the defendant undertaking were found to consist of an unincorporated business (a single trader, with or without employees, who has not incorporated his or her business, or a professional exercising his or her profession alone and unincorporated, or several natural persons operating a single business without any employment relationship between them and without any form of legal person), the Commission would necessarily have to address its fining decision to the natural person or persons operating the business. However, in the more than one hundred decisions in which the Commission has imposed fines under Regulation No. 17 and Regulation No. 1/2003, this situation has never occurred: thus far, all fines have been imposed on companies or other legal persons. Indeed, the undertakings found to have committed violations of Article 81 or 82 invariably either coincided with a single company or some other legal person which was the obvious addressee of the fining decision, or else they consisted of a group of companies, in which case fines were imposed on one or more of the companies in the group.44 According to Article 23(2) of Regulation No. 1/2003, fines can be imposed on undertakings where they infringe Article 81 or 82 EC “either intentionally or negligently”. Criminal intent is thus clearly not required. Article 23(5) of the Regulation expressly provides that the decisions by which the Commission imposes fines “shall not be of a criminal law nature”. As explained above,45 this provision appears to have some relevance not for the question of whether Article 6 ECHR is applicable to the antitrust enforcement proceedings at the level of the EU institutions, but rather for the question of the compatibility of those proceedings with the substantive requirements of Article 6 ECHR. The EU Council of Ministers, when it adopted Regulation No. 1/2003, may also have considered that this provision

T-316/94, T-318/94, T-325/94, T-328/94, T-329/94 and T-335/94 Limburgse Vinyl Maatschappij a.o. v Commission [1999] ECR II-945, paragraph 978. 43 Judgment of the Court of First Instance of 17 December 1991 in Case T-6/89 Enichem Anic v Commission [1991] ECR II-1695, paragraph 236; judgment of the Court of Justice of 16 November 2000 in Case C-279/98 P Cascades v Commission [2000] ECR I-9709, paragraph 78. 44 In Pre-Insulated Pipes, one of the undertakings was controlled and managed by a natural person, Dr. W. Henss, but the Commission appears to have made an effort (for reasons unexplained) to avoid imposing the fine on him, identifying instead a collection of companies which the Commission held jointly and severally liable for the fine. See recitals 157–160 and Article 3(d) of the Decision of 21 October 1998, Pre-Insulated Pipe Cartel, [1999] OJ L24/1; and paragraph 105 of the judgment of the Court of First Instance of 20 March 2002 in Case T-9/99 HFB and Others v Commission [2002] ECR II-1530, in which the court found that the Commission had ‘intentionally’ not established Dr Henss’ liability. 45 Text accompanying notes 16 to 21.


278

12/10/07

13:10

Page 278


had some relevance as to the necessary legal basis under the EC Treaty.46 It may also have some importance as to the status of such fines under national law, for instance with respect to whether they are tax deductible.47 The affirmation that fining decisions “shall not be of criminal nature” could also be read as reflecting the absence of moral condemnation for violations of Articles 81 and 82. On the other hand, there appears to have been in the last years from the side of the European Commission an increase in the rhetoric of condemnation of cartels in particular. For instance, in a speech in 2000, Mario Monti, the then-Member of the European Commission in charge of competition, declared that “cartels are cancers on the open market economy, which forms the very basis of our [European] Community”.48 It may also be noticed that the European Commission’s Directorate General for Competition has in the last years started using the term ‘antitrust’ alongside the official term ‘competition law’.49 As Christopher Harding has pointed out, the North American term ‘antitrust’ reflects the proscriptive and negative character of US antitrust law, as opposed to the historically more tolerant European approach.50 As to the relationship between the size of the penalty for antitrust violations and the magnitude of the harm caused by them, Christopher Harding has made the following observations, starting from the ranking of ‘seriousness’ in the European Commission Guidelines on the method of setting fines:51 “What is especially notable in this ranking is the focus on elements of market impact, and especially – at the higher level of seriousness – impact on the functioning of the single market. There is no explicit mention of the ‘bad attitude’ elements which are so prominent in the American rules, so that the European ‘offence’ is very much an offence of outcome. A further examination of the Commission’s practice in determining the severity of fines (its ‘sentencing practice’) reveals that market impact factors supply the cardinal points on its sentencing tariff, while attitudinal factors (such as organized collusive activity, furtiveness and secrecy, antitrust awareness, recidivist behaviour) appear as aggravating factors, or ordinal points,

46

See further text accompanying notes 187 to 192 and footnote 192 below. See Principles of European Antitrust Enforcement, supra note 2, section 1.1.4.3.3, footnote 88, and Kerse C. and Khan N. (2005): EC Antitrust Procedure, 5th edition, Thomson—Sweet & Maxwell, London, at paragraph 7-090. 48 “Fighting Cartels Why and How? Why should we be concerned with cartels and collusive behaviour?”, speech at the 3rd Nordic Competition Policy Conference (Stockholm, 11–12 September 2000), available at http://europa.eu.int/comm/competition/speeches. 49 See the website of DG Competition: http://europa.eu.int/comm/competition/index_en.html. 50 Harding C. (2002): “Business Cartels as a Criminal Activity: Reconciling North American and European Models of Regulation“, 4 Maastricht Journal of European and Comparative Law 393, at p. 404. 51 Guidelines on the method of setting fines imposed pursuant to Article 15(2) of Regulation No 17 and Article 65(5) of the ECSC Treaty, OJ C 9 [1998]; these guidelines remain valid under Regulation No. 1/2003; see Article 43(3) of Regulation No. 1/2003. The ECSC Treaty expired on 23 July 2002. 47


12/10/07

13:10

Page 279


279

moving up and down the tariff within a range founded upon a base gravity derived from market impact.”52

As to the Commission’s powers of investigation, Regulation No. 17 already gave the Commission significant powers, in particular the power to conduct inspections at business premises, including the possibility to overcome opposition with the assistance of the police or other authorities of the Member State where the inspection takes place. But these powers are clearly much more limited than those of the US Department of Justice or those which are common for criminal investigations in the EU Member States. Since 1996, the European Commission has been operating a leniency programme, which, in its current version dating from 2002, offers full immunity to the first cartel member to cooperate with the Commission’s investigation.53 Such a leniency policy, which was clearly adopted following the US example, was traditionally reserved in the EU Member States (if known at all) for the most serious forms of organized crime.54 Of course, since the Commission can only impose fines on undertakings, the leniency programme only concerns them, and not individuals (unless the individual qualifies as an undertaking55). Regulation No. 1/2003 has strengthened the Commission’s powers of investigation in two respects.56 Firstly, Article 21(1) of Regulation No. 1/2003 empowers the Commission to conduct investigations in premises other than business premises, including the homes of directors and employees of the undertakings concerned, “if a reasonable suspicion exists that books or other records related to the business and to the subject-matter of the inspection, which may be relevant to prove a serious violation of [Articles 81 or 82 EC], are being kept in [those] premises”. Article 21(3) adds that a decision to conduct such an inspection can only be executed with prior authorization from

52 C. Harding, supra note 50, at 412. However, it should be pointed out that the EC Courts have held, also after the publication by the Commission of its Guidelines, supra note 51, that, in assessing the gravity of the infringement, “factors relating to the intentional aspect, and thus to the object of a course of conduct, may be more significant than those relating to its effects [. . .], particularly where they relate to infringements which are intrinsically serious, such as pricefixing”. Judgment of the Court of First Instance of 13 December 2001 in Joined Cases T-45/98 and T-47/98 Krupp Thyssen Stainless and Acciai Speciali Terni v Commission [2001] ECR II3823, paragraph 199. See also the judgment of the Court of First Instance of 30 September 2003 in Case T-203/01 Michelin v Commission, [2003] ECR II-4071, paragraph 259. But see also the judgment of the Court of First Instance of 21 October 2003 in Case T-368/00 Opel and GM Nederland, [2003] ECR II-4491, paragraphs 195 and 196. 53 European Commission Notice on immunity from fines and reduction of fines in cartel cases, [2002] OJ C45/3, replacing the earlier European Commission Notice on the non-imposition or reduction of fines in cartel cases, OJ C 207 [1996]. 54 For instance, leniency programmes apply in connection with enforcement of the criminal laws against the mafia in Italy. 55 See text accompanying notes 41 to 44 above. 56 For a detailed analysis of these changes, see K. Dekeyser and C. Gauer, supra note 10.


280

12/10/07

13:10

Page 280


the national judicial authority of the Member State where the inspection takes place. Secondly, Article 20(2)(e) of the Regulation provides that, during an inspection at the business premises of an undertaking, the inspectors are empowered “to ask any representative or member of staff of the undertaking [. . .] for explanations on facts or documents relating to the subject-matter and purpose of the inspection and to record the answers”.57 However, Regulation No. 1/2003 does not threaten the individuals thus questioned with any sanctions for a refusal to answer, nor for giving incorrect, incomplete or misleading answers. Article 23(1)(d) of the Regulation only provides that fines may be imposed on undertakings that “fail or refuse to provide a complete answer on facts”, “give an incorrect or misleading answer” or “fail to rectify within a time-limit set by the Commission an incorrect, incomplete or misleading answer given by a staff member”. As to the defendant companies’ rights of defence in antitrust proceedings conducted by the Commission, they have significantly increased over time.58 The Commission combines the investigative and prosecutorial function with the adjudicative function at first instance, but its decisions are subject to an exhaustive review by the Court of First Instance.59 As to the standard of proof which the Commission has to satisfy to find a violation of Article 81 or 82 and impose a fine, according to Chris Kerse and Nicholas Khan, it “is a high one, although it would be going too far to say that the infringement must be proven to the criminal standard of ‘beyond reasonable doubt’”.60

3. The EU framework for antitrust enforcement in the EU Member States Regulation No. 1/2003 frames antitrust enforcement in the EU Member States in two ways. Firstly, it allows Member States to provide for criminal

57 Under Regulation No. 17, Commission officials could only seek explanations relating to the books and records under examination. see the judgment of the Court of Justice of 26 June 1980 in Case 136/79, National Panasonic v Commission [1980] ECR 2056, paragraph 15; Order of the Court of First Instance of 9 June 1997 in Case T-9/97 Elf Atochem v Commission [1997] ECR II-919, paragraph 23. 58 See Lenaerts K. and Maselis I. (2001): “Procedural Rights and Issues in the Enforcement of Articles 81 and 82 of the EC Treaty”, 24 Fordham International Law Journal 1615, and Kerse C. and Khan N., supra note 47. 59 See notes 18 and 19 above. 60 Kerse C. and Khan N., supra note 47, paragraph 8-038. The standard formula used by the EC Courts is that “the Commission must produce sufficiently precise and consistent evidence to support the firm conviction that the alleged infringement took place”. See, e.g., the judgment of the Court of First Instance of 8 July 2004 in Joined Cases T-67/00 a.o. JFE Engineering Corp. v European Commission, [2004] ECR II-2501, paragraph 179.


12/10/07

13:10

Page 281


281

enforcement of Articles 81 and 82. Secondly, it restricts the possibilities for Member States to enforce their national competition laws in isolation from the enforcement of Articles 81 and 82 EC.

3.1. Member States can criminalize their enforcement of Articles 81 and 82 EC Article 5 of Regulation No. 1/2003 provides that “[t]he 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 [. . .] decisions [. . .] imposing fines, periodic penalty payments or any other penalty provided for in their national law”. Member States can thus provide in their national law for administrative or civil fines to be imposed on companies for violations of Articles 81 or 82, or criminal fines, or fines for individuals, of an administrative, civil or criminal nature, or imprisonment, or indeed “any other penalty”. That “any other penalty” also covers penalties on individuals, including imprisonment, is further confirmed by Article 12(3) of the Regulation. Whereas Article 12(1) provides that the Commission and the competition authorities of the Member States can exchange all kinds of information, and use the information thus exchanged in evidence, Article 12(3) adds the following limitation: ”Information exchanged pursuant to paragraph 1 can only be used in evidence to impose sanctions on natural persons where: —the law of the transmitting authority foresees sanctions of a similar kind in relation to an infringement of [Articles 81 or 82 EC] or, in the absence thereof, —the information has been collected in a way which respects the same level of protection of the rights of defence of natural persons as provided for under the national rules of the receiving authority. However, in this case, the information exchanged cannot be used by the receiving authority to impose custodial sanctions.”

This provision confirms what is already apparent from the terms “any other penalty” in Article 5 of the Regulation, namely that Member States can indeed provide for sanctions enforceable against natural persons, including imprisonment, for violations of Articles 81 and 82 EC. By thus allowing Member States to provide for criminal enforcement of Articles 81 and 82 EC, the Council of Ministers, which adopted Regulation No. 1/2003 unanimously, appears to have rejected the view, which has sometimes been suggested, that Articles 81 and 82 EC are somehow inherently non-criminal and that the EC Treaty thus excludes criminal enforcement of


282

12/10/07

13:10

Page 282


Articles 81 and 82 EC not only at the level of the EU institutions but also at the level of the Member States.61 This view indeed appears to be mistaken. It is of course correct that Articles 81 and 82 EC only mention “undertakings”. But, as is apparent from Article 83 EC, the prohibitions on restrictive agreements and abuse of dominant positions by undertakings, as set out in Articles 81 and 82 EC, are only “principles”, and the appropriate regulations or directives to give effect to these principles are to be laid down by the Council. Nothing in the EC Treaty suggests that the Council could thus not allow Member States to provide for penalties for individuals, including imprisonment, to enforce Articles 81 and 82, as indeed it did in Regulation No. 1/2003.62 The possibility offered by Regulation No. 1/2003 for Member States to provide for criminal enforcement of Articles 81 and 82 EC has in fact been used by a number of Member States.63 Ireland and Estonia have provided only for criminal enforcement, through fines for companies and individuals as well as imprisonment.64 In the United Kingdom, non-criminal fines are provided for companies, but there are also provisions for director disqualification for individuals as well as imprisonment and/or criminal fines for individuals, specifically for price fixing, limitation of supply or production, market sharing and bid rigging.65 In France, Cyprus and the Slovak Republic, the main sanctions provided for are non-criminal fines for companies, but imprisonment for individuals is also provided for, at least in theory.66 In Denmark and Malta, essentially only fines for companies are provided for, but these have a criminal character. In Malta, directors may be held jointly and severally liable with the company for these fines. In Greece, noncriminal fines for companies (with joint and several liability for managers) are complemented by criminal fines for individuals. In Germany, non-criminal fines are provided for individuals, and derivatively also for companies, and imprisonment penalties are specifically imposed for bid rigging. In Austria, non-criminal fines are provided only for companies, but as with Germany, imprisonment is specifically provided for in the case of bid rigging.67

61 See remarks by Gerard Hogan in Roundtable Debate Session IV—Rights, Privileges and Ethics in Competition Cases, the 31st Annual Fordham Corporate Law Institute Conference on International Antitrust Law & Policy (New York, 8 October 2004), in Hawk B., ed. (2005): 2004 Annual Proceedings of the Fordham Corporate Law Institute—International Antitrust Law and Policy, Juris Publishing, New York. 62 See also text accompanying note 189 below. 63 See supra note 5. 64 See also text following note 84 below on Ireland. 65 See also text accompanying notes 85–86 and 150 below on the United Kingdom. 66 See also text following note 78 below on France. 67 See also text following note 78 below on Austria.


12/10/07

13:10

Page 283


283

3.2. Member States cannot criminalize the enforcement of only their national competition laws The possibilities for Member States to enforce their national competition laws in isolation from the enforcement of Articles 81 and 82 EC are limited in three ways. Firstly, Article 3(2) of Regulation No. 1/2003 forbids Member States to prohibit under national competition law agreements or concerted practices which may affect trade between Member States but which do not restrict competition within the meaning of Article 81(1), or which fulfil the conditions of Article 81(3) EC or which are covered by a block exemption regulation applying Article 81(3) EC.68 Secondly, it follows from the principle of equivalence, a general principle of Community law developed in the case law of the Court of Justice,69 that the penalties which Member States impose for violations of Articles 81 and 82 EC pursuant to Article 5 of Regulation No. 1/2003 must at least be equivalent in effectiveness and dissuasiveness to the sanctions they provide for comparable violations of their national competition laws. For instance, if a Member State provides for imprisonment sanctions to enforce a cartel prohibition under national law, it must also provide for such sanctions to enforce the cartel prohibition contained in Article 81 EC. Thirdly, Article 3(1) of Regulation No. 1/2003 provides as follows: “Where the competition authorities of the Member States or national courts apply national competition law to agreements, [. . .] or concerted practices within the meaning of [Article 81(1) EC] which may affect trade between Member States within the meaning of that provision, they shall also apply [Article 81 EC] to such agreements, [. . .] or concerted practices. Where the competition authorities of the Member States or national courts apply national competition law to any abuse prohibited by [Article 82 EC], they shall also apply [Article 82 EC].”

Unless the practices concerned do not affect trade between Member States, national competition authorities and national courts thus cannot apply national competition law without simultaneously applying Articles 81 and 82. This means that they cannot avoid the application of the mechanisms of cooperation with the European Commission and/or within the EU network of competition authorities as provided for in Articles 11 and 15 of Regulation No. 1/2003.70 68 On Article 3 of Regulation No. 1/2003 and on block exemption regulations, see Principles of European Antitrust Enforcement, supra note 2, sections 1.2.8.2 and 1.2.1 respectively. 69 See, inter alia, the judgment of the Court of Justice of 15 September 1998 in Case C-231/96 Edis [1998] ECR I-4990, paragraphs 34 and 36–37; for a more detailed analysis, see Oliver P. (2005): “Le règlement 1/2003 et les principes d’efficacité et d’équivalence”, Cahiers de droit européen 343. 70 On these cooperation mechanisms, see Schnichels D. (2004): “The Network of Competition Authorities: How will it work in practice?”, in Geradin D., ed., Modernisation and enlargement: Two major challenges for EC competition law, Intersentia, Antwerp, 99–121 and Principles of European Antitrust Enforcement, supra note 2, chapters 1 and 2.


284

12/10/07

13:10

Page 284


Article 3(3) of Regulation No. 1/2003 clarifies that, “[w]ithout prejudice to general principles and other provisions of [European] Community law”, Articles 3(1) and 3(2) “do [not] preclude the application of provisions of national law that predominantly pursue an objective different from that pursued by [Articles 81 and 82 EC]”. Recital 9 of Regulation No. 1/2003 explains that Articles 81 and 82 “have as their objective the protection of competition on the market”, and gives “national legislation that prohibits or imposes sanctions on acts of unfair trading practice” as an example of legislation protecting “other legitimate interests”. The three limitations mentioned above can be illustrated by the cartel offence in the United Kingdom. Section 188 of the Enterprise Act 2002 provides that an individual is guilty of this criminal offence, sanctionable by imprisonment and/or a fine, if he or she dishonestly agrees with one or more other persons that undertakings will engage in one or more of the prohibited cartel activities, namely price fixing, limitation of supply or production, market sharing, or bid rigging.71 Firstly, given that price fixing, limitation of supply or production, market sharing and bid rigging—when they actually or potentially affect trade between Member States—are also prohibited under Article 81 EC, this prohibition under United Kingdom law is compatible with Article 3(2) of Regulation No. 1/2003. Secondly, Section 188 of the Enterprise Act 2002 does not distinguish in any way between, on the one hand, price fixing, limitation of supply or production, market sharing, or bid rigging which may affect trade between Member States and which thus falls under Article 81, and, on the other hand, prohibited cartel activities which do not affect trade between Member States and which thus only fall under the Chapter I prohibition of the Competition Act 1998 (the provision of UK national competition law comparable to Article 81 EC). Unless the UK authorities were to develop an administrative practice of only investigating and prosecuting cases which concern prohibited cartel activities that do not affect trade between Member States, the principle of equivalence is thus fully respected. Thirdly, when a case which concerns prohibited cartel activities that may affect trade between Member States is investigated, prosecuted and punished by the competent UK authorities and courts, it would in my view follow from Article 3(1) of Regulation No. 1/2003 that they will have to establish that the prohibited cartel activities indeed may affect trade between Member States and are thus contrary to Article 81, and they will have to comply with the mechanisms of cooperation with the Commission and within the EU network of competition authorities as provided for in Articles 11 and 15 of the Regulation. 71 The text of the Enterprise Act 2002 is available at http://www.legislation.hmso.gov.uk/ acts/acts2002/20020040.htm.


12/10/07

13:10

Page 285


285

No prosecution has yet been brought under Section 188 of the Enterprise Act 2002, but the UK authorities have indicated that they consider that Regulation No. 1/2003 would not apply to the cartel offence.72 In the view of the UK authorities: “The cartel offence in Part 6 of the Enterprise Act and the prohibitions of Article 81 or Chapter I are aimed at different legal persons and the former is concerned with what is essentially dishonest activity. The cartel offence aims to catch dishonest activity by individuals whereas the civil prohibitions are directed at undertakings. [Regulation No. 1/2003] does not apply to national laws, which impose criminal sanctions on natural persons, except to the extent that such sanctions are the means whereby competition rules applying to undertakings are enforced.”73 As already argued above,74 I believe that the idea that Article 81 is somehow inherently non-criminal, and could not be enforced by penalties on individuals, including penalties of a criminal law nature, is mistaken, and contradicted by Regulation No. 1/2003 as well as by the practice of other Member States, in particular Ireland and Estonia but also France, Cyprus, the Slovak Republic, Denmark, Malta and Greece. The UK authorities’ view that Regulation No. 1/2003 would not apply to the criminal cartel offence apparently also relies upon the very last sentence of recital 8 of Regulation No. 1/2003, which says that “this Regulation does not apply to national laws which impose criminal sanctions on natural persons except to the extent that such sanctions are the means whereby competition rules applying to undertakings are enforced”.75 However, this sentence in my view merely specifies, with regard to national laws which impose criminal sanctions on natural persons, what already follows from Article 3(3) of Regulation No. 1/2003, namely that Articles 3(1) and 3(2) of this Regulation do not preclude the application of provisions of national law that predominantly pursue an objective different from that pursued by Articles 81 and 82 EC. Articles 3(1) and 3(2) of the Regulation would however appear to apply to all national laws that do not predominantly pursue an objective different from that pursued by Articles 81 and 82 EC, irrespective of the question whether those laws impose 72 This view has also been expressed by an official of the European Commission’s Directorate General for Competition, speaking in a personal capacity: see remarks by Kris Dekeyser in Roundtable Debate Session IV—Rights, Privileges and Ethics in Competition Cases, the 31st Annual Fordham Corporate Law Institute Conference on International Antitrust Law & Policy (New York, 8 October 2004), forthcoming in Hawk B., ed. (2005): 2004 Annual Proceedings of the Fordham Corporate Law Institute—International Antitrust Law and Policy, Juris Publishing, New York. The European Commission has to my knowledge never taken a position on this matter. 73 Paragraph 10.16 of the United Kingdom Department of Trade and Industry’s Consultation on Modernisation (April 2003), available at http://www.dti.gov.uk/ccp/consultpdf/ compmodcon.pdf. 74 Text accompanying notes 61 to 67; see also text accompanying note 189 below. 75 Whereas the Italian, Spanish and Dutch versions of this sentence use the equivalent of the words ‘the means’ in the English version (gli strumenti, el medio, het middel ), the French and German versions rather use the equivalent of ‘a means’ (un moyen, Mittel).


286

12/10/07

13:10

Page 286


criminal sanctions on natural persons.76 I have difficulty seeing how one could consider that the UK cartel offence predominantly pursues an objective different from that pursued by Article 81. Indeed, it is perfectly clear from the history of the Enterprise Act 2002 that the cartel offence was introduced in UK law because it was considered that the existing fines for companies were insufficient to deter hard core cartels prohibited by Chapter I of the Competition Act 1998 and Article 81 EC, and that imprisonment was the most effective additional deterrent.77 The cartel offence in the UK should thus in my view be considered a means whereby competition rules applying to undertakings are enforced.

4. Decriminalization and criminalization of antitrust enforcement in the EU Member States Looking at antitrust enforcement in the EU Member States over the last 20 years, one can detect a double tendency of decriminalization and criminalization.78

4.1. Decriminalization on the one hand The tendency to decriminalize antitrust enforcement is illustrated by the examples of France, The Netherlands, Austria and Luxembourg. In France, antitrust enforcement under the competition acts of 1953 and 1958 was exclusively criminal. After a first reform in 1977, the competition act of 1986 introduced the current French system, which relies essentially on administrative fines for companies, imposed by the Competition Council, a specialized administrative authority. Criminal sanctions, including imprisonment, are however still provided for in French law. In The Netherlands, the current competition act, which came into effect in 1997, replaced an earlier system of criminal enforcement by a purely administrative system with only fines for undertakings, comparable to the enforcement system at the level of the EU institutions. In Austria, the 2002 competition law reform also replaced a system of criminal enforcement by a system of administrative fines for companies. A specific criminal offence of bid rigging was however maintained. In Luxembourg, the 2004 competition act abolished an earlier competition 76

See Principles of European Antitrust Enforcement, supra note 2, section 1.2.8.2. See, inter alia, Department of Trade and Industry, ‘A World Class Competition Regime’, presented to Parliament by the Secretary of State for Trade and Industry, July 2001, available at http://www.archive.official-documents.co.uk/document/cm52/5233/523301.htm. 78 See also text accompanying notes 63 to 67 above as to the current situation in various Member States. 77


12/10/07

13:10

Page 287


287

regime with criminal enforcement by a new regime with only administrative fines that may be imposed on companies. These movements to decriminalize antitrust enforcement reflect, I believe, two phenomena. The first explanation is that some of the older competition laws with criminal enforcement which were abolished in these Member States were in fact not comparable to Articles 81 and 82 EC, because they were not based on the prohibition principle. In Luxembourg, for instance, before the 2004 reform, there were in fact no sanctions at all for the mere fact of having concluded and operated a price cartel. Only if a cartel had been investigated, and only if the undertakings had been ordered to put an end to it, were the owners, managers and directors of those undertakings threatened with criminal sanctions if they continued the cartel notwithstanding the order. Luxembourg can thus hardly be considered to be leading a new trend towards decriminalization. Rather, it was the last to follow an earlier trend started by the European Communities and Germany in the 1950s. Indeed, looking at the whole of Europe over close to a century, specifically with regard to cartels, it would seem that most jurisdictions are historically moving through similar phases in the same direction. In a first phase, the one which Luxembourg only abandoned in 2004, there was no real cartel prohibition, only a system of control of abuse under which, once an abuse had been established, the undertakings concerned were ordered to put an end to this abuse for the future. Non-respect of this specific order for the future may have been punishable with criminal sanctions, including imprisonment. In a second phase, first reached by the European Coal and Steel Community in the early 1950s, and subsequently by Germany and by the European Economic Community in the late 1950s, cartels were prohibited, but the direct effect of this prohibition was softened through a notification system. The notification system allowed temporary immunity from sanctions,79 and violations of the prohibition were only punishable by fines imposed on undertakings. In a third phase, introduced at the EU level by Regulation No. 1/2003, the deterrent effect of the cartel prohibition is enhanced through the abolition of the notification system. In a fourth phase, reached long ago by the United States,80 and recently entered into by the United Kingdom,81 the threat of individual sanctions, in particular imprisonment, is added, so as to increase deterrence for those types of cartel infringements which are most profitable and most difficult to detect, namely secret price fixing, bid rigging and market sharing arrangements. 79 See Principles of European Antitrust Enforcement, supra note 2, section 1.1.4.4, and Marenco G. (2001): “Does a Legal Exception System Require an Amendment of the Treaty?”, in Ehlermann C.-D. and Atanasiu I., eds., European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, London and Portland, Oregon, 145–184, at 174–175. 80 See text accompanying notes 22 to 31 above. 81 See text accompanying notes 71 above and 84 below.


288

12/10/07

13:10

Page 288


The second explanation for the observed decriminalization in some EU Member States over the past 20 years is that, although some of these Member States had already moved to a system based on the prohibition principle, their criminal enforcement did not work in practice because it was badly conceived. Indeed, as will be argued further in this paper,82 criminal antitrust enforcement, in its strongest form, relying on imprisonment, can only work or only makes sense if: it is limited to hard core cartels, complementing other penalties for companies and for lesser antitrust violations; a dedicated investigator and prosecutor and sufficient powers of investigation are provided for; and judges or juries are willing to convict, all of which requires a sufficiently broad political or societal consensus that cartels are really bad and thus deserve severe punishment. These conditions do not appear to have been fulfilled in those Member States at the relevant time.83

4.2. Criminalization on the other hand The tendency to criminalize antitrust enforcement is primarily illustrated by the examples of Ireland and the United Kingdom, but also by other phenomena, such as the introduction of leniency programmes in an increasing number of EU Member States.84 In Ireland, since the entry into force of the Competition Act 1996, later replaced by the Competition Act 2002, violations of the national law prohibitions equivalent to Articles 81 and 82 EC, as well as violations of Articles 81 and 82 themselves, are criminal offences, punishable by fines for companies, and imprisonment and/or fines for individuals. A comparable system has been adopted in Estonia. In the United Kingdom, as already mentioned above,85 the Enterprise Act 2002 has added a criminal cartel offence, limited to hard core cartels, punishable by imprisonment and/or fines for individuals only, as well as a sanction of director disqualification for individuals,86 to a system which otherwise relies on administrative fines imposed on companies for breach of Articles 81 and 82 EC and the equivalent prohibitions under national law. The introduction of criminal penalties in Ireland and the United Kingdom has been accompanied by the grant of corresponding criminal powers of investigation to the investigating authorities. In the United Kingdom, for instance, the Office of Fair Trading has been given new powers of covert surveillance for the purpose of investigating the criminal cartel offence.87 The 82

Text accompanying notes 155 to 174. See further text accompanying note 155 below. 84 See also text accompanying notes 53 to 55 above. 85 Text accompanying note 71. 86 See further text accompanying note 150 below on director disqualification. 87 Sections 199–200 of the Enterprise Act 2002, supra note 71; see also Office of Fair Trading, “Powers for investigating criminal cartels—Guidance” (January 2004), available at http://www. oft.gov.uk/Business/Legal+Powers/Enterprise+Act/publications.htm. 83


12/10/07

13:10

Page 289


289

Irish and United Kingdom competition authorities also operate leniency policies similar to those of the US Department of Justice.88 As part of a wider trend toward criminalizing antitrust enforcement, it should also be mentioned that an increasing number of Member States, including France, Germany, The Netherlands and Finland, have followed the European Commission in operating a leniency programme for undertakings that help the competition authority to detect cartels.89 As mentioned above,90 such leniency programmes, where known at all in the EU Member States, traditionally were typically reserved for the most serious forms of organized crime. Similarly to what was pointed out above with regard to antitrust enforcement at the level of the EU institutions, it would also appear that in those EU Member States where antitrust enforcement remains based on administrative fines for companies, there is a more general, broad tendency both to strengthen the powers of investigation of the competition authorities and to strengthen the defendant companies’ rights of defence in antitrust proceedings, even if either level generally remains below the level observed in fully criminalized systems. On the specific issue of the separation of investigative, prosecutorial and adjudicative functions, several Member States, including France, Belgium, Sweden and Finland, have separated these functions to varying degrees, even where they only impose fines of an administrative law nature.91 The introduction of criminal penalties in Ireland and the United Kingdom, as well as the introduction of leniency programmes at the level of the EU institutions and in many Member States, has clearly been inspired by the US example.92 The fact that Europe follows the US example in matters of antitrust is of course not new. Indeed, the history of antitrust in modern Europe,93 and in particular the history of Article 81 and similar provisions in national law, is very much one of American influence. The prohibition on abuse of a dominant position, first in the ECSC Treaty of 1951 and subsequently in Article 86 of the EEC Treaty of 1957 (now Article 82), was not without precedent: in Germany, for instance, an ‘Ordinance against the 88 See text accompanying notes 38 and 39 above, and, as to the United Kingdom: Office of Fair Trading, “The cartel offence—Guidance on the issue of no-action letters for individuals” (April 2003), available at http://www.oft.gov.uk/Business/Legal+Powers/Enterprise+Act/ publications.htm. 89 At the time of writing, already 18 of the 25 Member States have a leniency programme; see the updated list on the website of the European Commission’s Directorate-General for Competition: http://europa.eu.int/comm/competition/antitrust/legislation/authorities_with_ leniency_programme.pdf. 90 Text accompanying notes 53 to 55. 91 See Principles of European Antitrust Enforcement, supra note 2, section 1.2.10. 92 See Kunzlik P. (2003): “Globalization and Hybridization in Antitrust Enforcement: European ‘Borrowings from the U.S. Approach’ “, The Antitrust Bulletin 319; and text accompanying notes 20 and 21 above. 93 See supra note 21 as to older European precedents.


290

12/10/07

13:10

Page 290


Abuse of Economic Power’ had already been enacted in 1923.94 However, the prohibition on restrictive agreements first laid down in Article 65 of the ECSC Treaty and then also in Article 85 of the EEC Treaty (now Article 81 EC), was “a fundamental innovation in Europe”.95 Before the Second World War, cartels were a widespread and highly esteemed institution throughout Europe.96 The insertion of the prohibition on restrictive agreements in European law, as well as around the same time in German national law, was due to American influence, if not pressure.97 As Joel Davidow has recently observed: “In the early years of the 21st Century, it has become evident that US antitrust, once nearly the sole source of influential antitrust enforcement, has become part of a “dynamic duo”, along with the European Union. The EU and its Member States have eclipsed the United States in terms of number of professionals employed, sometimes in size of fines and certainly in terms of countries imitating the basic enforcement scheme.”98 However: “The United States, even if no longer the world’s largest jurisdiction, remains the most influential in terms of policy and theory.”99 The trend to criminalize antitrust enforcement reflects this influence.

C. Is criminalization of antitrust enforcement desirable? In this third part I want to address the question of whether, why or under what conditions criminalization of antitrust enforcement is desirable. The question is addressed in general, irrespective of whether such criminalization takes place at the level of the EU Member States or of the EU institutions, or

94 Nörr K. (1995): “Law and Market Organization: The Historical Experience in Germany From 1900 to the Law Against Restraints of Competition (1957)”, 151 Journal of Institutional and Theoretical Economics / Zeitschrift für die gesamte Staatswissenschaft 5, at p. 10. 95 Monnet J. (1976): Mémoires, Fayard Paris, at p. 413. But see supra note 21 regarding the older European precedents. 96 Schröter H. (1996): “Cartelization and Decartelization in Europe, 1870–1995: Rise and Decline of an Economic Institution”, 25 Journal of European Economic History 129, at p. 137, who also writes at 140 that Yugoslavia was the only European country where cartels were prohibited at that time. 97 See J. Monnet, supra note 95, at 356–357 and 411–413, Spierenburg D. and Poidevin R. (1994): The History of the High Authority of the European Coal and Steel Community, Weidenfeld and Nicholson, at pp. 26–28, and, with a detailed description of the strong resistance to be overcome in Germany, Berghahn V. (1986): The Americanization of West German Industry 1945–1973, Berg. See also Marenco G. (2002): “The Birth of Modern Competition Law in Europe”, in von Bogdandy A., Mavroidis P. and Mény Y., eds, European Integration and International Co-ordination—Studies in Transnational Economic Law in Honour of Claus-Dieter Ehlermann, Kluwer Law International, London, 279–304, and The Optimal Enforcement of EC Antitrust Law, supra note 4, sections 5.3.2 and 9.5. 98 Davidow J., supra note 32, at 407. 99 Ibid.


12/10/07

13:10

Page 291


291

whether it is harmonized at EU level. The question of whether it is problematic that antitrust enforcement is criminalized at the level of individual EU Member States without parallel criminalization at the level of the EU institutions or without EU harmonization will be dealt with afterwards in part IV of this paper. As described above,100 criminalization is a phenomenon with different faces, including for instance the spread of leniency programmes in jurisdictions which continue relying exclusively on administrative fines for companies. However, I will focus on its strongest variant, namely the introduction of individual penalties, and more specifically imprisonment, as exemplified by Ireland and the United Kingdom. I will first consider five arguments in favour of imprisonment as a penalty for antitrust offences. Then I will consider the types of antitrust offences for which these arguments are valid. Next, I will examine whether there are no equally effective alternatives to imprisonment, and finally I will discuss five conditions which need to be fulfilled for criminal antitrust enforcement to be effective not only in theory but also in practice.

1. Five arguments in favour of imprisonment 1.1. Effective deterrence limited to fines for companies would require impossibly high fines In a paper presented at the 6th Competition Law and Policy Workshop at the European University Institute (EUI) here in 2001, I argued that the minimum level of fines required generally to deter price cartels and other antitrust offences of comparable profitability and ease of concealment would be on the order of 150% of the annual turnover in the products concerned by the violation.101 This argument was based on the following hypothetical: I took the figure of 10% of the selling price as an estimate of the average price increase from price fixing. This estimate had been relied upon by the US Sentencing Commission when drafting its Sentencing Guidelines and appeared generally accepted in the American literature, no European studies being available. As

100

Text accompanying notes 53 to 60 and 89 to 91. Wils W. P. J. (2003): “Does the Effective Enforcement of Articles 81 and 82 EC Require Not Only Fines on Undertakings But Also Individual Penalties, In Particular Imprisonment?’, paper presented at the 6th Competition Law and Policy Workshop at the European University Institute (Florence, 1–2 June 2001), later published 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, 411–452, and in The Optimal Enforcement of EC Antitrust Law, supra note 4, chapter 8, 188–237. 101


292

12/10/07

13:10

Page 292


the price increase caused by the cartel will normally depress demand for the cartel members’ products, I assumed conservatively that a price increase of 10% would lead to an increase in profits of 5% of turnover. On the basis of the findings of duration in a number of cartel decisions of the European Commission, and of estimates in the American literature, I assumed, again conservatively, a cartel duration of 5 years. Finally, I assumed a probability of detection and punishment of 16%, which I considered again to be a conservative estimate, given that the one existing study in the American literature, by Peter Bryant and Woodrow Eckard, based on a sample of 184 price-fixing cases for the period from 1961 to 1988, had produced an estimate of between 13 and 17%,102 and given that European competition authorities have weaker investigative powers than their American counterparts. Assuming a 10% price increase, and a resulting increase in profits of 5% of turnover, a 5-year duration and a 16% probability of detection and punishment, the floor below which fines will generally not deter price fixing would be on the order of 150% of the annual turnover in the products concerned by the violation.103 A slight variation on this hypothetical was subsequently used by the UK government in its White Paper setting out its case for the introduction of a criminal cartel offence.104 The hypothetical was also referred to by the OECD in its 2002 report on hard core cartels.105 Five years later, would these estimates still appear appropriate? As to the average price increase from price fixing, the OECD reported in 2002 the results of a survey it had carried out. In the limited data (14 cases) of that survey, the median overcharge was between 15 and 20%.106 Most recently, John Connor has made a survey of hundreds of published social-science studies of hard core cartels that contained 674 observations of long-run overcharges. His primary finding is that the median cartel overcharge for all types of cartels over all time periods throughout the world is 25%. He also conducted a survey of 24 final verdicts in decided US horizontal collusion cases, revealing an average median overcharge of 21% and an average mean overcharge of 30%.107 On the basis of these new surveys, I should have taken 20% rather than 10% as an estimate of the average price increase from price fixing in my 102 P.G. Bryant and E.W. Eckhard, ‘Price Fixing: The Probability of Getting Caught’ (1991) Review of Economics and Statistics 531. 103 I added that the figure would be even higher if one were to take into account the fact that fines are only paid several years after the gain from the violation is obtained. Indeed, my calculations assumed a zero interest rate. With a more realistic higher interest rate, the minimum fine necessary to achieve deterrence would be even higher. 104 ‘A World Class Competition Regime’, supra note 77, at 40 (box 7.3). 105 OECD (2002): Report on the nature and impact of hard core cartels and sanctions against cartels under national competition laws, DAFFE/COMP(2002)7 (9 April 2002), available at http://www.oecd.org, at 19 (note 20). 106 OECD, cited in previous footnote, at 9. 107 Connor J., supra note 33.


12/10/07

13:10

Page 293


293

hypothetical. Keeping the same conservative assumption as to the price elasticity of demand for the cartel members’ products, I would then take 10% of turnover as an estimate of the increase in profits resulting from the price fixing. As to the average duration of price-fixing cartels, the literature published in the last five years would not alter my conservative estimate of 5 years.108 As to the probability of detection and punishment, it has correctly been pointed out that the detection rate in the US has presumably improved since the period 1961 to 1988 on which Peter Bryant and Woodrow Eckard based their estimate of between 13 and 17%,109 because of the use of more effective investigative methods, in particular the use of optimized leniency programmes.110 Some have opined that, nevertheless, “there is little reason to believe that the true probability of detection is outside this range”.111 It would in any event still remain true that the probability is probably lower in Europe than in the US.112 I believe I would remain very safely on the conservative side by assuming the probability of detection and punishment in Europe not to be above 33%. Assuming then a 20% price increase, and a resulting increase in profits of 10% of turnover, a 5-year duration and a 33% probability of detection and punishment, I would thus still today reach the conclusion that the floor below which fines for companies will generally not deter price fixing would be on the order of 150% of the annual turnover in the products concerned by the violation.113 108 See Evenett S., Levenstein M. and Suslow V. (2001): “International Cartel Enforcement: Lessons from the 1990s”, World Economy 1221, at p. 1226; and Connor J., supra note 33, at 12. 109 Supra note 102. 110 See OECD, supra note 105, at 13; and Bush D., Connor J., Flynn J., Ghosh S., Grimes W., Harrington J., Hawker N., Lande R, Shepherd W. and Semerano S. (2004): “How to Block Cartel Formation and Price-Fixing”, amicus curiae brief in the US Supreme Court in HoffmanLaRoche v Empagran, 15 March 2004, available at http://aei-brookings.org/admin/authorpdfs/ page.php?id=931, at 28; the point was first made to me shortly after I presented my 2001 EUI paper by Bill Bishop. See also Principles of European Antitrust Enforcement, supra note 2, chapter 5. 111 Bush D. et al., supra note 110, at 28. 112 See also OECD, supra note 105, at 13. 113 I would again add that the figure would be even higher if one were to take into account the fact that fines are only paid several years after the gain from the violation is obtained. See supra note 103. I would further add that the figure is probably also too low because it ignores the likely effects of overconfidence bias. Indeed, it appears that people consistently tend to overestimate the probability of good things happening to them, and that they underestimate the probability of bad things happening to them, in particular when the event is question is perceived to be controllable. See Weinstein N. (1989): “Optimistic Biases About Personal Risks”, 246 Science 1232, and Korobkin R. and Ulen T. (2000): “Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics”, 88 California Law Review 1051, at pp. 1091–1095. Applied to deterrence through antitrust fines, this suggests that companies and individual decision-makers in those companies will tend to overestimate the gains from the antitrust violation and to underestimate the probability of being caught. See also Denger M., “Too Much or Too Little”, American Bar Association, Section of Antitrust Law, Antitrust Remedies Forum, available at http://www.abanet.org/antitrust/remedies, at 5: “most cartel participants don’t think they will get caught”.


294

12/10/07

13:10

Page 294


As I argued in more detail in my 2001 EUI paper,114 raising the general level of fines to such a high level would be impossible or unacceptable for various reasons. First (and least importantly, as statutory ceilings could be raised), such higher fines are likely regularly to breach the statutory ceilings on the amount of fines which can legally be imposed.115 Secondly, and most importantly, such high fines are likely often to exceed the undertakings’ ability to pay, with as a result not only incomplete deterrence but also costly side-effects. The fact that the companies that have engaged in a price-fixing cartel have profited financially from the violation offers no guarantee that they will also be able to pay the minimum fine required for effective deterrence. Indeed, in the estimated average case of a price cartel leading to extra profits of 10% of turnover for a period of 5 years, with a probability of detection and punishment of 33%, the cartel members would each have gained over the whole duration of the cartel 50% of their annual turnover in the products concerned by the violation. Even if they retained these profits until the fine were imposed, these profits would only pay for one third of the fine of 150% of annual turnover in the products concerned. It is more likely that these profits would not be retained but rather would have been paid out in taxes, dividends, salaries and wages.116 Even liquidating the assets of the firms concerned is often unlikely to generate enough revenues to pay the fine required for effective deterrence. Indeed, many companies’ annual turnover exceeds their assets. A fine of 150% of annual turnover in the products concerned by the violation could thus only be paid out of the assets of large, diversified companies which have taken part in cartels only for a small part of their activities,117 or companies with very high asset-to-sales ratios.118 If such high fines were really imposed, many of the companies concerned would be forced into bankruptcy.119 However, this would entail undesirable social costs because, in the absence of perfect markets, it would hurt not only 114

Supra note 101. For example, with respect to fines imposed by the European Commission, the statutory ceiling is 10% of the total worldwide turnover of the undertaking concerned. See Article 23(2) of Regulation No. 1/2003, supra note 1. 116 Werden G. and Simon M. (1987): “Why Price Fixers Should Go to Prison”, The Antitrust Bulletin 917, at p. 928 note 35, referring to empirical estimates that unions are able to capture most of the monopoly profits earned by US manufacturing firms. 117 When looking at the names of companies found to have been involved in cartels, one often sees the same names reappearing, suggesting that companies found to have been involved in a cartel in one market may have been active in cartels in their other markets as well. 118 Werden G. and Simon M., supra note 116, at 928–929. 119 Craycraft C., Craycraft J. and Gallo J. (1997): “Antitrust Sanctions and a Firm’s Ability to Pay”, 12 Review of Industrial Organization 171, in an empirical study based on a sample of 386 firms convicted of price fixing in the US between 1955 and 1993, have estimated that 58% of the firms would not have been able to survive the imposition of an optimal fine without becoming technically bankrupt. Because of data restrictions, the firms in their sample were the larger firms, which means that their result may very well underestimate the extent of the problem of inability to pay. 115


12/10/07

13:10

Page 295


295

managers and shareholders, on whom the bankruptcy may be considered to have a desirable deterrent effect, but also all other stakeholders in the firm: employees, suppliers, customers, creditors and tax authorities.120 Even if they remain below the level of inability to pay, the imposition of such high fines is likely to be costly. Very high fines may have undesirable side-effects. In the absence of perfect markets, very high fines imposed on companies will have an incidence on all the stakeholders in the firm. Bondholders and other creditors will suffer a diminution in the value of their securities. Employees may suffer from cost-cutting campaigns induced by the need to pay the fine. Tax receipts will be reduced. Finally, consumers may end up suffering. Indeed, if the firm competes in a product market characterized by imperfect competition (as will often be the case), the fine may be partly recovered from consumers in the form of higher prices.121

1.2. Fines for companies do not always guarantee adequate incentives for responsible individuals within the firm As I also argued in more detail in my 2001 EUI paper,122 the logic of fining undertakings or companies for antitrust violations is that the threat of the corporate sanction, if sufficiently high, will induce the undertaking or company to try to make its agents respect the law. This shifting of the enforcement function from the authorities to the firm makes sense because in general the firm has a relatively good ability to influence the behaviour of its agents. However, in a number of circumstances the firm may not, in reality, be capable of adequately controlling the behaviour of its agents, with the result that exclusive reliance on corporate sanctions will not lead to effective deterrence. This situation can arise because firms may not be able to impose sufficiently serious sanctions on their agents. It can also arise when firms are controlled by management. Finally, the managers responsible for the violation may have left the firm by the time the violation is detected, and there may be little a company can do to a former employee.

1.3. Increased effectiveness of leniency and whistle blowing programmes Sanctions against individuals have another beneficial effect, in that they provide an incentive for individuals to offer to co-operate in cartel investigations against the interests of their employers.123 120 Kraakman R. (1984): ”Corporate Liability Strategies and the Costs of Legal Controls”, 93 The Yale Law Journal 857, at 882. 121 Coffee J. (1981): “ ‘No Soul to Damn: No Body to Kick’: An Unscandalized Inquiry into the Problem of Corporate Punishment”, 79 Michigan Law Review 387, at pp. 401–402. 122 Supra note 101. 123 OECD, supra note 105, at 16.


296

12/10/07

13:10

Page 296


As Scott Hammond has explained with regard to the US Department of Justice’s practice of combining corporate and individual leniency programmes: “[T]he individual amnesty program helps prevent companies from covering up their misconduct. The real value and measure of the Individual Leniency Program is not in the number of individual applications we receive, but in the number of corporate applications it generates. It works because it acts as a watchdog to ensure that companies report the conduct themselves. [. . .] So long as one of its employees has individual exposure, the company remains at great risk. If the company self-reports the conduct under the Corporate Leniency Policy, then the company and all of its cooperating executives will avoid criminal prosecution. However, if the company delays or decides not to report, then the company puts itself in a race for leniency with its own employees. In this example, if the company does not report the conduct first, then the executive may come forward on his own and report the conduct for his own protection, thereby potentially leaving the company out in the cold. [. . .] If the secretary gets nervous, say, after talking to a relative who convinces her that she has real criminal exposure for her own conduct, she may decide to report the conduct.”124

As Donald Baker has further explained, also describing US practice: “The same process works between individuals at different companies involved in the suspected conspiracy. There certainly have been cases where an individual was indicted for a one-on-one conversation with someone who worked for a competitor and who then turned the defendant in. Of course, as we saw with the Lysine cartel, an individual who turns herself in to the government may become a government informant and continue to participate in conspiracy meetings, armed with recording devices in order to entrap the other conspiracy members.”125

Donald Baker has also pointed out that the system of imposing criminal liability on individuals generates not only incentives for whistle blowing based on fear, but also incentives for whistle blowing based on the desire for revenge: “The desire for revenge is more picturesque, but it is still very much present. In my experience, disgruntled current employees, fired employees, former trade association officials, and even ex-spouses and ex-lovers may be anxious to finger the individuals who they think have done them in.”126 As the OECD has pointed out: “There is an offsetting consideration relating to individual sanctions, however. In many countries, sanctioning individuals requires criminalisation of the conduct. In that context individuals have rights against self-incrimination, which makes it more difficult to obtain evidence from them unless they willingly co-operate. In this sense,

124 Hammond S. (2004): “Cornerstones of an Effective Leniency Program”, paper presented before the ICN Workshop on Leniency Programs (Sydney, 22–23 November 2004), available at http://www.usdoj.gov/atr/public/speeches/206611.htm, at 12–14. 125 Baker D., supra note 22, at 708. 126 Ibid.


12/10/07

13:10

Page 297


297

criminalisation of cartel conduct for individuals makes voluntary co-operation more possible, because of the threat of personal sanctions, but also makes it more necessary, because a right against self-incrimination attaches.”127

The US experience,128 though, suggests that the net result is an increase in the effectiveness of investigations. As will be discussed further in this paper,129 replication of the successful US experience in Europe would however depend on the ability of the competition authorities concerned to make similar use of leniency instruments, as is done by the US Department of Justice.

1.4. Imprisonment is a very effective deterrent There is ample evidence, based in particular on the US experience, that imprisonment is a very effective deterrent for potential antitrust offenders. In Arthur Liman’s words: “For the purse snatcher, a term in the penitentiary may be little more unsettling than basic training in the army. To the businessman, however, prison is the inferno, and conventional risk-reward analysis breaks down when the risk is jail. The threat of imprisonment, therefore, remains the most meaningful deterrent to antitrust violations.”130 Or as Joseph Bauer has more recently put it: “The sight of A. Alfred Taubman, the extremely wealthy chairman of the board of Sotheby’s, the world-famous auction house, convicted and sentenced, at the age of 78, to a one-year term of imprisonment and a substantial fine for participating in a price-fixing conspiracy, doubtless sent a message to other business executives about the risks and penalties for this kind of behaviour”.131 The success of the current criminal antitrust enforcement in the US is shown by several indicators. First, according to Terry Calvani, in the US “there is much less domestic cartel activity today than before”.132 The majority of 127

OECD, supra note 105, at 16. See Baker D., supra note 22, and S.D. Hammond, supra note 124; text accompanying notes 156 to 161 below. 129 Text accompanying notes 156 to 161 below. 130 Liman A. (1977): “The Paper Label Sentences: Critique”, 86 The Yale Law Journal 619, at pp. 630–631. In the same sense, see Baker D. and Reeves B. (1977): “The Paper Label Sentences: Critique”, 86 The Yale Law Journal 619, at p. 621; Mukasey M. (1997): “Comment on Lynch”, 60 Law and Contemporary Problems 71. See also Antoine Winckler’s remarks in the panel discussion on criminal sanctions at the 6th Competition Law and Policy Workshop at the European University Institute (Florence, 1–2 June 2001), published in Ehlermann C.-D. and Atanasiu I. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, 387, at p. 404: “My experience as a legal practitioner is that criminal sanctions are an extremely effective deterrent.” 131 Bauer J. (2004): “Reflections on the Manifold Means of Enforcing the Antitrust Laws: Too Much, Too Little, or Just Right?”, 16 Loyola Consumer Law Review 303, at 307. 132 Calvani T. (2004): “Enforcement of Cartel Law in Ireland”, in Hawk B., ed., 2003 Annual Proceedings of the Fordham Corporate Law Institute International Antitrust Law & Policy, Juris Publishing, New York, 1, at 6. 128


298

12/10/07

13:10

Page 298


cartels detected and prosecuted by the US Department of Justice in the last years have been international, often even global in membership and geographic spread.133 Second, as Scott Hammond has pointed out: “We [the US Department of Justice] are observing firsthand in some of our investigations how the threat of criminal prosecution in the United States has deterred a significant number of global cartels from extending their conspiracy into the United States. We have uncovered cartels that operated profitably and illegally in Europe, Asia, and elsewhere around the world, but did not expand their cartel activity to the United States solely because it was not worth the risk of U.S. sanctions. I am referring to cartels that had every opportunity to target U.S. consumers, because they sold in the U.S. market. Indeed, in some cases, the U.S. market was the largest and potentially most profitable but the collusive conduct still ceased at the border. Why? The answer, from the mouths of the cartel members and verified by our investigators, is that the executives did not want to risk getting caught and going to jail in the United States.”134

Third, as William Kolasky has indicated with regard to those international cartels that nevertheless continue to target US consumers: “The power of criminal enforcement is illustrated by another characteristic common to those cartels in which the conspirators know that their conduct is unlawful—namely, their efforts to stay away from the United States. International cartels often try to minimize their contacts in the United States by conducting their meetings abroad. This has been particularly true since 1995, when the lysine investigation became public. In fact, cooperating defendants in several recent cases have revealed that the cartels changed their practices and began avoiding contacts in the United States at all costs once the [US Department of Justice Antitrust] Division began cracking and prosecuting international cartels. Some cartel members go so far as to try to keep their cartel activity secret from all U.S.-based employees, even those responsible for carrying out their instructions as to the firm’s output and prices.”135

Fourth, as Margaret Bloom has pointed out: “Another factor that illustrates the effect of criminal powers is the extent to which international cartels are first uncovered through amnesty applications to the [US Department of Justice]. In these cases, it is the threat of criminal sanctions that drives executives to go first to the [US Department of Justice]—and only later to the European Commission and other relevant authorities. Without the U.S. criminal sanctions and active enforcement record, how many of these cases would be revealed by leniency applications to the European Commission? It is, of course, impossible to answer this question, but I suspect the answer might well be few.”136 133

See Connor J., supra note 33, at 10; Hammond S., supra note 32, at 2. Hammond S., supra note 124, at 8–9. 135 Kolasky W. (2004): “Criminalizing Cartel Activity: Lessons from the U.S. Experience”, fifth lunch talk of the Global Competition Law Centre, Brussels, 29 September 2004, available at http://www.gclc.coleurop.be/lunchtalk20040929.htm, at 11–12. 136 Bloom M. (2005): “The Great Reformer: Mario Monti’s Legacy in Article 81 and Cartel Policy”, 1 Competition Policy International 55. 134


12/10/07

13:10

Page 299


299

1.5. Imprisonment carries a uniquely strong moral message Corporate managers are not necessarily just maximizers of profits for themselves and their principals. They may feel a moral responsibility to live within the law whether or not they are likely to be caught, and this normative commitment could trump their interest calculus.137 Indeed, psychological research suggests that normative commitment is generally an important factor explaining compliance with the law.138 The public punishment of those who violate the antitrust prohibitions thus has not only a deterrent effect, in that it helps to create a credible threat of punishment for those who would be willing to commit violations on the basis of a profit calculation, but also a moral effect, in that it sends a message to the spontaneously law-abiding, reinforcing their moral commitment to the rules.139 Imprisonment is by far the most expressive sanction. Prison sanctions not only send a special message not conveyed by fines, they also send it more effectively, as prison sentences for businesspeople are much more newsworthy than fines and will thus ensure more publicity and will be more noted by other businesspeople.140

2. For which types of antitrust violations is imprisonment desirable? The above argumentation that effective deterrence limited to fines imposed on companies would require impossibly high fines, and that such fines should be supplemented by imprisonment to achieve effective deterrence, is based on the paradigmatic case of a price cartel.141 The conclusions reached only apply to those types of violations of Articles 81 and 82 EC which are comparable to price cartels as to their profitability and ease of concealment. For those types of infringements which are difficult to hide, or which are anyway not very profitable for the undertakings committing them, fines for companies, or fines for companies combined with director disqualification,142 could be sufficient to achieve effective deterrence. Secondly, because imprisonment is a much more onerous punishment than fines, the costs of imposing it erroneously are all the greater. These costs include not only the suffering of those unjustly punished, but also the risk-bearing cost 137 Stone C. (1991): “Sentencing the Corporation”, 71 Boston University Law Review 383, at p. 389. 138 See Tyler T. (1990): Why People Obey the Law, Yale University Press. 139 See supra note 6. 140 Werden G. and Simon M., supra note 116, at 934; Lynch G., supra note 6, at pp. 47 and 49; Kahan D., supra note 6, at 383–384. See also text accompanying notes 130 and 131 above. 141 See text accompanying note 101 and following above. 142 See text accompanying notes 150 to 152 below on director disqualification.


300

12/10/07

13:10

Page 300


to those risking unjustified punishment, the risk of lawful behaviour being undesirably deterred, as well as possible negative effects on other people’s normative commitment to the law.143 Prison sanctions should therefore only be imposed in the case of clear-cut violations. Alternatively, one could conclude that imprisonment should only be imposed for wilful violations of the law, where there is evidence that the individuals knew that they were violating the law or acted with flagrant disregard for the law, with wilfulness being presumed in cases of clear-cut violations.144 Applying these two sets of criteria, it would seem that imprisonment is desirable for horizontal, naked price fixing, bid rigging and market allocation schemes. As to all other types of antitrust violations, including vertical restraints, other horizontal agreements and violations of Article 82, it would seem that they are usually easier to detect, as the victims of the violation will very often be aware of it. Furthermore, the risk of undesirably chilling lawful behaviour would appear to be much more substantial, as the borderline between anticompetitive and pro-competitive behaviour is often less obvious. For those other types of violations, fines for undertakings, or fines for undertakings combined with director disqualification, may suffice.145

3. Are there no equally effective alternatives to imprisonment? The above argumentation in favour of imprisonment consists in saying, on the one hand, that hard core cartels cannot be deterred by imposing only fines on companies, and, on the other hand, that imprisonment constitutes an effective deterrent. But could effective deterrence not also be reached by adding other types of sanctions or remedies to fines for companies?

3.1. Fines for individuals Fines for individuals would not appear to be an equally effective alternative to imprisonment. The main reason146 is that companies can relatively easily indemnify their agents for any threat of fines or any fines effectively imposed, thus removing the deterrent effect of the penalty on the individuals con143 See also Kraakman R., supra note 120, at 865 and 887; Blair R. (1985): “A Suggestion for Improved Antitrust Enforcement”, The Antitrust Bulletin 433, at p. 444 footnote 32. 144 See D.I. Baker and B.A. Reeves, supra note 130, at 623–624. 145 The United Kingdom would thus in my view appear to have struck the right balance in providing for a criminal cartel offence added to fines for undertakings and director disqualification for all types of antitrust violations. See text accompanying notes 71 and 85 and 86 above and text accompanying notes 171 to 174 below. 146 As explained in my 2001 EUI paper, supra note 101, another reason is that, just as for fines imposed on companies, the level of fines required for deterrence risks being impossibly high.


12/10/07

13:10

Page 301


301

cerned. Companies can relatively easily compensate their agents in advance for taking the risk of being fined and/or indemnify them ex post when they have to pay the fine. The crucial advantage of imprisonment is that it is impossible to shift the penalty ex post, and it is more difficult to arrange for a premium to compensate the risk in advance.147 As Donald Baker has reported a senior corporate executive telling him: “as long as you are only talking about money, the company can at the end of the day take care of me – but once you begin talking about taking away my liberty, there is nothing that the company can do for me”.148 Fines for individuals may nevertheless have at least some stigmatizing effect. They could thus be considered as a third-best option, to be used when the more effective options of imprisonment and director disqualification are unavailable.149

3.2. Director disqualification As already mentioned above, in the UK, the Enterprise Act 2002 introduced not only the criminal cartel offence but also Competition Disqualification Orders (CDOs).150 On application from the Office of Fair Trading, the competent court must issue a CDO against a director of a company (including a de facto director) if that company commits a breach of competition law (Articles 81 or 82 EC or the corresponding prohibitions in UK competition law) and if the court considers that the person’s conduct as a director makes him or her unfit to be concerned in the management of a company. The maximum period of disqualification under a CDO is 15 years. During the period in which a person is subject to a CDO, it is a criminal offence for him or her to be a director of a company or, in any way, whether directly or indirectly, to be concerned or take part in the management of a company. Director disqualification does not appear to be as strong a deterrent as imprisonment. Depending on the age of the person concerned, director 147 See Werden G. and Simon M., supra note 116, at 931; Blair R., supra note 143, at 440–441; Kraakman R., supra note 120, at 876–879; Stone C. (1980): “The Place of Enterprise Liability in the Control of Corporate Conduct”, 90 The Yale Law Journal 1, at pp. 45–65. 148 Baker D., supra note 22, at 705. 149 Specifically with regard to the enforcement of Articles 81 and 82 at the level of the EU institutions, fines on individuals would have an advantage in that they could uncontestedly be introduced under Article 83 EC or Article III-163 of the Constitutional Treaty, provided that they are “not of a criminal law nature”, in the same way as the fines currently imposed on undertakings. See text accompanying notes 45 to 47 above. It would suffice to amend Regulation No. 1/2003. See text accompanying notes 187 and following; The Optimal Enforcement of EC Antitrust Law, supra note 4, section 8.7.4.5. 150 Sections 9A to 9E of the Company Directors Disqualification Act 1986, as amended by the Enterprise Act 2002, supra note 71. See also Office of Fair Trading, ‘Competition disqualification orders—Guidance’ (May 2003), available at http://www.oft.gov.uk/Business/Legal+Powers/ Enterprise+Act/publications.htm.


302

12/10/07

13:10

Page 302


disqualification may simply be the occasion to move into comfortable retirement, and the company could fully compensate its former director for any financial loss. It can also only be used for directors, and not for middle management. On the other hand, when it can be used, director disqualification is certainly more effective than imposing fines on individuals. If the person concerned is not ready for retirement, being barred from taking part in the management of a company may hurt the individual concerned in a way which the company cannot fully compensate financially. Director disqualification may also have a stigmatizing effect and may send a moral message—not to the same extent as imprisonment,151 but certainly more so than fines. While it is not an equally effective alternative to imprisonment, director disqualification would thus appear to be a useful complement, as well as a defensible second-best.152

3.2. Private actions for damages Could the problem of hard core cartels not being deterred by imposing only fines on companies be solved by adding private follow-on actions for damages? Here the answer is in my view clearly no.153 As explained above,154 the main problem with fines for companies is that the floor below which fines will generally not deter hard core cartels would be on the order of 150% of the annual turnover in the products concerned by the violation, and that fines of such a high level would often exceed the companies’ ability to pay. This problem exists just as much for damages as for fines, as well as for any combination of the two. Nor does adding private actions for damages against companies to fines for companies bring the advantage explained above of adding individual penalties, namely the increased effectiveness of leniency and whistle blowing programmes.

4. Five conditions for criminal antitrust enforcement to be effective not only in theory but also in practice For criminal antitrust enforcement (and in particular imprisonment) to work as an effective deterrent, it does not suffice to enact legislation creating a criminal offence. The effective operation of such legislation must also be ensured. Indeed, deterrence will only work if there is a credible threat of detection and punishment. 151 152 153 154

See text accompanying note 140 above. See also text accompanying notes 145 above and 180 and 181 below. See also text accompanying notes 171 to 174 below. Text accompanying note 101 and following.


12/10/07

13:10

Page 303


303

As explained hereafter, for imprisonment to be an effective deterrent for hard core cartel behaviour, four conditions would need to be fulfilled: (1) a dedicated investigator and prosecutor and (2) sufficient powers of investigation must be provided for, and (3) judges or juries must be willing to convict, all of which requires (4) a sufficiently broad political and public consensus that cartels are really bad and thus deserve severe punishment. Moreover, given that imprisonment can only work or only makes sense if it is limited to hard core cartels and complements other penalties for companies, it remains important to provide for such other penalties as well as for penalties to deter antitrust offences other than hard core cartels.

4.1. The need for a dedicated investigator and prosecutor For the threat of imprisonment to be credible, there must be a sufficiently well resourced and sufficiently dedicated body, or bodies, in charge of investigating and prosecuting criminal antitrust offences. In the US, the Department of Justice Antitrust Division has a National Criminal Enforcement Section in Washington, D.C., as well as seven field offices, in Atlanta, Chicago, Cleveland, Dallas, New York, Philadelphia, and San Francisco, dedicated to investigating and prosecuting criminal cases. The absence of a sufficiently well resourced and sufficiently dedicated investigator and prosecutor may to a significant extent explain the ineffectiveness of criminal enforcement in those EU Member States which have decriminalized their enforcement in the last two decades.155 Indeed, if antitrust offences have to compete with manslaughter and child abuse cases for the attention of a (typically already over-stretched) generalist public prosecutor’s office, they are most unlikely to be given much priority.

4.2. The need for adequate powers of investigation The success of US criminal antitrust enforcement in recent years is also the result of the strong powers of investigation which the US Department of Justice Antitrust Division has at its disposal.156 First, as Terry Calvani has described: “[T]he Division began using special agents of the Federal Bureau of Investigations (‘F.B.I.’) to assist in its cartel investigations. F.B.I. agents now routinely staff these investigations. Indeed, agents are actually seconded to the Division field offices where they work together with Division attorneys in jointly investigating cartels. [T]hese 155

See text accompanying notes 78 to 83 above. Apart from the three elements mentioned in text accompanying notes 157 to 161, Donald Baker also mentions the grand jury system, a secretive process favouring the investigators. See Baker D., supra note 22, at 693, 697 and 708. 156


304

12/10/07

13:10

Page 304


F.B.I. agents have brought with them the investigative techniques used to combat serious crime. Antitrust investigators now routinely employ computer forensics, ‘wire’ witnesses, conduct ‘ambush’ interviews, etc., in the course of investigations.”157

Second, as already explained above, the US Department of Justice has both a Corporate Leniency Policy and an Individual Leniency Policy, under which it can guarantee immunity from prosecution to co-operating companies or individuals.158 It can also make precisely calculated promises of reduced penalties to those co-operating at a later stage who cannot benefit from immunity.159 The Corporate Leniency Policy also includes a so-called “Amnesty Plus” option, under which a company can obtain a substantial additional discount in the fine it will have to pay for a first cartel being investigated (where the company was not the first leniency applicant and hence does not enjoy immunity from prosecution) if it reveals a second cartel in another market, for which it can get immunity.160 The Department of Justice also increases the pressure by asking the so-called “omnibus question” at the conclusion of a witness interview or grand jury interrogation. The question goes something like this: “do you have any information whatsoever, direct or indirect, relating to price fixing, bid rigging, market allocation, etc. with respect to other products in this industry or in any other industry?”. The witness must answer the question and must answer it fully and truthfully, or he or she would not only lose the leniency protection but would also be subject to the penalties of perjury or of making false statements or declarations. Third, as already mentioned above, in the US, companies cannot invoke the privilege against self-incrimination enshrined in the 5th Amendment to the US Constitution.161

4.3. Judges and juries must be willing to convict For the threat of imprisonment to be credible, it is also essential for judges and juries to be willing to convict antitrust offenders and sentence them to prison.162 157 Calvani T., supra note 132, at 7–8. It should be noted that the existence of a leniency programme does not dispense with the need to have other effective powers of investigation. Indeed, leniency can only work if potential leniency applicants know that the competition authorities are also able to detect cartels without co-operation from leniency applicants. See Principles of European Antitrust Enforcement, supra note 2, section 5.2.3. 158 See text accompanying notes 38 and 39, and 124 and 125 above. 159 However, it is an open question to what extent this is still possible following the judgment of the US Supreme Court of 12 January 2005 in United States v Booker, 543 U.S. 220 (2005), which has declared the US Sentencing Guidelines, on the basis of which such calculations could be made, to be non-binding. 160 See Spratling G. (19999): “Making Companies an Offer They Shouldn’t Refuse”, paper presented at the Bar Association of the District of Columbia’s 35th Annual Symposium on Associations and Antitrust (Washington, D.C., February 16, 1999), available at http://www.usdoj.gov/atr/public/speeches/2247.pdf; Hammond S., supra note 32, at 10. 161 See supra note 40. 162 See Kunzlik P., supra note 92, at 345–346; Green N., supra note 21, at 26–28.


12/10/07

13:10

Page 305


305

In the US, this is the case today, thanks at least in part to the Sentencing Guidelines.163

4.4. The need for political and public support Underlying the three above conditions is a more fundamental fourth one, namely that there needs to be a sufficiently broad political and public consensus that cartels are really bad and thus deserve severe punishment. In Donald Baker’s words: “When all is said and done, criminal liability ought to reflect the culture and values of the people of a country. The United States has a long tradition of fairly broad and sometimes noisy support of antitrust law and enforcement. After all, this is what created the Sherman Act in 1890 and the felony statute in 1974. Robber barons and cartel conspirators are bad people in the American popular lexicon.”164

As is well known, Europe does not have the same history.165 However, we are not condemned to remain stuck in our past. As John Coffee has pointed out: “the limited empirical evidence on public attitudes toward white-collar crimes suggests that the public learns what is criminal from what is punished, not vice versa”.166 Leadership, such as that being shown by the Irish and United Kingdom governments and competition authorities in criminalizing cartel enforcement, may thus produce results.167

4.5. Fines for companies and public penalties for antitrust violations other than hard core cartels remain necessary Imprisonment as a penalty for hard core cartels should only be used in combination with fines for the companies concerned.168 Indeed, if firms were not responsible for the antitrust violations committed by their agents, serious incentive problems would result. Given that the firm normally benefits from 163 Baker D., supra note 22, at 694. It was not always the case in earlier times. See Baker D. and Reeves B., supra note 130. See also supra note 159. 164 Baker D., supra note 22, at 714. 165 See text accompanying notes 92 to 99 above; The Optimal Enforcement of EC Antitrust Law, supra note 4, sections 8.6.7 and 9.5; Harding C. and Joshua J. (2003): Regulating Cartels in Europe, Oxford University Press. 166 Coffee J., supra note 7, at 1889. 167 See, inter alia, the speech by John Vickers, Chairman of the Board of the UK Office of Fair Trading at the British Chambers of Commerce on 31 March 2003, ‘Policy for markets and enterprise’, available at http://www.oft.gov.uk/NR/rdonlyres/62E7FFA8-3803-440C-811933FF119C0967/0/spe0203.pdf, at 4: “Since hard core cartels are like theft, criminalisation makes the punishment fit what is indeed a crime.” 168 See also text accompanying notes 123 to 128 above, as to the increased effectiveness of leniency and whistle blowing programmes resulting from the combination of corporate and individual liability, and my 2001 EUI paper, supra note 101, for other reasons why imprisonment should only be used in combination with corporate fines.


306

12/10/07

13:10

Page 306


the antitrust violations committed by its agents, it would have an incentive to encourage violations. Even if they risk individual penalties, the firm’s agents may be willing to take the risk if they expect to be sufficiently rewarded in terms of salary and promotion (or avoidance of demotion or dismissal). Assigning liability to the firm tends to remove the incentive for both the firm and its agents to commit antitrust violations on behalf of the firm.169 Firms are instead encouraged to strive for clarity in the messages conveyed to their agents as to the behaviour they expect from them.170 As argued above,171 imprisonment appears to be desirable only for horizontal, naked price fixing, bid rigging and market allocation schemes. However, all other types of antitrust violations, including anticompetitive vertical restraints, other horizontal agreements that restrict competition and violations of Article 82 also need to be deterred. As already mentioned above,172 in the US, apart from the possibility for the Federal Trade Commission to seek disgorgement of profits, deterrence of antitrust violations other than hard core cartels is left to private treble-damage suits. As appears to be recognised also in the US,173 this could be considered to be problematic. Indeed, as I have argued in detail elsewhere, private antitrust enforcement is inherently inferior to public enforcement.174 The United Kingdom, with its combination of a criminal cartel offence and a system of administrative fines and director disqualification for all types of antitrust violations, would in my view appear to have struck the balance exactly right.

169 Ulen T. (1997): “The Economic Case for Corporate Criminal Sanctioning”, in Lofquist W., Cohen M. and Rabe G., eds., Debating Corporate Crime, Anderson, 117, at p. 133, and Coffee J., supra note 121, at 410. 170 DeMott D. (1997): “Organizational Incentives to Care About the Law”, 60 Law and Contemporary Problems 39, at p. 47. 171 Text accompanying notes 141 to 145. 172 Text accompanying notes 33 to 35. 173 See letter from the Assistant Attorney General, Department of Justice Antitrust Division, supra note 35, at 2 (“Our system’s extreme weighting of the civil enforcement balance toward private lawsuits as opposed to federal government enforcement is not necessarily ideal.”). 174 Wils W. P. J. (2003): “Should Private Antitrust Enforcement Be Encouraged in Europe?”, 26 World Competition 473, and Principles of European Antitrust Enforcement, supra note 2, chapter 4. See also “MS.dos settlement—Microsoft’s pay-offs show why regulators are necessary”, Financial Times, 26 November 2004, p. 12; Kay J., “Antitrust concerns are not so easily dismissed” ’, Financial Times, 30 November 2004, p. 13; and Groom B., European Comment, ‘Microsoft’s Brussels trouble not over yet’, Financial Times, 10 November 2004, p. 16: “It is a comfort that competition policy in Europe is pursued largely by public authorities rather than through private lawsuits. Rival companies can be bought off, but the [European] Commission’s role is to safeguard consumers.”


12/10/07

13:10

Page 307


307

D. Is it problematic that antitrust enforcement is criminalized at the level of individual EU Member States without parallel criminalization at the level of the EU institutions, or without EU harmonization? 1. Regulation No. 1/2003 has been designed to accommodate criminalization in individual Member States The fact that individual Member States, such as Ireland, the United Kingdom and Estonia, have criminalized their enforcement of Articles 81 and 82, without parallel criminalization at the level of the EU institutions and without EU harmonization, is not problematic. Indeed, Regulation No. 1/2003 has been designed precisely to accommodate criminalization in individual Member States. As already explained above,175 Article 5 of Regulation No. 1/2003 leaves each Member State the choice of whether or not to provide for criminal penalties to enforce Articles 81 and 82. Article 12(3) of the Regulation, also already mentioned above,176 ensures that criminal enforcement in some Member States can coexist with noncriminal enforcement in other Member States. Indeed, whereas Article 12(1) of Regulation No. 1/2003 provides that the Commission and the competition authorities of the Member States can exchange all kinds of information, and use the information thus exchanged in evidence, Article 12(3) adds the following limitation: “Information exchanged pursuant to paragraph 1 can only be used in evidence to impose sanctions on natural persons where: —the law of the transmitting authority foresees sanctions of a similar kind in relation to an infringement of [Articles 81 or 82 EC] or, in the absence thereof, —the information has been collected in a way which respects the same level of protection of the rights of defence of natural persons as provided for under the national rules of the receiving authority. However, in this case, the information exchanged cannot be used by the receiving authority to impose custodial sanctions.’

Without this limitation, the free circulation of information between the competition authorities of all the Member States as well as the Commission would have caused problems both in the Member States which provide for individual penalties, in particular imprisonment, and in the other Member 175 176

Text accompanying notes 60 to 61. Text accompanying notes 60 to 61.


308

12/10/07

13:10

Page 308


States as well as at the level of the EU institutions. Indeed, if in Ireland or in the United Kingdom individuals could have been sent to prison on the basis of evidence collected by the Commission or for instance the Dutch competition authority in their administrative proceedings leading only to fines for companies, the rights of defence of the individuals concerned would have been weakened. Indeed, the Irish and United Kingdom courts probably would have refused the use of such evidence. On the other hand, the risk of evidence collected by the Commission or the Dutch competition authority in their administrative proceedings leading only to fines for companies being subsequently used in Ireland or the United Kingdom to send individuals to prison, could have led the EC and Dutch courts or legislators to raise the level of the rights of defence in proceedings by the Commission or the Dutch competition authority to the same level at that applying to criminal investigations in Ireland or the United Kingdom. This could have seriously harmed the effectiveness of the investigations by the Commission or the Dutch competition authority.177 All these problems have been avoided as a result of Article 12(3), whereas this provision does not hinder the competition authorities of Ireland, the United Kingdom, Estonia, and any other Member State which may follow their example, from exchanging all information between them and using it in evidence to have individuals convicted and sent to prison. Article 12(3) of Regulation No. 1/2003 prevents information collected in a jurisdiction that does not provide for imprisonment from being used in evidence in another jurisdiction to seek imprisonment. It does not, however, prevent the intelligence from being transmitted. The United Kingdom competition authorities could thus, for instance, after having received information from the Commission or the Dutch competition authority, start an investigation of their own to collect themselves the same or related information and subsequently use it in evidence to have an individual convicted and imprisoned. This possibility could make individuals, as well as the undertakings wanting to protect these individuals, less willing to cooperate with the investigations of the Commission or the Dutch competition authority. To the extent that cooperation under leniency programmes is concerned, this problem has been anticipated and addressed in the Commission’s Notice on cooperation within the network of competition authorities, which provides that information provided under a leniency programme to the Commission or a national competition authority will not be transmitted to another competition authority unless the latter commits itself not to use this information, nor any information obtained following that information, to impose sanctions on the leniency applicant or on any of its employees or former employees.178 177

See further Principles of European Antitrust Enforcement, supra note 2, section 1.2.12. Commission Notice on cooperation within the Network of Competition Authorities, OJ C 101 [2004], paragraphs 37 to 42. See also Schnichels D., supra note 70, at 114–115, and Principles of European Antitrust Enforcement, supra note 2, section 1.2.14. 178


12/10/07

13:10

Page 309


309

2. What would be the advantages of criminalizing the enforcement of EU antitrust law in all Member States, through EU harmonization, as well as at the level of the EU institutions? As explained above, Regulation No. 1/2003 has been designed to accommodate criminalization in individual Member States, without EU harmonization, and without criminalization of antitrust enforcement at the level of the EU institutions, with the result that criminalization in individual Member States is not problematic. However, this does not exclude that there may be advantages in criminalizing the enforcement of EU antitrust law in all Member States, through EU harmonization, as well as at the level of the EU institutions. I would personally expect that criminal cartel enforcement, for instance in the United Kingdom, will prove to be effective, even if enforcement in most other Member States and at the level of the EU institutions continues to rely only on fines for companies, in the same way as US criminal antitrust enforcement has for the past years been highly successful,179 even though most of the rest of the world did not have a similar enforcement regime. However, I see three or four advantages in criminalizing the enforcement of EU antitrust law in all Member States, through EU harmonization, as well as at the level of the EU institutions. First, as explained above,180 Article 12(3) of Regulation No. 1/2003 provides, for very good reasons, that the competition authorities in Member States such as the United Kingdom can only use information exchanged in the network of EU competition authorities in evidence in order to have individuals convicted and sent to prison if the information has been collected by the competition authority of a jurisdiction which also provides for imprisonment. If all EU Member States were to follow the United Kingdom example in criminalizing their enforcement of hard core cartels, either on their own motion or as a result of EU harmonization, and if enforcement at the level of the EU institutions could also be similarly criminalized, there would be an obvious advantage in that more competition authorities could more effectively work together in investigating criminal cartel offences. Second, more widespread or harmonized criminalization throughout the EU would enhance the effectiveness of sanctions, allowing, for instance, individuals suspected or convicted of a criminal cartel offence to be extradited between Member States, or allowing the effect of director disqualifications to be EU-wide. Third, as explained above,181 Article 12(3) of Regulation No. 1/2003 prevents information collected in a jurisdiction that does not provide for imprisonment 179 180 181

See text accompanying notes 130 to 136 above. Text accompanying notes 176 to 178. Text accompanying notes 176 to 178.


310

12/10/07

13:10

Page 310


from being used in evidence in another jurisdiction to seek imprisonment. As has further been explained, the Commission’s Notice on cooperation within the network of competition authorities provides that information provided under a leniency programme to the Commission or a national competition authority will not be transmitted to another competition authority unless the latter commits itself not to use this information, nor any information obtained following that information, to impose sanctions on the leniency applicant or on any of its employees or former employees. However, the risk of imprisonment in some Member States, and the exchange of intelligence within the network of competition authorities may still have an impact on the effectiveness of investigations by the competition authorities in the other Member States or by the Commission in the absence of leniency applications.182 Indeed, the risk of imprisonment may lead cartel participants to make even greater efforts to hide the evidence of their behaviour, and the possibility of intelligence being transmitted to the competition authorities in, for example, the United Kingdom may reduce the willingness of individuals, and of the undertakings wanting to protect them, to cooperate outside the framework of leniency with investigations by the Commission or, for example, the Dutch competition authority. Individuals may, for instance, be less forthcoming in answering on-the-spot questions during inspections,183 or companies may be more evasive in their answers to written requests for information. Fourth, criminalization in some EU Member States, without criminalization in others or at the level of the EU institutions, could result in those Member States that have criminal enforcement playing a disproportionate role, or carrying a disproportionate share of the burden, in the overall task of enforcing EU competition law, in the same way that the US in recent years has played a disproportionate role in deterring and punishing worldwide cartels.184 Specifically with regard to the role of the Commission in the overall task of enforcing EU competition law, Peter Kunzlik has made the following conjecture of what might happen if criminal enforcement in Ireland and the United Kingdom is successful, and other EU Member States follow their example: 182 The ability of competition authorities to detect cartels in the absence of leniency applications is also important for the effectiveness of leniency programmes. Indeed, if companies believed that the competition authorities would not be able to detect a cartel on their own, they would have no reason to apply for leniency. See supra note 157. 183 See text accompanying note 57 above as to such questioning under Article 20(2)(e) of Regulation No. 1/2003. See also Venit J. and Louko T, “The Commission’s New Power to Question and its Implications on Human Rights”, paper presented at the 31st Annual Fordham Corporate Law Institute Conference on International Antitrust Law & Policy, New York, 8 October 2004, as well as remarks by several participants in Roundtable Debate Session IV— Rights, Privileges and Ethics in Competition Cases at the same conference, in Hawk B., ed., 2004 Annual Proceedings of the Fordham Corporate Law Institute—International Antitrust Law and Policy, Juris Publishing, New York. 184 See The Optimal Enforcement of EC Antitrust Law, supra note 4, section 8.7.3.3.


12/10/07

13:10

Page 311


311

“[T]he situation might well arise in which cartels will be uncovered and prosecuted more efficiently at the national level than at the European level. In such a scenario the European Commission’s view that its proper vocation is primarily in the enforcement of Community competition law against hard-core cartels—a central premise of its own modernization proposals—may need to be revisited. By the time that the Commission’s investigations will have cranked into gear, the national systems may already have taken the initiative. The Commission may in practice lose its central role in European cartel enforcement to quicker, more effective, and betterresourced national competition authorities operating under [systems such as the ones in Ireland and the United Kingdom].”185

In Margaret Bloom’s words: “If criminal sanctions do provide far more effective deterrence, why has Commissioner Monti apparently not considered introducing such sanctions as part of the fight against cartels? Should such powers be considered now? One obvious answer is the fact that only six out of the 25 Member States have such powers in their national law and these have not yet been used on their own for any custodial sentences. Also, the European Commission does not currently have criminal powers in any area. At this time, it seems unlikely that there would be the necessary support within the European Union for such a change. But if the United Kingdom or another Member State demonstrates effective use of its powers and increases deterrence significantly as a result, the Commission should consider criminal powers seriously.”186

E. Would it be legally possible to criminalize antitrust enforcement at the level of the EU institutions, or to criminalize antitrust enforcement in all EU Member States through EU harmonization? 1. Under the current EC Treaty As I have argued in more detail in my 2001 EUI paper,187 under the current EC Treaty, it would be legally possible to criminalize antitrust enforcement at 185 Kunzlik P., supra note 92, at 352–353. see also ibid., at 347–348: “If the approach adopted in [Ireland and the United Kingdom] were to be emulated by other member states then transatlantic cartel enforcement cooperation might become substantially refocused upon the national rather than the European level. This would, in turn, affect the prestige of the European Commission within the European system. It might potentially take back from the Commission, and return to the authorities of the member states concerned, the de facto role as primary collaborators with the U.S. in terms of an important aspect of transatlantic cartel enforcement cooperation.” 186 Bloom M., supra note 136. 187 Supra note 101.


312

12/10/07

13:10

Page 312


the level of the EU institutions. In particular, it would be legally possible to provide for a system in which the Court of First Instance, upon investigation and prosecution by the European Commission, could convict individuals to prison for hard core violations of Article 81.188 Indeed, as is apparent from Article 47 TEU (“nothing in this Treaty shall affect the Treaties establishing the European Communities”) and from the very first words of Article 29 TEU (“without prejudice to the powers of the European Community”), the existence of the Treaty on European Union, with its Title VI on police and judicial cooperation in criminal matters, does not affect the possibility or impossibility of introducing under the EC Treaty criminal sanctions to enforce EC competition law. Apart from two specific provisions in Articles 135 and 280 EC, concerning customs cooperation and fraud affecting the financial interests of the European Community, the EC Treaty does not contain any provision excluding criminal law from its scope. As already mentioned above,189 the use of the term ‘undertakings’ in Article 81 EC leaves open the question of what sanctions, on undertakings and/or individuals, are to be provided for. This choice is to be made by the Council of Ministers, when using its broad legislative powers under Article 83(1) EC to lay down “the appropriate regulations or directives to give effect to the principles set out in Articles 81 and 82”. More problematic is the explicit reference to fines in Article 83(2)(a) EC, which provides that the regulations or directives referred to in Article 83(1) “shall be designed in particular: (a) to ensure compliance with the prohibition laid down in Article 81(1) and in Article 82 by making provision for fines and periodic penalty payments”. It is clear that the authors of the Treaty thought that compliance with the antitrust prohibitions would be induced by the threat of financial penalties, not imprisonment. However, one could argue on the basis of a teleological interpretation that, if the Council came to the view that prison sanctions were necessary for effective deterrence, the reference to fines in Article 83(2)(a) EC would not prevent it from adding such sanctions, given the wide mandate in Article 83(1) EC to lay down any “appropriate regulations or directives to give effect to the principles set out in Articles 81 and 82”. In any event, even if it were concluded that Article 83 does not provide a sufficient legal basis for introducing prison sanctions, the solution would be to use Article 308 EC. This provision empowers the Council, acting unanimously on a proposal from the Commission and after consulting the European Parliament, to take the appropriate measures whenever action by 188 The same conclusion was reached by Manfred Zuleeg in his paper for the same EUI Workshop, ‘Criminal Sanctions to Be Imposed on Individuals As Enforcement Instruments in European Competition Law’, published in Ehlermann C.-D. and Atanasiu I., eds. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland, Oregon, 453–462. 189 Text accompanying note 62.


12/10/07

13:10

Page 313


313

the Community proves necessary to attain, in the course of the operation of the common market, one of the objectives of the Community and the Treaty has not provided the necessary powers. Of course, the introduction of prison sanctions would require the separation of the powers of investigation and decision currently held in common by the Commission, and the transfer of the latter power to an independent judicial body, such as the Court of First Instance. The Commission would then act as a prosecutor before the Court of First Instance.190 Such a transfer of the decisional power from the Commission to the Court of First Instance would be possible under the existing EC Treaty. Indeed, as already mentioned above, Article 83(1) EC gives a broad mandate to the Council, acting by a qualified majority on a proposal from the Commission and after consulting the European Parliament, to lay down “the appropriate regulations or directives to give effect to the principles set out in Articles 81 and 82”. Article 83(2)(d) adds that the Council’s regulations and directives “shall be designed in particular: (d) to define the respective functions of the Commission and of the Court of Justice applying the provisions laid down [under Article 83(1)]”. Moreover, Article 229 EC provides that regulations adopted by the Council may give the Court of Justice “unlimited jurisdiction with regard to the penalties provided for in such regulations”. The fact that such “unlimited jurisdiction” can take the form not only of an unlimited review of a Commission decision but also of imposition of penalties directly by the EC Courts is apparent from the German language version of Article 229 EC, which refers to giving the Court a competence “which includes the power of unlimited control of judgment and the power to modify or impose such penalties”.191 The above reasoning to the effect that Article 83 EC (if necessary combined with Article 308 EC) would provide a sufficient legal basis to criminalize antitrust enforcement at the level of the EU institutions, can also be transposed to criminalization of antitrust enforcement in all Member States through EU harmonization, which thus appears equally possible. This conclusion also finds support in the recent judgment of the Court of Justice 190 This requirement of separation flows both from Article 5 ECHR and from Article 6 ECHR. See text accompanying notes 17 to 19 above. See also note 9 above. 191 Emphasis added; my translation from ‘eine Zuständigkeit [. . .] welcher die Befugnis zu unbeschränkter Ermessensnachprüfung und zur Änderung oder Verhängung solcher Maßnahmen umfaßt’. That the Commission would prosecute undertakings before the Court appears also to have been the understanding of the Spaak Report preceding the adoption of the EEC Treaty; see Comité intergouvernemental créé par la Conférence de Messine, Rapport des Chefs de Délégation aux Ministres des Affaires Etrangères, Mae 120f/56 corrigé (Bruxelles, 21 avril 1956), at 56: ‘Pour l’application concrète, il conviendra d’établir une procédure qui évite autant que possible une multiplication de procès devant la Cour. A cette fin, la Commission européenne constituerait un comité consultatif des ententes et discriminations qui l’aiderait dans une tâche de conciliation et d’arbitrage. A défaut d’une solution acceptée par les parties à l’expiration d’un délai qui pourrait être de deux mois, les Etats ou la Commission européenne elle-même pourraient introduire une plainte devant la Cour. Dans tous les cas, l’affaire sera instruite par la Commission européenne’.


314

12/10/07

13:10

Page 314


confirming EC competence as regards protection of the environment through criminal law.192

2. Under the signed but not ratified Constitutional Treaty It would appear doubtful whether the same possibility to criminalize antitrust enforcement at the level of the EU institutions would still exist under the signed but not ratified Constitutional Treaty.193 The crucial difference is that the Constitutional Treaty integrates the areas currently covered by the EC Treaty and by the other Titles of the EU Treaty into a single Treaty, without any hierarchy of the former over the latter. It can thus no longer be said that the presence of provisions on judicial cooperation in criminal matters (in Articles III-270 to III-274 of the Constitutional Treaty) does not affect the scope of the provisions on competition, and in particular of Article III-163 of the Constitutional Treaty, which is the (for all practical purposes identical) successor of Article 83 EC. Article III-271(2) of the Constitutional Treaty reads as follows: “If the approximation of criminal laws and regulations of the Member States proves essential to ensure the effective implementation of a Union policy in an area which has been subject to harmonisation measures, European framework laws may establish minimum rules with regard to the definition of criminal offences and sanctions in the area concerned.”

Article III-271(1) further provides for the possibility for European framework laws to “establish minimum rules concerning the definition of criminal offences and sanctions in the areas of particularly serious crime with a crossborder dimension resulting from the nature or impact of such offences or from a special need to combat them on a common basis”.194 European framework laws, which correspond to directives under the current EC Treaty, are, according to Article I-33(1) of the Constitutional Treaty, legislative acts “binding, as to the result to be achieved, upon each Member State to which it is addressed, but shall leave to the national authorities the choice of form and methods”. Whereas Article III-271 of the Constitutional Treaty could undoubtedly be used to criminalize the enforcement of EU antitrust law in all the EU Member 192 Judgment of the Court of Justice of 13 September 2005 in Case C-176/03 European Commission v EU Council, [2005] ECR I-7879. See also Communication from the Commission to the European Parliament and the Council on the implications of the Court’s judgment of 13 September 2005 (Case C-176/03 Commission v Council ), COM(2005)583final/2 of 24 September 2005. 193 Treaty establishing a Constitution for Europe, OJ C 310 [2004]. 194 Cartels are not among the areas of crime listed in the second subparagraph of Article III-271(1), but they could be added through use of the third subparagraph.


12/10/07

13:10

Page 315


315

States, the fact that this Article only allows the use of European framework laws (and not European laws, the equivalent of regulations under the current EC Treaty) appears to close the door to criminalization of EU antitrust enforcement at the level of the EU institutions. Given that Article III-271 is the provision of the Constitutional Treaty specifically providing a legal basis for criminalization, it would appear doubtful whether this limitation to the instrument of European framework laws could be overcome by using Article III-163 instead. A possible avenue forward could be provided by Article III-274 of the Constitutional Treaty, which provides for the possibility to establish a European Public Prosecutor’s Office, and to give this new body the task of investigating and prosecuting, in the competent courts of the Member States, the perpetrators of “serious crime having a cross-border dimension”. However, this provision would not allow the Commission to act as prosecutor, nor would it endow the EU Courts with the appropriate competence. As already mentioned just above, criminalization of the antitrust enforcement in all Member States through EU harmonization, albeit limited to the establishment of “minimum rules”, would clearly appear to be possible under Article III-271 of the Constitutional Treaty. In the end, of course, the main issue is whether the Commission and the EU Member States would want to criminalize the enforcement of EU competition law at the level of the EU institutions and, through harmonization, in all Member States. The day there is a consensus that this outcome would be desirable, it will also be possible to make whatever Treaty changes might be needed.


12/10/07

13:10

Page 316

(P) Ehlermann 07 Calviño

12/10/07

13:10

Page 317

II Nadia Calviño 1 Public Enforcement in the EU: Deterrent Effect and Proportionality of Fines

“Laws too gentle are seldom obeyed; too severe, seldom executed.” Benjamin Franklin “The worst form of inequality is to try to make unequal things equal.” Aristotle “All animals are equal, but some animals are more equal than others.” Animal Farm, G. Orwell.

A. Introduction Fines are important instruments for fighting against anticompetitive conduct, and they are the most significant ones in systems that do not entail criminal punishments, such as those prevailing in the EU. It is generally accepted that the fining system should be such that it aims at minimizing infringements of the law within the limits of scarce public resources. This requires the simultaneous fulfilment of two main objectives that might conflict with one another in a particular case: disincentive for future infringements (deterrence); and right punishment for the damage caused (proportionality). It is also common ground that fines should be set on the basis of objective and transparent criteria so that the rational criminal will decide to infringe the law or not on the basis of a cost-benefit analysis where the costs are sufficiently weighted and known. There are a wide variety of essays of a theoretical nature that have analyzed different sanctions and tried to establish the requirements for an optimal fine. There are also various papers that have critically assessed the EU and US fines imposed in administrative decisions and court judgments. Whereas all legal systems can always be improved (and pointing out their limitations is undoubtedly useful in this regard), this paper will take a 1 Deputy Director General with special responsibility for mergers, DG Competition, European Commission. At the time of the Workshop, the author was Director General for Competition, Ministry of Economy and Finance, Spain.


318

12/10/07

13:10

Page 318


constructive and practical approach in order to review the different elements or criteria that can be taken into account for setting the fine from the double perspective of proportionality and deterrence. With this aim, Section B briefly reviews the theoretical background for assessing fines. Following the two phase structure common to most fining proceedings, Section C reviews different principles for setting the basic amount when calculating fines, and Section D deals with the main criteria used for increasing or reducing the basic amount in order to fix the final fine. Finally, Section E draws some conclusions on these issues.

B. Proportionality and deterrence: is there an optimal fine? The main approach to the optimal fine under economic theory was initially formulated in Becker (1968) and is based on the decision function of a rational entrepreneur who aims at maximizing his utility function and decides to infringe the law when the gain derived from his action outweighs the cost of getting caught for it. Under this microeconomic approach, which focuses on the deterrent effect of sanctions and compares it with the benefit of the violation, the optimal fine will be such that it minimizes infringement by making the second element (cost of punishment) outweigh the first (gains from infringement). This is reflected in the following simple equation, whose first limb is the deterrence effect and the second the gain for the un-detected or nonprosecuted offender. p U (Y – c – F) > (1 – p) U (Y) where p U Y c

= = = =

probability of detection and enforcement of sanctions; utility function of the offender; net income from illegal activity; costs incurred in case of detection, excluding fines (lawyers, stigmatizing effect, etc.); and F = level of fines. Since the return from the offence (Y) cannot be controlled by the enforcers, the deterrence effect of the antitrust system depends on the level of the legal threat (mainly F and part of c2) and on the probability of enforcement (p), which can be managed by competition authorities. 2 For instance, the publicity of fines is an additional deterrent element because it can lead to a loss of public credibility, which is particularly important for brands which base their marketing strategies on the reliance, closeness or “lovability” of the product.


12/10/07

13:10

Page 319

Deterrent Effect and Proportionality of Fines

319

From this initial perspective, it could be argued that the legal system should theoretically take a maximalistic approach and include the strongest possible criminal or administrative penalties, because this would ensure that the cost outweighs the benefit of infringement. But fines are not socially costless. The cost of enforcement and the possibility of legal error should be considered. Moreover, the deterrence effect would be stronger where the effectiveness of competition policy in investigating and effectively imposing high penalties are enhanced, yet these two elements (i.e., high fines and effective enforcement) do not necessarily work in the same direction,3 so that an excessively “strong” legal system could paradoxically lead to under-deterrence. An excessively tough penalty would also not fulfil the proportionality test, since it would not bear any relation with the illegal profit obtained by the offender or to the effective harm caused to competition. If imposing too high a fine implies different inefficiency costs, the objective should be to have the minimum optimal fine that has the right deterrence effect by offsetting the expected benefit of the infringement. Moving to the second part of the equation, the fine should be related to the likely profit, and should thus increase in proportion to the returns obtained by the offender. It is usually very difficult to have access to accurate data on these returns, so competition enforcers can in practice use different proxies or criteria for estimating them: volume of sales in the affected market, market share, type of illegal behaviour, duration of the infringement, characteristics of the affected market, etc. The offender’s decision-making equation described above sets the basic scheme for explaining the rationale of an optimal fining system from the microeconomic theory perspective. But this reflects a one-time individual infringement, and does not take into account two elements which can be relevant from the proportionality perspective: the economic position of the offender (which will be considered in Section D below); and one of the core elements of competition policy, i.e., harm. From a general perspective, anticompetitive behaviour reduces consumer surplus and can create a deadweight loss for the welfare of society as a whole: the social loss from offences. This explains why, notwithstanding private action, most countries have created specialized administrative bodies for enforcing competition policy with a view to protecting the public interest underlying these infringements.

3 According to the experience in some countries, where competition infringements are punished as crimes, the requirement that fines or prison penalties be imposed by criminal courts can require a “beyond reasonable doubt” standard of proof that is difficult to sustain by competition authorities, thus leading to a less stringent application of the law than in purely administrative systems with apparently lesser penalties. Also, higher fines can reinforce the efforts of offenders in hiding their infringements, thus making prosecution more difficult.


320

12/10/07

13:10

Page 320


The proportionality of the fining system demands that it take into account not only illegal profit but also harm caused by the different infringements to competition and society in general. This is particularly clear in the case of cartels, where harm exceeds the illegal gains obtained by offenders.4 This perspective is formalized in an alternative economic model for setting the optimal fine, described in Posner & Easterbrook (1981). This model focuses on proportionality with regard to harm caused by the increase in prices and reduction in output caused by the departure from perfect competition, and also takes into account the possibility that infringements may lead to increased productive efficiency. This is reflected in the following simple graph of Williamson (1969), which sets forth a classical partial equilibrium model: P E2

P2

E1 AC

P1

D X2

X1

X

Graph 1

Here, D is the demand curve for a product, X is the exchanged quantity and P is the price paid. E1 is the perfect competitive equilibrium and E2 is the equilibrium in the non-competitive market structure. The AC curve stands for the average production costs. Numbers reflect the costs, quantities and prices in the two competitive (1) and non-competitive (2) equilibria. In order to ensure the deterrence effect, the penalty should be greater than the excess profit derived from the anticompetitive conduct. Considering a consumer surplus welfare standard, the optimal fine should be equal to the excess profit perceived by the illegal activity plus the deadweight loss of efficiency for society, which means the shaded area P1P2E1E2 in Graph 1. This model can be developed in order to take into account that anticompetitive conduct can result in increased productive efficiency. From a total 4 Because the harm is not only projected in the price increase for those consumers who do buy the affected product but also in the loss of consumption of those who cannot afford the higher prices.


12/10/07

13:10

Page 321


321

welfare point of view, fines should not stop conduct which leads to increased efficiency for society as a whole (i.e., where the increase in profits derived from the reduction of average costs exceeds the loss in consumer surplus from higher prices and lesser output). The main problem with this approach is the difficulty of having accurate data and of modelling each product’s market: measuring the excess profit and the deadweight loss depends on elasticity of demand and requires not only knowledge of the resulting output and prices in the relevant market but also an estimate of those which would have prevailed in the counterfactual situation, i.e., the competitive environment existing if the breach had not taken place. Furthermore, this approach can lead to optimal fines well above the offenders’ ability to pay. Finally, none of the theoretic approaches takes into account the costs of investigating and convicting offenders or their relative responsibility or role in the infringement. From a broad perspective of proportionality, fines should therefore take into account different elements of a quantitative and qualitative nature in order to fulfil the different objectives of the enforcement system. Needless to say, this is without prejudice to the fact that anticompetitive conduct can hurt individual competitors, clients, suppliers and consumers. Such offended parties should be able to try and obtain compensation for damages suffered before the courts via private enforcement of the competition laws. Finally, one additional aspect of the optimal fining system should be stressed: transparency. Knowing the likely punishment for illegal activities is essential for deterrence purposes. It also affects the costs (other than the cost of the fine) of getting caught, such as costs associated with damage to the public credibility of a brand or a firm. Publishing all individual decisions imposing sanctions, guidelines and judgments, and explaining the reasoning underlying the final fine imposed, is undoubtedly of the utmost importance for the effectiveness of competition enforcement. Of course, it is also relevant from the perspective of legal certainty for individual firms and the system as a whole. So, following the quote by Benjamin Franklin at the beginning of the paper, competition enforcers should not aim at imposing the highest possible penalty but the right one; a penalty which maximizes deterrence while at the same time ensuring proportionality from a double perspective: considering the profit derived by the individual offender and also the harm caused to society as a whole.


322

C.

12/10/07

13:10

Page 322


Setting the basic amount: gain versus harm

Most competition systems have two main elements for setting fines. First, they fix maximum fines, which can be absolute or relative to the volume of sales of the offender.5 Second, competition authorities use different criteria for fixing the fine corresponding to each individual offender, which can be more or less concrete and established by law, or which can just be specified in each decision. The fixing of the fine usually follows a two-step system: first, a basic amount is set according to different criteria; and secondly, that basic amount is adjusted in order to take into account the particular circumstances of the individual infringement or the offender, the maximum legal threshold, and any existing leniency programmes. Setting the basic amount can be essential for striking the right balance between the two different objectives of the fining system that were described in Section B, i.e., deterrence and proportionality (the latter of which, as explained above, incorporates the double perspective of returns to the offender and harm to society). This basic amount fixes from the start the reference for all additions or subtractions derived from aggravating or attenuating factors, and so sends economic operators and society a clear message as to the deterrence effect and proportionality of the whole system. According to current experience, there are two main methods for setting the basic amount, which can illustrate an approach different from the abovementioned objectives and the analysis described in Section B: per se rules based on the type of infringement; and a more economic approach based on the relative returns obtained by each offender or harm caused by each infringement.

1. Per se rules The traditional approach to fixing the basic amount of fines or even relative maxima has been based on the graduation of infringement types according to their gravity. This stems from the consideration that some illegal activities are more harmful than others and thus that infringements should be classified by categories and fined accordingly in order to ensure proportionality, coherence and non-discrimination. 5 The most common relative limit in the EU is 10% of the total volume of sales in the previous year (Article 23 of Regulation 1/2003), whereas the maximum in the US Sherman Act (as modified in 2004 by the Antitrust Criminal Penalty Enhancement and Reform Act) is USD 100 million or twice the gain of the violator or twice the loss of the victims for firms. Individual sanctions can be up to USD 1 million and up to ten years of imprisonment.


12/10/07

13:10

Page 323


323

Since it is not obvious that the illegal returns obtained by the offender are directly linked to the type of conduct, this approach seems to focus mostly on the objective of proportionality and especially on its harm-related element: some infringements are per se more harmful for competition than others. Classifying the different infringements is a very difficult task because the relative harm of the conduct depends not only on its nature (e.g., anticompetitive agreements versus unilateral conduct) but also on a number of quantitative and qualitative elements. Therefore, the per se rules approach needs to take into account additional criteria for fixing the basic amount, such as the duration, the actual impact on the market, or the size of the relevant geographic market. This system of per se rules based in the nature of the infringement has clear advantages in terms of legal certainty: the offender knows in advance the gravity of its actions and the likely fine, and there is a clear message to society as a whole as to the relative seriousness of each violation. The main drawback of this approach is that it addresses proportionality from a rather inflexible perspective and does not always take the economic perspective, that is, it does not depend on the different elements of the decision-making function of the individual offender. It focuses on one of the aspects of proportionality without addressing deterrence from the start. Of course, an ex ante, graduated scale of infringements with corresponding absolute fines is in principle less flexible than another where the basic amount is fixed as a percentage of the volume of sales of the offender.

2. Economic analysis As explained above, the main drawback of the ex ante, graduated categorization of infringements is the lack of underlying economic analysis. But the perfect fining system, based in an estimation of the decision-making function of offenders and the assessment and quantification of the actual returns obtained and harm caused to consumers or society in general, is a theoretical construct. The economic approach would demand an estimation of the utility function of the violator, and the actual returns to offenders would have to be assessed by calculating the mark-up applied with respect to the price that would have prevailed in a competitive situation. It is unrealistic to expect competition authorities to carry out this exercise in order to set the right fine. In practice, competition authorities may not have the means, resources or information for calculating the most efficient fine for attaining deterrence and proportionality in each particular case for each particular offender. It is therefore necessary to turn to variables that can serve as proxies for the optimal fine, and the volume of sales in the relevant market or markets is undoubtedly the first and main candidate in this quest. This is because sales


324

12/10/07

13:10

Page 324


volumes are a good proxy for the quantity of goods affected by the anticompetitive conduct (harm) as well as the total income generated by the offenders (return), and the market share of each party in each market can be used for the sake of proportionality. This is a very relevant turn from the traditional approach, for it does not consider that the nature of the infringement leads per se to a certain degree of harm. Nor does it consider that the total volume of sales of the offender should be taken into account as a starting point for setting the fine, although the latter should be nevertheless be considered when fixing the maximum level of fines in order to take into account the relative financial capacity of each individual firm, group or association. The main limitation of this approach is that it focuses on returns and might not take into account the actual public interest in offsetting the actual harm caused by the anticompetitive conduct to competition and society as a whole (i.e., the negative external effects of infringements). This could be overcome by exceptionally recovering the traditional per se approach and increasing the basic amount in the case of the most harmful anticompetitive conduct (hard core cartels) in order to send the right signal to economic operators, to avoid under-deterrence and ensure proportionality.

3. Duration The time element is very relevant for determining the return for offenders and harm to consumers. This element is therefore at the core of all fine determination systems, be they based on the nature of the infringement or on other indicators, such as the volume of sales on the relevant market. Although most illegal activities are of an indefinite temporal nature, some of them have an unstable nature that, amongst other things, explains the success of leniency programmes. Although there are clear exceptions, it is generally assumed that anticompetitive agreements age, and so the longer a cartel runs, the smaller the marginal profit derived from the agreement, the stronger the prisoner’s dilemma and hence the greater the risk of defection and whistle blowing. It is obvious that returns from illegal activities are not necessarily constant throughout the years. Considering only the most recent data could thus lead to an under-assessment, since extra profits usually diminish with the age of the cartel. Nevertheless, it is very difficult from a practical perspective to have access to accurate data for each year, so it is common practice to take the last year and estimate the volume of sales or even the extra profits throughout the whole life of the infringement.


12/10/07

13:10

Page 325


325

4. The initial amount in the EU and the US models Before the 1998 Guidelines, fines imposed by the EC Commission were generally based on a certain percentage of turnover in the relevant market. In order to limit litigation based on the definition of the relevant sales to be included in these calculations, the Commission published Guidelines on the method of setting fines, which establish a four-step proceeding for calculating fines: first, an absolute basic amount is set. Second, this number is increased or reduced by applying aggravating or mitigating circumstances applicable to the case and the individual offender. The application of the leniency programme and particular circumstances of the individual offender will determine the final fine.6 The basic amount of the fine will be determined according to the gravity and duration of the anticompetitive conducts, which are classified in three groups: minor, serious or very serious infringements—not only depending on the nature of the infringement but also on other quantitative or qualitative factors such as their impact, the rigour of their application, the dimension of the product and geographic affected market or the market share of the offender. This apparently clear-cut system is not automatic, it foresees different finetuning criteria (such as the effective economic capacity of offenders, which can be taken into account not only with regards to the maximum final fine but also when determining the basic amount) in order to give the Commission sufficient flexibility for ensuring proportionality. As pointed out before, the main limit of this ex ante approach is its lack of flexibility, which may lead to different, not always coherent, adjustments of the initial graduation, and even to the overestimation of aggravating and attenuating circumstances in order to avoid disproportionate fines.7 The application of these Guidelines has certainly led to a greater detain in the sanctioning decisions, to increased legal certainty, and has coincided with—or maybe enabled—a substantial increase in the level of fines, certainly consistent with the greater proportion of decisions on cartel cases. As to the final enforcement of sanctions, there is general agreement in the different studies reviewing jurisprudence on antitrust sanctions as to the

6 “Guidelines on the method of setting fines imposed pursuant to article 15(2) of Regulation No 17 and Article 65 (5) of the ECSC Treaty”, OJ C 98 [1998]. 7 For instance, some studies underline the incoherent classification of certain infringements as serious or very serious according to the dimension of the geographic market affected, which could precisely respond to the need to align the final fine with the relative financial capacity of the different offenders, and thus to guard the principle of proportionality and avoid imposing the disproportionate fines which would be derived from an automatic application of the principles of the 1998 Commission Guidelines.


326

12/10/07

13:10

Page 326


absence of judgments of the Court of First Instance or the Court of Justice increasing fines imposed by the EC Commission.8 Although the EU legal system does not foresee criminal penalties for the breach of antitrust rules, we should note the fact that they can be imposed for the breach of Articles 81 and 82 EC in certain Member States that do have these sanctions, including imprisonment of individuals, this being somewhat easier after the coming into force of Regulation 1/2003. Fines in the US must be based on the highest of the actual returns or harm derived from the infringement, although quantitative maxima are also foreseen in view of the difficulty of assessing these variables.9 Only horizontal collusion is to be prosecuted criminally, and the 2005 US Federal Sentencing guidelines establish fining criteria in this field. Although they are initially based on the type of infringement or its “offence level”, which resembles the per se model explained above, in practice these lead to establishing a basic amount of 20% of the “volume of affected commerce” of each firm involved, which is then increased or reduced according to various individualised criteria applicable.10 This system has historically led to higher fines than in the EU, although this trend seems to have been reversed in the last years.11 Nevertheless, there are two other aspects beyond the scope of this paper which certainly play a more prominent role in the deterrent effect of the fight against anticompetitive conduct in the US: the risk of prison penalties for executives, and the relevance of private litigation, which can lead to treble damages being paid to the plaintiff.12 There are two outstanding cases that can be used as clear examples of the relative penalty systems at both sides of the Atlantic.13 In the Vitamins cartel, the amount of fines and private damages to be paid in the US was of roughly 2 billion USD, and nine executives were imprisoned, whereas fines imposed in the EC amounted to 855 million €.14 The Lysine cartel was also found and convicted in the US, and led to fines and private damages awarded amounting to around 230 million USD,15 plus three executives imprisoned, compared to the 20 million € imposed in the EU.16 8

See Geradin & Henry (2005). See supra note no. 4. 10 See US Federal Sentencing Guidelines Manual, concerning fines imposed on individuals and organizations in judgements on horizontal antitrust infringements, at www.USSC.gov. 11 See Wehmhörner (2005) and Buccirossi and Spagnolo (2005). 12 See “Civil Remedies—Damages and Liability Discussion Memorandum” of the US Antitrust Modernization Committee staff of May 4, 2006, at www.amc.com. 13 See Camilli (2005). 14 Commission Decision of 21 November 2001 in Case COMP/E-1/37.512 Vitamins, and Judgement of the CFI of 15 March 2006. 15 Around 82 million fines agreed with DOJ and 148 millions due to private claims, which represents around 189% of the affected commerce. 16 Commission Decision of 7 June 2000 in case COMP/36.545/F3—Amino Acids, and Judgment of the CFI of 9 July 2003 in Case T-224/00 Archer Daniel Midland Co. v. Commission. 9


12/10/07

13:10

Page 327


327

D. Relevant criteria for setting the final fine Once the competition authority has fixed the basic amount from which the fine calculation will start, it will usually take into account different elements or criteria which try to take on board the particular circumstances of each infringement and offender in order to maximize the deterrent effect while safeguarding proportionality. Some of these elements have a more objective nature whilst others take a more subjective approach and try to punish not only the actual harm but also the extent to which the infringement is deliberate.

1. Affected market This is undoubtedly one of the main elements for determining the fine. In fact, it is the starting point for setting the basic amount under the more economics-based systems, but it can also be a relevant criterion for setting the final fine from a two-sided perspective: the geographic dimension and the characteristics of the market. Although the geographic dimension is already taken into account if the basic amount depends on the turnover in the relevant or affected market, it may nevertheless be possible to differentiate infringements according to the broadness of their application in different regions of countries and, in the case of the EC Commission, according to their partitioning effect within the common market.17 With regard to the nature of the relevant market, this can be taken into account in order to ensure deterrence and also proportionality with respect to the harm to society as a whole. The fact that a market has been recently liberalized, or that the infringement affects a sensitive input for the industry of for consumer welfare, can be a relevant element when determining the basic amount or the final fine.18

2. Reiteration Most fining systems consider that the persistence of the same violator should be taken into account when fixing the fine in order to increase its deterrent 17

Joined Cases C-2/01 P and C-3/01 P BAI v Bayer and Commission [2004] ECR-I 26. See Commission Decision of 11 June 2002, Case COMP 36.571/D-1 Austrian Banks— ‘Lombard Club’, OJ L 56 [2004]. At the national level, see the judgment of the Spanish Audiencia Nacional of 26 September 2005 in Suresa/Correos. 18


328

12/10/07

13:10

Page 328


effect. Nevertheless, this criterion could also have played a somewhat attenuating role when considered to be limiting the economic ability to pay of offenders (the payment of one fine is taken into account in order to reduce the second one). The effective application of this criterion may require the identity of the article infringed or of the affected sector or even take into account the time lapse between the different violations.19

3. Effect on competition Article 81 EC prohibits all agreements which have as their object of effect the prevention, restriction or distortion of competition within the common market, and therefore does not require an actual effect on competition in order for an infringement to be declared. Nevertheless, it is obvious that the whole competition defence system is directly oriented to avoiding harm caused to competition in the market and to consumer welfare. Therefore, the actual effects of the conduct in question should be taken into account when fixing the fine. In this regard, non-implementation of the agreement is relevant, as in this case there might be no effects on competition, and a smaller or even symbolic fine might thus be justified. Furthermore, the fact that a violator ceases to engage in the illegal conduct or adopts measures to offset the negative effects can be taken into account for deterrence as well as proportionality reasons, since this is obviously related to the actual harm caused.20 In this regard, some systems require that the cessation of illegal activities take place once the investigation of the competition authority is launched.21

4. Animus in the infringement Regulation 1/2003 establishes that fines may be imposed where individuals or undertakings intentionally or negligently break any of the rules or obligations imposed by Articles 81 and 82 EC and by the Regulation itself. There is therefore a requirement of a certain degree of intent or at least knowledge that the 19 See Commission Decision of 20 June of 2001 in case COMP/E-2/36.041/PO-Michelin. At the national level, see the judgment of the Spanish Supreme Court of March 23, 2005 in case BT/Telefónica. 20 At the national level, see the judgment of the Spanish Audiencia Nacional of 16 October 2003 in case Estate Agents. 21 See Commission Decision of 21 October 1998 in Case IV/35.691/E-4 Pre-insulated Pipe Cartel, OJ L 24 [1999]; Commission Decision of 21 November 2002 in Case COMP/E-1/37.512 Vitamins, OJ L 6 [2003].


12/10/07

13:10

Page 329


329

conduct was illegal and thus a rejection of pure objective responsibility arising from infringing conduct. Nevertheless, this knowledge can be presumed in enterprises such as those usually subject to an investigation under Community law.22 One interesting aspect of this criterion is the possibility that the existence of a reasonable doubt as to whether certain conduct is illegal may be accepted as an attenuating factor by competition authorities and courts. Although a certain firm may have breached the law by negligently adopting a certain decision, it may have been convinced that it was doing no wrong because there had been no previous published decisions imposing fines for such conduct,23 or because the public sector had transmitted the idea that it was not contrary to the legal system. The role of sector regulators, government offices or even national laws and regulations can bring about very interesting issues in this regard.24

5. Role in the collective infringement Although basic economic models of the optimal fine do not distinguish amongst offenders depending on their active or passive role in the infringement, most competition laws do consider that this is a relevant element for the sake of deterrence and of proportionality. It is obvious that playing an active role can be an aggravating factor. The greater the involvement in the illegal activity, the greater the degree of responsibility, as it results in a higher participation in the harm. This includes, for instance, being the founding member of a conspiracy, playing the role of “secretariat”, implementing retaliation measures against those deviating from an illegal agreement or even forcing other competitors to join the cartel.25 On top of being a criterion for increasing the fine, playing a leading role can mean exclusion from leniency programmes. 22 This “animus” requirement is also generally present in the laws of the Member States. This is the case in Spain, for example, where the jurisprudence has clearly stated that conduct which does not include negligence or intent is not forbidden and thus cannot be punished. See the judgment of the Supreme Court of 9 March 2005, General Council of the Official Colleges of Estate Agents; judgment of 23 March 2005, Telefónica; judgment of 13 December 2004, Tabacalera. 23 See Commission decisions of 24 June 2004, Belgian Architects (non-confidential version of decision available at http://ec.europa.eu/comm/competition/antitrust/cases/index/by_nr_77. html#i38_549), and of 21 May 2003, case COMP/37.578 - TeleBel+7/DT+Deutschland (Deutsche Telekom) OJ L 263 [2003]. 24 See, e.g., the judgment of the ECJ of 9 September 2003 in Case C-198/01 Consorzio Industrie Fiammiferi [2003] ECR I-8055. At the national level, see the judgment of the Audiencia Nacional in Spain of 6 May 1999, Cider cartel. 25 See Commission decisions in: Case IV/35.814 Alloy Surcharge, OJ L 100 [1998]; Pre-insulated Pipe Cartel and Vitamins (supra note 10); Case No IV/30.178 Napier Brown—British Sugar, OJ L 284 [1988]; Case IV/37.614/F3—PO/Interbrew, OJ L 200 [2003]; etc.


330

12/10/07

13:10

Page 330


6. Cooperation with investigations The optimal fine theories take into account the likelihood of the violator being prosecuted (p) but do not account for the administrative costs of investigating breaches and convicting offenders. These elements are generally considered in the different fining systems, whether as an aggravating factor (obstruction to inspections) or as an attenuating one (cooperating throughout the investigation can be considered for the reduction of the fine in addition to or in lieu of leniency programmes). Obstruction or lack of cooperation is usually considered an infringement in itself 26 and can be subject not only to fines but also to criminal punishments in many jurisdictions. Nevertheless, the prosecution of all procedural breaches would consume scarce public resources, which might otherwise be more efficiently devoted to fighting against substantive anticompetitive conduct. It thus seems necessary to keep the right incentives in place and consider cooperation—or its opposite—as a relevant criterion for determining the final fine imposed in the substantive decision.

Leniency programmes Leniency programmes are reductions in fines for cartel members that give the competition authority relevant information and conviction elements for prosecuting them. The US established the first leniency programme in 1993 and the EC followed in 1996 although the community programme was revised in 2002, giving way to a sharp increase in the number of cartel cases prosecuted by the Commission since then. Although they are considered to be very valuable instruments for fighting against cartels and result in a reduction of the optimal fine from the deterrence perspective, they pose relevant questions from the perspective of proportionality of the fining system. These leniency programmes mean prizes for whistle blowers and might lead to some maybe less active cartel members being punished whilst other founding or more prominent ones get away without a fine. Nevertheless, they have two main positive effects on the deterrence side, which explain their success and progressive introduction in initially reluctant countries. First, they can reinforce the instability of a cartel by generating lack of confidence between its members and augmenting the cost of remain26 For instance, supplying incorrect or misleading information, failing to answer a formal request for information or to produce required records, refusing to submit to inspections, or breaking seals are infringements subject to fines and periodic penalty payments under Articles 23 and 24 of Regulation 1/2003. Such activities are also typically infringements in various EU Member States’ competition laws.


12/10/07

13:10

Page 331


331

ing within the agreement and the benefit of deviating from it. Secondly, they can give the authorities proves and data which are indispensable for discovering and prosecuting the cartel, thus reinforcing the effectiveness of competition policy, the probability of conviction and thus the deterrent effect of the fining system.

8. Ability to pay Proportionality of the fining system usually demands that each participant in an infringement be fined according to its capacity to pay and also that the fine imposed is not such that endangers the financial viability of the offender, since this would have substantial economic and even social costs. This is usually projected in the absolute maximum fines set in the different national law systems (in most EU Member States and at Community level there is an absolute limit of 10% of the worldwide turnover of the undertakings in the year preceding that in which the decision is taken) but may also be taken into account when fixing the basic amount or the final fine itself.27 Past decisions in the EU usually classify different members of a cartel into fining levels or groups according to their relative total turnover, thus showing the relevance assigned to the capacity to pay criterion from the proportionality perspective. Also US and EU courts, at the national or Community level, have considered the firms’ ability to pay as a relevant element for limiting fines. Nevertheless, the overweighing of the total turnover might lead to unwanted results from the point of view of deterrence and even to proportionality given its lack of relation to either return to the offender or harm to consumers, and thus it should not have a leading role in setting the initial amount. The use of this quantity makes more sense in relevant markets and infringements of a worldwide nature. The ability to pay criterion may demand the consideration of the total turnover of the whole economic group and not of the individual company that has broken the law. Another alternative proxy is the total “market value” of the firm although this depends on the evolution of capital markets and not so much on the actual strength of the firm. The use of other financial indicators could let to unwanted incentives and results from the proportionality perspective, penalising more efficient operators. This criterion also poses interesting issues when the offender is not an enterprise but an association that might not have goods of its own. According to the 1998 Commission Guidelines, in this case decisions should as far as 27 At the national level, see the judgment of September 22, 2003 of the Spanish Audiencia Nacional in case Driving schools of Tenerife.


332

12/10/07

13:10

Page 332


possible be addressed to and fines imposed on the individual undertakings belonging to the association. If not, the fine will be imposed on the association will be equivalent to the total of individual fines that might have been imposed on each of the members28. Finally, a relevant implication of the ability to pay criterion can the consideration of a sector crisis as a mitigating factor, an element very rarely considered in an explicit manner in laws or decisions.

9. Ne bis in idem The Court of Justice clearly stated in its Walt Wilhelm judgment29 that the special system of shared jurisdiction between the Community and the Member States with regard to cartels may result in a double sanction imposed by the Commission and by national competition authorities. If two proceedings are conducted separately, and if they could 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”. This rule does not necessarily apply to sanctions imposed in other jurisdictions outside the EU.30 Nevertheless, the judgment clearly foresees that the same conduct can be punished by the Commission applying Articles 81 and 82 EC and by national competition authorities applying their own national laws. The situation has changed somewhat with the entry into force of Regulation 1/2003, which: requires the application of the EC Treaty by national competition authorities to all competitive conduct which could affect trade between Member States; sets limits to the scope of national rules governing anticompetitive agreements and serves as a basis for the European Competition Network; and aims at the efficient use of public resources and the coherence of the application of antitrust rules throughout the EU by encouraging the principle of “one-stop shop” and by discouraging the simultaneous or successive analysis of a case by two or more authorities. In this context, it could be that a particular national competition authority is the only one to investigate and prosecute an anticompetitive agreement that is implemented in different countries. But since that authority can only apply the sanctions provided for under its national rules, it may not always address the effects of the agreement in the whole of the EU.

28 At the national level, see judgment of the Spanish Audiencia Nacional of April 11, 2003, in case Autoescuelas Alcalá. 29 Judgment of the Court of 13 February 1969, Case 14/68 Walt Wilhelm and Others v Bundeskartellamt [1969] ECR 1. 30 Supra note no. 10.


12/10/07

13:10

Page 333


333

Another relevant twist in this field relates to anticompetitive conduct in the regulated sectors. In most EU Member States, there are sector-specific regulators in recently liberalized markets such as energy or telecoms. By virtue of the relevant Community Directives, these regulators have the power to set prices, impose certain restrictions or conduct, and to impose sanctions for non-compliance with orders or general rules. This could lead to double sanctioning of the same conduct from two different perspectives.

10. Fines imposed on natural persons International studies such as that of the OECD (2004) usually stress the insufficiency of corporate sanctions for deterring anticompetitive conduct and the usefulness of sanctions on individuals, including imprisonment, for anticartel enforcement. It is pointed out that fines on companies are finally paid by their shareholders in the form of lesser profits or a fall in the stock value. Thus, the punishment imposed does not affect persons directly responsible for infringements. To put it in institutional terms, there are internal agency problems that do not create the right incentives for deterrence and proportionality. From the perspective of economic theory, imprisonment penalties make sense only if the level of the expected penalties derived from fines is too low, but they are periodically defended on the basis of their stronger deterrent effect from the individual and collective perspectives. Nevertheless, this kind of sanction also has disadvantages resulting from the judicial nature of decisions, the higher standard of proof, the high direct costs and the substantially higher social welfare loss of errors. Although many competition laws do not provide for criminal penalties for competition breaches, most do include fines for executives that play an active role in them. The existence of this kind of criminal penalty raises interesting issues related to the use of information exchanged between competition authorities under the different international or EU level cooperation and coordination mechanisms.

E. Conclusion The objective of a legal system is to set the right incentives for individuals and organizations and fines play a relevant role in this regard. Antitrust penalties have to maximize deterrence without incurring in inefficiency costs and always protecting proportionality. If expected sanctions against a crime are


334

12/10/07

13:10

Page 334


too low to have relevant deterrence effects, enforcement activity is unproductive and the investigation and prosecutions costs are deadweight losses from a total welfare perspective. If they are too high they can lead to inefficiency costs and lack of proportionality. Deterrence and proportionality do not always go hand in hand: some criteria for the determination of individual fines which can be very useful for deterrence, such as leniency programmes, do not clearly respond to the principles of proportionality. Striking the right balance between the different objectives and defining the optimal sanctioning system is not an easy task. Administrative fines are the main instrument at the disposal of competition authorities in most jurisdictions for the time being, although the issue of imprisonment penalties is periodically brought forward in the EU and will probably continue to be so as fines imposed in front-page cases increase. Whereas there are different models in economic theory defining the optimal fine, they are not applicable in practice in the broad majority of cases by the average competition authority. It is thus necessary to take a pragmatic approach and to use general criteria and proxies which lead to a fining scheme as predictable, effective and consistent as possible: the challenge is not so much to impose the optimal fine but the best possible one within the given legal and factual framework. Most fining schemes follow a two phase proceeding by which a basic amount is fixed on the basis of de facto rules or economic criteria responding to actual returns to violators or harm to society derived from the illegal activity and then adjusted by applying aggravating and attenuating criteria of the concrete case of the offender. The 1998 Commission Guidelines have been until now the most developed fine-setting scheme within the EU. They were based on the fixing of a basic amount corresponding to the gravity and duration of the infringement, which is then fine-tuned according to aggravating and attenuating factors. This qualitative system, essentially focused on theories of harm, has proven to be a relevant guide for the economic operators and competition authorities, has sent clear messages about the relative gravity of the different infringements and enabled a relevant reinforcement of the deterrence effect of fines in the EU. Nevertheless, such a per se rules-based model can be rather rigid when assessing actual returns and harm produced by anticompetitive activities, and might not correspond to the more economic and effects-based approach that underlies the process of modernization of the EC competition rules. Beyond the fixing of the initial amount, the fining system must take into account the various quantitative and qualitative criteria reflecting the variables at stake for achieving deterrence and proportionality. They should be applied in a consistent manner, without losing sight of the different angles, perspectives and issues that can be relevant in each particular case for each particular offender. The case-by-case analysis demands a system that provides the right equilibrium between flexibility and legal certainty.


12/10/07

13:10

Page 335


335

Finally, we must not forget that fines are just one of the instruments of the enforcement system, and their effectiveness is directly related to enforcement efforts, to transparency, and to the use of other means for stopping infringements and achieving deterrence and proportionality, such as structural conditions, termination with undertakings, and the imposition of interim measures, not to forget the relevance of stigmatization of the offender vis-àvis its consumers, which can be more effective than fines in putting an end to anticompetitive conduct.


12/10/07

13:10

Page 336

(Q) Ehlermann 07 Venit

12/10/07

13:10

Page 337

III James S. Venit* Modernization and Enforcement—The Need for Convergence: On Procedure and Substance

Given the impact of, and the publicity attending merger cases, the primary topic in the transatlantic dialogue has, not surprisingly, tended to be substantive convergence. However, convergence has been an important force in other areas as well. Indeed, one way to understand the Commission’s modernization program is through the lens of convergence, the convergence in this case being the procedural changes introduced into European competition law in order to facilitate the international alignment of antitrust enforcement agendas. The decision to eliminate the system of prior notification and approval (and thus the Commission’s monopoly over the granting of exemptions) can be understood as one of the means (the leniency program being another) by which the Commission has sought to gain greater control of its own enforcement agenda in order to focus it on the three pillars of competition law: merger control, anti-cartel enforcement and abuse of marker power. These procedural reforms have in effect introduced a system of self-assessment similar to the approach in, inter alia, the US and Canada. Seen from this perspective the modernization program looks like a “best practices” adaptation and it would probably be mistaken to ignore the role that the interaction with other nonEuropean antitrust enforcement agencies had on the decision to modernize. Indeed, learning from other enforcers and the adoption of best practices— and the direction of the exchange is hardly one way—has been one of the significant by-products of the greatly increased enforcement agency interaction over the past ten years. And yet, despite the frequent inter-agency contact and interaction, important procedural differences remain. In general, because they are more formal, EU procedures are more time consuming than those in the US where, for example, the use of plea bargaining, sworn testimony and affidavits greatly accelerates the pace of the agencies’ enforcement activities. Another major difference is the relative absence of private enforcement in the EU. There is currently a Green Paper and there is an open discussion on this subject. But ten or even five years ago private enforcement was not regarded positively by the Commission which wished to retain the guiding hand in all matters affecting EU competition law. More recently, the * Partner, Skadden, Arps, Slate, Meagher & Flom LLP and Affiliates.


338

12/10/07

13:10

Page 338


Commission has come to perceive private enforcement as a way of strengthening its own enforcement efforts. It is less clear whether it also perceives private enforcement as another tool to facilitate control over its agenda, although this is clearly another attraction. I would now like to focus on three areas where there is clearly a need for greater convergence. These are (A) the scope of legal privilege; (B) leniency programs; and (C) the impact of criminal sanctions in trans-Atlantic cases.

A. Legal privilege The approach to legal privilege is far more restrictive in the EU than in the US. In particular, in contrast to the US, the EU does not recognize privilege for outside lawyers who are not admitted to an EU member State bar. Nor does it recognize privilege for documents containing legal advice originating from in-house counsel, although this issue is now being litigated in the Akzo case and the President of the Court of First Instance has suggested that the position on in-house counsel could change. The EU’s limited recognition of privilege for counsel admitted to a bar elsewhere than in another EU member State and for in-house counsel can have consequences that may go far beyond the treatment of the documents within an EU proceeding in which they first surface. This is because the refusal of the Commission to recognize privilege and its use of what in the US would be privileged documents can result in the loss of the privilege in the US not merely for the documents in question but, potentially, for the subject matter to which they relate. In such a case, thousands or hundreds of thousands of other documents that would otherwise have been privileged in the US may lose this protection. In practice this means that if the Commission, after having seized documents during an on-site inspection, refuses to respect the privilege that would exist in the US, a US enforcement agency or a private litigant may be able to get access in the US to those documents and also to all other privileged documents covering the same subject matter. These are draconian consequences and they raise the issue of whether the EU or US legal systems are capable of providing a remedy. One obvious solution would be international harmonization of the law of privilege. Another would be unilateral Court or Commission mandated change to the EU rules. A third would be a US court-developed rule that would limit the effects in US administrative or legal proceedings of another jurisdiction’s refusal to recognize the scope of legal privilege recognized by the US agencies and courts. It is difficult to judge which of these solutions is the least likely.


12/10/07

13:10

Page 339

Modernization and Enforcement—The Need for Convergence 339

B. Leniency programs A second problem area involves leniency programs. Both the US and the EU programs, once they were revised to provide greater certainty, have been highly successful. Unfortunately, they also contain contradictory conditions that can have serious repercussions and result in the exclusion of a leniency applicant from one or the other program. One major difference is the inability of EU applicants to put down a “marker” as they can in the US (and in other jurisdictions like Australia) thus preserving their priority while they gather the material necessary for a formal leniency application. A second major inconsistency relates to discovery and termination of the infringement. In the US, the leniency applicant is required to have taken “prompt and effective” action to terminate the illegal conspiracy once it has come to its “knowledge.” “Prompt and effective action” and “knowledge” at the corporate level are, of course, terms of art. In contrast, in the EU, the leniency applicant is required to have terminated its participation in the conspiracy on the day it applies for leniency, an event that may occur considerably later than corporate discovery of the infringement. A third difference is in the approach taken to termination of the conspiracy. The importance that the Commission attaches to preserving the element of surprise that is the precondition for a successful dawn raid makes it less tolerant of attempts by a conspirator to communicate its abandonment of the conspiracy to its partners. This concern is of sufficient magnitude that the Commission views the obligation not to take any action that could prejudice a successful dawn raid as an essential part of the leniency applicant’s duty to cooperate with the Commission, and thus a condition for admission to the EU program. In contrast, the US requirement of “prompt and effective action” may force a leniency applicant to communicate its withdrawal, thus alerting its partners of their possible exposure and rendering it “uncooperative” from an EU perspective. It may be possible to resolve such difficulties through consultation with both authorities when the leniency application is made. But success cannot be guaranteed and enforcement agencies have a tendency to focus on, and to want to preserve, the peculiar features that are part of their system rather than to conform them to the requirements of another system. It is not clear to what extent bilateral cooperation agreements between enforcement agencies provide enforceable rights for private parties. The better argument is that, probably, they do not. When it exists, the horns of this dilemma provide little shelter.


340

12/10/07

13:10

Page 340


C. The impact of criminal sanctions Any practitioner who has dealt with a cartel case involving a jurisdiction that imposes criminal sanctions and one that does not knows that the conduct of the case in the non-criminal jurisdiction will be strongly influenced, and indeed driven, by the exigencies and choices imposed by the existence of criminal sanctions in the other jurisdiction. This truism applies down the line from the decisions such as the decision whether to seek leniency to decisions to respond to oral questions during an inspection. These problems are not limited to EU/US cases; the may also occur where other EU competition authorities have opted for criminalization as in the case, for example of the UK, Ireland and Estonia. But whereas the Commission can rely on the duty of member states duty of loyal cooperation to avoid conflicts, it has, despite the existence of bilateral agreements, no such leverage with non-EU jurisdictions that impose criminal sanctions. EU fact gathering procedures are, in essence, consensual. That is, under EU law, firms under inspection, have an obligation of active cooperation subject to limited protection against blatant self-incrimination. Employees, as individuals, have no such obligations and no personal liability for their breach although the firm may be held responsible. The existence of criminal sanctions can have a significant impact on EU inspection procedures and firm’s obligations to cooperate. Take, for example, the Commission’s new power to question individuals during a dawn raid. Under EU law the individual has no exposure since under EU law it is the firm and not its individual managers or agents who are liable to fines for infringement of EU competition law or for failure to cooperate during an inspection. Rather, an employee’s failure to cooperate with the inspection, for example, by refusing to answer questions during a dawn raid, can only create liability for the firm. However, since the employee faces no Community law sanction for the firm’s infringement (he may face other sanctions if he has knowingly violated the firm’s compliance program), the EU rules draw reasonably clear lines, which tend to incentivize cooperativeness during an inspection. The incentives under the EU system are, however, thrown completely off kilter if there is a parallel investigation in a non EU jurisdiction that imposes criminal sanctions. In that jurisdiction the individual’s willingness to answer the EU inspectors’ questions may generate evidence that may ultimately be used against the individual in a parallel proceeding or, indeed, may be deemed to constitute a waiver of the individual’s protection against self-incrimination. Once this factor enters the calculus, the incentives change substantially: the interests of the individual and the firm are no longer aligned as potential exposure to criminal sanctions eliminates the employee’s incentive to cooperate. The change in the employee’s incentives creates a conflict between the firm’s


12/10/07

13:10

Page 341

Modernization and Enforcement—The Need for Convergence 341 obligation to cooperate and the interest of its employees to remain silent. One practical consequence is that employees may need to have separate counsel and the firm’s counsel may have an ethical obligation to recognize this necessity. Another is that the criminal jurisdiction, because of its more powerful penalties, is likely to reduce or eliminate the Commission’s ability to gather information. In the same way that bad money drives out good, stricter sanctions tend to impose themselves and replace or limit the effects of less severe ones. Solutions to this type of problem are not easy. Short of uniform of adoption of a criminal system, one obvious remedy is the existence or adoption of rules in the criminal jurisdiction preventing the use of evidence gathered in a noncriminal jurisdiction that does not afford the same procedural protections as the criminal jurisdiction. The law in this area may not be clear and it would be interesting to determine the extent to which such rules already exist and provide effective protection. As I think the foregoing examples have illustrated, the absence of convergence can have serious consequences for firms active in the international arena. Whereas the asymmetries that are created by the existence in one jurisdiction of criminal sanctions may be the most difficult to overcome, the three areas surveyed above are all ones in which convergence, or compensating measures to deal with the consequences of the lack of convergence, are needed.


12/10/07

13:10

Page 342

(R) Ehlermann 07 Calkins

12/10/07

13:10

Page 343

IV Stephen Calkins1 Coming to Praise Criminal Antitrust Enforcement

Writing a paper in praise of criminal competition enforcement in Europe, for a conference session that includes Professor Wouter Wils, might seem something of a bootless exercise. The paper has already been written—at great length with extensive citation of authorities2—and then supplemented,3 and presumably Professor Wils will further supplement it for this conference. So also, others at the conference with great European experience have peered into the future and pronounced that criminal enforcement will play a central role.4 Rather than offer yet another set of arguments for criminal competition enforcement, this paper will briefly comment on those already familiar.5 In particular, it will stress the importance of disaggregating the argument and considering separately the importance of individual liability and criminal treatment. It will also note the irony that at the same time that US criminal antitrust enforcement is flourishing, doubts have arisen about the role of criminal corporate enforcement in other areas. Somewhat less well developed than the argument for expanding competition remedies is an appreciation of how fundamentally such an expansion can change the nature of the enforcement process and the substantive legal standards that are enforced.6 My comparative advantage lies in the US 1 Professor of Law and Director of Graduate Studies, Wayne State University Law School, Detroit, MI, USA. The author is a former General Counsel to the Federal Trade Commission serves Of Counsel to Covington & Burling, but writes solely in his individual capacity. He thanks Margaret Bloom, David Hull, Julian M. Joshua, and Spencer Waller for sharing thoughts with him, and Adele Ice and Kristin Pell for research assistance, but is personally responsible for the paper and especially any errors and misjudgments therein. 2 E.g., Wils W. P. J. (2005): “Is Criminalization of EU Competition Law the Answer?”, 28 World Competition 117; see also Calvani T. (2004): “Competition Penalties and Damages in a Cartel Context: Criminalization and the Case for Custodial Sentences”, paper presented at the ACLE workshop on Remedies and Sanctions in Competition Policy, Amsterdam, February 2005, text available from http://www.acle.nl; (Dec. 1004); Calkins S. (1998): “Corporate Compliance and the Antitrust Agencies’ Bi-Modal Penalties”, 60 Law & Contemporary Problems 127 (US law). 3 E.g., Wils W. P. J. (2006): “Optimal Antitrust Fines: Theory and Practice”, i28 World Competition 117. 4 E.g., Bloom M. (2005): “What Is the Future for International Antitrust?”, paper presented at the ABA Antitrust Section Fall Forum, November 15–16, 2005. 5 To avoid overkill, the paper will refrain from citing Professor Wils’ and others’ extensive writings. 6 Some of the points discussed herein have been explored in different ways my previous work, which I will generally refrain from citing. See Calkins S. (2006): “Civil Monetary Remedies Available to Federal Antitrust Enforcers”, 40 University of San Francisco Law Review; 567


344

12/10/07

13:10

Page 344


experience, which conveniently offers the most robust experience with both criminal enforcement and another expanded remedy, private enforcement. The second part of this paper will explore the ways that competition remedies may have influenced substantive standards and enforcement processes in the US.7 The third, brief part of this paper will draw upon the US experience to offer some very tentative thoughts about the European situation.

A. Reflections on the case for criminal enforcement Deciding on what is commonly thought of as criminal enforcement is not one decision, but three: —Should the consequences imposed on individuals or entities be ones that are denoted as criminal rather than civil. —Should enforcers punish individuals instead of (or in addition to) entities. —Should enforcers impose the unique punishment of incarceration.

1. Individual versus entity One pronounced change has been the recognition of the importance of punishing individuals and not just entities.8 As the Antitrust Division’s Scott Hammond has observed—reviewing progress in Australia, Japan, Israel, and Ireland—there is a “global trend toward greater individual accountability.”9 For many reasons, this makes good sense. Two arguments have particular appeal. First, individual liability results in greater deterrence. If a company employee is told that something untoward may happen to his or her employer Calkins S. (2003): “Perspectives on State and Federal Antitrust Enforcement”, 53 Duke Law Journal 673; Calkins S. (1998): “Corporate Compliance and the Antitrust Agencies’ Bi-Modal Penalties”, 60 Law and Contemporary Problems 127; Calkins S. (1998): “In Praise of Antitrust Litigation: The Second Annual Bernstein Lecture”, 72 St. John’s University Law Review 1; Calkins S. (1986): “Summary Judgment, Motions to Dismiss, and Other Examples of Equilibrating Tendencies In the Antitrust System”, 74 Georgetown Law Journal 1065. 7 For another effort that considers some of these issues, see Kunzlik P. (2003): “Globalization and Hybridization in Antitrust Enforcement: European ‘Borrowings’ from the U.S. Approach”, 48 Antitrust Bulletin 319. 8 See, e.g., OECD, Hard Core Cartels: Third Report on the Implementation of the 1998 Recommendation, at p. 27 (“A strong case can be made that cartel enforcement will be more effective if sanctions against individuals are part of that mix.”), available at http://www.oecd.org/ dataoecd/58/1/35863307.pdf. 9 Hammond S. (2006): “Charting New Waters in International Cartel Prosecutions”, remarks presented to the Twentieth Annual National Institute on White Collar Crime, March 2, 2006, at p. 2, available at http://www.usdoj.gov/atr/public/speeches/214861.htm.


12/10/07

13:10

Page 345

Coming to Praise Criminal Antitrust Enforcement

345

at some point in the possibly distant future, the employee may well experience relatively little concern. The employee may have left the company by then, whether to retirement or another position, especially now that we live in an age of early retirements and increased corporate mobility. An individual consequence inflicts individual pain—the kind that is most acutely felt. Everyone who has engaged in counseling has a relevant anecdote. Mine is of a corporate general counsel expressing his pleasure that the FTC had extracted very substantial payments from individual executives of a firm in a related industry, because he could point to that to get the attention of his top management. The second reason is the flip side of the first. Often a belated corporate punishment will harm individuals who did nothing wrong.10 Shareholders who may have enjoyed seeing enhanced profits may well have sold out to newcomers doubly hurt, once by paying inflated share prices and again by suffering the consequences of any corporate penalty. Investors should exercise caution, of course, but one cannot expect them to uncover concealed wrongdoing. If a company is punished by a remedy that drives up the cost of doing business, consumers may suffer, as may employees unrelated to any wrongdoing. At the extreme, should a company be put out of business, massive numbers of innocent employees and customers as well as governments and others who have to step in to help, will be harmed, as happened when Arthur Anderson was indicted (even though the indictment was later overturned). Indeed, the editorial page of the Wall Street Journal, which has long viewed the class-action firm Milberg Weiss Bershad & Schulman as a curse, expressed regret that the implosion of that firm that may result from its recent indictment will inflict unneeded harm on secretaries and paralegals who share none of the very great blame the Journal heaps on the responsible partners.11 Some disagree about the necessity of targeting individuals. All that is needed, the argument goes, is an ideally calibrated corporate penalty—which, they might note, can be lessened to reflect the greater chance of detection flowing from leniency programs.12 Presumably they would argue that any unfairness, considered ex post, is fair when considered ex ante. And in theory they are right. The problem comes because reality still seems to depart from theory; agency costs loom larger, not smaller, than they once did; and time horizons have, if anything, shortened. Under the circumstances, it makes eminent sense to impose external consequences on wrongdoing individuals as well as entities.13 10 See, e.g., Coffee J. (2006): Reforming the Securities Class Action: An Essay on Deterrence and Its Implementation, Columbia Law and Economics Working Paper No. 293. 11 “Very Rough Justice”, Wall Street Journal, May 22, 2006, at p. 12. 12 See, e.g., Buccirossi P. and Spagnolo G. (2005): Optimal Fines in the Era of Whitleblowers. Should Price Fixers Still Go to Prison?, Lear Research Paper 05-01, at p. 41. 13 Since excessive reliance on punishing individuals may make detection more problematic, this is not to suggest that sanctions be limited to individuals. See Langlais E. (2006): Corporate Criminality: Should We Rely on Gatekeepers’ Shortcomings to Deter Dishonest Principals (mimeo).


346

12/10/07

13:10

Page 346


The US Department of Justice (“DOJ”) Antitrust Division firmly believes in the importance of punishing individuals, and misses no opportunity to mention the number of individuals who suffered serious consequences. The Division has acted upon this belief, and in the past decade has charged more individuals than corporations with criminal responsibility:14

140

Individuals v Corporations Criminally Charged Individuals Charged Corporations Charged

Number Charged

120 100 80 60 40 20 0

1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005

Year As reviewed further below, the Justice Department in criminal securities fraud cases has taken even more dramatic steps to impose responsibility principally on individuals.

2. Criminal versus civil? As Professor Wils and I and others have pointed out, one principal advantage of criminal penalties is in communicating a signal. It is hard not to stand morally condemned when one has been charged with a crime. A recent example is provided by the state of Connecticut, which announced that insurance broker R.C. Knox “accepted $415,058 in improper concealed compensation from insurers” for placing state insurance purchases. Attorney

14 Source: Department of Justice, Workload Statistics FY 1996–2005 (available at http://www.usdoj.gov/atr/public/workstats.htm); Antitrust Division Workload Statistics FY 1986–1995, 70 Antitrust & Trade Regulation Report (BNA) p. 299 et seq. (referring to individuals and corporations “indicted” but apparently meaning the same thing as criminally charged). All graphs included herein take date from these same sources.


12/10/07

13:10

Page 347


347

General Richard Blumenthal announced that a consent order resolving this, in which R.C. Knox paid $754,804 (the concealed compensation plus interest), “ ‘sends a consistent, insistent message: We will eradicate concealed illegal payments and bid rigging in this industry.’ ”15 Similarly, comptroller Nancy Wyman declared, “ ‘These secret commissions are an abuse of taxpayer money and an abuse of public trust.’ ”16 Yet how did the company respond? By declaring, in its press release, that “R.C. Knox believes its receipt of contingent commissions was entirely lawful.”17 Had R.C. Knox pleaded guilty to a crime, it would probably not have boldly declared that its actions were entirely lawful. (Of course, for that same reason Connecticut could not have settled the case so easily.) Ironically, while the DOJ Antitrust Division roars ahead with criminal enforcement especially against international cartels,18 the rest of the DOJ has gotten cold feet about applying other criminal laws against major corporations. DOJ just indicted the controversial class-action law firm, Milberg Weiss Bershad & Schulman, but this was widely viewed as a remarkable exception. Especially since a later-overturned indictment of Arthur Andersen19 drove that major American institution rapidly into bankruptcy and permanently out of business, the Justice Department has been reluctant to indict major business entities for non-antitrust white-collar crimes.20 The proof of this reluctance is in the actions of the Justice Department’s Corporate Fraud Task Force, which has helped bring the celebrated cases involving Adelphia Communications, Enron, Healthsouth, and WorldCom—cases that the DOJ says it brings “against those who have traded on their positions of trust to defraud their investors, their employees and the public.”21 The Task Force’s Web Site listing of “Significant Criminal Cases” lists actions involving 92 companies (including those four). It lists

15 Press Release, “Attorney General Announces Settlement With Insurance Broker R.C. Knox For Accepting Undisclosed Payments”, May 18, 2006, available at http://www.ct.gov/ag/ cwp/view.asp?A=2341&Q=314836. 16 Id. 17 Press Release, “R.C. Knox And Company, Inc. Enters Into Settlement Agreement With State Attorney General’s Office Regarding Contingent Commission Inquiry”, May 18, 2006 (“R.C. Knox agreed to resolve this matter, without admitting any wrongdoing, on the terms the Attorney General presented in order to avoid the uncertainty and expense of contesting possible claims the Attorney General may have raised in connection with this issue.”), available at http://www.peoples.com/pressroom/article/0,8401,14171,00.html#. 18 See, e.g., Hammond S., supra note 9. 19 United States v. Arthur Anderson, LLP, 334 F.3d 281 (5th Cir. 2004), overturned, 544 U.S. 696 (2005). 20 See also Laufer W. (1999): “Corporate Liability, Risk Shifting, and the Paradox of Compliance”, 52 Vanderbilt Law Review 1343, at p. 1392 (“There is a sense that it is imprudent to be unnecessarily harsh when the social costs of a criminal indictment against an organization with multiple stakeholders may be great.”) (footnote omitted). 21 Significant Criminal Cases and Charging Documents, at http://www.usdoj.gov/dag/cftf/ cases.htm.


348

12/10/07

13:10

Page 348


indictments or informations of 301 individuals (including, of course, Kenneth Lay and Jeffrey Skilling).22 Yet with respect to corporations, it lists only a single indictment: the ill-fated indictment of Arthur Anderson. In addition, it lists seven corporate “deferred prosecution agreements,” two settlement agreements, and one notation of a criminal complaint dismissed. These “deferred prosecution agreements” and “non-prosecution agreements have become a major pattern because they serve (a) to avoid criminal liability, while (b) imposing consequences on individuals.23 Thus, when seven executives of Symbol Technologies were indicted for corporate fraud, the corporation was not. Instead, it acknowledged wrongdoing and agreed to cooperate fully, paid $139 million in redress, and agreed to terminate all employees at fault, dismiss its accounting firm, restructure its board of directors, appoint a monitoring examiner, and establish a training program.24 When two former Computer Associates executives were indicted, the DOJ announced a “deferred prosecution agreement” with the company: it accepted responsibility for the illegal conduct and agreed to cooperate; agreed to terminate responsible officials, appoint new management and reorganize; and agreed to pay $225 million in redress.25 When two Bristol-Myers Squib executives were indicted for securities fraud, the company’s deferred prosecution agreement called for it to, among other things, appoint a particular individual as nonexecutive chairman of its board of directors, adopt internal controls and engage a particular independent monitor, pay $300 million in additional redress, and endow a chair at a law school.26 Perhaps most controversially, KPMG entered a deferred prosecution agreement that saw nine individuals indicted for tax fraud while the company accepted responsibility for a massive tax scheme, agreed to pay $456 million, terminated two of its tax practice 22 Web site visited August 20, 2006. For an earlier count the makes the same point, see Abril P. and Morales Olazabal A. (2006): “The Locus of Corporate Scienter”, Columbia Business Law Review 81, at pp. 109–110. 23 See, e.g., Christopher J. Christie & Robert M. Hanna, A Push Down the Road of Good Corporate Citizenship: The Deferred Prosecution Agreement Between the U.S. Attorney for the District of New Jersey and Bristol-Myers Squibb Co., 43 Am. Crim. L. Rev. 1043 (2006) (government attorneys laud the flexibility); Eugene Illovsky, Corporate Deffered Prosecution Agreements: The Brewing Debate, 21 Crim. Just. 36 (summer 2006) (explaining deferred prosecution agreement as a way of imposing probation before conviction); Christopher A. Wray & Robert K. Hur, Corporate Criminal Prosecution in a Post-Enron World: The Thompson Memo in Theory and Practice, 43 Am. Crim. L. Rev. 1095 (2006) (discussing principles and experiences with deferred prosecution agreements and non-prosecution agreements). 24 Press Release, US DOJ, “Former Senior Executives At Symbol Technologies Indicted In Massive Corporate Fraud Scheme”, June 3, 2004, available at http://www.usdoj.gov/usao/nye/ pr/2004jun3.htm. 25 Press Release, US DOJ, “Former Computer Associates Executive Indicted on Securities Fraud, Obstruction Charges”, September 22, 2004, available at http://www.usdoj.gov/opa/pr/ 2004/September/ 04_crm_642.htm. 26 Press Release, US DOJ, “Bristol-Myers Squibb Charged with Conspiring to Commit Securities Fraud; Prosecution Deferred for Two Years”, June 15, 2005, available at http://www.usdoj.gov/usao/nj/publicaffairs/ J_Press/iles/bms0615_r.htm; see also Christie Hanna, supra note 23 (defending arrangement) .


12/10/07

13:10

Page 349


349

areas, restricted its acceptance of fees not based on hourly rates, and installed a government-appointed monitor.27 Earlier in the proceeding, KPMG had fired or threatened to fire, and refused to pay the legal fees, of employees not cooperating with the government.28 Through these rather drastic means and extensive individual liability, entities have avoided indictment.29 Needless to say, these steps have been controversial. Some critics complain that major corporations are getting a pass, whereas others accuse the government of extreme meddling in corporate affairs.30 Certainly the threat of a criminal prosecution that may be a corporate death sentence gives the government extraordinary power, and one always has to worry about abuse of such power. The striking point for antitrust, however, is how very different are the worlds of DOJ antitrust and DOJ securities fraud—although in both worlds, the criminal label matters a great deal.

3. Incarceration Criminal liability is not the same thing as incarceration. Nor is incarceration the only way substantially to punish individuals. As part of a systematic campaign to cause individuals to bear consequences for their actions, the Justice Department’s controversial “Thompson Memorandum” has indicated that decisions about charging corporations may turn on factors such as the corporation’s willingness to replace responsible management, discipline or terminate wrongdoers, and refrain from voluntarily paying the legal fees of culpable employees.31 Through its deferred prosecution agreements, the DOJ fraud operation has inflicted quite serious consequences on individuals wholly apart from jail time. That said, jail concentrates the attention like no other penalty. As has been noted elsewhere, it is the uniquely extreme punishment for a corporate official. I recall showing up at a substantial corporation long ago to help prepare 27 Press Release, US DOJ, “PMG to Pay $456 Million for Criminal Violations in Relation to Largest-Ever Tax Shelter Fraud Case”, August 29, 2005, available at http://www.usdoj.gov/opa/ pr/2005/August/05_ag_433.html. 28 Hasnas J. (2005): “Ethics and the Problem of White Collar Crime”, 54 American University Law Review 579, at p. 626. 29 Kennealy K. and Breen K. (2005): “White Collar Crime”, 29 Champion 44, at p. 44. 30 Goldstein H. (2005): “Corporate Crime: Government Prosecutors Are Going to Extremes”, 234 New York Law Journal 5; Michael E. Horowitz & April Oliver, “Foreword: The State of Federal Prosecution,” 43 A. Crim. L. Rev. 1033 (2006) (reviewing controversy); Illovsky, supra note 23 (same). 31 Thompson D. (2003): “Principles of Federal Prosecution of Business Organizations”, US DOJ, January 20, 2003, available at http://www.usdoj.gov/dag/cftf/corporate_guidelines.htm. For a thoughtfully qualified defense of the memo, see Wray & Hur, supra note 23. Judge Kaplan recently held that the Thompson Memo, as applied in the KPMG proceeding, created unconstitutional pressures. United States v. Stein, 2006 U.S. Dist. LEXIS 50723 (S.D.N.Y. July 25, 2006); United States v. Stein, 435 F. Supp. 2d 330 (S.D.N.Y. 2006).


350

12/10/07

13:10

Page 350


14000

Total # of Actual Days of Incarceration Imposed by Court

12000

Days

10000 8000 6000 4000 2000 0

1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005

Year the chief executive officer to appear before a grand jury that might or might not lead to a prison sentence, and for that day, no one in the building was more important than I was. We know from cartel results and from reports of cartel participants that the U.S. jails are feared far more than any other possible consequence of cartel behavior.32

800

Average Days of Incarceration Time Per Person Sentenced to Time

700 600 Days

500 400 300 200 100 0

1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005

Year 32 See, e.g., Bloom M. (2005): “The Great Reformer: Mario Monti’s Legacy in Article 81 and Cartel Policy”, 1 Competition Policy International 55; Hammond S., supra note 9.


12/10/07

13:10

Page 351


351

Certainly this is the view of the DOJ Antitrust Division, which has worked systematically to increase the imposition of jail time for cartel offenses. The campaign has succeeded at least in terms of the penalties imposed, as is made clear by DOJ’s record: The rising line noted above is the product of the number of sentences imposed and the length of those sentences. The latter has been a special interest of the Division, one in which it has achieved very substantial success:

A little math shows that prison sentences now regularly average more than a year, which must seem a very long time for a successful businessperson. It is no wonder that the Antitrust Division finds that cartel participants try hard to avoid the reach of the U.S. Justice Department.

B. Review of the consequence of expanded competition remedies Remedies have consequences beyond their immediate application. The structure of US remedies has fundamentally affected the nature of the two US competition agencies and the legal standards that those agencies enforce. Of particular significance, only one of the two US competition agencies can engage in criminal enforcement, and that has profoundly influenced the agencies and their respective roles. A brief review of US antitrust enforcement is required in case a novice has stumbled upon this paper. The Department of Justice’s Antitrust Division enforces the Sherman Act and the Clayton Act by bringing cases in federal district courts. It seeks what I have termed bi-modal penalties: when it proceeds civilly, it obtains a non-punitive injunction; when it proceeds criminally, it may obtain (for offenses committed on or after June 22, 2004), up to $100 million per offense or possibly even more from corporations and impose on individuals up to 10 years in prison and fines of up to $1 million or possibly even more.33 The FTC enforces the Clayton Act and the Federal Trade Commission Act, although FTC Act Section 5’s prohibition of “unfair methods of competition” has been interpreted to include everything covered by the Sherman Act, and possibly a little more. The classic FTC competition proceeding features an administrative hearing before an Administrative Law Judge, which, if resolved against the private party, results in a non-punitive injunction.34 In unusual 33 15 U.S.C. § 1. Those statutory maximums are supplemented by the Justice Department’s general authority to seek fines in amounts that are twice the gain or twice the loss from criminal conduct. 18 U.S.C. § 3571(d). In addition, the Justice Department may seek treble damages for injury to the United States in its role as a purchaser. 15 U.S.C. § 15a. Finally, the Justice Department has taken the position (in a non-antitrust case) that it may seek monetary equitable relief in a proper case. 34 For a recent such case, see Polygram Holding, Inc. v. FTC, 416 F.3d 29 (D.C. Cir. 2005).


352

12/10/07

13:10

Page 352


cases, the FTC may seek disgorgement or other monetary equitable relief.35 In addition to federal enforcement, the Sherman Act and Clayton Act (but not the FTC Act) may be enforced by private parties seeking injunctions and/or treble damages, and injunctions and/or damages may also be sought by state attorneys general enforcing the Sherman Act or Clayton Act or a state statute that parallels one or both of those acts or the FTC Act.

1. Dominance of criminal enforcement in non-merger cases Given the attraction of the criminal remedy, it would be surprising if criminal cases did not tend to dominate the non-merger work of the Justice Department. (Mergers are different: an injunction typically completely prevents a merger, so it is a highly effective remedy.) A comparison of the number of criminal and civil non-merger cases filed by the Antitrust Division is striking:

Criminal v Civil Cases Filed 100 90

Total Criminal Cases Filed District Court Sherman ß §1–2 Cases Filed

Number of Cases Filed

80 70 60 50 40 30 20 10

198 6 198 7 198 8 198 9 199 0 199 1 199 2 199 3 199 4 199 5 199 6 199 7 199 8 199 9 200 0 200 1 200 2 200 3 200 4 200 5

0 Year Another way of measuring a litigating enforcement agencies’ activity is by looking at the cases in which it files appellate briefs. In the almost ten years 35 See Calkins S. (2006): “Civil Monetary Remedies Available to Federal Antitrust Enforcers”, 40 University of San Francisco Law Review 567.


12/10/07

13:10

Page 353


353

since calendar 1995, the Antitrust Division has filed appellate briefs in the following: —46 cases (overwhelmingly private) in which one or more amicus briefs were filed; —26 criminal cases; and —15 civil cases.36 The briefing in a majority of those civil cases focused on what might loosely be called procedural issues, typically discovery or the review of a consent decree. Only six cases resulted in a judicial opinion on the antitrust merits— although several of those are doctrinally important.37 Having suggested that the availability of the criminal remedy has understandably tended to concentrate the efforts of the Antitrust Division on criminal cases, one has to recognize a major caveat. Conservatives and liberals alike agree on a core of antitrust enforcement against hard core cartels and anticompetitive mergers (although they may differ somewhat on which mergers are anticompetitive). There is not uniform agreement, however, on what non-merger activity short of a hard-core cartel is anticompetitive. Of those six cases in which an appellate court handed down a merits decision on an antitrust case, for instance, one was a merger case and the other five were all cases filed under President Clinton. It seems likely that the availability of criminal remedies tends to concentrate non-merger enforcement on activity appropriate for criminal punishment, but one cannot be sure of how large is the universe of non-criminal conduct that various Assistant Attorneys General have considered deserving of challenge at all.

2. Consequences for law development At one time, Justice Department cases comprised much of what was meant by U.S. antitrust law. On issue after issue, the leading case had been brought by the Justice Department: in mergers,38 boycotts,39 joint ventures,40 36 Source: http://www.usdoj.gov/atr/public/appellate/appellate.htm (computations done August 22, 2006). Each case was counted only once, even if the Division filed multiple briefs. 37 The six cases are United States v. Dairy Farmers of Am., Inc., 426 F.2d 850 (2005); United States v. Dentsply International, Inc., 399 F.3d 181 (3d Cir. 2005); United States v. LSL Biotechnologies, 379 F.3d 672 (9th Cir. 2004); United States v. VISA USA, Inc., 344 F.3d 229 (2d Cir. 2003); United States v. AMR Corp., 335 F.3d 1109 (10th Cir. 2003); United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001). 38 United States v. General Dynamics Corp., 415 U.S. 486 (1974); United States v. Philadelphia National Bank, 374 U.S. 321 (1963); Brown Shoe Co. v. United States, 370 U.S. 294 (1962). 39 United States v. Topco Associates, Inc., 414 U.S. 801 (1973); Eastern States Retail Lumber Dealers’ Ass’n v. United States (1914). 40 United States v. Penn-Olin Chemical Co., 378 U.S. 158 (1964); Associated Press v. United States, 326 U.S. 1 (1945).


354

12/10/07

13:10

Page 354


monopolization,41 attempted monopolization,42 market division,43 proof of agreement,44 information exchanges,45 distribution,46 vertical refusals to deal,47 tying,48 price fixing,49 and the rule of reason.50 In large part, US antitrust law was the law that was successfully enforced by the Justice Department. The best-selling antitrust casebook’s selection of leading cases51 shows DOJ’s dominance when looking at cases decided before 1977 (the landmark year of Continental T.V., Inc. v. GTE Sylvania, nc.,52 and Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc.53): Leading Cases Decided Before 1977 25

Cases

20 15 10 5 0 DOJ

FTC

Private

State

41 United States v. Grinnell Corp., 384 U.S. 563 (1966); United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377 (1956).; United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir. 1945). 42 Lorain Journal Co. v. United States, 342 U.S. 143 (1951). 43 Timpkin Roller Bearing Co. v. United States (1951); United States v. Sealy, Inc. (1967). 44 Interstate Circuit, Inc. v. United States, 306 U.S. 208 (1939); United States v. American Tobacco Co., 328 U.S. 781 (1946). 45 United States v. Container Corp. of America, 393 U.S. 333 (1969); Sugar Institute, Inc. v. United States (1936); Maple Flooring Mfrs. Ass’n v. United States (1925); United States v. American Linseed Oil Co. (1923); American Column & Lumber Co. v. United States, 257 U.S. 377 (1921). 46 United States v. Arnold, Schwinn & Co., 388 U.S. 365 (1977). 47 United States v. Colgate & Co., 250 U.S. 300 (1919). 48 International Salt Co. v. United States, 332 U.S. 392 (1947). 489 United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940); United States v. TransMissouri Freight Ass’n, 166 U.S. 290 (1897); United States v. Addyston Pipe & Steel Co., 85 F. 271 (6th Cir. 1898). 50 Chicago Board of Trade v. United States, 246 U.S. 231 (1918). 51 Pitofsky R., Goldschmid H. and Wood D. (2003): Cases and Materials on Trade Regulation, 5th edition, Foundation Press, New York, xxiii-xxxvii (data limited to US cases). 52 433 U.S. 36 (1977). 53 429 U.S. 477 (1977).


12/10/07

13:10

Page 355


355

Thereafter, two changes occurred. First, the Division’s criminal enforcement program gradually expanded.54 In 1974 Congress promoted antitrust violations from misdemeanors to felonies and dramatically increased criminal penalties. Then in the 1980s the Reagan Administration made criminal antitrust a centerpiece, and the dramatically strengthened leniency programs introduced in 1993 (corporate)55 and 1994 (individual)56 helped make “international cartel” a common term. During the past 20 years, DOJ antitrust enforcement has been at least numerically overwhelming criminal, as indicated in the chart on cases filed. Second, the 1980s saw Ronald Reagan bring the Chicago School to Antitrust Enforcement, and whether the years thereafter saw swings or continuity, the agencies never returned to the swashbuckling days of yore. Today what is known as US antitrust law no longer is exclusively or even principally the consequence of Justice Department enforcement. The leading modern cases on monopolization,57 attempted monopolization,58 joint ventures,59 proof of agreement;60 boycott;61 other horizontal restraints of trade,62 resale price maintenance,63 territorial restraints,64 vertical boycott claims,65 tying,66 price discrimination,67 jurisdiction,68 and exemptions69 are 54 See Baker D. (2001): “The Use of Criminal Law Remedies to Deter and Punish Cartels and Bid-Rigging”, 69 George Washington Law Review 693; Kovacic W. (2003): “The Modern Evolution of U.S. Competition Policy Enforcement Norms”, 71 Antitrust Law Journal 377, at pp. 418–23. 55 DOJ Antitrust Division Corporate Leniency Policy, August 10, 1993, available at http://www.usdoj.gov/atr/public/guidelines/0091.htm. 56 DOJ Antitrust Division Leniency Policy for Individuals, August 10, 1994, available at http://www.usdoj.gov/atr/public/guidelines/0092.htm. 57 Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, 540 U.S. 398 (2004); Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (June 21, 1993); see also Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., 126 S. Ct. 2965 (2006) (granting certiorari in monopolization case). But compare United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) (en banc) (unusually significant court of appeals opinion). 58 Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447 (1993). 59 Texaco Inc. v. Dagher, 126 S. Ct. 1276 (2006). 60 Matsushita Electric Indus. Co. v. Zenith Radio Corp., 475 U.S. 574 (1986); see also Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752 (1984) (single entity); Bell Atlantic Comp. v. Twombly, 126 S. Ct. 2965 (2006) (granting certiorari in agreement case). 61 Hartford Fire Ins. Co. v. California, 509 U.S. 764 (1993); Northwest Wholesale Stationers, Inc. v. Pacific Stationery and Printing Co., 472 U.S. 284 (1985). 62 California Dental Ass’n v. FTC, 526 U.S. 756 (1999); FTC v. Superior Court Trial Lawyers Ass’n, 493 U.S. 411 (1990); FTC v. Indiana Federation of Dentists, 476 U.S. 447 (1986); NCAA v. Board of Regents of University of Georgia, 468 U.S. 85 (1984); Broadcast Music, Inc. v. CBS, Inc., 441 U.S. 1 (1979). 63 State Oil Co. v. Khan, 522 U.S. 3 (1997); Business Electonics Corp. v. Sharp Electronics Corp., 485 U.S. 717 (1988). 64 Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36 (1977). 65 NYNEX Corp. v. Discon, Inc., 525 U.S. 128 (1998). 66 Illinois Tool Works Inc. v Independent Ink, Inc., 126 S. Ct. 1281 (2006). 67 Volvo Trucks North America, Inc. v. Reeder-Simco GMC, Inc., 126 S. Ct. 860 (2006); Texaco v. Hasbrouck, 496 U.S. 543 (June 14, 1990); Great Atlantic & Pac. Tea Co. v. FTC, 440 U.S. 69 (1979). 68 F. Hoffman-La Roche Ltd. v. Empagran S.A., 542 U.S. 155 (2004); Summit Health, Ltd. v. Pinhas, 500 U.S. 322 (1991). 69 E.g., Professional Real Estate Investors, Inc. v. Columbia Pictures Industries, Inc., 508 U.S.


356

12/10/07

13:10

Page 356


almost all the result of litigation brought by someone other than the Justice Department.70 Indeed, in the past twenty years, no Justice Department case has been decided by the Supreme Court.71 In comparison, during that same time period the Court has decided four FTC cases72 and one state case.73 The leading cases in that antitrust book now come from very different sources:

Leading Cases Decided 1977 and After 35 30

Cases

25 20 15 10 5 0 DOJ

FTC

Private

State

Another way of approaching this issue is to identify the leading unsettled issues in antitrust and observe the role being played by the two agencies. Any list of such issues would likely include the following: a. Mergers. In one sense merger standards are reasonably settled, in that the Guidelines have remained unchanged through two Presidential administrations. On the other hand, it is widely understood that the numerical thresholds contained in those Guidelines do little to indicate enforcement probabilities. More fundamentally, the antitrust agencies’, recent string of 49 (1993); FTC v. Ticor Title Insurance Co., 504 U.S. 621 (June 12, 1992); City of Columbia v. Omni Outdoor Advertising, Inc., 499 U.S. 365 (1991); FTC v. Superior Court Trial Lawyers Ass’n, 493 U.S. 411 (1990). 70 Except for the very important Microsoft opinion, examples are limited to Supreme Court cases, which by definition tend to be the most important opinions. For an earlier list, see Calkins S., supra note 2, at 155. 71 The preceding ten years saw only three Supreme Court opinions in Justice Department cases: Southern Motor Carriers Rate Conference v. United States, 471 U.S. 48 (Mar. 27, 1985) (upholding an exemption); United States v. United States Gypsum Co., 438 US 422 (June 29, 1978) (affirming reversal of criminal conviction); and National Society of Professional Engineers v. United States, 435 U.S. 679 (Apr. 25, 1978) (bidding agreement unlawful in civil case). Of course the antitrust agencies and especially the Justice Department routinely file amicus briefs in Supreme Court antitrust cases, so their voice can be heard even if they are not litigants. 72 California Dental Ass’n v. FTC, 526 U.S. 756 (1999); FTC v. Ticor Title Insurance Co., 504 U.S. 621 (1992); FTC v. Superior Court Trial Lawyers Ass’n, 493 U.S. 411 (1990); FTC v. Indiana Federation of Dentists, 476 U.S. 447 (1986). 73 Hartford Fire Ins. Co. v. California, 509 U.S. 764 (1993).


12/10/07

13:10

Page 357


357

court defeats74 leaves uncertain just how much evidence is needed to block a merger. Here, the two antitrust agencies bring the vast majority of challenges, and to the extent that law is made in courts, the agencies tend to make it and fully to share the responsibility for doing so. Of course, it is with mergers that the agencies have the most effective civil remedies (at least when courts can be persuaded to grant relief ). b. Dominant firm conduct. When competition authorities on both sides of the Atlantic are working on reports and revisiting the proper approach to dominant firm behavior, the unsettled nature of the legal standards is evident. In one sense, the Justice Department has played only a minor role in litigating these issues: in the most recent 20 fiscal years, it has filed a total of only 11 Section 2 (dominant firm) cases.75 This picture is deceptive, however, since nine of those 11 were filed during the eight years of the Clinton administration, which saw the filing of significant cases such as Microsoft, AMR, and Dentsply. The Clinton years also saw an FTC more active on the dominant firm front, filing cases such as Intel,76 Dell Computer,77 and Schering-Plough. At the FTC the picture is more mixed, however, since the Bush II FTC issued important dominant firm complaints in Union Oil and Rambus. Thus, although the key issues in dominant firm law are being litigated by private parties,78 the government has not been entirely absent—and to some extent its relatively modest role may simply reflect a policy preference by Republican enforcers. c. Proof of Agreement. Given that horizontal agreements are often considered suspect at least when they concern price or output, the distinction between unilateral and concerted action is fundamental. This heavily litigated issue sees private parties at the forefront.79 The FTC has brought some cases, settled by consents, that addressed signaling or attempted price fixing.80 Justice Department cases have focused on proving criminal conspiracy, which demands a higher standard of proof and focuses much less on drawing inferences and much more on evaluating disputed testimony.81 74 United States v. Oracle Corp., 331 F. Supp. 2d 1098 (N.D. Cal. 2004); FTC v. Arch Coal, Inc., 329 F. Supp. 2d 109 (D.D.C. 2004); United States v. Sungard Data Systems, Inc., 172 F. Supp. 2d 172 (D.D.C. 2001). 75 Antitrust Division Workload Statistics. 76 Intel Corp., Dkt. No. 9288, 1999 FTC LEXIS 38 (consent order 1999). 77 Dell Computer Corp., 128 F.T.C. 151 (1999) (consent order). 78 See, e.g., Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, 540 U.S. 398 (2004); Spirit Airlines, Inc. v. Northwest Airlines, Inc., 431 F. 3d 917 (6th Cir. 2005); Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., 1265. cf. 2965/2006 (granting certiorari in monopolization case); LePage’s, Inc. v. 3M, 324 F.3d 141 (3d Cir. 2003) (en banc). 79 See, e.g., Bell Atlantic Corp., v. Twombly, 1265. cf. 2965 (2006) (granting certiorari); Blomkest Fertilizer, Inc. v. Potash Corp. of Saskatchewan, 203 F.3d 1028 (8th Cir. 2000). 80 Toys “R” Us, Inc. v. FTC, 221 F.3d 928 (7th Cir. 2000); Valassis Communications, Inc., Dkt. C-4160, 2006 FTC LEXIS 25 (April 19, 2006); Stone Container Corp., 125 F.T.C. 853 (1998) (consent order). 81 See, e.g., United States v. Rose, 2006 U.S. App. LEXIS 11679 (5th Cir. 2006) (affirming verdict but vacating sentence); United States v. Therm-All, Inc., 373 F.3d 625 (5th Cir. 2004) (examining testimony to apply statute of limitations).


358

12/10/07

13:10

Page 358


d. Quick Look/Structured Rule of Reason. The great, on-going question of whether there is some middle category of analysis—between the per se rule and what I have called the “Full Monty” rule of reason82 is closely associated with the Federal Trade Commission, which has brought cases such as Polygram,83 California Dental,84 Indiana Federation of Dentists,85 and Superior Court Trial Lawyers.86 Private cases also have played a significant role.87 Although a seminal 1978 DOJ case helped launch this issue,88 since then DOJ cases have been conspicuous by their absence.89 One gets the sense that the DOJ’s priority is understandably to preserve the power of the per se rule and criminal enforcement of clear agreements, rather than to develop more nuanced ways to analyze more ambiguous behavior. e. Intellectual Property. In intellectual property, too, the case law is being developed by private parties and the FTC. Leading current cases include Illinois Tool Works,90 Tamoxifen,91 and other private cases,92 plus such FTC cases as Schering 93 and Rambus Inc.94 f. Vertical Restraints and Price Discrimination. Currently developing case law for vertical restraints and price discrimination is almost exclusively private.95 The FTC has entered some consent orders96 but has not contributed to the developing case law, although the Toys R Us case,97 which ended up treated as a horizontal group boycott, originated in a complaint that raised vertical issues. 82 Calkins S.: (2000): “California Dental Association: Not the Quick Look But Not the Full Monty”, 67 Antitrust Law Journal 495. 83 Polygram Holding, Inc. v. FTC, 416 F.3d 29 (D.C. Cir. 2005). For a recent application of Polygram, see North Texas Specialty Physicians, FTC Dkt. 9312 (Dec. 1, 2005, modified Jan. 20, 2006) (on appeal). 84 California Dental Ass’n v. FTC, 526 U.S. 756 (1999). 85 FTC v. Indiana Federation of Dentists, 476 U.S. 447 (1986). 86 FTC v. Superior Court Trial Lawyers Ass’n, 493 U.S. 411 (1990). 87 NCAA v. Board of Regents of University of Georgia, 468 U.S. 85 (1984); Continental Airlines, Inc. v. United Airlines, Inc., 277 F.3d 499 (4th Cir. 2002). 88 National Society of Professional Engineers v. US, 435 US 679 (1978). 89 But compare United States v. Brown University, 5 F. 3d 658 (3d Cir. 1993) (unusual case holding that abbreviated analysis was inappropriate for reviewing Ivy League agreement not to compete on financial aid packages because of the special nature of higher education and the agreement’s benefits). 90 Illinois Tool Works Inc. v Independent Ink, Inc., 126 S. Ct. 1281 (2006). 91 Joblove v. Barr Laboratories, Inc., 429 F.3d 370 (2d Cir. 2006). 92 E.g., ISO Antitrust Litigation, 203 F.3d 1322 (Fed. Cir. 2000); Image Technical Services v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir. 1997). 93 Schering-Plough Corp. v. FTC, 402 F.3d 1056 (11th Cir. 2005) (petition for cert. pending). 94 Rambus Inc., Dkt. 9302 (FTC Aug. 2, 2006). 95 E.g., Volvo Trucks North America, Inc. v. Reeder-Simco GMC, Inc., 126 S. Ct. 860 (2006); State Oil Co. v. Khan, 522 U.S. 3 (1997); Stop & Shop Supermarket Co. v. Blue Cross & Blue Shield, 373 F.3d 57 (1st Cir. 2004). 96 McCormick & Co., 2000 FTC LEXIS 43 (consent order) (Robinson-Patman Act exclusive dealing case); Time Warner Inc., 2000 FTC LEXIS 40 (consent order) (resale price maintenance); RX Care of Tenn., Inc., 121 F.T.C. 762 (1996) (consent order) (most favored nation’s clauses). 97 Toys “R” Us, Inc. v. FTC, 221 F.3d 928 (7th Cir. 2000).


12/10/07

13:10

Page 359


359

Because two changes occurred at the Justice Department—criminal enforcement came of age and the Chicago School changed antitrust thinking—one cannot be sure how much difference to attribute to which cause. On some key issues, private litigants are making law because they are bringing cases that the DOJ does not support. Only this term, the Supreme Court decided three private cases, each time agreeing with the DOJ’s (and FTC’s) recommendation to side with the defendant. Hence, antitrust policy preferences surely help explain why the DOJ has played a less prominent role in developing law through litigation. On the other hand, the FTC, headed by persons appointed by the same Presidents appointed DOJ officials, seems to bring more cases that develop antitrust law. Even here the difference may be explained by policy preferences, as suggested by the DOJ Solicitor General’s opposition to the FTC’s petition for certiorari in Schering.98 Yet so regularly does the FTC, under liberal and conservative leadership alike, explore novel issues on the edges of antitrust law, that one must conclude that the two agencies behave differently in part because only one can bring criminal cases and impose incarceration. (One can leave for another day the question whether this is good or bad for antitrust.)

3. Consequences for legal standards When the US Supreme Court recently rejected the notion that one can presume market power from a patent for tying purposes, it pointed out that the applicable statute (Sherman Act Section 1) made “the conduct at issue a federal crime punishable by up to 10 years in prison” as a “felony.”99 In context, this was quite a stretch, since there is no chance that any DOJ that we could imagine would ever challenge mere tying as a criminal violation.100 But the point was simple and indisputable: consequences matter, and as the consequences of finding a violation become more serious, courts become correspondingly more reluctant to find a violation. That antitrust is subject to this normal “equilibrating tendency” is a point I have made before and that I will not develop at any length herein. Precisely because proof of a vertical price agreement is per se illegal under US law, it is almost impossible for a plaintiff to prove a vertical price agreement.101 Because allowing a jury to find a horizontal agreement could result in the imposition of treble damages, US courts are remarkably willing to use summary judgment to keep ambiguous evidence away from juries.102 US 98

Amicus Brief (May 2006), available at http://www.usdoj.gov/atr/cases/f216300/216358.htm. 126 S. Ct. at 1291. 100 But compare United States v. Cuisinart, Inc., 1981–1 Trade Cas. (CCH) 63,979 (D. Conn. 1981) (consent decree in criminal challenge to resale price maintainance). 101 See Business Electonics Corp. v. Sharp Electronics Corp., 485 U.S. 717 (1988). 102 See Matsushita Electric Indus. Co. v. Zenith Radio Corp., 475 U.S. 574 (1986). 99


360

12/10/07

13:10

Page 360


predatory pricing law deliberately errs on the side of defendants out of fear that treble damages awards could discourage aggressive pricing.103 One would think that if the existence of private enforcement and treble damages awards have driven courts to adopt legal rules protective of defendants, the even more severe sanction of criminal penalties would have done the same. The Supreme Court has emphasized the importance of proving the mens rea needed to commit an antitrust crime.104 Beyond that, however, little of what is know as modern antitrust law has been developed in the criminal context, because the modern DOJ has used its criminal authority only in cases too clear to contribute to the development of antitrust doctrine. Were criminal powers to be used more aggressively, however, the courts could be expected to respond.

C. Preliminary reflections about the European situation One ought always to be nervous about drawing comparisons and finding lessons in other situations. The following four points are offered in that light: what a review of the above might suggest as points of departure for Europe.

1. Equilibrating tendencies matter No more in Europe than in the US can one change penalties without affecting other parts of the competition law system. When Europe’s competition enforcement system was dominated by a single, largely administrative approach without serious penalties, legal standards could be general, leaving the public comfortable in the knowledge that the standards would be enforced with sensible discretion. Now that has changed: penalties are increasingly severe, the law is being applied by national courts, and private enforcement is being encourages. Unless Europe is very different from the US, it follows inevitably that legal standards will become increasingly clear. Additional changes would have additional consequences. In particular, any ratcheting up of penalties and other consequences will lead to increased demand for clarity.

2. Criminal and individual penalties are more powerful deterrents than civil and corporate ones, and jail is the ultimate penalty If the only goal is deterrence, as many competition authorities as possible would impose criminal penalties, including jail time, on individuals. 103 104

See Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993). United States v. United States Gypsum Co., 438 U.S. 422 (1978).


12/10/07

13:10

Page 361


361

Deterrence is not the only goal, however. One also has to work within the existing tradition and achieve what is possible. It may well be that the European Commission simply will not be given criminal competition authority—that such a move would not be accepted by the politicians or the public. Even if true, the Commission could still increase deterrence by emphasizing individual responsibility. Some fines on individuals may be undone by supportive employers, but individuals who lose their jobs, are barred from serving as directors, and do not enjoy the luxury of having their attorneys fees paid may still suffer very real penalties. Especially intriguing is the question of corporate criminal liability. In the US, substantial corporations regularly plead guilty to the crime of price fixing while continuing to function otherwise effectively in the interests of their shareholders, whereas criminal entity liability for securities fraud is often regarded as potentially fatal (and, indeed, a mere indictment ended Arthur Anderson). Of critical consequence may be what follows from a conviction: an accounting firm or law firm may cease being able to function; a health care firm may become ineligible to be paid by the government. Before proceeding too far to make criminal corporate enforcement the norm in Europe (including in an individual country), one would want to know the consequences of criminally charging and/or convicting various possible defendants.

3. Greater penalties may tend to dominate There are two senses in which the availability of greater penalties may come to dominate. First, an agency that enjoys a range of possible sanctions will tend to use the more serious ones. If an agency can expend resources challenging either behavior that can result in jail time, or behavior that can result in nothing more than an injunction not to sin in the future, inevitably it will tend to favor the former. Public servants try to make a difference, so given a choice between two courses of action, one of which will make a very large difference, one a very small difference, the choice is usually easy. Dominance is also seen between agencies. The FTC consciously refrains from challenging naked cartels because the DOJ enjoys such a serious remedy. Conversely, defense lawyers jump a little quicker when getting a call from a DOJ lawyer working on a grand jury than when the FTC is calling about a matter that can lead only to a non-threatening injunction. Certainly there is a chance that European enforcers with more potent remedies may come to dominate.


362

12/10/07

13:10

Page 362


4. The role of the European Commission may be very different in the future This is such an exciting time in the development of competition enforcement in Europe. As others have noted, the role of the Commission is changing and may change quite drastically. If criminal enforcement with serious penalties becomes a reality in several national systems, would the Commission, like the US FTC with DOJ, feel obliged to refer evidence of cartel behavior to another enforcer? Might the Commission wait for national enforcement to end, and then supplement national enforcement by seeking monetary payments— much as US state or private enforcers follow federal government actions? Or would national courts achieve full relief—leaving the Commission, like the US FTC, to challenge some mergers and questionable conduct less egregious than cartels? Would such an assignment push the Commission to experiment with more novel theories than it otherwise might? The working out of equilibrating tendencies is especially fascinating. The U.S. antitrust system has largely managed the trick of treating Sherman Act Section 1 as two different statutes, one criminal and narrow, one civil and much broader. It has been able generally to ignore the inconvenient fact that Sherman Act Section 2 declares that monopolization and attempted monopolization are felonies: the modern DOJ does not bring such cases and courts interpret Section 2 as if it were an entirely civil statute (albeit one for which treble damages are available). What will happen in Europe if the Commission continues to develop competition law as largely administrative law with relatively modest penalties, while national courts apply the same treaty provisions in criminal cases, cases with severe individual penalties, and/or cases brought by private parties? Those courts would tend to constrict the reach of the relevant Articles. It would be ironic if all of these major changes occurring in Europe with the assistance or at the instigation of the Commission helped move us to a world in which the Commission felt constrained to defer to national enforcers while national courts work to constrict the reach of Articles 81 and 82 EC.

D. Conclusion Criminal enforcement and private enforcement are both powerful forces for increased competition law deterrence. They also are powerful forces for changing the roles of competition agencies and the substantive standards applied by them. The US experience offers an intriguing set of questions about how matters may evolve in Europe.

(S) Langus and Motta

16/10/07

13:59

Page 363

V Gregor Langus1 and Massimo Motta2 On the Effect of EU Cartel Investigations and Fines On the Infringing Firms’ Market Value3

A. Introduction 1. Motivation Fines are a crucial tool for competition authorities. Only if the fines (and more generally the costs that firms incur when found guilty of antitrust infringement) are large enough will the firms be deterred from engaging in cartels and other anticompetitive behaviour. Under EC competition law, there are no criminal penalties, and private actions are very rare, but fines can be very large; up to 10% of the annual aggregate worldwide turnover of the group. Yet anecdotal evidence often suggests that the impact of antitrust investigations and fines may not have a considerable impact on the firms which are caught infringing EC competition law. Indeed, a large number of firms (and indeed, firms from the sample we analyze in this paper) are repeat offenders. Moreover, adverse Commission decisions and Community Court judgments do not seem to trigger management changes very often. This raises the question of the extent to which firms are seriously affected by the fines they receive, and by the expectation of receiving such fines. In this paper, we carry out an empirical analysis to explore the effect of Commission antitrust fines and of the opening of Commission investigations on the share prices of a sample of firms which have been involved in cartel cases. Our research strategy is the following: we look for the exact dates on 1

European University Institute, Florence, Italy. European University Institute, Florence, Italy and Università di Bologna. 3 This paper is a shorter and non-technical version of our research paper (Langus and Motta, 2006), which examines all antitrust infringements, whereas here we focus only on cartels. As with the main paper, we wish to acknowledge comments by a number of people who helped us in various ways on this project. In particular, we are grateful to seminar audiences at K.U. Leuven, the EUI Economics Department and in Tel Aviv, as well as Walter Mölls, Richard Spady, Frank Verboven, and especially Andrea Ichino, Liliane Karlinger and Kirti Mehta. Comments from and discussions with participants at the 11th Annual EUI Competition Law Workshop were also particularly helpful. Finally, we are grateful to Francesca Bassetti and Giacomo Pansolli for their research assistance. The law is stated as of 31 May 2006. 2


364

16/10/07

13:59

Page 364


which some critical events of an Commission antitrust investigation take place, and—by using standard event study techniques—we estimate the impact that this new information (the event) has on the market prices for shares of the firms involved. The objective of the paper is to promote a debate on whether antitrust law infringements carry large enough penalties for the antitrust offenders, and therefore more generally on whether there is enough antitrust deterrence.

2. Literature To our knowledge, no previous work attempts to estimate the effects of EU fines and infringement decisions on offending firms.4 In contrast, Bosch and Eckard (1991) use a similar methodology on US data to estimate the effect an indictment for price fixing has on a firm’s stock market price. These authors find that the shares of indicted firms in their sample lose, on average, 1.08% of their value in the days immediately after the public announcement of the indictment. They also estimate that fines and damages account for only 13% of the total loss of the stock market value caused by the firm’s antitrust indictment, the remaining loss being due to loss of profits, which could result from any of the sources identified below.5 As far as our methodology is concerned, there is an ample literature, especially in finance, which uses event study techniques to assess the impact of particular events on share prices (or other variables). Standard references in this literature are Brown and Warner (1980, 1985) and Campbell et al. (1997).

3. EU fining policy The European Commission is the main enforcer of competition law in the EU. The Commission’s investigative powers are established by Regulation 1/2003, and they include the possibility to conduct inspections (so-called dawn raids) not only on the firms’ premises but also, under certain conditions, in the homes (and private vehicles) of the firms’ managers and employees. Under EC competition law, fines can be imposed only on firms as opposed to natural persons (although national laws in some European countries do 4 For empirical analyses of the effects of EU merger notifications and decisions, which make use of the event study methodology also used here, see Duso, Neven and Röller (2006), and Duso, Gugler and Yurtoglu (2006). 5 A comparison of these results with any EU results would be difficult, since in the US—unlike the EU—there are prison sentences for conspiracies (cartels), as well as a wide use of private actions for damages. Both raise the deterrence power of investigations well above the expected effect of the fines themselves.


16/10/07

13:59

Page 365

EU Cartel Investigations and the Infringing Firms’ Market Value

365

allow criminal penalties, and/or administrative sanctions, to be imposed on firms’ managers), and Regulation 1/2003 establishes that fines may not exceed 10 percent of the firm’s annual aggregate worldwide turnover (although actual fines rarely come anywhere close to that ceiling), and that they should be fixed with regard to the gravity and the duration of the infringement. The Commission has considerable discretion in the determination of the fines. In practice, it is widely believed that the Commission has toughened its stance against cartels over time (for a detailed study of the Commission’s recent fining policy, see Geradin and Henry, 2005). Recently in the EU there have been some discussions about making collusion a criminal offence. If this were the case, executives found guilty of collusive agreements could be imprisoned, as in the US or in the UK. Arguably, this might provide a stronger deterrent against collusion, as riskaverse managers would find it very dangerous to collude. However, it is unlikely that there will be sufficient support by EU Member States for a reform of EC competition law that introduces criminal penalties. More feasible is perhaps another way to increase a firm’s expected costs if caught colluding, namely, the recovery of damages in private actions before national courts (which in the US can amount to treble damages). So far, private actions have been rare in EC competition law, but there seems to be a tendency towards an increase of such actions (see other contributions to this Workshop). If this leads to important damages being recovered by clients and consumers hurt by cartels, the effect will be to substantially add to the fines that firms have to pay, thereby increasing deterrence. Another important recent development in the fining policy of the Commission is the adoption of a leniency programme, which grants total or partial immunity from fines to firms that collaborate with the Commission in cartel investigations.6

4. Reasons why firms may be hurt by (actual or expected) fines and negative decisions As indicated above, the main idea of this paper is to estimate the impact of some typical antitrust events on share prices, such as a dawn raid, a Commission Decision and a Community Court Judgment (if there is an appeal of the Commission’s Decision). There are several reasons why news about actual or expected fines may have a negative effect on a firm’s market value.

6 For an assessment of the Commission leniency programme, see Geradin and Henry (2005), as well as the contributions to this Workshop. For a textbook analysis of leniency programmes, see Motta (2004).


366

16/10/07

13:59

Page 366


First, the payment of a fine will diminish the cash held by the firm, directly affecting its value. Second, under imperfect financial markets and limited financial resources, outside investors will give credit to a firm depending on the assets it owns. The payment of a fine will reduce the firm’s assets, and hence reduce its possible credit, leading the firm to give up some profitable projects for lack of funding. Third, the opening of an antitrust investigation, and a fortiori a negative Commission Decision, may signal to the market that a profitable activity (e.g., a cartel) will have to end. Fourth, an antitrust investigation may be extremely lengthy, and it may lead the firm not only to incur huge legal costs but also to devote internal (managerial) resources to the legal proceedings rather than use them in more productive activities. Finally, an antitrust investigation may in some circumstances give the firm some negative publicity—which could reduce consumer demand and brand loyalty.

B. Events in a typical investigation, and event study methodology 1. Events in an EC antitrust investigation The Commission begins an investigation either at its own initiative or on the basis of a complaint from a third party (although, if a complaint is lodged, the Commission has no obligation to start an antitrust procedure). There is generally no announcement that an investigation has started, and there is no precise time frame in which to carry one out. Indeed, years may pass between the moment the Commission starts to think of a possible infringement and the moment formal steps, such as a Statement of Objections, are taken. If, during the preliminary stages, the Commission has serious doubts that there has been an antitrust infringement, it can carry out a surprise inspection, also called (as noted above) a dawn raid, on the premises of the firm (or firms), to gather documentary evidence. Such evidence is absolutely crucial in cartel cases, but it is also relevant for abuse cases. Next, the Commission has to send a Statement of Objections (SO) to the firms under investigation. In the SO, the Commission sets forth its allegations in relation to the practices of the firm and calls on the firm to respond. After analyzing all the evidence and hearing from the parties, the Commission will take a Decision, which may be reached long after the SO is issued. Note that the Decision is a collegial decision of the entire college of Commissioners, and that before taking it several bodies are consulted, such as representatives of national competition authorities and members of other


16/10/07

13:59

Page 367


367

directorates general within the Commission. Although all the people involved are bound by confidentiality clauses, leakages about (or speculations on) the content of the decision and the level of the fines are common. Firms which have been fined can appeal to the Community Courts, which can review the merits of the Commission’s Decision, and whose Judgments can annul, reduce, uphold or even increase the fine.7 Of course, the Community Courts can also annul or uphold, completely or partly, the overall Decision. The last column of the Table in Annex I summarizes the fines as they appeared in the first Court judgments. To a large extent, the decisions taken by the Court are secret until the moment they are announced, although in some cases there may be signs of the judges’ views.8

2. The event study methodology We use standard event study methodology to investigate the effect of the antitrust investigation on the firm’s share price. More particularly, we try to do so by analyzing the effect of the three main events in the investigation procedure identified above: (i) the dawn raid, (ii) the Commission Decision, and (iii) the Court’s judgment. For each of these events, there is a precise date on which they occur, even if in some cases we do not know it (and when this happens, the observation is dropped). Note also that surprise inspections do not always take place and firms may decide not to appeal. However, these events differ in the extent to which they represent a genuine surprise to investors. In other words, some of the events may have been expected and reflected in the price of the relevant securities. In such circumstances, the event dates are not good proxies for the time when the news about the (expected) event reached the market. Thus, our analysis would lead us to reject the null hypothesis of no effect less often than it should be (when the news of the event reaches the market it may trigger an effect, but by using the date of the event, which is anticipated by the market, we are unable to catch this effect). (i)

From our discussion above, the date of the surprise inspection, or dawn raid, seems to represent a genuine surprise for the investors. To test

7 We are aware of only one case in which a fine was increased. In this case, the Court of Justice raised the amount of the fine imposed on SGL Carbon AG, saying that the Court of First Instance had granted an excessive reduction of the fine in its earlier judgment concerning the Decision of the Commission. Cited in NERA, Global Antitrust Weekly, 24–30 June 2006, issue 393. 8 In particular, the opinion of the Advocate General often (though not always) anticipates the judgment of the Court. However, there are Advocates General only among the members of the Court of Justice; the Court of First Instance does not have any.


368

16/10/07

13:59

Page 368


this conjecture, we examined past issues of the Financial Times for any news about a (potential) investigation before the inspection took place, but we found no such news concerning any of the firms for which we have dates of the raid.9 Because the surprise inspection may allow the Commission to find incriminating evidence, and because it is typically done only after the Commission already has some reason to suspect an infringement, this event is likely to signal that a negative Decision of antitrust infringement will be taken.10 Accordingly, a dawn raid should induce investors to revise downward their valuation of the firm. (ii) Next, we investigate the effect of the Commission Decision. As explained above, the market is aware that the Commission is investigating the firm as early as the date of the dawn raid or the Statement of Objections, but the amount of the fine imposed by the Commission in the Decision may be over- or under-estimated. On the other hand, there may be rumours about the level of the fine before the Decision is announced (although there is always some uncertainty, given that it is a collegial decision). (iii) Finally, we investigate the effect of the Court’s judgment. When the judges significantly reduce or annul the fine, this should have a positive effect on the firm’s valuation, and when they confirm it, this should have a negative effect on the firm’s valuation, unless of course the effect of the final Court outcome was already correctly forecast by the investors at the time of the dawn raid or the Decision. To investigate the effects of these events on the firms’ share prices, we use the event study methodology. The main assumption behind this methodology is that financial markets are efficient, in the sense that a firm’s stock market price correctly reflects the value to the shareholders of all the relevant available information about the fundamentals of the firm (balance sheet, investment opportunities, competitive environment and so on). Moreover, under the efficient market hypothesis, any news about the fundamentals is immediately reflected in the firm’s share price. Given this hypothesis, if we knew (1) the moment at which the news about a changed fundamental, or event (for instance, a likely antitrust infringement) 9 It is of course possible that investors may have nonetheless anticipated that an investigation would take place. This may be the case in particular for some of the international cartel cases which appear in our sample, where a US antitrust case precedes an EU investigation. 10 We were unable to check in how many cases the firms that were raided were later found “not guilty” and were thus not fined, because Decisions are adopted only when an infringement is found and the firms are fined. If, in the course of an investigation, the Commission becomes convinced that a firm has not violated the law (or that it does not have enough evidence to prove a violation), there is no public decision or announcement that the investigation has ended.


16/10/07

13:59

Page 369


369

becomes available to the investors, and (2) the share price that would prevail in the absence of this news, we could compute the “value” of the news (and of the change of the fundamental, or event) to the investors and the firm. This is the difference between the counterfactual price and the actual price. It will be immediately obvious to the reader that there are two main difficulties behind the methodology described above. The first is that one has to be sure of the moment when the changed fundamental or event becomes known to the market; the second is that calculating the counterfactual price is not a straightforward exercise, and requires the use of econometric techniques. In particular, to compute the counterfactual share price, we use (see Langus and Motta, 2006 for a formal description of the technique used) the so-called market model, where the return Ri of a particular share i is assumed to be a linear function of the return to the index of the stock market in which that share is quoted, Rm. The econometric model to be estimated is therefore: Rit = ai + biRmt + eit where ai and bi are the parameters and eit is the residual. In order to estimate the parameters ai and bi we use data from a period window prior to and away from the event of interest (e.g., the dawn raid). The econometric estimates so obtained allow us to generate the counterfactual returns in the period of interest, as a function of the actual market return Rm, being a period after the event has taken place: Rˆ it = aˆi + bˆiRmt We can then construct the abnormal returns as the difference between the actual return and the counterfactual return. Finally, we aggregate abnormal returns and compute test statistics (for details, see Langus and Motta, 2006).

C. Data and results 1. Data Our data are culled from the Commission Decisions published in the Official Journal of the European Communities, and from the judgments of the Court of First Instance and the European Court of Justice (published in the European Court Reports and other sources). We looked at all the Decisions resulting in a fine from 1969 until 2005.


370

16/10/07

13:59

Page 370


We retained only decisions involving firms quoted in a stock exchange for which data on share price is available in the Datastream database.11 Our final sample refers to 100 observations, an observation being an infringement by a firm. Some of the firms were repeat offenders. We have exact dates of Commission Decisions for all 100 infringements. We also have dates of Court judgments for 55 infringements, and dates of surprise inspections for 43 infringements. Note that the final data on surprise inspections have been obtained by making the assumption that stock market operators know the industry well, and that they understand that a surprise inspection at the premises of one or more possible participants of a cartel signals a high probability of involvement, and discovery of infringement, also as regards the other firms that later appear in the decision. Therefore, if we had the date of a surprise inspection for one firm, we assigned this date to all the firms that appear in the same cartel case. We thereby had gained 7 additional observations for the dawn raid. If our presumption were false, we would only reduce the significance of our estimates.12 However, due to excessively short time series for 6 observations, we further excluded 6 observations for the dawn raid. The Table in Annex I lists the firms in our sample, and indicates the type of antitrust infringement as well as the dates of the relevant events. The firms in our sample are quoted in different stock exchanges. The majority are quoted in Frankfurt and Tokyo, followed by New York, London and Paris. The remaining stock exchanges where the firms from our sample are quoted are Amsterdam, Korea, Hong Kong, Singapore, Stockholm, Oslo, Brussels, Copenhagen, Milan, Luxembourg, Taiwan, Malaysia, Athens and Vienna.

2. Results The Table in Annex II summarizes the estimation results. Panel 1 contains the average abnormal return by day for a window of 11 days (the event day, indicated by t = 0, as well as 5 days before each event, and five days after the event), for each of the four events considered, and the associated test statistic 11 We are aware that sample selection is a possible concern of our analysis, to the extent that publicly quoted firms tend to be large, multi-product, and possibly multinational firms, for which the effect of a fine related to one particular product and geographic market may well be smaller than for a smaller, single-product firm operating in a domestic market. However, it should also be recalled that the Commission can impose fines up to 10% of the annual aggregate worldwide turnover of a firm, and that proportionality is one of the most important criteria in imposing fines. Thus, other things being equal, a larger multi-product and multinational firm would generally face a larger fine. 12 We have checked that the results are robust to the exclusion of the additional observations obtained in this way.


16/10/07

13:59

Page 371


371

(columns marked by J). Asterisks indicate those estimated returns that are statistically significant. This Panel shows that most daily abnormal returns (that is, the difference between actual and counterfactual returns) are not significantly different from zero. Notably, however, the very same day of the dawn raid there is an estimated average abnormal return of –.011, meaning that there is a decrease of roughly 1.1% in the returns of the share of companies which are raided, and the following day there is a decrease of about .6%. But the unexpected news about a fundamental, in the particular form of the news of an antitrust event, does not necessarily exhaust its effect in two days. Panel 2 of Table 2 shows the cumulative average abnormal returns. The estimated returns are statistically significant for the dawn raid, and indicate that the aggregate (over 11 days) effect of this event is to reduce the return of raided firms by 2.4% on average. We also carry out the same exercise on a shorter window of five days (from t = –1 to t = 3) and find that the cumulative average abnormal return for the dawn raid is of the same order of magnitude.13 There is also a significant effect, on the order of –1.5%, of the Commission’s Decision on the share prices of the raided firms. A Court judgment which annuls or significantly reduces the fine also has a significant impact on the share prices of the firms, which increase by 2.4% following such a judgment. By contrast, if the Court upholds the Commission’s Decision, this has no statistically significant effect on the firms’ valuation.

D. Conclusions Our main result is that the dawn raid (i.e., the surprise inspection of the firm’s premises carried out by the Commission), which is the first piece of information received by market operators indicating that the Commission intends to investigate an antitrust infringement, has a strong and statistically significant effect on the firm’s share price. On average, on the same day as the dawn raid the firm’s return is around 1.1% lower than the counterfactual return provided by the econometric market model; furthermore, the cumulative average abnormal return due to the dawn raid is approximately –2.4%. To this effect, one should add an estimated (statistically significant) decrease of 1.5% when the Decision is adopted, whereas the annulment of the Decision would result in a statistically significant increase by 2.4%. It can be seen that an annulment of the fine does not restore the initial position of the firm. This can be 13 In Langus and Motta (2006), we also perform other robustness tests which confirm the significant drop in returns due to the dawn raid.


372

16/10/07

13:59

Page 372


explained by considering, for instance, a case in which the Court annuls a Commission Decision on procedural grounds: in this case, the fine is annulled, but the possibility that that firm will be able to continue a profitable infringement is considerably lower. If the negative impact of the dawn raid and of the Decision was due more to the market’s anticipation that a profitable (but illegal) business practice would have to be terminated than it was due to the value of the expected fine as such, then even a favourable Court judgment will not bring the value of the firm back to its pre-investigation levels. Another interesting and related question is to understand to what extent the loss in the firm’s market value is due to the fine itself or due to other reasons (e.g., cessation of a profitable activity, legal and managerial costs, and so on). We computed that, on average, the fine represents around 1.6% of firms’ market value as given by Datastream. This value is computed as the value of outstanding shares of primary issue and is weakly lower than the true market capitalization of the firm (which might also have issued other types of shares). By using the estimates of –2.4% as the negative impact of the dawn raid, and of –1.5% as the effect of the Decision, yielding a total loss of 3.9%, the fine would account for about 40% of the total loss in market value.14 Even considering the exploratory nature of this study and all the necessary qualifications, it is difficult to resist the temptation to draw some implications from our results, and to inquire whether the costs a firm expects to bear in the face of an Commission investigation provide enough deterrence for antitrust violations. Deterrence is determined by the probability that an infringement will be uncovered and prosecuted by the Commission multiplied by the costs that the firm will incur if the investigation does take place. While our paper is completely silent on the former factor (which is extremely difficult to estimate), our analysis does tell us something about the latter, which is estimated to be a drop in value of 3.9%. The question is whether a loss in capitalization of this order of magnitude, a loss which of course occurs only if the investigation is started, would provide a sufficient deterrent to antitrust infringements. According to the works surveyed by Connor (2005), a conservative estimate of mean overcharges for cartels is about 25%, and their average duration of 5 years, whereas from Geradin and Henry’s data we see that the average duration of infringements of Articles 81 and 82 EC is of no less than 3.3 years. To fix ideas, let us focus on an overcharge of 25% and a duration of four years for the average infringement. The cost for consumers would be given by: 14 In the US, Bosch and Eckard estimated that fines and damages accounted for 13% of the total loss in the firm’s stock market value. In Motta and Langus (2006), we also considered Article 82 EC cases and found that the fine would account for, at most, 30% of the total loss in market value.


16/10/07

13:59

Page 373


373

4[(⌬p)q+dwl], i.e., the excessive price (⌬p) paid by consumers multiplied by the yearly quantity (q) sold by the firm plus the deadweight loss (dwl), multiplied by the number of years. Since the yearly overcharge is 25% of the relevant market turnover, a fine amounting to 100% of a firm’s annual turnover (in the relevant market) would still be less than the consumer loss, since it does not take into account the deadweight loss. Additionally, 100% of relevant turnover would be a lower boundary, also because, in order to provide enough deterrence, the fine should be augmented to take into account the probability of uncovering antitrust infringements, which is certainly much lower than 1. It is difficult to say how this value compares to the current theoretical rule in the EU, which establishes that fines should be no larger than 10% of the annual aggregate worldwide turnover of the firm: for multi-product firms operating in dozens of markets, this might simply be an exceedingly high fine. In practice, it is likely that the Commission imposes fines by having in mind not the overall turnover of the firm, but rather the relevant market turnover. If 10% of the latter represented the base fine, there is little doubt that this would be an exceedingly low starting point for a fine. Perhaps the Commission should adopt a clear rule based on the relevant market turnover, but use a much higher percentage as a starting point for the calculation of the fine. It is also difficult to say how much deterrence is provided if the antitrust investigation results in, say, a –3.9% loss in stock market value, in part because a portion of this loss is simply due to lower future profits (which is not a penalty), and in part because it is unclear how capitalization relates to the firm’s aggregate turnover (not to speak of the relevant market turnover). To conclude, our paper shows that an antitrust investigation does have a significant economic (and statistical) negative impact on the stock market valuation of a firm. However, whether these costs are large enough compared with the damage that the firm’s practices inflict on society, and whether they constitute a sufficient deterrent, are questions that should be investigated further.


374

16/10/07

13:59

Page 374


Legend for table: Decision: decision date of the Commission; Surp. insp: the date of the surprise inspection; Dec.Court: date of the decision of the Court; fine/capit.: the ratio between fine as set by the commission and capitalization approximation; Fine Comm: fine of the commission in million Euro; Fine Court: fine as set by the Court. Decision

25/11/1980 23/04/1986 23/04/1986 23/04/1986 23/04/1986 21/12/1988 21/12/1988 21/12/1988 21/12/1988 21/12/1988 21/12/1988 21/12/1988 21/12/1988 19/12/1990 19/12/1990 1/04/1992 13/07/1994 21/12/1994 21/12/1994 21/12/1994 21/12/1994 21/12/1994 16/09/1998 16/09/1998 16/09/1998 16/09/1998 16/09/1998 16/09/1998 16/09/1998 14/10/1998 9/12/1998 8/12/1999 8/12/1999 8/12/1999 16/05/2000 16/05/2000 16/05/2000 16/05/2000 16/05/2000 16/05/2000 16/05/2000 16/05/2000

Surp. insp.

13/10/1983 13/10/1983 13/10/1983 13/10/1983

Dec. Court

24/10/1991

21/11/1983 21/11/1983 21/11/1983 21/11/1983 21/11/1983 21/11/1983

27/02/1992 27/02/1992 27/02/1992 27/02/1992 27/02/1992 27/02/1992 27/02/1992 27/02/1992 29/06/1995 29/06/1995

23/04/1991

14/05/1998

27/05/1994 1/12/1994 1/12/1994 1/12/1994

30/09/2003 30/09/2003 30/09/2003 30/09/2003 30/09/2003 30/09/2003 30/09/2003 12/07/2001 11/12/2003 8/07/2004 8/07/2004 8/07/2004 19/03/2003 19/03/2003 19/03/2003 19/03/2003 19/03/2003 19/03/2003 19/03/2003 19/03/2003

index

Firm

116 Johnson & John. 111 BASF 110 Hoechst 118 Shell 98 Solvay 106 Norsk Hydro 105 Solvay 104 BASF 103 Bayer 102 Dow Chemical 101 Hoechst 100 ICI 118 Shell 98 Solvay 97 ICI 96 Nedlloyd 94 SCA Holding 90 Kawasaki Kisen Kai. 89 Mitsui OSK 88 Neptune Ori. 87 Nippon Yus. 86 Orient Overs. 82 Moller-Maersk 81 P & O Nedlloyd 80 Orient Overs. 79 Neptune Orient 78 Nippon Yusen 77 Hanjin Shipp. 76 Hyundai Merch. 75 Tate & Lyle 74 Minoan Lines 72 Vallourec 71 Sumitomo Met. 70 Nippon Steel 69 Kawasaki Kis. 68 Moller-Maersk 67 Malaysia Int. 66 Mitsui OSK 65 Neptune Orient 64 Nippon Yus. Kai. 63 Orient Overs. 62 P & O Nedlloyd

Fine Fine Comm. Court 0.2 2.5 9 9 9.000 2.500 0.75 0.75 3.5 3.5 5.5 0 2.5 0 2.25 0 1 0 3.5 0 0.85 0 7 0 7 0 0.026 2.2 2.2 0.01 0.01 0.01 0.01 0.01 27.5 0 41.26 0 20.63 0 13.75 0 20.63 0 20.63 0 18.56 0 7 5.6 3.26 3.26 8.1 8.1 13.5 10.935 13.5 10.935 0.62 0 0.836 0 0.134 0 0.62 0 0.368 0 0.62 0 0.134 0 1.24 0

fine /cap.

0.00095 0.00038 0.00473

0.00197 0.00073 0.00025 0.00086 4E-05

5E-05

0.00542 0.02432 1E-05 4E-06


16/10/07

13:59

Page 375

EU Cartel Investigations and the Infringing Firms’ Market Value Decision

16/05/2000 16/05/2000 7/06/2000 7/06/2000 7/06/2000 7/06/2000 5/12/2001 5/12/2001 5/12/2001 18/07/2001 18/07/2001 18/07/2001 18/07/2001 18/07/2001 18/07/2001 18/07/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 21/11/2001 5/12/2001 11/12/2001 11/12/2001 11/12/2001 11/06/2002 2/07/2002 2/07/2002 2/07/2002 24/07/2002 24/07/2002 24/07/2002 24/07/2002 30/10/2002 30/10/2002 27/11/2002 27/11/2002 27/11/2002 27/11/2002 1/10/2003 3/12/2003

Surp. insp.

11/06/1997 11/06/1997 11/06/1997 11/06/1997

5/06/1997 5/06/1997 5/06/1997 5/06/1997 5/06/1997

Dec. Court 19/03/2003 19/03/2003 9/07/2003 9/07/2003 9/07/2003 9/07/2003

29/04/2004 29/04/2004 29/04/2004 29/04/2004 29/04/2004 29/04/2004

15/06/2000 25/02/2003 25/02/2003 25/02/2003 25/02/2003 25/02/2003 25/02/2003 25/02/2003 25/02/2003 25/02/2003 25/02/2003 26/10/1999 16/02/1999 16/02/1999 16/02/1999 16/06/1999 16/06/1999 16/06/1999 11/12/1997 11/12/1997 11/12/1997 11/12/1997

25/11/1998 25/11/1998 15/01/2001 15/01/2001

14/10/2004 14/10/2004 14/10/2004 14/10/2004

index

Firm

Fine Fine Comm. Court

61 Evergreen Mar. 0.368 60 Hanjin Shipp. 0.62 59 ADM 47.3 58 Ajinomoto 28.3 57 Kyowa Hakko Kogyo 13.2 56 Daesang 8.9 53 ADM 39.69 215 Bayer 14.22 201 Roche 63.5 49 SGL 80.2 48 Showa Denko 17.4 47 Tokai 24.5 46 Nippon Car. 12.2 45 SEC 12.2 44 Carbide Graphite 10.3 43 SAS 39.375 40 BASF 296.16 39 Aventis 5.04 38 Takeda 37.06 37 Merck 9.24 36 Daiichi 23.4 211 Lonza 0 34 Solvay 9.1 33 Eisai 13.23 32 Sumitomo 0 31 Tanabe 0 202 Roche 462 30 Groupe Danone 44.043 29 Commerzbank 28 28 Dresdner Bank 28 27 Bayer. Vereinsbank 28 26 Erste Bank 37.69 25 Aventis SA 24 Degussa 118.125 23 Nippon Soda 9 19 Air Liquide 3.64 20 Air Products 2.73 21 BOC Group 1.17 22 Linde 12.6 17 Christie 0 16 Sotheby 20.4 15 Lafarge 249.6 14 BPB 138.6 13 Aventis 2.85 12 Merck 0 9 Hoechst 99 8 Carbone Lorraine 43.05

0 0 43.875

7.128

69.114 10.44 12.276 6.274 6.138 6.48

375

fine /cap.

0.00371 3E-05 3E-05 3E-05

0.02885 5E-05 0.0002 0.00023 0.00039 0.00777

296.16 5.04 37.06 9.24 23.4 0 9.1 13.23 0 0

0 0 0 0 0

0.01522 0.00254 0.00225

0.01965 0.0001 0.00037 0.00026 0.00275

0.02826

0


376

16/10/07

13:59

Page 376


Table Cont. Decision

3/12/2003 10/12/2003 10/12/2003 10/12/2003 16/12/2003 16/12/2003 3/09/2004 3/09/2004 3/09/2004 29/09/2004 29/09/2004 9/12/2004 9/12/2004 9/12/2004

Surp. insp.

Dec. Court

22/03/2003 22/03/2003 22/03/2001 22/03/2001 22/03/2001 25/01/2000 25/01/2000

index

Firm

Fine Comm.

7 6 5 203 4 207 206 208 209 204 205 212 213 214

SGL Akzo Degussa Atofina Outokumpu KME-group KME Outokumpu Halcor Danone Heineken AKZO BASF UCB

23.64 0 16.73 43.47 18.13 18.99 32.75 36.14 9.16 1.5 1 20.99 34.97 10.38

Fine Court

fine /cap.

0.01243 0.08943 0.07635 0.03692 0.02281 9E-05 6E-05

Panel 1. Event:

Dawn raid

Comm. dec.

Annul

Upheld

43 obs.

100 obs.

27 obs.

28 obs.

time Abn. ret. –5 –4 –3 –2 –1 0 1 2 3 4 5

J

–0.00049 –0.16184 0.00395 1.30569* –0.00372 –1.23128 0.00299 –0.98965 0.00291 0.96477 –0.01133 –3.74904*** –0.00628 –2.07657*** –0.00223 –0.73782 –0.00445 –1.47345* 0.00070 0.23085 0.00007 0.02441

Abn. ret.

J

–0.00362 –0.00174 –0.00110 0.00001 –0.00026 –0.00374 –0.00300 –0.00074 0.00353 –0.00174 –0.00241

–1.49878* –0.71816 –0.45739 0.00559 –0.10638 –1.53905* –1.24364 –0.30518 1.45394* –0.71617 –1.00053

Abn. ret.

J

–0.00361 –0.85540 0.00373 0.87716 0.01203 2.82779*** 0.00222 0.52488 0.00874 2.05229*** 0.00134 0.31766 0.00350 0.82719 0.00028 0.06627 –0.00248 –0.58390 0.00111 0.26374 –0.00293 –0.69397

Abn. ret.

J

–0.00527 0.00100 0.00193 0.00311 –0.00480 –0.00157 0.00012 –0.00192 –0.00251 –0.00670 0.00026

–1.42780* 0.27217 0.52300 0.84121 –1.30148* –0.42480 0.03317 –0.52174 –0.68155 –1.81865** 0.07061

CAR

J

Panel 2.

CAR

J

CAR

J

CAR

cum. –0.02386 –2.27320*** –0.01482 –1.76180** 0.02393

J

1.62190* –0.01634 –1.27150

(T) Ehlermann 07 Session C

12/10/07

13:11

Page 377

SESSION C

TOOLS OF ENFORCEMENT

1 PANEL DISCUSSION

Participants: Rafael Allendesalazar Thomas Barnett Christopher Bellamy Margaret Bloom Jochen Burrichter Stephen Calkins Philip Collins Claus-Dieter Ehlermann Ian Forrester Nicholas Forwood Calvin Goldman Christopher Harding

Joe Harrington Julian Joshua Donald Klawiter Assimakis Komninos William Kovacic Jon Lawrence John Ratliff Patrick Rey Peter Roth James Venit Wouter Wils Donncadh Woods


12/10/07

13:11

Page 378


12/10/07

13:11

Page 379

Panel Discussion

379

Session C: Tools of Enforcement Panel IV: Immunity/Leniency/Financial Incentives/Plea Bargaining 䉴 DONALD KLAWITER—Leniency is indisputably the most effective tool in the history of competition law. This success did not happen over night; it has been an evolution in time that will continue in the future. In my presentation, I will first review how we got there, and second, I will comment on where we are likely to go. In the US we have had leniency since 1978, and we had early successes with a programme that only allowed leniency in cases where the government had not initiated any investigation and had not detected any illegal activity. In 1993 that changed dramatically with a new policy that allowed for leniency after an investigation in very specific circumstances. Looking back at that time, very little happened for the first two or three years. It was very interesting: the programme was there, everybody was very excited about it—and nothing happened. I remember vividly doing a programme at the annual ABA meeting in 1994, where we were to put up a model of leniency under the new programme—five people showed up to listen. Today, when we do that, we have several hundred attend, so there has been a change. Why did it take some time for the programme to get started? First of all, the programme was not enough; there needed to be indications that it would work effectively. Secondly, in the early days there were issues at the staff level at the Antitrust Division. I remember the first time I walked in and asked if leniency was available, and a staff member answered: “I do not believe in leniency.” The third important issue was detection. Because the Antitrust Division had little success with its early international cases, detection was a significant concern. All of this changed dramatically in 1995, when the Lysine cartel investigation was announced.1 There was the immediate impact of large international cases, and as a direct result, flowing directly from large cases, huge fines, of the kind there never had been before, there was the introduction of amnesty plus, which was the incentive for the second case, and there was DoJ “marketing”—a concerted effort the get out there by way of speeches, presentations, etc., to tell people about the leniency programme and also to warn them that there were efforts of detection under way. The tone that the DoJ was using in those days, and still uses largely, was “Cartel profiling”. The first time they announced that everyone thought of TV shows which dealt with 1 For a list of the Lysine cases brought by the Antitrust Division of the US Department of Justice, see http://www.justice.gov/atr/public/4523d.htm. A detailed account of the Lysine investigation is to be found in Hammond S. (2005): “Caught in the Act: Inside an International Cartel”, speech delivered at the OECD Competition Committee on 18 October 2005, available at http://www.usdoj.gov/atr/public/speeches/212266.htm.


380

12/10/07

13:11

Page 380

Session C – Tools of Enforcement

criminal profiling, so everybody was focused on that set of issues, and it was clear that a fair amount of activity was going on in terms of detection. The fourth element was the de-trebling legislation in 2004, which again added a major incentive for people to come in under the leniency programme. Clearly, after the first few years of quiet, the programme has taken off: it has destabilized cartels, and it certainly destabilized the joint defence agreements that occur when an investigation starts. It used to be that you would have a joint defence agreement, everybody would sign up, you would have a written document, you would exchange documents, and everybody would talk to everybody else under the privilege, because you were in this together. Since the 1990s there have been very few signed joint agreements. There were sort of get-together meetings where you talked informally, but the type of agreements we had previously are simply non-existent in criminal cases. It is very interesting, a huge change, and again, destabilization—it really is a means of destabilizing cartels. You are now looking at everyone around that table suspiciously, as you should in this kind of situation. That is where we are. And where are we going in the near future? First of all, the fact that there has been great success in the leniency programme suggests that there will be success in the future, but there is also a lot that we ought to be looking out for. Amnesty plus, a great programme, has worked very effectively, but it is also an opportunity for a company to come in after an internal investigation and throw everything on the table, saying, “Well, we found this cartel which had very little substance to it and we are giving it to you.” The DoJ has been very good about holding off many of these cases, not putting them up for leniency, but in a few instances there have been investigations that have flowed from Amnesty plus and turned out to be complete wash-outs. That is going to happen, but the more frequently it happens, the more it will adversely affect the system. I mentioned yesterday at one point that if you come in and say, “leniency is not available, we have a leniency applicant, and there is really no evidence in the case”, this is going to cause people to take it less seriously than they did before—it is not an issue now, but it is something we should keep in mind. Secondly, if people see that getting into the leniency programme is not quite as transparent as they thought it was going to be, will it hold somebody back from applying for leniency? We do not know. Third, the whole area of damage law and cases—that has been a great benefit, but it is also an issue, because the 2004 legislation is written in such a way that it is not entirely clear how the procedure is going to work. We had one case that worked very well, but we also had plaintiffs’ lawyers saying they are not going to accept only the evidence that the DoJ has got, so they want you to waive attorney-client privilege, they want you to waive economic expert evidence that is going to make our case, they want you to do all the work or else we will tell the court you did not comply, and you would not get single damages treatment. Besides, the question of whether you get single-damage


12/10/07

13:11

Page 381

Panel Discussion

381

treatment may not be decided until the end of the case. That is another area of potential danger that we have to watch for carefully. The fourth area is the international context, and I will not comment much about this because it is detailed in my paper and other speakers will touch upon this—just two quick points. One, the paper-less process is essential, and that goes to all the reform undertaken now in the EC and in other places as well. Second, the one-stop shop is essential, and I cannot emphasize this enough. From the perspective of someone trying to get a party to go for leniency at a certain point, the fact that you then tell the party that it has to go not only to the US, but also maybe to Canada, the EU and a number of countries within the EU, and then it has to get separate lawyers in each of those countries, translate documents into all languages, etc. . . . In certain circumstances this has caused clients to prefer to be second in. In a very real sense they have looked at this as something that is a huge burden and cost in terms of manpower, etc. Finally, corporate governance: these are indeed the issues that can be a challenge to the programme in the future. Corporate governance is a great thing, it has moved these programmes ahead, but we are now in an environment where, if a cartel is detected, even if you get leniency, what is the Board of Directors going to do with the offending employee? What will happen to the person who was the cartel actor and who is also the main witness in the amnesty proceeding? Is he out the door? When is he out the door? How is the company going to deal with him? That is an important issue to consider for the future. JOCHEN BURRICHTER—I would like to make a few comments on the implementation of plea bargaining, or direct settlement, in the EU. Last year Commissioner Neelie Kroes announced European success in cracking down on cartels, to the point that there is concern that the Commission is becoming a victim of its own cartel-busting success, through an increasing workload. One of the possible changes to the existing system is the proposal for a simplified case-handling arrangement similar to the US plea bargaining system. The Commissioner announced that the Commission intends to simplify the proceedings in cases where both the undertakings and the Commission concur on the nature and scope of the anticompetitive behaviour. The Commission prefers the term “direct settlement” instead of “plea bargaining”. I would like to report briefly about the German experience with a similar procedure, which we have had in the past few years. The Bundeskartellamt (BKA) for a long time unofficially applied a quasi-plea bargaining proceeding to what we call “consensual fining decisions”. In clear-cut cases, where the BKA can demonstrate the anticompetitive behaviour, discussions with the undertakings concerned and their legal advisors are initiated, and in the course of the discussions the BKA generally reveals the core elements of evidence. The undertakings can also present exonerating evidence, to reduce the 䉴


382

12/10/07

13:11

Page 382


future fine, and after an understanding has been reached including on the amount of the fine, the BKA will normally provide to the undertakings a draft fining decision, that once more they can comment upon, and finally it will render a decision in the anticipation that the addressees will not appeal it. The decision contains only a short summary of the facts of the case and the legal assessment of the anticompetitive behaviour. In the context of these negotiations, part of the fines have sometimes been either reduced or waived. However, a waiver of the right to appeal is void under German law, so the undertakings are not banned from appealing the decision, although they have been negotiating it. Due to the shortened statement of the facts and curtailed legal reasoning, the consensual fining decisions may not withstand judicial review, so in some cases, at the stage of appeal, the BKA has withdrawn the initial abbreviated decision, and thereafter issued a normal decision. This abbreviated procedure is an example that the Commission could also follow. The Commission could probably issue a short-form decision concentrating on the core of the infringement, the allegation that the Commission has against the companies, and in the settlement process, the Commission could confront the undertakings with the alleged anticompetitive behaviour, either as we have done it in Germany, by discussions, or with a short-form Statement of Objections and a short description of the facts. The undertakings could then be invited to admit to these allegations, and it is expected that the hearing officers would also be involved at this stage of the proceedings. What are the interests of the Commission and of the undertakings in such a procedure? For the Commission it is quite clear that the envisaged direct settlement would likely reduce the workload and thus free resources for other cases. The undertakings concerned may profit from such an accelerated proceeding too. When admitting to an alleged infringement of EC competition rules and waiving their rights to a fully reasoned decision, they can expect a quicker and predictable disposition of the case, so that they save on internal and external expenses. Furthermore, the undertakings would be able to limit the negative publicity associated with antitrust proceedings even if the Commission upheld its practice of publishing a press release. Finally, an undertaking expects that its concessions will eventually lead to a reduction in the fines imposed, even if only by concentrating on the core infringement, and leaving aside any additional aspects which the Commission could take into account in the case of a full investigation. Of course, there are some legal aspects that need to be observed in connection with such a procedure. First, the principle of legality, meaning that such a direct decision would have to be based on some rule of procedure providing that the undertaking concerned can be fined for the alleged anticompetitive conduct it admitted to. In order to safeguard this, the Commission should issue a Notice in which it explains the procedure, so that the undertakings can adapt their behaviour and the negotiations have clear rules. Second, legal certainty: any system of direct settlement would have to take into account legal


12/10/07

13:11

Page 383

Panel Discussion

383

certainty, which as a matter of fact proves beneficial for both parties to the settlement. Regarding the undertakings concerned, legal certainty would require some predictability as to the amount of potential fines the undertaking is facing. In order to guarantee the highest possible level of legality and certainty, it seems desirable that the Commission and the undertakings concerned might negotiate the exact amount of the fine on the basis of the hopefully clear new Fining Guidelines.2 But even if the Commission were willing to set a maximum level of the fine, the undertaking would still be able to ponder this. In this respect, the Commission should exploit its fining range, as it seems rather unlikely that at that stage the undertakings will be able to present exonerating evidence, if they admitted to the charges in the first place. Legal certainty is also important to the Commission when settling antitrust cases. In order to achieve its goal of reducing its caseload, the Commission will try to bargain for a factual waiver of the right to file an appeal. I do not want to comment on whether such a waiver would be permissible, because most national laws do not allow that, but I think that as a matter of fact such a procedure would at least have the effect that, if the decision were appealed, the court would carefully evaluate the fact that the parties have admitted to the allegations. I cannot cover now further aspects of due process, I only want to comment on a couple of open questions. First of all, I think what has to be considered is the relationship to leniency programmes. I do not see a contradiction between such a direct settlement and leniency, because it is an additional element, and it should be granted in addition to cooperation under the Leniency Notice.3 Second, it is very important to prevent abuse. The Commission should not force any undertaking to enter into such a settlement, and the court would have the right to check that. On the part of the undertakings, the abuse could also be that, in order to reduce the risk of exposure to a full-scale antitrust investigation, they admit to something in order to avoid investigation into other matters. CHRISTOPHER BELLAMY—Your presentation was most interesting, especially since we know that bargaining usually takes place in the common law system, whereas leniency is an administrative proceeding.

䉴

䉴 JOHN RATLIFF—I have the pleasure to follow Jochen Burrichter on the same subject, and I am happy to say that our written contributions are largely complementary. We agree on many issues. I propose here to highlight just a few of the main ones. 2 New Commission Guidelines on the method of setting fines were published in September 2006—OJ C 210 [2006]. 3 See Commission Notice on immunity from fines and reduction of fines in cartel cases OJ C 298 [2006].


384

12/10/07

13:11

Page 384


First: as Jochen Burrichter mentioned, the Commission is suffering from an excess of success with the leniency programme. The figures speak for themselves: there are some 40 applications outstanding, and the Commission deals with about 8 cases per year. This is a considerable achievement if you consider the previous rate of performance, but the process is still too slow, so something needs to be done to enhance the Commission’s effectiveness. Second: can you imagine the Commission doing direct settlements now? Could they do it under Article 9 of Regulation 1/2003, which deals with current settlement procedures? The answer is, no. Article 9 is not designed for it; there is a Recital saying that this procedure should not be used for cases involving fines. However, it is interesting that, in this procedure, there are a number of elements which we may imagine being grafted into some Notice, or hybrid Regulation, to establish a procedure for direct settlements involving fines. The hallmark of settlement at the moment is that it has no horizontal effect: it would not block the actions of other competition authorities, so it is useless for these purposes. Could you then imagine the Commission using Articles 7 and 23 of Regulation 1/2003? Article 7 confirms the simple ability of the Commission to take a decision imposing a fine, and to impose a remedy if there is an infringement. Article 23 is the right to impose a fine. I think the Commission might be able to go forward under these provisions, but that would be adventurous, One suspects, rather, that the Commission will issue a Notice, or even seek to modify Regulation 1/2003—they can propose a change to the Council that may lead to an amendment. Third: we have to assume that the defendants and their lawyers will be willing to do an awful lot in order to lower the fine, and that they will be willing to be flexible, within limits. That means that they will be looking for a clearly defined discount, a short procedure, as little paperwork as possible, nothing that the plaintiffs will use—we learned from the US experience and want a paper-less process. They will also want fining predictability. How do we negotiate unless we know what the proposed fine would be? It is very interesting that Jochen Burrichter mentioned that in Germany there would be a draft of the fining decision in the procedure. This is something that I thought that the Commission should have been doing for ages already, because if we had a better idea earlier on and could make comments on the fine and the related reasoning, I think a lot of appeals would never be brought in the first place. Fourth key issue: what is the minimum that the EC legal order requires for the Commission to take a decision imposing a fine, considering the existing case law of the European courts? Or does there have to be an amendment of Regulation 1/2003, in which case the Member States would have to be persuaded to put plea bargaining into the Regulation? This raises a number of issues. First, should plea bargaining be admitted for a European administrative enforcement system, when it is not accepted in general for national criminal systems? I noticed with great interest two refer-


12/10/07

13:11

Page 385

Panel Discussion

385

ences in the papers to a French procedure for plaider coupable; I have noticed a settlement in the UK after a Statement of Objections in the public school fees case; and Jochen Burrichter has spoken of a German example. On the other hand, I believe the Italians are against this sort of bargaining. Maybe they would not object in competition law—that is something to be explored. Second, a big issue is whether the ECN should move as one on this, because the whole idea is to have procedural neutrality within the system. If we get a divide where you can do a plea in Brussels whereas you cannot do one in Stockholm, that would not be helpful for the network. Third, even if you have a short Commission decision, how much reasoning should be in it? We still have an obligation under Article 253 EC (the old Article 190 EEC) to put in a minimum of reasoning. Now could we put in what would look like an American agreement, maybe expand it a bit? Fourth, it was interesting that Olivier Guersent spoke about a “mini” Statement of Objections. That would meet many of the objectives here. Do we need a Statement of Objections at all? Apparently we do. There seems to be a consensus about this, although I have wondered about it, because as a defence lawyer, in this scenario it appears to me one could just waive it. Fifth, access to file: do we need it in order to know what the Commission checked? Normally we always have access to file as of the moment of the Statement of Objections. Sixth, what is the position versus other co-defendants and third parties? This is critical, because one party might settle and others might not, and then their procedural rights could be affected—cartels are conspiracies, and the position of one person affects others, unless you say the plea decision does not have any value for the process, it just settles one case. I recommend that my European colleagues go to the DoJ’s website, where you find a manual, called “The Grand Jury Manual”.4 There is a section on plea-bargaining that is very clear, with all sorts of information regarding the system (largely summarized in my paper). It is a very serious and interesting system. There are big differences: it is a criminal law system, where the DoJ is prosecuting and recommending the plea to a court, which may or may not agree. There are negotiations, to some extent. There is no right of access to file in the US. People in the US seem to believe that their fining rules are clearer, particularly because there is a reference to the effect on interstate commerce—although, when you look at the fining guidelines and the way they are applied, you see there is still a lot of room for judgment in the US system. What I took out of the US analysis is that there are differences, but the system appears to work well, and I do not see why we should not be thinking about it seriously now in Europe. I come back to where I started from, that is, to suggest that the Commission has no choice but to look at it quickly. This is something to my surprise—I did not feel that way when I started this 4

Available at http://www.usdoj.gov/atr/public/guidelines/4371.htm.


386

12/10/07

13:11

Page 386


exercise—but I am convinced that there is a need which is of real interest, for both the enforcers and the companies concerned, and I think there will be more and more cases coming into Brussels. Most big cartels are of a European dimension; so unless the Commission starts giving out bigger cases to groups of Member States, they will continue to be swamped. So I think there will be a need for a new EU settlement procedure, and I support efforts to develop it. 䉴 JULIAN JOSHUA—I will concentrate in this presentation on two of the issues raised in my paper: the first is certainty and predictability, and the second is what I call “looking for the holistic solution”—i.e., not looking at leniency in isolation, but as part of a more complex, anti-cartel enforcement system. On certainty and predictability: a cartel leniency programme has two essential characteristics: it has to be generous, and predictable and transparent in outcome. When I say “generous” I do not imply any degree of kindness or weakness, I mean it must not be too difficult to qualify. Scott Hammond has put it very well in his speech, “cornerstones of an effective leniency program” in 2004,5 that wherever possible the policy of the DoJ is to tilt in favour of finding ways to make companies eligible for the programme, rather than finding ways to keep them out. I think those are very wise words that should be taken into account in certain quarters. As regards predictability, when you go into the programme you have to know exactly what to do in order to qualify. The hardest thing that US prosecutors had to swallow when they introduced the second version of the leniency programme in 1993 was giving up their discretion, making immunity pretty well automatic. Let us look now at the Commission’s policy: notwithstanding the generous accolades when it was introduced, the message behind that being that the previous immunity programme was considered to be a chill on US efforts, because if you did not have predictability in Europe you might not want to come in to apply in the US either. At the time there were the “8(a)” and the “8(b)” standards. The ABA warned that the “dawn raid sufficient” test was too high: there could be cases where you could substantially advance the investigation without necessarily meeting that test, and if you had that kind of uncertainty it would discourage rather than encourage parties from cooperating. We know that there is no objective legal requirement of probable cause or some threshold of evidence for a dawn raid. It is basically whatever the Commission wants to do, and dawn raids may be conducted on a wing and a prayer. Usually they work out, sometimes they do not. Even at the time the Commission said it needed to have complete discretion, for two reasons: we do not want to be bounced into doing dawn raids when we do not want to,

5

Available at http://www.usdoj.gov/atr/public/speeches/206611.htm.


12/10/07

13:11

Page 387

Panel Discussion

387

and we do not want people to appeal our decisions to the court. In practice, to begin with, the “dawn raid sufficient” test was applied generously, and if you came in you were pretty certain to get immunity. There has been a sea of change in the last half year or so, the bar is now being placed much higher, and Olivier Guersent actually said the Commission would like to get rid of “8(a)” applications. That is indeed what is happening in practice now. How did this change come about? We all know about the huge caseload, and for an administrative mindset, one way to address caseloads is to find administrative solutions to clear what you have already and heavy obstacles to new cases coming in, so you keep your backlog down. There is a feeling in DG COMP—perhaps I am exaggerating slightly for the purpose of effect—that in a number of cases they gave away immunity far too easily in the beginning: what happened was that the company that came in first gets its immunity and banks it. The Commission does a dawn raid, does not find very much, then goes back to the immunity applicant reminding it of the duty of continuous cooperation, but suddenly the applicant has an attack of amnesia and is unable to help. At this stage the Commission is in a dilemma: does it drop the case, or does it do what it has been doing, which is to recycle 8(a) evidence into proof of the cartel? The number two who came in is in a difficult position, because it has to be neutral in these circumstances, but those at the tail end do not have much to lose by contesting. The result is that the Statement of Objections gets taken apart in a hearing. In a number of cases the Commission had to drop all or part of the allegations, or put the case on ice, because there is no procedure for a guilty plea. I know that a stock-taking of existent cases is going on at the moment, to prioritize them. New cases are very difficult to qualify . . . I understand why: there is no way to keep applicants on track once the Commission has given them immunity. Commission officials have published an article identifying four different potential outcomes for an immunity application. This is not what you want; you want certainty. This absence of certainty has huge implications for global enforcement. You would be in an impossible situation if you could qualify in the US in ten days, and then the Commission takes months and there are still no guarantees of outcome. So it is very difficult to handle a multi-jurisdictional leniency strategy: you cannot even get a marker, you risk having your application declared incomplete and someone else can get ahead of you. To conclude with “the holistic solution”: the Commission must establish itself as a number one hunter of cartels in Europe. The modernization maze is just too confusing, there has to be a one-stop shop. You have to have negotiated settlements, the statistics are clear—90 decisions so far, four years to get out a decision, which is even longer than before leniency was introduced. Fining Guidelines: the Commission still subscribes to the idea that you must not have predictability in setting fines because otherwise cartel participants


388

12/10/07

13:11

Page 388


would do a cost-benefit analysis. But if your fines are always going to be high, then potential offenders will calculate that they are never to be going to benefit from it. We should have something more like the “volume of commerce” test in the US, rather than an “X” that the Commission plucks out of the air. That would help with negotiated settlements, because then you have a benchmark against which to work on a principled basis. Qualifications for immunity: we will have to get rid of “8(a)”, but actually I would suggest in its place the kind of standard used by the OFT and the DoJ, which is basically: “You can come in, you have to tell us everything you know, but there is no formal evidential threshold that you have to meet, you have to stay on track, and cooperate.”. MARGARET BLOOM—What I would like to address is actually a follow-on from what Julian Joshua was speaking about, which is how the Commission leniency programme could be made more effective. To have an effective leniency programme, you need to satisfy three conditions. Firstly, you need to have transparency and certainty in the programme. Secondly, you need to have effective sanctions against the cartels. Ideally they would be sanctions for individuals, and I think imprisonment is clearly the most effective sanction, plus director disqualification, and possibly fines if they are not going to be paid by the company. Thirdly, you need to have effective powers of investigation. If you look at the 2002 Leniency Notice, that is clearly more certain and transparent than the 1996 version, but could it be improved further? I have seven suggestions in this sense, which are set out in more detail in my written contribution. I just want to mention here that I endorse Julian Joshua’s statement about generosity—a programme needs to be generous if it is to work effectively. My first suggestion is this question of the paper-less process. We know about the risk of discovery, the Commission has for a number of years been accepting oral applications. Incidentally it is not just the risk of discovery from the US, I think there will also be increased risk of discovery requests from some European jurisdictions. The question is: do the formal proposals that the Commission has put out satisfy all the issues? Yes and no. They address some of them, but there are still some significant concerns—which I comment on in my written contribution. My second suggestion is: what is the standard of evidence? I would like to highlight the point of what is meant by “significant added value”? If you are not the first applicant, you need to satisfy this test, and there seem to be different standards as to what that means. The third suggestion relates to the transparency of fines: there are some perhaps unduly sceptical individuals who say, “Maybe the fines are just increased in order to give scope to make a reduction for leniency.” That may or may not be the position, but if there are enough people who think that is the position, it may have an effect and we need to resolve that. 䉴


12/10/07

13:11

Page 389

Panel Discussion

389

Fourth, a point that has already been raised in relation to all the different programmes within the ECN: highly desirable. Almost everybody has a programme, I think we’re up to 19, plus the Commission programme.6 My paper sets out a good number of differences between these programmes. It is surprising how different they are. If there are differences, particularly conflicting ones, that is a real difficulty in terms of legal certainty. The most obvious conflicting difference is that some programmes (including the EC one) require applicants to stop participating in the cartel at the moment they apply, while others (including the OFT one) have an arrangement whereby the authority may require the company to continue participating. Fifth suggestion, the one-stop shop: highly desirable. I would not want to underestimate the challenge, but I would like some sort of global policy on leniency. I am conscious, with Thomas Barnett sitting here, that is an even bigger one, but I think having a more consistent approach world-wide would aid in terms of certainty. Sixth, in Europe there are a number of different programmes and in some cases they produce conflicts between individuals and the companies, inter alia because of different sanctions—under certain circumstances, that can discourage applications. Thomas Barnett said, in his written contribution, that in the US there is a difference between corporate and individual amnesty, which can create a certain dynamic and a race to the DoJ. If you have a race, that is really effective, but in Europe, if in certain circumstances we discourage applications, that is not good. Seventh, on the paper-less process, is it really necessary to have specific citations from the leniency application in the Statement of Objections, and the same in relation to the decision. Can the leniency application really be a roadmap for the Commission to get evidence, or is it the evidence itself? The problem is, the more it is the evidence itself, the greater the applicant is at disadvantage in relation to other members of the cartel in follow-on private actions. I will conclude with a few brief remarks on sanctions. Yesterday a number of participants argued—rightly, in my view—that criminal sanctions are the most powerful incentives to apply for leniency. There are eight Member States who have powers to impose custodial criminal sanctions. There are a few that regard their fines as criminal. If these are to be effective, first of all, the powers need to be used and there also need to be effective leniency programmes to go alongside the criminal sanctions. Last comment: if there is a significant increase in private actions, as the Commission is hoping, then there 6 As of early 2007, the number of Member States with leniency programmes had increased to twenty-two. Twenty-one of these are listed at: http://ec.europa.eu/comm/competition/ecn/ leniency_programme_nca.pdf. In addition, the Italian Antitrust Authority recently began to operate a leniency programme. See Comunicazione sulla non imposizione e sulla riduzione delle sanzioni ai sensi dell’articolo 15 della legge 10 ottobre 1990, n. 287, available at http://www. agcm.it.


390

12/10/07

13:11

Page 390


needs to be some incentive to maintain applications for leniency. One obvious incentive would be to introduce more criminal powers throughout the Member States, and possibly in due course also for the Commission. 䉴 WILLIAM KOVACIC: I want to discuss possible future steps in the arms race between cartels and the competition agencies, and to do that, I will draw on work that I did jointly with Cecile Auber and Patrick Rey,7 and in doing so I will give you my own thoughts, and not necessarily those of my employer. As we suggested in the discussions of the past day and a half, in deciding whether to comply with the law, firms basically have to ask themselves six questions: what is the legal rule, what evidence will establish guilt, will violations be detected, will they be prosecuted, will adjudication tribunals find guilt, and how will violations be punished. The key feature of this taxonomy is that if you change any single variable you affect the decision to comply and all our acceptable strategies in attempting to induce firms to comply. The implications are that a jurisdiction can think of all these variables, and the key policy issue is, which adjustment, or mix of adjustments, builds the greatest compliance at the lowest social cost. As I was suggesting yesterday, commenting on Stephen Calkins’ contribution to thinking about how antitrust systems operate as systems, any adjustment has to account for connections and interdependence among the variables. One of the lessons of modern experience about the detection and prosecution of cartels is that cartels are ingenious and adaptive. We are so focused on the Stiglerian conditions for effective coordination that we forget what Ronald Coase said about firm coordination.8 It is possible to stipulate that coordination in fact causes problems, but if you give firms large prizes for solving them, they try very hard to do so. And the same imagination that brought them to solve difficult contracting problems in the area of legitimate transactions, one may expect, will be brought to bear in solving problems relating to illegitimate transactions. Over a century of experience with competition policy enforcement has shown the adoption of enforcement steps being followed by countermeasures to block the effectiveness of every single enforcement tool. Nonetheless, despite their imagination, cartels have vulnerabilities, which my proposal seeks to exploit. The coordination solutions that we have seen in international cases like the Vitamins cartels 9 are extraordinarily complex, they require massive information sharing with individual cartel members and across members, and there is an indicator of lasting records—you simply cannot keep it your head, you have to write it down in electronic or written form. 7 Aubert C., Kovacic W. and Rey P. (2005): “The Impact of Leniency and Whistleblowing Programs on Cartels”, 21 International Journal of Industrial Organization 347. 8 Coase R. (1932): “The Nature of the Firm”, in Economica. 9 See e.g. Connor J. (2006): “The Great Global Vitamins Cartels”, text available at http://ssrn.com/abstract=885968.


12/10/07

13:11

Page 391

Panel Discussion

391

Why use bounties to detect this activity and exploit it? First, once you decide to provide leniency, you’ve already crossed a fundamental policy threshold; you’ve decided to pay people something to come forward—this is just a different form of payment. It is the proximity to the information that matters: it is the person in the accounting department who discovers odd transfers that are being used to true up accounts at the end of the cartel’s accounting year; it is the salesperson who detects a strange adjustment in the firm’s behaviour; it is the secretary who happens to note the continuing references to a Mr K’s phone calls, simply referring to the conversation that is going on with the cartel members off the official line of the organization. To sit in a room, as the cartel members did, and to imagine that every person there could be a snitch, is a powerful source of instability. The fact that anyone could defect at any single moment and turn in the cartel might prove to be decisive. The key aim here is not so much to find the persons who will be witnesses in a trial, but to provide the vector of information that will allow further investigation to track down in detail how the firm operated. The possibilities for the revelation increase because you divest management of control of all these intriguing decisions that we talked about this morning—about when to walk in the door. If we take the start date of this case, not the plea period from 1989 to 1998, but the case that started in the mid or early 1990s, and ran for 15 years plus. Imagine if a bounty system were available. It has been tested to some extent: in the US there is experience with the Civil False Claims Act, which addresses fraud against the federal government. The bounties may be up to 30% of the funds recovered. From 1986 to 2000 there were over 3000 cases, and recoveries resulting from these cases were over $ 3.2 billion; whistleblowers recovered $ 500 million. Korea has adopted a limited experiment of this for its competition system. What are the key design elements? If you focus the system on cartel detection—I am not talking here about Article 81 EC or Section 2 infringements, just about hard core cartels—the price for informing has to be large if the informant has a promising career ahead within the organization. Imagine that you offered a 10% bounty on the $ 500 million recovery in the Vitamins cartel cases, that buys a lot of personal satisfaction for liquidating one’s own career. Indeed, as part of this mix you have to provide incentives for the informant, as I detailed in my written contribution. Think about the elaboration of methods over time. . . why not simply pay the money to the person inside? 䉴 CHRISTOPHER BELLAMY—Before we start the discussion I will abuse my privilege as a chairman to tell you about the main case we have had so far in the UK on leniency problems at the level of the courts, and that is the problem of a failed leniency application. We had a cartel case in which there were four appeals, and the fourth appellant said he was only appealing against the penalty. “We only got one point, we made a leniency application, it did contain added value, we did cooperate, and our application was refused. We


392

12/10/07

13:11

Page 392


should have been fined less because we made a leniency application.” Furthermore, he asked for a secret appeal, because nobody knew about the leniency application, and if that became known he would face commercial reprisals in the marketplace. We answered that he could not really have a secret appeal: there is a right to judgment and there is a court proceeding. There was a further problem, in that in the course of the failed leniency application the applicant had put in several leniency statements by employees, which of course were disclosed as annexes to the motion of appeal, and it became perfectly apparent when we started to look at them that the statements in question were hopelessly inconsistent with later statements made by the same employees. That was relevant to the rights of defence of the other three appellants, who were denying there had ever been a cartel in the first place. So, making what was in many ways quite a different call, we decided at the level of the court that the court proceedings had to be transparent, and the only thing to do was to disclose the fact of the failed leniency application and the statements that went with it. The result was that the authors of the statements in the course of the appeal were cross-examined by the cartelists who denied their participation. In the end we believed the latest of the series of statements and convicted everybody, and we dismissed the appeal by the fourth appellant who was seeking the fine discount because it was apparent that they had withheld information at an early stage and they had not fully cooperated. This is just a small illustration of the complexities you can get into in this area when you have to balance the rights of defence of the other parties and the position of the leniency applicant. JAMES VENIT—For anybody practicing in this area it is clear that there is a need for the adoption of similar standards throughout the different jurisdictions, which is indispensable to the survival of leniency programmes. Frankly, it is virtually unimaginable, but the worst situation a company can be in is to be admitted in one jurisdiction and not in another. It is an impossible conflict, that cannot be resolved. Having said this, there has to be recognition of a need for harmonization within the programmes. There are various features of the programmes that differ in significant ways: the EC does not have a marker system; the US requires prompt and effective action upon the discovery of the cartel, whereas in Europe you have to stop the moment you come in for the application. In Europe there is a greater desire not to have the leniency applicant reveal what it has done, including withdraw, until the Commission has performed its dawn raid, so there are a lot of conflicting areas between programmes that make life very difficult. This is a sort of plea: this is one area where all enforcers have a common set of interests and it is very important to get the programmes aligned so that they can function in an effective way. Private parties should not be victimized by those inconsistencies.

䉴


12/10/07

13:11

Page 393

Panel Discussion

393

䉴 JOE HARRINGTON—I would like to address a recommendation to William Kovacic. First of all, I very strongly support his proposal. It is not that difficult to imagine that the involved companies might have employees who have that sort of information, and in fact if you read the Commission decision for Carbonless Paper 10 and Christie’s and Sotheby’s,11 they both make reference to employees who were not involved with the cartel but were nevertheless suspicious. Furthermore, in the Lysine case12 they did not want to involve the sales representatives—ADM was explicit about not trusting them—so I think there are lots of grounds to think that, given the proper incentives, some company employees would come forward. A bit of a challenge arose at a symposium in which I participated in January in Tokyo. Japan recently introduced a leniency programme and there was a discussion there too about bounties for employees. And particularly in Japan, there is this issue of disloyalty to the company. I can imagine this becoming even more of an issue if there is criminalization. Do you really want to see your boss go to prison? One thought I had in this connection would be to offer the bounty but the authorities give the company the opportunity to opt for leniency. In doing so, they put in the employees’ mind that they may not just be acting in self-interest, but actually in the best interest of the company. I could imagine in certain circumstances that some other facts might discourage employees from coming forward.

IAN FORRESTER—I read with very great interest the tactical analysis offered by various people in counselling clients, particularly in the US. I would like to make a comment from a Brussels practitioner’s point of view about the problems we are discussing. I have found it tiresome, in cases in which I have been involved, that Commission decisions condemning and fining undertakings describe facts as if those facts have been discovered by heroic Commission investigators using their skills, when in truth those facts have been delivered on a silver plate by companies and their counsel. Why is that question of aesthetics relevant? Because when one reads the final decision condemning one’s client and imposing a fine, the decision presents everything in a bad and uncooperative light as justification for a higher fine. That prompts me to make two remarks. One is that settlements in which there are short versions of the facts would be highly desirable, because this speeds things up. Second, and this is a real problem, the fining process, at least

䉴

10 See Commission fines ten companies for carbonless paper cartel, IP/01/1892, text available at http://europa.eu/rapid/pressReleasesAction.do?reference=IP/01/1892&format=HTML&aged= 0&language=EN&guiLanguage=en. 11 See Commission rules against collusive behaviour of Christie’s and Sotheby’s, IP/02/1585, text available at http://europa.eu/rapid/pressReleasesAction.do?reference=IP/02/1585&format= HTML&aged=1&language=EN&guiLanguage=en. 12 See supra note no. 1.


394

12/10/07

13:11

Page 394


in the majority of cases in my experience, or the process by which the Commission decides how much of a fine to impose, is too non-transparent, last minute, and personalized. Therefore, there is a reticence on the part of clients as to the trustworthiness of an offer of a discount, because the mechanisms on the basis of which the 100% are calculated are non-transparent, and one does not know whether 50% or 70% of an unknown figure is a good or bad thing. As a result of this, I have been approached on a number of occasions by clients with an embarrassing problem: what should they do about it? I have more frequently counselled not approaching the Commission than doing it. I am confident that my colleagues in the US take a different view. Last, the threat of a fine in Europe, particularly for a company that is not an industry leader, is much less troubling—in my experience—than the threat of treble damage entanglement in the US. Therefore, the international consequences of an initiative in Brussels, for example, are more troublesome than the Brussels consequences are. THOMAS BARNETT—I certainly agree that consistency across jurisdictions is a desirable thing, and a goal that we all work toward. I am not sure I entirely agree, though, that the worst possible set of affairs is to be accepted in one jurisdiction and not in another. That presumes that you might not have been in any jurisdiction if you had not gone in. If you accept the possibility or even the probability that someone in the cartel is going to report it, you are going to end up subject to the penalties available in each jurisdiction, and you are better off if you have leniency in at least one of them. An observation: to the extent that the US programme has been successful, it is built on a remarkable degree of trust. Virtually nothing in the system is written in a statute. The definition of what is a criminal act, the entire act is criminal, so it is only through the exercise of prosecutorial discretion that we have limited that. Who is going to prosecute you is a matter of trust. We have individual states who have the right to initiate criminal antitrust actions, and who typically do not do that, but who are not covered by our grant of leniency or plea bargains. It is a matter of trust that they will not act. We were talking here about the calculation of fines and discounts: we have a set of sentencing guidelines that help provide some guidance, and as an initial matter there is discretion as to how many counts the company should be charged with. The DoJ has a lot of ability to affect which range of the guideline the company falls into. Once you have accepted a guideline range, that is only advisory; the court is not bound by that advice, and under our typical plea agreements it is made quite clear and express that the court can do anything it wants, and you are not allowed to withdraw your plea. But through a history of consistent practices, people come to accept that the courts are likely to do something within the guidelines range and recommendation. To some extent, we trust the parties when they come in to obtain a marker. We give

䉴


12/10/07

13:11

Page 395

Panel Discussion

395

them a marker based on their representation that they are going to follow through and cooperate. At times, we enter into plea agreements and the parties may even be sentenced before all that cooperation is complete. So, through a consistent set of practices, speeches and transparency, you can accomplish a lot even without having this legislated, and this is something to consider. WOUTER WILS—A first point on direct settlement: to some extent there was a partial precedent to that under the old leniency programme: whereas the 2002 Notice only gives a reduction of fines for the provision of evidence, in the 1996 Notice there was also a reduction of fines for companies who, after receipt of the Statement of Objections, did not substantially contest the facts. At that time the habit was to give a 10% reduction for that, and it was confirmed not to infringe any fundamental rights. So the idea is just to go further on that, not substantially contesting the facts but confirming them and giving a commitment to pay the fine. If you look at it this way it seems less revolutionary than it would otherwise seem. Second point, on the debate on whether or not the bar/threshold should be raised for the evidence to be provided by an incoming immunity applicant: it is important to keep in mind what the objective is. The number of applications is not the objective, it is of course deterrence. It was said yesterday that we cannot really measure deterrence, but at least we can do better than just count the number of incoming applications and look at what we can do with them, and how many fines can be imposed on other companies thanks to the immunity applications. I think that there is a perception that there is a problem there: many applications come in but some bring so little evidence that it is difficult afterwards to construct a case. I think it is reasonable to expect a minimum of evidence in order to build a case on it. To the extent that this would constitute a problem in international cartels, maybe one should think about the point I made yesterday, that in the beginning the Commission was usually following on cases brought in the US, and this is different from European cartels, where the Commission will have to establish the facts and evidence on its own. Third point: Margaret Bloom commented about first raising the level of the fines and then reducing them for leniency. I think these are two different things to be distinguished. It is a good thing, and it has always happened historically, that around the time that you introduce a leniency system you also raise the general level of the fines, so as to ensure that leniency does not have an overall effect of lowering the level of fines, and the higher fines are probably also justified because leniency has an option of escape in it. What would not be alright is a system where one would have a fine irrespective of the result, and if there is a reduction for leniency the same result would be reached as when doubling and reducing. Of course, the possibility for that depends again on the new Guidelines. I seem to gather from your comments that there

䉴


396

12/10/07

13:11

Page 396


is a perception of too much uncertainty under the current Guidelines, which are, as you know, being reviewed.13 䉴 PATRICK REY—To follow up on the standard of evidence issue, everybody has emphasized the need for transparency and certainty, which means that you have to make a decision up front, which in my view means that you have to bring in all the evidence up front and establish whether it is sufficient to bring the case. In my view, the need for certainty therefore calls for quite a high standard of evidence. On top of that, I would like to come back to the point that one of the merits of this system is to encourage maintaining course and keeping evidence. If you send the message that you had better come forward with enough evidence, that will also have an impact on this incentive. I emphasized the idea of destabilizing cartels, and incentives to keep evidence is a perfect example. The cartel will obviously be better off if nobody keeps evidence, so that it will never be found, but the leniency system introduces high incentives for maintaining evidence in order to apply for leniency. This is the perfect example of where the leniency programme can exacerbate the tensions and contribute to destabilizing cartels. I take the point that you have to be generous, and I think that bounties for individuals is a way of being generous. I also support Joe Harrington’s suggestion to combine bounties for individuals and corporate leniency. At the same time, we should avoid sending the wrong message: we do not want companies to think that through all these systems the cost of cartels goes down because you can benefit from leniency and reductions of fines. If you want to be generous, you should probably focus on the first applicant with enough evidence, rather than being generous with weak standards of evidence and giving leniency to several cartel members.

䉴 ASSIMAKIS KOMNINOS—On Article 9 of Regulation 1/2003, indeed, unfortunately this mechanism cannot be used for commitments in the area of cartels, and this has been criticized by some commentators, for example Prof. Laurence Idot. In the past the Commission apparently has accepted commitments even in cartel cases. In the Woodpulp II case 14 commitments were taken by one of the parties. However, I have attended conferences where Commission officials indicated informally that they would be ready to accept commitments, even in cartel cases. The mechanism of Article 9 would obviously not be used, but this would be done informally. My second point is about plea bargaining. I think that there is a big issue here with regard to judicial review, and the rights of individuals and com-

13

See supra note no. 2. Joined Cases C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85, A. Ahlström Osakeyhtiö and Others v. Commission (Woodpulp II ) [1993] ECR I-1307. 14


12/10/07

13:11

Page 397

Panel Discussion

397

panies to go to court. On the one hand, I see a problem with the effectiveness of the mechanism if companies later go to Luxembourg. On the other hand, it is quite clear from the Treaty itself and Regulation 1/2003 that companies cannot lose their right of judicial review. Having heard what Wouter Wils said, I am not sure that a commitment can be given to the Commission where there will be no judicial review. Perhaps I have misunderstood, but it is quite difficult to have this kind of commitment between the justiciable and the administrative authority. It is not really a private litigation between parties, where you can have civil settlements and perhaps go to arbitration and forego the rights to go to court. This is something to think about, because it has to do with the effectiveness of the proposed mechanism. WOUTER WILS—I think it is not regrettable at all that Article 9 in Regulation 1/2003 cannot be used. It is very good, because the important thing about this Article is that it does not establish an infringement. Do we want a system whereby people pay fines without having established the infringement? Indeed, when fines are involved, it will not be under Article 9, it will be in the line of what exists already in leniency. As I said, it builds upon the old leniency system, so it will be an old Article 293 decision which finds the infringement. On the point of giving away the right to judicial review, I do not think that I have said anything of the kind. The only thing I said was that there would be a recognition of the infringement and a commitment to pay the fine. Of course, when someone has recognized the infringement and made such a commitment, the probability that this person will go to court is reduced, and second, even if he does, I doubt that he would be very convincing or successful. 䉴

䉴 CLAUS-DIETER EHLERMANN: I have two points, one on leniency and the other on plea bargaining. On leniency, there are lots of suggestions as to what DG COMP should do and what the national authorities should do. But leniency is also a matter for management, as written very convincingly by some of the participants. It is probably even more so in Europe because of the cultural context. Individuals coming forward to blame their superiors, either before a competition authority, or even reporting within the company to their management in order to motivate the company to come forward. This, in my view, raises very difficult problems of management. I wonder if there is management literature to address these problems for antitrust, and the issue is not specific to antitrust only—there are similar issues in areas like financial securities and fraud. If one wants to help management, particularly in Europe, to overcome its hesitations and be more willing to help, independently of the question of whether there is a criminal or administrative proceeding, the question is whether there is anything out there in other areas besides antitrust from which we could learn and spread the news. My second point, with respect to plea bargaining, is related to the point raised before about the minimum requirement. I believe that, fundamentally,


398

12/10/07

13:11

Page 398


there is a notion of waiver in plea bargaining, and if you waive something, do there have to be fundamental constitutional requirements? I found it very interesting in Jochen Burrichter’s paper that the Bundeskartellamt practices something of the kind. If, contrary to expectations, people later renounce the waiver, you withdraw the whole thing and substitute it with the mechanism which gives you the fundamental rights protection that the legal system of the country requires—and I say that particularly in view of the fact that this is yet another case where DG COMP might very well hesitate to go forward with a formal amendment of any existing text without having tried it in the field. It seems to me, in view of European experience, that it is to a certain extent much easier to experiment with something on the ground, in practice. Having worked it out if it functions, then you consolidate, and a proposal to the Council has a much higher chance to fly then. But again, my question is: in a system which is based fundamentally on consent between two parties who both see advantages to the deal, why do you need these constitutional requirements, which are there for conflictual situations, in which you do not strike a bargain? PHILIP COLLINS—A few short points. I would avoid the term “plea bargaining” in the European context, I think it points to the wrong kind of consent. Perhaps “settlements” is better, otherwise you get all kinds of associations with criminal proceedings. Secondly, it is absolutely clear to me that we are not talking about settlements in all cases. There are many cartel cases that are very complex, the facts are hugely in dispute, and you have to go through the whole process. But there is a category of cases in any jurisdiction—and I do not know how big or small it is—where there is a desire by the business people to put the thing behind them, to settle and move forward. Those are the cases where people ask why things cannot be solved quicker. We are looking at a group of consensual cases where at least certain or all facts about the scope and duration and identity of the parties is probably not in dispute, but there may be huge scope for dispute about all sorts of other things, and indeed the case is turning into other areas. Thirdly, no doubt there has to be due process, but we are talking about some sort of abbreviated or accelerated process; there are reasons for getting on with the case quicker and there are means for doing it. Fourthly, you need to ensure that in those kinds of cases you are actually engaging with the senior management concerned, because one of the problems is that sometimes lawyers become a barrier between the business and the case officers. One person told me last week in a business meeting, “If only I had gotten in there earlier, I am sure we could have sorted this.” Also, I entirely agree with Claus-Dieter Ehlermann: the idea that we need to sit down and write some magnificent Notice to describe how it is going to work is a fool’s paradise. You do need to test it out in a number of different

䉴


12/10/07

13:11

Page 399

Panel Discussion

399

cases, with different lawyers and business people, and then you write the Notice, you put it out for consultation, and hopefully that becomes an established process. You cannot write it down before this. Last point: I would also like to see built into any settlement mechanism appropriate redress measures for consumers. Yesterday we discussed about mediation, arbitration or some other scheme in which those who suffered loss receive some benefit, and that is reflected in the amount of the fine. STEPHEN CALKINS—A remark addressed to William Kovacic: on the notion of deputizing all secretaries in the world to turn in cartels, I really think it is a terrific idea. Let me suggest three refinements based on a nonscientific sample of one—my own secretary. First, to echo Joe Harrington’s point, she really does believe that all of the faculty for whom she works mean to do the right thing, and she regularly causes us to do things that maybe we did not really want to do. There are lots of people out there who would like to help their bosses do the right thing. Second, she has a very short time horizon: given the choice of an instant tax refund of $ 10 today or $ 100 in a month, she will take the 10 $ every time. So, put a little time perspective into your payout scheme, and you will be able to attract far more secretaries: prompt payment is better than delayed payment. Third, in terms of your payment, she would rather have $ 10,000 than any percentage that you offer, because $ 10,000 is a much bigger prize than she ever imagined winning—and she is also a little math-challenged, so she’d rather take $ 1 million than 9% of 23 million. So, you do not need to give her the exorbitant sums to which you were referring before; what you need to offer is a quick and guaranteed pot and a chance for her to help the world. Then you would have an army of secretaries out there serving justice.

䉴

JOHN RATLIFF—First, on the question of whether a company would appeal after entering into a direct settlement. I do not think that parties are very likely to do that. The whole idea is consensual: they want to get finished with the proceedings, why would they then engage in another 3 years in front of the European courts? So I do not think you would actually need to have a provision excluding the right of appeal (and I question whether legally you could exclude that fundamental right). What is more complex is whether other parties would appeal, arguing unequal treatment. I think the short answer is that all defendant companies must have clear, equal opportunities to enter pleas. Second, I never liked the way in which the Commission gave a 10% reduction for those not contesting the facts. I believe it led companies not to contest the Statement of Objections on various details, because they did not want to put themselves into nervous anxiety as to whether that was going to be treated as a helpful clarification and as positive cooperation with the Commission or contesting the facts leading to a fine reduction. I think that, instead, the 10%

䉴


400

12/10/07

13:11

Page 400


reduction should have been given for further constructive cooperation in clarifying the facts, even at the time of the Statement of Objections (and only withheld if there was a clearly unacceptable change of position by the defence). DONNCADH WOODS—On the Article 9 issue, I wanted to draw your attention to Point 13 in the Preamble to Regulation 1/2003, the reading of which makes clear that you cannot use Article 9 to settle cartel cases.

䉴

䉴 JOCHEN BURRICHTER—Four points: first, why has it happened that decisions of the Bundeskartellamt have been appealed? Just for the reasons that John Ratliff described, namely that the parties had the impression that they had not been treated equally. It is very important that the Commission makes clear how the procedure works. This brings me to the second point: it is necessary for the Commission to publish a Notice making the procedure very clear, although I agree with you that the procedure has to be developed first. But the basic rules on how to proceed, when and how one could contact the Commission, what the Commission can do, etc., need to be spelled out. I think that such a Notice also has another effect, that of an invitation—the companies see that there is the possibility of such a procedure, and this is an incentive. On the minimum requirements for such a decision: even if it is very concentrated, I think it is necessary to precisely describe what a company is admitting to, in order to separate this violation from other possible violations, which finally come up in another procedure. Final remark: I would like to point to the problems and constraints that relate to a procedure such as the one that has been practiced by the Bundeskartellamt, in the context of the Commission. The Bundeskartellamt negotiates with those who decide, whereas DG COMP will negotiate, but it is the collegiate body of the Commission that decides. I think this is a problem that the Commission will have to solve, so that the undertakings can rely upon what has been said.

䉴 WILLIAM KOVACIC—Claus-Dieter Ehlermann raised a corporate governance issue that has been litigated on several occasions under the Civil False Claims Act. The scenario is as follows: I am an employee of a firm, my employment contract obliges me to report instances of apparent wrongdoing to an internal hotline, an ombudsman within my company, but instead of placing this phone call, I go to the DoJ with a complaint about procurement fraud. With the information I provided, the DoJ brings a successful false claims act case, and recovers a large sum of money. I step forward and claim my reward, and then I am sued by my employer for breaching my employment contract, and the relief that is sought is the recovery of the complete bounty. The larger question is whether or not—and I think this would arise in the case of leniency too—I am obliged by my contract to report first to the internal ombudsman before I can do anything else. The courts in the US have


12/10/07

13:11

Page 401

Panel Discussion

401

been troubled by this question, but most of them have concluded that the preference built into the statute—which discourages retaliation and indicates a strong preference for the revelation of the information rather than giving the company a chance to try and fix this on its own—trumps the employment contract obligation. Try to play this in the Lysine case.15 Suppose that one of Mark Whitaker’s colleagues had discovered what was taking place, or that the person who was arranging the meetings in the hotel rooms had known about this. Would that person have the duty to report first to the ADM ombudsman? Or are you relieved of that duty once you learn that the chief executive and founder of the company and the vice-president are in fact backing the whole scheme? Do you have to take a reporting path that seems perfectly futile, and may in fact be simply the mechanism for targeting the space between your shoulder blades for your own dismissal? 䉴 THOMAS BARNETT—Just one last comment, related to what William Kovacic said, I think the trust issue in the leniency context is the most critical element. Traditionally, maybe in the “good old days” until 5 years ago, trust was the client, the lawyer and the government working together. The point from our debate is, the number of people you have to trust and the amount of trust you have to manifest has increased dramatically. Number one, if you are under a marker rather than under a letter agreement, which is more common these days because the process takes longer, for a lot of good and legitimate reasons, you are trusting the government more and more in the US, and that is fine. Secondly, you are trusting other governments, and the issue there is the one-stop shop kind of problem, the paper-less issue. Again, the principle there is that, at the end of the day, leniency applicants should be no worse off than if they had not applied for leniency in the first place, that is the minimum standard. In the US, under the de-trebling provision you need to trust the plaintiff’s lawyers. The issue here is again an important one: what are they going to demand? Are they going to say, “we gave the government all that evidence, but we want you to give us more and more”, or we are going to say this case is not fit for single damages, so that is a trust factor. The last one is actually the most interesting: you have to trust the employees in the company, especially in situations where corporate governance comes into play, and you are now saying to an employee, “we thank you very much for being forthcoming with us and ultimately with the government in this setting, but you also violated the law, you are out of here, you have no pension . . .”, but of course you need to continue to cooperate with both the company and the government in this context. That is going to be a hard one, and as the corporate governance rules develop and really set in, I think that is one that we are going to see as a real challenge, but at the end of the day, it is all trust.

15



402

12/10/07

13:11

Page 402


䉴 CHRISTOPHER BELLAMY—I wonder if I can close this session with one final anecdote, which goes to the issue of trust, since it has been mentioned. The question of the Commission as a collegiate body deciding on the fine, the question of the discount, and the whole question of legal certainty, is what happened in one of the steel cartel cases in which I was the juge rapporteur. The Commission had decided to meet on Friday morning to fix the fine, and the Commissioner had arranged to give a press conference at 12:00 noon on that day. By 11:55 the Commission had not yet formally adopted the fines, but the Commissioner decided to give the press conference anyway, so he went downstairs with the staff for the press conference, and unfortunately he left behind his reading glasses. He had various pieces of paper in front of him, and he picked up one and read out the list of the fines. Unfortunately, when the decision came out, the fines were different. Two companies’ fines had gone up, and another’s had gone up by a very substantial proportion. That raised some fairly obvious questions as to who had decided the fines, etc. What in fact happened is that the Commission carried on with their discussion in the afternoon, and the fines in their final version were adopted in the afternoon, but being Friday afternoon in Brussels, by the time the fines were adopted, most of the Commissioners had actually gone away for the weekend, and the question arose as to whether there had actually been a quorum to adopt the final fines. It was all a mistake, but it does raise the question of how it all works when there is a collegiate body deciding the fine, and this is to some extent a political decision.


12/10/07

13:11

Page 403

Panel Discussion

403

Panel V: Ensuring that effectiveness of enforcement does not prejudice legal protection: rights of defence PETER ROTH—At this stage of our workshop, after one and a half days of stimulating discussions, it is tempting to see fundamental rights as the legal equivalent of Coco Chanel’s little black dress, which, as you may recall, she described as perfect for every occasion. At least in Europe, the same long worn principles—Article 6 (right to a fair trial), Article 7 (no retroactive offences), Article 8 (right to privacy)—are relevant to almost every area of the law, and competition law is naturally no exception. Now the costume is being updated in the new fashion of the Charter of Fundamental Rights. The legal status of the Charter is unclear, at least since the EU Constitution, to which it was appended, is in hibernation, if not permanently buried. But it is seen as a reflection or reaffirmation of fundamental rights already established in EC law, and Regulation 1/2003 says in Recital 37 that it is to be interpreted and applied with respect to the rights set out in the Charter. So clearly it does have a potential role in cartel investigation and enforcement. The Charter incorporates many of the rights in the Convention, but it also sets out additional rights, and the two that are perhaps worth noting as relevant are Article 41 (the right to good administration)—which I think could be relevant for complainants in cartel matters, and Article 50 (the ne bis in idem provision)— which is framed, as Wouter Wils has pointed out in one of his writings, in different terms from the Convention. While the Convention speaks in Protocol 7 about prohibition against second proceedings within the same state, in the Charter that is expressed as second proceedings within the European Union, and therefore it has potential implications for parallel proceedings or subsequent proceedings in different states regarding the same cartel. In the quest for detection and sanction of cartels, it is tempting to downplay the relevance of fundamental rights, as they have not been developed for this particular area of the law and can frustrate effective detection and prosecution in the public good. Tempting, but dangerous: there is inevitably that tension between fundamental rights protection and the prevention of a threat to society. This has become prominent recently with the struggle against terrorism, and there can be few greater threats to society than those posed by terrorism, and that is making clear how important respect for fundamental rights is in the way that they adapt to counter-terrorism—an area in which the record of my country, the UK, has not been particularly good, and that of the US is not very impressive either . . . Even when one confines one’s attention to cartels, if you consider the kind of cartels that they talk about in Columbia—very hard to detect, very difficult to gather evidence for legal proceedings, with a need to offer immunities, a potential for bounties, increasing criminal sanctions—protection of the fundamental rights of the accused is clearly very important. 䉴


404

12/10/07

13:11

Page 404


In saying that, I am not seeking to be mischievous, but the fact that the more penetrating and intrusive the methods of investigation become, and the heavier the sanctions applied—and one of the conclusions that came out most strongly yesterday is that we do need harsher sanctions—the greater the engagement of fundamental rights will inevitably be. None of that means that we should disregard the particular features of competition law, and specifically cartel enforcement, when one considers the application of fundamental rights in this area. I will mention a few that come to mind. First, undertakings are usually legal, and not natural persons. I think that is relevant both to the scope of the right and to questions of justification of the limitation of qualified rights, and most of the rights concerned are qualified rights. This applies when one talks about the privilege against selfincrimination, and I think in the US the privilege in fact does not apply to corporate bodies, but it may apply when one thinks about procedural safeguards for dawn raids, etc. They will not always be natural persons, but there is an increasing focus on application of the competition rules in the field of the professions. Professionals will very often be natural persons, and there will be businesses that are not legal entities, such as in English law a partnership. . . . There is no cartel decision yet at Community level concerning natural persons, but it may come, and certainly in national law, where one may be dealing with quite narrow local markets but a significant adverse effect on consumers. So one cannot assume that natural persons will not be the subject of decisions. Second, to note what was described as the trend of criminalization, both in terms of criminal sanctions generally—I think that Thomas Barnett in his slides had 16 countries that have full criminal sanctions for cartels—and fines in Community law are already classified as criminal for the purpose of the European Convention. If one moves to penalties on individuals, that hugely complicates the picture. If one uses the notion of interdependent variables that Stephen Calkins and William Kovacic talked about, one of those is going to be the effect on fundamental rights protection. Eventually these issues are going to come before the courts—not only specialist competition courts, because they are going to work their way up the appellate process and they will come before supreme courts, which will see cartels in the broader perspective of the full range of the law. Third, the development of increasing cooperation between competition authorities—the Commission and national authorities, the various national authorities of the Member States within the ECN, the Commission and nonEuropean authorities under the cooperation agreements and memoranda of understanding—lead to an encouraging sharing of information, but that raises particular issues when there are very different rules of process in the different jurisdictions. Fourth, the situation whereby the competition authority is investigator, prosecutor, and effectively judge of first instance determining the infringement and the level of penalty. The EC model that has been adopted in many


12/10/07

13:11

Page 405

Panel Discussion

405

of the Member States has significant implications both for protecting rights in that process and for the appellate process that follows. Fifth, the globalization of markets and the international nature of many cartels means that businesses in one state will need and receive advice from lawyers in many different jurisdictions in and outside the EU, and so questions of protection of confidentiality become ever more pressing. And the fact, therefore, that proceedings may be brought against those suspected of being participants in the cartel in a series of states, and not just by the Commission and the US authorities, but also by different NCAs now required to apply EC law. . . . The case allocation system within the ECN is still under-developed, there will be national competition laws that may also apply, and so rules about double jeopardy become important. The application of fundamental rights in the context of cartel enforcement needs to take account of these various particular factors. I have listed in the written paper some areas in the investigation and enforcement process that I think directly raise fundamental rights issues. Anyone could add to the list, coming out of this morning’s discussion the issue of leniency and fines raises issues of equal treatment that were mentioned, settlement agreements raise questions of rights of disclosure to other parties, fining guidelines raise questions of potentially retroactive application. I will mention three areas in more detail. The first is evidence gathering. Dawn raids are clearly an important weapon in the armoury of the competition authorities. The Roquette Frères judgment of the ECJ16 was, I think, hugely important in establishing safeguards, and it is interesting how the ECJ waited for the judgment of the Strasbourg court in the Colas est case17 in deciding how to establish its principles and bring them into line, and it represents a sort of compromise between ex ante control by the national court and ex post control by the EC Court. Dawn raids is an area where the distinction between natural persons and legal persons is very significant, not in the context of who is an undertaking, but in the direct practical question of where the dawn raid is carried out, as there is a significant difference if it is a private home or business premises. Proportionality seems to be the key message that comes through. But the fact that the national judge has no right to see any of the evidence, under the Roquette Frères compromise (which is picked up and enacted in the modernization regulation),in exercising control over the issuing of a warrant, is an interesting limitation, I wonder how that compares with the situation in Canada or the US in getting judicial authority for search and seizure. The second area is attorney-client privilege, which has not surfaced in our discussion so far, except when mentioned by James Venit. “Privilege” is an 16

Case C-94/00 Roquette Frères [2002] ECR I-9039. Colas est and Others v. France, ECHR judgment of 12 April 2002, text available at: www.echr.coe.int/echr (HUDOC database). 17


406

12/10/07

13:11

Page 406


old term for this concept, it is not really the privilege of the lawyer at all, it is a principle that rests on a rule of law rationale, the importance of being able to get advice on your rights and to do so with complete candour, and the importance of that in maintaining a law-abiding society. It is clearly a right that has been recognized to be within the scope of Article 8 of the Convention. We have two significant limitations under the AM&S rule.18 First, the exclusion of protection of advice from in-house counsel, which is now pending in a case before the CFI.19 One can see the problem of situations where the lawyer is not subject to rules of conduct, ethical rules of a bar association or legal professional body, but some employed lawyers are, and there is a great variation from jurisdiction to jurisdiction—they are subject to professional rules, I think, in Belgium, Ireland and other countries within the EU and outside. Maybe that points to a way of solving that problem, and having not one clear rule, but a rule that takes account of the ethical obligations that employed lawyers have or do not have. But the second limitation, not raised in the AKZO Nobel case,20 is the lack of protection of advice from non-EU lawyers. That seems to me still more problematic: a company wants to exchange information with its competitors about shortages of raw materials, so they get advice from their US lawyer, or their lawyer in the UK, Canada, or Germany, and only the advice from the first two lawyers is protected from seizure by the Commission. Or they are directly under investigation for alleged cartel violation in the US and the EU, and the advice from the UK lawyer is subject to seizure. James Venit mentioned yesterday the issue of what happens when such advice is seized by the Commission and then becomes disclosable in US proceedings. Even before that qualification problem, it is just a basic issue of the right to confidentiality in advice, and why it should be limited only to lawyers admitted to practice in the EU. I think it is very hard to see the European Court of Human Rights (ECHR), which is not limited in its composition to judges from the European Union, accepting the distinction that is being made in the current state of EC law, and if the instance arose—as I think eventually it will, in one Member State or another—it puts the national court in the difficult position of dealing with the situation, with conflicting superior norms. The third of the three areas is self-incrimination, where we still have in the EU the Orkem rule,21 fashioned before a number of cases were decided by the ECHR. This rule works fairly well when it comes to documents, separate existing information and evidence, and it may work well when you come to questions that clarify what is in the document—what does an abbreviation mean, whose initials are there etc.—but when you go beyond that and start asking questions of a more general nature, I think it gets rather messy, even 18 19 20 21

Case155/79 AM&S v. Commission [1982] ECR 1575. Cases T-125/03 and T-253/03 Akzo Nobel v. Commission (not yet decided). See supra note no. 20. Case 374/87 Orkem v. Commission [1989] ECR 3283.


12/10/07

13:11

Page 407

Panel Discussion

407

when questioning undertakings, but when one goes further, to coercive questioning of individuals, which is something that the Commission cannot do but many national authorities can, the self-incrimination problem becomes still more complicated, and once you have personal sanctions, whether criminal or director disqualification etc., I think there is a very serious issue of selfincrimination. The Orkem rule, which is not always very practical anyway, as it is not very clear, needs reconsideration. The CFI effectively ducked the issue in the Mannesmannröhen case,22 and it is pending before the ECJ in one of the Graphite Electrode appeals.23 There are many other questions that can be raised, and perhaps will be raised in our discussion. I will conclude by saying that much in our discussion in the previous sessions has focused on the creation and improvement of effective means of anti-cartel enforcement, but when one designs these new means, and that is an endeavour that will clearly continue because the methods and techniques can always be improved and taken further, let us remember that whatever we fashion has got to match the principles enshrined in that “little black dress”. CHRISTOPHER HARDING: I am very conscious that, being an ultimate presenter here, I am at a stage where what I have to say might have been said before, and indeed, there have obviously been important themes in our discussion so far which do touch upon some of the points which I have raised in my paper, but I still think there are a couple of new points I can raise. I started reflecting on the subject of my paper by thinking about the modalities of enforcement and the changes that have taken place over the last few years in the way in which enforcement is carried out. We have heard a lot about leniency programmes, and the changes they have brought to the methods of enforcement, and other issues coming out of enforcement. My general point is that changing modalities of enforcement means, very often, new issues of legal protection. We heard numerous anecdotes this morning about issues raised in relation to leniency applications, fairness of proceedings, etc. This very well illustrates that new questions will continue to arise as the methods of enforcement evolve. This is good news for all of us—new issues means work for lawyers, economists and academics. What I wanted to do was pick up on what seemed to me to be some of the ongoing and perhaps even future, prospective issues in relation to legal protection. In the paper I took three areas, or contexts, in which to explore these questions. First of all there is the whole question of fines—calculation of the fines, the way in which they will be imposed—this has clearly become an area of intense legal activity. We heard about the rate of appeals against decisions imposing fines—at the

䉴

22

Case T-112/98 Mannesmannröhen-Werke v. Commission [2001] ECR II-729. Case C-301/04 P Commission v. SGL Carbon, judgment of 29 June 2006 (decided subsequent to this discussion and upholding the Orkem rule). 23


408

12/10/07

13:11

Page 408


moment there is a great deal of legal argument before the EC courts in relation to the way in which fines are calculated and way in which decisions imposing fines are taken. Second, I wanted to say something about what I refer to as “the temporal dimension” of enforcement proceedings, issues coming out of the fact that these processes take a long time, several years, to complete their course. It seems to me that now we see some interesting and sometimes very difficult questions arising out of this long time span from the time the investigation is opened to the decision imposing sanctions and the appellate procedure, remembering that now these matters go to the CFI first and then to the ECJ. The third context is what I call the interface between legal orders, the complexity arising from the more globalized process of legal enforcement. Within Europe we already have the complication of the EC legal system and national systems. If we extend beyond that to the application of rules against cartels which have an international dimension, we bring in other national legal systems elsewhere as well. Taking those three main contexts, I then identified some more specific issues. Starting with the area of fines, what seemed to me to be a very interesting debate going on in a number of cases is the issue of transparency versus discretion. It seems to me that it does raise some interesting questions about what is the purpose of the procedure we are talking about. The claims of transparency and predictability in the calculation of fines seem to me not to be as much driven by the need of the appellants against fines to win their cases for a reduction or annulment, but also by the enforcement needs of the Commission, with its heavy workload, and now we are hearing indeed about the possibility of direct settlement as well—which of course is going to produce greater transparency and certainty in the calculation of any sanctions, if you agree by that kind of track. So there are clearly pressures from both sides to move towards greater transparency and certainty in the calculation of fines. On the other hand, the Commission and the EC Courts still claim that there is a need for discretion, that this is a deterrent, and again, if we see part of the process as being involved with some kind of retributive justice against wrongdoing, then discretion is an important part of that. I think it is something that is not only productive of specific legal questions, but something that should make us reflect on the nature of the process. Greater transparency and predictability of sanctions seem to me to be bringing the process more towards a kind of business deal, whereas the emphasis on discretion is more important if you see the objective of the process as being the long-term eradication of cartels, where issues of deterrence and retributive justice are much more at the core. I think this debate, which is clearly taking place now in legal practice and in litigation, does touch upon what the ultimate objectives of the process are. For instance, from the perspective of the Commission, is the objective to clear up as many cases as possible, or is it, as is claimed in rhetorical statements, the eradication of cartels? Also in relation to fines, I am interested in some of the discussion on the socalled “financial constraint discounts”, where companies are pleading that


12/10/07

13:11

Page 409

Panel Discussion

409

they should have a lower level of fine because of the financial difficulty that might otherwise result from the imposition of a very heavy fine. This is possible under the 1998 Guidelines. There is some interesting literature on the way in which this financial constraint plea is being taken into account. There are two concerns that have been expressed by some writers on this subject. The first is the reality of the financial constraint. By looking at the way in which this argument is being taken into account in different cases, there are some issues of unequal treatment there. Finally, on fines, there is the issue of soft law. What we are talking about in relation to the fining guidelines and leniency programmes is a lot of decisionmaking based on soft law instruments. That raises the question as to what extent those instruments, as a legal basis for decisions, are subject to arguments about legal protection. Is it just a matter of discretion exercised, or are there normatively harder rules which in fact are being applied? This is something that the EC Courts now have to address in some of these appeals, trying to work out whether there are elements, for example in the Fining Guidelines and the Leniency Notice, which have achieved a normative hardening so as to become rules whose application would then clearly be subject to legal protection principles in a way in which the exercise of discretion would not be. Those are just a few points coming out of the whole issue of how fines are being calculated and imposed. Turning to the second context, the temporal dimension, there are two interesting points because of the length of proceedings. . . . First of all, what happens if the law changes between the point when the offending conduct took place and the time when the sanctions are decided and imposed? The principle concerning retroactive application of new rules, enshrined in Article 7 of the ECHR, has been brought forth in some of the legal arguments. It seems clear that the EC Courts will apply this principle by analogy with Article 7. The question is then to what extent we can see a change in the law, particularly in the context of the application of soft law instruments? There we see the EC Courts working out a distinction between, for instance, a new interpretation of the law which would qualify as a “change” and therefore bring into play the principle against retroactivity, and what is described as a gradual “clarification”, which is not such a legal change as to make it possible to invoke that principle. The other situation, which is interesting in relation to the temporal dimension, is what happens when the identity of the persons subject to proceedings changes. This is becoming a fairly frequent problem: after the occurrence of the offending conduct the legal identity of the defendant changes, the company disappears, or merges with another company, or it is taken over or whatever. Here the guiding principle is the so-called principle of “guilt”, which means that it is personal, individual responsibility which should attach to the original actor. So, for instance, if the company is restructured in the sense of becoming the subsidiary of another company but retains its original


410

12/10/07

13:11

Page 410


legal personality, the sanction follows the original legal actor there because the latter is still identifiable. And then there is the case of a merger where the original legal identity is lost, and so the merged entity would then take over the responsibility. In one of the appeals coming out of the Cement Cartel case24 we have the interesting situation where one of the original companies, Portland Cement, changed its identity and became a newly-created subsidiary through the transfer of some of the original assets of the company, and there the issue was: who takes over the liability for the sanctions? Clearly, from the point of view of effective enforcement, there is a need to ensure that corporate restructuring is not used as a way of evading the imposition of the sanction. Finally, on issues arising from the interface of legal orders, just to identify two particular problems: first of all, the double jeopardy issue. We heard already that this is very much on the agenda now as far as EU law is concerned—the traditional application of that principle would be in relation to the same company being dealt with in different legal systems. The EC Courts already visited that problem long time ago, when companies started arguing that they had already been dealt with under US law or under Member State law, and were subsequently punished by the Commission under EC law. To what extent should be the earlier proceeding and sanction be taken into account? That issue has been largely resolved. There is another aspect of double jeopardy which interests me, and this I think is a much less easily resolved issue: a company and the individuals working for it being dealt with under the same legal order or different legal orders. This we can describe as “the alter ego problem”, the fact that you may have sanctions imposed both on the corporate person and the individual when you could argue that, in a sense, the company and the individual are the same actor. The individual works for the company and represents its interests; is there any difference between the two actors? I think this is something which might very well occur now as an argument exploited by the defence, in relation to sanctions being imposed on the company under EC law and criminal sanctions being imposed on the individual under another system. Finally, there is the issue of having consistent and equal legal protection across different legal orders, now that we are talking about the imposition of both sanctions against companies under EC law and criminal sanctions against individuals under national law. There is of course a very strong argument in terms of legal protection, that there should be consistent legal protection, procedural consistency and equality in terms of defence rights in those different systems. But of course we are talking about different legal orders: EC law with its own legal protection principles, and Member State criminal law being subject rather to the kind of legal protection that arises under the ECHR. Of course, the two systems try to ensure consistency 24 Joined Cases C-204/00 P, C-205/00 P, C-211/00 P, C-213/00 P, C-217/00 P and C-219/00 P Aalborg Portland A/S v. Commission [2004] ECR I-6453.


12/10/07

13:11

Page 411

Panel Discussion

411

between the two approaches, but we need to look at what is actually happening in practice, on the ground, in terms of the use of procedure and the detail of legal protection that is available across those different legal orders. 䉴 NICHOLAS FORWOOD: I would like to begin by saying that I have found this debate of particular value from the perspective of a judge at the CFI. One of the reasons why I say that at the beginning is because it is sometimes forgotten how removed we are in Luxembourg—not only physically removed from the hurly-burly of Brussels discussions on Commission proposals, but also temporally removed. I was reflecting that the most recent judgments that we had to deal with involving the leniency programme were still concerned with the original leniency programme. The first cases involving applications of the 2002 Leniency Notice have only now been lodged before the court. To a certain extent, we are living in strange sort of parallel world, that is about 3 to 4 years behind. What follows are really some reflections in the light of what I have heard so far. The first is that, although many of the remarks have been made in the context of how existing cartels are to be dealt with, both from the point of view of the lawyers and the competition authorities, I think it is worth remembering that the primary measure of effectiveness of the laws prohibiting cartels is the extent to which those laws—and the way in which they are applied—combine to deter companies from entering into cartels in the first place. Sometimes, as understandable as the concerns with the problems of how to deal with an existing cartel may be, we need to stand back and remember that the purpose of anti-cartel laws is to prevent their formation. Second, one has to remember that potential cartelists do act with a high degree of rationality in deciding whether or not to enter into a cartel. Donncadh Woods made the point that these are very complex business judgments made by individuals who are used to making similar judgments in other areas and there is no reason to think, therefore, that these signals will not be appropriately read by them in relation to the decision on whether or not to enter into a cartel. Third, in relation to the level of fines, there was a point in Massimo Motta’s presentation in this connection: one should perhaps think more about building into the appreciation of the level of fines something to reflect the likelihood of detection. The common list is, I think, a perfectly justified system in which the fine is a multiple of the perceived benefits of the cartelists, and that multiple is the perceived risk of detection. We all understand the problems that this would involve, but it seems to me that, at least from a policy point of view, that is something that could be brought in even under the framework of the existing Guidelines, that fines may be set at a lower level. Having said that, the other thing that struck me is that, in discussing what changes do need to be made for the future in order to reinforce the effectiveness


412

12/10/07

13:11

Page 412


of what has become by now the primary instrument of cartel detections and enforcement, the leniency programme—in deciding what changes need to be made, it is worth remembering that we are in a period of transition. The transition is seen in Donncadh Woods’ remarks about how, in the early stages of the leniency programme, people need time to get familiar with it, and we then have another transitional effect, which concerns end-of-life cartels. But remember that the ultimate objective is to create the right atmosphere to prevent the formation of new cartels, and thus some of the phenomena being described as end-of-life cartels are perhaps understandable in a process of transition. I do not necessarily mean, however, that this will continue to be the case. Let me now turn to something more open in the framework of this session, the administrative or criminal approach. A few remarks: first, on the other side of the Atlantic, the criminal approach is clearly the focus of activity and the system has developed there in a way that appears to be effective. People understand how it works, it is based on trust, but nonetheless everyone understands broadly how it works. Here, I am personally concerned that we do not have the same clarity of approach. There is, on the one hand, the administrative approach. It is just unthinkable that we are going to be in a situation in which we will be able to envisage criminal sanctions taken within the EC legal order against undertakings. We know the problems that we had with Treaty changes of a less contentious nature, but introducing a criminal enforcement system for cartels is unthinkable for the foreseeable future. That being so, we are stuck with the core administrative enforcement structure. The more I listen to some of the problems that have been identified, the more concerned I am about the conflict in signals and incentives that may be appearing with the setting up of parallel systems, in which some Member States have criminal enforcement against individuals. Several speakers have referred to the fact that, in international cartels, it is the US system that drives the attitude of cartel participants in deciding the nature of their cooperation. Again I see a parallel: the same thing happening here, the risk of the Commission losing control over this aspect. . . . This does require a lot of reflection. Another point that occurred to me is the tension that may exist within a firm in relation to cooperation, between the individuals involved, and their superiors, the highest levels of management. If there is a situation in which there is a single choice to be made as to whether the company goes for cooperation and leniency, that may make cooperation more likely. On the other hand, if individuals at the same time are saying, “What is my future within the firm? Am I going to get a reward if I go to the authorities? Am I going to avoid criminal proceedings?”, I am unclear whether the structure that is being created is conducive to maximizing the likelihood of the cartel being informed on, and thus the perceived risk of a person denouncing the cartel will be reduced. Listening to the debate, I am just concerned that we are getting a situation of a lack of clarity, whereas the one thing that is necessary in order to


12/10/07

13:11

Page 413

Panel Discussion

413

deter cartels is a high degree of clarity and a high perceived risk that any cartel formed will sooner or later be sanctioned. Another thought relates to the question of a 100% reduction for the first to go in. If the purpose is to discourage the formation of cartels, it is worth asking the question whether the 100% should be available to the initial founders of the cartel, as opposed to subsequent members. I would not go as far as Mario Siragusa went, to say that [establishing and then later blowing the whistle on] a cartel is an anticompetitive strategy, but it does seem to me to be a question worth considering. As far as procedural questions are concerned, Peter Roth has rightly touched upon some of the difficulties with self-incrimination. One of the great pities, when Regulation 1/2003 went through, is that the occasion was not then taken to introduce a power to require individuals to provide information about cartels, a personal obligation on the management and so on to respond accurately, and there, of course, no question of self-incrimination would arise, since they would not be exposing themselves. More generally, in relation to the application of fundamental rights principles, one fact which is worth bearing in mind, and which does justify a different approach for cooperation on the part of undertakings as opposed to individuals, is that corporate existence is a privilege granted by the laws of the countries in question. I do not see anything wrong in principle if those privileges were given at a particular price, one of them being an obligation to respond with information as to the legality of their activities. That does not take us all the way home, but it is a factor that does justify a different approach. RAFAEL ALLENDESALAZAR—I wanted to comment on one of the issues that Peter Roth has mentioned, legal privilege. There is a very interesting Spanish case on legal privilege. The background is the Pepsi-Coca Cola fight, where there had already been a complaint in Brussels, and then the case came up in Spain. The Spanish competition authority obtained, in a dawn raid, a questionnaire sent by Coca Cola Europe to its bottlers in Spain, which are independent companies, on the kind of practices they were applying and their importance for the business, trying to evaluate, in order to negotiate with the Commission. This questionnaire was seized by the Servicio (the Spanish competition authority is divided between a Servicio, which carries out the investigation, and the Tribunal, which decides). The Servicio got the questionnaire, not the answers, and of course it also requested those. We denied this request, arguing this was legal privilege, because Coca Cola Europe had asked for them, it is European lawyers who are requesting this information in order to negotiate with the Commission. At the beginning, the Servicio did not really believe that there were any lawyers behind that, but then we proved it, and then the Servicio argued that it was not privileged information, because it was a document provided by Coca Cola Europe to its independent Spanish bottlers. Interestingly enough, the Servicio initially had also requested those

䉴


414

12/10/07

13:11

Page 414


responses from Coca Cola Spain, but in the end it only requested them from the bottlers. The lawyers of Coca Cola Europe were not their lawyers. We made an appeal to the competition authority, and surprisingly, considering that it is an administrative authority, the Tribunal accepted the appeal, recognizing legal privilege. It said that the foundation of legal privilege is not only to protect persons who need legal counsel, to defend their rights and liberties, it is also to guarantee the proper administration of justice. Consequently, the principle does not only serve private interests, it is also demanded by the general public interest. Then it went on to see whether the fact that there was not clearly a lawyer around the place mattered or not. It established a material criterion which is essential, it said it is not the fact that the lawyer is there, but that the document has been produced within the framework of the defence of the company, the information was exclusively intended to support, in the most appropriate way, the legal defence of the company’s interests. I think this decision is extremely important for several reasons. First of all, it demonstrates the advantages of having two competition bodies, one conducting the investigations and the other taking the decisions, because the latter may have a fresh way of looking at things. This decision was taken prior to the AKZO Nobel ruling,25 which is also important—the Spanish decision was adopted in early 2002. This decision rejects the idea of having a formal lawyer there as the essence of legal privilege. Would including a lawyer copied in an e-mail suffice to obtain legal privilege? Of course, this material criterion is not always easy to apply, but it is much more intelligent than just saying that, if an EU lawyer is the addressee of a document, that is legal privilege, and otherwise it is not. 䉴 MARGARET BLOOM—One of the issues that Christopher Harding mentioned was the discretion of enforcement authorities versus transparency or predictability. He was talking about this in relation to fines, but it is also an issue in relation to how you devise leniency programmes. Back in 1999, in the days when there were few leniency programmes in Europe besides the Commission program, in the UK we were devising a leniency programme for the first time. We were much impressed by the success of the DoJ, and they helped us considerably, making a very strong point that you need to make sure you have as little discretion and as much certainty as possible for the applicant, because otherwise you discourage applications. To the OFT that seemed quite a straightforward proposition, but interestingly, when we were discussing this with other colleagues in Europe, they said we cannot give up discretion. At the time, there was some significant discretion in the Commission’s program . . . the Germans decided to follow the Commission, 25 Cases T-125/03 R and T-253/03 R, Azko Nobel v. Commission [2003] ECR II-4771; on appeal: Case C-7/04 P(R), [2004] ECR I-8739.


12/10/07

13:11

Page 415

Panel Discussion

415

and so did the Dutch, and there was this interesting tension within Europe, whether to go US-style or keep it within the control of the enforcers. In the UK, we decided to go for as little discretion as seemed sensible for the OFT, to encourage applications. I think this reflects to a certain extent a cultural difference between the two sides of the Atlantic, but also, I suspect, some evolution, because since then the Commission programme has been amended, and is now much less discretionary. The Bundeskartellamt also just changed its leniency programme, and there is less discretion in it too. My second quick comment: Nicholas Forwood questioned whether we should have this 100% amnesty for applicant number one. I would very much advocate such an approach, because what you want to do is have something that is so attractive for number one, ideally number one would not be prosecuted at all, in order to have this race to get in first. Number two can also get something pretty good. 䉴 NICHOLAS FORWOOD—In reality, I was expressing doubt as to whether the founding members of the cartel should receive the same favourable treatment as other number one applicants.

MARGARET BLOOM—I must have misunderstood you. In that case, I would still say, be very generous with number one and be good to number two.

䉴

CALVIN GOLDMAN—I certainly agree with Nicholas Forwood’s comment in that there is a large continuing hiatus between the EU and the US/Canadian approach, that criminalization or its absence is a major distinction which cannot be overlooked. Australia is going through that painful exercise, crossing that bridge right now. The OECD spent a lot of time trying to come up with a set of recommendations on information exchange, and I want to comment on that for the purpose of responding to the very important points raised by Peter Roth about legal privilege. One of the focal points of this exercise—which took 4 years at the OECD, to work through the principles of information exchange between the authorities performing cartel investigations—was the protection of privilege. In so doing, the recommendations made it very clear that both the requested and the requesting jurisdictions must take all steps to protect, in accordance with the relevant rules, each legal privilege. There was a bit of a breakthrough in ensuring that, when a request is made from a common law jurisdiction to one in the EU, the higher standard will prevail in the course of information exchange among the authorities involved in the pursuit of an international cartel. That is a significant acknowledgement, at least by the OECD—all this is a recommendation, but it is a step forward in trying to achieve the higher standard in these difficult times. It is also very important because, in this era

䉴


416

12/10/07

13:11

Page 416


of dawn raids, when computer data are downloaded and all kinds of information comes out of the hard-drive which may contain privileged e-mails, the US cannot take this information and give it to Brussels or vice versa because there is no mutual assistance treaty to allow that kind of information exchange, and it can only occur where an amnesty applicant gives a waiver. But it can flow now, and this is happening, from Washington or Ottawa to London, because the UK has criminalized its system. So these are real world issues that are now evolving, and the principle of the OECD is one that I hope will see greater application in the future. 䉴 JULIAN JOSHUA—First point: the present system of administrative enforcement is I think, in the end, the way it is done. The problem of course comes with modernization and criminalization in some Member States. Christopher Harding and I wrote an article in the English Criminal Law Review—the first antitrust article ever published in that conservative English journal—and I think we had a picture of this official of the OFT accompanying Commission officials in a dawn raid. And the OFT had several hands and feet in this system of administrative and criminal law enforcement system, as a prosecution authority assisting the Commission . . . and while the Commission officials were hassling the employees and executives of a company in a corner and attempting to extract all explanations from them in a way that we would not even have thought possible in my days at the Commission, the OFT officials would be busy having to “Mirandize” them, to use a US term. So there are lots of complications; I think we have to try to simplify matters rather than making them more complicated. I think that, with all due respect, my colleague Peter Roth was a bit unfair to the judiciary in relation to privilege and self-incrimination, because I think there is a tightrope to be walked between the jurisprudence of the ECHR and its application in practice, especially in administrative procedures involving incorporations. As Nicholas Forwood said, it is a privilege to be incorporated. It seems rather strange to make human rights arguments for companies. I think the system is quite balanced—there are no areas off limits, you just have to ask the question correctly. With regard to attorney-client privilege, I really do think again we were walking down the wrong road in Europe in sticking to this distinction of whether you are a member of the bar or not and so on. You can have a picture again of a French company employing a UK solicitor, and the French juriste d’entreprise sitting in the same office, and the advice of the UK solicitor is privileged whereas that of the French juriste is not. I think we should have a test similar to the one in the US, which is a test of functionality, in the sense that it does not matter where the advice is being given or to what bar the lawyer belongs or not—if he is giving confidential legal advice to the company, that ought to be covered by privilege. As far as I recall, even in English law the idea of whether one legal adviser belongs to one discipline or another


12/10/07

13:11

Page 417

Panel Discussion

417

is something of a red herring, the main point is whether one is giving confidential legal advice. 䉴 PATRICK REY—I was a bit struck when Christopher Harding advised discretion versus transparency, associating discretion with the policy objective of deterrence and transparency with the objective of clearing up cases upfront. I think that one of the benefits of transparency is indeed to encourage applicants to come forward, and this is the key element for deterrence purposes. To the extent that we agree that deterrence is the purpose, Massimo Motta said something about the current wave of dying cartels. I think it is also good that dying cartels see detection as being much more costly than otherwise anticipated. This also contributes to the deterrence purpose, and helps to provide an environment where cartel members and prospective cartelists become aware that they cannot just walk away from the law. Last point, again related to deterrence: I would like to refer to the possible exception for the founder of the cartel. Given the point made here, that you want to create distress and you want to make the ones approached by prospective cartelists suspicious, I would support Margaret Bloom’s point about the race. Removing one or a couple of the founders from this race may lessen the incentive to come in. It is a good idea to have as many potential runners in the race as possible, to make sure someone will decide to come forward.

䉴 CLAUS-DIETER EHLERMANN—I wanted to ask a question. We all seem to agree on the necessity to create incentives, to avoid the creation of cartels, and to make them come forward. The old ECJ case law on privilege, which limits the privilege to EC-admitted practitioners and excludes in-house counsel and non-EU practitioners—does that fit into this landscape of creating incentives for companies to avoid cartels and to come forward? My feeling is that it does not fit anymore, and there are good reasons for those who created that obstacle—the judges—to reconsider.

䉴 NICHOLAS FORWOOD—The distinction was created by the judges at the “instigation” of the Commission, which benefited from it at the time, and it was created in entirely different circumstances. Two observations on privilege. First, the principle mentioned by Rafael Allendesalazar sounds to me more like what we call “litigation privilege” in the UK context rather than legal professional privilege or functionality, because you need to have some kind of constraints and obligations in the system, to make sure that those giving legal advice are genuine lawyers and exercising their function according to certain professional ethics rules and professional standards. In the UK, both the Bar and the Legal Society have created special rules for employed lawyers. In my experience, the difficulty of employed lawyers is often that they wear


418

12/10/07

13:11

Page 418


several different hats—they give legal advice about 30% of their time and they do essentially executive jobs of all kinds in the rest of the time. On this alleged tension between cleaning up cases and the eradication of cartels, I think that is an absolutely critical point—we cannot allow a situation to evolve where the main motivation for settling is just to get the cases through the door and tick boxes about the number of cases we have resolved in one year. The whole purpose has to be eradicating cartels. What we are trying to do is to demonstrate to the business world that cartels are unacceptable, and we can do that in combination, by some very short decisions in some cases, where the parties come in and talk sensibly, and some pretty tough decisions in other cases. I think that the tension can actually be resolved; it is not as obvious as it may seem. Finally, I happened to be the lawyer in the case mentioned by Christopher Bellamy at the end of the previous session, and I was quite surprised to see our fine go up from 20 to 30 million Euros, and even more surprised when, despite two orders of the court, the minutes of the Commissioners’ meeting were not produced until the day of the oral hearing. DONNCADH WOODS—Philip Roth took the words out of my mouth on the issue of functionality. The only other issue that I wanted to mention is that Regulation 1/2003 is based on a particular concept of ne bis in idem—it is based on the concept that you can allocate work between different authorities. If you had a cartel which initially had effects in one Member State, and let us say France prosecuted it, and there was a decision finding an infringement, and then it became apparent that the cartel was operating in other Member States as well, if you were to adopt a strict interpretation of the ne bis requirement, in the sense of ne bis in facts, then you would not be able to act on the cartel in the rest of the EC or at EC level. The whole Regulation 1/2003 system is based on a concept of ne bis taking into account territorial effect. I think that, especially in competition law, this effects-based concept is a workable way to go. I simply wanted to mention the recent ECJ judgment in Archer Daniels Midland ,26 which discusses the relationship between EC law and third countries, and in this case the ECJ equated idem with effects.

䉴

䉴 THOMAS BARNETT—If I could briefly take up your question, to agree with and reinforce your perspective, my own take is that respecting attorney-client privilege is very important in terms of encouraging compliance with the law, and not only for encouraging people to get legal advice as things are forming. Based on my personal experience, I think that in-house counsel can play a key role in getting business people to trust somebody enough so that they will

26 Case C-397/03 P Archer Daniels Midland Company and Archer Daniels Midland Ingredients Ltd v. Commission [2006] ECR I-4429.


12/10/07

13:11

Page 419

Panel Discussion

419

open up and confess something that may be a criminal act. Sometimes it is easier to talk to the legal counsel with whom they have been working for a number of years than it is to persuade them to go to the outside counsel. But I would also underscore that, after the criminal conduct has occurred and we are in the leniency negotiation or investigation process, external legal counsel who have a lot of experience and integrity can be great facilitators in the process: inside information is helpful, and the Antitrust Division in its cartel enforcement does not request waivers of attorney-client privilege. We respect that and really go out of our way in many ways to preserve it because we see a value to it. ASSIMAKIS KOMNINOS—I have a question for Judge Forwood, although I appreciate that it might not be possible for him to comment upon that. Do we see in Europe the possibility to challenge acts by the Commission relating to leniency before the final infringement and fining decision is taken? In other words, would it be possible for a company that has seen its leniency status withdrawn to challenge this act before the CFI prior to the actual final decision?

䉴

䉴 NICHOLAS FORWOOD—I think under our present case law it would be difficult. There are a number of decisions taken prior to the final decision, which have been accepted subject to the possibility of review, for example, refusal to grant confidential status to a document, and so on. But by and large the line of our case law suggests it would be very difficult, at least until the final fine is imposed—after all, the claim may be purely theoretical—we do not know whether, at the end of the day, the Commission is going to find an infringement and impose a fine, and if does neither of those things, the question is theoretical.

䉴 PETER ROTH—I just wanted to respond to something that Julian Joshua said about self-incrimination, because perhaps I had not made myself very clear. I was not suggesting that privilege against self-incrimination should not take account of the difference between corporations and individuals; on the contrary, I think it should reflect that in the scope of the privilege. What I was saying is, first, that the Orkem rule 27 as such and the distinction it seems to make between permissible and impermissible questions seems to me not to be very satisfactory and murky. One can take the view that companies should not have the privilege against self-incrimination at all—you could say that one of the obligations you assume as a quid pro quo of legal personality is that you do not have such privilege. That would be quite an extreme step to take,

27



420

12/10/07

13:11

Page 420


and if one took it, it would not apply only in the field of competition law, and one would have to consider the implications of that across the board. As for the point on individuals: it arises once you have a situation where you have coercive questioning of individuals—which we do, not at the Commission level, but in some Member States. Secondly, you have a situation of exchange of information between national competition authorities within the network. Thirdly, you have a situation with criminal sanctions against individuals in some Member States. When you have those three conditions in place you have a real problem of self-incrimination of individuals, and you have to have effective safeguards and protection mechanisms to deal with that. The articulation in Article 12 of Regulation 1/2003, which tries to find some way through that problem, is perhaps a good faith effort but it is not a satisfactory answer. DONALD KLAWITER—First, I wanted to underscore what Thomas Barnett said about the US process, because of the complexity of these kinds of cases where there is in-house counsel and outside counsel, and the extraordinary fact that, in most white collar criminal cases in the US other than antitrust, the prosecutors are making a very aggressive push to get you to waive attorney-client privilege as a condition of cooperation. The ABA has protested about that to whoever would listen to us, because when you waive the attorney-client privilege there is a risk of taking out the heart of candid attorney-client communications. So I commend the Antitrust Division on its position, which makes sense in this context, and I do not think it deprives it of any single shred of information as a result of doing it that way. Indeed, in the course of leniency programmes we attorneys have a better relationship and offer more information than if we were obliged to renounce the privilege. Second, in the context of the European system, where you have to worry about the attorney-client privilege issues, we are in a structure nowadays where many of these investigations are global—the Air Cargo investigation, and others in the chemical industry, have been the subject of five, six or more jurisdictions coordinating. It would seem to me that, in such cases, there should be an opportunity for defence counsel to be able to do the same thing—within some limits, of course. The idea that you can communicate freely among that group of people for a global defence is, I think, a very important point and one that deserves some consideration from the Commission and the EC Courts.

䉴

JON LAWRENCE—Two short points, one on the attorney-client privilege, one on the UK position, and a bit of a tease for Philip Collins. On the EU position: my clients simply cannot understand what I call the “dawn raid dilemma”: a client is faced with a Commission raid, and the Commission seeks access to a document for which it is unclear under the AKZO Nobel rule

䉴


12/10/07

13:11

Page 421

Panel Discussion

421

whether it is disclosable or not. The client cannot understand that, if he discloses the document, the Commission can look at it without infringing its rights of defence, even if the document subsequently proves to be privileged. The alternative for the client is to refuse to hand over the document, with the risk of being accused of obstruction. As I understand, the Commission takes the view that it can access servers overseas, so it potentially can access privileged material outside the EU. This kind of issue is arising increasingly frequently in dawn raids. Now, we can increase powers of investigation and incrimination, but then, even if we do not have greater rights of self-defence, we do need clear rights of self-defence. Secondly, I come from a privileged English background, where we have litigation privilege and broad protection from the courts, and this favours the administration of justice. But it is with some concern, now that we have criminalization, that I read in the OFT guidelines that they take the view that, with their new code of surveillance powers, they can listen in—admittedly, in exceptional circumstances—to lawyer-client conversations. I take the view that this must be a breach of the ECHR, but nonetheless there are public statements out there. PHILIP COLLINS—Just one comment on the point about servers. I think that just goes to the need for companies to make it absolutely clear, when they retain information in electronic form or physical form, that there are certain classes of documents that may be privileged. The ability of a well-informed lawyer, in negotiations, to ensure that certain information is actually protected while the discussion goes on, is something that I always found very valuable. In relation to surveillance powers, these are powers that have not been used to a great extent so far, and they are subject to the most rigorous constraints from the Office’s Surveillance Commission. I can assure you that, to the extent that there are issues out there about human rights, they will be addressed, if they have not been already.

䉴

䉴 JOHN RATLIFF—First, I would suggest that Rafael Allendesalazar’s focus on the privilege in relation to defence simply is not wide enough. The whole concept here is getting companies to comply—they need to have recourse to their lawyers to find out what they are doing. I think this is what Julian Joshua was saying, partly, when he was saying, “focus on the function of the test”. It is not a question of the proceedings having started, it should be that a company is able to go and get advice in advance, which was an open issue in Mannesmannröhen.28 Hopefully in AKZO Nobel 29 it will be made crystalclear that you can go to your lawyer and have a privileged conversation and

28 29

See supra note no. 23. See supra note no. 20.


422

12/10/07

13:11

Page 422


work out your position in advance of any proposed conduct under that privilege. I regard that as a position of principle, and as a positive in promoting compliance. Sometimes I think that in these cases people get lost in nittygritty details and forget fundamental principles. The dawn raid dilemma mentioned by Jon Lawrence is the same: the investigating authority says: “I am going to read the document, because from an administration point of view the investigation must be able to go on. I have to do this now, it is a simultaneous dawn raid across Europe; we have to go on.” But it is offensive to everybody who thinks about it that the investigation official should read a document and make up his or her mind about whether it is actually privileged or not. If they read it, the privilege is lost, and they cannot just forget what they read! I heard one practical suggestion, informally, from an official, which was, “put it in an envelope and give it to the hearing officer at the Commission”, who will not be allowed to sit in the proceedings . . . or otherwise disclose the contents inside DG Competition. It is not a perfect solution, but it may be a practical one. Something like this has to be done; otherwise it is offensive to basic principles. 䉴 CLAUS-DIETER EHLERMANN—I do not want to draw any conclusions, just one or two reflections. In 2003, we discussed monopolization in this forum. There were of course divergences of approach between the two sides of the Atlantic, and last year (in 2005) we discussed intellectual property, and again there were notable differences between the two sides. This time I find there was convergence to an extraordinary degree. In antitrust there usually is a West-East wind of inspiration—we learn so much from the US, and we will continue to do so (except perhaps in the area of private enforcement, at least when it comes to treble damages and ambulance-chasing). Secondly, the discussion yesterday and today was a perfect illustration of the enormous benefits from having economists among the lawyers. I am very grateful to John Fingleton and Massimo Motta for helping us bring in this valuable input.

(U) Ehlermann 07 Session B Walbroek&Slater

16/10/07

13:59

Page 423

SESSION B


2 PANEL V: PRIVATE ENFORCEMENT

WORKING PAPERS


16/10/07

13:59

Page 424


16/10/07

13:59

Page 425

I Denis Waelbroeck and Donald Slater1 The Commission’s Green Paper on Private Enforcement: “Americanization” of EC Competition Law Enforcement?

A. Introduction In this paper, we do not want to dwell at length on just how rare damages actions are, nor even on why they are so rare. This work was largely done in the so-called “Ashurst Study”, which gave—as far as possible (more than 800 pages!)—a neutral and ‘photographic’ picture of the current situation across Europe and of the problems encountered. Instead, and after briefly reiterating the main findings of the above study as regards the prevalence of damage actions in Europe (see Section B), we wish to focus on the more contentious issues of why damages actions should or should not be encouraged (Section C), and on the possible changes to the conditions for obtaining damages (Section D). In this regard, it is our view that some prudent and well-balanced changes can be justified—provided they follow a strict ‘compensatory’ model and provided they do not promote the excesses of a US-style litigation culture.

B. Just how rare are damage actions for breach of the competition rules in Europe? Although there is no way of measuring the figure accurately, it would seem that the number of private, competition-based damages actions that proceed to judgment is extremely limited. As is well known, the Ashurst Study found that, in the whole 50-year history of the EC, the number of such cases is well below 1002 (and indeed, the number is below 30 if cases based on national 1 Ashurst. The authors participated in the writing of Ashurst’s “Study on the conditions for claims for damages in case of infringement of EC competition rules” (“Ashurst Study”, 2004), available at: http://ec.europa.eu/comm/competition/antitrust/others/actions_for_damages/ comparative_report_clean_en.pdf. 2 Supra note 1, at page I-1. This figure includes only actions for damages, and does not include cases where injunctive relief was claimed. Nor does it include contractual disputes where the competition rules were invoked by a party in order to challenge the validity of the contract or of one of its clauses.


426

16/10/07

13:59

Page 426


competition law are ignored and only those based on EC competition rules are taken into account). True, that figure probably masks a much larger number of cases that are settled or withdrawn before judgment, and there is undoubtedly an even greater number of disputes in which damages actions are threatened but where a settlement is reached before the formal commencement of court proceedings.3 Moreover, there are also anecdotal accounts of cases that are resolved through arbitration, the outcome of which is kept secret. Other than the rather limited numbers of cases that are fully litigated, there is therefore no real way of even beginning to measure the level of private enforcement in the form of (threatened) damages actions. In the US, it is generally recognized that only some 15% of cases proceed to judgment, while the others are settled. If this proportion is similar in the EC,4 then the number of cases still remains rather limited when compared with the US.5 In addition, not only is the number of cases particularly low—the amount of compensation granted has remained very limited, rarely exceeding € 1 million and usually falling far short of that figure.6 The compensation paid out can therefore in no way be compared with those awarded, e.g., in the US. However, although this relative “rarity” of damages actions in Europe is a fact that is difficult to deny, the rationale for encouraging damages actions is less clear, as we will see below. In any event, the prospect of promoting such actions raises many contentious issues. This is not to suggest that all of the options in the Green Paper7 should be controversial. For example, measures aimed at enhancing the efficiency of procedures, or improving the level of judges’ expertise should generally be approved. However, a lot of the policy choices the Commission is faced with are extremely contentious and basically involve either making it easier for one party (generally the claimant) to win or increasing the winner’s potential reward. Any policy choices that entail a shift in the risk/reward balance that currently exists between potential claimant and potential defendant risk seri3

See the Ashurst Study, supra note 1, at page I-114. We tend however to believe that, if anything, fewer cases are settled in the EC than in the US, as the incentives to settle are necessarily less intense in the EC (lower amounts of damages are granted; there are no class actions; there is no discovery; there are no trials by jury; treble damages cannot be recovered; etc.). According to the comments of Cefic on the Commission’s Green Paper on damages actions for breach of EC antitrust rules, “in the USA—unlike in European practice—cartel cases are almost never concluded by a court decision but practically always by settlement” (emphasis added). 5 However, it is true that, even in Europe, there seems to be a current a trend towards an increase in the number of damages actions. 6 The amount exceeded € 1million in the French Mars case (€ 5.2 million) and in the Italian Telesystems case (€ 1.8 million), but recoveries of damages are usually for far lesser sums (e.g., £ 131,000 in the Crehan case). 7 Available at: http://ec.europa.eu/comm/competition/antitrust/others/actions_for_damages/gp.html. 4


16/10/07

13:59

Page 427

The Commission’s Green Paper on Private Enforcement 427 ously changing the negotiating positions of parties to potential litigation as they seek to settle,8 which apparently is the objective pursued by the Commission. In this regard, we note here the comment by Neelie Kroes about the need “to motivate victims of antitrust infringements to bring damages claims”, and that therefore “the potential benefits of such an action must clearly outweigh the possible costs”.9

C. Objectives pursued by the Commission in trying to facilitate damage actions According to the Working Paper accompanying the Green Paper,10 the main aims that the Commission associates with private enforcement of EC competition law are (1) compensation, (2) deterrence, (3) detection of infringements, (4) bringing competition law closer to the citizen, (5) the possibility for successful plaintiffs to recover costs, and (6) competitiveness. A number of observations can be made regarding these advantages and objectives.

1. Compensation Compensation in itself is a perfectly legitimate end. Indeed, we would argue that it is the main objective to be pursued in this context. Moreover, it is fully in line with the legal systems and traditions of the 25 Member States. Accordingly, damages actions should mainly pursue a compensatory—and not a punitive role. Indeed, in drafting our Ashurst study, one of the key messages obtained from almost all our national reporters related to the fundamentally compensatory nature of damages in the various national legal systems, which excludes any form of “over-compensation” or multiple damages. Interestingly, this basic and fundamental characteristic of damages actions across the Continent can in fact be traced back to the development of criminal law in Europe. Under Roman law, at a time when public enforcement was weak, the main purpose of tort law was clearly to punish and deter. Penalties potentially amounting to several times the actual damages sustained were 8 See Neelie Kroes, opening speech at the conference Private Enforcement in EC Competition Law: the Green Paper on Damages Actions, Brussels, 9 March 2006, available at: http://www. europa.eu.int/rapid/pressReleasesAction.do?reference=SPEECH/06/158&format=HTML&age d=0&language=EN&guiLanguage=en. 9 Ibid. 10 See Section 1.1 of the Working Paper on the “Background and Objectives of the Green Paper”.


428

16/10/07

13:59

Page 428


directly paid to the victim. However, as the systems and society overall matured, and particularly as criminal law developed, the latter took over this function and civil damages became exclusively compensatory in nature.11 Of course, legal systems can be changed along with their underlying philosophies. However, in considering any reforms in this area, one should not lose sight of the deep-rooted nature of the compensation philosophy in the civil law traditions of the Member States, which is so strong that it is often perceived as a form of public policy. Thus, in Europe, US court decisions imposing treble damages are generally considered to be unenforceable (and US suits aiming at such damages as inadmissible), as being contrary to the ordre public. In this regard, reference is generally made to Articles 13 and 23 of the Hague Convention on the taking of evidence abroad in civil or commercial matters, which contain a “public policy” reservation. Two points should therefore be underlined. Firstly, certain measures envisaged in the Green Paper are in clear and open conflict with the compensatory philosophy underlying damages actions in the Member States. The most obvious of these is the possible introduction of punitive damages. However, also in contradiction with the principle of compensation is the possible suppression of indirect purchasers’ standing to bring such actions, even if they have suffered harm.12 It is clear that a shift from a compensatory to a more deterrence-oriented system might unavoidably contribute to creating the “litigation culture” that the Commission claims to wish to avoid at all costs.13 Secondly, while the Commission presents compensation and deterrence as the twin pillars of private enforcement, thereby implying some form of “equality” between the two aims, this does not appear to reflect the current state of the law in the Member States. And even under EC law, there is scant evidence in precedents either way on this question, although the relevant passages from the Crehan judgment14 would tend to indicate that deterrence results from the compensatory aims of private damages actions, rather than being an end in itself. This is not just a theoretical debate. If deterrence is not considered a standalone objective of EC competition law-based private damages actions, then initiatives such as double damages would be fundamentally flawed.15

11 See, for example, in the case of Belgian law, H. De Page (1967): “Traité Elémentaire de Droit Belge”, vol. II, 3rd edition, n° 1031. Punitive damages virtually do not exist on the Continent anymore, and where they do still exist, it is almost by oversight. Thus, in Belgium, the only (rather anecdotal) example is Article 7 bis of the Belgian Act of 28 February 1882 relating to damages caused by rabbits. 12 See Option 22 of the Green Paper. 13 As Neelie Kroes has stated: “We must avoid excessive levels of litigation. We must avoid speculative lawsuits prompted by ambulance-chasing lawyers. We must avoid an avalanche of unmeritorious claims.” Cited supra note 8. 14 Case C-453/99 Courage v Crehan [2001] ECR I-6297, paragraphs 26 and 27. 15 See Option 16 of the Green Paper.


16/10/07

13:59

Page 429

The Commission’s Green Paper on Private Enforcement 429 In this regard, any move towards non-compensatory damages16 would not only be at odds with the general approach taken in the Member States,17 and indeed at odds with the current trend in the United States, but it would also contradict the approach taken by the Commission itself on this question in other areas, where it has proposed legislation that would render such awards contrary to public policy.18 Interestingly, in his opinion of 26 January 2006, in the Manfredi case, Advocate General Geelhoed made the point that damages should essentially be “compensatory”, i.e., that damages suffered by the claimant should be fully compensated. However, as the Advocate General pointed out, punitive damages are not required under EC law, and they are certainly not necessary to apply the principle of effectiveness.19

2. Deterrence As indicated above, we find compensation to be a perfectly legitimate objective in itself. Nevertheless, the objective of achieving deterrence indirectly through damage actions appears more problematic to us, unless it is perceived as a “side benefit” of a proper compensation system. In addition to the comments above, we would make in particular the following observations in relation to deterrence as an aim of private enforcement. Firstly, the point has often been made that the current system of enforcement of Articles 81 and 82 EC already has, in itself, a strong deterrent effect. No less than twenty-six competition authorities are responsible for detection in the EC,20 to which further private enforcement is arguably rather unlikely 16 While the Green Paper envisages a range of options for defining damages, it is interesting to note that elsewhere the Commission appears very tentatively to be coming out in favour of some form of multiple damages (see SPEECH/05/533 by Neelie Kroes on 22 September 2005). 17 Only three of the Member States allow for a system of punitive damages (namely, the United Kingdom, Ireland, and Cyprus). 18 In the Commission’s original proposal for a second regulation on the law applicable to noncontractual obligations, Article 24 explicitly excluded non-compensatory damages as being “contrary to Community public policy” (emphasis added) (Proposal for a Regulation of the European Parliament and the Council on the law applicable to non-contractual obligations (“ROME II”), COM/2003/0427 final). In its recent revised proposal, the Commission practically executed a complete volte-face, stating that it could agree to Article 24 being removed, subject to drafting changes, “to make clear that punitive damages are not ipso facto excessive”. See the preamble to the amended proposal. 19 See Opinion of Advocate General Geelhoed, 26 January 2006, in Case C-295/04 Manfredi, [2006] ECR I-6619 at para. 70. 20 On the deterrent effect of this multiplication of authorities, see Waelbroeck D. (2005): “Twelve Feet All Dangling Down and Six Necks Exceeding Long. The EU Network of Competition Authorities and the European Convention on Fundamental Rights”, in C.-D. Ehlermann and I. Atanasiu, European Competition Law Annual 2002: Constructing the EU Network of Competition Authorities, Hart Publishing, Oxford and Portland, Oregon, pp. 465 et seq.


430

16/10/07

13:59

Page 430


to contribute much.21 Moreover, the leniency system has largely contributed to destabilizing existing cartels. Finally, the fines imposed on undertakings for violations of the competition rules are significant and the trend appears to be upwards (nearly € 4 billion imposed by the Commission since 2001), with national competition authorities also regularly imposing record fines in their jurisdictions. In addition, criminal sanctions such as imprisonment are now possible in certain Member States. In any event, if additional deterrence is necessary at all, it is unclear whether it should take this form (i.e., amounts of compensation which can be highly unpredictable and unsuitable to achieve a proportionate effect of deterrence) and not other forms (such as higher fines, imprisonment, etc.).22 Secondly, and maybe more fundamentally still, undertakings cannot be deterred from violating the law when there is significant lack of clarity as to what behaviour constitutes a violation of the law. In practice, there are a vast number of situations where it is far from clear whether particular behaviour is contrary to EC competition law, in particular in relation to Article 82 EC, to vertical agreements under Article 81 EC, and even to many horizontal agreements under Article 81 EC (such as information sharing and joint R&D). Where undertakings take a bad risk in this grey zone (or a judge simply takes a different view on the legality of their behaviour—see comments below at point 6 on the cost of “bad decisions”), they could find themselves severely penalized by private damages actions. Where they choose not to take risks, they and society as a whole may miss out on a wholly legitimate opportunity to increase competitiveness and improve conditions of competition. The potential for such increased stakes to chill competition is obvious.

C. Detection Another objective of the Commission’s drive to develop private enforcement is to complement public enforcement. However, in doing so, one must not lose sight of the material differences that exist in the nature of private and

21 It is noted that, generally speaking, private actions in the US either follow on from prior national competition authority decisions, or they are actions under Section 2 of the Sherman Act (the equivalent of Article 82 EC). 22 In addressing these issues, there is a need to consider further the practical motivations of individuals involved in and responsible for antitrust violations. In this regard, see, e.g., Pitt (2006): “Markets are Instruments Not Values”, 27 European Competition Law Review 1. See more generally Wils W. P. J. (2003): “Should Private Antitrust Enforcement Be Encouraged in Europe?”, 26 World Competition—Law and Economics Review 473.


16/10/07

13:59

Page 431

The Commission’s Green Paper on Private Enforcement 431 public enforcement of the competition rules, a fact that is already recognized in EC law.23 One major difference between public and private enforcement is their underlying philosophy. On the one hand, private enforcement is used by individuals for individual ends, whatever these might be (compensation, “specific” deterrence,24 etc.). On the other hand, public enforcement is undertaken in the interests of the population at large and seeks to enhance the overall common good. While the Commission insists that the purpose of facilitating damages actions is not to replace public enforcement but to complement it, one has the impression from the Working Paper that the Commission sees damages actions as a means to achieve a greater volume of enforcement. In other words, it seems that such actions are conceived to be an extension of the Commission’s own efforts and those of the 25 national competition authorities in this area, without necessarily giving sufficient weight to the differences inherent in the two types of enforcement. Indeed, the fact that public enforcement is generally selective and assigns priority to the cases that are most likely to contribute to overall welfare means that it tends to shy away from those where there remains significant controversy. Conversely, private enforcement is far more indiscriminate. Cases are brought for individual ends and are heard by instances with far less expertise than that of competition authorities. This means that increased private enforcement entails costs that either do not exist or are far more limited in the case of public enforcement. Generally, it seems to us that the Commission is particularly keen to encourage damages actions in order to uncover more secret cartels. However, even in the US, where so much is done to encourage damages actions, such actions serve more often as either follow-on suits or as actions under Section 2 of the Sherman Act (the equivalent of Article 82 EC) than as means to discover unknown cartels. If anything, one may therefore expect non-hardcore violations to become the main area of private enforcement (in addition to mere “follow-on” actions, which add nothing with respect to this objective of detecting additional infringements). A final key issue relating to the complementarity between public and private enforcement, which the Commission has identified clearly, is the impact increased private enforcement will have on leniency applications. This relates to the fact that the prospect of increased exposure to private litigation at national level clearly reduces the incentive for undertakings to contribute to 23 “[W]here a natural or legal person asks the national court to safeguard his individual rights, national courts play a specific role in the enforcement of Articles 81 and 82 EC, which is different from the enforcement in the public interest by the Commission or by national competition authorities (see Case T-24/90 Automec [1992] ECR II-2223 at point 85).” See Commission Notice on the co-operation between the Commission and the courts of the EU Member States in the application of Articles 81 and 82 EC, OJ C 101 [2004]. 24 By “specific” deterrence we refer to an individual business strategy aimed at deterring specific undertakings from engaging in particular types of behaviour.


432

16/10/07

13:59

Page 432


detection by blowing the whistle and requesting leniency.25 One possible response aired by the Commission on this point26 is the introduction of double damages from which leniency applicants would be exempt. However, although this would achieve a relative difference in exposure, this does not get round the fact that private actions against leniency applicants will otherwise be facilitated in absolute terms. More importantly, it does not address the question of whether the advantages from facilitated private enforcement outweigh the reduction in detection rates resulting from a possible drop in leniency applications. In this regard, while a key advantage to the claimant is that he can more easily claim compensation through the courts, any reduction in detection rates will mean that other groups of potential claimants might never even be made aware that they have suffered a loss as a result of a violation of the competition rules.27

4. Bringing competition law closer to the citizen A further objective pursued by the Commission in facilitating damages actions is “to bring competition law closer to the citizen”. However, in seeking to facilitate and encourage private damages actions, there is, as the Commission itself recognizes, a difficult line to tread between, on the one hand, encouraging a ‘culture of competition’ and, on the other hand, fostering a ‘culture of litigation’, in particular where the costs in terms of resources spent in obtaining justice far outweigh any benefits to the individual litigant (or indeed to society as a whole). In other words, the main beneficiary ends up being the lawyer rather than the victim!28

5. Cost recovery Another reason why the Commission believes that increased levels of private enforcement are desirable is that successful claimants may be able to recover costs. However, the issue of costs is a very complex one. In relation to costs, the Commission has publicly stated that, “if we want to motivate victims of 25

Options 28–30 of the Green Paper. The Commission does not make this explicit in the Green Paper itself. See SPEECH/05/533 by Neelie Kroes, delivered on 22 September 2005. 27 A significant proportion of private damages actions in the US are follow-on actions, i.e., actions that rely on prior detection of a competition law violation by the US authorities. 28 The costs of an excessive tort system can be considerable, as shown by the US model, where it is estimated that it costs the economy roughly USD 250 billion a year (i.e., USD 800 per US citizen per year. See http://www.towersperring.com/tillinghost/publications/reports/Tort_2004/Tort.pdf). 26


16/10/07

13:59

Page 433

The Commission’s Green Paper on Private Enforcement 433 antitrust infringements to bring damages claims, the potential benefits of such an action must clearly outweigh the possible costs. We can achieve that by reducing the costs involved, for example by spreading them in case of a collective action. Or by reducing the risk by guaranteeing a reasonable claimant that he will not have to bear the bill. Or by increasing the benefits of a damages action.”29 A number of points should be made in relation to this general statement that potential benefits must clearly outweigh costs, and in relation to the solutions considered by the Commission. Firstly, while very general considerations of ‘fairness’ in a perfect world would indicate that an unsuccessful plaintiff should not have to pay the costs of litigation so long as his claim was reasonable, it is difficult to see who else should pay the bill in such cases. Imposing such costs on the taxpayer might be a theoretical option.30 However, it would then arguably be reasonable to require successful plaintiffs to contribute part of their award to the maintenance of such a system. On the other hand, short of emptying a judgment of much of its meaning, there does not seem to be any justification for imposing the plaintiff’s costs on a defendant who has prevailed in the litigation. As regards the defendant’s own costs, the general rule in the Member States is that, if the plaintiff loses, he must also pay these costs.31 This rule appears justified, to the extent that the defendant has been found not to be liable and therefore should not be required to pay for the experience of being unsuccessfully sued.32 In any event, if the Commission wishes to give specific guidance on who— other than the plaintiff—should bear the costs in an unsuccessful (but reasonable) suit, it is clear that such guidance will have to justified. It should also specify how reasonable claims should be distinguished from unreasonable ones. Secondly, objectively speaking it is no doubt true to state that if potential benefits clearly outweigh costs, this will motivate more private actions. However, this does not address the preliminary question of whether it is at all desirable generally to encourage more such claims.33 Nor should it be overlooked that encouraging more claims by ‘true victims’ in this way will also 29

See Speech of Neelie Kroes, 9 March 2006, supra note 8. In the recent amendment to the German Competition Act, it is envisaged in some cases that the Bundeskartellamt may fund the costs of the procedure. 31 Even where this is the rule, in practice the defendant may not be able to recover all costs for various reasons, e.g., insolvency of the plaintiff, general caps on costs imposed by regulation, or case-specific caps imposed by the court itself. 32 Requiring each party to bear its own costs was essentially the approach advocated recently in the UK by Colman J. in Arkin v. Bochard Lines [2003] Comm EWHC 2844. However, this ruling on costs was overturned by the Court of Appeal in Arkin v. Borchard Lines [2005] EWCA Civ 655. 33 In this regard, the Commission sometimes appears more interested in increased levels of litigation in the courts than in behind the scenes settlement of disputes, even though both can lead to compensation of victims and deterrence of anticompetitive behaviour. Thus, the Commission has stated that it wants “to use the [current] debate to find ways to increase deterrence, while avoiding the situation where defendants settle simply because litigation costs are too high”. Speech of Neelie Kroes, 22 September 2005, supra note 26. 30


434

16/10/07

13:59

Page 434


inevitably encourage more unfounded claims. Finally, it is unclear why competition law, as opposed to other complex areas of economic law, such as securities, requires specific rules on costs.

6. Competitiveness A final reason given by the Commission for encouraging damages actions is that it will enhance the competitiveness of the European economy (see the ‘Lisbon agenda’). At the same time, in its Working Paper the Commission stresses the need for a “balanced approach”, which involves taking into account the social costs of increased private enforcement. There are a number of such costs, some of which have already been alluded to above, e.g., the indirect cost to consumers through increased prices (which fund the damages awards), the cost to the taxpayer arising from a higher level of litigation, the ‘cultural’ cost of increased litigation, the negative impact on public enforcement, etc. However, in our view, the main changes may lie in the cost of ‘bad decisions’. Our definition of ‘bad decisions’ here are judgments that require undertakings to pay damages for alleged breaches of competition law where the undertaking’s impugned actions actually had no negative effect on competition or consumers whatsoever. We consider that, in the field of competition law, more litigation implies a greatly increased risk of such ‘bad decisions’ (or ‘false positives’), in particular due to the complex nature of the field, the lack of expertise of many courts and—perhaps most importantly—basic disagreements as to whether certain behaviour is actually anticompetitive or not (which, moreover, are subject to great shifts over relatively short periods of time34). One major cost of bad decisions is that they wrongly punish an undertaking by requiring it to pay damages for innocuous behaviour (damages that will most likely be paid in the long run by consumers that purchase the goods of the unsuccessful defendant). Another cost that is harder to measure, and potentially much more damaging, is the deterrent effect of bad decisions. Indeed, although private enforcement may undeniably be useful as a deterrent to anticompetitive behaviour and a boon to competitiveness, one should not ignore the extent to which bad decisions will potentially deter undertakings from competing aggressively and thus chill competition. Of course, public authorities may also make mistakes in enforcing competition law. However, the courts enforcing competition law in private actions do not have the same level of 34 Take, for example, the change in the EC law approach to vertical restraints over the past forty years, which will no doubt continue to evolve considerably over the coming forty years.


16/10/07

13:59

Page 435

The Commission’s Green Paper on Private Enforcement 435 expertise as the Commission or national competition authorities, and they will very likely be more prone to errors as a result. This problem is exacerbated by the fact that courts do not have a choice as to which cases they take. Unlike competition authorities, they cannot prioritize and focus resources on the most obvious and egregious abuses. They are also compelled to deal with the most borderline and arguably insignificant cases. Moreover, the consequences of errors by national courts are potentially far greater than mistakes by competition authorities, not only in financial terms but also in terms of the potential chilling effect on competition: private enforcement may make the citizen more aware of the virtues and financial opportunities of competition law, but it will also make undertakings more aware of its vices.

7. Conclusion In view of the above, it seems to us that the main objective of damages actions is and should remain compensatory. If such actions result incidentally in additional deterrence or detection of infringements, this may be a welcome side effect, but it should never be the main objective of any reform in this area. However, a major concern in any reform, which seems so far to have been largely overlooked (by the Commission, at least), should be the complexity of the law in this area and the risk that ‘bad decisions’ will not only result in injustice in individual cases but will also have a major chilling effect that will negatively impact the economy as a whole.

D. Conditions of competition law-based damages claims On the basis of our comments above on what we consider to be the proper goals of any reform in this area, we consider below a number of aspects of competition law-based damages claims and how and whether they should be altered.

1. Illegality—The need for specialized courts? There are two aspects of the illegality condition that should be highlighted here: firstly, in relation to prior competition authority decisions; and secondly, in relation to per se rules of illegality. Prior competition authority decisions are discussed further below (at point 6.2.). As regards per se rules of illegality, the term is used here to describe types of behaviour that carry a very


436

16/10/07

13:59

Page 436


strong presumption of illegality, such as price fixing or other restrictions ‘by object’ in the context of Article 81 EC cases or resale below average variable cost in the context of Article 82 EC cases. There are arguments to suggest that the existence of such per se rules is in fact the major factor which influences the prevalence of private enforcement. Indeed, in the context of private enforcement in the United States, Ginsburg and Brannon argue 35 that the evolution of the level of private enforcement there correlates closely with the “rise and fall” of the per se rule in the US, particularly with reference to the per se condemnation of non-price vertical restraints that took place in the 1960s.36 If it is indeed the case that the prevalence of per se rules is the major influence on the level of private enforcement, then the Commission’s apparent focus on costs, damages and standing in the context of its current investigations into the subject is somewhat misplaced. Focus should rather be, for example, on the parallel Article 82 EC modernization project, which seeks to define the rules governing abuse of dominance (and in doing so, creates a plethora of more or less strong presumptions of illegality, based on the form of conduct involved). If one is keen to avoid the kind of excessive litigation common in the US, the lesson might be that the Commission should be far more wary of creating presumptions of illegality than it has been in the past. Indeed, the damage done to competition from an overly zealous creation of presumptions of illegality can only be compounded where individuals are actively encouraged to sue on the basis of such misguided approaches. Another lesson to be learned is that it might now be useful to envisage—at the current level of development of competition law, and given its increasing complexity—the creation of specialized courts, on the model of the CAT in the UK. (Given the limited number of damages actions, one idea might be to create regional competition courts, having jurisdiction inter alia for damages actions.)

2. Fault There are a number of different approaches taken in the various Member States to the fault requirement.37 Some Member States require fault (in some 35 Ginsburg D. and Brannon L. (2005): “Determinants of Private Antitrust Enforcement in the United States”, 1 Competition Policy International 29. 36 See, e.g., United States v. Schwinn, 388 U.S. 365 (1967) (territorial restraints); Albrecht v. The Herald, 390 U.S. 145 (1968) (maximum resale price maintenance); Siegel v. Chicken Delight, 448 F.2d 43 (9th Cir. 1971) (sourcing of products from franchisor). 37 See Ashurst Study supra note 1, at pp. I-51 et seq. Where fault is required, the degrees of fault vary from one Member State to another—or at least the terminology after translation differs—e.g., negligence, gross negligence, intention, etc. Moreover, the yardstick against which an undertaking’s behaviour is measured may also differ (e.g., average person, average professional).


16/10/07

13:59

Page 437

The Commission’s Green Paper on Private Enforcement 437 form or another),38 whereas others do not require fault at all—effectively creating strict liability for harm caused by breaches of competition law. Still others create rebuttable presumptions of fault.39 One way in which the Commission might facilitate damages actions would be to request Member States to remove any fault requirement—where they impose one—or to create a presumption of fault. However, there are two compelling reasons why such an approach would be mistaken. Firstly, the case law of the ECJ clearly distinguishes between the question of the existence of a breach of EC competition law (where the question of culpability and fault is irrelevant) and the question of civil liability for such a breach (where the issue of attributing fault can be considered). Thus, in Crehan, the ECJ stated that: “[. . .] the Commission and the United Kingdom Government also rightly point out that a contract might prove to be contrary to [Article 81(1)] of the Treaty for the sole reason that it is part of a network of similar contracts which have a cumulative effect on competition. In such a case, the party contracting with the person controlling the network cannot bear significant responsibility for the breach of [Article 81], particularly where in practice the terms of the contract were imposed on him by the party controlling the network. Contrary to the submission of Courage, making a distinction as to the extent of the parties’ liability does not conflict with the case-law of the Court to the effect that it does not matter, for the purposes of the application of [Article 81] of the Treaty, whether the parties to an agreement are on an equal footing as regards their economic position and function [. . .] That case-law concerns the conditions for application of [Article 81] of the Treaty while the questions put before the Court in the present case concern certain consequences in civil law of a breach of that provision.” (emphasis added)40

The fact that Mr Crehan was a party to an anticompetitive agreement did not make him liable for his behaviour. On the contrary, he was perceived as the victim of the behaviour. Although it is well known that fines are conditional on fault, many antitrust infringements exist independently of any fault, as the ECJ has made clear since Continental Can.41 Secondly, the fairness to the defendant of excluding or assuming the fault condition is highly questionable, particularly in cases where the application of competition law can lead to novel or unpredictable outcomes. Examples of

38

Ibid. Ibid. 40 It is clear from this quote that consideration of culpability in the context of competition law based damages actions does not conflict with the description of abuse under Article 82 EC as an “objective” concept (see, e.g., Case 62/86 AKZO v Commission [1991] ECR I-3359, para. 69). 41 In Case 6/72 Continental Can v Commission [1973] ECR 215 the ECJ stated that the abuse existed “irrespective of any fault” (para. 27). In Case 85/76 Hoffmann-La Roche v Commission [1979] ECR 461, the ECJ declared that the concept of abuse “is an objective concept” (para. 91). 39


438

16/10/07

13:59

Page 438


this would include a vast number of cases of alleged abuse of dominance42 and also cases involving vertical agreements having ambiguous competitive effects, or of agreements that are normally legal but become illegal as a result of the specific economic circumstances in the market, such as the cumulative effect of ‘small’ agreements (although the issue is likely to be of lesser importance in cases involving secret cartels). In any event, it can legitimately be asked how suppressing the fault requirement in such cases can in any way contribute to deterrence (unless it encourages undertakings to be overly cautious—i.e., over-deterrence—with the obvious dampening effects this has on the economy as a whole). Finally, it is also noted that, as a general rule, strict liability in Member State tort law is the exception rather than the rule. One must query the underlying justification for introducing strict liability in the sphere of competition law.

3. Damages/passing on In addition to the remarks made above (at points B.1. and B.2.) on punitive damages, identifying the amount of compensation is also a most complex issue, as shown, inter alia, by the discussion on the ‘passing on defence’. While it is true that strict adherence to the compensatory nature of awards in private damages actions means that the passing on defence cannot be ignored, one problem with this approach is that the question of passing on is notoriously difficult to prove either way with any acceptable degree of certainty. Indeed, the main reason that the ECJ in San Giorgio43 objected to a presumption of passing on was that such a rule would impose an impossible burden (a probatio diabolicum) on the claimant.44

42 The law on Article 82 EC is in such a state of disarray that it is currently almost impossible to give undertakings comprehensive and risk-free advice. (Indeed, in a recent speech, the Director General of the Commission’s Competition Directorate, Philip Lowe, described Article 82 EC as “a mess” (speech given in Brussels on 4 October 2005)). It is acknowledged that in certain areas there is a degree of clarity (e.g., the rules on the legality of predatory pricing—although this clarity unfortunately does not extend to the existence of predatory pricing). Nevertheless, even this is of little comfort given the total absence of any kind of systematic approach to the most pervasive behaviours (e.g., discriminatory pricing or tying). Furthermore, there are also truly novel cases, where predicting problems arguably requires recourse to the occult sciences (see, e.g., Case 223/01 AstraZeneca [2003] ECR I-11809. 43 Case 199/82 Amministrazione delle finanze dello Stato v San Giorgio [1983] ECR 3595. 44 On this complex question, see Waelbroeck M. (1982): “La nature du droit du remboursement des montants payés contrairement au droit communautaire” in Liber amicorum Josse Mertens de Wilmers, Antwerp and Zwolle, pp. 429–439; Hubeau F. (1981): “La répétition de l’indû en droit communautaire”, in Revue Trimestrielle de Droit Europeen, 442–470; Schermers H. and Waelbroeck D. (2001): Judicial Protection in the European Union, Kluwer Law International, London, paras. 401–402, 416 and 1131.


16/10/07

13:59

Page 439

The Commission’s Green Paper on Private Enforcement 439 In light of these practical difficulties, it is tempting to consider “abolishing” the passing on defence as envisaged by, for example, Option 22 of the Green Paper (as is well known, the German legislator recently opted for this solution45). However, it is clear that this change could result: (i) in the claimant receiving damages that exceed the harm he has suffered; and (ii) in ‘double claims’ against the defendant where indirect purchasers have standing and are able to show that higher prices have been passed on. The contentious policy choices involved here and the interaction between the passing on and indirect purchaser questions can be seen in the approaches taken in the United States. Whereas at federal level the passing on defence is excluded and indirect purchasers may not claim damages,46 at state level the problem has often been resolved differently. In considering these issues, one point that the Commission must take into consideration is the impact that an abolition of the passing on defence might have in other areas of the law. There seems to be no good reason why, if the passing on defence is abolished in the competition law context, it should not be similarly abolished in other areas, such as fiscal law, since the same practical difficulties also arise in those areas (in those cases, the ECJ has refused to exclude the passing on defence, although it was repeatedly invited to do so by its Advocates General in the San Giorgio and Comateb 47 cases). The fact that changing the basic rules for passing on in the context of competition lawbased damages actions may have a knock-on effect in how we treat the defence in other types of litigation is highly important, and indeed this concern is present in all measures considered in the Commission’s Green Paper that single out competition law-based damages actions as a ‘special case’.

4. Legal basis The approach to legal basis for competition-based damages actions varies across the Member States: some Member States have no discrete legal basis for such actions; others do; and still others mix the two approaches (generally a discrete legal basis for national law-based actions and another non-specific one that provides for EC law-based actions). These differences can give rise 45 See Seventh Amendment to the GWB. In the United Kingdom, there are indications that the CAT may also be inclined to reject the defence. At least in its decision on security for costs in the Vitamins case, published on 17 February 2005, where the CAT argued that the only reason for awarding costs against the claimants would be if the passing on defence was established as a good defence and if it led to significantly lower damages. As this was not the case, the CAT held that this risk should not be borne by the dominant firms and refused to award security for costs to the defendants. 46 See Illinois Brick Co. v. Illinois, 431 U.S. 720 (1977). 47 Case C-192/95 Comateb and Others v Directeur général des douanes and droits indirects [1997] ECR I-165.


440

16/10/07

13:59

Page 440


to confusion and uncertainty, particularly where two separate legal bases exist for national and EC law-based actions.48 A quick and simple way of ‘fixing’ the problem might be to require Member States to introduce a single discrete legal basis (as many already have done). However, this approach relies on the assumption that competition law-based damages actions are generally different from other civil damages actions (which is an approach that has not yet been adopted by the legislature in a number of Member States).49 While the creation of a discrete legal basis per se may appear to be anodyne, the assumption clearly paves the way for further distinguishing the regime of civil liability for harm caused by breach of competition rules (see point C above in relation to the passing on defence). Although it is clear that competition law-based damages actions are in some ways particular, especially as regards their complexity, the case has yet to be made that that they are so different that they justify being singled out in this way.

5. “Group” actions We use the term ‘group’ action to refer to legal mechanisms that are used to allow a number of claimants to bring cases together before national courts. In practice, there are a bewildering number of different types of group actions in the Member States, including, inter alia, those designed to cope (or at least those potentially capable of coping) with situations of diffuse harm.50 Moreover, there has been a rise in the number of initiatives taken by Member States to address the issue of group actions,51 although as yet no Member State has quite the same notion of ‘class action’, as in the US. The Commission, for its part, appears to favour at least some forms of group actions. However, it has also intimated that it does not wish to see the adoption of class action mechanisms (exactly) like those used in the US, and 48 In Italy, for instance, this has resulted in litigation to decide which court was competent in each case, as EC competition rules could be invoked before any court but Italian competition rules could only be invoked before the Court of Appeal. Given that it is often unclear whether a given behaviour is contrary to national or to EC competition rules, this has given rise to a lot of confusion. See the discussion of this problem in the Opinion of Advocate General Geelhoed of 26 January 2006, Case C-295/04 Manfredi, [2006] ECR I-6619, at paras. 41 et seq. 49 The altering of specific conditions such as passing on or standing also makes such an assumption. However, it does so only in relation to specific aspects of competition-based damages actions, whereas introducing a discrete legal basis implies that competition-based actions are different in general. 50 There is no standard terminology in this area, and therefore a list of types of such actions is not useful here. However, see Ashurst Study, supra note 1, at pp. I-43 et seq., where an attempt is made to categorize such actions according to the identity of the plaintiffs and the identity of the beneficiaries of any award. 51 Initiatives have been taken by a number of Member States over the past years, including Italy, the UK and Sweden. See Ashurst Study, supra note 1, page I-46.


16/10/07

13:59

Page 441

The Commission’s Green Paper on Private Enforcement 441 indeed it appears to view the US approach to class actions as problematic.52 The obvious question, to which the Commission has not yet proposed an answer, is: under what circumstances do group actions cease to be a legitimate deterrent and cross the line to become “unmeritorious” and/or “speculative lawsuits prompted by ambulance-chasing lawyers”?53 Perhaps the positioning of this line is somehow related to the chances of success and/or the level of the claim involved (and maybe the relative size of the plaintiff’s lawyers’ fees?). For example, if each individual would only receive a ‘small’ amount of money even if the group won (and/or if there were little ‘objective’ chance of winning at all), then the claim would become unmeritorious. In any event, it is clear that establishing this line is a difficult task for a legislator, even when it does have some guiding principles, and this is particularly so on a continental scale, where there are huge differences in litigation culture.

6. Rules of evidence: discovery and prior competition authority decisions Following publication of the Green Paper, Competition Commissioner Neelie Kroes stated that one of the “central messages of the Green Paper [is that] it’s unacceptable that victims of an antitrust infringement decide not to bring a damages claim or lose such a case in court simply because they do not have access to evidence which is in the hands of the defendant. It must be possible to find reasonable ways to bring that evidence to the attention of the court.” Moreover, “[c]ases that follow public enforcement decisions should be able to rely on those decisions as proof of the fact of the infringement”. These two aspects of rules of evidence—discovery and prior decisions—are discussed below.

6.1. Discovery Currently among the EC Member States, only the UK and Ireland, and to a lesser extent Cyprus, have broad discovery rules. Other Member States’ discovery rules tend to be far narrower, the most prevalent model being that the 52 See, e.g., Mario Monti’s reference to the “excesses” of other jurisdictions (a thinly veiled reference to the US). Arguably, the Commission appears to view class actions as being more concerned with litigation per se than as a means to an end (see Speech of Neelie Kroes, 22 September 2005, (supra note 26), where she remarks that “we [the Commission] want to use the debate to see how we can do the European economy a favour. How we can foster a competition culture, not a litigation culture”.) (emphasis in original). Even in the US, the excesses of class actions are a major preoccupation, and a major reform was introduced recently with the Class Action Fairness Act of 2005. 53 Speech of Neelie Kroes, 9 March 2006, supra note 8.


442

16/10/07

13:59

Page 442


judge may request relevant, specific and identified pieces of evidence from a party at another party’s request. The Commission, quite rightly, notes that a plaintiff’s job would be made easier if it had much easier access to evidence held by the defence. However, going beyond such truisms, the real policy and legal culture question here is: what breadth of discovery can be justified by a balancing of, on the one hand, a plaintiff’s legitimate interest in obtaining access to evidence and, on the other, the defendant’s right to be presumed innocent and not to have that presumption de facto undermined by an overly intrusive justice system which de facto opens the door to fishing expeditions? Over the centuries, this question has been resolved in different ways in different Member States. The Commission therefore has the double challenge of striking a balance and striking one that fits within the legal cultures of the 25 Member States, keeping in mind that extensive discovery systems could in some cases conflict with the basic principle of non self-incrimination or raise complex issues of confidentiality. Another dimension to this question—which is in fact relevant for most, if not all, areas of procedural reform envisaged by the Green Paper—is whether it is legitimate to single out competition law and apply to it specific rules on discovery that are (potentially) very different from the rules applied in other areas. Finally, while discovery arguably brings added value in terms of deterrence, it is unlikely to contribute much to increased detection of cartels, as the Commission seems to hope. As explained earlier, private actions in the US generally either follow on from prior national competition authority decisions or else they are brought as monopolization cases under Section 2 of the Sherman Act.

6.2. Prior decisions The Commission has also come out clearly in favour of making competition authority decisions formally binding on courts as regards the finding of an infringement. In practice, such a situation already exists to a certain extent in relation to Commission competition decisions, by virtue of the Masterfoods judgment.54 However, formalizing the binding nature of Commission decisions to extend it to actions for damages and also to national competition authority decisions again raises serious policy and legal culture questions that must be addressed before taking such a step. 54 Case C-344/98 Masterfoods v HB [2000] ECR I-11369. See also Article 16 of Regulation 1/2003, according to which national courts “cannot take decisions running counter to the decision adopted by the Commission. They must also avoid giving decisions which would conflict with a decision contemplated by the Commission in proceedings it has initiated”.


16/10/07

13:59

Page 443

The Commission’s Green Paper on Private Enforcement 443 A central issue in this regard is conformity with Article 6 of the ECHR55 on access to the courts, as even where there is extensive judicial review of decisions by an administrative court, this is only to assess the legality of the decision itself. However, in reviewing the mere legality of a decision, several points of acts which may be decisive for a separate action for damages may not have been looked at by the administrative court. The principle that civil actions for damages and administrative actions for annulment are entirely autonomous and different in scope and purpose has been repeatedly recognized in EC law,56 and there are similar principles in the national laws of Member States. In other words, this implies that, even after the expiry of time limits for actions for annulment or even after a court has ruled on judicial review, the legality of the decision in question may still have to be assessed in the different context of an action for damages. In any event, it is not clear to us that the issue of the binding effect of decisions of national competition authorities constitutes a serious hurdle to damage actions. We tend to believe that it will be very rare that courts will deviate from findings of well-established competition authorities in other Member States. Moreover, making national competition decisions binding on national courts as regards a finding of infringement requires important changes not only to national legal systems but also to the approach of the Commission and the ECJ to such decisions. In this regard, both institutions have on occasion simply ignored certain national competition authority decisions finding the absence of an infringement.57 It would appear contradictory to make such decisions binding on the judicial authorities of the Member States but not on the administrative (or judicial) organs of the EC.

7. Leniency Another issue raised by the encouragement of private enforcement is how to reconcile this with leniency policies. The essence of the problem here is that, the greater the risk of private actions following on from the exposure of cartels through leniency applications, the less incentive cartel members have to make the leniency request in the first place.

55

European Convention for the Protection of Human Rights and Fundamental Freedoms. See, e.g., Case 5/71 Schöppenstedt [1971] ECR 983, para. 3, stressing that the action for damages is “an independent form of action with a particular purpose to fulfil within the systems of action”, and that it “differs from an application for annulment in that its end is not the abolition of a particular measure, but compensation for damage”. 57 See, e.g., Case T-203/01 Michelin II [2003] ECR II-4071. A non-exhaustive list of contradictory decisions of national authorities is provided in D. Waelbroeck, “Twelve Feet All Dangling Down and Six Necks Exceeding Long”, supra note 20. 56


444

16/10/07

13:59

Page 444


There is a real smorgasbord of solutions in relation to this question, but they all have the same underlying purpose, i.e., to (partly) neutralize the reduced incentive to blow the whistle by somehow giving favourable treatment to leniency applicants in subsequent related private actions. The two most obvious—but not necessarily the best—policy options here would be: (a) to grant successful leniency applicants immunity from paying damages in subsequent related private actions; or (b) to reduce the potential liability of such applicants, for example by excluding the award of multiple damages.58 Nonetheless, both of these approaches have their problems. Solution (b) clearly presupposes the introduction of multiple (or at least non-compensatory) damages, which, inter alia, brings into question the basic philosophy of compensation underlying damages remedies in the Member States.59 Solution (a), on the other hand, brings into question the principle underlying damages remedies in the Member States (and explicitly included in the civil codes of many Member States60 as well as the EC Treaty itself 61), according to which persons that have caused harm should in principle be held liable for it.62 Although these solutions cannot be totally excluded, serious consideration of them will require clear justifications. Moreover, implementing either of these solutions implies the resolution of a number of complex technical issues such as: how to reconcile conflicting leniency policies in different Member States; the issue of what constitutes a ‘related’ private action; how to deal with unsuccessful leniency applications; how to handle compensation of victims in case of insolvency of other cartel members; etc.

8. Contingency fees In many Member States, the rules of the Bar prohibit contingency fees. However, it is clear that in damage actions—more than in any other area of the law—some form of contingency fee or at least a ‘success’ fee may often prove to be a condition for plaintiffs to start even considering such actions. This is so particularly for ‘smaller’ victims of infringement. 58 The Commission could of course abstain from intervening in this area, in which case a balance would not need to be struck in the first place. 59 See above point III.A. 60 See, e.g., Article 1382 of the Belgian and French Civil Codes. 61 See Article 288 EC, which has been confirmed as a more generally applicable principle of EC law by the case law of the ECJ (see Joined Cases C-46/93 and C-48/93, Brasserie du Pêcheur and Factortame [1996] ECR I-1029; see also Crehan, supra note 14). 62 Totally excluding a person’s civil liability is clearly a different matter from holding two or more entities jointly and severally liable, since the plaintiff no longer has the discretionary choice of making claims against any of the potential defendants.


16/10/07

13:59

Page 445

The Commission’s Green Paper on Private Enforcement 445 Contingency fees are perceived positively by plaintiffs, as they show a ‘belief’ in the case by the lawyer involved, and his interest in the outcome is perceived as a guarantee that he will involve himself fully in seeking the success of the action. We therefore fail to see why contingency fees or success fees should be regarded as reprehensible, and tend to believe that rules preventing this form of remuneration are not only anticompetitive but are unjustified. In our opinion, the excesses of the US system cannot be attributed to the contingency fee system as such but to many other imbalances between the interests of plaintiffs and defendants.

E. Conclusions The EC finds itself at a critical juncture as regards its approach to private antitrust enforcement. Although the Commission has so far given little away about what it plans to do next, it is clear that in principle it favours more private damages actions. However, the justification for this is still unclear and, apart from general improvements in the speed and efficiency of Member States’ justice systems, any steps towards facilitating private damages actions could call into question fundamental aspects of the Member States’ rules on non-contractual liability. Any reform that follows on from the Commission’s Green Paper must therefore be carefully thought through if one wants to avoid—in the words of the Commission—an “Americanization” of EC competition law enforcement. This is particularly so at a time when the Americans themselves are reconsidering their own system, which has led to obvious abuses. Still, the main merit of the Green Paper may be that it has started a discussion and created a momentum that in itself has led Member States and their courts to reconsider carefully and on a case-by-case basis certain aspects of their compensation system that appeared unjustified—even with a mere ‘compensatory’ objective in mind.63

63 Legislation has been introduced in some Member States in that regard, most notoriously in Germany with the Seventh Amendment to the GWB (on issues such as the binding effect of decisions of other national authorities, collective actions for consumer groups, the passing on defence, the limitations period, etc.). In France, a decree of December 2005 reduced the number of courts allowed to decide on competition cases. In Sweden, ‘opt in’ class actions were introduced, etc. Furthermore, the number of successful actions has tended to increase in Spain, France, Germany, The Netherlands and the UK.


16/10/07

13:59

Page 446

(V) Ehlermann 07 DeSmijter&Woods

12/10/07

13:12

Page 447

II Eddy De Smijter and Donncadh Woods1 The Commission Green Paper On Damages Actions for Breach of the EC Antitrust Rules

A. The importance of private enforcement of the EC competition rules As early as the 1970s, the Court of Justice stated that Articles 81(1) and 82 EC “tend by their very nature to produce direct effects in relations between individuals (. . .) which the national courts must safeguard.”2 Despite this clear statement, national courts so far have been rarely invited to apply the EC competition rules. Part of the explanation lies in the fact that until the entry into force of Regulation 1/2003,3 national courts were not entitled to apply Article 81(3) EC. This implied that a party to a national proceeding could de facto block that proceeding by making a notification to the Commission. If it were not abundantly clear that Article 81(3) EC was not applicable, the national court would have to suspend proceedings.4 Being aware of this, complainants generally went directly to the Commission. The reflex of victims of competition law infringements to seek redress with a competition authority may seem a rational choice, it has some important drawbacks too. For the victim itself, the main drawback is that competition authorities only have limited tools to remedy the infringement.5 One of the main tools that is lacking, is the power to award damages to compensate for the loss suffered as a result of a competition law infringement. That power is left exclusively to national courts. Also from an enforcement perspective, the clear preference of antitrust infringement victims to address themselves to a competition authority has the drawback of limiting enforcement of the competition rules to what the competition authorities are able and willing to do. 1 European Commission, DG Competition. This paper is an elaborated version of the contribution on the Green Paper that was published in the EC Competition Policy Newsletter of spring 2006. All views expressed in this article are strictly personal and should not be construed as reflecting the opinion of the European Commission or of DG Competition. The authors can be reached at [email protected] and [email protected]. 2 Case 127/73 BRT v. SABAM (BRT I) [1974] ECR 51, para 16. 3 Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ L 1 [2004]. 4 Case C-234/89 Stergios Delimitis v. Henninger Bräu [1991] ECR I-935, para. 50. 5 See Article 5 of Regulation 1/2003 for the maximum list of tools at the disposal of national competition authorities and Articles 7 to 10, 23 and 24 for the tools the Commission can use.


448

12/10/07

13:12

Page 448


Complementing the public enforcement of competition rules by private enforcement actions before national courts would widen the enforcement scope of Articles 81 and 82 EC and thus create “a powerful additional incentive for companies to comply with the EC competition rules.”6 Facilitating private enforcement of the EC competition rules, in particular through actions for damages would thus serve a double purpose: it allows victims of antitrust infringements to get compensation for the loss suffered and it increases the incentive for companies to comply with the law. That double objective of antitrust damages actions has been explicitly recognised by the Court of Justice in its 2001 judgment in Courage v Crehan.7 The Court held that the full effectiveness of Article 81 EC and, in particular, the practical effect of the prohibition laid down in Article 81(1) EC would be put at risk if it were not open to any individual to claim damages for loss caused to him by a contract or by conduct liable to restrict or distort competition. The Court also held that the existence of such a right strengthens the working of the EC competition rules and discourages agreements or practices, which are liable to restrict or distort competition and that from that point of view, actions for damages before the national courts can make a significant contribution to the maintenance of effective competition in the Community.8

B. Facilitating private enforcement of the EC competition rules Shortly after Courage v Crehan, the Council made a first step in the direction of facilitating private enforcement of the EC competition rules by adopting Regulation 1/2003, which eliminates the notification system and the Commission’s exclusive power to apply Article 81(3) EC. One may wonder, though, whether Regulation 1/2003 is in itself sufficient to substantially increase the level of private antitrust enforcement. Indeed, Regulation 1/2003 largely leaves it to the Member States to determine the procedural framework within which the EC competition rules are to be enforced by national courts. It therefore requires an analysis of these national procedural rules to know whether they constitute structural impediments or incentives to private enforcement of the EC competition rules. Such an analysis not only has an empirical value, it also has a normative impact, since the Court reminded Member States that their national procedures may “not render practically 6 Monti M. (2004): “Competition for Consumers’ Benefit”, speech delivered at The Netherlands European Competition Day, Amsterdam, 22 October 2004, available at http://europa.eu.int/ comm/competition/speeches/ text/sp2004_016_en.pdf. 7 Case C-453/99 Courage Ltd v. Bernard Crehan [2001] ECR I-6297. 8 See ibid., at paras. 26 and 27.


12/10/07

13:12

Page 449

Green Paper On Damages Actions for Breach of EC Antitrust Rules

449

impossible or excessively difficult the exercise of rights conferred by Community law (principle of effectiveness)”.9 It is in this context that the Commission adopted on 19 December 2005 its Green Paper on damages actions for breach of the EC antitrust rules.10 The Commission considered it appropriate to adopt this Green Paper because there have been very few damages awards for breach of EC antitrust law so far.11 Many of the victims of antitrust infringements seem to refrain from bringing damages actions. Moreover, where they do bring damages claims, such actions often fail to be successful for a variety of reasons. The Green Paper identifies the main obstacles in current national procedural law to a more efficient system of antitrust damages claims and sets out, for further reflection and possible action, different options to remove or diminish these obstacles. When formulating these options, the Commission kept in mind that one should only facilitate actions by genuine victims of competition law infringements. Procedural rules that would incite unmeritorious claims are costly for society as a whole and for the economy in particular and should thus be avoided. The exercise of the Green Paper is thus a fine balancing act, but that complexity cannot be a reason for not raising the relevant issues and remedying the current situation when that would prove to be appropriate.12

C. The main issues of the Green Paper on antitrust damages actions 1. Access to evidence Actions for damages in antitrust cases regularly require the presentation of a broad and complex range of factual evidence. The particular difficulty of this 9

Ibid., para. 29. The Commission Green Paper on damages actions for breach of the EC antitrust rules (COM(2005)672) is available at http://europa.eu.int/comm/competition/antitrust/others/actions _for_damages/gp.html. The Green Paper is accompanied by a Commission Staff Working Paper (SEC(2005)1732), which is available at http://europa.eu.int/comm/competition/antitrust/others/ actions_for_damages/sp.html. The staff working paper gives background to and elaborates the options mentioned in the Green Paper. 11 See the study that was commissioned by the Commission and published in 2004, available at http://europa.eu.int/comm/competition/antitrust/others/actions_for_damages/study.html. 12 “I don’t think that simple fact that these potential risks are out there should stop us taking any action to make things better in Europe. To put our heads in the sand would not protect the rights of European consumers, and it would not help strengthen the European economy.” See Kroes N. (2006): “More Private Antitrust Enforcement Through Better Access to Damages: An Invitation for an Open Debate”, opening speech at the conference ‘Private enforcement in EC competition law: the Green Paper on damages actions’, Brussels, 9 March 2006, http://ec.europa. eu/comm/competition/antitrust/others/actions_for_damages/index_en.html. 10


450

12/10/07

13:12

Page 450


kind of litigation is that the relevant evidence is often not easily available to the injured party, for example because it is held by the party committing the anti-competitive behaviour or by third parties. Questions of access by victims to such evidence are key to making damage claims more effective. The Green Paper presents several options aimed at facilitating access to such evidence or alternatively alleviating the claimant’s burden of proving the infringement. With regard to access to the relevant evidence, it is important to draw a distinction between damages actions that are brought after a competition authority has found an infringement (follow-on actions) and actions that are brought independent from any public enforcement initiative (stand-alone actions). In case of a follow-on action, one may wonder whether it should be necessary for the claimant to prove the infringement once again. It flows from Article 16 of Regulation 1/2003 that a Commission decision finding an infringement of Article 81 or 82 EC constitutes a binding finding for the national court.13 In that scenario the claimant can limit himself to proving the damage he suffered and the causal link between the infringement and that damage. A number of Member States have extended the principle of Article 16 to final infringement decisions taken by a national competition authority, be it the domestic authority 14 or another European one.15 The Green Paper wonders whether it isn’t appropriate for the other Member States to take inspiration from such best practice. As far as the information in a Commission file is concerned, access can be granted to a national court, to a party to the Commission proceedings or to a third party. It goes without saying that the accessibility of information differs according to who is asking for access and at what stage of the Commission proceedings. Community procedures moreover provide for appropriate protection of confidential information and of leniency applications.16 Access to Commission documents being particularly limited for parties to the private litigation that are/were not a party to the related Commission proceedings, the Green Paper suggests creating a specific right for private litigants to obtain directly from the other parties in the private litigation a copy of the documents they supplied to the Commission in the course of the administrative procedure. In order to protect the leniency programme, the leniency applications should not be divulged, though. Issues relating to 13

See also case C-344/98 Masterfoods [2000] ECR I-11369, para. 52. See e.g. Sections 18 and 20 of the UK Enterprise Act 2002, inserted as sections 47A and 58A respectively into the UK Competition Act 1998. 15 See e.g. Section 33(4) of the German Competition Act. 16 For requests by national courts, see the Commission notice on the co-operation between the Commission and the courts of the EU Member States in the application of Articles 81 and 82 EC, OJ C 101 [2004], paras. 21–26. Requests by parties to the Commission proceedings are handled under the Commission notice on the rules for access to the Commission file in cases pursuant to Articles 81 and 82 of the EC Treaty, Articles 53, 54 and 57 of the EEA Agreement, and Council Regulation (EC) No 139/2004, OJ C 325 [2005]. Requests by third parties are covered by Regulation 1049/2001 of the European Parliament and of the Council of 30 May 2001 regarding public access to European Parliament, Council and Commission documents, OJ L 145 [2001]. 14


12/10/07

13:12

Page 451


451

disclosure of business secrets and other confidential information would have to be addressed by the national court on basis of its national law. Finally, in order to achieve the objective of private antitrust enforcement as a complement to, rather than an amplifier of public antitrust enforcement,17 one may have to consider procedural means in order to improve the difficult situation victims of an antitrust infringement find themselves in when they have to adduce the necessary evidence to convince the court. Indeed, in stand-alone actions the claimant usually faces an information asymmetry compared to the defendant. The Green Paper suggests rebalancing the relationship between claimant and defendant. That can be achieved either by allowing for some kind of disclosure of relevant documents held by the defendant or by a third party or by shifting the burden of proof to the defendant as soon as the claimant has made a prima facie case.

2. Damages and costs of action When deciding whether or not to bring a damages case, a potential claimant usually balances the likely benefits of an action with the costs of such an action and the risk of having to bear those costs. In most Member States, that cost/benefit balance currently tends to discourage large categories of potential claimants. That structural imbalance could be restored by lowering the costs of an action in court or lowering the risk of having to bear those costs and/or by increasing the benefits of an antitrust damages action. In view of the fact that Community law as well as the European Convention on Human Rights demand an effective access to courts for civil claims, the Green Paper considers how cost rules might facilitate such an access. Lowering the risk of having to bear the costs of litigation is a delicate exercise because ideally, the cost rules should be such that claimants with a strong damages case may be less hesitant to start an action, while frivolous claimants are still not encouraged. It is with that precise objective in mind that the Green Paper suggests that unsuccessful claimants would have to pay costs only if their claim was manifestly unreasonable. Alternatively, one may consider increasing the benefits of a damages action. The main policy question in that regard is whether or not damages should be limited to mere compensation. That sensitive question becomes already somewhat less sharp if one accepts that ‘mere compensation’ also includes compensation of lost profits and pre-judgment interest from the moment of the infringement onwards. However, the question remains 17 Compare with Case 26/62 van Gend & Loos [1963] ECR (English special edition) 1: “The vigilance of individuals concerned to protect their rights amounts to an effective supervision in addition to the supervision entrusted (. . .) to the diligence of the Commission and of the Member States.” (emphasis added)


452

12/10/07

13:12

Page 452


whether that is sufficient to bring injured parties to the court. One could further increase the incentive for claimants to bring an action by allowing for higher levels of interest or by introducing a multiplier. In that respect, the Green Paper suggests awarding double damages in the case of horizontal cartels, thereby stimulating the victims of these most serious competition law infringements to bring a damages claim. Finally, the quantification of damages appears to be a thorny issue. The Green Paper presents several economic models in order to provide for the calculation of damages in complex situations and wonders whether the accuracy of complex calculation models outweighs the intrinsic difficulty of their application.

3. The passing-on defence and indirect purchaser’s right to claim damages The ‘passing-on defence’ concerns the legal treatment of the fact that a buyer who purchases from a supplier engaged in anti-competitive behaviour may be in a position to mitigate his economic loss by passing on the overcharge to his own customers. The damage caused by anti-competitive behaviour may therefore be distributed down the supply chain or may even be suffered in its entirety by the ultimate purchaser, the final consumer. The Green Paper asks the question whether the infringer should be allowed to raise such a passingon as a defence against a damages claim brought by a direct purchaser. Community courts have not yet had the occasion to examine this matter in the area of antitrust. They did, however, address the issue in the area of illegal taxes and levies. It flows from that case law that rather than qualifying it as a ‘passing-on defence’, the Court accepts an ‘unjust enrichment defence’ which requires the defendant to prove that the claimant passed on the damage and did not suffer any type of income reduction, such as a reduction in sales. It seems arguable that the same reasoning could be applied in antitrust cases. Closely related to the question of the passing-on defence, is the issue of standing for the indirect purchaser and ultimately for the consumer, to whom the overcharge may or may not have been passed on. Although the Community courts have not yet taken a clear position on the standing of indirect purchasers in antitrust actions, one could read Courage v Crehan as preventing an absolute bar to actions being brought by indirect purchasers.18 The question is therefore not that much one of standing for indirect purchasers, but rather whether their damages actions should be facilitated or not. 18 Supra note 7, at para. 26: it should be “open to any individual to claim damages for loss caused to him by a contract or by conduct liable to restrict or distort competition” (emphasis added).


12/10/07

13:12

Page 453


453

One could argue that facilitating damages actions brought by indirect purchasers who suffered a loss is a fair policy decision because it allows, at least in theory, all those who have suffered a loss to be compensated. Others may, however, argue that facilitating damages actions brought by indirect purchasers—at the cost of actions brought by direct purchasers—may lead to a lower level of enforcement, because direct purchasers have higher chances in winning a damages case since they have more access to the relevant evidence. The Green Paper has not taken any stance on this issue and leaves all options open.

4. Defending consumer interests Because of their small and scattered claims, final consumers are not encouraged to take damages actions. Still, they are the ones who in the end probably suffer the most from an antitrust infringement. The Green Paper wonders whether the recent Commission proposal for a European Small Claims Procedure is sufficient for final consumers to bring an antitrust damages action.19 Alternatively, the Green Paper presents some options as to how their interests could be better protected by collective and representative actions. Beyond the specific protection of consumer interests, collective actions can serve to consolidate a large number of smaller claims into one action, thereby saving time and money. It is clear that the options presented in the Green Paper are far removed from a US style opt-out class action.

5. Coordination of public and private enforcement Some wonder why the Commission, being a public enforcement authority, wants to facilitate private antitrust enforcement. That reaction seems to start from the assumption that there is a contradiction between fostering both public and private enforcement of the EC competition rules. There is not necessarily such a contradiction, though. Since both public and private enforcement of the EC antitrust rules share the objective of increasing the respect of those rules, it should be possible to optimise the co-ordination between these two kinds of enforcement. This is especially true for the coordination between leniency applications and damages claims. The question is whether both can be stimulated at the same time. The starting point is that the effectiveness of the leniency programmes has to be preserved; not only in order to safeguard the fight against cartels by 19 Commission proposal for a Regulation of the European Parliament and of the Council establishing a European Small Claims Procedure, COM(2005)87 of 15 March 2005.


454

12/10/07

13:12

Page 454


public competition authorities, but also with the purpose of ensuring successful follow-on actions against the members of the cartel. The various options formulated in the Green Paper show that a well-devised antitrust damages system can not only be reconciled with the protection of a leniency programme, it may even stimulate leniency requests. According to one of those options the successful leniency applicant, on top of the reduction in fines, would get a reduction in damage liability. The latter should of course not be to the detriment of the victim, but rather to the detriment of the other cartelists, who remain jointly and severally liable for the whole damage. There is thus room to improve the private enforcement of the competition rules while maintaining and even strengthening the effectiveness of public enforcement.

6. Other issues The Green Paper addresses a few more technical issues which are considered necessary to guarantee that damage claims can be brought more effectively. We limit ourselves to three of those technical issues: the fault requirement, rules on jurisdiction and applicable law and the suspension of limitation periods. As a tortious action, damage claims in many of the Member States require fault to be proven. In some of these Member States fault is presumed if an action is illegal under competition law. In other Member States, however, such a presumption does not exist. The Green Paper therefore considers both a rebuttable and an irrebuttable presumption of fault where illegality is shown. The Green Paper also invites comments as to the introduction of a legitimate defence in case of an excusable error on the side of the defendant. The study referred to in footnote 11 has shown diversity amongst the procedural rules of the Member States. In order to reduce the forum shopping that may result from such diversity, one needs clear rules on jurisdiction and on applicable law. The rules on jurisdiction are laid down in Regulation 44/200l,20 whereas the rules on applicable law are the subject of the Commission proposal for a Regulation which is currently being discussed in the European Parliament and in Council.21 Limitation periods can impose significant restrictions on the recovery of damages. The Green Paper therefore asks for specific consideration of the relationship between limitation periods and procedures before competition authorities. It is suggested suspending the limitation period for damages 20 Council Regulation 44/2001 of 22 December 2000 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters, OJ L 12 [2001]. 21 Commission proposal for a Regulation of the European Parliament and the Council on the law applicable to non-contractual obligations (“Rome II”), COM(2006)83 of 21 February 2006.


12/10/07

13:12

Page 455


455

claims from the date proceedings are instituted by the Commission or any of the national competition authorities. Alternatively, the limitation period could start running after a court of last instance has decided on the issue of infringement.

D. Conclusion Facilitating actions for damages is a logical next step after Regulation 1/2003, which enhanced the involvement of the national competition authorities and national courts in the enforcement of the EC antitrust rules. In addition, by being able to effectively bring a damages claim, the individual citizen in Europe, be that a company or a consumer, is brought closer to the competition rules and will be more actively involved in the enforcement of these rules. Moreover, by increasing the level of enforcement of the EC competition rules, actions for damages contribute to the respect of those rules and thus to effective competition in Europe. They are thus important tools in creating and sustaining a competitive economy, a key element of the “Lisbon strategy”, which aims at making the economy of the European Union grow and create employment for Europe’s citizens. The Green Paper is meant to launch an ample and open reflection on how to improve the level of successful actions for damages caused by an infringement of the EC antitrust rules. It looks as if the Green Paper has indeed triggered the intended wide debate: numerous conferences have been organized on the topic, both the European Parliament and the European Economic and Social Committee are preparing own-initiative reports on the issues raised by the Green Paper and to date, the Commission has received over 140 comments to the public consultation. On the basis of all these reactions, the Commission will assess what actions, if any, are necessary to promote further facilitation of actions for damages of EC antitrust law.


12/10/07

13:12

Page 456

(W) Ehlermann 07 Lawrence

12/10/07

13:12

Page 457

III Jon Lawrence1 Seeking the Perfect Balance: Some Reflections on the Commission’s Green Paper on Damages Actions for Breach of the EC Antitrust Rules

A. Introduction It may well be that the national court procedural systems around Europe will in due course undergo change. The underlying assumption of the Commission’s Green Paper2 is that private damages actions in Europe do not act as an effective means of compensating victims of antitrust infringements, and that they do not serve as an effective enforcement mechanism in the antitrust arena and that the position should be improved. A variety of options for change have been put forward and comments were invited to be submitted to the Commission by 21 April 2006. Views on the appropriate way forward vary. Possible outcomes range from the establishment of a harmonized system (perhaps through an international treaty and/or regulatory measures at Commission level) to general encouragement and persuasion of Member States towards national judicial reform. In this paper, a further possible outcome is also considered, namely, that efforts be focused on the development of one or more ‘model’ jurisdictions with a view to encouraging others to emulate such a ‘model’ approach. This possibility is explored in more detail at pages 10 to 12. Whatever position is taken, a careful balance needs to be struck between encouraging claims to increase the effectiveness of the enforcement (and, perhaps, deterrent) mechanism, and creating a litigation culture that favours the bringing of unmeritorious claims with an adverse effect on the economy. To reach an appropriate solution by whatever method, careful thought needs to be given to how to achieve the appropriate balance—one that achieves the enforcement/deterrent objective, but avoids potential adverse 1 Partner at Freshfields Bruckhaus Deringer specializing in contentious antitrust cases. Thanks to Nicholas Gibson (associate at Freshfields Bruckhaus Deringer) and Margaret Bloom (former head of competition enforcement at the OFT and senior consultant at Freshfields Bruckhaus Deringer). The views expressed in this paper are entirely personal and do not necessarily reflect those of Freshfields Bruckhaus Deringer, or any of my colleagues at that firm. 2 European Commission’s Green Paper on damages actions for breach of the EC antitrust rules (the Green Paper), and the Commission Staff Working Paper (the Working Paper, together the Papers), 21 December 2005.


458

12/10/07

13:12

Page 458


consequences and excesses. This paper seeks to identify the most relevant elements to be addressed in designing any reformed system for the future and applies it to some of the options put forward in the Green Paper. To that end: (1) Section B explains in greater detail the perceived need for equilibrium in any reformed system; (2) Section C addresses the pairs of interests that are to be weighed against each other in seeking to produce a balanced system; and (3) Section D makes some suggestions as to particular options in the Green Paper which could perhaps be adopted in order to achieve the Commission’s objectives.

B. The need for equilibrium The objective of the Green Paper is to identify how to facilitate private damages actions. The assumption is that wherever private enforcement is in Europe today, it needs to be somewhere different tomorrow. Two questions present themselves in trying to work out the way forward. First, why—and how—is the current system not fulfilling the objectives desired? Secondly, how—and why—can we in Europe make it better? This paper seeks to explain that the answer to both questions lies in finding the right equilibrium.

1. Where are we now? The Report commissioned by the Commission to study the present state of private enforcement in Europe offers a partial—or, more precisely, the first part of a—response to the first question.3 It provides evidence of the paucity of private enforcement in Europe, but does not analyze how this situation has come to pass. The Report indicates that, by August 2004, there had been only around 60 judged cases for damages actions on the basis of either European or national competition law—only 12 of which were brought on the basis of EC law alone.4 Very few claims reach the courts, and claimants are resolving disputes before litigating. This amounts only to a description of the symptoms: it demands further diagnosis of the cause. 3 Ashurst, Study on the conditions of claims for damages in case of infringement of EC competition rules: Comparative Report, 31 August 2004 (hereafter “the Report”). 4 The Report, page 1. After the Report was published, its authors were informed that two cases were omitted (one German, one Dutch), according to Denis Waelbroeck, Partner, Ashurst, “Actions for Damages in Competition Law”, paper presented at the Conference of the Joint Working Party of the Bars and Law Societies of the United Kingdom on Competition Law, 7 April 2006.


12/10/07

13:12

Page 459

Reflections on the Commission’s Green Paper 459 The so-called “total underdevelopment” of private damages actions in the enlarged EU appears to be attributable in large part to an imbalance between claimants and defendants. The present system poses too many challenges to claimants to merit the effort and inherent risk of bringing a claim. The author’s own experience is of a multitude of claims settling at an early stage, either to avoid precedent or publicity, or to cement for the future a commercial relationship between supplier and customer. Claimants are reluctant to proceed to trial because of costs risks, lack of certainty as to their substantive rights, procedural hurdles and delays. The balance of the risk-reward system is perceived to be favourable to defendants—and strongly so. This picture of scant private enforcement stands in stark contrast to the United States, which boasts a century of private enforcement and a rich and complex jurisprudence on the subject as a corollary. Incentivized by treble damages and Alabama jury trials, spurred on by contingency fees fuelling a well-developed plaintiff bar, and with class action mechanisms and plaintiff friendly cost rules to reduce the risks, the claimant in the United States is ever eager to initiate proceedings. Instead, it is the defendants who often bail out of litigation at an early stage, with stupendous settlement awards5 providing windfall gains for claimants.

2. Where are we going? The Commission now seeks to strengthen the enforcement of antitrust law by facilitating damages claims for breach of that law. It wants to ensure that it is easier for consumers and firms who have suffered damages arising from an infringement of the antitrust rules to recover their losses from the infringer. At the same time, the Commission seeks to avoid the excesses of the US system. Neelie Kroes claims, in this respect, to have left her cowboy hat at home.6 The Commission’s Green Paper, therefore, put forward some 36 options, on which comments have now been submitted.7 Commenting on or approving of 5 For example, seven of the world’s largest vitamin makers, who agreed to pay US$1.17 billion to food and animal-feed companies to settle a class-action lawsuit in November 1999 (Warren S. and Wilke J. (1999): “Vitamin-Firm Accord to Pay $1.17 Billion”, Wall Street Journal, 4 November 1999; Barboza D. (1999): “$1.1 Billion To Settle Suit on Vitamins”, New York Times, 4 November 1999). 6 Kroes N. (2006): “More Private Antitrust Enforcement Through Better Access to Damages: An Invitation for an Open Debate”, speech delivered at the Conference on private enforcement in EC competition law: the Green Paper on damages actions (SPEECH/06/158), 9 March 2006, page 1: “As for importing ideas . . . I’ve left my cowboy hat at home. But as for the rest, I can only say that I am glad I stuck to my guns.” 7 Comments were due for submission by 21 April 2006. Comments received by the Commission will be posted on its website. Available from http://europa.eu.int/comm/competition/antitrust/ others/actions_for_damages/gp.html.


460

12/10/07

13:12

Page 460


those options is not a box-ticking exercise. The inter-relationship between the options (ranging from the introduction of double damages in cartel cases, contingency fee arrangements and class actions to clarification of the jurisdictional and governing law questions) must be fully understood before an informed selection of options can be made. The options interlink and it is only through the adoption of the correct combination that the appropriate balance will be struck. If the Commission, with Member States’ support and assistance, does not achieve the appropriate balance, Alabama jury style litigation could well hit European shores. Defendants would potentially be forced into early, and sometimes unwarranted, settlement. On the other hand, if it fails to achieve a refined facilitation of damages claims, claimants will not take the risks involved and will continue to refrain from pursuing private claims. Hence, private enforcement will not be effective as either a deterrent or to compensate those harmed by anticompetitive behaviour. If, then, the diagnosis of the present failings of European private enforcement is attributable to an imbalance in favour of defendants, it is vital to ensure that the prescription for the future does not tilt the balance too far in the other direction. The solution lies in finding an equilibrium such that both claimants and defendants are assured a fair and efficient system. The next section of this paper offers some reflections on how that balance might be achieved. It assesses the objectives and interests to be weighed against each other. It also makes some general observations as to how to calibrate the balance.

C. The interests to be weighed against each other The competing interests of claimants and defendants in private litigation are self-evident. Yet, these are not the only interests to be weighed against each other in achieving equilibrium in antitrust damages actions in Europe: there are various countervailing issues and interests at stake. Any lasting prescription for the future of European antitrust enforcement should, it is suggested, strike an appropriate balance between the following pairs of potentially competing priorities: —claimants and defendants; —compensation and deterrence; —public and private enforcement; —fairness and efficiency; —Community and Member State interests. The remainder of this section considers each of these pairs of interests in turn.


12/10/07

13:12

Page 461

Reflections on the Commission’s Green Paper 461

1. Claimants and defendants As identified in the introduction, there is a need for balance between claimants and defendants. In striving for the appropriate equilibrium, it is important not to make any assumption that all claimants are victims, and all defendants are guilty. Similarly, it should be appreciated that there are at least two distinct means of approaching the project of private enforcement reform: there are changes designed to alleviate the burden on the claimant in bringing an action; and, conversely, there are those that would increase the reward for the claimant in doing so. Both types of change would alter the risk-return analysis that a claimant undertakes in considering whether to bring proceedings. However, while the former address the problems inherent in the existing system, the latter could introduce new problems. Changes positively to incentivise litigation would not only provide a potential advantage to claimants, but, simultaneously, a real disadvantage to defendants. This could well tip the balance between claimant and defendant too far in favour of the former, unless carefully managed. It will be crucial for the Commission to be alert to this distinction as it attempts to navigate the fine line between a ‘competition culture’ and a ‘litigation culture’—a litigation culture which the Commission has expressly, and (it is submitted) rightly, stated that it wishes to avoid.8

2. Compensation and deterrence Compensation and deterrence are generally recognized as being the two primary goals of antitrust enforcement.9 Certainly, any system of enforcement should aim at both prevention (by deterring infringements) and cure (by fairly compensating the victims of any infringements that occur). Private enforcement—as part of the overall system of antitrust enforcement—can contribute towards the achievement of both of these goals. Depending on how the system is designed, there will not necessarily be any tension between compensation and deterrence in the context of private enforcement. The fact of improving the means by which victims are able to claim compensation for losses suffered may itself contribute to the overall deterrent effect. Private damages payable to victims can add to the public fines levied 8 See “European Commission Green Paper on Damages Actions for Breach of EC Treaty Anti-trust Rules—Frequently Asked Questions,” MEMO/05/489, 20 December 2005, p. 5. 9 See the Working Paper, paras. 4–6, for the European Commission’s own statement of its commitment to these goals. They also constitute the twin goals of antitrust enforcement in the United States (Illinois Brick v Illinois 431 U.S. 720 (1977) at [25]).


462

12/10/07

13:12

Page 462


by authorities, compounding the financial consequences for infringers of their illegal behaviour. To this extent, compensation has a ‘passive’ deterrent effect per se. However, it is open to question whether private damages should extend beyond a strictly compensatory measure: how appropriate is it to seek to add an ‘active’ deterrent effect in private damages actions? This question is the subject of long and ongoing debate among antitrust academics, practitioners, and the business community.10 Such ‘active’ deterrence may lead to unwarranted and undesirable side effects. Doubts also arise as to the effectiveness of private enforcement as a deterrent. Punitive damages arguably constitute a disproportionate, and therefore unjustified, interference with a defendant’s property rights.11 Even if that is not correct, there are certainly some potential unwanted side effects, such as the additional costs to business unrelated to losses caused. In light of these potential consequences, it is all the more important to be certain that ‘active’ deterrence of this nature through private damages actions really does achieve its aim. Yet this remains doubtful. Arguments for ‘active’ deterrence rely on assumptions as to the deterrent effect of financial penalties levied against the corporate defendant. This, in turn, relies on conflation of individual and corporate decision-making. Financial penalties undoubtedly affect a company’s bottom line. Fines and damages may, therefore, contribute to deterrence to a degree (and there are already fining regimes at European and national level). However, a law of diminishing returns perhaps applies in respect of their effectiveness in preventing infringements altogether. This is because they have only an indirect impact on those responsible for making decisions to enter into infringing conduct. By contrast, individual directors and managers are far more readily persuaded by penalties that impact on them personally: imprisonment; personal fines; disqualification from holding directorships. When a man knows he is to be held personally responsible, it concentrates his mind wonderfully.12 It is submitted that a greater deterrent effect will be achieved in this way, than through damages awards in litigation. On balance, therefore, it seems preferable to limit the deterrence goals of private enforcement to those inherent in the compensatory function. To seek ‘actively’ to deter anticompetitive behaviour through punitive awards in private damages actions again risks tilting the balance too far in favour of claimants. The specific example of double damages is considered in more detail below. 10 It is not proposed to revisit this debate in full in this paper. For one (polemical) position in the debate, refer to Shughart W. (1990) “Private Antitrust Enforcement: Compensation, Deterrence, or Extortion?”, 13 Regulation 53. 11 Article 1, Protocol 1 of the European Convention on Human Rights (1950). 12 To borrow from Samuel Johnson, who was, of course, speaking of being hanged: this may be a step too far, even in light of the present tendency towards criminalization of hard core cartel activity!


12/10/07

13:12

Page 463

Reflections on the Commission’s Green Paper 463

3. Private and public enforcement In designing a system of private enforcement in Europe, it will be important to view private damages actions in the wider context of antitrust enforcement as a whole. The Commission rightly observes that the two arms of antitrust enforcement serve complementary functions.13 An effective enforcement system must achieve an appropriate balance between the two. The move towards private enforcement should also be understood in the context of a wider change in the focus of antitrust enforcement in Europe. The modernisation initiative, ushered in by Regulation 1/2003/EC, has altered the balance of responsibility for enforcement of EC antitrust between the European Community and its constituent Member States—devolving power from the European competition authority to its national counterparts and allowing the Commission to refocus its resources as a result.14 It has also devolved power to national courts, with a view to fostering a climate for individuals to enforce their rights—a climate for private enforcement.15 Accordingly, it might be fair to say that the current debate is not only a recognition of how private enforcement can benefit individuals, but also how it can improve the position of competition authorities. However, it would be a mistake to view private enforcement as capable of substituting the role of public enforcement. Private parties—like national competition authorities—can supplement the enforcement effects of the Commission, but the two arms of antitrust enforcement are not completely fungible. The Commission has acknowledged as much in principle,16 and this should also be reflected in the role allocated in the future to private enforcement. The earlier discussion of the appropriate balance of compensation and deterrence alluded to some of the respective advantages and disadvantages of each of the private and public arms of enforcement policy. It is important to recognize the respective strengths of each, and to play to those strengths. A balance must, therefore, also be struck between these.17 13

Working Paper, paragraph 13. The UK Department of Trade and Industry (DTI) provides a useful summary of the background to and motivation for Regulation 1/2003/EC, noting that “being able to focus resources on serious infringements will allow the Commission . . . to enforce EC competition law more efficiently” (Regulatory Impact Assessment: Giving Effect to Council Regulation 1/2003 Available from: http://www.dti.gov.uk/files/file23036.pdf). 15 Recital 7 and Article 6 of Regulation 1/2003/EC, and Working Paper, paras. 23 to 27. 16 Working Paper, para. 13: “Private action should not and cannot replace public enforcement.” The Commission also deals with the issue of how to mediate between the two aspects of enforcement where they potentially conflict (e.g. the effects on leniency programmes in the Working Paper, Section VIII). However, this is only part of the picture. It is important also to recognize that each is inherently more or less suited to particular enforcement objectives. 17 On the subject of the relative merits of private and public enforcement and the appropriate balance between the two, please refer to the instructive debate between Wils W. P. J. (2003): 14


464

12/10/07

13:12

Page 464


Private enforcement has certain advantages over public enforcement. Most pertinently, private damages allow victims to make good their losses. The Commission, quite correctly, has pointed to the advantages to individuals of litigating a claim in the national courts. As Competition Commissioner Neelie Kroes has succinctly summed up the disadvantages to individuals of being a complainant instead of a litigant:18 “Public authorities are there to act in the general interest, and they cannot themselves repair the damage caused by violations of competition law.”

The courts, however, can do so. Of course, the differences between courts and public authorities in enforcement do not end with the ability to compensate victims for the losses they have suffered. The Commission has also noted that, among other advantages, national courts allow litigants the opportunity:19 —to recover their legal costs where they are successful; —to apply more easily for interim relief; and —to combine claims under European and national competition law—or even entirely separate heads of claim. And, equally importantly, national courts are obliged to hear the cases brought before them, while an administrative body has the discretion to pursue other priorities. Private enforcement, therefore, is arguably more responsive to individual victims than public enforcement. However, private enforcement, as noted above, is less effective in detecting and deterring hard core cartels.20 Whereas private enforcement can offer a more effective compensatory function, public enforcement is better suited to deterrence. Criminalization and the other sanctions against individuals are likely to be more effective in deterring anticompetitive behaviour in practice. The future system for private antitrust enforcement in Europe should therefore marry private compensation to public deterrence. Such a division of “Should Private Antitrust Enforcement Be Encouraged in Europe?”, 26 World Competition 473, and Jones C. (2004): “Private Antitrust Enforcement in Europe: A Policy Analysis and Reality Check”, 27 World Competition 13. 18 Kroes N. (2005): “Damages Actions for Breaches of EU Competition Rules: Realities and Potentials”, opening speech at the conference ‘La reparation du prejudice cause par une pratique anti-concurrentielle en France et à l’étranger: bilan et perspectives’, Cour de Cassation, 17 October 2005. 19 On the advantages of private enforcement, see: Monti M. (2004): “Private Litigation as a Key Complement to Public Enforcement of Competition Rules and the First Conclusions on the Implementation of the New Merger Regulation,” IBA—8th Annual Competition Conference SPEECH/04/403, 17 September 2004. 20 There is, of course, a tension between the interests of private enforcement and the operation of public enforcement through leniency programmes. The incentive to take advantage of a leniency programme and making admissions to the antitrust regulators may be reduced if the risks of substantial damages through private action is thereby increased (perhaps more than offsetting the benefits in any fine reduction). This is a tension that must be taken into account in assessing the extent to which private actions are to be encouraged.


12/10/07

13:12

Page 465

Reflections on the Commission’s Green Paper 465 complementary functions is most likely to strike the appropriate balance between the parties to private actions, and between the competing priorities outlined above. This marriage of private and public enforcement could take various forms in practice. The Commission invites consideration of how to achieve the optimum coordination between the two, suggesting a special regime regarding discoverability of leniency applications in order to guard against any potential conflict between leniency programmes and the encouragement of private enforcement actions, among other proposals.21 However, it would be possible to develop this relationship further as follows. As an additional condition for gaining leniency, applicants would be required to submit to mediation in respect of any actions brought by parties affected by the applicant’s anticompetitive behaviour. This mediation could be independent or managed or supervised by the regulator, but would be subject to a condition that the leniency applicant mediate with other parties in good faith. Any breach of this condition would carry the threat of revocation of leniency status, with the possibility of appeal to a judicial body, such as the UK’s Competition Appeal Tribunal (CAT). This would have the benefit of encouraging resolution of private antitrust actions by harnessing the success of public enforcers’ leniency programmes. This possibility is only mooted here by way of tentative suggestion; the detailed workings of such a system would require further consideration. However, in principle, it would facilitate resolution of private actions while maintaining an appropriate balance between the public and private arms of antitrust enforcement—it would, as suggested above, play to each of their strengths.

4. Fairness and efficiency Inevitably, a system of law must also strike a balance between fairness and efficiency. This pairing is a recurrent theme in law: the need for legal certainty, but without rigidity. It is, therefore, also important to recognize the limitations of antitrust enforcement in, for example, compensating victims. These hold true even in a system in which private actions have been introduced in order to serve the compensatory objective better. Antitrust enforcement does not—and probably cannot—compensate all the wrongs done by anticompetitive behaviour. First, it does not compensate the biggest victim—the economy as a whole, which suffers from the resulting misallocation of resources. Secondly, it (arguably) does not compensate the next biggest victim—those who do not 21

Options 28 to 30 of the Green Paper.


466

12/10/07

13:12

Page 466


buy at the anticompetitive price because it is above the relevant reservation price.22 However, recognition of the inherent limitations of the system should not be used to justify compromising the compensatory function that can be provided. The Commission appears to consider that efficiency might take precedence over fairness—“in the sense of full recovery for all those who have suffered a loss from an illegal practice”.23 Certainly, there is a need for balance in order to ensure that the system of private enforcement is workable in practice, but any limitation on the principle that every party should be entitled to defend its rights must be justifiable and proportionate. An appropriate balance between fairness and efficiency need not require sacrifices to pure policy. Arbitrary rules are no basis for a fair and efficient system. Instead, it is submitted that the application of the principle that every party should be entitled to claim for provable loss should allow standing for indirect purchasers, a passing-on defence, and rules to facilitate consumer actions. These points are revisited below.

5. Community and Member State interests Achieving a balanced system for private antitrust claims in Europe will require delicate handling of the respective interests of the Community as a whole (on the one hand) and individual Member State interests (on the other hand). There seems to be a common assumption that the Green Paper is intended to lead towards European harmonization of national procedural (as well as substantive) rules in relation to private antitrust damages claims. The unwritten assumption underlying the Green Paper appears to be that the same approach ought to be taken in all jurisdictions in relation to this type of claim. Leaving aside the question of whether antitrust claims should be dealt with in a different manner from other tort claims, it is questionable whether harmonization is achievable, or even desirable. It may not be achievable, because historic approaches to litigation will be difficult to displace. Approaches to the presentation of pleadings and witness evidence, the treatment of expert evidence and the extent of disclosure are procedural aspects that have historically been specifically aligned to national traditions. They are regarded as being part of the national preserve. Any defence of that national preserve in this area could, if faced with Community level legislation, point to the principle of subsidiarity to resist change. 22 Calvani T. (2004): “Competition Penalties and Damages in a Cartel Context: Criminalisation and the Case for Custodial Sentences”, Dublin Law Society, Dublin, 28 June 2004 (particularly pp. 15–23), available from http://www.tca.ie/speeches/tc_law_soc_ jun04.pdf. 23 Working Paper, para. 179.


12/10/07

13:12

Page 467

Reflections on the Commission’s Green Paper 467 Whether harmonization is necessary is also debatable. It may well be that, with appropriate encouragement from the Commission, a number of Member States would take the initiative to create a more balanced environment for the bringing of private antitrust damages claims. If these Member States were also able to establish jurisdictional rules, particularly within the existing framework of the Brussels Regulation,24 it might well be that the majority of claims in this area would ultimately be brought in the courts of those Member States. The approach of facilitating claims to be brought in follow-on damages claims arising from cartel decisions appeared to be given a boost by the English courts in Provimi.25 This case provides a possible example of how the existing system under the Brussels Regulation might assist the consolidation of claims in one (or several) key jurisdictions. —In Provimi, the claimants, all of which were direct purchasers of vitamins from the defendants, brought separate proceedings in the English courts against certain English, Swiss, German and French companies belonging to the defendant company groups. The claimants alleged breach of statutory duty for infringements of EC competition law arising from the defendants’ participation in cartels. —The defendants applied to strike out the claims on the grounds that, among other arguments, the English courts did not have jurisdiction to hear the claims against most (though not all) of the defendants, since certain of the defendants had jurisdiction clauses in their standard terms and conditions in favour of Swiss, German and French courts. —The court was, therefore, called upon to decide the proper application of the Brussels Regulation26 where one party alleged that a foreign jurisdiction clause applied. The relevant provisions of the Brussels Regulation were, therefore: —Article 5(3), which provides that a person domiciled in a Member State might be sued in another Member State in matters relating to tort in the courts of the place where the harmful event occurred; —Article 6(1), which provides that such a person may also be sued, where he or she is one of number of defendants, in the courts of the place where any one of them is domiciled, provided the claims were so closely connected that it was expedient to hear and determine them together to avoid the risk of irreconcilable judgments resulting from separate proceedings; and 24

Regulation 44/2001/EC. Provimi Ltd v Aventis Animal Nutrition SA [2003] 2 All ER (Comm) 683. 26 The case also concerned the Lugano Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968, which applied to Switzerland and contained broadly similar provisions to most of the relevant provisions of the Brussels Regulation. In this paper, only the Brussels Regulation is discussed. 25


468

12/10/07

13:12

Page 468


—Article 23(1), which provides that, where the parties had agreed that the courts of a Member State were to have jurisdiction to settle any disputes, those courts would have exclusive jurisdiction. —The court held that the proper approach was for the court to satisfy itself that factors existed which allowed it to take jurisdiction. On the facts, the defendants had to accept that, absent the jurisdiction clauses, the claimants would be entitled to invoke jurisdiction under one or both of Articles 5(3) and 6(1) of the Brussels Regulation. However, a burden remained on the claimants to satisfy the court that they had “much the better of the argument” that the court should take jurisdiction despite the jurisdiction clauses. In the event, the court accepted that the claimants had satisfied this test by showing that the scope of the jurisdiction clauses was not sufficiently wide to include the present claims, thereby depriving the defendants of the effect of Article 23(1).27 —The case illustrates how claimants might use Articles 5(3) and 6(1), among others, to select a suitably balanced jurisdiction and/or to consolidate claims against multiple defendants.28

27 This was at an interlocutory stage (i.e. before trial) and the appeal was settled without being heard. 28 Advocate General (AG) Léger’s Opinion of 8 December 2005, in Case C-539/03 Roche Nederland et al v Frederick Primus et al, [2006] ECR I-6535 considered the application of Article 6(1) and is therefore, also relevant to the present discussion, although the case did not relate to antitrust infringements. In Primus, the claimants (two US holders of a European patent) brought actions before a Dutch court alleging patent infringements by nine companies incorporated in the Netherlands and a number of other Member States. The claimants, therefore, sought to rely on Article 6(1) of the Brussels Regulation. The AG noted that Article 6(1) should apply where it is expedient to hear and determine cases together in order to avoid the risk of irreconcilable judgments resulting from separate proceedings (Case 189/87 Kalfelis [1988] ECR 5565, paragraph 12). This test of ‘irreconcilability’ is broadly similar to that found in Article 28 of the Brussels Regulation, which provides that, where related actions are brought in courts of different Member States, any court other than the court first seised may stay its proceedings pending the outcome of the case in the court first seised. In Tatry (Case C-406/92 [1994] ECR I-5439), in the context of Article 28’s identical predecessor, the ECJ interpreted the test of ‘irreconcilability’ broadly to cover all cases where there is a risk of conflicting decisions, even if the judgments can be separately enforced and their legal consequences are not mutually exclusive. In Primus, the AG had to decide, therefore, whether this broad interpretation of ‘irreconcilability’ under Article 28 should also apply in the context of Article 6(1). The AG found that it should not, and that a narrower test should apply, since the two Articles had the following material differences. First, they have very different relationships with Article 2—the default determinant of jurisdiction under the Brussels Regulation (by domicile of the defendant). Whereas Article 6(1) systematically deprives defendants of their right to courts of their domicile, Article 28 is discretionary. Secondly, Article 28 provides a discretion to a judge, who can be expected to decide in light of good administration of justice, whereas Article 6(1) may be invoked at the request of a claimant in his or her own best interests. If the ECJ follows its AG, this case may therefore narrow the application of Article 6(1) and, with it, the ability of claimants to consolidate claims brought against multiple defendants into one ‘model’ jurisdiction.


12/10/07

13:12

Page 469

Reflections on the Commission’s Green Paper 469 Such a process, of course, would involve forum shopping. Yet this may be viewed as ‘positive’ forum shopping—leading by example—and is therefore to be encouraged. Conversely, the Commission must strive to avoid the development of a single jurisdiction which is unduly claimant-friendly. The experience of the United States is, as ever, instructive. The reputation of certain state jurisdictions precedes them, particularly the example of ‘Alabama jury trials’. Such a reputation has invited ‘negative’ forum shopping—led by greed—and typifies the excesses of the US litigation system. Clearly, such an outcome would be pernicious and run counter to the balanced solution that must be sought in creating the future system of European private antitrust enforcement. The key, then, as suggested in the introduction to this paper, may be to produce one or more ‘model’ jurisdictions that embody the equilibrium discussed in this paper. A system predicated on this principle would have the further benefit of avoiding lengthy and potentially intractable debates regarding the peculiarities of procedure and litigation culture in all 25 Member States and the inevitable barriers to full procedural harmonization. Indeed, it also seems questionable whether the Commission even has competence to legislate in this field.29 The Green Paper, rather than a blueprint for a European harmonizing Regulation, may be better characterized as a manifesto—seeking to create consensus—or the first step in an exercise in persuasion.

D. Ensuring equilibrium Creating the necessary consensus will be no mean feat. Any resulting reforms must combine and respect all of the various competing priorities listed above. They will also require careful consideration as to: how best to calibrate the balance; how particular options interrelate with each other; and how the options might be best applied in order to achieve the necessary equilibrium.

29 Article 65 EC offers only a very limited legislative basis for European intervention in national judicial proceedings, providing for the elimination of obstacles to the good functioning of civil proceedings by promoting compatibility of the rules on civil procedure applicable in the Member States. ICC Comments on the Commission’s Green Paper on damages actions for breach of the EC antitrust rules (April 2006), p. 3, also note “issues to be resolved concerning subsidiarity and the division of authority between the Community and member states.” Nonetheless, at least half of the national competition authorities apparently believe it would be difficult to provide a better framework for private action without some instrument at EU level (see Philip Lowe, Director General of DG Competition, European Commission, “Actions for Damages in Competition Law”, at the Conference of the Joint Working Party of the Bars and Law Societies of the United Kingdom on Competition Law, 7 April 2006).


470

12/10/07

13:12

Page 470


Some general observations follow regarding the calibration of the balance, addressing selectively some of the options put forward in the Green Paper.

1. Identifying the existing imbalance Before attempting to recalibrate the existing balance, the Commission should first assess what imbalance presently exists. To return to the introductory theme, therefore, an accurate diagnosis is important, if the prescription for the future of antitrust enforcement in Europe is to remedy any problems, and avoid any unwanted side effects. In this regard, it is noteworthy that European-level changes with a view to facilitating private enforcement have only been implemented relatively recently.30 It would be premature to suppose that the full effect of these changes has yet been felt. Any equilibrium in the system has probably yet to re-stabilize, and the findings in the Report may yet alter as a result. Any prescription should, therefore, factor in the eventual effects of these changes at the European level. Any plans for the future should also be tailored to fit the particular problem: a ‘one size fits all’ approach would be unattractive. This concern arises particularly in connection with the notion that the answer to the present failings would be to create a special system—a lex specialis—for antitrust. It is certainly true that complexities arise in the antitrust context that demand special consideration, but it is questionable whether it is appropriate to suppose that these complexities are uniform or unique. That all antitrust cases can or should be treated as a separate class of action benefiting from special rules of procedure seems odd.31 Antitrust actions can take a number of forms—not all are equally amenable to individualised rules for antitrust proceedings. Follow-on actions, for example, more naturally lend themselves to separate procedural mechanisms,32 whereas other cases may involve a mix of antitrust and commercial issues.33 Practical difficulties are likely to arise if a separate system for

30 Regulation 1/2003/EC and the Working Paper, paras. 23 to 27. However, some Member States have made changes to facilitate private enforcement, e.g. the UK through sections 47A, 47B and 58A of the Competition Act 98; Germany through s33(4) of revised German ARC, which says courts are bound by decisions of Commission, the BKA and other NCAs—see http://www.bundeskartellamt.de/wEnglisch/download/pdf/GWB_7_e.pdf. Further such changes can be expected. 31 In this regard, and possibly for the reasons set out below in the text, Ministries of Justice and Home Affairs of Member States are apparently reluctant to see a lex specialis adopted with regard to procedures at national level (Philip Lowe, supra note 39). 32 CAT damages actions and specialized rules of procedure. 33 ICC Comments on the Commission’s Green Paper on damages actions for breach of the EC antitrust rules (April 2006), p. 3.


12/10/07

13:12

Page 471

Reflections on the Commission’s Green Paper 471 antitrust issues necessitated parallel proceedings with different procedural rules and timetables. Instead, a better approach may be to introduce a specialized approach only where necessary in light of particular problems arising in the context of antitrust enforcement—looking at the problem narrowly, as well as broadly. Particular improvements may only be appropriate for certain aspects of private enforcement. Such an approach could be said to adhere more closely to the principle of proportionality, and be consistent with the balanced, more nuanced solutions achieved by the ECJ through the principles of effectiveness and equivalence. In this regard, it may assist to highlight the following distinctions within antitrust, rather than between antitrust as a whole and other areas of law: —follow-on versus stand-alone actions; —hard core cartel cases versus less clear-cut areas of substantive law; and —actions brought by consumers versus claims by intermediate indirect purchasers or competitors. A nuanced approach may well better ensure that the balance is properly calibrated and, in particular, responsive to the needs of claimants and defendants alike. It may also be that the answers to the problems inherent in the present system lie outside the realm of national court procedures. The discussion above has concentrated on private enforcement mechanisms and their capacity to offer a compensatory function. However, it could be feasible to introduce compensatory powers for public enforcement authorities. This could serve as an alternative or supplementary route to achieving an appropriate balance between the compensation and deterrence goals of enforcement, by conceiving a different equilibrium between the private and public enforcement functions. Part of the answer may also lie not in national judicial procedure, but national judicial training. The complexities in the antitrust context arise at least in part as a result of the (European) substantive law, together with the underlying economic theory or analysis. In light of the inherent difficulties in creating special rules for antitrust cases, an alternative solution may be to ensure that national judges have an even better understanding of competition law and the economics underpinning it. Training to improve such expertise among the judiciary could, where it has not already happened, be introduced. Specialist judges may also offer a more flexible solution than specialist courts or special rules. Indeed, the lack of sufficient judicial familiarity with competition issues in some jurisdictions may currently dissuade prospective litigants from bringing a case because of uncertainty as to its outcome.


472

12/10/07

13:12

Page 472


2. Establishing clearly where the balance should lie The success of any programme of reforms may well depend not only on ensuring that there is equilibrium, but also that those involved—claimants and defendants—know that there is, and how it operates in practice. It seems possible that lack of certainty is at least part of the reason why so few parties pursue proceedings in Europe at present. The US Department of Justice also notes the value of greater clarity as to both substantive and procedural rules, which “lowers the costs to business of complying with the law by reducing uncertainty about the parameters of legal behaviour”.34 Clarity is needed in respect of, for example: quantification of damages; availability of punitive damages; availability of the passing-on defence; availability of standing for indirect purchasers. However, clarity is also lacking at present in respect of the substantive principles of competition law. The ongoing debate regarding the role and remit of Article 82 is a good example of the continuing challenges in this regard.35 This is not a question of national procedure: it is important not to confuse or conflate national procedural and European substantive impediments to the bringing of antitrust claims.

3. Calibrating the balance—weighing the options up The selection of options identified in the Green Paper is not merely a box-ticking exercise. The system of private enforcement is the sum of its parts. So, all of the elements of any programme of reform must be considered together. In choosing the appropriate options, therefore, the Commission should have regard to the aggregate effect it is seeking to achieve, and how each of the options selected will contribute to achieving the relevant balance. The appropriate balance may be achieved by more than one arrangement of the options. This is why it seems preferable to focus on the end balance to be achieved. This approach also has the merit of accommodating the “astonishing diversity” within European litigation (as described in the Report) with the desire to foster further private enforcement expressed in the Green Paper. With these considerations in mind, the remainder of this paper considers certain of the options put forward in the Green Paper, and how they may contribute to achieving a desirable equilibrium in any new system. These options have been selected on the subjective basis of what appears to represent the key 34 US Department of Justice, Antitrust Division, available from http://www.usdoj.gov/atr/ overview.html. 35 The European Commission’s parallel consultation process in respect of the future of Article 82 EC, available from http://europe.eu.int/comm/competition/antitrust/others/article_82_review.html.


12/10/07

13:12

Page 473

Reflections on the Commission’s Green Paper 473 changes required—particularly those that are likely to have a substantial effect for the business community in litigating private damages actions. No attempt is made to review the full range of options set out in the Green Paper.

3.1. Changes to the rules relating to cost exposure The Green Paper, at option 27, notes the disincentive effect of the ‘loser pays’ principle, and, therefore, suggests insulating the claimant from exposure to the defendant’s costs by means of a ‘cost protection order’. In jurisdictions allowing significant cost recovery, it is true that the ‘loser pays’ principle can operate as a significant disincentive to bringing a claim because of the exposure to additional costs it entails. The extent to which this principle operates as a disincentive will depend on a number of factors, including: court and lawyers’ fees; the complexity and length of the case; and the financial strength of the paying party. Some change to the existing rules, particularly those where costs can be considerable, would seem to be a high priority in seeking to foster private antitrust enforcement and remove a substantial disincentive to claimants. Yet a balanced approach remains important. Introducing a uniform rule to relieve the claimant from any obligation to pay the defendant’s costs would potentially tip the scales too far in favour of claimants, beyond what is strictly necessary to encourage private enforcement, and towards the ‘litigation culture’ that the Commission seeks to avoid. One need only consider the example of an American-style plaintiff bar and possible vexatious claims being settled as a result not of the merits of the claim, but of the costs risk in continuing the action as defendants must pay their costs whether they win or lose (plus plaintiffs’ costs when they lose). This would not result in the required equilibrium. Instead, a balanced solution, which addresses and redresses the specific mischief in the present system, is necessary. One means of doing so might be by replacing the ‘loser pays’ rule with a presumption that, unless the defendant makes a reasonable offer to settle which is refused, the successful claimant is able to recover costs. This redresses the present imbalance by offering the claimant the possibility of avoiding higher cost exposure; but retains balance for both parties by penalising claimants where they bring unmeritorious claims. The English Civil Procedure Rules (CPR) contain a possible model that might serve as a starting point for such a procedure, stipulating detailed rules relating to offers to settle and payments into court36 and the provision of security for costs.37

36 37

Civil Procedure Rules, Part 36. Civil Procedure Rules, Part 25.


474

12/10/07

13:12

Page 474


3.2. Clear, certain, and balanced quantification of loss The Commission sets out a range of options relating to the definition and calculation of damages, inviting comments on the suitability of: a compensatory measure of loss;38 recovery of illegal gain;39 and double damages for hardcore cartel activity.40 As noted above, private enforcement offers the opportunity to claimants of recovering their losses. It is submitted that a valid approach is to regard the compensatory (rather than deterrent) objective as lying at the heart of private enforcement, and that it should primarily dictate choice of options relating to the quantification of loss. On this view, option 14—a compensatory measure of damages—seems preferable. Double damages, as suggested under Option 16, would be inappropriate, disproportionate and unnecessary—risking a skew of the scales of justice in favour of one party and away from a balanced approach to enforcement. The ongoing and fierce debate in the United States includes scathing criticism of the undesirable side effects treble-damages have had in that jurisdiction.41 Here in the EU also, recent judicial comment expresses doubt as to the necessity of such a remedy.42 In addition, double damages would punish the shareholders at the time of litigation, instead of the wrongdoer at the time of the infringement and potentially work against leniency programmes. The need for greater clarity and certainty regarding the applicable substantive and procedural regime has been emphasised above: and this is particularly pertinent in the context of the quantification of damages. Presently, there is a high degree of uncertainty in this regard. The following examples from the English courts suffice by way of illustration of the current difficulty facing claimant and defendant alike in assessing the value of any claim. It remains unclear: —whether claims could include punitive or exemplary damages; —whether actions may be brought on bases that are not purely compensatory (such as for recovery of illegal gain);

38

Option 14 of the Green Paper. Option 15 of the Green Paper. 40 Option 16 of the Green Paper. 41 For a brief summary of both sides of the debate, see, for example: Comments of the Section of Antitrust Law and the Section of International Law of the American Bar Association in Response to the Commission of the European Communities’ Request for Public Comment on Damages Actions for Breaches of EU Antitrust Rules (April 2006), pages 26 to 30. 42 Opinion of Advocate General Geelhoed in C-295/04 Manfredi, [2006] ECR I-6619: as a matter of Community law, punitive damages are not necessary to give the EU competition rules proper effect. However, the ECJ took a slightly different position, accepting the notion of “specific damages” if appropriate to ensure equivalence between remedies granted for claims based on Community law and those based on national law, subject to the prerogative of national courts to prevent unjust enrichment of claimants. See Paras. 92–94 and 99 of the Court’s judgment. 39


12/10/07

13:12

Page 475

Reflections on the Commission’s Green Paper 475 —whether defendants may rely on evidence that the claimant ‘passed-on’ any overcharge to those further down the supply chain; and —whether any subsequent (indirect) purchasers have any standing to bring a claim against the defendant. Such a situation is far from satisfactory, operating as a disincentive to bring or continue claims, and creating a perception of arbitrary settlement of cases: this benefits neither claimants nor defendants. The proposal at Option 19— for the Commission to publish guidelines on the quantification of damages— would seem to be one means of addressing the present lack of certainty. The persuasive/binding nature of any such guidelines would, however, remain an open question—the solution is not, therefore, ideal. Finally, in order to ensure that the cost of making a claim does not outweigh the advantages to end consumers of doing so, some procedural mechanisms to facilitate consumer claims appear necessary. The Green Paper includes proposals for a cause of action for consumer associations and collective actions by groups of purchasers.43 On the basis of the author’s own experience, the English system may again recommend itself in both respects.44

3.3. Balanced rules applied evenly The Green Paper invites debate as to whether a passing-on defence should be allowed, and whether both direct and indirect purchasers should have standing to sue.45 That every party should be entitled to claim for provable loss (but only for provable loss) has been advocated above in connection with the need to achieve balance between fairness and efficiency. Application of this rule would resolve some of the uncertainty in the area of quantification of claims and militates in favour of a passing-on defence, as well as standing for indirect purchasers. This approach follows the proposal at Option 21 of the Green Paper. Such a system could not only achieve balance, but also make the resulting benefits available to all those who have suffered loss including end consumers, for whom special rules may be necessary in order to counter any risk of ‘atomization’ of their relatively small claims. The necessity of pursuing a balanced approach dictates in favour of such a system, although it is recognised that an efficient means of overcoming precise quantification of losses down the supply chain (and perhaps based on broad legal presumptions) may also be required.

43

Options 25 and 26 of the Green Paper respectively. Section 47B, Competition Act 1998, which allows for claims to be brought on behalf of consumers by specified bodies stipulated by the Secretary of State, or Group Litigation Orders under Civil Procedure Rules, Part 19. 45 Options 21 to 23 of the Green Paper. 44


476

12/10/07

13:12

Page 476


To achieve this, it seems that, while a passing-on defence should be available, the burden for proving it should lie with the defendant. Balance requires, however, that, as a corollary, the defendant be equipped with the means by which to prove pass-through. To this end, the courts could impose mandatory or discretionary disclosure obligations on the claimant to enable the defendant to understand the claimant’s cost structure and pricing policies. Secondly, in seeking to achieve balance, sophisticated principles and rebuttable presumptions for determining causation and loss are likely to be preferable to govern the determination of who is the most appropriate claimant than an inflexible rule of standing.46 These principles allow for the court to make an assessment in light of the specific facts at issue—a solution that (it is submitted) strikes a balance between efficiency and fairness (and claimant and defendant) and eschews unwarranted rigidity and the arbitrary results this may entail.

E. Concluding remarks Balanced enforcement is better—and fairer—enforcement. The primary deficiency in the system of private enforcement in Europe today is a lack of balance. Having diagnosed this as the problem, it is vital not to overcompensate—creating an imbalance in prescribing a cure. The analogy to a level playing field is one familiar to all antitrust practitioners—and this is what a successful programme for reform of private antitrust enforcement must seek to achieve. Those planning that programme must have one eye on the imbalance inherent in the past and present system, and the other eye to creating an appropriate equilibrium for the future. A successful ‘model’ jurisdiction adhering to these principles may provide the necessary impetus for these wider reforms in European private antitrust enforcement. The reflections set out in this paper can, hopefully, make a small contribution towards achieving this ambitious aim.

46 In this regard, there is no need to provide harmonized rules regarding causation: the difference in national causation principles does not preclude the possibility that each, by different means, assures an adequate degree of balance between the parties to litigation.

(X) Ehlermann 07 Siragusa

12/10/07

13:12

Page 477

IV Mario Siragusa and Marco D’Ostuni 1 A Reflection on Some Private Antitrust Enforcement Issues

A. Introduction The European Commission’s Green Paper on Damages actions for breach of the EC antitrust rules (and the annexed Commission Staff Working Paper) promoted an intense debate on national procedural rules which might unduly hinder the development of private antitrust enforcement.2 The Green Paper draws significantly from the so-called Ashurst report, i.e., a study requested by the European Commission comparing the status of private antitrust enforcement in all of the Member States.3 According to the Court of Justice, enhanced private antitrust enforcement is in itself desirable, as it helps to ensure the full effectiveness of EC competition rules and to preserve effective competition (and competitiveness) in the European Union.4 Therefore, the Green Paper suggests options aimed at fostering the development of private antitrust enforcement and requests comments in order to identify the possible areas of legislative or regulatory intervention. Currently, it is up to the Member States to ensure the protection of private rights based on EC law, subject to the principles of equivalence and effectiveness. Since the Crehan judgment started a comprehensive debate on how to foster private antitrust enforcement, many Member States have adopted legislation to that effect. It is still unclear, therefore, whether the issues addressed in the Green Paper should be regulated at the EU or Member State level. In her recent opening speech at a conference called Private Enforcement in EC Competition Law: The Green Paper on Damages Actions, Competition Commissioner Kroes stated: “I can assure you I know my basics—the fundamental principles of subsidiarity and proportionality—and will apply them. I will only recommend European actions if they add demonstrable value and 1

Cleary Gottlieb Steen & Hamilton LLP. Green Paper Damages Actions for Breach of the EC Antitrust Rules, Commission of the European Communities, COM (2005) 672 final. Commission Staff Working Paper, Annex to the Green Paper Damages actions for breach of the EC antitrust rules, Brussels, 19 December 2005, SEC(2005) 1732. 3 Waelbroeck D., Slater D. and Even-Shoshan G. (2004): Study on the conditions of claims for damages in case of infringement of EC competition rules, 31 August 2004, text available at http://ec.europa.eu/comm/competition/antitrust/others/actions_for_damages/comparative_ report_clean_en.pdf#search=%22waelbroeck%20damages%20conditions%22. 4 Case C-453/99 Courage Ltd v. Crehan [2001] ECR I-6297, paras 26–27. 2


478

12/10/07

13:12

Page 478


deliver something which Member States alone cannot. We will look carefully at all the possible options and their impact.”5 This short paper aims at discussing whether the fault requirement and current national rules on discovery and the value of administrative decisions in private liability actions should really be regarded as obstacles to the development of private antitrust litigation and, if so, whether the possible solutions suggested by the Green Paper would really achieve their goal.

B. The fault requirement In most Member States (i.e., 20 out of 25, according to the Green Paper), successful damages claims usually require proof of fault (intentional or negligent), according to established principles of tort or civil liability. For purposes of the Green Paper, negligence is understood as a company’s failure to exercise the requisite degree of care (a definition that encompasses the concept of negligence in many Member States, allowance made for differences in terminology). The Green Paper asks whether it is really necessary to require fault in damages claims based on infringements of competition law. Three policy options are suggested: (i)

introducing a general strict liability standard (possibly in the form of an irrebuttable presumption of fault), whereby proof of an infringement would automatically imply fault; (ii) introducing a strict liability standard only with respect to hard core violations (i.e., cartels or egregious abuses); and (iii) introducing a general presumption of fault, which the defendant could rebut by proving excusable errors of law or fact (in this respect, the requisite degree of care could be set at an appropriate level, e.g., that of a prudent market operator as opposed to a prudent layman). The case against maintaining the fault requirement is set forth in the Staff Working Paper,6 according to which the fault requirement could be an obstacle to private antitrust enforcement in those seven Member States where the claimant actually needs to prove fault, because it does not benefit from any 5 Neelie Kroes, “More private antitrust enforcement through better access to damages: an invitation for an open debate”, Speech/06/158, 9 March 2006, at p. 2. 6 See Commission Staff Working Paper, Section III. Some of the arguments discussed in the Working Paper with respect to fault were anticipated by Norberg S. (2005): “Some Elements to Enhance Damages Actions for Breach of the Competition Rules in Articles 81 and 82 EC”, paper presented at the 32nd Annual International Antitrust Law & Policy Conference, Fordham, New York, 14 September 2005, p. 7. Text available at http://ec.europa.eu/comm/competition/ speeches/text/sp2005_019_en.pdf.


12/10/07

13:12

Page 479

A Reflection on Some Private Antitrust Enforcement Issues

479

legal presumption.7 First of all, under EC competition law, any conduct having the effect of restricting competition is prohibited, regardless of fault. Intention or negligence is required only regarding the imposition of fines. Second, Member States’ liability for breach of EC law cannot be made conditional on “any concept of fault going beyond that of a sufficiently serious breach of Community law”.8 Third, competition law infringements are serious breaches of law. Fourth, under Council Directive 85/374/EEC of 25 July 1985 (the Product Liability Directive), liability without fault was perceived as the “sole means” of adequately solving the problem of “a fair apportionment of the risks inherent in modern technological production”, also due to information asymmetry between manufacturers and consumers. As noted, the principles of subsidiarity and proportionality require an analysis as to whether the fault requirement really is an obstacle to private damages actions (so that action at EU level would add demonstrable value). Indeed, any additional prong of a legal test that needs to be satisfied by the plaintiff could be perceived as an additional obstacle to successful legal actions. However, there are several reasons why the fault requirement can hardly be considered as a significant impediment to private antitrust enforcement: —All restrictive agreements or concerted practices under Article 81 EC imply an intentional element, and many abuse cases are based on evidence of an abusive strategy by the dominant company. If a claimant is able to prove the existence of any such infringement, proving fault would not be an obstacle; —Cartels and the most egregious abuses normally imply fault, because the defendants cannot be excusably unaware of the restrictive nature of their conduct; —Once a claimant proves an infringement, judges often presume fault, unless the defendant is able to prove otherwise or unless national legislation is so crafted as to deny judges any discretion in that regard. According to the Ashurst Report, “where a hard core violation is shown fault will often not be contested at all—a view which is reflected in existing case law”;9 —Apart from leniency programmes, the Commission and national competition authorities recognize immunity from fines under very limited circumstances. 7 The Member States in question are Denmark, Greece, Spain, Poland, Portugal, Finland and Sweden. Italy is not listed among them, but it should be: under Italian law, the claimant actually needs to prove fault in antitrust damages actions (and does not benefit from any legal presumption). The Commission’s position might reflect the reference made in the Ashurst Report to the opinion of some Italian authors, according to whom the rebuttable fault presumption set forth under the unfair competition rules might apply by analogy to antitrust damages actions. However, it has been clarified, since the publication of the report, that unfair competition law cannot apply to antitrust cases (see, e.g., Italian Corte di Cassazione, judgment No. 2207/2005). 8 Joined Cases C-46/93 and C-48/93 Brasserie du Pêcheur v. Bundesrepublik Deutschland and The Queen v. secretary of state for Transport, ex parte: Factortame and Others [1996] ECR I-1029. 9 Ashurst Report, supra note n. 3, at p. 3.


480

12/10/07

13:12

Page 480


There is no reason to think that the situation would be any different with respect to civil liability. Excusable errors of fact or law would normally be recognized in very limited scenarios, such as in cases involving the state action defence, legitimate expectations, non clear-cut or new infringements, or—at least, theoretically—cases where the defendant can prove that it adopted all of the precautions required under the applicable standard of care (e.g., implementation of frequent antitrust compliance programmes, active surveillance, creation of efficient supervisory structures, consistently compliant behaviour, reliance on legal advice and the like); and —As recognized by the Green Paper, in the vast majority of Member States the fault requirement is not an obstacle because either there is no such requirement, or fault is already rebuttably or irrebuttably presumed once the infringement is proven. That makes it hard to believe that the perceived underdevelopment of private antitrust enforcement in Europe might be due to any significant extent to the fault requirement. As to the legal arguments supporting the Commission’s options, not all of them seem entirely convincing. In Courage, the ECJ made a clear distinction between the principles governing the application of the EC competition rules and the consequences in civil law of a breach of such provisions.10 In fact, by recognizing that even a party to an anticompetitive agreement may be entitled to antitrust damages, this judgment clearly stands for the proposition that civil or tort liability need not mirror the application of the EC competition rules. From this perspective, we see no obvious reason why public fines should remain conditional on fault, while tort or civil liability consequences should not. In cases of cumulative foreclosure effects caused by networks of agreements, for instance, it would be unfair to hold all parties to the agreements equally liable to their competitors, regardless of their actual bargaining power and contractual position vis-à-vis the other party. Furthermore, the comparison with Member State liability does not seem entirely appropriate. In Brasserie du Pêcheur, the ECJ confirmed that “certain objective and subjective factors connected with the concept of fault under a national legal system may well be relevant for the purpose of determining whether or not a given breach of Community law is serious”; thus, “the obligation to make reparation for loss or damage caused to individuals cannot [. . .] depend upon a condition based on any concept of fault going beyond that of a sufficiently serious breach of Community law”.11 We read this to mean that the concept of “serious breach” may include some elements of fault, but is substantially a much broader and different concept. That renders it difficult to make meaningful comparisons. Furthermore, this passage of the judgment can be explained by the fact that Member States have both the 10 11

Courage v. Crehan, supra note n. 4, at para. 45. Brasserie du Pêcheur, supra note n. 8, at paras. 78–79.


12/10/07

13:12

Page 481


481

power and incentive to limit their own liability under EC law, by imposing additional conditions for private actions. Therefore, they should not be allowed to do so, particularly because any relevant aspects of fault are already taken into consideration within the sufficiently serious breach test under EC law. However, the same rationale would clearly not apply to private parties. Thirdly, while antitrust infringements are indeed serious breaches of law, antitrust law itself is not always clear. It should not be forgotten, for instance, that the Commission is currently trying to clarify and streamline the rules on the abuse of a dominant position, based on the underlying assumption that those rules do not always make sense. Moreover, the enlargement and modernization of EC competition law have notoriously increased the scope for diverging—and not always reconcilable—interpretations of the law. At least with respect to genuine cases of uncertainty as to the applicable rules, therefore, it would seem that the fault requirement still has a role to play. Finally, the strict liability standard in product liability stems from a fundamental need for corrective justice, pursuant to which it is socially acceptable to allocate the risk of defective products to companies with deep pockets, regardless of the degree of care they actually exercise in supervising the manufacturing process. However, such a need is not necessarily felt in competition law, where damages suffered by individual consumers may often be very small and may have much less dramatic consequences.12 Based on the foregoing, it would seem that the option of introducing a strict liability standard should be adequately assessed in light of the objectives that the Commission intends to pursue. A strict liability standard would certainly not increase deterrence, as companies would be liable for damages whatever the degree of care exercised. To the contrary, it may moderately facilitate compensation of injured parties because defendants would be liable even in those limited cases where they would otherwise have an absence-offault defence. However, this would hardly give sufficient added value for the European Commission to encroach upon established principles of tort or civil liability in the Member States. The second option would not add much value, because fault is already normally implied or presumed in cases of hard core violations. Therefore, assuming that the Commission decides to intervene in this area, the third option would be preferable. However, most probably, the introduction of a rebuttable presumption of fault would not significantly change the current scenario, where proof of infringements is in most cases already treated as implying a shift to the defendant of the burden of proving the absence of fault. 12 See also Wils W. P. J. (2005): Principles of European Antitrust Enforcement, Hart Publishing, Oxford and Portland, Oregon, at p. 125 et seq., where the author concludes that, “in the antitrust context, attempts to achieve corrective justice through damages actions are likely to be very costly or to lead to results which do not really serve corrective justice”.


482

12/10/07

13:12

Page 482


C. Discovery According to the Green Paper, access to evidence is key to making damages claims effective, particularly in stand-alone actions, where private applicants cannot benefit from the results of public authorities’ investigations.13 However, the Commission intends to strike a balance between allowing private parties sufficient access to evidence and avoiding the introduction of US-style discovery, which could encourage vexatious litigation. Extensive discovery allows the parties to better evaluate the strengths and weaknesses of their respective positions, and to get the facts right from the outset. On the other hand, they can be expensive and time-consuming, and they can potentially be abused in order to obtain business secrets. In the US, the parties normally conduct pre-trial discovery without the judge’s intervention. Document requests may therefore be massive, particularly in complex antitrust cases, and in order to avoid the ensuing costs, defendants are often willing to settle, regardless of the merits of the claim.14 To avoid encouraging unmeritorious litigation, all of the discovery options suggested by the Commission are subject to a fact pleading and require some form of court intervention to mitigate the extent of the parties’ requests. The Green Paper defines “fact pleading” as a document in which the party requesting discovery sets out the relevant facts of the case in detail and presents reasonably available evidence in support of its allegations. However, the contents of the fact pleading are not specified in any further detail. Ideally, the pleading should not allow the claimant to embark on a ‘fishing expedition’, in which claims too general and unsupported are put forward only to acquire useful evidence or business secrets from another party. The Commission envisages three possible options of discovery, all of which are subject to fact pleading. Under the first option, a party may request that the court order disclosure of relevant and reasonably identified documents. Under the second option, a party may ask the court to order disclosure of entire classes of documents, under some form of court supervision. Under the third option, a party may ask the other party to serve a list of relevant documents in its possession and make them available, unless a court is satisfied that disclosure may be harmful to the party who furnishes the list. The first option seems very close to current forms of discovery in place in Member States with a civil law tradition. In Italy, for instance, based on general procedural rules the judge may already order the disclosure of reasonably 13

Green Paper, at Section 2.1; Commission Staff Working Paper, Section II. See Comments of the Section of Antitrust Law and the Section of International Law of the American Bar Association in Response to the Request for Public Comment of the Commission of the European Communities On Damage Actions for Breaches of EU Antitrust Rules, at pp. 3 et seq. 14


12/10/07

13:12

Page 483


483

identified documents, deemed relevant for the pending case, by one of the parties or a third party. The second and third options would be more in line with discovery as applied by Member States with a common law tradition. The introduction of a harmonized discovery procedure throughout the EU could add considerable value to the regulation of private antitrust litigation. The information asymmetry between the defendants and the private claimants is indeed one of the major obstacles to the development of private antitrust litigation. However, if private enforcement is to become a credible alternative (or complement) to public enforcement, it is submitted that option one (and perhaps option two) would probably be too moderate. For instance, in the struggle against cartels, smoking guns are notoriously difficult to find even for the Commission and the national competition authorities, despite their considerable investigative powers. Therefore, it would seem unlikely that from the outset a private applicant would be able to acquire sufficient information for a fact pleading and to identify relevant evidence of the existence of a cartel. Option three could be much more effective in this regard. However, the requirements for the fact pleading and dynamics of court intervention should be carefully defined, in order to minimize the risk of tipping the balance towards a US-style culture of fishing expeditions.

D. Binding value of administrative decisions To alleviate the burden of proof on private applicants, the Green Paper suggests that infringement decisions of EU national competition authorities could be made binding on civil courts. Alternatively, the burden of proof could be reversed in favour of the party producing them in court.15 In Germany, previous decisions by the Commission, and by all of the Member States’ national competition authorities (subject to certain conditions) and courts have recently been given binding precedential value in private damages actions.16 In the UK, ordinary courts are bound by decisions of the national competition authority and by judgments of the Competition Appeals Tribunal. However, this is a highly controversial issue in countries where constitutional law enshrines the principle of separation between the administrative and judiciary powers. Furthermore, the focus of public and private enforcement may be radically different; while national competition authorities may be satisfied with proving infringements whose object or potential effect is inherently anticompetitive, 15

Green Paper, at p. 6, Option 8; Commission Staff Working Paper, at p. 27. Böge U. and Ost K. (2006): “Up and Running, Or Is It? Private Enforcement—The Situation in Germany and Policy Perspectives”, 27 European Competition Law Review 197. 16


484

12/10/07

13:12

Page 484


private damages actions are necessarily concerned with the actual effects of antitrust infringements. In complex cartel cases, where a national competition authority has proven the existence of a cartel as a whole, without focusing on every single episode, the final decision may contain several unchecked errors of fact. However, in the context of (administrative) judicial review of those decisions, these mistakes might never be corrected, as most often they are irrelevant to the solution of the case, or because the court might not have sufficient resources to carry out an extensive review of the file. As a result, civil courts may be the only place where the defendants have a chance to set the record straight, and to limit the extent of their liability. If infringement decisions of EU national competition authorities were made binding on civil courts, the latter would be deprived of this essential function. Moreover, in principle, civil courts assess decisions by national competition authorities, like any other evidence, on a case-by-case basis. However, those decisions have an undeniable evidential value, because in practice the judge will regard the facts described in such decisions as proven unless the defendant is able to rebut them. In follow-on actions, therefore, the burden of proof with respect to the infringement normally shifts to the defendant. It follows that a reversal of the burden of proof, as suggested by the Commission, would not, in practice, provide significant added value.

E. Conclusions The Green Paper sets forth the findings of an extensive review of EU Member States’ procedural rules in order to identify potential obstacles to the development of private antitrust litigation, and to pinpoint possible areas of intervention by the Commission. At the outset, we submit that, in most Member States, the perceived underdevelopment of private antitrust actions is partly a cultural factor, due to a general lack of awareness about the rights conferred by competition law and the remedies to enforce them. The Commission’s initiative is most welcome in this respect, precisely because it draws attention to private antitrust enforcement. However, this also means that, so far, the current procedural framework of private antitrust enforcement in the various Member States has not been tested to its fullest. Therefore, many national procedural rules currently considered inadequate may actually prove to be perfectly suitable once they are put to the test. An intervention at the EU level would not seem necessary with respect to the fault requirement, because the latter can hardly be considered a significant obstacle to private antitrust enforcement. Furthermore, there are no obvious policy reasons that would justify the application of a strict liability standard to tort or civil liability for antitrust infringements.


12/10/07

13:12

Page 485


485

By contrast, the introduction of an EU discovery procedure for private antitrust cases might prove very useful. However, the Green Paper does not provide much detail as to the fundamental features of the envisaged discovery procedures. To avoid excessive and vexatious litigation, such procedures should not in any event allow private applicants to set sail on fishing expeditions. Finally, national competition authorities’ decisions should not be made binding on civil courts because this would disrupt the balance of several national legal systems to a significant extent, without providing significant benefit.


12/10/07

13:12

Page 486


12/10/07

13:13

Page 487

SESSION C


2 PANEL VI: IMMUNITY/LENIENCY/ FINANCIAL INCENTIVES/PLEA BARGAINING

WORKING PAPERS


12/10/07

13:13

Page 488


490

12/10/07

13:13

Page 490


and legal communities by announcing that the Division would not to prosecute a company that reported its illegal conduct before any investigation was initiated and cooperated fully to achieve the prosecution of others in the industry.5 The program was an immediate success because it resulted in several important criminal prosecutions and effectively destabilized the cartel business. However, as time passed, the number of applications and expressions of interest declined. The intervening Reagan years were not the time for many companies to self-report antitrust violations; many companies probably believed that the policy and the criminal prosecution of antitrust offenders had become ancient history. Even before the Antitrust Division moved decisively into the international cartel business, Assistant Attorney General Anne Bingaman returned the leniency program to the antitrust spotlight in August 1993, announcing a revised leniency program that added a new and significant variation—leniency after the investigation has begun.6 The new policy revised the original policy to allow a company to come in after the initiation of the investigation, provide complete cooperation to the Division, and escape criminal prosecution for itself and its cooperating executives.7 To qualify for post-investigation leniency, a company was required to meet seven conditions: 1. The company is the first one to come forward and qualify for leniency with respect to the illegal activity being reported; 2. The company comes forward before the Division has evidence against it that is likely to result in a sustainable conviction; 3. The company, upon discovery of the illegal activity being reported, takes prompt and effective action to terminate its part in the activity; 4. The company reports the wrongdoing with candor and completeness and provides full and continuing cooperation that advances the Division’s investigation; 5. The confession of wrongdoing is a truly corporate act, as opposed to the isolated confessions of individual executives; 6. Where possible, the company makes restitution to injured parties;

5 John H. Shenefield, Assistant Attorney General, Antitrust Division, Statement before the 17th Annual Corporate Counsel Inst. (October 4, 1978), noted in Trade Reg. Rep. (CCH) para. 50,388. 6 Anne K. Bingaman, Assistant Attorney General, Antitrust Division, Antitrust Enforcement: “Some Initial Thoughts and Actions”, address before the ABA Section of Antitrust Law, New York (August 10, 1993). 7 Although this article’s primary emphasis is on the post-investigation leniency, it is important to remember that leniency is always available to a company reporting its conduct before any investigation occurs. Pre-investigation leniency is non-discretionary with the Division. Corporate counsel should not forget that this policy continues and should carefully consider it if an internal investigation uncovers serious antitrust violations.


12/10/07

13:13

Page 491

US Corporate Leniency After the Blockbuster Cartels

491

7. The Division determines that granting leniency would not be unfair to others, considering the nature of the illegal activity, the confessing company’s role in it, and when the company comes forward.8 The grant of post-investigation leniency by the Division is discretionary and, as is evident from the requirements, several of the criteria are very subjective. The Division has made it clear that the burden of satisfying these subjective conditions will be low if the company comes forward before the investigation begins or very early in the investigation. The burden will increase, however, as the Division gets closer to having sufficient evidence that would result in a sustainable conviction.

2. How leniency operates in practice The seven leniency requirements can really be viewed as three principles. The first principle distills the first five leniency considerations into a simple formula—first in and full cooperation. The Division must be certain that the company was the first in, was in before the Division had evidence likely to result in a sustainable conviction, took prompt and effective action to terminate the conspiracy, and cooperated fully and completely as a truly corporate act. The importance of this principle is underlined by the Division’s revocation of leniency in the Stolt-Nielsen matter. Importantly, the Division appeared to determine that the company did not terminate its role in the illegal activity when it assured the Division that it had.9 The second principle requires restitution to injured parties, where possible, and requires only that restitution be made to injured parties in the United States.10 The settlement of damage actions filed against the company is certainly an acceptable satisfaction of the restitution requirement. The third principle is that the Division will determine the fairness and propriety of granting leniency to the company. Essentially, if the Division determines that granting leniency to a certain company would be unfair to others, considering the illegal activity, the company’s role in it and the timing 8

Corporate Leniency Policy, supra note 1. Even after a district court and Court of Appeals adjudication, the facts in the Stolt-Nielsen matter remain confused and confusing. The fundamental point of the exercise is that the Antitrust Division acted when it believed that an applicant had failed to meet one or more terms of the leniency policy. Given that the Division’s consistent practice had been to make it possible for applicants to enter and remain in the program, this was certainly a major event that bears watching as it moves to the next level of the judicial process. The Division sought and obtained an indictment of Stolt-Nielsen, two subsidiaries and two executives on September 6, 2006. Stolt moved to dismiss and the hearing is scheduled to begin in May of 2007. 10 Gary R. Spratling, Deputy Assistant Attorney General, Antitrust Division, “The Corporate Leniency Policy: Answers to Recurring Questions,” paper presented at the ABA Antitrust Section Spring Meeting, Washington, DC, April 1, 1998, at p. 5. 9


492

12/10/07

13:13

Page 492


and extent of cooperation, the Division has the discretion to reject a leniency application.11 The gravest concern that Division officials face is the prospect of giving leniency to the organizer and instigator of the illegal conduct or to the “bully” who coerced others into the conduct. From a law enforcement perspective, the idea that the instigator or “bully” would be granted leniency while the others would be required to pay significant fines and subject their executives to prison sentences is a catastrophic situation for the Division. Further, from a trial strategy perspective, having the cooperating witnesses testify that his or her company instigated the conspiracy and coerced or threatened others to participate but was given a complete pass by the Division would certainly make cross-examination of the cooperating witnesses an uplifting moment for defense counsel. Such situations would make the Division’s case against any other defendant very difficult, if not impossible, to try successfully. This principle, therefore, allows the Division to make a very subjective determination of whether leniency would be fair to others—the most important indicator of the credibility and effectiveness of the leniency program. It is important to note that the Division makes a distinction between the instigator or organizer, and one of two or more “founding” companies of the conspiracy. If two or more companies come together to form an agreement, as is often the case, any one can seek leniency (if it is available) because it is not the organizer or leader.12 The distinction means that the sole instigator or the company that attempts to coerce others does not get leniency, yet one or more organizers that are equal partners in forming the conspiracy are all fully qualified to seek leniency. There have also been cases where the Division has offered the possibility of leniency to one company but not another because the other was clearly viewed as the organizer or instigator of the agreement (and, in some cases, the “bully”) who coerced, threatened or intimidated others. This principle is important to the subjective determination of leniency, but it could also provide opportunities, however unintentional, for a Division staff to make judgments that would treat one company differently than others have been treated by other Division staffs on these subjective issues. Division senior officials have been very careful to deal with these issues when they have arisen and to make centralized and consistent decisions. The Division needs to be vigilant to assure consistency of policy among its offices and from industry to industry. This is not something new, but it is something that merits the Division’s continual efforts to improve the decision-making process. With the high stakes confronting companies today, transparency, which means consistency and equitable treatment throughout the Division, must be the hallmark of its enforcement—and forgiveness—policy. 11 12

Corporate Leniency Policy, supra note 1, at p. 3. Spratling, supra note 10, at p. 7.


12/10/07

13:13

Page 493


493

3. Variations on a theme by the corporate leniency policy The Division has expanded or modified portions of the Corporate Leniency Policy on several occasions to make it far more attractive for a company on several levels. These expansions are part of the natural evolution of the leniency policy, especially as the major international cases of the past ten years allowed for evolution of the policy and experimentation with new ideas. There is a clear “carrot and stick” dynamic to the development to these tools. The first development was “Amnesty Plus,” where the Division provided leniency to a company for reporting a violation in a completely separate market from the one under investigation.13 The “Amnesty Plus” policy provides that even if a company cannot obtain leniency for product A, it can obtain leniency for product B and, in addition, receive an additional reduction from its corporate fine in product A as a result of its cooperation in product B. There is no limit to the number of products an applicant can bring in under “Amnesty Plus.” Following from “Amnesty Plus” is “Penalty Plus” which provides that if the Division finds a conspiracy that the party to an earlier case failed to report to the Division, the Division will argue aggressively to the court that the fact that it was not disclosed is an aggravating factor at sentencing. That generally means that the Division would recommend a fine at the upper end of the Sentencing Guidelines range, as well as increasing the “criminal history” Guidelines score because the company was a recidivist and should have found and reported the other conduct. With an executive, the upward departure will directly affect the amount of incarceration that the executive will serve by many months. The “Amnesty Plus” expansion of the policy came early in the initial international cartel enforcement phase in the mid–1990s. It was the enforcement “carrot.” The “Penalty Plus” provision came a little later and completes the cycle—it is truly the “stick.” It provides a powerful incentive for a company to expand its internal investigation of cartel behavior and come clean as to all potential antitrust violations that it uncovers. The Division has developed an extraordinary number of major cartel cases from the “Amnesty Plus” policy and corporate participants have benefited greatly from the policy. In fact, the Division reports that roughly half of international cartel investigations were initiated by “Amnesty Plus”-type evidence obtained as a result of an investigation in a completely separate market.14 The Division should be proud of the success of this policy, but should remain vigilant to assess and evaluate “Amnesty Plus” and “Penalty Plus” 13

Spratling, supra note 10, at p. 7. Scott D. Hammond, Deputy Assistant Attorney General, Antitrust Division, “Measuring the Value of Second-In Cooperation in Corporate Plea Negotiation,” ABA Section of Antitrust Law Spring Meeting, Washington, DC (March 29, 2006), at p. 9. 14


494

12/10/07

13:13

Page 494


situations with great care. First, the Division must maintain the requirements of the leniency policy and its investigative standards—and not accept indefinite or incomplete “markers” or fragments of evidence under “Amnesty Plus.” The “Amnesty Plus” policy has been criticized for accepting “leniency applications” from companies supplying only shreds of evidence in some markets—in other words, the company dumps all its findings on the Antitrust Division even when the evidence is neither strong nor clear. While the company is to be commended for coming clean with all of its evidence to the Division—and, indeed, that is the safe course of action—the consequences may be that evidence is stretched or exaggerated to make the proffer more attractive to the Division so a marker will be perfected into a leniency letter. There have been several situations where the evidence offered is far from clear and where it is only through counsel’s “interpretation” of the facts that a violation can be alleged. The Division is seeing a number of these cases and must carefully distinguish at the early stages what is going on. This has been a serious issue in some investigations. While the Division has been very careful to decline cases based on limited or flawed evidence, the problem is that it has opened investigations based on incomplete or interpretive “Amnesty Plus” proffers, spent time and resources investigating, and causing the parties to the investigation substantial anxiety, not to mention the huge expense to defend the investigation and sometimes meritless damage actions. At the same time, the Division must not rush to judgment on its “Penalty Plus” policy. There will be situations where, no matter how intensive an internal investigation takes place, the violation will not be uncovered. We have all been surprised by the ingenuity of the rogue employee who has kept hidden his conduct that clearly violates the law. That rogue employee, if he is acting alone within the company, should clearly be prosecuted in such a case and the company, despite its best efforts to uncover the conduct, will likely be prosecuted as well. If the “Penalty Plus” assessment is made against the company solely because it should have known about this conduct in which its executive was involved, the company should at least be entitled to provide rebuttal by presenting the careful efforts it made to uncover the conduct. It seems illogical that a company conducting the omnibus audit of all its products would hold back from reporting conduct in a certain product. It is far more likely that the executive who is running the scheme alone was successfully hiding that conduct from the company. The company should, as a matter of fairness, be able to establish its level of cooperation to the Division’s satisfaction. A final—and more recent—variation on the leniency policy is “Affirmative Amnesty.” This involves a situation where the Division discovers likely cartel conduct in a market and decides to approach one of the subject companies with the information and give it with the opportunity to cooperate in a covert investigation in return for leniency. The other companies will not be aware of the Division’s action. The Division has used this strategy on several occasions, selecting “candidates” that have previously obtained leniency,


12/10/07

13:13

Page 495


495

“amnesty plus” or cooperated as “second in.” Essentially, the Division is offering this great advantage to companies that it has worked with and trusted not to disclose the investigation to others in the market.15 While this may be an effective policy, there are concerns with the Division selecting a party, even one known and trusted, since it may look like the Division is playing favorites. More importantly, if the party selected is a recipient of leniency or “Amnesty Plus,” why is the Division turning to that party which clearly failed to report this conduct and should be an excellent candidate for “Penalty Plus” under the Division’s rationale for that policy? This policy needs to be transparent—and, since it was only recently announced, needs to be watched vigilantly in the future.

B. Leniency and the US investigation 1. How leniency has changed the investigative landscape In practice, the leniency program has revolutionized the way criminal investigations are conducted by the Antitrust Division and by defense counsel. In the past, conventional wisdom was for counsel to sit back, meditate on the Division’s subpoena, and casually begin to plan the document review, witness interviews, and a first meeting with the Division staff. Indeed, usually the only reason to call the Division’s staff early in the investigation was to extend the subpoena production return date. Today, that approach is exactly the wrong way to proceed. Instead, the word “leniency” likely will be prominent in counsel’s initial communications with the Division.16 If the staff advises that leniency is available (and occasionally, in the past, Division staff have left a subtle hint by attaching a copy of the Corporate Leniency Policy to its grand jury subpoena), or if defense counsel determines that it is available, there could be a race to the Division’s door to obtain leniency. If leniency is not available in the investigation, two other races could possibly begin: the race to be second in and obtain the best prosecution deal, and the race to obtain “Amnesty Plus” by bringing evidence of the second conspiracy to the Division in exchange for leniency on that conduct and a reduction of the fine level for the 15

Id. at 11. Some have argued that by raising the subject of leniency with the Antitrust Division in this initial conversation, you are telegraphing your culpability to them. From numerous conversations with Division staff over time it is perfectly clear that they understand the “is leniency available?” question to be an attempt to find out where the party stands in the investigation and how to conduct its investigation. It is in the Division staff’s interest to begin a cooperation race among the parties and that’s what may result from this conversation. 16


496

12/10/07

13:13

Page 496


first conspiracy.17 There have been, however, many situations where there is no race at all. After the parties conduct thorough internal investigations and discover no violations, there is often a very long period when the Division and the parties are discussing the evidence. These are often situations where it becomes evident that the leniency applicant provided very limited information or exaggerated the significance of a meeting or encounter among competitors. These situations are inevitable in circumstances where an agreement or meeting of the minds is the essential element of the violation of law. Leniency in other jurisdictions also requires quick action. Although other jurisdictions have somewhat different leniency procedures and standards, it has been interesting to watch the various policies “converge” over time. Over the period from 1997 to 2004, we have gone through the first full wave of problems that have resulted from divergent leniency policies. While a great deal still needs to be harmonized, including the important need for a real “one stop shop” for leniency in the European Community, the process has come a very long way very quickly.

2. How to approach the leniency decision 2.1. Educating the corporate decision makers Even five years ago, after the major international cartel matters such as vitamins, lysine and graphite electrodes had been publicized, company counsel and outside counsel often sought to discuss the criminal investigation, leniency, and criminal sentencing with corporate executives so they were aware of the possibilities. The consistent response was: we don’t want to hear about it—that will never happen to us, so do not waste the time or the company’s legal budget talking about hypotheticals. After watching the absolute chaos that accompanies the beginning of a cartel investigation and the dangers of obstruction of justice, false statements and failure to understand important information, my own practice is to insist, respectfully, that training about the criminal investigation be one of the essential elements of an antitrust compliance presentation. That presentation describes the initiation of an investigation, the perils of destroying documents, making false statements to prosecutors who visit the executive at home, or alerting counsel to actions that signal an investigation.18 That compliance session also must 17 Gary R. Spratling, Deputy Assistant Attorney General, Antitrust Division, “Making Companies An Offer They Can’t Refuse,” presented at the Bar Association of the District of Columbia’s 35th Annual Symposium on Associations and Antitrust, Washington, DC (February 16, 1999), at pp. 7–8. 18 For a more detailed description of this type of compliance presentation, see Klawiter D. (2005): “Antitrust Compliance in 2005: It’s Not The Numbers, It’s The Results”, The Antitrust Review of the Americas 2005 (Global Competition Review) at p. 14 (available upon request).


12/10/07

13:13

Page 497


497

explain the corporate leniency program so the executive knows that there may be a practical and effective way out of the criminal investigation, if one begins. If you explain leniency effectively, the executive instinctively understands why document destruction is not only illegal but stupid because it eliminates evidence that could help the company. The executive also understands how any false statement can jeopardize the executive’s admission into the leniency program. Most importantly, the executive understands the value of quick action to report questionable conduct. If the executive understands the documents are valuable to the company, false statements compromise the company’s and the executive’s ability to qualify for leniency, and that time is critical in organizing the investigation, counsel has been successful in his compliance program. A brief, pointed discussion of what to do if the prosecutor calls sets the stage for understanding and appreciating the investigation and the corporate risk it brings. In the vast majority of situations, it will be fifteen minutes of compliance training. In the situation where the prosecutor does knock on the door, the executives and counsel know what to do—and how to do it—quickly.

2.2. The “four day drill” In an investigation where leniency is available, the first four days following the service of the subpoena or the raid must be an intensive examination of the company’s information—documentary and testimonial. How can counsel conduct a full-scale investigation in four days or less? The simple answer is that counsel cannot. It can, however, conduct an efficient, bare bones probe in that time period. What counsel is looking to do is determine if there were significant communications with competitors that relate in any way to price or price-related issues. This is chiefly accomplished by short (30 minute) interviews of key executives. The executive should be advised that if there was significant competitor contact, this is the time and place to disclose it because the company could obtain leniency, which means that the company and the executive will have no criminal risk. Gaining the executive’s confidence at this stage is key—and that is the most difficult aspect of this process. This process also requires very quick review of basic corporate records, including e-mail files of key executives. Importantly, as corporate executives know that the lawyers have access to their e-mail files and documents, it should increase their level of candor considerably. Having the executive’s records as a check on what he says is a very valuable psychological tool to establish that counsel has substantial knowledge of the executive’s activities and that the matter is very serious and poses great personal risk. This “four day drill” is critical and the company needs to marshal the resources to get this done as efficiently as possible.


498

12/10/07

13:13

Page 498


2.3. The corporate decision Assuming one or more executives provides evidence of collusion—or enough evidence to pursue a leniency marker—the company must make an informed decision on what it means to apply for leniency. The Stolt-Nielsen19 controversy provides a sober reminder that leniency is a very serious step for a company and its executives. In my own experience, this is not a decision for the General Counsel alone. This is a Board issue, consistent with corporate governance practice—and consistent with common sense. The Board and the executives need to appreciate that if the company seeks leniency, the company is about to partner with the Antitrust Division, and it needs to do what the Division asks of it. The CEO and the Board must realize that they can never make a statement—public or private—critical of the investigation or its findings. A random “we don’t think this conduct was so terrible” remark on the front page of a trade journal, or as an aside to a cooperating witness, is considered an unfriendly act by the Division and could have grave consequences. Counsel need to make absolutely certain the illegal conduct has stopped when it represents to the Division that it has stopped—and that there are no communications with any of the competitors. Tipping off the competitors about the investigation is not only an unfriendly act, it opens the specter of obstruction of justice and places that executive in substantial jeopardy of being summarily thrown out of the leniency program, thrown out of a corporate position, and thrown into jail for the substantive offense and obstruction of justice. The Antitrust Division is generally quite helpful in clearing the way for the company to qualify for leniency. That assistance and welcoming attitude should not be mistaken for tolerance of short cuts or excuses. A company needs to be committed to devote resources to the leniency process and play perfectly by the rules. Failure to do so has consequences, as Stolt-Nielsen found out.

2.4. The presentation to the division Following the company’s decision to proceed with leniency, counsel should immediately arrange a call or meeting with the Antitrust Division staff responsible for the matter. The call to set up a first meeting should state that the company is seeking leniency and would like a marker to hold its place as first in line. The Division staff will generally give you some amount of time to come in and present your evidence. At times, the call to the Division staff will simply request a marker and explain that the company still needs some time 19 Stolt-Nielsen was the first public expulsion from the US Leniency Policy and resulted in significant litigation. Essentially, the Antitrust Division believed that Stolt-Nielsen gave the Division incorrect information about the timing of its withdrawal and about the evidence it provided.


12/10/07

13:13

Page 499


499

to collect its evidence in a coherent way. The staff will almost always provide a marker with a time limit, and there have also been many situations where the marker is extended. Generally, if the timetable is reasonable, the marker will run for some time—unless there is another party that is also seeking leniency and is ready to come in with its evidence. A second applicant allows the Division the luxury to put pressure on the first applicant. After all, competition is what the Division is all about! Counsel’s presentation to the staff should do two things. First, it should lay out the evidence that counsel has discovered that either suggests or establishes an antitrust conspiracy. Counsel should present this orally—and never, ever provide a written statement to the Division. If there is a writing, plaintiffs in antitrust damage actions will demand such a document because it is unprivileged and likely discoverable. Counsel should also be careful not to read from a script. Plaintiff’s counsel could attempt to argue that there is a writing of what was presented to the Division that was read—chapter and verse—by counsel. While that is a difficult basis to obtain the document by way of discovery, it complicates the discovery issues significantly. The presentation to the Division staff need not provide every small detail—it should identify the meetings, the discussion, the attendees, the results of the meeting and the follow-up conversations. It should be short, crisp and clear. If the company witnesses only take the evidence so far, it is critically important to inform the staff of that at the first meeting. If there is an inconsistency between witnesses, inform the staff candidly. If counsel has further work to do, that should be explained to the staff. This is, after all, a process, not a final examination. Second, counsel should briefly describe how the company meets the leniency criteria. The Division staff will look to counsel to provide a justification of why the company qualifies and how the evidence assists the Division in developing a sustainable case against others in the industry. The Division has taken leniency proffers on a variety of different terms. Some counsel have immediately provided a full and detailed explanation of the evidence, the companies and individuals involved, and its impact in the United States. At the end of the proffer, counsel will often request the conditional leniency letter. In other settings, counsel has only limited information and is interested in maintaining its “marker,”—that is, providing limited factual information, while advising the staff that they are looking into the matter that they believe will result in a case and they will get back to the staff very quickly. This is intended to “preserve the company’s place in line.” This process has evolved over the life of the 1993 Leniency Policy. In earlier years when many of the major blockbuster cartels were developing, there was much more immediate decision-making by the Division on leniency applications. There was also, generally, a far greater degree of certainty on counsel’s part that they had a clear violation of the law. In my own view, I also believe that counsel were more comfortable making a more complete presentation to the Division when the evidence was very decisive.


500

12/10/07

13:13

Page 500


As time has passed, counsel have increased the speed of the process and now initially approach the Division with a much less complete record. This is true for several reasons: First, counsel tend to approach the Division when they have a reasonably clear picture of the conduct, but not the full detail expected in earlier leniency applications. Anxiety about a “race” has clearly influenced counsel and companies to move quickly—if not completely. Second, many of today’s leniency proffers are much less “decisive” than the earlier blockbuster cases. Although those cases still exist to be sure, there are many more where the evidence is not complete or where there are missing witnesses (often an executive who left the company’s employ under less than optimal circumstances). This complicates matters for the Division and the applicant. Third, in the aftermath of Stolt-Nielson, both the Division and the applicant seek to be more careful about getting the leniency decision right. The Division acutely understands that expulsions from the leniency program decrease the likelihood that more companies will come forward—especially where the companies perceive Stolt-Nielsen as coming out of the experience far worse off than when they came in. This is far from an attractive situation and the Division appreciates this. From a company perspective, the decision makers must evaluate the pros and cons carefully and make certain they stay in once they get in. Fourth, with the passage of the Antitrust Criminal Penalty Enhancements and Reform Act of 2004,20 discussed infra, there is a far greater premium on obtaining leniency than there was before. There have reportedly been a number of leniency applications that relate to conduct that is very clearly a civil violation, not a criminal violation. Counsel and their clients are focused on the opportunity to qualify for de-trebling in its civil action by self reporting to the Division and cooperating with plaintiffs. Although the Antitrust Division has not commented publicly on this development, the increasing use of the marker to determine whether there is a criminal violation appears to be a direct reaction to the increased number of leniency applicants flowing from civil litigation situations. This could become a matter of great concern, and it is the important responsibility of the Antitrust Division to dispose of these purely civil matters without placing additional burdens on the true criminal matters that leniency applicants should continue to bring to the Division in good faith.

2.5. Random fire drills After the preliminary proffer is made and the Division makes its internal decision whether to proceed with the leniency application and. thus, a full-scale investigation, counsel must produce documents and witnesses that will substantiate the proffer and provide the Division with the type of evidence that 20

Pub. L. No. 108-237 § 213, 118 Stat. 661, 666–67 (2004).


12/10/07

13:13

Page 501


501

it promised at the time of the initial meeting or meetings. Although this process often takes several months, the Division should insist that it be as quick and direct a process as possible. The reason for having a leniency program is to save the Division time and effort in detecting violations of the law. A party that gives a detailed proffer and then takes several months to get its documents and their witnesses together does not enhance the Division’s enforcement efforts, although there are often problems with developing the evidence, especially when relying on former employees. Companies seeking leniency have an obligation to provide information quickly and efficiently and the Division has a responsibility to make sure that happens. The fairness of the program is based on decisive action and not delay or opportunistic tactics by an applicant or the Division. There will be days when the Division and the applicant will be at odds or when a random answer or comment by a proffered witness will cause controversy. The circumstance of company executives cooperating with the Division against individuals who are often close friends is exceedingly difficult. Similarly, the tension between a senior executive’s desire to place his company in the best light with investors and the public is antithetical to admitting to violating the law and receiving leniency. The desire to say “we did not do anything really bad” is irresistible, yet it—or anything like it—cannot be said. A company’s Board and executives must be advised of how difficult this process is likely to be. Effective counsel will navigate the company and its executives through these dangerous waters. The key in a leniency application is to keep your eye on the goal—no criminal prosecution, no executives prosecuted or in jail and single damages in civil actions. Those are powerful incentives to avoid the fire drills. Counsel should ensure that company employees and executives understand from the outset that once the leniency decision is made, it is an ongoing obligation where cooperation is expected when requested and, even if the investigation goes on for several years, the company is committed to assisting the Division in every way it can consistent with the leniency agreement. This ultimately is a major long-term corporate responsibility, not a one-shot deal.

2.6. The true partnership In the end, the leniency program cannot work without the full cooperation of both the Division and the company. At the point that a company is accepted conditionally into the leniency program or has a marker firmly secured, the Division and the company are allies and will work together against whatever obstacles remain in the investigation. The Division must help the leniency recipient to meet the requirements of the program, and the recipient must meet them swiftly, completely, and consistently over time. Neither party should create obstacles for the other—this is especially valuable advice to the


502

12/10/07

13:13

Page 502


Division when it is dealing with so many applications, many of which may not involve actual criminal conduct. The Division should make certain it is building effective partnerships and removing obstacles for the sincere applicants with the serious evidence of criminal conduct. Ultimately, it is a question of commitment, trust and good faith that both parties have a solemn obligation to uphold. If either party attempts to exploit the process to the detriment of the other party, the integrity of the entire leniency policy will be compromised.

3. Leniency does not come without costs and risks While leniency in the United States exempts the company and its cooperating executives from criminal prosecution, there are still serious risks and consequences to the leniency decision. When a company is accepted into the US leniency program, it faces three major risks that it must confront with eyes fully open: civil damages actions, multi-jurisdictional investigations and corporate governance and internal operational issues.

3.1. Antitrust private damage actions First, the major consequence of obtaining leniency is the virtual certainty of the company being sued in antitrust civil damage actions in the United States and the disruption to corporate life that such litigation will cause. Ten years ago, private plaintiffs would generally bring antitrust class actions only when the Antitrust Division brought a criminal action against the participants in a market. Since that time, plaintiffs have become more skilled at finding out about investigations in the US, Europe and elsewhere because of two factors: more diligent and creative reporting by news sources and greater likelihood that a company would report the fact of the investigation in its securities disclosure statements. The first is the result of a more public focus on antitrust cases. The front-page stories describing the ADM investigation started a trend that has continued. Enterprising business reporters are looking to be the next Kurt Eichenwald.21 Details of dawn raids, industry rumors and reports on corporate disclosure statements are now everyday news events. The ADM scandal started it; the Enron-World Com-Tyco events have expanded it enormously. The second is a direct result of more focus on corporate governance issues, by which companies are far more likely to report the service of a grand jury subpoena or a dawn raid than 21 See Eichenwald K. (2000): The Informant, Broadway Books, New York. Mr. Eichenwald produced the detailed account of the ADM saga and later went on to chronicle the Enron story in Eichenwald K. (2005): Conspiracy of Fools, Broadway Books, New York.


12/10/07

13:13

Page 503


503

previously. Today, potential plaintiffs are on the lookout for these signs of an investigation and will often file actions on the basis of the news articles or disclosures.22 In some situations, plaintiffs still hold back to see what happens next. In those situations, the disclosure that a company has sought leniency provides far more credible evidence that plaintiffs should pursue their claim. As we will see, however, the fact that a company applied for leniency is no longer the assurance that a criminal case is coming—but it is certainly a better bet than if no leniency applicant comes forward. The leniency recipient, of course, is not prosecuted by the Division. The plaintiffs’ bar, however, will find out about the company’s role by a securities act disclosure or from industry sources. Thus, in virtually all cases, the leniency recipient will be named in the damage actions. Often, in considering the leniency option, the major problem the company has in obtaining leniency is the prospect of massive damage actions by federal class plaintiffs, federal direct action plaintiffs, and state indirect purchaser plaintiffs. In the wake of vitamins, graphite electrodes, lysine, citric acid, DRAM and other huge international cases, corporate decision makers are justifiably concerned about bringing this paper avalanche on itself with the resulting catastrophic economic consequences. Although treble damage actions pose a substantial risk, the company should understand that, regardless of whether it applies for leniency, the damage actions are highly likely to occur. While there is always the chance that the company’s decision to forego leniency will result in no criminal prosecution, the fact that the Division has opened a grand jury investigation means that the Division has a respectable chance to develop a case—if not from the company contemplating leniency, then from other corporate competitors or their employees. If the treble damage cases will be filed anyway, the leniency applicant should consider three very important advantages that leniency could bring. First, because it has a head start, the leniency applicant has the time and the ability to plan a restitution strategy. The Division does not announce leniency deals, so the company has the time, if it wishes, to approach its customers before others are charged—and at least attempt to get the best deal it can from its customers. By making a voluntary approach, the company is also likely to generate some good will—or at least not as much bad will. Second, because the company receiving leniency will not be prosecuted, it will not plead, or be found, guilty. The important result is that there will be no prima facie effect, pursuant to 15 USC § 16. The plaintiffs will quickly realize that they are at a disadvantage in proving liability, because they cannot simply rely on the government’s case and the resulting finding of guilt, as they could 22 There are often civil damage complaints alleging billions of dollars in damages that base their allegations on Wall Street Journal reports about an Antitrust Division investigation or a European Commission dawn raid.


504

12/10/07

13:13

Page 504


with a guilty plea. The leniency applicant should be able to settle on somewhat more favorable terms than it would if it had been required to plead guilty. This may result in a lower percentage of damages applied to its sales— on top of its already great advantage of not paying any corporate fine. While the leniency applicant gains a considerable benefit from the head start and the absence of a prima facie effect in damage actions, plaintiffs increasingly scan the defendant pool for leniency candidates and attempt to obtain admissions or other evidence that the leniency candidate has given to the Division in the course of its cooperation. Document requests demanding the leniency “application,” the signed leniency agreement, and materials provided to the Division in the work product format in which the company presented the information to the Division are common in civil damage litigation. While plaintiffs have not been—and should never be—successful in establishing that leniency is an admission of guilt, nor have they been permitted to present the fact of a grant of leniency to a jury, the leniency recipient must always be concerned that any material it provides to the Division—or any other sovereign as part of its cooperation23—may be discoverable in the civil actions. The best course is to carefully consider what you provide in writing since a neatly packaged roadmap organizing documents and witness testimony may be far more helpful to the plaintiffs than to the Division. The leniency recipient is best served by proceeding by interviews and oral discussions of the evidence to avoid additional obstacles in the civil discovery process. In an effort to increase the incentives to apply for leniency and to lower the risk of substantial damages for the leniency applicant, the US Congress passed legislation in 2004 to provide the leniency applicant with the prospect of paying only single damages and not being subject to joint and several liability.24 Under the terms of this five-year experimental legislation, the applicant must have received leniency and cooperated with the plaintiffs in the litigation. This is a strong incentive—which further reduces the most serious monetary risks to the company—treble damages and the prospect of being held liable for the share of the settling co-defendants. While this is a great step forward and an incentive to seeking leniency and obtaining a much more reasonable “peace” with the private plaintiffs, the proposal has obtained little definition from Congress or the courts. On a positive note, one reported ruling has been issued under the Act. In July 2005, the US District Court for the Northern District of Illinois held that two affiliated companies were entitled to the Act’s protection because they had provided satisfactory cooperation in the civil litigation following the Antitrust Division’s sulfuric acid investigation. The court found “satisfactory coopera23

Spratling, supra note 17, at p. 5. The Antitrust Criminal Penalty Enhancements and Reforms Act of 2004, Pub. L. No. 108-237 § 213, 118 Stat. 661, 666–67 (2004). 24


12/10/07

13:13

Page 505


505

tion” based on witnesses provided and documents disclosed, as contemplated by the law.25 While the Sulfuric Acid case creates a procedure for evaluating “satisfactory cooperation,” the legislation itself does not set out such a procedure for determining when a leniency applicant qualifies for single damages. Under the statute, the applicant must provide “satisfactory cooperation” to the civil claimant, where cooperation “shall include providing a full account of the facts known to the applicant or cooperating individual . . . that are potentially relevant to the civil action that are in the possession, custody or control of the applicant.”26 The court determines the level of damages based on statements by the claimant and information provided by the leniency applicant. Without any more transparent standard, this procedure could allow the plaintiff to set a higher standard in some cases and make it enormously difficult to qualify for de-trebling even after the applicant already cooperated fully with the plaintiff. While most plaintiffs will fairly advise the court of the scope of cooperation and allow for the court to make an easy and straightforward decision, as it did in Sulfuric Acid, there is nothing to prevent an unscrupulous plaintiff from demanding more and more evidence, including evidence that the Antitrust Division did not require. Plaintiffs could demand things such as waiver of attorney-client privilege (something the Division has never requested in leniency situations), or require levels of economic expert analysis that would be very costly and potentially harmful to the applicant. This determination could become an adversary setting where the courts will need to articulate some standard. The outcome of that process will determine whether this highly praised tool will actually become the incentive for future leniency applicants that its drafters contemplated. As noted above, there is increasing concern that the availability of detrebling will overwhelm the Antitrust Division with leniency applications for conduct that is, at worst, a civil violation of the antitrust law. The concern is that applicants will be sued for treble damages, or be threatened with suit, and will rush to the Antitrust Division for leniency. Although there is no data available on this issue, it appears that there are at least some situations where this is happening; however, the floodgates have not yet opened. The Antitrust Division’s response to this potential problem will be critical to making certain that the leniency program is not overwhelmed by those applications and the original intent of finding criminal violations of law is maintained.

25 Id., Minute Entry, No. 1:03-CV-04576, In re Sulfuric Acid Antitrust Litig. (N.D. Ill. July 7, 2005) (identifying order granting Agreed Motion for a Finding of “Satisfactory Cooperation” and Limitation of Damages Pursuant to the Antitrust Criminal Penalty Enhancement and Reform Act of 2004) ; ABA Section of Antitrust Law, 2005 Annual Review of Antitrust Law Developments, at 173–74 (2006). 26 Pub. L. No. 108-237 § 213, 118 Stat. 661, 666–67 (2004).


506

12/10/07

13:13

Page 506


3.2. Enforcement actions in other jurisdictions Obtaining leniency in the United States does not help the company with any other jurisdictions. In international cartel cases, there is a strong likelihood that Canada, the European Commission and perhaps some of its member states, Japan, Korea, Brazil and others will open investigations as well. These jurisdictions require separate and distinct leniency applications that should be made as close in time to the US application as is practical for the company. The fact that a company is first in under the US program means nothing to other jurisdictions. While obtaining leniency in the United States is of enormous value because of the size of criminal penalties in the United States and the high probability of jail sentences for individuals, other jurisdictions also have severe penalties, civil and criminal, that need to be faced directly. One of the most significant benefits to obtaining US leniency in this environment, however, is that the Division has made clear that it will not share the information and materials a company gives it under the leniency program with any foreign sovereign absent a waiver by the applicant.27 This is a powerful incentive that allows the company to proceed to the next jurisdiction without fear that the United States will provide the other sovereign with a company’s own leniency information, to the detriment of the company. While this advantage will buy some time for the applicant to go to other sovereigns, the advantage is limited in time, since securities disclosure statements or the public filing of a criminal action against another party as a result of the company’s cooperation will notify the other sovereigns and allow them to open investigations promptly. Delay may also permit another party to run to these other sovereigns and obtain leniency leaving the US leniency recipient open to liability in those jurisdictions. This has occurred on numerous occasions. While other commentators will deal with the procedures of other jurisdictions in detail, it is worth noting that the proliferation of leniency programs makes life far more complicated for the leniency applicant. The lack of a “one stop shop” in Europe is especially difficult when a prospective applicant sees that it may need to make applications to as many as twenty jurisdictions in the Community, as well as the Commission. This is the type of procedure that, quite honestly, causes companies to question seriously whether they should apply for leniency—or question why to apply twenty times over when there is a single central body in Brussels. Not only would the applicant need twenty applications, it would be required to hire expert counsel in each separate jurisdiction. This is the type of situation that causes company executives and boards to question to value of the leniency program to them and to others in 27 Gary R. Spratling, Deputy Assistant Attorney General, Antitrust Division, “International Cartel Prosecutions: Antitrust Division Policies Relating to Plea Agreements in International Cases,” presented at the ABA Criminal Justice Section’s Thirteenth Annual National Institute on White Collar Crime, San Francisco, California (March 4, 1999) at pp. 10–11.


12/10/07

13:13

Page 507


507

the business world. This is an issue that must be resolved for the future success and credibility of the European program. Commissioner Kroes has proposed the idea of one “stop shop” and her efforts are to be commended.28 Despite Modernization, the major cases will still be handled by the Commission, or the Commission will effectively determine where they will be prosecuted. A single leniency application that can be provided to each of the interested member states, but centrally decided by the Commission, makes enormous sense and would certainly remove the strong objections of corporate executives for not seeking leniency. The final issue regarding multi-jurisdictional leniency is simple and direct. There should never be a situation where the leniency applicant is worse off for seeking leniency than the parties that did nothing to cooperate with the enforcers. This is a special concern in dealing with the need for a “paperless” leniency process that cannot ever subject the applicant to a disclosure that harms it far more than the other parties to the investigation. Whether it be the jurisdiction’s requirement of a written leniency statement that is discoverable in US civil litigation or the use of the recorded counsel statement that finds its way, in verbatim form, into the Statement of Objections, all jurisdictions must be vigilant to protect the leniency applicant’s disclosures.29 If they are not protected effectively, the incentive of parties to seek leniency will erode quickly with disastrous consequences to enforcement.

3.3. Corporate governance and internal operation issues Companies often find the leniency decision so difficult because of the nature of the evidence it must disclose, the corporate governance procedures it must enforce and the uncertainty of the result. First, a company must report not only its own “corporate” conduct, it must report the conduct of its employees, its competitors and the industry generally. Many corporate officials, especially outside of the United States, find it very distasteful to inform on their employees and industry colleagues. In an age of interdependence in the business world, a company may be taking action that will damage its joint venture partner in some other business, a fellow member of its strategic alliance, or its supplier of various important product components. While the Antitrust Division does not encourage closeness among corporate rivals (and antitrust counsel should not, either), it is difficult for a company to make a decision to implicate another company and its officials, knowing that there 28 Nellie Kroes, European Commissioner of Competition, “The First Hundred Days,” International Forum on European Competition Law (April 7, 2005). 29 ABA Section of Antitrust Law, “Comments In Response To The Commission of the European Communities’ Request For Public Comment On The Draft Amendment To The 2002 Commission Notice On Immunity For Fines And Reduction Of Fines In Cartel Cases” (April 24, 2006), http://www.abanet.org/antitrust/at-comments/2006/04-06/comm-ec-leniency.pdf.


508

12/10/07

13:13

Page 508


could be severe and irreparable consequences to it in the market place. In many respects, this is the most difficult step of the leniency decision-making process. Notwithstanding that difficulty, companies increasingly are discovering through antitrust investigations and litigation that each company must set its own course and look out for itself. It has become much more common for companies to report conduct and implicate their corporate colleagues than was the case even five to ten years ago. Nevertheless, companies remain reluctant to take such action, except in the most severe circumstances. Education and experience will probably be the most effective way to dissolve this reluctance, but, for the moment, it remains a serious issue that counsel and the Antitrust Division need to appreciate and understand. Second, in the corporate governance structure of today in the United States, the futures of the corporate executives who were instrumental in the company’s quest for leniency will be very much in doubt. Granted, the company executives who cooperated are free from the risk of indictment, criminal conviction and jail, but can they continue their positions and do they have futures in the corporate world? These are difficult questions, and there are usually no guarantees. Just as leniency is a board decision, the fate of the senior company executives who participated in the illegal conduct is also left for deliberations in the boardroom. Surely their executives were the heroes who cooperated fully so the company could qualify for leniency, but they were also the villains who got the company into the mess, which, in the best of situations, will cost many millions in legal fees and single damages in restitution. The prospect of this outcome is also something that may affect the front end cooperation of the executives in this increasingly complicated environment. Beyond the executives, securities law violations, shareholder actions and difficult shareholder and board meetings are among the risks and worries of the leniency applicant. Finally, one of the more frequent trends in leniency applicants in the current environment is that after the internal decisions, the securities disclosures, the payment of restitution to customers and the public perception as corporate sinner, the Antitrust Division or other enforcement jurisdiction decides not to proceed with an enforcement action. Often, the applicant self reports the conduct it can establish, but the remainder of the evidence never materializes. Sometimes, it is the missing witness who did not cooperate under the leniency umbrella. Sometimes, the evidence proffered was slightly exaggerated—or it was too old, or too limited jurisdictionally—or just did not meet the high standards of the Principles of Federal Prosecution. These situations do occur—and with more parties seeking leniency, they will occur more often. This is the ultimate risk in applying for leniency and should always be discussed with management and the board before deciding to go ahead. In most situations where the evidence is not complete or the jurisdictional nexus is tentative or the statute of limitation is approaching, the decision is a difficult


12/10/07

13:13

Page 509


509

one. In the majority of situations, the company will still opt for leniency because of the enormous downside of guessing wrong. It is, however, a very difficult decision. It is appropriate to end with the discussion of the failure to prosecute to make certain the corporate decision makers do not rush to a decision and are advised of all of the possibilities. These decisions are very serious and very hard. Although there is enormous value to the company and its executives from leniency, there are also risks and uncertainties. Make sure there is water in the pool before diving.


12/10/07

13:13

Page 510

(Z) Ehlermann 07 Joshua

12/10/07

13:13

Page 511

II Julian M. Joshua* That Uncertain Feeling: The Commission’s 2002 Leniency Notice

In upbeat articles in the Competition Policy Newsletter, Commission officials laud the success of its leniency programme1 in exposing huge price fixing cartels that would otherwise have remained undetected. To be sure, only last month, fines of a hefty €388 million were imposed on the peroxides and perborate cartel.2 But headline-catching fines may not be the only measure for judging success. Many of the applications received derived from applications for US immunity in worldwide cartels, so it is not necessarily the intrinsic “attractiveness” of the Commission’s leniency notice that brought the applicants through the door. In attributing the flood of applications to endogenous factors the Commission may be falling into the post hoc fallacy. Moreover, the statistics reveal that the Commission risks being overwhelmed by the sheer numbers: at least eighty applications for full immunity have been filed under the “new” notice, but only four decisions have emerged as a result. Given that one of the basic objectives of leniency policy is the rapid disposal of cases by encouraging cooperation against a reduced penalty, there are worrying signs that the current system is overstretched. Soon after assuming her post, Commissioner Neelie Kroes announced in her barnstorming “First Hundred Days” speech on 7 April 2005 3 her intention to look at wider systemic changes in the cartel area. Top of the reform agenda was the setting up of a one-stop leniency shop, but the review was also to include negotiated settlements and revamping the 1998 Fine Guidelines. The only proposal to emerge so far from DG Comp however is for an amendment of the Notice formalising the existing (and problematic) practice of “oral only” applications as the benchmark method for taking evidence from * Partner, Howrey LLP, 9–31 avenue des Nerviens, B-1040 Brussels. Tel: +32 2 741 1011. E-mail: [email protected]. The views expressed are personal to the author. 1 Commission Notice on immunity from fines and reduction of fines in cartel cases (the “Leniency Notice”), OJ C 45 [2002]. 2 IP/06/560 of 3 May 2006. This cartel was part of what seems like a much wider arrangement so Degussa’s decision to apply in the EC cannot be entirely disconnected from its potential criminal exposure in the US. Akzo and Solvay agreed to plead guilty in the US to charges of arrangements more limited in scope than the Commission press release would suggest were applied in Europe: see US v Akzo Nobel Chemicals International, Joint Sentencing Memorandum, Case N° CR 06-0160 MMC, ND,CA, 17 May 2006 ($32 million); US v Solvay, Plea Agreement, Case n° CR 06-0159 MMC, ND,CA, 20 April 2006 ($40.87 million). 3 “The First Hundred Days”, speech delivered to 40th Anniversary session of Studienvereinigung Kartellrecht, Brussels, 7 April 2005.


512

12/10/07

13:13

Page 512


cooperating undertakings.4 For many stakeholders, however, both would-be applicant companies and antitrust practitioners, the time is right for a root and branch review of the leniency framework set out in the Notice of 14 February 2002. Leniency is a global growth industry. Competition authorities across the world have climbed—albeit with different degrees of enthusiasm—onto the Justice Department’s enforcement bandwagon and adopted their own leniency programmes, all more or less modelled on the hugely successful US policy.5 The mushrooming of leniency programmes across the world brings its own set of problems, notably in assuring what could be called “interoperability” between the 30 or more different systems, but it does allow for benchmarking and comparison of one policy against another. There is a general consensus among enforcement agencies that the three pre-requisites to successful implementation of a leniency programme are: (1) the threat of significant sanctions for hardcore cartel participants who fail to self-report; (2) a heightened risk of detection by the authorities if they do not self-report; and (3) transparency and predictability throughout the whole of the jurisdiction’s cartel enforcement programme, so that companies can predict with a high degree of certainty how they will be treated if they seek leniency and what the consequences will be if they do not.6 As this paper will show, despite the accolades7 when it was unveiled in 2002, the terms, incentives and operating framework of the Commission’s leniency programme may in practice have fallen short of the ideal in terms of transparency. To be sure, there has been no shortage of takers under the new Notice, but the back up of pending cases does nothing to improve the “attractiveness” of coming in for immunity or leniency. The credibility of the programme depends upon every serious case that comes in also going out within a reasonable timeframe. There is a widening gap between the US and European timeframes in the same investigation. Applicants are baffled by the fact that while they have had “closure” in the US several years ago, their case in Europe is still in limbo. This paper proposes a “holistic” solution integrating the leniency regime, direct settlements and transparent fine guidelines. 4 Communication from Commission: Draft amendment of the 2002 Commission Notice on Immunity from fines and reduction of fines in cartel cases, released on 22 February 2006, available on http://ec.europa.eu/comm/competition/antitrust/legislation/leniency2_en.pdf. 5 Antitrust Division, US Department of Justice Corporate Leniency Policy (1993), text available at http://www.usdoj.gov/atr/public/guidelines/0091.htm. 6 “Drafting and Implementing an Effective Leniency Program”, in International Competition Network, Anti-Cartel Enforcement Manual, Cartel Working Group Subgroup 2, Chapter 2, p 3, largely reproducing the insights of Scott Hammond, then Director of Criminal Enforcement at the DoJ, in Cornerstones of an Effective Leniency Program, ICN Workshop on Leniency Programs, Sydney, 22–23 November 2004. 7 Hammond S. (2002): “A Review of Recent Cases and Developments in the Antitrust Division’s Criminal Enforcement Program”, The Conference Board, 2002 Antitrust Conference, New York, 7 March 2002.


12/10/07

13:13

Page 513

The Commission’s 2002 Leniency Notice 513

A. Benchmarking the “effective” leniency programme When the Commission issued its first leniency notice in 19968 it was only the second enforcement authority after the US Justice Department to articulate the bold policy of exonerating cartel members from the penalties that would otherwise apply as the reward for reporting their illegal activity and denouncing their fellow participants. Now no self-respecting antitrust agency with any aspiration to effective enforcement is without a leniency policy.9 Some fundamental characteristics are common to all leniency programmes worth the name: —automatic full immunity for the first in to cooperate where the enforcers are unaware of the cartel’s existence; —availability of immunity even after an investigation has begun (but with a higher burden of persuasion); —only one first prize; —far less attractive prizes (but ones still worth competing for) for the runners up. Programmes co-exist but do not mingle: nobody gives credit for getting in first somewhere else. Of course it is easy enough to draft a leniency policy that looks good on paper. The effectiveness of a policy, in the sense of the written set of rules or principles issued by the agency to govern the leniency process, can however only be evaluated in the wider context of the authority’s leniency programme, which includes the internal agency processes and most critically, the way in which it applies and administers its own policy.10 The global enforcers’ club, the “International Competition Network” (ICN) has itself set out the benchmark for drafting and implementing an effective programme in its Anti-Cartel Enforcement Manual,11 and especially its list of “good practices”. As the Guide is careful however to observe, practices depend on the peculiarities of each jurisdiction’s cartel regime, and the features of one leniency policy may not translate directly into another with different legal, competition and public policy considerations. What can however be said, especially if we look to the transatlantic model, is that the state of the art immunity programme unashamedly sets up a “race to the courthouse door”.12 Cognoscenti stress two basic requirements: 8

Commission Notice on the non-imposition or reduction in fines in cartel cases, OJ C 207 [1996]. At last count, the EC Commission and seventeen EU Member States, as well as fourteen jurisdictions outside the EC, had a cartels leniency programme. 10 A distinction succinctly made in the ICN’s Anti-Cartel Enforcement Manual, Chapter 2, p 2 (supra note 6). 11 See supra note 6. 12 Hammond S. (2000): “Detecting and deterring Cartel Activity through an Effective Leniency Program”, International Cartels Workshop, Brighton, 21–22 November 2000. 9


514

12/10/07

13:13

Page 514


certainty and “generosity”, the latter not implying any moral concession, rather meaning that any barriers the applicant has to cross in order to qualify are low.13 When an applicant is considering going in for amnesty, nothing should argue in favour of hesitation. As Scott Hammond, the chief of criminal enforcement at the Antitrust Division puts it, the guiding principle is: “How do you create an enforcement regime that causes management to conclude after it discovers a violation that it has but one viable choice—the company must self-report and cooperate or face certain and severe penalties?”14

These elements are particularly important in defining the choices faced by the would-be applicant in an international case. The standard—and invariably sound advice—is that applications for immunity should be made simultaneously or as close to simultaneously as possible in all the jurisdictions where there is significant exposure.15 This used to mean the US, Canada and the EC but other enforcement agencies are starting to flex their muscles and now that they also have leniency programmes, are required ports of call as well. Watershed decisions may have to be taken against a ticking clock. The nightmare scenario for any immunity applicant is to come in first in one jurisdiction and lose the race in another. Thus, the procedures and qualification requirements of the programmes in the main jurisdictions should not be too far out of line, otherwise applicants might hesitate to go in at all, and indeed the United States has not eschewed robust diplomacy when it perceives an overseas leniency policy to operate as a “chill” on its own programme. Leniency is not only supposed to bring applicants in through the door, it is also intended to speed up the resolution of their cases. With a confession from an insider and the incrimination of the other participants, the whole idea is “to obtain evidence more quickly, and at lower direct cost compared to other methods of investigation, and leading to prompt and efficient resolution of cases.”16 While lawyer/moralists see leniency in terms of a Benthamite hedonic calculus, economists view it as a welfare enhancer: more cases resolved more quickly at lower cost = enhanced deterrence.17

13

As Mr Hammond puts it, “The Division’s philosophy has always been that wherever possible, we will tilt our program in favour of finding ways to make companies eligible for our program rather than looking for ways to keep them out.” (Hammond, “Cornerstones” speech, supra note no. 6 ante);

Hammond speeches, passim. 14 “Cornerstones” speech, supra note 6. 15 Joshua J. and Klawiter D. (2002): “International Cartels Investigations: Practical Lessons in Multi-jurisdictional Coordination”, GCR Antitrust Review of the Americas. In case of doubt in deciding priorities, the clincher is the question: “Where will it hurt most to come in second?”. 16 ICN Anti-Cartel Enforcement Manual, Ch 2, p 2 (supra note 6). 17 See e.g. Motta M. and Polo G. (2003): “Leniency Programs and Cartel Prosecution”, 21 International Journal of Industrial Organization 347; Møllgaard P. (2002): Must Trust Bust?, Copenhagen Business School, Department of Economics Working Papers 02/2002.


12/10/07

13:13

Page 515


B. The 2002 Notice The Commission unveiled its new programme on 14 February 2002.18 Probably the single most important and certainly necessary step from the perspective of effectiveness was the guarantee of automatic, if conditional, immunity to the first applicant in to qualify. The other crucial change was the discarding of the “decisive evidence” rule and the adoption instead of two separate evidential thresholds.19 Making the applicant jump whichever of the two separate evidential hurdles was applicable to the particular circumstances however was in marked contrast to the United States (and Canadian) systems where all a candidate had to do before getting the conditional grant of immunity was to report the violation, offer to “tell all it knows” and leave the rest to the law enforcement process.20

C. Certainty and transparency are the prerequisite of an effective programme “Certainty” is said to be one of the primary requisites of a successful leniency policy, but it is worth exploring the nuances, since uncertainty as to the state 18

See supra note 1. The most important features of the Commission’s Valentine card were:

—An “upfront” guarantee of full immunity from fines for the first company to cooperate with the Commission in the detection and prosecution of a secret cartel; —The replacement of the much-maligned “decisive evidence” test by two different evidential thresholds for obtaining immunity in different situations; —The possibility of amnesty even after an investigation has begun (other of course than as a result of a leniency application). 19

The two standards laid down in the Notice are as follows:

—Under point 8(a), a full pass from fines is promised to the first cartel member that submits evidence which in the Commission’s view may enable it to order a surprise investigation or “dawn raid”; —Under 8(b), a more demanding though still ambiguous evidential threshold is required, with the applicant being obliged to provide evidence which—again “in the Commission’s view”— may enable it to find an infringement of Article 81 EC. 20 As the OFT also does in its Leniency programme: see OFT’s Guidance as to the Appropriate Amount of a Penalty, Para 3.9, p. 14 (December 2004). The Commission Notice does not actually say in terms when each of the “alternative” standards will apply. The execution of a dawn raid is not itself the watershed, rather the Commission’s internalized perception of its state of knowledge; a candidate coming in before the investigation may be held to the higher standard if the Commission already suspects the existence of the cartel in question. And it goes without saying that even if a dawn raid carried out following an 8(a) application is a failure, rushing in the next day with the full set of cartel minutes and quota scheme accounts will win at best a fifty per cent reduction.


516

12/10/07

13:13

Page 516


of the investigator’s knowledge and the intentions of other cartel members is of course the driver of the decision to self-report. What is needed to encourage the applicant to come in is certainty of outcome, which requires abdication of discretion, perhaps the hardest thing for a prosecutor to accept. As Mr Hammond has said, “Prospective amnesty applicants come forward in direct proportion to the predictability and certainty of whether they will be accepted into the program. If a company cannot accurately predict how it will be treated as a result of a corporate confession, our experience suggests that is far less likely to report its wrongdoing, especially where there is no ongoing government investigation. Uncertainty in the qualification process will kill an amnesty program.”21

Cooperation from violators in the leniency process is dependent upon the programme and its administrators providing the security in all areas of its application that will allow the company to “predict with a high degree of certainty how it will be treated if it reports the conduct and what the consequences will be if it does not.” Not only the articulation of the standards for qualifying, but also the operation and application in practice of those standards and policies have to be transparent. US prosecutors bit the bullet and made the grant of immunity in their 1992 leniency program automatic instead of discretionary, at least in cases where a the Justice Department was unaware of the conspiracy. Transparency does not however stop there. To make the programme “attractive”, the Division has also sought to provide certainty across the board, setting out and sticking to transparent standards for opening an investigation and filing charges, transparent enforcement priorities, transparent application of the immunity programme and transparent policies on sentencing and negotiating plea agreements.22 Given that for all the emphasis on “certainty”, the grant of immunity after an investigation has started is discretionary, and indeed the conditions for “automatic” amnesty call for some degree of subjective assessment, the credibility of the programme depended on building up the trust of applicants by consistent application and fair dealing. In the congratulatory atmosphere surrounding the announcement of the Commission’s new notice, Mr Hammond lauded it as “the single most important contribution to US antitrust enforcement of 2002”.23 To be sure, the insecurity as to whether full immunity as opposed to a 75 per cent cut had been removed, but how do the Commission’s alternative qualification criteria, and their application in particular cases, measure up against the benchmark of certainty? 21

Hammond’s “Cornerstones” speech, supra note 6. Hammond speeches, passim. 23 Oral remarks of Scott Hammond, ABA Section of Antitrust Spring Meeting, Washington DC, 2 April 2003. The belief had gained currency among American lawyers that the original Commission Notice had put potential applicants off the whole idea of leniency altogether, so they were taking the conscious—and probably suicidal—decision to stay out in the US. 22


12/10/07

13:13

Page 517


D. The Commission’s Leniency Notice examined 1. Subjectivity of the 8(a) and 8(b) tests Qualifying for immunity before the authority is aware of the cartel should be a near-automatic process, although obviously it is the agency itself that has to decide whether the objective criteria are met. Both the “dawn raid sufficient” test under 8(a) and the 8(b) “sufficient for a final decision” test are however on their face entirely subjective, as the rider “in the Commission’s view” underlines.24 As regards 8(a), there is no reviewable rule or standard as what level of evidence is sufficient to warrant a dawn raid, and no requirement in the regulation or case law for any “reasonable ground to suspect/believe” or any probable cause.25 Under point 9 of the Notice 8(a) immunity is only available if the Commission did not have sufficient evidence to adopt a dawn raid decision at the time of the application. To the extent that the Commission may already have some knowledge or suspicion that a cartel is existent, thus potentially making any immunity applicant meet the equally subjective but higher evidential standard of 8(b) instead, there is room for cognitive bias. Since the Commission does not even need any “evidence” at all to adopt a decision under Article 20(4), the precise circumstances in which the bar comes down so as to exclude an 8(a) application are hazy. Absent a clearly stated rule on the standards for opening an investigation, the Commission must be vigilant to ensure consistency in its treatment of companies and their advisers. While some competition authorities explicitly state in their leniency policies that they will err in favour of the applicant where there is a genuinely close call,26 24 During consultations on the new draft Notice, the ABA Section of Antitrust warned, with some prescience, that information provided by an applicant could substantially advance the investigation while technically not being considered enough for a dawn raid. “The proposed standard would in those circumstances discourage, rather than encourage, those that possess relevant information about possible violation from cooperating.” Commission officials maintained that the sufficiency of the evidence needed to launch a dawn raid had to remain within the Commission’s discretion as it did not want to be put in the position of conducting dawn raids “against its better judgment” or face challenges in court from an applicant, a justification which begs the question of whether the subjective “dawn raid sufficient” test is appropriate in the first place: whether an applicant qualifies ought not to depend on the Commission’s inclination to pursue investigations or the availability of its resources. 25 “Necessary” as used in the Regulation (“the Commission may carry out all necessary inspections”) does not mean indispensable. Whether or not to order a dawn raid is a matter for the Commission’s discretion, with which the Courts are loath to interfere: National Panasonic v Commission [1980] ECR 2033; for commentary, see e.g. Joshua J. (1983): “The Element of Surprise”, 8 European Law Review 3. 26 See e.g. Office of Fair Trading, Leniency and No-action, July 2005 (OFT803). Also the DoJ’s guiding principle as articulated by Mr Hammond, see supra note 13.


518

12/10/07

13:13

Page 518


the Commission is more tight-lipped. With so many applications coming in, ensuring consistent treatment across the board may be a priority task for the new Cartel Directorate, as greater predictability would go a long way towards encouraging doubters to come in. Evidence which “in the Commission’s view may enable it to find an infringement of Article 81 EC in connection with an alleged cartel affecting the Community” is an equally elastic standard. It obviously denotes a higher standard than 8(a), but remains elusive. How conditional is “may”? Is all that is needed a prima facie case? Does “enable” imply that no further investigative steps are needed at all, or merely that the information has significantly advanced or is the clincher in the investigation? The “decisive evidence” test of the old Notice did not require the applicant to drop the whole of the evidence in the Commission’s lap: a “smoking gun” document that would crack the defence wall of silence was enough.27 Towards the end of its life, the standards of the 1996 Notice were notably relaxed; the first company through the door in Vitamins obtained a full pass from fines although it did not provide contemporaneous documents, merely a “corporate statement”.28 Now, what constitutes evidence “enabling the Commission to adopt a decision” has become even more difficult to predict as the Commission has turned increasingly to using oral lawyer recitation not just as the vehicle for providing an 8(a) roadmap, but also to serve as probative “testimony” for use in a Statement of Objections.29 Indeed, the boundary between “dawn raid sufficient” and “sufficient for a decision finding an infringement” has also become increasingly blurred in two ways. First, practitioners coming in under 8(a) are finding that they are sometimes required by officials to provide far more evidence than what ought to suffice to enable the Commission to mount a dawn raid.30 Secondly, if a dawn raid produces only slim pickings, statements originally made by the amnesty applicant’s lawyers to support the 8(a) application may well be used in the Statement of Objections as proof of the substantive violation. This secondary use may account for the increasingly high level of detail demanded, but it 27 For all the deficiencies of the old Notice, the decisive evidence rule itself did at least have the advantage of certainty. Everybody knew in practice what it meant: see Joshua J. and von Hinten R. (2001): “Rethinking Leniency at the European Commission”, GCR, October/ November, pp 15–18; Carle J., Lindeborg S. and Segenmark P. (2002): “The New Leniency Notice”, 23 European Competition Law Review 265. The vast majority of the huge fines collected by the Commission in the past five years resulted either from “own initiative” investigations or the old 1996 Leniency Notice, although once again it is likely that many of the applications were derived from going in the US. 28 Rhone-Poulenc in Vitamins, OJ L 6 [2003]. 29 In some recent cases the Commission has apparently abandoned part of the case after the probative value of narrative provided by the lawyers of immunity applicants had been challenged at the hearing by other addressees of the SO. The issue would not arise if 8(a) “roadmap” material were not recycled into “evidence”. 30 Although practice is not always consistent, the Commission may temporarily withhold the “acknowledgement of receipt” until the application is on its face complete.


12/10/07

13:13

Page 519

The Commission’s 2002 Leniency Notice 519 adds to the uncertainty and may well further negatively impact the already complex operation of coordinating global applications.31 Contrast the procedure with the Justice Department’s. The moment the violation is discovered by the appropriate instances in the company, counsel can “put down a marker” with the Department. A telephone call will do. A few days may suffice to uncover enough information to be able to make a “proffer”, i.e. an outline by the company’s lawyers of what it will be able to tell the agency going forward. Verification may be needed and there will be discussion on the terms of the amnesty agreement, but the grant will be forthcoming in short order.32 Crucially, the proffer does not involve providing actual evidence, just a good “cartel story”, and “evidence” only has to be provided after the applicant is granted immunity. In the US system, the process of making good on the proffer is then largely driven by the DoJ as it investigates and develops the case, seeking to interview witnesses or calling for documents to obtain the full picture to put the case before a grand jury, but immunity has been granted upfront. But in Europe, there is no provision for a marker and the Commission will not even give the conditional grant of immunity against an outline “proffer”. The procedure under Point 13 of the Notice requires the applicant to provide already packaged “evidence” meeting the relevant standard at the time it makes its application. As the Commission official responsible for overseeing the operation of the leniency programme has put it, “in order to be recognized as an application, the application has to be substantial and provide all the evidence and information the applicant has at its disposal at that point in time.”33 Recognising however that “it does happen” that a company is in a great hurry to apply, the Commission may grant some leeway, but any concession will be limited and very much case-specific. The “model” application expected by officials requires a fairly extensive production,34 unlike the US procedure, which does not call for evidence or testimony as such to be submitted before the grant of immunity is made. Obviously there is a duty of full cooperation, there must be full disclosure at some stage, and applicants must not be able to present a selectively filtered view of the cartel or their role in it,

31 On the coordination of global applications, see e.g.: Joshua J., Low D. and Spratling G. (2001): “International Cartels Investigations: How to Contain the Pain”, GCR Cartel Regulation 2001, pp 3–5; Joshua J. and Klawiter D. (2002): “International Cartels Investigations: Practical Lessons in Multijurisdictional Coordination”, GCR Antitrust Review of the Americas, pp 7; Mobley S. and Arakistain M. (2000): “Seeking Leniency Worldwide—A Global Strategy”, Global Competition Review 16; Turner V. (2005): “International Cartel Enforcement: Multijurisdictional Defence Strategies”, paper given at ABA Section of International Law Annual Conference, Brussels 26–29 October 2005. 32 In the meantime, pending issue of the provisional grant of amnesty, the position of the applicant is reserved by a “marker” letter issued by the DoJ as a matter of routine. 33 Van Barlingen B. (2003): “The European Commission’s 2002 Leniency Notice After One Year of Operation”, 2 Competition Policy Newsletter 18. 34 Ibid., pp 18–24.


520

12/10/07

13:13

Page 520


but the level of detail and documentation required up front from the applicant by the present Notice is at the expense of speed. The point of having two different tests was to allow for immunity still to be available after an investigation had begun, a notable omission from the old Notice. Obviously getting full immunity at this later stage should be less easy, so that the incentives to come forward in a timely fashion are maintained. However, after a dawn raid—or indeed if the Commission believes it has enough to start one—the requirement for the applicant to produce evidence to allow the Commission to adopt a decision would seem to call for a volume and standard of evidence that could take months to assemble. The standard seems rather like the old “decisive evidence” test ratcheted up a good few notches—and made more discretionary. Rather than making immunity “automatic” on qualification but calling for a standard that could be dissuasive in practice, other antitrust authorities operating successful programmes make the grant of immunity discretionary after the commencement of the investigation, with transparent standards as to their ability realistically to take the case forward at the time of the applicant’s approach.35

2. Playing poker with the Commission? It is not just the elastic formulation of the 8(a) and 8(b) tests that creates uncertainty. Adding to the discombobulating effect is a lack of transparency in the process. Instead of the unambiguous “fork in the road” offered by the US system, the Notice invites the applicant to enter a poker game in which the Commission sees the other side’s cards but keeps its own close to its chest.36 For applicants it is very important to know whether full immunity is still available when they make their application. As drafted, the Notice makes the appropriate concessions in favour of transparency in the qualification process. The Commission is supposed to tell the applicant immediately that immunity is not available if it becomes apparent that it does not meet the conditions in Points 8–10. An applicant who does not meet the conditions 35 See the US Corporate Amnesty Program, supra note 5. This is the method adopted by the OFT’s programme, acknowledged by many commentators to be the most effective of the leniency programmes operated by the NCAs: see e.g. Henry D. (2003): “Leniency Programmes: An Anaemic Carrot for Cartels in France, Germany and the UK”, 26 European Competition Law Review 13. The OFT provides a clear rule defining the “cut-off point” when full immunity after the investigation is no longer available: the OFT must have issued a written r. 14 Notice stating that it proposes to take action finding that Chapter1/Art. 81EC have been infringed. 36 Joshua J. and Camesasca P. (2005): “Cartels and Leniency: Hold or Fold? The Commission’s New Leniency Policy”, GCR Antitrust Review 1; Joshua J. and Camesasca P. (2005): “Where Angels Fear to Tread: the Commission’s ‘New’ Leniency Policy Revisited”, GCR European Antitrust Review 10. See also Carle, Lindeborg and Segenmark, supra note 27, opining that the dawn raid sufficiency test is “arbitrary and open to manipulation” and that the Commission “seems to take away with one hand what it gives with the other”.


12/10/07

13:13

Page 521

The Commission’s 2002 Leniency Notice 521 for immunity under 8(a) or 8(b) is then allowed to withdraw the evidence disclosed for the purposes of the application.37 To be sure, making the dawn raid to a large extent the pivotal point brings some risks to the “integrity” of the procedure.38 However, rather than relying on the robustness of the system and properly refined incentives to deter any would-be manipulators, the Commission tries to deal with this by making an applicant for immunity commit in advance to converting the 8(a) or 8(b) approach into a request for a fine reduction if told that immunity is not available. This is not what the Notice says, but some applicants who are riskaverse may accept. It is not however so easy to convert an 8(a) application into one for a fine reduction. In the pre-dawn raid scenario, (a) the company coming in with an 8(a) tip-off might not in fact even have sufficient evidence immediately available to meet the higher “value added” or “sufficient for a final decision” requirement (whichever may be applicable is not immediately apparent) and (b) the idea of talking one’s self into paying a fifty per cent fine even before any dawn raid has been carried out lacks a certain attraction. If Commission officials cannot give a straight answer to a good faith enquiry what should be a transparent process is turned into guessing game. Both before and after a dawn raid, the incentives to come forward and confess are thus less than universally compelling, absent some endogenous factor like exposure to the criminal prosecution of individuals in the United States. The uncertainties of the process are such that if nobody is going to go to jail some companies may think twice about upsetting the industry status quo and hope their fellow cartelists are thinking the same.39 In practical terms, the Commission will probably never be able to wield the big stick available to the Justice Department, but the fact that management discovering a cartel may still regard “stay out and hope the others do the same” as a viable option unless there is US criminal law exposure could ultimately undermine the foundations for the Commission’s status as a world class antitrust authority in its own right. 37 2002 Notice, point 17. According to Commission officials Arbault and Peiro, writing in June 2002: “If immunity is no longer available for the infringement in question, the applicant will immediately be informed. Should this be the case, the applicant may still request the Commission to consider its application under section B of the Notice.” (emphasis added) Arbault F. and Peiro F. (2002): “The Commission’s New Notice on Immunity and Reduction of Fines in Cartel Cases: Building on Success”, 2 Competition Policy Newsletter 15 (Emphasis added). 38 If a second company approaches the Commission between the time immunity has been granted under 8(a) and the dawn raid (if any) occurs, a straight “no” answer would alert the inquirer to an impending dawn raid and allow the unscrupulous to warn their fellows to turn on the shredder. Post-investigation, there are further opportunities for the canny to get out of the prisoner’s dilemma: a “yes” response to an enquiry by a cartel member as to whether immunity is still available could convey the useful intelligence that (a) there was no 8(a) immunity applicant; and (b) the dawn raid had failed. It is easy enough to work out how the system could be manipulated if the applicant is allowed to backtrack at this stage. 39 For the application of game theoretic insights and analysis to the decision on whether to seek amnesty or leniency see Zane P. (2003): “The Price Fixer’s Dilemma: Applying Game Theory to the Decision of Whether to Plead Guilty to Antitrust Crimes”, 48 Antitrust Bulletin 1.


522

12/10/07

13:13

Page 522


3. Absence of a marker In many systems where getting in quickly is at a premium, the authority allows the applicant to reserve its place by putting down a marker. The Commission has so far resisted the idea of giving a “marker” in the US sense of keeping a place in the queue.40 The applicant for immunity in Europe is thus vulnerable to losing first place right up to the actual moment it receives the conditional amnesty grant. If the applicant is put on hold at first meeting without getting an immediate acknowledgement of receipt, as does sometimes happen,41 another company coming in later could still get first place. Moreover, even after accepting the application, and lacking the full powers of the criminal law enjoyed by the US authorities to keep applicants on track after the grant, it is only natural for officials to squeeze the maximum amount of useful information from applicants. Although time might be allowed to complete the application, hanging over the putatively “incomplete” application is the Sword of Damocles of a rejection decision denying immunity on the ground that the substantive conditions are not met, or of another company getting in ahead while the applicant is still trying to qualify. US practice is for the DoJ to give the applicant a “marker” while its application is being considered. In Europe no such security is available. Case handlers assessing the application will no doubt contact the applicant to fill in any obvious gaps, and the Commission will not consider another immunity application while it is still looking at the first. However, if another applicant gets in with an offer of the “smoking gun” while the officials are still mulling over the first application, there is a risk the first application could still be declared “incomplete”. Commission practice (not explicitly stated in the Guidelines) is that “the moment the second applicant submits evidence, the first applicant can no longer supplement its application with further evidence. Its application will be evaluated on the basis of the evidence it had submitted until the moment the second application was made.” (emphasis added) 42 The uncertainty this engenders no doubt provides 40 For an explanation of the US marker system, see Hammond S. (2001): “When Applying for Corporate Amnesty, How Do You Put a Price Tag on an Individual’s Freedom?”, speech, 8 March 2001; Hammond speeches, passim. The Canadian authorities have described in detail their marker system in Immunity Program: responses to Frequently Asked Questions, available at: http://www.competitionbureau.gc.ca/internet/index.cfm?itemID=1980&lg=e. 41 See supra note 30 and accompanying text. The “acknowledgement of receipt” should not be confused with a marker. 42 This potential trap for the unwary is not apparent in the Notice and is counter to the concept of certainty. Some officials however see it as a positive strength of the programme. According to Van Barlingen and Barennes, the “rule” encourages completeness, and they even warn of the dangers of breaking for lunch when presenting an application in case somebody else gets in and the first company’s application is “frozen” as it stands at that moment; see Van Barlingen B. and Barennes M. (2005): “The European Commission’s 2002 Leniency Notice in Practice”, 3 Competition Policy Newsletter 10.


12/10/07

13:13

Page 523

The Commission’s 2002 Leniency Notice 523 an added incentive to file the maximum of evidence immediately. The marker system as used in the US and now by an increasing number of national competition authorities is however more attuned to the underlying goals of a leniency policy.43 While the US Justice Department constantly refines the incentives in order to make it more “attractive” for applicants to come in, the Commission seems to think that promoting insecurity will do the trick. There have been hints at a review, but so far, no concrete proposals have been made to reform the currently less than satisfactory system.

4. No amnesty plus One of the most prolific generators of new cases for the US Justice Department has been the ingenious and pro-active enforcement strategy of “immunity plus”. A company losing the race in product A but coming in as the first with another cartel will not only earn immunity in product B but will be given a substantial additional reduction (the “plus”) in its negotiated fine in the first case.44 The OFT has also adopted this incentive in its own programme.45 Yet the Commission appears to have reservations about adopting this inducement for encouraging companies already under investigation to make a clean breast of all their hidden cartel activity and “put matters behind them”.46 Officials claim that there is no provision for it in the Notice, but should not ignore the possibilities afforded by the Fine Guidelines for “cooperation outside the leniency notice”.47

5. A transatlantic disconnect? The mixed signals apart, the major disadvantage of the Commission’s programme for a company involved in a global cartel and thus exposed in 43 See, e.g. Reynolds M. and Anderson D. (2006): “Immunity and Leniency in EU Cartel Cases: Current Issues”, 27 European Competition Law Review 82. 44 For an explanation of Amnesty Plus, see Spratling G. (1999): “Making Companies an Offer They Shouldn’t Refuse: The Antitrust Division’s Corporate Leniency Policy”, speech of 16 February, 1999. 45 OFT Guidance as to the Appropriate Amount of a Penalty, Ch 3.16–3.17, p 16. 46 For a detailed discussion see McElwee D. (2004): “Should the European Commission Adopt Amnesty Plus in Its Fight Against Hard-Core Cartels?”, 25 European Competition Law Review 558. 47 The Commission’s hesitations towards adopting a leniency plus programme while allowing a fifty per cent reduction under this head to a company that had allegedly imperilled the success of the dawn raids by informing its competitors that it had been granted immunity may confuse practitioners struggling to gauge what the Commission expects by way of “cooperation” See Raw Tobacco Italy, Commission Decision of 20 October 2005, available at: http://ec.europa.eu/comm/ competition/antitrust/cases/decisions/38281/en.pdf.


524

12/10/07

13:13

Page 524


multiple jurisdictions lies in the discontinuity between the optimal timing each system ordains for making the application. The incentives in the US are unambiguously all loaded on getting first through the door. In the EC the choices are not so clear. The applicant must decide between the conflicting dictates of completeness and priority. It is of course unlikely that any serious applicant would hold off approaching the US authorities until its file in Europe was complete enough notionally to satisfy the Commission. Since the risks in the US are simply too great, the optimal strategy there is to apply as soon as possible, taking advantage of a marker if need be. An 8(a) application in Europe however takes time to prepare, and given the lack of clear guidelines, different officials have different exigencies and expectations. Waiting in Europe until there is sufficient “evidence” to satisfy Commission officials runs counter to all best practice and leaves the potential applicant liable to be beaten at the post if news of the US procedure leaks out first. The applicant may not however be in a position to apply immediately to DG Comp even under 8(a) without giving chapter and verse, since Commission officials may insist on a “who, what, when and where” showing before they will even issue an acknowledgement. Coordinating the applications where there is multijurisdictional exposure thus poses tricky challenges for the most experienced of transatlantic practitioners.

6. Which courthouse door do I have to race to? If there is to be a race to the courthouse door, the contestants have to know which door to run to. In the United States, the federal system means that each state has its own antitrust laws, but there is only one enforcer of the Sherman Act. Violators of Section 1 know exactly where to go in order to self-confess. This was basically the situation in Europe before the Commission’s ambitious “Modernization” project. The Commission used to be the only authority entitled to enforce Article 81 of the Treaty, a monopoly reinforced by its exclusive competence to exempt agreements under Article 81(3), although of course no question arose of hard-core cartels getting clearance—in recent years at least.48 That has all changed. An express rationale of Modernization was the need to free the Commission for major tasks like hunting down Europe-wide cartels, but the clarity of that vision was not enhanced by the Commission ceding its position as the par excellence enforcer to a loose network consisting of itself and the 25 national authorities. Oddly, despite the priority supposedly 48 Although the CFI goes out of its way to say that there is no such thing as a per se violation of EC law and no that arrangement is inherently incapable of benefiting from Art 81(3) EC: see Matra-Hachette v Commission [1994] ECR II-595.


12/10/07

13:13

Page 525

The Commission’s 2002 Leniency Notice 525 accorded to the fight against cartels, no special regime was foreseen in the Regulation to allow the Commission to track down and prosecute hardcore offenders. At the core of the decentralisation vision is the system of parallel competence.49 Articles 81 and 82 EC are now enforced by the national competition authorities and courts, as well as by the Commission, alongside national competition laws in a Rubik’s Cube of competences, powers and jurisdictions. Instead of being the primary cartel hunter in Europe, the Commission is now just one of 26 agencies entitled to investigate and penalise price fixing activities under Article 81 EC. While the Commission’s powers to impose penalties are set out in Regulation 1/2003, those of the Member States are left to their respective national laws. There is no uniform regime either,50 so enforcement by the NCAs can be administrative or criminal and may even involve the imposition of custodial sentences for a breach of EC rules.51 The decision as to whether or not to adopt a leniency programme at all is also purely a matter for national law, and while most Member States have indeed adopted one, no two are exactly the same. The leniency landscape is thus a very varied one. There is no “single stop shop”, and just as getting in first in the US or Canada does not get any credit in Europe, an application to one “European” authority will not count as an application for leniency to another despite their all administering the same law.52 Any one of the NCA’s could in theory have power to open an investigation and offer immunity, assuming it has a leniency programme. From the general principles set out in the Network Notice it is likely that the Commission will be considered “best placed” to handle any suspected cartel with a more than “regional” dimension, but the principles are not binding rules, and while the Network is no doubt anxious to avoid unseemly “turf wars”, it is quite possible to imagine some confusion of competences at the early stages of a cartel being uncovered, especially if a race for 49 Gauer C. (2005): “Due Process in the Face of Divergent National procedures and Sanctions”, paper for Antitrust Reform in Europe Conference, 9–11 March 2005, also published in Lawyers’ Europe, Autumn 2005, p 14. The system of parallel competence derives from Regulation 1/2003 of 16 December 2002, Arts 3, 5–6. For a detailed explanation from the Commission viewpoint of the effect of modernisation on leniency, see Blake S. and Schnichels D. (2004): “Leniency Following Modernisation: Safeguarding Europe’s Leniency Programmes”, 2 Competition Policy Newsletter 7. For the practitioners’ view see e.g. Swaak C. and Mollica P. (2005): “Leniency Applicants Face to Modernisation of EC Competition Law”, European Competition Law Review 507; Joshua, The European Cartel Enforcement Regime PostModernisation: Is It Working?, Lawyers’ Europe, Autumn 2005, p. 24. For the principles of coordination between the NCAs and Commission, see Network Notice, infra note 52. 50 See Blake and Schnichels, supra note 49. 51 So far it seems that only Ireland has managed the not inconsiderable feat of turning a breach of EC competition rules into a criminal offence despite the CFI’s insistence that there are no per se violations of Art 81 EC. The UK did everything possible to distance its new “cartel offence” under s. 188 Enterprise Act 2002 from Article 81 and EC enforcement. 52 Commission Notice on Cooperation within the Network of Competition Authorities, OJ C 101 [2004], (“Network Notice”) Point 38.


526

12/10/07

13:13

Page 526


leniency develops.53 What to one low-level participant may be a purely national cartel could in fact be the local end of a much broader Europe-wide conspiracy.54 On the other hand, when time is of the essence, difficult decisions have to be taken in a short time frame and the cartel is likely to be global, applying for leniency to the Commission only may seem the top priority. Article 11(6) provides some—but not complete—reassurance: the initiation by the Commission of proceedings relieves the NCAs of competence to apply Article 81 (and it seems, if read with Article 3 of Regulation 1/2003, to apply national competition law as well, though this practical interpretation is contested by some). In a major case the Commission is likely to be the “best placed” authority, but with the huge backlog of cases threatening to overwhelm the Commission’s hard-pressed services, there is no guarantee that the Commission will in fact initiate proceedings. As officials helpfully point out, to apply for leniency only to the Commission and rely on Article 11(3) alone leaves the applicant open to the risk that another cartel member may in the meantime go in for leniency to the relevant NCA or NCAs. Moreover, there is no way to force the Commission to take a case or initiate proceedings under Art 11(6) even where there are more than three NCAs investigating: the case allocation mechanism does not create any rights for the parties to have the case dealt with by any particular authorities.55 Commission officials note cheerfully that under the general case allocation principles set out in the Network Notice, “[. . .] an applicant would be unlikely, at least in the vast majority of cases, to need to consider applying for leniency to more than a maximum of four authorities within the ECN, including the Commission.”56

Practitioners who advise today’s cost—and value conscious clients that it is necessary to apply not only to the Commission but also to another three or more national European jurisdictions are likely to find themselves calling a taxi for the client to go to a competitor firm. While in most large “global” cartel cases, it will be the Commission, not an ad hoc combination of Member 53 While the Commission is obliged by Art 10 of Reg 1/2003 to tell the NCAs of important evidence that it has obtained, there is no reciprocal obligation on the Member States. They are not even obliged to inform the Commission before they carry out an investigation, only “without delay after”. 54 Several Europe-wide cartels supplemented their “bosses” and “experts” meetings in Zurich with “local” meetings for each national market, in which the participant might well be an independent exclusive distributor that might have no knowledge of the broader picture: see e.g. Cartonboard, OJ L 243 [1994]. In the Pre-Insulated Pipe Cartel case, the CFI awarded an Italian company a reduction in fine in addition to one already given by the Commission (for the fact its participation was confined to the Italian market) since it was not shown to have been aware of the wider geographic extent of the conspiracy: Sigma Tecnologie v Commission [2002] ECR II-1845. 55 Swaak and Mollica, supra note 49. 56 Blake and Schnichels, supra note 49.


12/10/07

13:13

Page 527

The Commission’s 2002 Leniency Notice 527 States that will be handling the case, the associated risks and uncertainty are one more unnecessarily complicating factor in what ought to be a very obvious choice.

7. Oral leniency One area where Commission officials have proved sensitive to the interaction of different leniency policies is in the exposure to civil discovery in other jurisdictions—basically the United States—of any statements generated for the purposes of its leniency programme—whether by the first company in under 8(a) or the last to finish in the run off for a fine reduction in the “up to 20 per cent” band. To avoid generating written admissions that might be discoverable in US civil proceedings, so-called “paperless applications” have been accepted by the Commission from applicants virtually from the inception of the 2002 Notice.57 Lawyers are allowed to deliver into a tape recorder a narrative account of the facts required by the Commission to support the application, although of course they are not dispensed from the obligation to produce all the relevant documentary evidence in their possession.58 The exposure arises from the mathematical certainty that any announcement of an antitrust investigation in the US will trigger a welter of expensive and burdensome treble damages claims. Discovery in US litigation is an industry in itself and a huge burden on defendants.59 Having to make any discoverable “admission” in the course of its application could be a huge disincentive to coming in first. The amnesty procedure is therefore deliberately structured by the Justice Department in such a way that nothing is said in the negotiations that could be regarded as “evidence”. Of course, for the other cooperating companies, the price to pay for a reduced fine and less severe jail 57 Guersent O. (2002): “Cartels and Leniency”, speech to IBC Annual Conference, 21 November 2002. See also Reynolds and Anderson, supra note 42. 58 The US lawyers who first advocated the paperless procedure were primarily concerned with avoiding exposure for full immunity candidates. In the US, the successful amnesty applicants are spared having to enter a guilty plea that would be a serious damper on any negotiating strategy with the class action plaintiffs. It should also be noted that under the Antitrust Criminal Penalty Enhancement and Reform Act of 2004 the successful immunity applicant now only faces single damages. 59 Under the US Federal Rules of Civil Procedure, Rule 26, the district court may order discovery of any documents in the possession, custody or control of a party that relate to “any matter, not privileged, that is relevant to the claim or defence of any party”. Relevance is defined broadly, and discovery may be had of material that would not be admissible in evidence at trial but could lead to the discovery of admissible evidence. The court may limit discovery if the discovery sought is unduly burdensome, duplicative or is obtainable from some other source. For a discussion of the discoverability in US litigation of so-called “corporate leniency statements” see Levy N. and O’Donoghue R. (2004): “The EU Leniency Programme Comes of Age”, 27 World Competition 75, at pp 86–90.


528

12/10/07

13:13

Page 528


terms for executives is a guilty plea. As prima facie evidence of guilt in any civil action, this tends to limit the negotiations with plaintiffs to contesting causation and quantum. Accepting that the need to provide a potentially discoverable “corporate statement” in writing that had been imposed on immunity applicants by the old Notice was a major disincentive to coming forward in the EU, Commission officials welcomed the idea of the oral application. Of course US civil plaintiffs cannot obtain the statement by discovery from the Commission itself. Discovery requests or orders addressed to defendants routinely however demand copies of all documents submitted to “any antitrust authority, including the European Commission”, and while practice differs from circuit to circuit, that can include the corporate statement, basically providing plaintiffs with a confession of guilt. At first, American commentators (and enforcers) enthusiastically hailed the oral procedure as another laudable step towards transatlantic convergence. No official rules or procedures had been devised to deal with the paperless application, and it was only years later when Statements of Objections began to filter down that it was realized that the discovery risk had not been removed, only postponed. The Commission practice of annexing the transcript of the oral application to the Statement of Objections (as required by the rules on “access to file”) meant that the recipients would face demands for discovery at the very moment that their negotiations with the more serious opt-out claimants in the US were at their most critical juncture.60 Advocates of the paperless procedure have therefore urged the Commission to treat such applications not as proof of the infringement but purely as a roadmap pointing it in the right direction and thus enabling it to gather the hard evidence in successful dawn raids.61 By this simple means, the Commission would be relieved of the obligation to disclose the transcript on access to file as material evidence. Indeed, without actually explaining why, the Commission amended its access to file rules in 2005 so as to ensure that unattested transcripts were treated as no more than an “internal resource”, hence not evidence of the violation and thus not liable to be produced to defendants with the SO—and so far less prone to US discovery.62 Had the Commission confined the oral procedure to receiving “dawn raid sufficient” information this would have provided an elegant and effective solution to the discovery problem. The use of the paperless application had however gradually evolved from a practical means of obtaining an 8(a) roadmap to the standard method for taking testimony from the fine reduction applicants to serve as “added value evidence”. Indeed, in its first and only

60 See Joshua J. (2004): “Oral Statements in EC Competition Proceedings: A Due Process Short-Cut?”, 26 Competition Law Insight 1. 61 Mansfield P., Baer W. and Spratling G. (2004): “The EU Leniency Program: The Need for Procedural Reform”, ABA International Cartel Workshop, New York, 5–6 February 2004. 62 Commission Notice on the Rules for Access to the Commission File, OJ C 325 [2005].


12/10/07

13:13

Page 529

The Commission’s 2002 Leniency Notice 529 official proposal to amend the Leniency Notice,63 the Commission is now proposing to officialise the paperless method as the norm for adducing evidence proving the cartel, as the Commission is still required by law to do in the absence of any procedure for guilty pleas. Not only has an elaborate method been devised for turning such narrative into usable evidence; critical changes have also been made to the access to file process to obviate the need to deliver a copy of the transcript into the hands of any addressees of the Objections. Whatever inroads may have been made by the introduction in the Modernisation Regulation of an “interview” procedure, and an enhancement of the powers to demand oral explanations during a dawn raid,64 the Commission’s procedure is still an administrative process, and the prime method of proving infringements is by contemporaneous documentary evidence. There is no procedure for taking present sworn oral testimony from witnesses at a trial or hearing as in a judicial procedure. True, the Commission has sometimes used written statements as evidence65, even before the 2002 Leniency Notice, and the Court has recently endorsed the use as evidence against all the members of the alleged cartel of statements made during the procedure by an executive of one of them, provided such statements are (a) reliable and (b) corroborated.66 However in such cases, the primary evidence was still largely in the form of contemporaneous written documents and the statements were generated mainly in order to fill in gaps in the documentary evidence. Further, they were made not by lawyers seeking immunity or a fine reduction but by executives possessing actual knowledge of the events and in circumstances that guaranteed their reliability (declaration against interest). Using attorney narrative as “testimony” may have its administrative attractions as a substitute to having to prove the infringement by hard evidence, but the Commission’s latest proposal lays it open to criticism on the grounds of both denial of due process and ineffectiveness. The “due process” arguments need no elaborate rehearsal here, but in a nutshell relate both to the adequacy of the proof and the right to be heard.67 In the absence of a system of negotiated settlements, the Commission has still to prove the existence of the cartel by clear, precise and coherent evidence. There is no procedure for taking oral testimony, no penalty for giving false witness, and no real opportunity for cross-examination or confrontation. 63

See http://ec.europa.eu/comm/competition/antitrust/legislation/leniency2_en.pdf (undated). Reg 1/2003 Art 19 (Power to take Statements); Art 20(2)(e) and 23(1)(d) (Oral explanations on facts or documents). 65 As in Pioneer OJ L 60 [1980]. 66 JFE Engineering Corp v Commission [2004] ECR II-2501. The UK’s Competition Appeal Tribunal emphasized in JFB Sports v. OFT [2004] CAT 17 that contemporaneous documents, prepared at the time, are likely to be more credible than evidence given later. The Tribunal expressed caution about accepting at face value the testimony of individuals who might “have an axe to grind”. 67 See e.g. Joshua J., supra note 60. 64


530

12/10/07

13:13

Page 530


Unless all parties admit the violation, so that their status is limited to that of confessions, hearsay accounts given by lawyers fall short of any generally accepted evidential standards, especially if they are the only proof adduced by the Commission. The proposed procedure for access to file where oral statements are to be relied on as proof is similarly disturbing. If the proposal goes through, disclosure will be confined to a visual review of the transcripts on the Commission’s premises, with no possibility of taking copies. The limitations of such a procedure in terms of permitting detailed analysis of the evidence do not need to be spelled out. In a complex case the statements are likely to be of a complexity depth and level of detail that could require weeks of thorough examination. “Access to file” is not a pro forma gesture but is designed to allow the addressees of the statement of objections to test and challenge the evidence adduced against them in a meaningful way.68 It is by no means entirely clear that the oral statements will prove immune against determined discovery demands from US class action lawyers. It is not the mode of delivery that protects from discovery statements by US lawyers making an application to the Antitrust Division. Commission officials contend that if it works in the US, it should also do so in Europe. The conceptual differences between the two “oral” procedures are however controlling. In the US, the process is deliberately structured not only so as to avoid giving evidence but to avoid even giving any impression that what is being provided is “evidence”. Clients are never present, the lawyers are not witnesses as to fact, and in any event are careful to make only a “proffer”, an outline of what story the witnesses would tell if they had to give evidence or were to be interviewed by the government. In total contrast, the Commission sees the whole purpose of the oral application as the delivery of “evidence” that will be used as such against the other cartel participants. In its view, the company is not only confessing but also “testifying” through its lawyers. The process is deliberately structured not to avoid giving evidence but so as to constitute evidence. In such circumstances, some US district judges may be less than impressed by what could be presented by plaintiffs as a device to deprive the court of relevant testimony. After all, from one perspective, the primary facts in a conspiracy case are within the exclusive knowledge of the participants, who have deliberately concealed their agreement from their victims. “Work Product” privilege is not inviolate: it may be overturned on a showing of “particular need”.69 Thus, absent any other means of proving the existence of the conspiracy, plaintiffs 68 See e.g. Solvay v. Commission (Soda Ash) [1995] ECR II-1175 at point 59: “ The proper observance of that general principle (the rights of defence) requires that the undertaking concerned be afforded the opportunity during the administrative procedure to make known its views on the truth and relevance of the facts, charge and circumstances relied on by the Commission.” 69 Hickman v. Taylor, 329 US 495 (1947), codified in Rule 26 of the Federal Rules of Civil Procedure.


12/10/07

13:13

Page 531

The Commission’s 2002 Leniency Notice 531 could well demand production of the script of any oral submission, though whether they will succeed is of course another matter. The Commission justifies the new proposal by reference to the “no worse off ” principle: companies cooperating with the Commission should not be placed in a worse position than if they had not cooperated. As embodied in Point 23 of the present Notice, the principle is an eminently sensible trade-off to further the gathering of new evidence. But even if it works, the protection from US discovery is achieved at the expense of defence rights and the integrity of the procedure. Indeed, lowering the standards of proof and fairness may have the opposite effect from that intended and could in fact encourage contestation. If there is a good chance that the only evidence against them is going to be lawyer recitation, late finishers might be tempted to gamble on staying out entirely. To be sure, making cooperating companies provide a potentially discoverable roadmap would be undesirable, but instead of tinkering with the Notice, the whole discoverability issue could be resolved by allowing guilty pleas that did not generate testimony. The imbroglio could be avoided by an unequivocal declaration from the Commission that the oral procedure will in future be used merely to give a roadmap or a “proffer” and not as evidence, as suggested by US practitioners. In the longer term, the issue of its application to fine reduction applicants needs to be considered as part of a wider review, and indeed the present requirement that they have to provide “added value” evidence could be dispensed with if there were to be a system of negotiated settlements. If it will be necessary from a policy standpoint to protect applications already made from US discovery, legitimate expectations having been raised by the Commission’s encouraging oral applications from fine reduction applicants, a more fruitful road for opposing plaintiff demands could lie in the assertion by the Commission of comity arguments70 rather than leaving the defendant to rely on the technical accident of the mode of delivery.

8. The also-rans There are of course no “guilty pleas” in the Commission’s procedure that would allow finality in the shape of negotiated settlements and waiver of any right to challenge the fine on appeal. The Commission’s equivalent of the US process where the also-rans in the race to get in first have to scramble to make the best deal they can, with both 70 Arguments that the discovery sought by applicants under 28 USC § 1782 (judicial assistance) would circumvent and undermine the Commission’s procedures were favourably received by the district courts in e.g. In re Application of Microsoft, MBD N° 06-10061-MLW, (Mass.) 17 April 2006; in re Application of Microsoft, Case N° 06-80038 JF (PVT), (N.D. Cal.).


532

12/10/07

13:13

Page 532


positive and negative incentives to come in quickly is the system of “percentage bands”.71 Simply not contesting is not enough any more to get a reduction and the companies looking for a fine reduction have to earn it by providing “value added” evidence. The logic behind the “run-off” is simple enough: the longer companies hold off, (a) the more onerous is likely to be the burden of adducing evidence that adds something to what the earlier finishers have already handed in; and (b) the lower the band in which they are placed and the less the reduction. To secure even “top of the band” the applicant will have to provide a lot more than companies coming in before it, with the added disadvantage that it does not know what they have given the Commission. The value of the evidence counts for less than the finishing order. Again for the Commission it is the uncertainty, rather than transparency, which drives the process. One can question whether the current Notice has the incentives right. If the best the third or fourth company in can hope for, no matter how probative their cooperation, is a reduction of at best 30 and 20 percent respectively off an unspecified and indeterminate fine, they might well be tempted to take their chances and resist the SO line by line, especially now that the Commission is increasingly relying as “proof” of the violation on what might be no more than contestable lawyer hearsay.

9. The statistics speak Ultimately the success and effectiveness of the Leniency Notice has to be judged against its results on the ground. Nobody can gainsay that it has attracted many takers: since its entry into force, as of September 2005, 80 applications for immunity had been received, and 79 for a reduction in fine.72 The close parity between the two figures is however a clue to the policy’s shortcomings. With most cartels having more than two members, there should surely have been more takers for a fine reduction, but dawn raids have not exactly been followed by a race for the consolation prizes.73 By the end of September 2005, only eighteen applicants had even been informed of the Commission’s intention to place them in a particular reduction band, a step that indicates the issuance of a statement of objections. As a cartel on average has five or six members this epitomises the hold up in the case pipeline. Either the cases are stalled, or dawn raids are not being mounted and the other members of the putative cartel are staying out. 71

2002 Leniency Notice, Part B, points 20–27. Van Barlingen and Barennes, supra note 42. 73 Even if, as is further explained, conditional immunity was refused in 19 cases, leaving some 60 valid applications, this still shows that other companies besides the applicant are in no great hurry to come in, while raising further questions as to the difference in approach between the Commission and the DoJ: see Hammond remarks quoted supra note 13. 72


12/10/07

13:13

Page 533

The Commission’s 2002 Leniency Notice 533 One other telling statistic is that over four years after the entry into force, the sum total of cases actually disposed of by a Commission decision imposing fines on the cartel is just four.74 With no measures yet to streamline the Commission’s lengthy procedures in order to ensure timely dispatch of cases, a huge backlog has built up, and even if the Commission manages by focusing resources in the new Directorate to double to ten or so the cartel decisions that emerge annually, if immunity applications continue to come in at the rate of two or so per month, by the year 2012 the stockpile of unresolved cases could have reached the hundreds. Even in a case that promises to be uncontested, the Statement of Objections is still likely to run to over a hundred pages. There is no short form procedure and an oral hearing, with full “access to file” is inevitably demanded. In the absence of any system for negotiated penalties, the standard tactic of the alsorans is to talk up their “cooperation”, pocket their full fine reduction, and then all appeal to the Court of First Instance on the ground that the fine was too high, with every prospect of receiving a further cut from the Court that makes the investment worthwhile. It clearly pays for a firm to cooperate with the Commission and then challenge the decision in the CFI.75 The absence of a mechanism for the prompt settlement of cases is destabilising for several reasons. First, from the enforcer’s angle, the reward for letting one conspirator off free is normally speedier resolution of investigations by (a) obviating the need for lengthy, complex and possibly unsuccessful investigations; (b) shortcutting the prosecution process by securing admissions, confessions and “state’s witnesses”. The inability of the programme to achieve this welfare saving means that the efficiencies expected from “Modernization” are not being achieved. Secondly, as regards the incentives of the programme for making the decision to self-report an easy one, the long period in limbo could make companies question why they ever bothered to go in anyway.76 Their motivation in coming forward is often a desire to put the past behind them and “move on.” 74 Raw Tobacco Italy, Decision COMP/C. 38.281/B.2 of 20 October 2005, OJ C 60 [2006]; Rubber Chemicals, Decision of 21 December 2005, IP/05/1656; Bleaching Products, Decision of 3 May 2006, IP/06/560; Acrylics, Decision of 31 May 2006, IP/06/698. 75 According to Case Associates, Penalties for Price Fixers, Casenote May 2006, out of 39 cartels dealt with under the Commission’s two leniency programmes, fines were reduced or cancelled by the Commission in 35 of them; fines were appealed to the CFI in 34 out of the 39 decisions (87%); and in the cases decided by the Court, the CFI reduced the fine imposed on a firm litigating the decision by a further 18% on average. 76 The refusal of immunity in 19 cases, in thirteen of which “there was considerable doubt as to whether the conditions of the Notice were met and the case was not suitable for further investigation”, does not indicate a generous approach is being taken towards applicants either. This is in marked contrast to the US where the DoJ leans in favour of making the leniency grant (see supra note 13). The Commission’s use of “no action letters” as an administrative device to file leniency applications without taking a decision on the merits where it does not intend to carry out investigations does not provide certainty. For “No Action letters” see Van Barlingen and Barennes, supra note 42.


534

12/10/07

13:13

Page 534


Absent a rapid and visible resolution, the attractiveness of the leniency game will diminish. General Counsel will have a harder job to sell the leniency option to hard-bitten CEOs when cartel activity is uncovered by a routine audit. A further adverse effect on the policy is the dilution of deterrence. Leniency programmes destabilize cartels by ramming home the message that if the companies do not come in, they run an unacceptably high risk of getting caught. It enhances deterrence to know that even without the wild card of leniency, the investigating authority has the will and the means to track down cartels aliter. With resources so thinly stretched, despite a huge expansion in the numbers of officials investigating cartels, officials may be tempted simply to wait for leniency applications to come in rather than use “gumshoe methods” to go out to look for cartels. The trouble is that without the real risk of cartels being discovered and punished other than by a defecting informer, it is only a matter of time before cartel members work out that if internal solidarity prevails, their illicit activities could continue unhindered by the risk of discovery. Of course, such a gamble will not be worth taking in a case where individuals are at risk of losing their own liberty, a factor that could further widen the distance that is starting to emerge between the best strategies of cartelists in “global” and “European-only” cartels.

10. Proof of the pudding: decisions on the ground Statistics apart, the paucity of actual decisions (just four at the date of writing, and just one published so far) provide little indication as to how the Leniency notice has operated on the ground and affected the enforcement dynamic. The only published decision, Raw Tobacco Italy,77 shows that the programme hardly got off to an inauspicious start; the successful applicant, Deltafina, had its immunity taken away in the final decision for apparently freely volunteering to other cartel participants at the next industry association meeting that it had applied for and received conditional amnesty, prompting them both to rush in to the Commission for leniency the afternoon of the very same day within three hours of each other. It might be observed that far from complicating the Commission’s life, the disclosure by Deltafina had spared it the risk of having to find the evidence in the dawn raids, but the Commission took a dim view and in the end held that Deltafina had not only forfeited its immunity for failure to cooperate but also any claim under the Notice to a reduction in fines. In vain the applicant argued that the Notice did not in terms prohibit communicating the fact of its amnesty grant to the other players in the cartel. If the Commission meant to send a warning to compan77

See supra note 47.


12/10/07

13:13

Page 535

The Commission’s 2002 Leniency Notice 535 ies not to manipulate the leniency programme, the clarity of its message was somewhat blurred by its nevertheless awarding Deltafina a 50 per cent reduction as a reward for cooperation outside the Leniency Notice reflecting its “actual contribution to the establishment of the infringement”. Deltafina, along with all the other actors in the drama, has appealed the decision to the CFI. In terms of efficiency, the Notice can hardly be judged an enormous time-saver when despite its being purely a “national” case largely free of the usual translation problems, the facts of which were not seriously disputed, the decision runs to some 112 pages, the procedure took three years and nine months from application to decision, and with little prospect of either side being satisfied with the outcome in the CFI on the point of law central to the refusal of immunity, the vista opens ahead of another seven years or more elapsing before there is finality in the case.

E. An integrated solution What could be done to improve the Commission’s overburdened leniency programme? Piecemeal tinkering with the Notice or “administrative solutions” to dispose of applicants without a formal decision will not address the mounting problems. Only a bold and “holistic” solution will do the trick, encompassing the whole scope of the enforcement process. In the spirit of the generous invitation issued by Mrs Kroes and Philip Lowe a year ago,78 the following proposals are put forward as a serious contribution to the debate:

1. Re-establish the Commission as the primary hunter-down and prosecutor of cartels with a “community dimension” Although the uncertainty is probably manageable for those who are not riskadverse, the possibility that even in a global cartel case, a member state may open its own procedure after an immunity application has been made to the Commission is one more factor that complicates the taking of what should be an easy decision. Modernization was not an end in itself, so the argument that a “single stop shop” is somehow counter to its underlying “philosophy” is misplaced.79 Ideology should not stand in the way of pragmatism. If 78 Kroes N. (2005): “Taking Competition Seriously—Antitrust Reform in Europe”, and Philip Lowe, Chairman’s Closing Remarks, both at IBA/Commission Conference, Antitrust Reform in Europe: A Year in Practice, Brussels 10–11 March 2005. 79 See Cooperation within the European Competition Network, Competition Law Forum, Modernisation Expert Group, (2004) EBLR, p 263. Commission officials insist that the position


536

12/10/07

13:13

Page 536


effectiveness is to be enhanced, a clearer statement is needed that the Commission is the primary anti-cartel enforcement agency in Europe. A single application for immunity to the Commission should be sufficient to bring down the bar, at least temporarily, and ensure that other applicants do not gain an advantage by getting in first in a member state. While some member states have introduced provision for filing a pro forma application locally when the applicant is also proposing to inform the Commission,80 the device is a useful stopgap and not a permanent fix. It may have taken twenty years to persuade the member states to adopt a Merger Control Regulation, but the model is established, and the contrary case is unarguable on any rational basis.

2. Drop the requirement for an evidential threshold Any evidential threshold at the stage of the first approach negates the “get in first” imperative. Once there is a concrete basis for suspecting that a cartel exists and the decision has been taken at corporate level to cooperate, there should be no further obstacle to an application. All that should be required for a provisional grant is frank and candid reporting of the violation, although thereafter the applicant would of course have to provide its full cooperation as at present. Immunity should be automatic where the company is the first in to report the cartel and the Commission does not yet have information from any other source that would give it concrete grounds to suspect the existence of the cartel. Where the Commission has already begun an investigation (or is in a position to do so, which would require a clear definition of the standards necessary to open an enquiry) the grant of full immunity should be discretionary for the first company in to qualify, provided the Commission does not yet have sufficient evidence in its possession to sustain a decision making a finding of the existence of the cartel; Alternatively, instead of being discretionary, full immunity could be made automatic in limited cases even after the investigation had started or was about to begin, but the circumstances would need careful definition. Clearly, the hurdle should be higher than an application made before any investigation, but success should not depend in any event on providing all the evidence needed to prove the cartel on its own, a standard almost impossible to achieve as regards leniency is at least no worse than it was before Modernization, but they overlook the fact that few if any competition authorities (a) had the power to enforce Article 81 EC (as they all do now); or (b) had a leniency programme anyway. As Swaak and Mollica (supra note 49) point out, “for leniency applicants Modernisation means above all ‘complication’ ”. 80 See e.g. Bundeskartellamt, Notice No 9/2006: Leniency Programme of 7 March 2006. After the marker is placed the applicant has eight weeks to make its application.


12/10/07

13:13

Page 537

The Commission’s 2002 Leniency Notice 537 even if flexibly interpreted: substantially advancing the Commission investigation should suffice.

3. Increase certainty by allowing a marker Provision should be made for putting down a marker to reserve a place. The grant of a marker could be for a limited time, appropriate to the circumstances, and could also be extended. Where the applicant is able to put down a “smoking gun” document as a token of good faith, the time available to redeem the marker could possibly be more generous. It should be clear however that “making good” on the marker should not involve an immediate evidential threshold: a proffer outlining the proposed cooperation allowing the Commission to evaluate it should suffice. (This would presuppose that attorney narrative is used not as proof but simply as a roadmap.) If the evidential threshold is dispensed with, the provisional grant could be given within a relatively short time, an advance that could only enhance the attractiveness of the programme.

4. Enhance the incentives by a formal Leniency Plus system In order to stimulate the dynamic of the leniency policy in uncovering cartels, companies carrying out an internal review of their exposure as a result of their initially getting caught should have the added incentive of a substantial reduction of the fine in the first case.

5. Introduce a system of negotiated settlements The corollary of favourable treatment under any leniency policy is finality. This applies also to any benefit short of full amnesty that is accorded for cooperation. The present situation by which fines are reduced—sometimes very substantially—in recognition of cooperation is an absurdity if the beneficiaries are able to appeal the decision as a matter of course, with a more than even chance of gaining a further discount. There is no good reason in principle why negotiated settlements should be contrary to the interests of justice. For a properly advised company, represented by experienced counsel, to reach a considered and binding settlement with the authorities is not an affront to due process, especially where the procedure is administrative and no criminal penalties on individuals are involved.


538

12/10/07

13:13

Page 538


There would be a huge saving in time and resources would be freed to deal with the detection of cartels that do not take advantage of the leniency programme. Removing the anomaly of companies being able to apply to the CFI and have their reduced fines systematically reduced still further would also allow the Court to concentrate on more meritorious matters. The procedure could be formalized in a Commission implementing Regulation adopted under Article 33 of Regulation 1/2003. To the extent that Article 31 (“The Court of Justice shall have unlimited jurisdiction to review decisions whereby the Commission has fixed a fine or periodic penalty payment”) may be considered an impediment, and it is by no means certain that an agreement to exclude the court’s review is contrary to public policy, the conclusion of such an agreement in proper conditions should impose a heavy burden of persuasion on any applicant that wishes to resile from it.81

6. Introduce predictable Fine Guidelines “Guidelines” as applied to the Commission’s policy is something of a misnomer for US lawyers used to the comprehensive set of Organizational Sentencing Guidelines issued by the Sentencing Commission under the Sentencing Reform Act 1984.82 To be sure, instead of the previous largely intuitive system of basing fines on a percentage of turnover in the product, the Commission’s methodology is now standardised—a “start point”, possibly weighted for “deterrence”, increases for duration of basically 10 per cent per year to give the basic fine, and percentage uplifts and reductions for aggravating and mitigating factors—but crucially the choice of the start point, “x”, is what the use of algebraic notation suggests: it can be anything.83 Commissioner Kroes has promised to look again at the tendency of the present Guidelines to hit smaller one-product companies too harshly in comparison to multinational groups, but has expressed satisfaction with the overall scheme.84 Dismissing the criticism of insufficient transparency that have been levelled at the Guidelines, the Commissioner endorsed the standard justification wheeled out by officials in the past, who argued that potential infringers should not be allowed to make a cynical calculation of the cost/benefit ratio of a cartel in advance. This was an argument for a higher level of fines, not for a lottery. If 81 According to the CFI in Tokai Carbon v. Commission [2004] ECR II-1181, an undertaking is (only) stopped in principle from disputing facts where it has “clearly and unequivocally” accepted those facts in the administrative procedure. 82 US Sentencing Commission Guidelines Manual, Chapter 8, Sentencing of Organizations, pp 423 et seq. The Guidelines were amended in May 2004 so as to specify the minimum requirements that a compliance programme had to meet in order to qualify for a reduction in fine. 83 Joshua J. and Camesasca P. (2004): “EC Fining Policy Against Cartels After the Lysine Rulings: The Subtle Secrets of X”, European Antitrust Review 5. 84 Kroes N., “First Hundred Days” Speech, supra note 3.


12/10/07

13:13

Page 539

The Commission’s 2002 Leniency Notice 539 the fines are always set sufficiently high, predictability will enhance, not dilute the deterrent effect. If managers know that the level of fines discounted for the risk of detection will always be unacceptably high, the cost-benefit analysis will be persuasive. It does nothing to bring home the message that cartels do not pay when differences of something like three times or more are found in the start points (weighted for market value and market shares) adopted in not dissimilar cases,85 and purely “national” cartels have been punished with much larger fines than those imposed in pan-European cartels many times their size.86 It is also odd that if the fine is supposedly related to “impact”, a cartel that runs for ten years will only be fined twice one that stopped after a few months. In the US, basing the corporate fines on “volume of affected commerce” has not exactly handicapped the DoJ in its purge of cartels. Indeed, the Justice Department has skilfully exploited the provisions of the Guidelines to persuade corporate defendants in antitrust cases to plead guilty and accept huge “agreed” fines rather than take the chance of going to trial and risking an even higher penalty. Few defendants in the Federal criminal justice system take the chance of fighting the case when a reasonable deal is on the table. A staggering 97 per cent of all Federal prosecutions are disposed of in this way, in marked contrast to EC competition enforcement system where something like the same percentage take the reduction offered by the Leniency Notice and appeal anyway. True, the Commission does not have the big stick of being able to charge and jail individuals, but the certainty of very high fines and a high chance of detection will have a salubrious effect on antitrust miscreants in cartel-prone Europe. Predictable and near-mandatory fines but with a “range” of discretion for the Commission to choose a multiplier will also have to be introduced in case negotiated settlements ever become a reality. Any fine negotiation in the US takes place “in the shadow of litigation”, so the reduction agreed for cooperation and mitigating circumstances has a principled basis. The same should apply in the EU; without guidelines that are uniform and precise to provide the base level for negotiation, the spectacle of haggling over fines like carpets in a bazaar will not be edifying.87 85 Weighted pro rata, the start point fines in Choline Chloride ([2005] OJ L190/22) appear to be three times heavier than those imposed in MCAA six weeks later (Commission Decision of 19 January 2006 in Case COMP/E-1/37.773, not yet published). 86 Compare the fines in Raw Tobacco Italy (supra note 47), with those in Needles, Decision of 26 October 2004, Case COMP/F-1/38.338, available at http://ec.europa.eu/comm/competition/ antitrust/cases/decisions/38338/en.pdf. 87 Any Guidelines could do worse than emulate the US model of:

(a) working out the “culpability score” for the defendant based on the points system set out in a check list; (b) applying the resulting score to a table setting out for each score (on a scale from 0 to 10 or more) the maximum and minimum multipliers to be applied to an objectively assessed twenty per cent of volume of affected commerce to give a fining “range” (For a culpability score of 5, the multiplier is x 1 to x 2; for a “bad” score of ten, the sentencer can potentially double or quadruple the base fine; cont./


540

12/10/07

13:13

Page 540


In settlement negotiations, the Department has enough flexibility under the Guidelines ultimately to come up with a “reasonable” but still principled fine proposal that many corporate defendants will accept rather than run the risk of a tougher “top of the range” Guidelines sentence if they fought the case and lost. The scheme may sound complicated to administer, but at its core is not the “x” start point that the Commission plucks from the air but an objective volume of commerce.

F. Final thoughts The present leniency policy, while making gestures in the direction of “certainty”, in fact relies for its success on the very instability and insecurity of the qualification process. Game theory should stop once the decision has been taken to come in. Instead of blurring still further the fuzzy logic of its programme, the Commission should take to heart the clear linkage and interaction between the pre-requisites for a successful leniency programme: (1) laws that provide the threat of severe sanctions for those who participate in hardcore cartel activity and fail to self-report; (2) the clear perception by organisations of a high risk of detection if they fail to come in; and (3) transparency and predictability to the greatest extent possible throughout the whole programme. Although common sense dictates that whatever the uncertainties of the EC process, applicants involved in a global cartel must find a way to come in to the Commission at the same time as they apply in the US, it is vitally important for the credibility of the programme that its incentives and disincentives lead to the same outcome even where there is no US exposure. If companies begin to think it a viable option to stay out and gamble that the Commission will not be able to build a case against it, the status of the Commission as an antitrust agency will diminish. In order to enjoy independent credibility, (a) not only must the penalties be severe, but also (b) the chances of detection must be heightened so as to make the risk unacceptable. To avoid any scenario in which the cartels could work out that absent a defector, the risks of being caught are negligible, the Commission must ensure that it does not rely solely on leniency to bring in the cases. With its growing focus on prosecuting individuals, the US is increasingly looking to those Member States that have criminalized hard core cartels as its allies of choice in the “war on cartels”. If the Commission is unable to demon-

(c) setting the appropriate fine inside the range based on specified factors relative to the offender and the offence; (d) allowing upward or downward “departures” for aggravating or mitigating factors that are not adequately taken account of in the Guidelines.


12/10/07

13:13

Page 541

The Commission’s 2002 Leniency Notice 541 strate its “stand alone” credentials as a global enforcer, it could eventually even find itself on the sidelines.88 Mrs Kroes’s visionary First Hundred Days speech identified the problems but the Commission has still to come up with proposals for the cure. Rather than partial or temporary expedients, drastic surgery is called for, but it must be just one component in a holistic solution that will restore the Commission to its former position as the primary hunter of cartels in Europe. Modernization is here to stay, but its underlying rationale was the creation of an effective enforcement regime in a Community of 25 (now 27) members, not an end in itself. The Commission has many responsibilities in the competition field, and enforcing Article 81 EC against hard core cartels is just one of them, but it may ultimately have to decide whether it aspires to be an administrative competition regulator or an antitrust enforcement agency and identify its priorities accordingly. Otherwise, “Red Tape in the Qualification Process will Strangle its Amnesty Programme”.

88 Joshua J. (2004): “Brave New World: A North Atlantic Triple Entente Against Cartels”, Competition Law Insight, March 2004.


12/10/07

13:13

Page 542

(Za) Ehlermann 07 Bloom

12/10/07

13:13

Page 543

III Margaret Bloom1 Despite Its Great Success, the EC Leniency Programme Faces Great Challenges

A. Introduction Most of the EU Member States now have leniency programmes—some introduced only recently. A growing number of other countries also have leniency programmes following the great success of the US regime. The significant growth in leniency regimes reflects their considerable benefits as the most effective tool in the fight against cartels. However, there are a number of important issues in devising and operating effective programmes. This paper first discusses three issues primarily in an EU context: an evaluation of the European Commission program (Section B), harmonization—or even a onestop shop—for leniency within the EU (Section C), and the interface of these programmes in the EU with private and criminal enforcement (Section D). It then discusses some issues in a wider international context: international implications of leniency regimes (Section E), financial incentives for whistleblowers more generally (Section F) and plea bargaining (Section G). The term “leniency programme” is used in the paper to describe all programmes which offer either full (100%) immunity—that is, amnesty—or a significant reduction in penalties that would otherwise have been imposed on a participant in a cartel, in exchange for the freely volunteered disclosure of information on the cartel which satisfies specific criteria prior to or during the investigative stage of the case, and does not cover reductions in the penalty granted for other reasons.

1 Visiting Professor, King’s College London and Senior Consultant, Freshfields Bruckhaus Deringer. The views expressed in this paper are personal and do not necessarily reflect those of my colleagues. I am grateful for comments from Kris Dekeyser, Jon Lawrence, Vanessa Turner, Richard Whish and Wouter Wils.


544

12/10/07

13:13

Page 544


B. The 2002 Leniency Notice:2 an evaluation The main questions in evaluating a leniency programme are whether it is a well-devised programme and whether it is effective in uncovering cartels. Hence these two issues are discussed in order to evaluate the EC leniency programme. The first question can be considered specifically for the 2002 Leniency Notice. However, the second one needs to be answered in the context of the whole EC programme including both the 1996 and 2002 Notices as few infringement decisions have so far been made under the 2002 Notice.

1. Is the 2002 Notice well devised? The key elements of a successful leniency programme are identified in the 2003 OECD Competition Committee report3 on cartels and leniency (emphasis added): The US and EC experiences are instructive as to the necessary elements of a successful leniency programme. They could be summarized as follows: —Complete immunity from sanctions should be awarded to the first applicant. This maximizes the reward for co-operation. —Only the first to apply should receive complete immunity, and if the programme is extended to subsequent applicants, the gap in the rewards should be substantial. This maximizes the incentive to be the first to defect, thus destabilizing the cartel. If the returns to the second applicant approximate those that would accrue to the first then the result may be that no one would apply. —The programme should have maximum transparency and certainty. Wouldbe applicants should be able to predict as accurately as possible what the outcome of their application will be. —The programme should be available in circumstances in which the competition agency has already begun an investigation. If the members are aware of an investigation and of the possibility that one of them could benefit from leniency, the stability of their agreement is likely to be severely eroded.

2 European Commission (2002): Notice on immunity from fines and reduction of fines in cartel cases, OJ C 45 [2002]. 3 OECD (2003): Hard Core Cartels—Recent Progress and Challenges Ahead, p. 22, available from http://www1.oecd.org/publications/e-book/2403011E.PDF. See also International Competition Network (2006): “Drafting and Implementing an Effective Leniency Program”, in Anti-Cartel Enforcement Manual, available from http://www.internationalcompetitionnetwork. org/capetown2006/FINALFormattedChapter2-modres.pdf.


12/10/07

13:13

Page 545

The EC Leniency Program Faces Great Challenges 545 —The competition agency should accord confidentiality to leniency applications and the information resulting therefrom to the maximum extent possible. In addition to these elements, if a leniency program is to be effective at uncovering and deterring cartels the competition authority needs effective sanctions for cartels and effective powers of investigation. If sanctions are weak and there is a very low probability of being uncovered, even a very well designed program is unlikely to be effective. Cartel members will not fear that other members will apply for leniency or that the cartel will be uncovered some other way. Hence, there will be little incentive to apply for leniency. The European Commission has increased the size of its fines for cartels significantly over the past decade. This should have increased the incentive to apply for leniency. Does the 2002 Leniency Notice contain all the above elements? Yes—but certainty and transparency could be improved further as discussed below. In contrast, the 1996 Leniency Notice4 did not satisfy all these conditions. Lawyers and some competition authorities had raised concerns over its lack of certainty and transparency. In particular, there was no certainty that the first applicant would obtain complete immunity. This was in marked contrast to the very successful program run by the US Department of Justice. In practice, under the 1996 Notice immunity was granted in significantly less than half the cases. In addition, once the Commission had begun an investigation, it was not possible to obtain immunity—the maximum leniency possible was 75%. The 2002 Notice was published on 19 February, 2002. The key aspects of the new policy and the main changes over the 1996 policy were summarized in the 2002 Competition Report (emphasis added): “The new leniency notice improves on the 1996 leniency notice in several respects. It incorporates a number of changes designed to make it more attractive for companies to come forward, and thereby to make the Commission’s fight against cartels even more effective. Key elements of the new notice are: first, full immunity from fines is available to the first company that comes forward; second, the evidence supplied should be enough for the Commission to order an inspection; third, the Commission allows hypothetical applications, where actual evidence only needs to be supplied in a second stage; fourth, taking decisions granting conditional immunity within a matter of weeks provides up-front legal certainty to the applicant; fifth, even after the Commission has undertaken an inspection, immunity is still available under certain circumstances; sixth, if immunity has already been granted, or the Commission already has enough evidence to find an infringement, reductions of fines of up to 50 % remain possible for companies that provide significant added value to the Commission’s case; last but not least, with a view to introducing more 4 European Commission (1996): Notice on the non-imposition or reduction of fines in cartel cases, OJ C 207 [1996].


546

12/10/07

13:13

Page 546


certainty with respect to reductions, the Commission takes a preliminary decision on the band of reduction to be applied as soon as possible following the application.”5

The 1996 leniency notice had required a company to provide “decisive evidence” and excluded from immunity companies that had acted as an instigator of or played a determining role in a cartel. Both of these left room for interpretation. There was uncertainty as to what decisive information was and what it meant to be an instigator or play a determining role. The new Notice resolves these concerns. A year after the 2002 Notice had been published, Philip Lowe, Director General of DG Competition, confirmed that the new policy had been a “tremendous success”. He considered the main reasons for that success appeared “to be twofold: it offers full immunity to the first company providing the requisite evidence and it offers more legal security, in the sense that companies will know more rapidly what treatment will be accorded to them.”6 There are a number of ways in which certainty and transparency can be improved further over the 2002 Notice. One is currently being addressed by the Commission with the aim of ensuring that leniency applicants are not prejudiced in civil damages litigation by reason of their application. Private plaintiff lawyers are increasingly seeking to obtain access to corporate leniency statements, in particular, through US discovery provisions. These corporate leniency statements provide plaintiffs with a “road map” for a treble damages claim against the leniency applicant. The Commission has, until fairly recently, required a written statement. In contrast, the entire US amnesty process, other than the conditional leniency agreement, has been oral for over a decade, not least to reduce discovery. The scale of US treble damages—for example, totalling billions of dollars in the case of the vitamins cartels—is such that potential applicants for leniency for international cartels might decide not to come forward if the risk of very large private damages is increased by the discovery of corporate leniency statements. Some leniency applicants settle with private litigants before there is any possibility that they will be able to request or use the evidence given under leniency. But this may only be practical in some cases. The changes proposed by the Commission to reduce the risk of discovery are considered further in Section E below which looks at international implications of leniency programs. Some practitioners have commented that the Commission’s proposed reforms do not go far enough. In particular, they do not take account of the possible use of corporate leniency statements in litigation in Europe despite the fact that the European Commission is seeking to stimulate private actions. 5 European Commission (2002) XXXIInd Report on Competition Policy, p. 29, available from http://europa.eu.int/comm/competition/annual_reports/2002/en.pdf. 6 Lowe P. (2003): “What’s the Future for Cartel Enforcement?”, DG Competition, Brussels, 11 February 2003, text available from http://europa.eu.int/comm/competition/speeches/text/ sp2003_044_en.pdf.


12/10/07

13:13

Page 547

The EC Leniency Program Faces Great Challenges 547 As for other reforms to improve legal certainty and transparency there is a need to address the following issues: —What standard in terms of evidence, etc., does the Commission apply for “significant added value” which an undertaking must satisfy where it is not the first applicant to come forward? Some practitioners consider there has been a considerable variation in this standard. —Some practitioners also consider that the level of transparency of fines is not all it might be. 50% reduction of what? Sceptics could allege that the Commission doubles the number to arrive at the result it wants to achieve after the deduction. It is important to be as transparent as possible in order to minimize the scope for any such scepticism. —Harmonization of the leniency programmes in the European Competition Network (ECN)7 and the possible introduction of a one-stop shop. These issues are discussed in Section C. One particular problem is that where an application is made to the European Commission and other Member States (including applications under national criminal leniency programmes) the company may be faced with conflicting conditions on terminating participation in the cartel. Some authorities require continued participation in order not to alert the other cartel members to the possible investigation and some times as a source of covert intelligence. But the EC program requires involvement to be ended no later than when the application is made to the Commission. However, the ECN members recently agreed that an authority with discretion will exercise it to avoid such conflict. But there remain many differences between individual programmes. As for a one-stop shop, there is a powerful argument not just for a one-stop shop in the ECN but also for a global policy on leniency. —Is it necessary to provide specific citations from the leniency applicant in the statement of objections and decision? The greater this information the more the leniency applicant is put at a disadvantage compared with the other cartel members in any private actions. This needs to be avoided if the incentive to apply for leniency is not to be damaged. The impact of private actions on leniency applications is discussed further in Sections D and E. —The position of individuals and that of undertakings should be as consistent as possible in order to ensure there is no disincentive to a leniency applicant—or, indeed, unfairness. Criminal sanctions bite against individuals. Undertakings are punished with fines. Where there are different programmes operating, as for example in the UK, there is an increased need, at EU and at ECN level, to make sure that the systems of leniency and immunity are sufficiently transparent and certain to ensure that there is no conflict of interest between individuals and companies that will discourage applications. 7 The ECN comprises the competition authorities of European Commission and of the 27 Member States.


548

12/10/07

13:13

Page 548


Another reform which would increase the effectiveness of the EC leniency programme would be the addition of a provision for “leniency plus”. Under leniency plus an undertaking that has secured less than 100% leniency (immunity) in a cartel investigation in one market can increase the reduction in its fine if it advises the competition authority about another cartel in another market and is eligible for immunity in the second cartel. Nearly half of the US Department of Justice’s current international cartel investigations were initiated following evidence submitted under the Department’s Amnesty Plus programme.8

2. How successful is the EC leniency programme in uncovering cartels? The vast majority of EC cartel decisions which involved the grant of leniency to one or more companies have been under the 1996 Notice. The 2002 Notice builds on the results of the 1996 Notice. Hence any evaluation of the 2002 Notice should include experience with both notices. Given the improvements in the 2002 Notice, its effectiveness in uncovering cartels should be even better than its predecessor’s. In October 2003, the then Head of the Scrutiny and Coordination Unit in DG Competition, Olivier Guersent, stated that the EC leniency programme is a key measure in the European Commission’s fight against cartels (emphasis added): “[The 1996] Notice has proven to be a formidable tool to detect and prosecute cartels. Not only did it enable the Commission to uncover individual cartels, but the fear that one member could go to the authorities and obtain immunity proved to de-stabilise cartel activity in general. As a result, since 1996, the Leniency Program has been the most effective generator of important cases. About 100 companies have filed leniency applications under this program and, since 1996, the Commission has taken 24 formal decisions in cartel cases in which companies co-operated with the investigations. . . . Already the [2002 Notice] . . . is very successful. Since 19 February 2002 . . . more than 50 new leniency applications have already been submitted,9 amongst which 44 applications [were] for immunity. These include a number of simultaneous applications in both the US and the EU (and some simultaneous applications in Canada), which allowed in a number of cases, for close co-operation and/or simultaneous investigative measures.”10

8 Hammond S. (2005): “An Update of the Antitrust Division’s Criminal Enforcement Program”, Antitrust Division, US Department of Justice. Washington, 16 November 2005, text available from http://www.usdoj.gov/atr/public/speeches/213247.htm. 9 33 of the 53 applications for leniency were granted. 10 Guersent O. (2004): “The Fight Against Secret Horizontal Agreement in the EC Competition Policy”, in Hawk B., ed., International Antitrust Law and Policy, Fordham Corporate Law 2003, Juris Publishing, New York, pp. 43–54. The statistics cover the period up to 1 October 2003.


12/10/07

13:13

Page 549

The EC Leniency Program Faces Great Challenges 549 All the EC cartel decisions up to end July 2006 where leniency and/or immunity have been granted are listed in Table 1. The table shows for each case how much leniency was granted and the number of companies which received this leniency. The dates in the table are those for the adoption of the decision, which will normally be a few years after the leniency application(s). In some cases, one company received immunity. The cases listed all involved the 1996 Leniency Notice cases other than four which were under the 2002 Notice: Italian raw tobacco and Rubber chemicals in 2005, and the two decisions so far in 2006. Table 1: EC Leniency Notice cases11 by the year of the cartel decision Year

Decision

Reduction of fine and number of companies

1998 1998 1998 1998 1999 2000 2001 2001 2001 2001 2001 2001 2001 2001 2002 2002 2002 2002 2002 2002 2002 2002 2003 2003 2003 2003 2004 2004 2004 2004

Stainless Steel British Sugar Preinsulated Pipes Greek Ferries Seamless Steel Tubes Lysine Graphite Electrodes Sodium Gluconate Vitamins Brasseries Luxembourg Brasseries Belges Citric Acid Zinc Phosphate Carbonless Paper Methionine Industrial & Medical Gases Fine Art Auction Houses Plasterboard Methylglucamine Food Flavour Enhancers Rond à Beton Speciality Graphite Sorbates Carbon & Graphite Products Organic Peroxides Industrial Copper Tubes Copper Plumbing Tubes Spanish Raw Tobacco Needles, etc. Choline Chloride

40%: 2 companies, 10% : 4 50%: 1, 10%: 3 30%: 5, 20%: 3 45%: 1, 20%: 6 40%: 1, 20%: 1 50%: 2, 30%: 2, 10%: 1 70%: 1, 40%: 1, 30%: 1, 20%: 2, 10%: 3 80%: 1, 40%: 2, 20%: 3 100%: 1, 50%: 2, 35%: 3, 30%: 1, 15%:1, 10%: 1 100%: 1, 20%: 3 50%:1, 30%: 1, 10%: 4 90%: 1, 50%:1, 40%:1, 30%:1, 20%: 1 50%: 1, 40%: 1, 10%: 4 100%: 1, 50%: 1, 35%: 1, 20%: 1, 10%: 3 100%: 1, 50%: 1, 25%: 1 25%: 2, 15%: 2, 10%: 2 100%: 1, 40%: 1 40%: 1, 30%: 1 100%: 1, 40%: 2 100%: 1, 50%: 1, 40%: 1, 30%: 1 20%: 1 100%: 1, 35%: 6, 19%: 1 100%: 1, 50%: 1, 40%: 1, 30%: 1, 25%: 1 100%: 1, 40%: 1, 30%: 2, 20%: 1 100%: 1, 50%: 1, 25%: 1, 15%: 1 50%: 1, 30%: 1, 20%: 1 100%: 1, 50%: 1, 35%: 2, 15%: 1, 10%: 2 40%: 1, 25%: 2, 20%: 1, 10%: 1 100%: 1 30%: 2, 20%: 1

11 For 1998 to 2002 (Rond à Beton) see n 9. From 2002 (Speciality Graphite) to date see http://europa.eu.int/comm/competition/antitrust/cases.


550

12/10/07

13:13

Page 550


Table 1: (cont.) Year

Decision

Reduction of fine and number of companies

2005 2005 2005 2005 2005 2006 2006

Monochloroacetic Acid Industrial Thread Italian Raw Tobacco12 Plastic Industrial Bags Rubber Chemicals Bleaching Chemicals Acrylic Glass

100%: 1, 40%: 1, 25%: 1 Some reductions 50%: 1, 30%: 1 30%: 1, 25%: 2, 10%: 2 100%: 1, 50%: 1, 20%: 1, 10%: 1 100%: 1, 40%: 1, 30%: 1, 10%: 1 100%: 1, 40%: 1, 30%: 1

Under the 1996 Notice, leniency was granted to approximately 150 companies over five and a half years in just over 30 cartel cases.13 Of these, in 14 cases one company was granted immunity. The first case in which immunity was granted was in the Vitamins Cartel with the decision in 2001. However the first application which led to 100 per cent leniency was in the Carbonless Paper Cartel with the decision later in 2001.14 Under the 2002 Notice there were 167 applications up to the end of 2005, i.e. in the first nearly four years. Of these, 87 were immunity applications. By end 2005, the Commission had granted 51 decisions for conditional immunity (one of which was subsequently withdrawn), rejected or decided not to deal any further with 23 immunity applications and had 13 more recent immunity applications outstanding. In the same period, 17 applications for reductions in fines were provisionally granted (8 of which were confirmed), 34 applications were rejected and nearly 30 were outstanding. Applications for reductions in fines are decided much later in the procedure than those for immunity. Twice as many applications for leniency have been submitted during nearly four years of the 2002 Notice than during five and a half years of the 1996 Notice. Even more striking is the comparison for immunity where decisions granting conditional immunity under the new Notice have been running at over five times the rate at which immunity was granted under the old Notice.15 Chart 1 presents Table 1 graphically (up to 2005). It clearly shows the significant increase in the numbers of firms granted leniency since the first cases in decisions in 1998. Also striking is the high number of firms granted leniency 12 Deltafina was originally awarded conditional immunity. But immunity was withheld in the decision because of a failure by Deltafina to maintain confidentiality over its leniency application. In the decision, Deltafina was awarded a 50 per cent reduction in its fine for its effective cooperation outside the scope of the 2002 Notice. This 50 per cent reduction is not included in the table. Deltafina has appealed to the Court of First Instance. 13 This excludes a few cases where fines were reduced by 10 per cent solely for not contesting the facts or “some cooperation”. 14 Bloom M. (2005): “Benefits and Challenges of Leniency Programmes in the Context of EC Modernization”, in Raffaelli E., ed., Antitrust between EC Law and National Law, Bruylant, Treviso, pp. 183–197. 15 Kroes N. (2006): “Enforcement of Prohibition of Cartels in Europe”, speech presented at the 11th Annual EU Competition Law and Policy Workshop, EUI Florence, 2–3 June 2006.


12/10/07

13:13

Page 551

The EC Leniency Program Faces Great Challenges 551 in the 2001 cartel decisions. But the peak year for the number of firms granted immunity in cartel decisions was not until 2002. The average leniency granted each year has generally increased since 1998 as shown in Chart 2. Partly this will reflect the extent to which a company is granted immunity in each cartel.

30 100% 50–90% 35–45% 10–25%

No. of companies

25 20 15 10 5 0

1998

1999

2000

2001

2002

2003

2004

2005

Chart 1: EC Leniency Notice by the year of the cartel decision 50 45 40 35 30 % 25 20 15 10 5 0

1998

1999

2000

2001

2002

2003

2004

2005

Chart 2: Average percentage of leniency granted in cartel decisions in each year


552

12/10/07

13:13

Page 552


Another way of assessing the effectiveness of the leniency policy is to look at the generation of on-site inspections in cartel investigations. Between 1996 and 1998 none of the on site inspections16 carried out by the European Commission was based on a leniency request. However, as the EC policy became more widely known the position changed dramatically. Between 2001 and 2003 nearly two-thirds of inspections were based on leniency requests. Another striking fact is that throughout the eight year period to 2003 there were generally three or four inspections a year that were not based on leniency requests.17 But inspections based on leniency applications rose steadily from one in 1999 to fourteen in the first nine months of 2003. This demonstrates the strong ability of leniency programmes to uncover cartels. Almost all the inspections resulted in a decision.18 Unfortunately there are no available figures to compare the effectiveness of the two Notices in generating on-site inspections. One important factor that is likely to lead to an overestimate of the success of the EC leniency programme is where applications to the Commission either followed on from those to the US Department of Justice (DOJ) or were simultaneous. The prime aim of any applicant is normally to avoid US criminal sanctions. But once a US investigation is stimulated by an amnesty application, other authorities will start investigations as they become aware at some stage of the US one. Hence applications need to be made simultaneously to other authorities or as soon as possible after one to the DOJ. It is the US powers rather than the EC (or other jurisdiction) powers which drive these applications. However, if the applicants could not secure leniency in the EC as well as the US it is highly likely that a significant proportion of them would not apply for US amnesty as they would not be able to avoid heavy EC fines.19 In approaching half of the EC cases from 2000 there was a prior or simultaneous application for amnesty under the US programme. Of course, this primarily reflects the much greater effectiveness of criminal powers rather than any inherent weakness in the design of the EC leniency programme as such. Some international cartels are now “carving out” the US.20 Without the US criminal sanctions and active enforcement record, how many of these 16

Each inspection concerns a different case and involves numerous companies inspected. In 2002 there were six and in the first nine months of 2003 there were seven. 18 See supra note no. 10 for the facts in this paragraph. 19 This is analogous to the concerns that led to the US detrebling legislation (see page 15 and note 32) in order to avoid the risk of large US treble damages deterring amnesty applicants. 20 “We have uncovered international cartels that operated profitably and illegally in Europe, Asia, and elsewhere around the world, but did not expand their collusion to the United States solely because the executives decided it was not worth the risk of going to jail. I am referring to cartels that had every opportunity to target U.S. consumers. The cartel members sold in the U.S. market, and they were already getting together and fixing prices everywhere else they sold. Indeed, in some cases, the U.S. market was the largest and potentially most profitable, but the collusive conduct still ceased at the border. Why? The answer, from the mouths of the cartel members and verified by our investigators, is that the executives did not want to risk getting caught and going to jail in the United States.” Hammond, S. (2006): “Charting New Waters in International Cartel Prosecutions”, Antitrust Division, U.S. Department of Justice, San Francisco, 2 March 2006, text available from http://www.usdoj.gov/atr/public/speeches/214861.htm. 17


12/10/07

13:13

Page 553

The EC Leniency Program Faces Great Challenges 553 cases would have been revealed by leniency applications to the European Commission? It is, of course, impossible to answer this question but I suspect the answer might well be “few”. Leniency programmes have, in addition, generally benefited from increased regulatory disclosure obligations under statutes such as Sarbanes Oxley (US), the Proceeds of Crime Act (UK) and related money laundering legislation. Some applications to the Commission may also have been stimulated by the cartel rules and leniency programmes in Member States, such as criminalization in the UK. The conclusion is that the 2002 Notice is clearly effective in uncovering a significant number of cartels. This should mean that the regime is also an important deterrent to cartels. However, more cartels would be uncovered by the Notice and more would be deterred with some improvements in the certainty and transparency of the Notice, the addition of leniency plus and, in particular, if the European Commission had criminal powers with custodial sentences to sanction cartels. Another important improvement would be harmonization of all the programmes operated by the members of the ECN, and ideally, the introduction of a one-stop shop. This is discussed in the next section. Indeed, as already mentioned above, there is a powerful argument not just for a one-stop shop in the ECN but also for a global policy on leniency.

C. Harmonization of amnesty/leniency programmes? A one-stop shop system? 1. Harmonization of leniency programmes There is strong pressure from legal practitioners and business to harmonize the different leniency programmes in the Member States. But there is even greater pressure to introduce a one-stop shop. The former would be useful in the absence of the latter. A one-stop shop would, almost inevitably, only be possible if the individual programmes were essentially identical or part of a single overall programme. Practitioners consider that the uncertainty inherent in the differences between programmes and the lack of a one-stop shop reduces the effectiveness of the European leniency programmes. Uncertainty can seriously reduce the incentive to apply for leniency. There are many and significant differences in the leniency programmes within the ECN. These differences include: —Whether a programme exists. As of this writing, the Commission website lists 18 Member States with a programme.21 Those without, according to 21 European Commission Authorities in EU Member States which operate a leniency programme [online]. Available from http://europa.eu.int/comm/competition/antitrust/legislation/ authorities_with_leniency_programme.pdf [Accessed 1 August 2006].


554

12/10/07

13:13

Page 554


this list, are Denmark, Estonia, Italy, Malta, Portugal, Slovenia and Spain. However, Estonia has leniency provisions in its code of criminal procedure—although the implementing regulation and guidelines are not yet in place. Portugal adopted a law in May 2006 to establish a leniency regime but this has not yet entered into force. —Which infringements are covered by the programme. Some cover vertical price fixing as well as horizontal price fixing. —The scale of reductions in fines provided and the extent of discretion of the authority over this including whether there is automatic immunity for the first applicant. —Whether leniency plus is offered—mostly not. —The action required to secure a place in the queue. Whether an oral application can be made or only a written one. Whether a hypothetical/ anonymous application is possible. Whether markers are allowed. —Whether the applicant should continue participating in the cartel or stop immediately. —The evidence required for a valid application e.g. all evidence in an applicant’s possession; sufficient to find an infringement; sufficient to launch an investigation. —The extent of the on-going cooperation obligation in practice. —Whether ringleaders are carved out from immunity or only those who have coerced others to take part in the cartel. —Whether admission of an infringement is required. Practical experience is insufficient to assess for some programmes but an admission is likely to be required in around half of the programmes. —Whether the cartel legislation—and hence the leniency programme— includes sanctions on individuals. Some countries have civil fines (or similar sanctions, such as company director disqualification) for individuals and some have criminal sanctions as the next bullet discusses. —Whether the programme applies to civil or criminal sanctions. In Ireland and the UK there are criminal leniency programmes. The UK has both criminal and civil programmes. In Hungary there is a leniency programme for criminal sanctions against cartels involved in public procurement. In some Member States there are criminal powers which may be used against cartels and no leniency programme: Austria (bid rigging), Estonia,22 France,23 Germany (bid rigging), and Slovenia. —When an applicant is informed that it is likely to obtain leniency. An ECN working group is seeking to develop a model harmonization programme. This will, of course, only achieve harmonization to the extent that individual Member States make revisions to their programmes to 22 As noted above, Estonia has leniency provisions in its code of criminal procedure but the implementing regulation and guidelines are not yet in place. 23 In France, there is an informal leniency arrangement.


12/10/07

13:13

Page 555

The EC Leniency Program Faces Great Challenges 555 harmonize with the model. Before seeking to develop a model programme, the working group screened existing leniency programmes to detect discrepancies that may have an adverse effect on other ECN programmes. It had also explored different possibilities for alleviating the burden associated with multiple filings. One possible option where it is likely that the Commission will deal with the case is for applicants to file short summary applications with national competition authorities at the same time as a complete application is filed with the Commission. Some ECN members have implemented some of the results of these early ECN discussions. For example, this is reflected in the revised German programme.24 New features in this program are: guaranteed immunity for the first applicant that enables the Bundeskartellamt (BKA) to obtain a search warrant, subject to satisfying various conditions; the ability to obtain immunity after a search provided the applicant is the first and satisfies various conditions; the introduction of markers; and a system to enable companies to exit a European-wide cartel which have filed applications with both the BKA and the Commission. Under this new system, the marker is retained with the BKA until it is clear whether the Commission or the BKA will handle the case. On this last point, there is a need for greater certainty as to whether the Commission will take a case that is reported to it as part of a leniency application, and what the position is if it does not do so. In terms of harmonization the most important issues are where there are contradictory demands in the programmes and the burden on companies of making multiple (and different) applications. The ECN working group should address these as a priority. On the issue of contradictory demands, Kris Dekeyser, Head of Unit, DG Competition, recently stated in a paper on the ECN: “The only example of a truly conflicting demand that has so far been detected is where one authority would require the applicant to immediately stop its cartel activities whereas another would request it to continue in order not to endanger the investigation. Should this materialise in an individual case, the ECN members have agreed that the authority that has a discretion would use it in such a way that a conflicting demand would not arise in the concrete case.”25

Apparently, there would be no situation in which there was insufficient discretion for this agreement to be applied in practice.

24 Bundeskartellamt (2006): Notice no. 9/2006 of the Bundeskartellamt on the immunity from and reduction of fines in cartel cases, Bonn, 7 March 2006, text available from http://www. bundeskartellamt.de/wEnglisch/download/pdf/06_Bonusregelung_e.pdf. 25 Dekeyser K. (2006): “Almost 2 Years of Cooperation Within the ECN”, IBC Conference on Advanced EC Competition Law, London, 5 May 2006. See also OECD (2006): Competition Law and Policy in the European Union, p. 70, text available from http://www.oecd.org/dataoecd/ 7/41/35908641.pdf.


556

12/10/07

13:13

Page 556


2. One-stop shop Each of the national authorities with a leniency programme operates them separately from each other and from the Commission. Hence, the grant of immunity under one national programme has no effect under other national programmes or the Commission program even if the same cartel is concerned. Only the Commission can grant leniency that applies throughout the EU— and that leniency is not relevant if Member States, rather than the Commission, subsequently investigate the cartel case. Whether leniency can be obtained depends on which authority or authorities take the case, whether they each have a leniency programme and whether the applicant is first to apply to all those investigating the case. Practitioners are clear that the effectiveness of the leniency programs in uncovering cartels would be improved by a one-stop shop 26—but is this possible? Commissioner Kroes first spoke publicly about a one-stop shop for leniency applications in February 2005: “I would like to further explore the idea of a European-wide ‘one-stop-shop’ for leniency applications in cartel cases.”27 This idea was mentioned in subsequent speeches by the Commissioner and rather more detail was provided in her “First Hundred Days” speech in April 2005: “[. . .] the weaknesses identified in the present leniency application regime need to be addressed. I have already floated the idea of a one-stop shop for leniency applications in the EU. Initial reactions have been mostly favourable, from practitioners and businesses alike. Detailed work is underway with the cooperation and input of our colleagues in the European Competition Network. The design of any future system has to take full account of the specific needs of those who will eventually have to operate it. This means that rather than limiting ourselves to centralising solutions, we have to look at the options for one-stop leniency which fully exploit the dimension of the network.”28

Two options for a one-stop shop could be (a) a centralized system run by the Commission and (b) a coordinated system run by all members of the ECN, at least those with a leniency programme. The Commissioner and Philip Lowe have both indicated their preference for the second option, which is a system of parallel competences. The first option runs counter to the decen26 In addition to increasing effectiveness, there is the question of “fairness”, if a party applies for leniency everywhere, is denied in one location, or is too slow there, and another party gets immunity because it has been tipped of as to the leniency application in the meantime. 27 Kroes N. (2005): “Effective Competition Policy—A Key Tool for Delivering the Lisbon Strategy”, Brussels, 3 February 2005, text available from http://europa.eu.int/rapid/ pressReleasesAction.do?reference=SPEECH/05/73&format=HTML&aged=0%E2%8C%A9ua ge=EN&guiLanguage=en. 28 Kroes N. (2005): “The First Hundred Days”, Brussels, 7 April 2005, available from http://europa.eu.int/rapid/pressReleasesAction.do?reference=SPEECH/05/205&format= HTML&aged=0&language=EN&guiLanguage=en.


12/10/07

13:13

Page 557

The EC Leniency Program Faces Great Challenges 557 tralizing objective of EC Modernization and would not make good use of the resources of the national authorities. It would have resource issues for the Commission. But it is conceptually more straightforward and is likely to be viewed as more certain and predictable by legal practitioners and business. The second option of a decentralized and coordinated system is a significant challenge. For it to work effectively, all the programmes would need to be fully harmonized. The authorities would need to accept that information provided to one national competition authority is sufficient in order to receive immunity for the whole ECN. If so, the authorities would need to be prepared to receive information that might only or predominantly be of relevance for another national authority’s investigation. Consider the example of an immunity application before inspections had been carried out concerning an alleged cartel covering Germany, France and the Netherlands. An application could be filed with the BKA. If the French authority (Conseil) and the Dutch authority (NMa) were not willing to respect an immunity decision by the BKA unless the latter had gathered the information that the Conseil and the NMA would need to carry out their respective inspections, the BKA would need to demand and process information concerning the Dutch and the French markets. Measures would also need to be put in place to ensure that the duty of continuous cooperation not only applied towards the authority that had granted (conditional) immunity but towards all authorities that were (in parallel or subsequently) acting on the case. The example can be extended to applications in a post-inspection scenario (immunity or reduction of fines). The authorities would be aware that parties that had been subject to an ex officio inspection could seek to “game” the system by immediately applying for immunity to another national authority. Since that authority would have nothing in its file, immunity might be readily granted. Another alternative would be a system whereby all information which any ECN member possessed was taken into account when one authority decided whether immunity is still available. This approach would require authorities to share sensitive information at a very early stage. Applications for reductions of fines would be challenging to handle, if an authority needed to assess whether information provided by an applicant contributed to the investigation carried out by another authority. If a true one-stop shop is developed, it will be the first leniency regime in the world where applications can be made to more than one place—in this case in up to 26 places.29 A global policy on leniency is at least as desirable but is an even greater challenge.

29

Up to 28 places since Bulgaria and Romania joined the EU in January 2007.


558

12/10/07

13:13

Page 558


D. Leniency and the interface with private enforcement and criminal enforcement 1. Interface with private enforcement Neelie Kroes recognized the tension between leniency and private enforcement in a recent speech on the Commission Green Paper on damages actions.30 “I would also like to say a word on what some portray as a tension between a well functioning leniency programme and enhanced private enforcement of the European competition rules. I am convinced that the two objectives are compatible. Although we must make sure that damages actions are possible whether or not there has been public enforcement, we must also be realistic. Cartels are by their nature hard to detect. Damages actions for victims of a cartel will remain largely theoretical if we don’t also have a well-devised leniency programme that destabilises and uncovers them. And conversely, the options in the Green Paper show how the leniency programmes of the European competition authorities could actually benefit from some well devised measures to facilitate actions for damages.”31

The US has already addressed the tension between the DOJ programme and treble damages private actions through the de-trebling provision of the Antitrust Criminal Penalty Enhancement and Reform Act of 2004.32 This provision increases the incentive of cartel members to turn themselves in as it limits their potential damages in private lawsuits to single damages based on their own role in the cartel, provided that they also cooperate with plaintiffs in the private lawsuit (in addition to the cooperation required with the DOJ investigation and prosecution.) Other cartel participants remain fully liable for treble damages based on harm caused by the entire conspiracy on a joint and several basis. Hence, their liability is increased as they are also liable for the damages that the amnesty applicant no longer bears following the detrebling. This should further increase the incentive to be the first to seek amnesty despite the requirement to cooperate with plaintiffs which can be an onerous responsibility. It is understood that between two to three applications a 30 European Commission (2005): Damages actions for breach of the EC antitrust rules, COM(2005) 672, Brussels, text available from http://europa.eu.int/comm/competition/antitrust/ others/actions_for_damages/ gp_en.pdf. 31 Kroes N. (2006): “More Private Antitrust Enforcement Through Better Access to Damages: An Invitation for an Open Debate”, Brussels, 9 March 2006, text available from http://europa.eu.int/rapid/ pressReleasesAction.do?reference=SPEECH/06/158&format= HTML&aged=0&language=EN&guiLanguage=en. 32 Pub. L. No. 108-237, 118 Stat. 661 (codified in scattered sections of 15 U.S.C. and accompanying notes).


12/10/07

13:13

Page 559

The EC Leniency Program Faces Great Challenges 559 month are currently being made for amnesty in the US compared with one a month prior to the de-trebling legislation. If private actions increase materially in Europe this is likely to reduce leniency applications unless action is taken to ensure there remains a real incentive to apply for leniency. This risk is greater for the EC programme— and most national programmes—than for the DOJ one as the threat of a jail sentence is a strong incentive to apply for amnesty in the US. If the European Commission, as discussed earlier, was to obtain criminal sanctions with custodial sentences this would maintain the incentive to apply for leniency—as can such criminal sanctions in Member States accompanied by appropriate leniency regimes. The Green Paper on private actions puts forward three options for comment on the tension between increased private actions and leniency applications: —Exclusion of discoverability of the leniency application, thus protecting the confidentiality of submissions made to the competition authority as part of leniency applications. —A conditional rebate on any damages claim against the leniency applicant; the claims against other infringers—who are jointly and severally liable for the entire damage—to remain unchanged. —Removal of joint liability from the leniency applicant, thus limiting the applicant’s exposure to damages. One possible solution would be to limit the liability of the leniency applicant to the share of the damages corresponding to the applicant’s share in the cartelized market.33 The first option will be particularly necessary for national authorities which continue to require written corporate leniency statements. This has already been briefly mentioned in Section B in relation to the EC programme. However, will it be possible to prevent discovery where US courts—or perhaps those in the UK and Ireland—have jurisdiction over the leniency applicant? This is discussed further in Section E as part of the possible tension between EC leniency and US and some European private actions. The other two options are both features of the US de-trebling provision. But the difference between the damages payable by the amnesty applicant and the other cartel members is likely to be significantly larger in the US than with the two EC options. US damages are, of course, trebled. While one of the ideas floated in the Green Paper is double damages, we are not likely to see treble damages in European courts. Also, where pre-judgment interest applies this will significantly increase the single damages compared with the US where there is no pre-judgment interest. If private actions do increase substantially in Europe, and the Commission does not have criminal powers,

33

See Green Paper, cited supra note 30, at pages 9–10.


560

12/10/07

13:13

Page 560


there is a real risk that the adoption of all three of the above options will not be enough to ensure leniency programmes maintain their attraction.

2. Interface with criminal enforcement Some EEA countries have powers to impose criminal custodial sentences in cartels: Austria (bid rigging), Estonia, France, Germany (bid rigging), Hungary (public procurement), Ireland, Norway, Slovenia and the UK. Other countries have powers to impose criminal fines. So far only France and Ireland have imposed sentences on an individual. The sentences were almost entirely suspended. In France there have been at least five bid rigging cases since the early 1990s in which sentences have been imposed. Three of these involved suspended sentences and in two of the cases the sentences were partially suspended. In the first case in Ireland in March 2006, the chief coordinator of a heating oil cartel in Galway was sentenced to serve six months in jail and to pay a €15,000 fine. Other members of the cartel were also fined. The sentence was suspended for one year. In the UK, a cartel is being investigated under new statutory criminal powers. In another UK case, involving alleged price fixing in the supply of pharmaceuticals to the National Health Service, nine executives and five companies were charged on 5 April 2006 with the common law offence of conspiracy to defraud. This is the first time individuals in the UK have faced criminal charges for cartel activity. Experience in the US is that a criminal regime is a powerful incentive to apply for leniency. In a lecture at King’s College London in February 2004, James Griffin, then Deputy Assistant Attorney General for Criminal Enforcement, Antitrust Division, DOJ, illustrated this point with two anecdotes: Senior Executive: “As long as you are only talking about money, the company can take care of me—but once you begin talking about taking away my liberty, there is nothing the company can do for me.” “In 25 years of prosecuting individuals engaged in cartels, I have never had one lawyer for an executive I was prosecuting tell me that his client would spend a few extra days in jail for a reduction in the recommended fine.”34

But, clearly, this incentive only applies where there is a criminal leniency programme—and there is no risk that the discretion of a public prosecutor to pursue a case could undermine certainty for defendants that they will not be prosecuted if they apply for amnesty. The latter could arise where the authority to implement a leniency programme and the authority to prosecute criminal cases are divided. In the absence of either of these conditions the criminal 34

See supra note no. 14, at page 193.


12/10/07

13:13

Page 561

The EC Leniency Program Faces Great Challenges 561 powers will discourage leniency applications if applicants consider these powers may be used. If so, this is also likely to discourage applications to other authorities where the cartel involves more than one country or to the European Commission. In most of the EEA countries where there are criminal powers there is no criminal leniency programme: Austria, Estonia,35 France,36 Germany, Norway and Slovenia. However, so far, it is only in Ireland and the UK that potential leniency applicants consider there is currently a real risk that the criminal powers may be used to prosecute cartels—and there are criminal leniency programmes in both these countries.37 In the US, the DOJ has the authority both to grant amnesty and to prosecute. In Ireland, the Competition Authority, in conjunction with the Director of Public Prosecutions (DPP), launched its Cartel Immunity Programme in December 2001.38 Applications for immunity are made to the Authority. The DPP has sole responsibility to prosecute offences on indictment. Only the DPP can grant immunity. Subject to meeting the requirements of the program (terminate participation, not have coerced, first applicant to meet requirements, cooperate fully, etc.) the Authority will make a recommendation to the DPP to grant immunity. In the UK, there is close cooperation and consultation in criminal cases between the Office of Fair Trading (OFT) and the Serious Fraud Office (SFO); the latter is the lead prosecutor.39 Investigations are carried out jointly and there is exchange of staff. The OFT grants immunity (i.e. “no-action letters”) after consultation with the SFO. The OFT can also prosecute on its own.40 The UK leniency programmes also provide protection against disqualification as a company director for a breach of the UK and EC competition prohibitions.41 Some have suggested that executives may be more cautious about applying for civil leniency from the European Commission if there is also likely to be a criminal trial in, say, the UK or Ireland. Their argument is that the published infringement decision that is likely to result from the application to the Commission opens up the risk of a criminal trial. But where there is a sound criminal leniency programme—as in the UK and Ireland—the existence of 35

See note no. 22. See note no. 23. 37 There is also a civil leniency program in the UK. 38 The Irish Competition Authority (2001): Cartel Immunity Programme, available from http://www.tca.ie. 39 Office of Fair Trading (2005) Leniency and no-action, London, text available from http://www.oft.gov.uk/NR/rdonlyres/FE4F98BF-D8C5-48FB-9F7E-D6133F907538/ 0/oft803.pdf. 40 The leniency programme does not apply to the common law offence of conspiracy to defraud but the Director of the SFO has announced that the SFO will not prosecute a cartel under the common law powers where a no-action letter has been issued under the Enterprise Act powers. See paragraph 5.12 at note no. 39. 41 A grant of leniency under the EC programme also provides protection against disqualification as a director. 36


562

12/10/07

13:13

Page 562


the criminal cartel powers seems more likely to increase than reduce applications for leniency for the reasons indicated by James Griffin.

E. International implications of leniency programmes Non-EEA countries with leniency programs include Australia, Brazil, Canada, Israel, Japan, Korea, Mexico, New Zealand, Romania, South Africa, Switzerland and the US.42 Of these jurisdictions, there are criminal cartel powers with the possibility of custodial sentences in Canada, Israel, Japan, South Korea, Mexico and the US. Australia is expected to introduce legislation into Parliament later this year for criminal cartel powers. Custodial sentences have only been imposed so far in Canada, Israel and the US. The importance of criminal powers—particularly the US ones—in stimulating leniency applications has already been discussed above. This section considers two international issues from a European perspective, both of which involve the interface with private enforcement. These are: can treble damage claims be brought in US courts for foreign sales and what is the likely reach of US and any European discovery powers? Significant risks from either of these will reduce the attractiveness of European leniency programmes.

1 Can treble damage claims be brought in US courts for foreign sales? Prior to the Supreme Court’s F. Hoffmann-LaRoche Ltd. v. Empagran 43 decision, which limited the extraterritorial reach of US antitrust laws, there was a real concern that plaintiffs with injuries caused by conduct outside the United States could sue in US courts.44 In Empagran, the Court held, unanimously, that a case brought by foreign plaintiffs fell outside the scope of the Sherman Act where the foreign anticompetitive conduct and injury were entirely independent of the domestic anticompetitive effects.45 The case followed in the wake of the vitamins cartels scandal, in which the defendants have paid fines of approximately $2.2 billion in the US, the EU, Canada, Australia, and Korea and have settled civil lawsuits with US plaintiffs for approximately $2.4 billion. More than ten 42 See pages 15–17 of ICN Anti-Cartel Enforcement Manual, supra note no. 3, for web links to most of these programmes. 43 F. Hoffmann-LaRoche Ltd. v. Empagran S.A., 124 S. Ct. 2359, 2366 (2004). 44 Bloom M. (2005): “Should Foreign Purchasers Have Access to US Antitrust Damages Remedies? A Post Empagran Perspective from Europe”, 61 New York University Annual Survey of American Law 433. 45 F. Hoffmann-LaRoche Ltd. v. Empagran S.A., 124 S. Ct. 2359, 2366 (2004).


12/10/07

13:13

Page 563

The EC Leniency Program Faces Great Challenges 563 individuals have received sentences resulting in imprisonment. The plaintiffs in Empagran were foreign corporations domiciled in Ukraine, Panama, Australia, and Ecuador who purchased vitamins abroad for delivery outside the US.46 The US has by statute and judicial decisions established a bounty-based system of antitrust recoveries through treble damages for violations that goes well beyond any policy that any European country has chosen to adopt.47 This system is supported by plaintiff-friendly court procedures and costs arrangements that mean that almost any plaintiff that can plausibly squeeze its way into a US federal court will seek to do so.48 However, the Supreme Court’s decision and subsequent Appeal Court decisions suggest that the probability of facing treble damage exposure in US federal court for alleged injuries suffered “worldwide” has significantly reduced—but it has not completely disappeared. Thus, while the overall trend is toward limiting the jurisdiction of U.S. federal courts to what one would expect to be a very narrow band of cases in which the domestic impact of an alleged international cartel proximately caused the foreign plaintiff’s injuries, other courts of appeals may rule on this issue in the future. For parties seeking to evaluate potential U.S. civil litigation exposure related to purely foreign sales, the probabilities of facing treble-damage exposure in U.S. federal court for alleged injuries suffered “worldwide” have certainly reduced. They may reduce further as other Courts of Appeals revisit this issue in light of Empagran. Notably, both the DOJ and Federal Trade Commission, as well as the governments of Canada, Japan, Germany, and the Netherlands, filed briefs before the D.C. Circuit in Empagran urging a rejection of a theory that would convert U.S. courts into “world courts,” interfere with foreign sovereigns’ own enforcement regimes, and risk deterring—rather than encouraging—applications for leniency.49 If the extraterritorial reach of US antitrust laws had not been limited, the risk of a treble damages suit would have significantly discouraged leniency applications in Europe given that the sanctions are generally only fines. While leniency in Europe will deliver no (or reduced) fines, it exposes companies to private actions for damages, because detailed knowledge of the cartel will 46

See supra note no. 44 at page 433. Baker D. (2004): “Revisiting History—What Have We Learned About Private Antitrust Enforcement That We Would Recommend To Others?”, 16 Loyola Consumer Law Review 379, at p. 397. Also available from http://www.luc.edu/law/academics/special/center/antitrust/ symposium/baker.pdf. 48 Brief for United Kingdom, Ireland and the Netherlands as Amici Curiae in Support of Petitioners, at 13–15, Empagran (No. 03-0724) 49 Spratling G., Jarrett D. and Shephard A. (2006): “Making the Decision: What to Do When Faced with International Cartel Exposure. Developments Impacting the Decision in 2006”, paper for American Bar Association Section of Antitrust Law 54th Antitrust Law Spring Meeting, Washington, p. 42–47. 47


564

12/10/07

13:13

Page 564


become public when the fully reasoned infringement decision is published by the competition authority. A potential leniency applicant will also consider the risk that another member of the cartel might go to the competition authorities first. However, unless there are criminal sanctions, the other members will make the same financial assessment of fines against private damages. If there is a risk of US treble damages, this is likely to outweigh the risk of civil fines—and deter leniency applications by all members of the cartel.

2 What is the likely reach of US and any European discovery powers? In 2002, US plaintiffs sought discovery through the US courts of the confidential corporate leniency statements made by companies seeking leniency from the European Commission. A US court ordered discovery in one of the cases, but not in the other.50 The Commission had filed amicus briefs in both cases. There have been further efforts by US plaintiffs since then to obtain this kind of discovery.51 In addition to cases seeking discovery of corporate leniency statements through the US courts, there was an important Supreme Court opinion in 2004 concerning discovery assistance to foreign and international tribunals.52 In October 2000, AMD filed a complaint with the European Commission against Intel for alleged anticompetitive behaviour. AMD sought discovery (through the US courts) of Intel documents in order to provide them to the Commission in connection with the complaint. AMD used a statute that permits US courts’ assistance in discovery of matters before foreign and international tribunals.53 The case did not involve a leniency statement, but other plaintiffs might use this statute to seek such documents. The Supreme Court took a very broad view of the statute, finding that the European Commission is a “foreign tribunal” within its meaning. But it concluded that as the statute authorises rather than requires “discovery assistance [for matters before foreign and international tribunals], we leave it to the courts below to assure an airing adequate to determine what, if any, assistance is appropriate.”54 The Court set out four factors that must be considered by a trial court judge in deciding an application for discovery. However, two lower court decisions in 2004 applying the factors both denied discovery.55 In Spring 2006, three

50 In re Vitamins Antitrust Litigation, 217 F.R.D. 229 (D.D.C. 2002); In re Methionine Antitrust Litigation, 221 F.R.D. 1 (N.D.Cal. 2002). 51 See supra note no. 49 at page 69. 52 Intel Corp.v. Advanced Micro Devices, Inc., 542 U.S. 241 (2004). 53 §1782(a). 54 See Intel, 542 U.S. at –, 124 S.Ct. at 2484. 55 Schmitz v Bernstein Liebhard & Lifshitz, LLP, 376 F.3d 79 (2d Cir. 2004); Advanced Micro Devices, Inc. v. Intel Corp., No. C 01-7033, 2004 WL 2282320 (N.D. Cal., Oct. 4, 2004).


12/10/07

13:13

Page 565

The EC Leniency Program Faces Great Challenges 565 different US District Court judges rejected attempts by Microsoft to use this statute to obtain discovery of documents from Oracle Corp, Sun Microsystems, Novell and IBM. In “The first hundred days” speech in April 2005 Commissioner Neelie Kroes shared some initial thoughts on five aspects of cartel enforcement. One concerned the possibility of a one-stop shop for leniency, which has already been discussed in Section C. Another of these thoughts was how to address the “interface between EU immunity and procedures in non-EU jurisdictions.” “[W]e need to address the challenges arising from the interaction of European administrative immunity and civil procedures beyond the borders of the Union. In some other jurisdictions undertakings found guilty of anticompetitive behavior can be obliged to pay civil damages to the consumers affected. Discovery of evidence rules sometimes oblige undertakings to reproduce before the national court the information already provided to the Commission as part of an immunity application. This threatens the credibility of the European programme and limits the willingness of undertakings to come forward to expose illegal activities to the Commission. We are working on practical solutions to this question of compatibility, and I intend to incorporate these in a revised version of the 2002 Leniency notice.”56

In February 2006 the Commission issued proposals to amend the 2002 Leniency Notice, in particular, through adding an annex with a procedure to minimize the risk of discovery of corporate leniency statements. This annex formalises the existing arrangements for oral leniency applications. Another solution would be a globally harmonized system—but until, and if, this can be achieved other remedies are necessary. The Commission’s main proposals for the new special procedure are: —Corporate statements can be made in oral form. —Access to the file, including to corporate statements, will only be given for the purposes of administrative and judicial proceedings for the application of Article 81 EC.57 —Access to corporate statements will only be granted by allowing parties to read written corporate statements and transcripts of oral corporate statements at the premises of the Commission and to take notes—but not make mechanical copies. —Parties seeking access to corporate statements will be required to make a written commitment to use documents obtained through access to the file only for the purposes of administrative or judicial proceedings for the application of Article 81 EC. 56

See supra note no. 28. The Commission is legally required to grant access to the file to parties concerned by the proceeding when a Statement of Objections is issued to them. As there is no access before then, corporate statements are not accessible until a Statement of Objections is issued. OJ L 123 [2004], page 18. 57


566

12/10/07

13:13

Page 566


—The Commission may seek to sanction abuses of the right of access to the file by a complaint to the bar association of the lawyer who was allowed access to the file and/or by seeking a higher fine for the undertaking concerned either in the cartel decision or in subsequent Community Courts proceedings.58 Practitioners have generally welcomed the Commission’s proposals but there are some concerns that they may not be sufficient to ensure that leniency applicants are not prejudiced in private action proceedings as a result of their application. In particular, as mentioned in Section B, the amendments do not take into account the use of corporate statements in litigation that might be brought in Europe. “Administrative and judicial proceedings for the application of Article 81” appear to include actions in national courts.59 If so, to ensure leniency remains attractive as private actions increase in Europe, the Notice will need expressly to limit the use of corporate statements to proceedings brought by the Commission and to any appeals to the Community Courts. Even with this, there are questions over the ability of the Commission to provide effective restraint. Once documents are out in the public in any form, they will be available for use in all litigation. This is an important issue which needs to be addressed in a comprehensive way to ensure leniency applications are not unnecessarily discouraged. When, and if, a one-stop shop leniency regime is developed by the ECN this limitation would need to be widened to cover all ECN members. Until then, to ensure the confidence of leniency applicants, the ECN should make it clear that national procedural rules will not undermine the objective of protecting corporate leniency statements. Another of the concerns is that the sanctions proposed by the Commission may not be effective. A lawyer may not be able to control the behaviour of a client. National bar rules may require them to act in the interests of their clients rather than in accordance with policy statements of the Commission. An undertaking that has had access to the file following receipt of a Statement of Objections may not, in the end, be found to have infringed Article 81— hence there would be no fine to increase as a sanction. An appeal may not be pending before the Community Courts, for example where an undertaking has received immunity. It is fair to conclude that there are some risks to the attractiveness of European leniency programmes from future private actions in the US and Europe. 58 European Commission (2006) Draft amendment of the 2002 Commission Notice on Immunity from fines and reduction of fines in cartel cases, available from http://europa.eu.int/ comm/competition/antitrust/legislation/leniency_en.pdf. 59 Recital 7 to Regulation 1/2003 states “National courts have an essential part to play in applying the Community competition rules. When deciding disputes between private individuals, they protect the subjective rights under Community law, for example by awarding damages to the victims of infringements. The role of the national courts here complements that of the competition authorities of the Member States. They should therefore be allowed to apply Article 81 and 82 of the Treaty in full.” (emphasis added).


12/10/07

13:13

Page 567

The EC Leniency Program Faces Great Challenges 567

F. Financial incentives for “whistleblowers” Very few jurisdictions provide financial incentives for whistleblowers. However there are provisions for such awards in the US and South Korea. The US powers have existed since the mid–1980s to punish fraud against the government. The Korean reward system was introduced in April 2005 by the Korean Fair Trade Commission (KFTC) to facilitate the detection of secret violations of competition law. The qui tam mechanism of the US Civil False Claims Act offers substantial monetary rewards for individuals who inform the DOJ of fraudulent conduct by government contractors and others. If the Department successfully prosecutes the alleged misconduct, the whistleblower is entitled to 25% of all sums the government recovers. If the Department chooses not to prosecute, the informer may proceed with his own suit. In this case, the maximum reward to the informant is 30% of the funds recovered. The False Claims Act and its qui tam mechanism were adopted by Congress in 1863 and the powers were augmented in 1986. Through this mechanism, the US government recovered $3.5 billion over the period 1986 to 2000. Informers received over $500 million. Some informants have received millions of dollars. The first use of the qui tam provisions to expose a large multinational cartel was in a bid rigging case concerning USAID funded wastewater treatment projects in Egypt.60 Fines of more than $87m were imposed in 2001. In South Korea, the KFTC pays awards for those who report on competition law violations with supporting evidence. The size of the reward depends upon the severity of the violation and the strength of evidence submitted. A Reward Review Committee has been established to ensure a fair and transparent process for determining reward eligibility. The identification of informants is kept confidential and “anonymous informants can receive reward money if they apply for the payment within six months after the date of the FTC’s ruling”.61 A reward of 66.87 million won (about $63,700) was paid in June 2005 to a person who provided decisive evidence in a welding rod cartel case—the largest reward so far to have been given. Until the informant provided this evidence the KFTC had no evidence or decisive information about the cartel and had not begun any investigations. The KFTC “expects this case will serve as a good example, encouraging outsiders like consumers, retired executives and staff of cartel participants, companies doing business with cartel participants to provide cartel evidence.”62

60 61 62

See http://www.wilmerhale.com/robert_bell . See http://www.ftc.go.kr/data/hwp/rewardsystem.doc. See http://www.ftc.go.kr/data/hwp/informant_reward.doc.


568

12/10/07

13:13

Page 568


Will such incentives be adopted in Europe? This is highly unlikely in the near future given concerns over the acceptability of “informers”—particularly paid ones. Indeed, as noted earlier, a few European countries have not yet introduced a leniency regime. But the following assessment of William Kovacic, now Commissioner, US Federal Trade Commission, seems a wise one. “The absorption of the qui tam technique into competition policy may be a slow growth, but it is not inconceivable. Twenty years ago, one would not readily have predicted that jurisdictions outside the United States would embrace the American approach of treating cartel offences as crimes. Nor was it inevitable that most competition policy systems would come to regard cartels as worthy of severe civil sanctions. If nations are willing to use a reduction in punishment (leniency) to “pay” cartel participants to reveal unlawful collusion, it is not a long conceptual step to “pay” informers with cash to do the same.”63

G. Plea bargaining In “The first hundred days” speech in April 2005 Commissioner Neelie Kroes introduced the idea that the Commission might consider some form of plea bargaining—or direct settlements—in order to deliver “swift enforcement with timely punishment” in cartel enforcement: “We have seen that the burden of immunity applications falls heavily on the European competition authority. Partly this is because in each and every case brought before us, we are obliged to investigate every last substantive detail. Our final decisions have to be fully reasoned on the basis of our own analysis of the facts. Unlike some other systems—that of our American colleagues, for example— there is no arrangement for simplified handling of cases in which the parties to the cartel and enforcer concur as to the nature and scope of the illegal activity undertaken and the appropriate penalty to be imposed. For the time being, I make this comment as a simple comparative observation. The measures we are already committed to taking to improve anti-cartel enforcement need to bed down and have a chance to demonstrate their effectiveness. But if, despite these efforts, the Commission is not able to deliver swift enforcement with timely punishment, we may need to look at how some form of plea bargaining procedure could bring advantages in the context of European competition law.”64

This is an interesting idea which is well worth investigating. Plea bargains, as such, are not generally acceptable in national courts in Europe. But a direct settlement with an administrative authority would be possible. Indeed, this is 63 Kovacic W. (2003): “Private Participation in the Enforcement of Public Competition Laws”, London, 15 May 2003, available from http://www.ftc.gov/speeches/other/030514biicl.htm. 64 See supra note no. 28.


12/10/07

13:13

Page 569

The EC Leniency Program Faces Great Challenges 569 the process with commitments under Article 9 of Regulation 1/2003. Article 9 commitments are, of course, not used for cases where the Commission intends to impose a fine—as in hard core cartel cases. Very few of the US DOJ’s cartel cases go to trial. Instead they are settled through a plea bargain which must then be accepted—or rejected—by a court. Almost invariably the judge accepts the plea bargain without variation. A very important driver for cartel members to settle their case quickly is the desire to obtain as short a jail sentence as possible. The first applicant to the DOJ from a cartel will obtain amnesty (provided that the various conditions on cooperation, not coercing other cartel members etc are satisfied). Subsequent applicants will obtain increasingly harsh conditions in a plea bargain as it will be increasingly difficult to provide the Department with evidence that it does not already have.65 There is an argument that plea bargains—or direct settlements—in Europe may hinder private damages actions in follow-on cases, as there may be no formal decision on which reliance can be placed. Defendants would, no doubt, try to bargain to avoid a formal decision. If the formal decision is not retained in the case of a settlement, claimants would find it hard to bring claims unless there are disclosure rules akin to those in the US—or there is a mechanism for redress as a requirement of the settlement. The latter would, however, be likely to reduce the attractiveness of settlements to defendants. While few EU countries have criminal sanctions with jail sentences for cartels there would be an incentive to settle rapidly if this clearly reduced the fine significantly—and by a reasonably predictable amount. Indeed, a significant number of companies granted leniency under the 1996 Notice were granted it for “not substantially contest[ing] the facts on which the Commission base[d] its allegation”66 and the others cooperated with the Commission’s investigation to a greater or lesser degree. Predictability of likely fine reductions would be important. Scott Hammond, Deputy Assistant Attorney General for Criminal Enforcement, Antitrust Division, DOJ, recently gave a speech with the aim of providing “more transparency as to the potential rewards and incentives available for second-in companies and the factors considered in determining the size of the cooperation discount.”67 Another important driver in the US for cartel members to settle their case without an open trial is their desire to minimise the information available to plaintiffs in treble damages suits. As indicated by the discussion of discovery 65 The DOJ’s Corporate Leniency Program only covers amnesty. The reductions in sentences and fines for the applicants after the first one are granted through plea bargaining. In contrast, European civil leniency programmes cover immunity for the first applicant and the reductions in fines for subsequent applicants. 66 See supra note no. 4 section D2. 67 Hammond S. (2006): “Measuring the Value of Second-in Cooperation in Corporate Plea Negotiations”, Antitrust Division, US Department of Justice Washington, 29 March 2006, text available from http://www.usdoj.gov/atr/public/speeches/215514.htm.


570

12/10/07

13:13

Page 570


in Section E this could be a driver for rapid settlement in Europe too even if— at least currently—a less powerful one than in the US.

H. Conclusion Despite the great success of the EC leniency programme, there are some significant challenges to ensuring it is fully effective now and in the future. Transparency and certainty will remain of key importance in maximizing incentives to apply for leniency. This paper has identified a number of ways in which certainty and transparency could be improved. EC Modernisation has highlighted the desirability of a one-stop shop within the EU. Indeed, a global policy on leniency would be at least as desirable. But that is an even tougher challenge. If private enforcement grows significantly in Europe it will almost certainly reduce the incentive to apply for leniency. As will any material increase in the ability of private plaintiffs in the US or Europe to obtain discovery of corporate leniency statements. One way to maintain—or increase—the incentive would be the introduction of criminal sanctions in more Member States or even for the European Commission.

(Zb) Ehlermann 07 Kovacic

12/10/07

13:14

Page 571

IV William E. Kovacic1 Bounties as Inducements to Identify Cartels

A. Introduction Competition policy today treats horizontal schemes to fix prices, rig tenders, allocate customers, stabilize market shares, or divide geographic territories as grave transgressions.2 Modern efforts to challenge cartel arrangements have taken two basic forms. One is to raise sanctions for culpable parties.3 The second is to increase the detection of forbidden activity. The foremost recent innovation in detection has consisted of providing powerful inducements for cartel participants to inform government competition agencies of wrongdoing.4 Among other features, these leniency or amnesty measures significantly reduce the punishment for the first cartel member to reveal the cartel’s existence. This paper makes the case for taking the practice of giving rewards to informants a step further.5 It argues that competition agencies should pay 1 Commissioner, US Federal Trade Commission. The views in this paper are the author’s alone and do not necessarily reflect the views of the Federal Trade Commission or its individual members. 2 International Competition Policy Advisory Committee to the Assistant Attorney General for Antitrust, Final Report 164 (2000) (noting a “heightened degree of international consensus among enforcers that cartels should be detected and prosecuted”); “Cartel Enforcement Goes Global: Jail, Fines, & Videotape”, 14 Antitrust 6 (Summer 2000) (expansion of efforts by competition enforcement bodies to attack cartels); Connor J. (2004): “Global Antitrust Prosecutions of Modern International Cartels”, 4 Journal of Industry, Competition and Trade 239 (reviewing antitrust penalties imposed on 167 international cartels discovered between 1990–2003); First H. (2001): “The Vitamins Case: Cartel Prosecutions and the Coming of International Competition Law”, 68 Antitrust Law Journal 711 (prosecution of vitamins cartel suggests broader international acceptance of anti-cartel norm). 3 The expansion of antitrust sanctions for cartel behavior is traced in Calkins S. (1997): “Corporate Compliance and the Antitrust Agencies’ Bi-Modal Penalties”, 60 Law & Contemporary Problems 127; Wils W. P. J. (2005): “Is Criminalization of EU Competition Law the Answer?”, 28 World Competition: Law and Economics Review 117. 4 On the development and operation of modern leniency programs, see Chen J. and Harrington J., “The Impact of the Corporate Leniency Program on Cartel Formation and the Cartel Price Path”, forthcoming in Ghosal V. and Sennek J., eds., Political Economy of Antitrust; Joshua J. (2000): “Leniency in U.S. and EU Cartel Cases”, 14 Antitrust 19 (comparing US and EU leniency programs); Klawiter D. (2000): “Corporate Leniency in the Age of International Cartels: The American Experience”, 14 Antitrust 13 (describing the development of US leniency policy); Motta M. and Polo M. (2003): “Leniency Programs and Cartel Prosecution”, 21 International Journal of Industrial Organization 347. 5 The paper draws heavily on proposals presented in Aubert C., Rey P. and Kovacic W., (2006): “The Effect of Leniency and Whistleblowing Programs on Cartels”, 24 International


572

12/10/07

13:14

Page 572


bounties to individuals, including employees of the cartel members, who provide information that leads to the successful prosecution of cartels. A bounty system could provide a source of cartel instability more acute and disruptive than existing leniency programs and, by raising the likelihood that government prosecutors would discover illegal collusive schemes, would deter their formation or impede their operation. The paper first describes the US Civil False Claims Act (CFCA), which relies heavily on bounties to attack fraud against the government. The paper then discusses the rationales for using private monitoring and prosecution to enforce public laws. Using the CFCA as an illustration, the paper considers how private participation interacts with other elements of a legal regime to inform a firm’s decision to obey the law. The paper concludes by examining the possible contributions to the deterrence of cartels by bounty systems that boost incentives for private monitoring.

B. An overview of the Civil False Claims Act In 1986, Congress amended the Civil False Claims Act (CFCA)6 and bolstered a mid-19th Century practice by enhancing incentives for private individuals to bring qui tam actions to challenge fraud against the government.7 The 1986 Amendments transformed the enforcement of rules involving the payment or collection of funds by the US government.8 Before 1986, qui tam actions under the CFCA amounted to several each year. From 1986 to 2000 alone, over 3000 qui tam cases were filed, resulting in over $3.5 billion in

Journal of Industrial Organization 1241; Kovacic W. (2001): “Private Monitoring and Antitrust Enforcement: Paying Informants to Reveal Cartels”, 69 George Washington Law Review 766; Kovacic W. (1993): “The Identification and Proof of Horizontal Agreements Under the Antitrust Laws”, 38 Antitrust Bulletin 81; Kovacic W. (1997): “Antitrust Policy and Horizontal Collusion in the 21st Century”, 9 Loyola Consumer Law Reporter 97, pp. 104–07. For another recent treatment of bounties as an inducement for cartel detection, see Spagnolo G., “Fines, Leniency and Whistleblowers in Antitrust”, forthcoming in Buccirossi P, ed., Advances in the Economics of Antitrust Law. 6 False Claims Amendments Act of 1986, Pub. L. No. 99-562, 100 Stat. 3153 (1986) (codified at 31 U.S.C. §§ 3729–31 (1994)). 7 Kovacic W. (1996): “Whistleblower Bounty Lawsuits as Monitoring Devices in Government Contracting”, 29 Loyola of Los Angeles Law Review 1799, at pp. 1801–07 (1996) (hereinafter Whistleblower Bounty Lawsuits) (discussing adoption of 1986 Civil False Claims Act Amendments). “Qui tam” is an abbreviation of the phrase “qui tam pro domino rege quam pro si ipso in hac parte sequitur,” which means “who sues on behalf of the King as well as for himself ”, Black’s Law Dictionary 1251 (6th edition, 1990). 8 See Kovacic W. (1998): “The Civil False Claims Act as a Deterrent to Participation in Government Procurement Markets,” 6 Supreme Court Economic Review 201, at pp. 217–23 (1998) (hereinafter Civil False Claims Act) (describing significance of 1986 CFCA reforms).


12/10/07

13:14

Page 573

Bounties as Inducements to Identify Cartels

573

recoveries to the U.S. Treasury and the payment of over $500 million in rewards to the qui tam informants.9

1. Prohibited conduct The CFCA creates civil liability for any person who knowingly submits a false claim for payment to the US government, knowingly uses a false statement to induce the government to pay a false claim, conspires to defraud the government to pay a false claim, or knowingly uses a false statement to decrease an obligation to pay money to the government.10 “Claims” include all requests for payment of money or property where the federal government pays for any of the money or property in question.11 “Knowing” conduct embraces actual knowledge of a falsehood, as well as “deliberate ignorance” or “reckless disregard” of the truth.12

2. Enforcement mechanism The CFCA delegates the decision to prosecute to the Department of Justice (DOJ) and to any “private person.”13 Courts have interpreted the law to give standing to a broad range of “relators”—individual citizens, employees of entities that receive federal funds, and private firms.

2.1. Jurisdictional limits on private suits The CFCA precludes private suits that are: —brought by a former or present member of the armed forces against a member of the armed forces arising out of such a person’s service in the armed forces; —brought against a member of Congress, a member of the judiciary, or a senior executive branch official if the action is based on evidence or information known to the government when the action was brought;

9 Stuart E. Schiffer, Deputy Assistant Attorney General, Civil Division, US Department of Justice, Statement Before the Subcommittee on Energy and Mineral Resources of the US House Committee on Resources Concerning Oil Royalty Overpayment False Claims Act Litigation (May 18, 2000), available at http://www.usdoj.gov/civil/speeches/schiffer051800.htm. 10 31 U.S.C. section 3729(a) (1994). 11 Id. at section 3729(c). 12 Id. at section 3729(b). 13 Id. at 3730(b).


574

12/10/07

13:14

Page 574


—based on allegations that are the subject of a civil suit or an administrative civil money penalty action in which the government is already a party; or —based on public disclosure of allegations in a criminal, civil, or administrative hearing, in a congressional, administrative, or Government Accountability Office report, hearing, audit, or investigation, or from the news media, unless the action is brought by the Attorney General or the person bringing the action is an original source of the information. The “original source” exception to the last of these limits applies to “an individual who has direct and independent knowledge of the information on which the allegations are based and has voluntarily provided the information to the government before filing an action . . . which is based on the information.”14

2.2. The government’s role in private suits The relator begins a qui tam action by serving DOJ with a copy of the complaint and “written disclosure of substantially all material evidence and information the person possesses.”15 The complaint is placed under seal with the US district court for at least sixty days and is not served on the defendant until sixty days elapse. The waiting period allows DOJ to investigate and evaluate the relator’s allegations. DOJ may seek extensions of the sixty-day waiting period. After reviewing the relator’s complaint, DOJ has four options: prosecute the action;16 let the relator proceed independently, while retaining the right to intervene later in the proceedings;17 ask the court to dismiss the action after notice to the relator and an opportunity for the relator to be heard;18 and, subject to court approval, settle the action following notice to the relator and an opportunity for the relator to critique the proposed settlement.19

2.3. Controls on abusive or baseless suits Beyond seeking to dismiss qui tam suits, DOJ can move to restrict the relator’s participation if DOJ prosecutes the case.20 This lets DOJ bar the relator from pursuing duplicative or fruitless evidentiary paths. The court also may limit the relator’s role if the defendant shows that unrestricted participation

14 15 16 17 18 19 20

31 U.S.C. at section 3730(e)(4)(B). 31 U.S.C. section 3730(b)(2) (1994). Id. at section 3730(b)(4)(A). Id. at section 3730(b)(4)(B). Id. at section 3730(c)(2)(A). Id. at section 3730(c)(2)(B). 31 U.S.C. section 3730(c)(2)(C) (1994).


12/10/07

13:14

Page 575


575

“would be for the purposes of harassment or would cause the defendant undue burden or unnecessary expense.”21 If DOJ does not prosecute and the relator proceeds alone, the defendant can receive reasonable attorneys fees and expenses if it prevails and the court finds that relator’s claim was “clearly frivolous, clearly vexatious, or brought primarily for the purposes of harassment.”22 Few decisions have imposed these sanctions.23

3. Sanctions The CFCA penalizes violations with (1) a civil penalty of at least $5000 and not more than $10,000 for each violation and (2) three times the actual damage the government has sustained from the defendant’s misconduct.24 The first measure is important where the defendant engages in many activities that constitute claims. For example, if a supplier submits numerous vouchers for payment under a government-funded program, each voucher tainted by fraud is a separate claim for which a civil penalty of $5000 to $10,000 can be assessed.

4. Rewards for the relator The CFCA entitles relators to a share of amounts recovered. If DOJ prosecutes, the relator receives 15–25% of the recovery, plus reasonable attorneys fees, costs, and expenses.25 The bounty varies according to “the extent to which the person substantially contributed to the prosecution of the action.”26 The court may limit the bounty to 10% or less if the action rested mainly on public disclosures of information (other than information provided by the relator) concerning “allegations or transactions in a criminal, civil, or administrative hearing, in a congressional, administrative, or Government Accountability Office report, hearing, audit, or investigation, or from the news media.”27 If DOJ does not prosecute the action, the bounty is 25-30% of the recovery.28

21

Id. at section 3730(c)(2)(D). Id. at section 3730(d)(4). 23 E.g., United States ex rel Herbert v. Nat’l Acad. of Sciences, 1992 WL 247587 (D.D.C. Sept. 15, 1992) (sanctioning qui tam plaintiff who stated invalid claim and sued to harass defendant). 24 31 U.S.C. section 3729(a) (1994). 25 Id. at section 3730(d)(1). 26 Id. 27 Id. 28 Id. at section 3730(d)(2). 22


576

12/10/07

13:14

Page 576


One reason for paying generous bounties is to compensate the informant for endangering her investment in her career and damaging her standing in the community.29 Even with safeguards against retaliation by an employer, informing likely precludes an employee’s advancement within her firm and, possibly, future employment in the industry. Informing also may expose an individual to social ostracism. A bounty arguably must be large enough to compensate the employee for liquidating her career and accepting the costs of social stigma that might result from informing. The CFCA curtails recoveries by relators who participate in the challenged misconduct. The relator gets nothing if she is convicted of criminal conduct that also violates the CFCA. If there is no criminal conviction, the court may reduce the bounty of a relator who plans and initiates violations. The reduction must account for the relator’s role “in advancing the case to litigation and any relevant circumstances pertaining to the violation.”30 The CFCA provides safeguards for employees who are punished by their employers because of the employees’ lawful acts in investigating, initiating, testifying for, or otherwise assisting in a qui tam action.31 The aggrieved employee may obtain “all relief necessary to make the employee whole,”32 including restoring seniority, two times back pay, interest on back pay, special damages, litigation costs, and reasonable attorneys fees.

C. Rationales for monitoring and prosecution by private parties A legal command’s impact depends on the quality of implementation,33 which encompasses various actions through which legal commands are applied to specific conduct. Four key steps stand out.34 Monitoring is the observation of conduct by parties subject to the law to detect violations and to collect proof of misconduct. Prosecution involves filing charges, presenting evidence before a judicial tribunal, proposing remedies, and negotiating settlements. Adjudication involves making decisions about guilt or innocence and selecting remedies. Punishment is the execution of remedies.

29

See Kovacic, Whistleblower Bounty Lawsuits, at p. 1819 (rationale for substantial bounties). 31 U.S.C. at section 3730(d)(3). 31 Id. at section 3730(h). 32 Id. 33 Posner R. (2001): Antitrust Law, 2nd edition, The University of Chicago Press, Chicago, at p. 266 (“It is not enough to have good doctrine; it is also necessary to have enforcement mechanisms that ensure, at reasonable cost, a reasonable degree of compliance with the law.”). 34 A complete list would add activities such as the physical apprehension and, in some cases, detention of alleged violators. For example, a jurisdiction could recruit private bounty hunters to find and apprehend fugitive offenders. 30


12/10/07

13:14

Page 577


577

Which individuals or institutions should perform these functions? A jurisdiction could make the state the monopoly provider of all these services. Under a purely public implementation scheme for a law forbidding cartels, government investigators would identify suspicious conduct by studying industry developments and would use compulsory process to gather documents and testimony that prove a violation. A government prosecutor would initiate and litigate charges against offenders, a public court would decide guilt or innocence, and public institutions would execute the remedy. By contrast to a state monopoly model, a law could engage private parties to perform all or some implementation functions. In practice, governments routinely enlist private parties to perform implementation tasks. Through various devices, public laws enlist private parties to supplement efforts by government bodies to detect and prosecute illegal conduct.35 Governments rely on voluntary disclosures of information by insiders,36 and some laws give private parties financial rewards to give public agencies information about violations of statutes or regulations.37 Other laws decentralize prosecutorial power by creating private rights of action.38 Some government programs hire private parties to operate prisons and, in effect, contract out the administration of remedies.39 As described above, the CFCA decentralizes implementation by giving private parties financial incentives to perform the monitoring and prosecution functions of law enforcement.

1. Rationales for private monitoring The enforcement of a legal command and the punishment of offenders give potential wrongdoers incentives to take precautions to avoid creating observable evidence of misconduct. Prospective offenders may plan and execute illegal behaviour by using subtle, covert measures that government prosecutors find difficult to detect. As the participants’ skill in concealing misconduct 35 See Landes W. and Posner R. (1975): “The Private Enforcement of Law”, 4 Journal of Legal Studies 1 (rationale for private enforcement mechanisms); Polinsky A. (1980): “Private Versus Public Enforcement of Fines”, 9 Journal of Legal Studies 105 (same). 36 An unsolicited tip from an Archer Daniels Midland official (Mark Whitacre) set in motion the DOJ inquiry that unveiled a global cartel of food additives producers and yielded massive criminal fines for the corporate participants and convictions for a number of individuals. The investigation and prosecution of the food additives cartel is recounted in Eichenwald K. (2000): The Informant: A True Story, Broadway Books, New York. 37 See, e.g., 26 C.F.R. section 301.7623-1 (2000) (allowing payment of bounties to individuals who provide information to the US Government leading to recovery of underpaid federal taxes). 38 Private rights to enforce antitrust laws once were unique to the United States. In recent years other jurisdictions have created private rights for antitrust and still others are considering whether to do so. Private rights also are common in US environmental protection and civil rights statutes. 39 E.g., Brown K. (2001): “New Prisons to Be Built By Private Companies”, Financial Times, October 2, 2001, at p. 9 (privatization of prison programs in England and Wales).


578

12/10/07

13:14

Page 578


increases, the likelihood of being prosecuted and punished for violations fades.40 Private monitoring can be an antidote to concealment. When confronted with the possibility of severe sanctions if their conduct is detected, a cartel’s members will exercise care to conceal the fact of their unlawful collaboration from any jurisdiction that will prosecute them A government enforcement body could rely exclusively upon its own personnel to detect the cartel. Public enforcement agencies could monitor an industry to detect patterns of behaviour that appear consistent with supplier collusion, and government investigators could obtain warrants to search business premises for documents, to wiretap telephones, or engage in other forms of surveillance. Monitoring by cartel insiders can be a more effective means of detection than observation by government officials alone. The chief virtue of private monitoring is that it gives monitoring tasks to individuals closest to the relevant information. An insider’s greater knowledge of the cartel’s activities may enable her to identify and assess relevant information at lower cost than external government monitors. For example, compared to external observation by a public enforcement agency, a cartel member’s employees may more easily detect that their employers meet covertly with their rivals to discuss plans to coordination behaviour. Where private monitoring enhances the likelihood and speed of detection, it may deter misconduct more powerfully than public monitoring. As discussed more extensively below, the likelihood of detection influences a party’s decision to obey or violate the law. By increasing the likelihood of detection, robust private monitoring can decrease the expected payoff from violating the law and boost the incentive to obey. Private monitoring can be destabilizing where the success of an illegal practice requires covert collective action. Adherence to a common scheme may be harder to achieve if each participant recognizes that employees entrusted with the cartel’s tasks, or employees who simply happen to learn of the cartel’s existence, may act as a private monitor. Will antitrust tribunals regard informants as credible witnesses in the trial of a cartel’s members? When a cartel’s participants refuse to settle charges of misconduct and contest the prosecutor’s allegations in court, they may depict informant witnesses as incredible because the informants’ testimony is shaped by the pursuit of personal financial gain. In a legal system that gives cooperating witnesses immunity from criminal sanctions, defendants often assert that the immunity recipient testified falsely to avoid or reduce her own punishment. Attacks on an informant’s credibility will be no less intense

40 To achieve optimal deterrence, deficiencies in the ability of prosecutors to detect violations can be offset by increasing punishments for offenders. Setting sanctions to account correctly for concealment requires hard judgments about detection rates. See Posner R. (2001): Antitrust Law, supra note no. 33, at pp. 269–73 (importance of probability of detection in setting punishment levels in antitrust cases).


12/10/07

13:14

Page 579


579

where the anticipated benefit of cooperation is a cash payment rather than the mitigation of criminal sanctions.41 Defence arguments that attack the testifying informant’s motives sometimes may resonate with judges and juries. Yet it is unclear that efforts to avoid liability by destroying an informant’s credibility inevitably or even regularly succeed.42 Nor is it likely that an informant can make useful contributions to the prosecution of a cartel only if she testifies at trial. Informing can facilitate successful enforcement even if the informant never appears in court. The informant may provide internal company records that document the cartel’s operations and establish how the conduct has injured consumers. An informant may help the government develop a blueprint for the investigation by describing the cartel’s organization and operations, by identifying key participants, and suggesting records that the antitrust agency might obtain through compulsory process. Some informers may be willing to cooperate with government prosecutors by recording comments by cartel members. A non-testifying informant sometimes may help the government amass such conclusive evidence of illegality that the participating companies and individuals agree to plead guilty and forego trials.

2. Rationales for private prosecution At least two reasons can justify the delegation of prosecutorial power to private individuals or institutions. The first is that private rights can vindicate legal commands where public institutions have weak incentives to challenge violations. For example, when it detects supplier fraud, a public purchasing body might not prosecute because it fears that a scandal will endanger funding for a favoured program, or because the supplier has captured its regulator. By bolstering private rights of action, the 1986 CFCA Amendments reduced reliance on public bodies that might not attack government contractor 41 An informer seeking a cash reward might be accused by defendants of having a still greater incentive to distort testimony if the reward is paid only if the prosecutor obtains a conviction. One could argue that a testifying informant who is promised a payment contingent upon the defendant’s conviction would face greater temptations to misrepresent events than a cooperating witness who receives criminal immunity (or a payment of cash) without regard to the success of the prosecution. 42 The individuals prosecuted in the US cartel cases involving art auction houses and lysine made the credibility of informants a key part of their defense and were convicted nonetheless. See United States v. Andreas, 216 F.3d 645, 658–62 (7th Cir.) (affirming convictions of Archer Daniels Midland executives for participation in food additives cartel; rejecting argument that misconduct of government informant, Mark Whitacre, tainted guilty verdict); Blumenthal R. and Vogel C. (2001): “Ex-Chief of Sotheby’s Is Convicted of Price-Fixing”, New York Times, December 6, 2001, at A6 (attacks on credibility of government’s cooperating witnesses, who received dispensations from criminal prosecution, failed to sway jury).


580

12/10/07

13:14

Page 580


misconduct aggressively. The amended law counteracted the possibility of lax public prosecution of misconduct in two ways. First, it gave DOJ greater access to fraud-related information that procurement agencies might suppress or, owing to half-hearted monitoring, might not discover. The CFCA compels the private enforcer to serve DOJ with the complaint and supporting data. This tripwire alerts DOJ to misconduct that purchasing agencies might not independently reveal to the Department. Second, by deputizing private parties to sue on the government’s behalf, the CFCA decreased the likelihood that meritorious cases will languish because the purchasing agency or DOJ—due to sloth, negligence, or deliberate efforts to protect specific programs from needed scrutiny—declines to investigate and attack apparent fraud. A second reason to create private rights is to use private initiative to supplement public law enforcement resources. Private suits not only can increase total enforcement of a legal command but also may be more efficient enforcement tools. Given a choice between spending resources on monitoring by private parties or monitoring by public officials, one might prefer private monitoring when private parties can identify misconduct at lower cost than public employees.43

3. Models of participation by private parties Law enforcement systems vary in how they elicit private parties to identify and prosecute misconduct. The principal models of private participation are described below.

3.1. Monitoring as pure volunteerism One model of private participation in law enforcement is to rely solely on volunteers whom the government neither protects against retaliation nor compensates for giving information. In this model, private parties voluntarily to supply information that helps identify misconduct and establish guilt in the judicial process. To attract volunteers, government prosecutors might appeal to the private party’s sense of civic duty or patriotism. Some private actors may act as monitors without assurances of protection or compensation. Where the potential costs to the private monitor are great, a system that relies on volunteerism alone may elicit low levels of private monitoring. Informing can expose the private party to physical danger, professional retaliation, and social stigma, and it may require an expenditure of time and emotional capital to detect behaviour, share information with 43 See Becker G. and Stigler G. (1975): “Law Enforcement, Malfeasance, and Compensation of Enforcers”, 3 Journal of Legal Studies 1, at p. 15 (private enforcement may be more efficient that public enforcement).


12/10/07

13:14

Page 581


581

prosecutors, or testify in judicial proceedings. Many potential monitors may refuse to incur these costs if the government declines to insure against risks or pay some of the monetary benefits that flow from a private party’s contributions to the detection and prosecution of violations.

3.2. Monitoring with compensation Law enforcement systems can encourage private monitoring by compensating such effort. One form of compensation is to protect informants against physical harm or to bar retaliation by employers. The government could promise to protect private monitors against physical attacks by wrongdoers and shield private monitors from retribution by employers.44 By such protection, the government insures against some adverse consequences that may befall private monitors if they inform. The public prosecutor could supplement these measures by giving amnesty or immunity from prosecution to private monitors who have engaged in illegal behaviour. This is the strategy of antitrust policies that give leniency to individual or corporate wrongdoers. Or a law enforcement body could give private monitors cash for providing information. Payments could consist of a predetermined sum announced publicly,45 or the government could negotiate cash rewards with each private party who offers to supply information.46 Alternatively, the payment could be set as a percentage of amounts ultimately recovered by public prosecutors from wrongdoers.47 To encourage the swift detection of misconduct, a private monitoring system might emphasize priority in providing information that leads to the successful prosecution of wrongdoers. Eligibility for rewards might be restricted to the first individual to come forward. A difficulty with a strict rule of priority is that the first informant might not be the best source of information; a second informant might provide information that advances a government investigation more than the contributions of the initial source. A compromise between priority and completeness would be to provide the largest reward to the first informant but also give a substantial bounty to later informants who supply valuable data.48 Whether the monitoring mechanism compensates only the first mover or permits rewards to multiple parties, the body responsible for assigning bounties (e.g., an enforcement bureau or court) will face 44 One safeguard consists of programs that relocate witnesses and give them new identifies in return for cooperating with prosecutors. 45 An example would be a public announcement that said the government will pay a specific cash award for information leading to the apprehension of an offender. 46 Here the government would announce a general policy of paying rewards to informants and would delegate authority to negotiate the amount of bounties to law enforcement officials. 47 This is the approach used in the CFCA. 48 The CFCA does not limit rewards to the first mover. It sets a cap on the bounty to be paid to all informers and lets the court allocate rewards according to the relative contributions of each.


582

12/10/07

13:14

Page 582


disputes concerning which informant achieved priority or, if multiple awards are possible, how to weigh the contributions of each informant.

3.3. Monitoring with compensation and the power to prosecute The first two forms of monitoring described above assume that a public agency alone has power to prosecute violations. If the public agency declines to prosecute, a prospective private monitor who desires insurance against retaliation or cash payments must negotiate such arrangements with the prosecutor as a condition of revealing the information initially. Without such an agreement, a prospective monitor who anticipates recovering a percentage of fines or penalties ultimately paid to the government bears the risk that the monopolist public enforcement body will decline to prosecute, leaving the monitor with no compensation. A more robust model of private participation would compensate private parties for monitoring (via insurance, immunity from prosecution, or cash payments) and give them the right to prosecute. The delegation of prosecutorial authority occurs in three basic forms. The law creating the private right of action might: —Require the private party to inform the public prosecutor of apparent misconduct and give the prosecutor an option to undertake the matter or allow the private party to proceed alone; —Adopt the notification requirement described immediately above and authorize the public prosecutor to seek dismissal of any private suit, regardless of whether the prosecutor assumes the case; or —Allow the private party to sue independently and require no prior consultation with or subsequent participation by the public enforcement body. The first model resembles private rights under US statutes banning employment discrimination,49 the second describes the CFCA, and the third governs private suits under the U.S. antitrust laws.50 Laws that combine private monitoring with private prosecution vary in the quality of insurance and monetary compensation they offer the private monitor. Private enforcement systems sometimes compensate prevailing plaintiffs for their attorneys’ fees and litigation costs.51 More powerful incentive schemes (such as the US antitrust laws) entitle the prevailing plaintiff to attorneys fees and costs and to monetary recoveries based upon the harm caused by the wrongdoer.52 The CFCA not only provides a share of amounts 49

42 U.S.C. section 2000e-5(f)(1) (1994). 15 U.S.C. sections 15, 26. Under the third model, a public enforcement body usually retains the right to advise the court in an amicus capacity and may have the power to intervene by right. 51 This is the mechanism used in the U.S. Clean Air Act, 33 U.S.C. section 1365(a) (1994). 52 15 U.S.C. section 15. 50


12/10/07

13:14

Page 583


583

recovered by the US Treasury and mandates payment of the relator’s attorneys fees, but also supplies anti-retaliation safeguards.

4. Defining the universe of eligible private monitors and prosecutors Beyond determining the scope of the private party’s role in monitoring or enforcing the law, designing a private participation system requires one to define the universe of eligible participants. A private participation mechanism could make any individual eligible to qualify for rewards as monitors or prosecutors. For a hypothetical corporate wrongdoer, a system that broadly defined eligibility might entertain monitoring by the firm’s employees, its customers, its suppliers, and individual government employees engaged in regulating its affairs. In practice, private enforcement systems often screen potential participants according to several criteria.

4.1. Employment status Some private enforcement systems limit the eligibility of certain government employees. US Internal Revenue Service tax regulations that authorize bounties for private monitoring of tax delinquencies deny eligibility to employees of the Department of the Treasury or any other federal employee who acquired information relating to the underpayment of taxes in the course of official duties.53 Such limits seek to discourage perverse conduct by government bodies and their employees. Bounties might provide a greater incentive for public officials to threaten to attack comparatively minor violations54 or to challenge benign conduct as a way to extract side payments from the violator. Such threats are more credible if the government employee can sue unilaterally rather than having to gain the approval of superiors to whom the alleged wrongdoer can appeal to discourage the prosecution of legitimate conduct or behaviour that amounts to an insignificant violation.

4.2. Nature of and proximity to the injury Many decentralized enforcement systems determine eligibility by examining the nature of the claimant’s injury and proximity to the misconduct. Judicially created antitrust doctrines impose formidable standing limitations upon private plaintiffs and generally deny standing to employees who allege 53

27 C.F.R. section 301.7623-1 (2000). By “minor” violations I have in mind harmless departures from stated regulatory requirements. 54


584

12/10/07

13:14

Page 584


that their employers have engaged in antitrust violations.55 The private cause of action for employment discrimination limits private standing to individuals who have wrongfully been denied employment.56 Public institutions responsible for enforcing these laws do not pay rewards to employees who provide information about their employers’ illegal hiring practices, and the statute does not give a right of action to third parties who observe unlawful behaviour.57

4.3. Participation or acquiescence in illegal conduct Some law enforcement systems restrict participation in private enforcement by individuals who engaged in illegal conduct or passively acquiesced in its execution. For example, the CFCA precludes eligibility for persons who are convicted of the illegal behaviour and reduces (but does not eliminate) the reward for those who helped plan the misconduct. Such safeguards seek to curb incentives for individuals to cooperate in illegal schemes. Yet many systems designed to detect crimes elicit assistance from those engaged in illegal behaviour. Prosecutors in criminal cases give immunity to individuals in return for their promise to provide evidence that assists in apprehending or convicting other wrongdoers. The US antitrust regime grants criminal amnesty to a cartel member that discloses the cartel’s existence and is not the cartel’s ringleader. The US system gives full amnesty from criminal prosecution only to the first party to inform. Immunity from prosecution, not a cash payment, constitutes the informant’s reward.

D. How private monitoring and enforcement affect the impact of a legal command The impact of any legal command depends heavily on five variables:58 55 The ban on employee standing assumes that employees of antitrust violators benefit from the firm’s anticompetitive conduct because the firm gives some of its monopoly profits to workers via better wages and amenities. See Apperson v. Fleet Carrier Corp., 879 F.2d 1344, 1352 (6th Cir. 1089) (standing denied to employees; employees could have been economic beneficiaries of violation). 56 42 U.S.C. section 2000e-5(f)(1) (1994) 57 Suppose that a company’s hiring committee refuses to offer a job to a candidate on grounds that law forbids employers to consider. The excluded candidate would have standing to sue under U.S. law, but a hiring committee member who objected to the committee’s decision would not. 58 For a further elaboration of this taxonomy, see Gavil A., Kovacic W. and Baker J. (2002): Antitrust Law in Perspective: Cases, Concepts and Problems in Competition Policy, Thomson West, American Casebook Series, at pp. 923–27; Kovacic W., (2006): “Competition Policy and Cartels: The Design of Remedies”, in Cseres K., Schinkel M. P. and Vogelaar F., eds., Criminalization of Competition Law Enforcement, Edward Elgar, Cheltenham.


12/10/07

13:14

Page 585


585

—The substance of the command (e.g., a prohibition upon cartels). —The legal standards by which an adjudicatory tribunal decides whether the defendant has transgressed the command. —The likelihood that violations will be detected. —The likelihood that detected violations will be prosecuted. —The power of the sanctions imposed for violating the command. A legislature can determine a law’s impact by adjusting any of these variables. It can expand or curb the law’s substantive requirements; bolster or weaken the legal standard that plaintiffs must satisfy to establish culpability; increase or reduce the probability that violations will be observed, strengthen or weaken the means for prosecuting violations; or augment or diminish the consequences that flow from a violation. Any single adjustment alters the law’s significance. The creation or enhancement of private monitoring and prosecution affects the third and fourth variables by raising the probability that violations of the law will be detected and prosecuted. Even if one holds the other three variables (substantive requirements, legal standard, and sanctions) constant, giving private parties stronger inducements to serve as monitors and prosecutors will tend to increase the incentives of affected actors to comply with the law.

1. Criteria for expanding private participation in implementing a law Three major factors are relevant to deciding how much to engage private parties in implementing a legal command. The first is one’s judgment about the wisdom of the underlying substantive commands to be enforced. The sensibility of implementing a law more vigorously, such as by eliciting greater private involvement in detecting or prosecuting violations, depends on one’s confidence that the command forbids truly harmful behaviour. The more egregious the misconduct, the greater the potential value society might derive from expanding implementation efforts through private participation. Careful assessment of the wisdom of the substantive rules to be enforced is important to evaluating an expansion of private participation where the existing legal regime relies heavily on the exercise of discretion by public prosecutors. In drafting legal commands, legislators inadvertently may condemn behaviour that the legislature regards as acceptable. Or experience with enforcing a law may suggest that selected prohibitions contradict existing social norms or contemporary conceptions of sound policy. In these and other circumstances, a public prosecutor might choose not to challenge conduct that violates a nominal legal command. Robust private participation, especially independent rights of action that eliminate a public prosecutorial monopoly, reduce or eliminate the ability of government officials to use prosecutorial discretion as a non-legislative tool


586

12/10/07

13:14

Page 586


for altering the law. Private prosecutors motivated by strong rewards may try to enforce all existing rules aggressively, not simply the rules that public prosecutors currently deem sensible. Thus, a regime with private rights of action must ensure that the rules subject to private enforcement are not overinclusive in their definition of forbidden conduct. Beyond analyzing the soundness of the substantive law, a second factor affecting the choice among low-powered and high-powered public participation should be an evaluation of why existing public enforcement mechanisms are inadequate. Suppose that the source of weak enforcement is the incompetence, corruption, or capture of public prosecutors. In such circumstances, one might prefer to establish private rights. Private enforcement can offer an important implementation safeguard if there is a serious possibility that public prosecutors will default in executing their duties. Where the source of inadequate public enforcement is weak access to evidence, different forms of enhanced private participation would be appropriate. The payment of rewards for providing information—without conferring a private right to prosecute—may suffice. Public prosecutors with a demonstrated proclivity to prosecute cartels are likely to use information provided by an informant effectively. Thus, decisions about how to expand private participation in implementing the law requires an assessment of the quality of existing enforcement institutions, including the capability, intentions, and motivations of public prosecutors. A third factor for consideration is the adequacy of safeguards against mistaken or perverse private efforts to detect or prosecute violations.59 A private participation mechanism might impose stronger safeguards—e.g., by giving a public prosecutor a greater gate keeping function—where the possibilities for error or counterproductive behaviour by private parties are substantial. In the case of the U.S. CFCA, injecting powerful forms of private participation into the regulation of contracts between the government and its suppliers without strong safeguards against error, opportunism, and malice may damage the government’s ability to attract an optimal range and quality of contractors.60

2. Relying on judicial adjustment Where the legislature authorizes powerful forms of private participation, courts have been known to interpret statutes in ways that seek to reduce the 59 On the importance of accounting for potential adverse side effects in creating a private rights, see Marshall R., Meurer M. and Richard J.-F. (1991): “The Private Attorney General Meets Public Contract Law: Procurement Oversight by Protest”, 20 Hofstra Law Review 1 (analyzing benefits and costs of bid protest oversight in government procurement). 60 See Kovacic W., Civil False Claims Act, at p. 235 (describing how excessively broad CFCA oversight can discourage desirable levels of private participation in public procurement markets).


12/10/07

13:14

Page 587


587

adverse consequences of private oversight. Judicial retrenchment of modern antitrust doctrine in the United States has resulted substantially from rulings that have imposed restrictive standing and injury tests for private plaintiffs.61 Automatic trebling of damages in private antitrust cases in the United States may have induced courts to toughen liability tests and may have increased the willingness of judges to grant defendants’ motions to dismiss or motions for summary judgment.62 These developments have been possible because the open texture of the principal US antitrust statutes gives courts considerable discretion to apply restrictive procedural screens and to alter requirements governing liability and standing.63 Relying on judicial interpretation to limit private enforcement of overly broad statutory commands presents risks. Courts might try to compensate through distorted constructions that forestall the socially harmful private claims but also impede the pursuit of valid private and public cases. Failures in the legislative design of a private participation mechanism can create irresistible pressure for courts to impose their own solutions. Strained interpretations to restrict participation by some parties can set legal principles that unduly bind worthy private claimants as well.

E. Expanding private participation through bounty hunting As the framework from the preceding section suggests, a jurisdiction seeking to increase efforts to defeat cartels can do so in five basic ways: (1) strengthen substantive prohibitions against collusion (e.g., treat certain acts as illegal per se); (2) lighten the evidentiary burden that plaintiffs must bear to prove a violation (e.g., treat parallel conduct as an indication of unlawful agreement); (3) take measures to increase detection (e.g., give cartel insiders stronger inducements to inform); (4) increase the likelihood that misconduct, once discovered, will be prosecuted (e.g., create private rights of action); (5) boost sanctions for infringements (e.g., imprison culpable 61 The restrictive impact in private suits of judicially developed standard and injury doctrines is analyzed in Page W. (1989): “The Chicago School and the Evolution of Antitrust: Characterization, Antitrust Injury, and Evidentiary Sufficiency”, 75 Virginia Law Review 1221. 62 See Breit W. and Elzinga K. (1986): Antitrust Penalty Reform: An Economic Analysis, American Enterprise Institute for Public Policy Research, Washington DC, at p. 62 (courts have reacted to “inefficiencies of private actions” by imposing “stricter standards regarding liability, standing, and damage estimation”); Calkins S. (1988): “Equilibrating Tendencies in the Antitrust System, With Special Attention to Summary Judgment and Motions to Dismiss”, in White L., ed., Private Antitrust Litigation, MIT Press, Massachusetts, at pp. 185, 229 (“One of the ways in which courts have adjusted to the treble damage remedy is by being relatively more willing to keep cases from going to trial.”). 63 Kovacic W. (1992): “The Influence of Economics on Antitrust Law”, 30 Economic Inquiry 294, at p. 295.


588

12/10/07

13:14

Page 588


individuals). This section discusses how a bounty system could increase the rate at which cartels are detected and, by making detection more likely, deter the formation of cartels or impede their operations. At first glance, the widespread adoption of leniency programs might seem to exhaust possibilities for improving detection rates. This paper makes two assumptions about the capacity of leniency systems to take the process of detection to its furthest boundaries and ultimately supply a decisive, lasting deterrent to cartels. First, the history of competition policy efforts to control cartels—reflected in studies of old and newer cartels, alike—illuminates the extraordinary ingenuity of cartels and their remarkable capacity for adaptation.64 The attention given to the obstacles to successful coordination has obscured the creativity of cartel members in surmounting such barriers when the gains to be had are substantial. This history gives good reason to fear that any single system of defences, no matter how impressive at the moment, can become a Maginot Line as the arms race between cartels and government prosecutors inspires cartels to devise countermeasures. This paper also assumes that bounties can disrupt cartel coordination more powerfully than leniency. With a system of bounties, each company insider who becomes aware of the cartel is a potential informant. The firm’s managers not only must worry about the possible moves of their cartel partners, but about the actions of each employee who is enlisted to assist in the illegal plan or who learns about the plan inadvertently. Each meeting or conversation or act in furtherance of the cartel is beset with doubt about whether each knowledgeable individual can be trusted to remain faithful to the illegal enterprise, or whether employees who merely observe irregular activities will keep their suspicions to themselves. Compared to leniency, a bounty system denies the managers of the cartel firms control over the timing of when to reveal their illegal operations to prosecutors. The choice of when to inform no longer is governed simply by conjectures about whether other cartel members will seek leniency. With bounty hunting, the managers of the cartel participants confront the possibility that any employee of any cartel member with knowledge of the cartel will reveal the illegal arrangements to prosecutors. In any event, bounty hunting and leniency can be viewed as complements rather than substitutes. Because it gives each employee the capacity to inform, a bounty hunting system may accelerate recourse by cartel members to leniency. 64 See, e.g., Grossman P., ed. (2004): How Cartels Endure and How They Fail: Studies of Industry Collusion, Edward Elgar, Cheltenham (essays on operation of cartels); Connor J. (2001): Global Price Fixing: Our Customers Are Our Enemies, University of Purdue; Expert Report of R. Douglas Bernheim, M.D.L. No. 1285, In Re Vitamins Antitrust Litigation, Misc. No. 99-0197 (TFH) (May 24, 2002); Evenett S., Levenstein M. and Suslow V. (2002): “International Cartel Enforcement: Lessons from the 1990s”, in OECD Global Forum on Competition 2002; Genesove D. and Mullin W. (2001): “Rules, Communication and Collusion: Narrative Evidence from the Sugar Institute Case”, 91 American Economic Review 379; Marshall R., Marx L. and Raiff M. (2005): Cartel Price Announcements: The Vitamins Industry, Duke University Working Paper.


12/10/07

13:14

Page 589


589

1. Exploiting the cartel’s operational vulnerabilities The adoption of legal standards condemning cartels and the establishment of severe sanctions for violators has motivated cartel members to take ever greater precautions to avoid detection. One means to this end is to adopt coordination strategies that are less likely to arouse suspicion.65 For example, a coordination rule that allocates market shares is superior to an allocation of customers or territories because the latter arrangements require the cartel member to instruct its sales staff not to make sales to specific buyers or within specific areas. Rather than give this suspicious guidance to a sales force, cartel members instead can allocate market shares, tell sales personnel to accept orders from any source, and make side payments to ensure that market share targets are honoured. In executing a coordination strategy, cartel members also will strive to avoid creating evidence that readily would prove an unlawful agreement. As the power of an enforcement scheme increases, cartel members will try not to generate corporate records that a prosecutor could use to show illegal concerted action. These precautions can limit the amount of direct evidence—testimony or documents—that show the existence of concerted action. In addition to measures to curb the creation of written or electronic records, a cartel member can curtail the distribution of information about the collusive arrangement within its own organization. Attempts to control the flow of information on a “need to know” basis can reduce the number of company insiders who have a full, informative picture of the cartel. Even with precautions in structuring operations and in managing information, the cartel inevitably must rely to some degree on written or electronic records and must share information with a number of individuals. For example, despite efforts to avoid detection, the members of the vitamins and food additives cartels found it necessary to employ complex coordination mechanisms. In vitamins, for example, coordination could be achieved only through a careful accounting of actual sales in order to match sales to the market share targets that supplied the foundation for cooperation among the vitamins producers. Additional records were maintained to document the transfer payments—often masked in the form of buy-sell agreements or asset transfers— that were necessary to compensate cartel members whose actual sales lagged behind the market share targets. The payment of bounties to informants could place extreme pressure on some of the most vulnerable joints of the cartel’s operations by making the execution of each cartel management task an opportunity for various individuals—managers, sales people, secretaries, accountants, administrators—to 65 See Kovacic W., Marshall R., Marx L. and Raiff M. (2006): “Bidding Rings and the Design of Anti-Collusion Measures for Auctions and Procurements”, in Dimitri N., Piga G. and Spagnolo G., eds., Handbook of Procurement, Cambridge University Press 381.


590

12/10/07

13:14

Page 590


acquire information about the existence and operation of the cartel. No cartel of any consequence can avoid the need to formulate and execute operational plans. Even with policies that seek to limit the internal dissemination of such information, to select seemingly loyal employees for participation in the cartel’s affairs, and to provide special compensation to employee participants, the cartel is unlikely to bond everyone to silence. To the degree that the firm must pay larger salaries or bonuses to individual employees to discourage them from seeking bounties for informing, the costs of the cartel’s operations increase and the rewards to the firm for illegal coordination decline.

2. The design of a system of bounties to undermine cartels A jurisdiction can design a bounty hunting system that exploits the vulnerabilities described above in a manner that is compatible with larger goals of competition policy and sound public administration. The basic ingredients of an anti-cartel bounty hunting system are outlined below.

2.1. The desirability of the substantive rule Rewards for private participation are likely to generate the largest social benefits when incentives focus private initiative upon detecting or prosecuting unambiguously harmful conduct. Most commentators regard the enforcement of stringent rules against cartels as antitrust’s most important positive contribution to economic efficiency.66 By contrast, the prosecution of antitrust prohibitions against abuses of dominance, mergers, and distribution practices pose greater analytical ambiguity and arouse greater debate, particularly when private lawsuits are the means for implementing such commands.67 A nation contemplating an expansion of private participation might best begin by focusing private activity exclusively upon the enforcement of rules governing antitrust’s most serious offences—namely, hard core horizontal restraints. 66 See Bork R. (1978): The Antitrust Paradox, Basic Books, New York, at p. 263 (praising the per se ban against horizontal price-fixing and market divisions; concluding that “[i]ts contributions to consumer welfare over the decades have been enormous”); OECD (1998): Recommendation of the Council Concerning Effective Action Against Hard Core Cartels, OECD C(98)35 (1998) (hard core cartels are “the most egregious violations of competition law”), available at http://www.oecd.org/daf/clp/Recommendations/Rec9com.htm. This view is not universal—see Jenkins W. (2001): “Silly Trial, Silly Law,” Wall Street Journal, December 12, 2001, at A19 (criticizing U.S. prosecution of art auction cartel and attacking law against horizontal price fixing). 67 See Snyder E. and Kauper T. (1991): “Misuse of the Antitrust Laws: The Competitor Plaintiff”, 90 Michigan Law Review 551, at pp. 596–98 (proposing limits on capacity of competitors to pursue various antitrust claims).


12/10/07

13:14

Page 591


591

2.2. Analyzing the causes of inadequate enforcement The second criterion seeks to pinpoint the inadequacies of the implementation status quo. In some jurisdictions, the main obstacle to executing an effective anti-cartel policy is the government prosecutor’s inability to obtain information that reveals the existence of cartels and aids their prosecution. Where the problem is poor access to information and not a deficiency in the commitment, resources, or effort by public agencies to implement the law, the appropriate solution may be to strengthen incentives for private parties to act as monitors. In such circumstances, it is not necessary to go beyond creating a system of bounties for detection and take the further step of creating private rights of action for insiders. A diligent, competent, and adequately funded public enforcement body is likely to prosecute cartels identified by information supplied by the insider. Experience in the United States provides an example of this scenario. Since the mid-1930s, DOJ has actively prosecuted cartels. Despite variations over time in the mix of government initiatives, anti-cartel measures have formed the core of modern US federal enforcement policy.68 Given the history and norms of US public enforcement policy, one can confidently expect DOJ to prosecute illicit conduct identified by private informants. For DOJ and for other competition agencies with similar anti-cartel preferences, there is no need to give the private informant an independent, action-forcing right to bring lawsuits if the public prosecutor has declined to pursue the matter. In these conditions, the public agency is likely to desist only if the information provided failed to identify a cartel. An expanded bounty system, consisting of rewards for informing and privates right of action, may be desirable in jurisdictions where the public authority is unlikely to intervene even when an informant supplies information that reveals a cartel. Various causes—among them, corruption, sloth, weak technical capacity, overpowering political pressure, or inadequate resources—might lead a public agency to default. If the public prosecutor is likely to ignore an informant’s data, it may be necessary to give the informant a reward and a right to sue if the government does not.

2.3. Safeguards against counterproductive private participation Experience with designing and implementing the CFCA in the United States suggests possible difficulties with a system of bounties for private monitoring.69 68 See Kovacic W. (2003): “The Modern Evolution of U.S. Competition Policy Enforcement Norms”, 71 Antitrust Law Journal 377, at pp. 417–25 (describing development of US enforcement program against cartels). 69 See Kovacic, Civil False Claims Act, at pp. 227–35 (describing possible adverse effects of private enforcement under CFCA); Kovacic, Whistleblower Bounty Lawsuits, at pp. 1825–41 (describing problems with CFCA whistleblowing system).


592

12/10/07

13:14

Page 592


Some of the more important potential problems and possible solutions are identified below. The Class of Eligible Potential Informants is Excessively Broad. In adopting the 1986 Amendments to the CFCA, Congress inadequately delimited the categories of individuals who might serve as qui tam relators.70 The CFCA and its legislative history do not preclude qui tam actions by government employees or advisors (such as attorneys) who have confidential relationships with the firm to be monitored. Attorneys who are willing to breach a duty of confidentiality (and be barred from the practice of law) conceivably could recover a CFCA bounty for revealing their clients’ misconduct. A bounty system can solve this problem by disqualifying certain categories of informants, including government employees and the alleged violator’s confidential advisors. Informants May Erroneously Perceive the Existence of a Cartel. Insiders may misapprehend the significance of behaviour they observe—for example, by erroneously concluding that a company engaged in conduct properly associated with a legitimate joint venture is participating in a cartel. Faulty insider perceptions could lead public authorities to expend resources in chasing down fruitless leads or, if there is a private right of action for informants, such errors could cause companies to spend resources defending against baseless claims. Several antidotes to the problem of informant mistakes are available. Take the case of a jurisdiction that has a capable, well-motivated public enforcement body and creates rewards for informing, but does not give the informant a private right of action. Although it may spend some resources on pursuing the mistaken informant’s suggestions, the government prosecutor is likely with modest effort to test the basis for the informant’s suspicions. As long as the government alone has the right to act upon information provided by the informant, there is only a limited risk that the informant’s error will inspire a challenge to benign conduct. Note that the problem of false leads already exists when, as today, government prosecutors receive and act upon tips from mere volunteers. Where the jurisdiction creates rewards and a right of action for private insiders, the informant might proceed with a matter that the prosecutor correctly perceives to be baseless. Perhaps the best cure for false positives in such matters is to give the government prosecutor the right (as does the CFCA) to intervene in the private case and ask the court to dismiss the private suit. The government may not be able to forestall the initiation of a mistaken cartel case, but it would have an express statutory basis to seek dismissal of a matter that it has reviewed and found wanting. A more drastic safeguard would be to impose a “loser pays” rule for costs and attorneys fees if the government enforcement body has decided not to pursue the case. 70 See Kovacic, Whistleblower Bounty Lawsuits, at pp. 1834–38 and 1846–48 (describing problems associated with the failure to preclude standing for government employees under CFCA).


12/10/07

13:14

Page 593


593

Informants May Deliberately Fabricate Claims. Some individuals might use the private participation scheme vexatiously. A company insider might knowingly give false information to the government body or initiate a case based upon a fabrication of events. For example, an employee facing the loss of a job due to scheduled layoffs or a termination for good cause might take (or threaten to take) such measures to elicit more favourable severance terms from an employer. If the bounty scheme only pays rewards for detection and gives the insider no private right of action, the bounty system is unlikely to increase incentives for vexatious behaviour beyond those existing under the status quo. Under existing US enforcement policy, which relies partly upon voluntary tips about possible misconduct, company insiders already could give government officials false information. A bounty system is unlikely to increase the number of deliberately frivolous leads given to the government. A private participation mechanism that combines rewards for informing and a private right of action may give malevolent insiders greater capacity to credibly promise to impose costs on the employer by threatening pursue a lawsuit that the employer will have to spend resources to defend. To address this possibility, the private participation system could provide sanctions (as does the CFCA) for filing vexatious or frivolous suits. Allowing the government to intervene to seek dismissal of baseless suits also provides a means for screening out such actions. Bounties Might Decrease Incentives to Oppose Misconduct. A reward scheme that is tied to the amounts ultimately recovered by the government might weaken the incentive of cartel insiders to oppose or report misconduct. Insiders might conclude that, by allowing misconduct to continue, the total size of recoverable penalties will increase and will boost the award. Rather than oppose or report misconduct at early stages of the cartel, an insider might wait until the harm increases. A countervailing factor is the use of a rule that provides awards exclusively or disproportionately to the first person to inform. An insider who observes misconduct and delays informing runs the risk that another insider will act first and capture all or the largest share of the reward. Another scenario for concern involves the possibility that an insider might encourage or facilitate misconduct in the hope of ultimately capturing a reward. One antidote is a rule that denies any compensation to those who play a central role in orchestrating the misconduct.71 For individuals who play a secondary or peripheral role in the misconduct, the solution would be for the government to reduce (but not eliminate) the informant’s share in the proceeds.72 71

By comparison, many leniency programs deny eligibility to the cartel ringleader. By comparison, the CFCA permits the relator’s share to be reduced where the relator was a participant in the misconduct. 72


594

12/10/07

13:14

Page 594


3. Operational elements of a private monitoring system A system of rewards for private parties who inform on cartels would pay a percentage of amounts ultimately recovered by the government where the informant’s cooperation contributes significantly to the identification and successful prosecution of a collusion offence. Such a bounty scheme would have the following elements: Presentation of Information. The informant would be required to present information about suspected collusion to the public competition authority. In systems that create no independent right of action for the informant, the government would have total discretion to act upon the information or withhold prosecution. In systems that give the government a right of first refusal and then allow the insider to proceed independently, the procedure under the CFCA might be followed. The informant would simultaneously file an action under seal with the court and give the information to the government. After a set period of time, if the government takes no action, the informant could proceed with the private suit. Any system that creates a private right of action should give the government authority to petition the court to dismiss groundless suits. Calculating Rewards. Rewards for informants should be a significant percentage—the CFCA maximum is 30%—of all criminal fines or civil penalties recovered by the competition authority. The reward scheme could provide the entire bounty to the first to inform, or could give a disproportionately large share of the total award to the first informant and divide the rest among subsequent informants who provide valuable information. The amount paid to any informant could be adjusted downward where the informant has participated in executing the scheme. Individuals who orchestrated the illegal behaviour would be ineligible. Administering the Reward Mechanism. The public enforcement agency would make the initial decision about the amount of the reward to be paid to individual informants. The agency’s decisions would be subject to judicial review. Informants would be represented by counsel, who would be entitled to payment (by the violators) of reasonable attorneys fees and costs for their assistance to the informer. Retaliation Safeguards. The statute creating the private monitoring mechanism would protect informants against retaliation by their employers by forbidding dismissal or similar forms of punishment for participating in the monitoring process. As discussed above, the most important inducement for an insider to inform is likely to be the prospect of obtaining a substantial award and not the assurance that the statute forbids retaliation. Duration. The bounty system could be adopted as an experiment. The originating statute could authorize the monitoring system for, say, a ten-year period. The mechanism would sunset unless the national legislature chooses


12/10/07

13:14

Page 595


595

to extend it. A ten-year trial period would provide a suitable time to assess the efficacy of the informing system in eliciting useful information about covert cartels.

F. Conclusion Properly focused systems that pay bounties to informants have the potential to defeat efforts by cartel members to conceal their misconduct. By facilitating the efforts of public competition authorities to detect and prosecute cartels, bounty systems can make an important contribution to deterring the formation of illegal collusive schemes. The modern US Civil False Claims Act provides a useful model for considering the design and operation of bounty systems for the enforcement of other public laws. In particular, experience with the CFCA since 1986 suggests how anti-cartel policies might benefit from providing financial rewards for private parties to participate in monitoring compliance with antitrust prohibitions upon cartels.


12/10/07

13:14

Page 596

(Zc) Ehlermann 07 Ratliff

12/10/07

13:14

Page 597

V John Ratliff * Plea Bargaining in EC Anti-Cartel Enforcement— A System Change?

The object of this paper is to consider the possibility raised by Mrs Kroes in April 2005 that the European Commission might seek a simplified form of settlement procedure for cartel cases. Since we do not know precisely what the Commission has in mind, my approach here will be to explore considerations which appear relevant.

A. Mrs Kroes’ speech, April 2005 Under the heading “Direct settlements?” Mrs Kroes noted that it is a considerable burden to the Commission that “in each and every case brought . . . we are obliged to investigate every last substantive detail. Our final decisions have to be fully reasoned on the basis of our own analysis of the facts”. She then noted “(u)nlike some other systems—that of our American colleagues, for example—there is no arrangement for simplified handling of cases in which the parties to the cartel and enforcer concur as to the nature and scope of the illegal activity undertaken and the appropriate penalty to be imposed”. She went on to say that if, despite other efforts to improve anti-cartel enforcement, the Commission were not able to deliver swift enforcement with timely punishment, “we may need to look at how some form of plea bargaining procedure could bring advantages”.1 The Commissioner’s comments immediately provoked a reaction in the Press. They were unexpected, at a time when the focus of European Commission activity on anti-cartel enforcement systems appeared to be on developing and harmonizing EU national competition authority (“NCA”) leniency programmes. Some immediately suggested that, in order to introduce such a system, Regulation 1/20032 might have to be changed. Others welcomed the idea as it “would cut unnecessary cartel work massively”.3 * WilmerHale, Brussels. 1 Mrs Neelie Kroes, EC Commissioner for Competition, “The First Hundred Days”, SPEECH/02/205, 7 April 2005, at p. 5. 2 OJ L 1 [2003] at p. 1. 3 Global Competition Review News, 8 April 2005.


598

12/10/07

13:14

Page 598


B. Why suggest such a change? As the Commission has explained, it is now becoming overburdened by the success of amnesty/leniency applications. It is understood that the Commission now has at least 35–40 cartel investigations underway and clearly that is a huge burden, when one considers how detailed the EC procedure is. To just touch on the main tasks: there are applicants’ statements to be taken, “dawn raids”, requests for information, witness statements, Statements of Objections, hearings, confidentiality reviews, evidentiary reviews and complex fining considerations to evaluate, resulting in detailed decisions, often with hundreds of footnotes and files which often contain thousands of pages. Clearly there are also related translations and, since there are numerous defendants, much of this has to be done in multiple versions. The result is that Commission cartel proceedings usually take years and considerable resources, even with amnesty/leniency. The number of decisions per year is only some seven or eight. That may be no mean achievement in the circumstances (and already faster than fully contested cartels without amnesty/leniency), but clearly it is not a good result, if cases are coming in faster, creating an ever greater backlog. The number of cartels revealed is also in itself disturbing, because it may suggest that, despite all enforcement efforts, there are still many. As a result, there is an argument for a faster case handling rate so that enforcement resources can be spread more widely. Beyond consideration of the burden of the Commission’s proceedings, it is also argued that the subsequent appeals are a drain on Commission resources. Even with leniency, a large number of Commission decisions are still appealed to the European Courts. Again, for each cartel there may be multiple appeals because there are multiple defendants. Such appeals are mainly brought because, in some way or another, the companies concerned feel hard done by in terms of the amount of fine imposed and because, in a significant number of cases, appeals are successful in all or part. One may think that this is just normal “due process”, but it is argued that it is a further burden on the Commission, as it has to justify its position through two levels of appeal, again taking up resources.

C. The infringing company’s perspective One may think that many infringing companies would share the Commission’s desire to deal with these cases faster.


12/10/07

13:14

Page 599

Plea Bargaining in EC Anti-Cartel Enforcement 599 Many cartel cases result from the behaviour of maverick individuals who manage to slip through a group’s efforts at competition compliance. Such proceedings are a huge nuisance to such groups, which consider them as a stain on their general reputation for legitimate competition and such companies are very keen to resolve the issues quickly. Clearly there may be others which were more aware of infringements. Often they also do not want to be dealing with such proceedings for at least 10 years, which unfortunately is all too possible taking into account Commission proceedings, European Court appeals and related damages claims, up to now mainly in the United States. Practically, by the end of the case (to quote an old joke) only the lawyers may still be around on the defence side! With modern employment mobility, everyone else has changed jobs, if not companies. It is not always like this, but the general theme is often correct, so faster procedures should be welcome. Clearly companies may also be expected to be interested in any way to reduce the huge fines which they face. If there is a possibility of appreciable discounts to such fines, because a company’s cooperation extended beyond leniency to agreeing to a simplified “direct settlement”, that is something to be considered seriously, provided it is a fair and secure result.

D. Settlements under Regulation 1/2003 Another reason why the Commission may be tempted to suggest “direct settlements” at this time is the existence of new settlement procedures under Regulation 1/2003. As is apparent from Recital 13 to Regulation 1, these were not designed to deal with cases involving a fine. However, such new procedures were prompted by similar desires to improve the effectiveness and timeliness of Commission intervention in appropriate cases.4 It may be useful to recall how these settlements work and consider whether some features of the system may be useful for cartel cases. First, under Article 9(1) of Regulation 1/2003 the Commission has the ability to settle an alleged competition infringement, accepting commitments from the alleged offender, which are made binding through a Commission “commitment decision”. We have only seen a few examples so far, but the format appears to be:

4 The last sentence of the recital reads: “Commitment decisions are not appropriate in cases where the Commission intends to impose a fine.”


600

12/10/07

13:14

Page 600


—A very short decision in the EC Official Journal (a few paragraphs only). —Commitments published on the Commission’s website, first in draft form with a related notice inviting comments, then in definitive form. —A more detailed, but still lightly reasoned non-confidential version of the decision published on the Commission’s website.5 One would think this sort of decision attractive both to the Commission and infringing companies in cartel cases, involving less work and less explicit public findings. A key issue may be, however, whether this is enough for a cartel case. Secondly, it is important to bear in mind that, while these commitment decisions are binding on those offering the commitments and on the Commission, they do not have “horizontal effects” vis-à-vis other possible proceedings by other NCAs.6 NCAs are not bound to follow such decisions, as explained in Recital 13 to Regulation 1/2003, although in practice one may think that members of the European Competition Network (“ECN”) may have regard for Commission decisions, unless there are other, more specific effects to take into account in their territories, justifying further or other considerations. However, one may think that this is not enough for infringement settlements, where clearly infringing companies will not want to settle in Brussels, only to risk further proceedings before one or more EU Member States. Thirdly, it is also key that these settlements do not involve fines. On the contrary, they are “without prejudice” settlements, where the companies concerned do not accept any past wrongdoing, but just give undertakings as to future conduct. The conclusion in such a commitment decision is that “there are no longer grounds for action by the Commission”, as is made clear in Recital 13 and Article 9(1) of Regulation 1/2003. In cartel cases, that would not appear to be possible. Companies seeking amnesty/leniency will already be admitting the infringement. Yet the Commission still has to find an infringement by decision in order to impose a fine.7 Fourthly, one may note that not much weight can be attached to the Commission’s reasoning in such cases, since the decisions explicitly state that they are only based on a preliminary assessment of the facts and clearly they have not been fully argued by the companies concerned. As the legal saying goes, they have not been fashioned on the “anvil” of the law, through the detailed contradictory process of full proceedings. Again it is a good question whether that is enough for the imposition of a fine. One would think not, since the Commission has to do enough investigation adequately to reason its decision on the infringement and the imposition 5

See, e.g. German Bundesliga, OJ L 134 [2005]; Coca-Cola, OJ L 253 [2005]. As debated in relation to the parallel Spanish Coca-Cola proceedings, see Brokelmann H. (2006): paper presented at the IBC Competition Conference, London, May 2006. 7 See Article 23(2) of Council Regulation 1/2003, supra note 2. 6


12/10/07

13:14

Page 601

Plea Bargaining in EC Anti-Cartel Enforcement 601 of the fine itself, in order to comply with Article 253 (ex Article 190) of the EC Treaty, even if under a “direct settlement” procedure, fines are agreed. Fifthly, these settlements are also not based on a very detailed procedure. The Commission appears to send the company alleged to have infringed a “preliminary assessment” of the allegations and case concerned. However, it is not clear whether that should be treated like a Statement of Objections in the circumstances, triggering access to file, co-defendant access rights and third party rights to a non-confidential version. The Commission’s current approach appears to be that this is not the case. Such an approach may also not be enough for a cartel case. It may be thought normal that the infringing company would wish to have access to the file at the “preliminary assessment” stage to know what it faces and to decide whether to agree. It may also be that other co-defendants and third parties, who may include complainants, would wish access to at least a non-confidential version of the preliminary assessment. However, whether that is advisable is an open question, since it might be counter-productive to persuading a company to offer its plea, accepting the infringement. (As we shall see, in the United States third parties are not part of the plea bargaining process.) Sixthly, it appears that the Commission may accept commitments from some infringing companies, while proceeding against others. This could be a major issue in EC cartel cases. One could imagine cases in which only some parties agree a plea to settle the case and those which have settled might still be concerned by evidence as to the nature of the infringement suggested by others in continuing proceedings. Similarly, those continuing in the proceedings might want to contest what those bargaining have agreed! (This is discussed further below.) Finally, it may be useful to comment on some of the interests here. In such cases, the Commission is looking for a solution which occurs promptly and is therefore more relevant to the market concerned. The company is also looking for a prompt end to the proceedings, without any finding of infringement or sanction. Neither is likely to be interested in the procedure where a major point of principle is in issue. Overall therefore, there are a number of interesting parallels in the current commitment decisions procedure, which could be relevant to development of an “infringement settlement” procedure. However, it appears likely that something more developed will be required for cartel cases, since a plea agreement in such cases will normally involve findings of infringement and fines.


602

12/10/07

13:14

Page 602


E. Broader considerations for a new “infringement settlement procedure” I propose now to explore further what other issues appear to need clarifying if plea bargaining is being considered in EC law. First, there is the issue as to whether to accept plea bargaining at all. It is not clear that the idea will be generally accepted in the EU. Even with leniency, there has been reluctance in some countries to introduce such procedures (and there are only 19 programmes out of 25 Member States at the time of writing). Plea bargaining appears to go a step further and may not accord with the traditions of some Member States’ judicial systems. On the other hand one may note that there was recently a settlement in the United Kingdom after the “Statement of Objections”.8 As Jochen Burrichter and Daniel Zimmer explain in their paper for this Workshop, it appears that there is also already some settlement practice in Germany (and there appears to be a new procedure to “plead guilty” in France).9 So perhaps at least some of the Member States may be open to plea bargaining ideas. Some may object to any concept of bargaining away the sanction for what is, in EC law, an administrative infringement, but still a very serious one. All the more so if, in their systems cartel activity is a criminal offence and if, in their systems, plea bargaining is not accepted. Some may also object if there is any question of excluding a right of appeal, even by clear contractual waiver. It may be argued that there are compelling practical reasons for such a simplified “infringement settlement procedure”, in order to deal with the specific EU problem of a fast growing cartel backlog (as explained above), even if the Member States do not itself use it in their domestic enforcement systems. Even if anti-cartel enforcement in the EU has grown greatly with many Member States now actively pursuing cartel cases, that also does not resolve the Commission’s developing backlog. On the other hand, it may be argued that this will introduce even more variable geometry into an enforcement landscape which the Commission (and, it appears, the ECN as a whole) is trying to level through harmonizing leniency and other procedures. There are already major procedural differences, both between the Commission’s system and those of Member States and as 8 See, the Sevenoaks Survey case, UK Office of Fair Trading Press Releases 88-06, 19 May 2006 and 214/05, 9 November 2005, available on the UK Office of Fair Trading’s website. 9 See Burrichter J. and Zimmer D. (2006): “Reflections on the Implementation of a “Plea Bargaining”/“Direct Settlement System in EC Competition Law”, written contribution for this Workshop.


12/10/07

13:14

Page 603

Plea Bargaining in EC Anti-Cartel Enforcement 603 between the Member States. Is it sensible to create a new divide, where an infringing company could plea bargain in Brussels, but not if a NCA, which does not allow plea bargaining, were to take up the case? Should there not be the maximum possible procedural neutrality on such issues, so that either the ECN moves to accept plea bargaining as a whole or no-one should? Secondly, can the Commission introduce such a practice under Regulation 1/2003 as the Regulation stands? If not under Article 9(1), under a combination of Articles 7(1) and 23(2), which generally enable the Commission to take decisions requiring undertakings to bring an infringement to an end and to impose fines? The core issue appears to be what are the minimum elements required in the decision and the procedure for it to be acceptable in the EC legal order. For example, would a preliminary assessment as in current settlements suffice or would such a document be more akin to a Statement of Objections? Could an infringing company waive the need for a Statement of Objections and otherwise its right to be heard? Would the Member States have to be consulted? If not, in Article 44 of Regulation 1/2003, there is a mechanism to report on the operation of the Regulation to the EU Council, which may be one way to build the consensus for a revision proposal, introducing a more specific hybrid procedure, somewhere between Article 9(1) and the detail of the Implementing Regulation 773/2004,10 if that is what is considered necessary. Thirdly, what does the company concerned need to know about the case to be able to agree to the infringement? Again, one would think that the company needs to have at least a reasonably clear “preliminary” assessment from the Commission of the infringement. Otherwise, it would be offering a plea “blind”, based just on its own internal investigation. To come back to Mrs Kroes’ speech, that appears to mean that the Commission would still have to do some “own analysis of the facts” and also some reasoning of its decision, although how full both would have to be is a good question. It should still be less of a burden than a full decision in its current form. A related issue, as noted above, is whether an applicant should have access to file at this stage. This appears fair if the infringing company is to assess what finding of infringement it should accept. However, does it mean that all parties involved in a Commission proceeding should have such access at the same time? It should be said that earlier file access would be most welcome in all cases. As the law stands, access to file is only required when a Statement of Objections is sent.11 However, it is questionable whether that is the right rule. Having assisted counsel in Italian proceedings where we could obtain early access to file and “top it up” as the case developed, I think there is a strong case for letting the defence see what is building as soon as possible. This would 10 11

OJ L 123 [2004], at p.18. Article 15(1) of Regulation 773/2004.


604

12/10/07

13:14

Page 604


facilitate settlements in some cases (as sought now by the Commission). It might also allow for earlier and better defences, which in some cases may resolve key misunderstandings which can occur. Fourthly, what does the company concerned need to know about the possible fine which could be imposed in order to settle? Some argue that this means that the infringing company should at least know what is the maximum possible fine against it (as in the United States, see below) and that the current rules are not predictable enough. Defenders of the current system say that the 1998 EC Fining Guidelines are adequate. They argue that these work well, apart from minor cartels with small companies where the Commission has made adjustments in its practice. Commission decisions are also an indicator, even if not binding on the Commission for future cases. One notes also that in Mrs Kroes’ speech last April she appeared to say that she still does not want the Commission’s fines to be very transparent, out of concern that “potential infringers (would) calculate the cost/benefit ratio of a cartel in advance”.12 Others argue that there should be changes to the EC Fining Guidelines and Commission practice, addressing the mains areas in which they are considered unpredictable.13 For example, the Commission’s initial determination of fining level, based on the “gravity” of the infringement, where the Commission states that fines are not linked to the turnover in the products concerned, yet companies argue that fines should be, at least to some extent, if only to allow proportionality to be assessed. Or the Commission’s determination as to the relative level of fine amongst the cartel participants, adjusting upwards the fines of multi-product, conglomerate companies in the name of “deterrence”, which many consider unpredictable and unfair in its current form. At a matter of principle, it is also not clear why companies should not be able to calculate what fines they face if they infringe. Law is about establishing rules giving people choices as to how to behave and encouraging them to behave in a certain way. Why should cartel sanctions be any different to other sanctions? The Commission may need to retain some ability to adapt fines to particular circumstances, but it may be thought that a high degree of transparency is desirable and a necessary legal requirement. It is also arguable that if the EC Fining Guidelines were more predictable, less appeals would be brought, partially alleviating the Commission’s current backlog concerns, even without plea bargaining. The impression one has in practice is that in some cases companies appeal simply because they are caught by surprise and object to the ultimate fining calculation. A related idea is that the current procedure should give companies more of an opportunity to be heard on fines. For example, it was a surprise to some to 12

See supra note 1, at p. 5, under the heading “Reviewing the fines system”. See the European Commission’s new Fining Guidelines issued in September 2006, OJ C 210 [2006], at p. 2. 13


12/10/07

13:14

Page 605

Plea Bargaining in EC Anti-Cartel Enforcement 605 learn that there would be a specific hearing on remedies in the recent Article 82 EC Microsoft proceedings, when cartel participants only have short indications in the Statement of Objections as to what the fine might be and hear no more until the decision is taken. Clearly there are differences in the two cases, but again one would think that more specific indications as to proposed fines after the finding of infringement, with an opportunity to be heard thereon before the Commission’s ultimate decision, would reduce the number of appeals. Something like this appears to be intended in the plea bargaining concept, if an infringing company is to be able to settle its fine with the Commission. However, there is a case for doing it anyway for all cases. Companies will also need to know that the Commission’s offer is fully “mandated”. In other words, that the person agreeing the direct settlement has the power to do so from the full College of Commissioners who usually all deliberate, or at least are consulted on the fine to be imposed. Fifthly, the Commission also has to define what it will offer as an incentive to settle and to what extent “bargaining” is in issue. There is an additional saving to the Commission. How much discount will be given for that? It may be also that any incentive would vary according to when a plea is offered. One would think that a plea before the Statement of Objections would be of far greater value than one after it, but the latter might still have a value to the Commission worth recognising. Companies may also have something to say about the quality of evidence which the Commission has, going to the nature, scope and duration of the infringement. Is that a dialogue to which the Commission would be open in such plea bargaining (as appears to happen in the United States)? Sixthly, companies need to know that their Commission plea and “infringement settlement” will end the matter in terms of ECN enforcement once and for all. An Article 9(1) settlement has no such effect, whereas full Commission decisions do. A hybrid “infringement settlement decision” would have to have the latter character or the infringing company could settle in Brussels only to face a number of national cases. Finally, I would mention that it appears necessary to clarify how any “infringement settlement” rules would work in cartel cases as regards codefendants and third parties.14 For example, it may be that all companies should have the same opportunity to settle and gain the relevant benefits if they cooperate in a defined way, but that the Commission may want to retain discretion not to accept pleas in some circumstances (e.g. cases of principle). Given the number of appeals based on infringement of the principle of unfair treatment, one may also think that, as matters stand, there is a risk of one company arguing that it should have been able to obtain a plea 14

With thanks to Anne Vallery and Frédéric Louis for their assistance on this point.


606

12/10/07

13:14

Page 606


agreement, or a similar plea agreement, to another company. How is that to be dealt with? Another aspect is that in cartels, one company’s position may be closely affected by the position of others. What happens if two companies in a cartel plead to a short duration, but three others do not settle and the investigation continues, revealing ultimately a longer duration of infringement? Presumably, the Commission cannot go back on its plea agreement. However, what if those that settled might have further things to say on the case, which would affect the ultimate decision and any follow-on damages claim? As noted above a similar point may also arise the other way around, with those who have not settled, objecting to the admission in the plea. One may raise questions as to third party rights. How will these be dealt with? In the United States, as we shall see below, third parties are not involved in plea agreements. The usual position in the EU is very different, with third parties seeing a non-confidential version of the Statement of Objections. Is that not to apply to a preliminary assessment in a hybrid “infringement settlement procedure”? There may also have been a complainant about the cartel. What if the complainant contests the plea settlement, with his eye focussed on a follow-on damages claim? Will the Commission be able just to reject such action based on arguments of the Community interest in the style of Automec II? 15

F. US comparisons I propose here only to note aspects of the US system of plea bargaining, which appear most relevant for comparative purposes (and subject to the correction of the distinguished US colleagues present). First, in the United States it appears that plea bargaining is widely accepted. There is Supreme Court authority suggesting that plea bargaining is an “essential component of the administration of justice”.16 There are general rules in the Federal Rules of Criminal Procedure (e.g. at Rule 11) and specific guidance is to be found in the US Department of Justice (“DOJ”) Antitrust Division Grand Jury Manual17 (“The Division Manual”). The Division Manual states that it is the policy of the Division that, “absent 15

Case T-24/90, [1992] ECR II-2223. Santobello v. New York, 404, US 257 at p.260 (1971). 17 This is available on the DoJ’s website, www.usdoj.gov/atr. Various examples of plea agreements are also available there in the cases section. Plea bargaining in competition cases is also accepted in Canada, see Bériault Y. (2004): “Negotiating a Plea of a Corporate Defendant in Canada”, ABA International Cartel Workshop, New York, February 2004. (With thanks to Charles Stark for his assistance in directing me to this material.) 16


12/10/07

13:14

Page 607

Plea Bargaining in EC Anti-Cartel Enforcement 607 unusual circumstances, criminal cases may be disposed of pursuant to plea agreements”.18 Secondly, US plea bargaining in cartel cases occurs in a context very different from that of EC anti-cartel enforcement and current Commission settlements. Enforcement of the cartel prohibition in the US is a matter of criminal law, and the company or individual in question may agree with the prosecutor on the terms of a plea. An agreed joint recommendation is then made to a court, which is binding on the defendant and the US Government, but not on the court. (Nevertheless, it appears that courts often do respect such agreements.) The court itself is not involved in the plea bargain negotiations. Those are between the prosecutor and the defendant. Importantly, “outsiders [i.e. third parties] are never participants in plea negotiations”.19 Reinforcing the point, such negotiations are also privileged communications. The court also checks that the plea was offered “knowingly and voluntarily” and there is a requirement that the defendant be represented by counsel, since this is, in practice, a criminal confession and guilty plea and the defendant is waiving important rights. A factor for plea acceptance is the defendant’s willingness to accept responsibility for the unlawful conduct. (The DoJ does not accept nolo contendere pleas, where the defendant does not contest a charge, but does not admit it either.) This is rather different from the EC context where, amongst other things, the law is not criminal and the Commission is the investigator and judge, not just the prosecutor. Thirdly, the defendant agrees to waive the right to appeal the sentence imposed by the court, if that sentence is consistent with that jointly recommended to the court by the DoJ and the defendant. A plea bargain is therefore a contract going to core rights. Fourthly, agreement on a plea appears to be facilitated by fairly transparent Federal Sentencing Guidelines. These involve: —Suggested base amounts, antitrust offences being at a defined level. —Rules on the correlation of the fine to the volume of US inter-state commerce affected. It appears that in a cartel case the base fine may start at 20% of the affected commerce for the duration of the infringement20 (although there may be variations for international offenders, for whom it is argued that this understates the impact of the offence 21). 18

Chapter IX, p. 1. Division Manual, Chapter IX at p. 11. 20 ABA Antitrust Law Developments 2005, Fifth Edition, I-VIII, p. 8. 21 Spratling G. (1999): “Negotiating the Waters of International Cartel Prosecutions— Antitrust Division Policies Relating to Plea Agreements in International Cases”, speech of March 1999, at p.18. 19


608

12/10/07

13:14

Page 608


—Multipliers based on a “culpability score”, which can increase fines up to 80% of the volume of commerce involved. —Resulting “fine ranges”, with an ability for “upward or downward departures” (e.g. if the defendant offers substantial assistance with the investigation). It is not clear what value plea agreement judgements have as precedents, in particular since the recommendation to the court is negotiated.22 On the other hand, it appears that it is US policy that plea bargaining result in convictions which are uniform for defendants guilty of similar crimes and which reflect the seriousness of the crimes committed.23 It should also be noted that the Sentencing Guidelines are advisory in nature and not binding on the court. Some sentencing elements therefore appear clearer than in EU practice. However, the rules are complex and there still remain a number of elements where judgements may be made as the proportionality of any proposed fine for the defendant is assessed. A key point is that the defendant knows what is the maximum possible sentence which the court could impose on it/him. It should also be noted that the US amnesty/leniency programme is integrated with the agreed sentencing recommendation. Thus, the defendant is not fined on the basis of the increased level of commerce affected, if it/he provided the evidence thereof and substantial assistance to the DoJ’s investigation is taken into account in any recommendation for a downward departure from the Sentencing Guidelines. The plea agreement also appears to repeat continuing cooperation obligations of the defendant under such programmes. Fifthly, there are a wide range of factors which fall to be considered by the DoJ when deciding whether to accept a plea. In other words, the prosecutor appears to have discretion whether to do so and does not have to accept such a plea. Interestingly, given Mrs Kroes’ comments in her speech last year, two of these factors are: —The desirability of prompt and certain disposition of the case. —The need to avoid delay in the disposition of other pending cases.24 The Division Manual states “this . . . may be the most important factor to consider in favor of a plea agreement, because a plea may lead to a speedy

22 A similar point is made by Bériault, supra note 17, as regards the Canadian system at p. 7: “These decisions should be used very carefully, since the fine may represent a high percentage of the volume of commerce during the period of the indictment as a result of the negotiations to reduce the period of the charge in situations where the there may have been debate as to the actual length of the conspiracy”. 23 Division Manual, Chapter IX, p. 9. 24 Division Manual, Chapter IX, pp. 7–8.


12/10/07

13:14

Page 609

Plea Bargaining in EC Anti-Cartel Enforcement 609 resolution of other cases, including the filing of some which might never be made if the case is tried . . .”25 Sixthly, it appears that various issues may be negotiated. For example: —The nature and scope of the charge in the circumstances (e.g. could a plea to a lesser or related charge be accepted?). —The actual length of the alleged conspiracy (if that is in debate; which is clearly of particular importance in the United States given treble damage plaintiff actions). —Elements in the prosecutor’s discretion (e.g. as regards the fine range and the degree of substantial assistance). It is said that the prosecutor should not trade away readily provable charges,26 but defence counsel may make submissions on the evidence (e.g. as to what they think the DoJ could prove in court) and what the appropriate fine should be in the circumstances. It should also be noted that in the plea agreement the DoJ states that it has evidence sufficient to prove various facts in support of each charge.) The fact of entering into a plea bargain of itself leads to a downward adjustment of the assessment of an offence. It appears that pleas may often be raised after the preliminary (grand jury) investigation, but in principle they can be raised at any time during the investigation. Finally, according to general rules on plea bargaining set out in the US Principles of Federal Prosecution, a plea must meet certain criteria. Thus, a plea must: —Bear a reasonable relationship to a defendant’s conduct in nature and extent. —Reflect an adequate factual basis (meaning that the prosecutor should check that the defendant is not just pleading guilty to be rid of the case, when in fact it/he considers that he is not guilty). —Make likely an appropriate sentence in the circumstances of the case. —Not adversely affect the position of other corporate defendants.27 It may be all too clear why my eye has been caught by these points, given parallels to the current EU agenda here! I hope that captures the core considerations for present purposes.

25 26 27

At p. 8. Division Manual, Chapter IX, pp. 20–21. Division Manual, pp. 15 et seq.


610

12/10/07

13:14

Page 610


G. Conclusions Mrs Kroes’ speech has opened an interesting debate. It will be interesting to see where it leads. Plea bargaining is a relevant issue, but one which needs careful review, both as to core principles and practicalities. As matters stand, if the concept is introduced, I believe it will involve something of a system change, for a new hybrid “infringement settlement procedure” will be required. If so, it will be very welcome, provided that the appropriate legal checks and balances are retained.

(Zd) Ehlermann 07 Burrichter

16/10/07

13:59

Page 611

VI Jochen Burrichter and Daniel J. Zimmer* Reflections on the Implementation of a “Plea Bargaining”/“Direct Settlement” System in EC Competition Law

A. Introduction On 7 April 2005, the European Commissioner for Competition, Neelie Kroes, voiced the concern—in view of the Commission’s growing success in cracking down cartels—that “the Commission [. . .] risks becoming the victim of its own cartel-busting success”.1 As one of the possible changes to the enforcement system, she proposed an arrangement for a simplified handling of cases. Similar to the US “plea bargaining” system, the Commissioner intends to simplify proceedings in cases where both the undertaking and the Commission concur as to the nature and the scope of the anticompetitive behaviour. This concept will ease the burden of the Commission, which is generally obliged to investigate each case completely and to fully reason its decision, and will thus permit “swift enforcement with timely punishment”. In her speech at the 10th Annual Competition Workshop at the European University Institute on October 13, 2006, Commissioner Kroes confirmed her intention to further develop the instrument of direct settlements.2 Plea bargaining has been a long-standing practice on the other side of the Atlantic, where only a minority of cases actually proceed to trial. Ever since the judicial approval of “plea bargaining”,3 this practice has formed an

* Jochen Burrichter is partner at Hengeler Mueller, Düsseldorf/Brussels. Daniel J. Zimmer, LL.M. (Duke) was a junior associate (Rechtsreferendar) at Hengeler Mueller, Düsseldorf. 1 Kroes N. (2005): “The First Hundred Days”, speech delivered at the 40th Anniversary of the Studienvereinigung Kartellrecht 1965–2005, International Forum on European Competition Law, Brussels, 7 April 2005. 2 Kroes N. (2006): “Delivering on the Crackdown: Recent Developments in the European Commission’s Campaign Against Cartels”, speech delivered at the 10th Annual Competition Conference at the European University Institute, Fiesole, 13 October 2006, text available at http://europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/06/595&format=HTML& aged=0&language=EN&guiLanguage=en. 3 See Santobello v. New York, 404 U.S. 257, 260 (1971); Blackledge v. Allison, 431 U.S. 63, 76 (1977). See also the Plea Agreement Guidelines by the US Department of Justice, in US Department of Justice, Antitrust Division (1991): Antitrust Grand Jury Practice Manual, Chapter 9 (1st ed. 1991), available at http://www.usdoj.gov/atr/public/guidelines/207144.htm.


612

16/10/07

13:59

Page 612


integral part of the public enforcement of the antitrust laws, and the possibility of such plea bargaining has been equally important in advising undertakings faced with proceedings for allegedly anticompetitive behaviour. Recent developments in the laws and practice of certain European Member States show that the concept of consensual elements is viewed as a potential key to balancing effective cartel law enforcement with the precept of the rule of law even in criminal proceedings. After France introduced the new rules on “plaider coupable” in French criminal proceedings,4 similar legislative proposals in Germany have picked up on a recent decision by the German Federal Supreme Court (i.e., the Bundesgerichtshof). In its latest decision on “arrangements in criminal procedure”, the Court held that the previous practice, as accepted by the German courts, threatened to violate the constitutional limitations set by the separation of powers.5 Therefore, the Court identified an urgent need for legislative intervention. The new discourse about the implementation of plea bargaining in the EC competition law context falls on fertile ground, and in fact it is not without precedent in European antitrust enforcement practice. The following reflections will thus turn first to the German experience, involving the “consensual fining decisions” of the German Federal Cartel Office (Bundeskartellamt). The paper will then explore a potential European plea bargaining system, focusing on the reasons for its implementation and on its interaction with the precept of the rule of law. Finally, we will provide an outlook on certain aspects that merit further discussion.

B. The German experience In national competition proceedings, the Bundeskartellamt has in the past unofficially adopted a quasi-plea bargaining proceeding leading to a “consensual fining decision” (einvernehmliche Bußgeldbescheid). In “clear-cut” cases where the Bundeskartellamt was able to demonstrate anticompetitive behaviour, discussions with both the undertaking concerned and its legal counselors were initiated even before a Statement of Objections was issued. In the course of the discussions, the Bundeskartellamt generally revealed the main documents and other pieces of evidence in its possession which in its

4 For a German perspective, see Jung H. and Nitschmann K. (2004): “Das Bekenntnis zum Schuldbekenntnis—zur Einführung des plaider coupable im französischen Strafprozess”, Zeitschrift für die gesamte Strafrechtswissenschaft 116, pp. 785 et seq. 5 BGH, Case GSSt 1/04, Entscheidungssammlung des Bundesgerichtshofes in Strafsachen (BGHSt) 50, 40 = Neue Juristische Wochenschrift (NJW) 2005, 1440, 1443 = Neue Zeitschrift für Strafrecht (NStZ) 2005, 389.


16/10/07

13:59

Page 613

“Plea Bargaining”/“Direct Settlement” Sytem in EC Competition Law 613 view established an infringement of the antitrust rules. The members of the Bundeskartellamt and the undertaking and its advisors then started discussions concerning the appropriate fine. At this time, undertakings also sometimes presented evidence of mitigating circumstances in order to reduce an envisaged fine. The negotiations took place with the civil servants in charge of the case (i.e., those working in the Decision Department). Those participating in the negotiations were the same individuals who ultimately took a decision in the case in question. After an understanding had been reached, the Bundeskartellamt normally provided the undertaking’s legal advisors with a draft fining decision so that they could once more comment on it. Finally, the Bundeskartellamt would render a decision in the anticipation that the addressee would not appeal. Based on that assumption, the decision contained only a shortened summary of facts of the case and a curtailed legal assessment of the anticompetitive behaviour. With regard to the legal consequences of the allegedly anticompetitive behaviour, the Bundeskartellamt also tended to be more reluctant to impose fines on representatives of the undertaking who are primarily personally liable under German law (see §§ 9 and 30(1) of the Law of Administrative Offences, i.e., the Gesetz über Ordnungswidrigkeiten, or “OWiG”). It has emerged that in a recent case an undertaking was able to negotiate a maximum fine by arguing that it needed predictability of the maximum fine in order to be able to float its shares, which was necessary to rescue the undertaking. However, as any waiver of the right to appeal in the process of reaching the “consensual fining decision” is legally void, the undertaking concerned is not barred from appealing the decision. Due to its shortened summary of facts and curtailed legal reasoning, the “consensual fining decision” may not withstand judicial scrutiny in a subsequent appeal. If that is the case, the Bundeskartellamt in some cases has to consider withdrawing the initial decision and replacing it with a new, fully reasoned fining decision. In recent cases where only some of the undertakings concerned had agreed to such a “consensual decision”, they also decided to appeal the decision after it was appealed by the other undertakings. In view of these potential drawbacks in connection with this practice, and given the expanded jurisdiction of the Bundeskartellamt to also fine anticompetitive behaviour that infringes EC competition law under Article 81(1) EC after the entry into force of the 7th amendment to the German Act Against Restraints of Competition (the Gesetz gegen Wettbewerbsbeschränkungen, or “GWB”), this procedure will have to be reviewed critically. One of the particularities of German law consists in the fact that the procedure in administrative fine cases is governed by the Code of Criminal Procedure. If the decision of the Competition Authority is appealed, it is treated legally as a criminal indictment. The allegations in the indictment thus have to be proven under the strict rules of criminal procedure.


614

16/10/07

13:59

Page 614


C. “Plea bargaining” in EC competition cases? Article 9(1) of Regulation No 1/20036 allows commitment decisions entailing some form of direct settlement with the Commission. However, this type of decision is not available in cartel cases,7 as they “are not appropriate in cases where the Commission intends to impose a fine”.8 Still, the experience gained in the use of commitments decisions in the recent Coca-Cola, German Bundesliga and De Beers cases may serve as an indication for a possible settlement procedure:9 the Commission first publishes a notice pursuant to Article 27(4) of Regulation 1/2003 in the Official Journal; then it poses a draft of the commitments on the DG Competition website and invites comments; and finally, it publishes an abbreviated, non-confidential version of the decision. Similarly, the envisaged “direct settlement” system under EC competition law would most likely consist of a “short form decision” according to Article 7 of Regulation No 1/2003, which would concentrate only on the “core” of the infringement. In the settlement process, the Commission would confront the undertaking with its alleged anticompetitive behaviour. In this respect, the Commission would issue a “Statement of Objections”, but in anticipation of this it could already present a preliminary assessment of the facts, together with a disclosure of the most important pieces of evidence. The undertaking would then be invited to admit to these allegations. The hearing officer(s) of DG Competition would also have to be involved at this stage of the proceeding. From the point of view of both the Commission and the undertakings concerned, there are different (but good) reasons warranting the implementation of such a “direct settlement” system in Europe (see point 1 below). However, the procedure will have to respect minimum guarantees of “due process” (point 2), and there remain a number of aspects that merit further discussion (point 3).

6 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 [2003]. 7 Reynolds M. and Anderson D. (2006): “Immunity and Leniency in EU Cartel Cases: Current Issues”, 27 European Competition Law Review 82, at p. 86; European Commission (2004): Commitment decisions (Article 9 of Council Regulation 1/2003 providing for a modernised framework for antitrust scrutiny of company behaviour), MEMO/04/217, 17 September 2004; Dalheimer/Feddersen/Miersch: EU—Kartellverfahrensverordnung, 2005 Art. 9 Ann. 4. 8 Council Regulation (EC) No 1/2003, cited supra note no. 6, Recital 13. 9 See Coca-Cola, OJ L 253 [2005]; German Bundesliga, OJ L 134 [2005]; De Beers, OJ L 205 [2006].


16/10/07

13:59

Page 615

“Plea Bargaining”/“Direct Settlement” Sytem in EC Competition Law 615

1. Reasons for implementation The recent caseload of the European Commission certainly seems alarming: a growing number of leniency applications (16 in 2003 compared to 29 in 2004) entails an increased workload for the Commission. Currently, the Commission is said to handle approximately 40 cartel investigations at a time but it can only render seven or eight decisions each year. Therefore, there is no “swift” and “timely” enforcement. As Commissioner Kroes has pointed out, one of the reasons for this steadily growing backlog is the Commission’s obligation to fully investigate each and every case. To do so, the Commission is obliged, among other measures, to take the leniency applicant’s statement,10 to conduct dawn raids at the offices of other undertakings, to request information from the undertakings concerned and other market participants, and to collect—if necessary—witness statements. This process is so time consuming that even in cases initiated on the basis of a leniency application, it takes on average about three years for the Commission to reach a decision.11 Furthermore, the Commission must fully reason its decision—in many cases in several official languages—in order to withstand the judicial scrutiny of the European Court of First Instance (CFI) and sometimes of the European Court of Justice (ECJ). These subsequent appeals further strain the Commission’s resources and thus limit its ability to render more fining decisions. Therefore, their number has already started to disproportionately diminish relative to the growing number of leniency applications: since its height in 2001, with almost € 2 billion in fines, the overall fines imposed by the Commission dropped to approximately € 390 million in 2004, but then increased again to approximately € 683 million in 2005. The envisaged “short form decision” will likely reduce the workload of the Commission and will thus free up resources. With the implementation of a form of “plea bargaining”/”direct settlement” system, “clear cut” cases could be more easily disposed of, leaving the Commission more flexibility to concentrate on the contested cases. The undertakings concerned may also profit from an accelerated proceeding: when admitting to an alleged infringement of the EC competition rules and waiving the right to a fully reasoned decision, the undertakings can

10 This makes the procedure even more complicated if the Commission has to take into consideration the potential exposure of the whistleblower to pre-trial discovery requests from private plaintiffs in other jurisdictions, primarily the United States. For a discussion of the “paperless application”, see Reynolds M. and Anderson D. (2006), supra note no. 7, at p. 83. See also the Draft Commission Leniency Notice, published 29 September 2006, available on the DG Competition website at http://ec.europa.eu/comm/competition/cartels/legislation/draft_revised_ notice_en.pdf, at pp. 11–12 (“IV. Corporate Statements”). 11 See Reynolds M. and Anderson D. (2006), supra note no. 7, at p. 85.


616

16/10/07

13:59

Page 616


expect an expedited and predictable disposition of the case, so that their internal and external efforts and legal expenses will likely be reduced. Furthermore, the undertaking may be able to limit the negative publicity associated with antitrust proceedings even if the Commission upholds its practice of publishing a press release.12 Finally, an undertaking can expect that its cooperation will eventually lead to a reduction of the fine to be imposed.

2. “Direct settlement” and “due process” The rule of “due process”, as it is expressed in Article 6 of the European Convention on Human Rights, also applies to Commission decisions imposing fines.13 Therefore, a “direct settlement” system in EC competition law will have to provide for a procedure that safeguards the different aspects of the principle of a “fair trial”. This has also been expressed by the German Federal Constitutional Court (Bundesverfassungsgericht) with regard to a plea bargaining procedure in criminal cases.14 It embraces aspects of legality (see a) below), of legal certainty (see 2.2. below), and of “due process” in its narrower meaning.

2.1. Principle of legality First, it is recognized as a general principle of Community law that “any intervention by the public authorities in the sphere of private activities of any person, whether natural or legal, must have a legal basis and be justified on the grounds laid down by law”.15 Therefore, a “direct settlement” system will have to be based on some rule of procedure providing that the undertaking concerned can be fined for the alleged anti-competitive conduct to which it has admitted. As noted above (under C.), a settlement reached in such cases cannot be based on Article 9(1) of Regulation 1/2003. The Commission therefore has to base its subsequent decision on Articles 7(1) and 23(2) of the Regulation. In order to explain and promote a new instrument of “direct settlement”, the Commission should issue a Notice regarding the relevant procedure. This 12 See Kauper M. (1993)” “The Use of Consent Decrees in US Antitrust Law”, in Slot P. J. and McDonnell A., eds., Procedure and Enforcement in E.C. and U.S. Competition Law, Sweet & Maxwell, London, at p. 105. 13 See Kerse C. and Khan N. (2005): EC Antitrust Procedure, 5th edition, Sweet & Maxwell, London, at paras. 6-017 et seq. 14 Bundesverfassungsgericht, Case 2 BvR 1133/86, Neue Juristische Wochenschrift (NJW) 1987, 2662 = Neue Zeitschrift für Strafrecht (NStZ) 1987, 419. 15 Case 46/87 & 227/88 Hoechst v. Commission [1989] ECR 2859, para. 19.


16/10/07

13:59

Page 617

“Plea Bargaining”/“Direct Settlement” Sytem in EC Competition Law 617 would help to establish transparency with respect to the corresponding conditions. In either case, the Commission will have to respect its obligations under the principle of legality if it bases its decision on Articles 7(1) and 23(2) of Regulation 1/2003 and stipulates in a Notice—similar to the Notice on the leniency programme—the conditions under which a settlement can be achieved.

2.2. Legal certainty Second, any system of a “direct settlement” will have to take into account the principle of legal certainty, which in fact proves beneficial for both parties to the settlement. Firstly, regarding the undertaking concerned, legal certainty will require some predictability as to the amount of the potential fine the undertaking is facing. However, it may be sufficient, as it is in criminal proceedings in the US, for example, to negotiate a maximum fine that cannot later be exceeded in the final decision.16 In order to guarantee the greatest possible legal certainty and predictability, it seems desirable that the Commission and the undertaking concerned might negotiate the exact amount of the fine on the basis of the Commission’s new fining Guidelines.17 In relation to the Commission’s decision-making procedure, it has to be born in mind that the negotiations will take place between the undertakings and their legal advisers and DG Competition, whereas the final decision is taken by the College of Commissioners. This constitutes a considerable problem for the predictability of the decision, which to a certain extent could be reduced by authorizing the Competition Commissioner to take the decision on behalf of the college, similar to decisions taken under Article 6 of the Merger Regulation. Secondly, it is necessary to have predictability with respect to the contents of the decision. This condition may be fulfilled by a Statement of Objections from which the final decision does not deviate. The prerequisite of predictability would be eased by negotiating the contents of the SO and of the final decision on the basis of drafts to be disclosed to the undertakings concerned before their adoption. Legal certainty is also of great importance for the Commission when settling antitrust cases. In order to achieve its goal of reducing its caseload and freeing up resources, the Commission will have to obtain at least a “factual” waiver of the right to file an appeal, bearing in mind that, at least in criminal proceedings, such a waiver to file an appeal is usually regarded as being 16 Compare Fed. R. Crim. Proc. Rule 11(b)(1)(H) (prior to accepting a guilty plea, the court must inform the defendant of “any maximum possible penalty, including imprisonment, fine, and term of supervised release”) and Bundesgerichtshof (BGH), Case 4 StR 240/97, Entscheidungssammlung des Bundesgerichtshofes in Strafsachen (BGHSt) 45, 193 = Neue Juristische Wochenschrift (NJW) 1998, 86. 17 Commission Guidelines on the method of setting fines imposed pursuant to Article 23(2)(a) of Regulation No 1/2003, OJ C 210 [2006].


618

16/10/07

13:59

Page 618


ineffective.18 And although Article 23(5) of Regulation 1/2003 explicitly states that Commission decisions are not “of a criminal law nature”, there remain serious doubts as to whether a formal waiver would be permissible under the provisions of the EC Treaty. On the other hand, the prospects of a successful appeal in such a case would be limited, given the fact that the parties would already have admitted to an infringement. In critical cases where an appeal is filed, the Commission will have to consider whether to withdraw the decision and render a new, fully reasoned one. This possibility could be expressly reserved in the settlement decision.

2.3. Due process Third, the “direct settlement” system will also have to observe due process requirements in a stricter sense, as provided for by Regulation 1/2003 or by the general principles of European law. These include the right to be heard, the obligation of a fair investigation of the facts of the case, and the need for a full reasoning of the decision rendered. The right to be heard has long been acknowledged by the ECJ19 and is now expressly provided for in Article 27(1) of Regulation 1/2003. When settling a decision with the undertaking concerned, this principle will generally be observed. But it needs to be borne in mind that the Court has extended this right to third parties who will be adversely affected by the decision.20 The standing of third parties in the “direct settlement” proceeding is therefore one of the open questions to be dealt with below (see point 3.4.)). An additional aspect of due process is the general obligation of the Commission to sufficiently investigate the facts of each case.21 However, this requirement of an all-embracing investigation collides with the general requirement of an expedited disposition of a case when both the enforcement agency and the addressee of the envisaged fining decision agree as to the nature, extent and duration of the illegal conduct. Under these circumstances, both parties to the settlement should be able to dispense with an in-depth factual investigation, bearing in mind that the accelerated proceedings will generally be beneficial for each of them. Finally, the ECJ ruled as early as 1963 that a Commission decision must “set out, in a concise but clear and relevant manner, the principal issues of law and of fact upon which it is based and which are necessary in order that the 18 BGH, Case 4 StR 240/97, Entscheidungssammlung des Bundesgerichtshofes in Strafsachen (BGHSt) 45, 193 = Neue Juristische Wochenschrift (NJW) 1998, 86. 19 Case 17/74 Transocean Marine Paint Association v. Commission [1974] ECR 1063, para. 11. See also Kerse C. and Khan N. (2005), supra note no. 133, at para. 4-004. 20 Case C-68/94 & C-30/95 France v Commission [1998] ECR I-1375, paras. 174 et seq. 21 For the US, see Fed. R. Crim. Proc. Rule 11(b)(3) (“Before entering a judgment on a guilty plea, the court must determine that there is a factual basis for the plea.”) and Plea Agreement Guidelines, supra note no. 3, at D. 1. (b) (“Selection to Charge or Charges”).


16/10/07

13:59

Page 619

“Plea Bargaining”/“Direct Settlement” Sytem in EC Competition Law 619 reasoning which has led the Commission to its decision may be understood”.22 However, neither the Treaty nor Regulation 1/2003 provides for the precise form of a Commission decision. Even if a truncated set of facts and abbreviated legal reasoning will theoretically limit the ability of the courts to ascertain the reasons why the measure was adopted,23 the court will also have to consider the fact that the undertaking has already admitted to the allegations and has agreed to the expedited procedure, including a “short form” decision. At any rate, in an extreme case, the Commission could withdraw the decision and issue a new, fully reasoned one if necessary to have some comfort in the appellate procedure.

3. Open questions Even in view of the compelling reasons warranting the implementation of a “direct settlement” procedures in Europe, and with the necessary safeguards imposed by the precept of the rule of law, there remain different and difficult open questions that merit further discussion. They concern—among others— the correlation between the “direct settlement” system and the leniency programme (see point 3.1.)), the necessary measures to prevent abuses (point 3.2.)), the potential effects on private enforcement (point 3.3.)), the position of co-defendants and third parties in the “plea bargaining” process (point 3.4.)), and, finally, the potential impediment to the cooperation of national competition agencies (point 3.5.)).

3.1. The correlation between the “direct settlement” system and the leniency programme Undertakings will generally be more willing to make concessions in a “direct settlement” if they receive a predictable reduction in fines, even if they have already successfully applied for leniency. But even if the Commission would not generally reward a company by reducing the fine, the mere fact that, in direct settlement cases, the Commission would only concentrate on the “core of the infringement” could lead to a reduction of the number of counts for the offence. The Commission will thus inevitably waive the fining of any “marginal infringements”, and the overall fine could be reduced. Ideally, the undertaking should be able to put a specific “price tag” on its concession. In any case, the relationship between the envisaged “direct settlement” system and the Commission’s leniency programme warrants a more detailed look at the following questions: 22 23

Case 24/62 Germany v. Commission [1963] ECR 63, para. 69. Case C-338/00 P Volkswagen AG v. Commission [2003] ECR I-9189, at para. 124.


620

16/10/07

13:59

Page 620


—Should settlement proceedings be available to undertakings where the Commission is already in possession of ample evidentiary material proving an undertaking’s anticompetitive conduct and where the undertaking thus could not benefit from the leniency programme? The question should be answered in the affirmative, as the Commission is nevertheless relieved of having to pursue the case by the simplified “direct settlement” procedure. —Should an undertaking which could only receive partial leniency through the leniency programme, be able to benefit from a direct settlement with the Commission, and thus further reduce the fine? Here too the answer should be affirmative. Although the undertaking participating in the leniency programme has already had to admit to the allegations in the leniency procedure, when agreeing to a settlement the company not only pleads guilty but also waives a number of important procedural rights, which in turn alleviates the strain on the Commission’s resources. Therefore, the settlement procedure should also be available in this case. —Should undertakings also enjoy a (virtual) reduction in fines by way of settlement if they were exposed by co-conspirators who in turn applied for the leniency programme? In order to uphold the incentives for conspiring undertakings to turn on their co-conspirators and to blow the whistle, the Commission should ensure that the effective reduction in fines through the “private settlement” procedure is not disproportionate compared to the reduction granted to successful leniency applicants. —Should the “direct settlement” system also be available to those undertakings who cannot benefit from the leniency programme because they took “steps to coerce other undertakings to participate in the infringement”?24 As the leniency programme is intended to uncover anticompetitive conduct and the settlement system is designed to facilitate the prosecution of such conduct, they serve different purposes. Therefore, these undertakings should also be able to benefit from a direct settlement, as their guilty plea to the allegation will significantly ease the burden on the Commission and guarantee “swift” and “timely” punishment.

3.2. Prevention of abuse Any procedure introducing a “direct settlement” system will have to provide for safeguards against potential abuses. Such abuses could emanate from either the undertakings concerned or the Commission or both. On the one hand, a “direct settlement” should not be exploited to sidestep the prohibition of self-incrimination, which has basically been acknowledged 24 See Commission Notice on immunity from fines and reduction of fines in cartel cases, OJ C 45 [2002], para. 11(c).


16/10/07

13:59

Page 621

“Plea Bargaining”/“Direct Settlement” Sytem in EC Competition Law 621 by the CFI and the ECJ.25 The undertaking concerned consequently may not be pressured into admitting to anticompetitive conduct, whether or not it did in fact commit it. To that end, the Federal Rules of Criminal Procedure in the US, for example, provide, in Rule 11(b)(2), that in the course of a public hearing a judge must ascertain that the plea agreement “did not result from force, threats, or promises”. It seems unlikely that the CFI would become involved in the course of the settlement proceedings, as any actions of the Commission at this stage would be very unlikely to qualify as “decisions”. Consequently, there will only be an ex post judicial control: the party to the “direct settlement” will have to appeal the decision, although the judges of the court will most likely ascribe to its guilty plea an important evidentiary value. However, the court will also be required to investigate—similar to their US colleagues—whether or not the settlement was reached voluntarily, i.e., without coercion. If the latter were the case, the guilty plea would be void or would otherwise be precluded from being used as evidence, and the Commission would have to fully prove the alleged anticompetitive conduct without being able to rely on the undertaking’s tainted admission. On the other hand, the undertakings concerned may also abuse the settlement process in order to reduce the risk of exposing the full scale of the anticompetitive conduct: When admitting to an infringement of the EC competition rules, the undertakings may in fact be trying to divert the Commission from further investigation of the matter, thus obscuring the full extent of the conduct. Under these circumstances, further discussion is warranted concerning the extent to which such a settlement is binding and under which conditions it may be revoked. Any abuses of this kind can be avoided if settlement negotiations only proceed after the main issues of the case have been sufficiently investigated.

3.3. Effects on private enforcement When establishing a “direct settlement” system, the Commission will also have to consider its possible effects on private antitrust enforcement. Depending on its conception, a settlement with the Commission might in fact hinder private enforcement in follow-on cases. If the settlement reached does not result in a detailed formal decision, the private plaintiff will not be able to alleviate its burden of proof with regard to the infringement.26 Consequently, 25 Case 374/87 Orkem v. Commission [1989] ECR 3283, at paras. 34 et seq.; Case T-112/98 MannesmannRöhren-Werke v. Commission [2001] ECR II-729. at paras. 78 et seq. See also Kerse C. and Kahn N. (2005), supra note no. 13, at para. 3-016 f. 26 See § 33(4) GWB; Article 16(1) of Regulation 1/2003; and the Commission’s proposals in its Green Paper, Damages actions for breach of the EC antitrust rules, COM(2005) 672 final, Question C Option 8–10.


622

16/10/07

13:59

Page 622


in order to gain a potential advantage in damages actions at the national level, it will likely be in the interest of the undertaking concerned to bargain with the Commission and avoid a fully reasoned decision. Direct settlements as formal decisions according to Article 16(1) of Regulation No 1/2003 can have legally binding effects on subsequent damage cases before national courts, as far as the violation of Article 81 as such is concerned, as, e.g., in Germany according to § 33 sub. para. 4, sentence 1 GWB. Having admitted to an infringement as a condition for the settlement would in any case preclude the undertaking concerned from arguing that there had been no violation. Settlement decisions as well as ordinary decisions normally have no or only a minor impact on the question of causation and the amount of damages. Having to admit to the infringement does not necessarily diminish the practicability of private settlements. The undertaking concerned will have to weigh the prospect of damage suits for the “core” infringements admitted to in the settlement against the possible reduction in fines for its cooperation and the amount of potential damages claims if the entire anticompetitive conduct were fully investigated and described by the Commission in the decision, thereby making the calculation of damages easier for the complainant.

3.4. “Direct settlement” with respect to co-defendants and third parties Furthermore, the “direct settlement” system will have to include provisions with respect to its effects on third parties and their participation in the proceeding. On the one hand, the Commission has to observe the fundamental principle of equality before the Law. For that purpose, the Commission should issue a Notice on how it intends to enter into settlement in future cases, thus creating predictability for the undertakings concerned. It needs to be clarified whether or not the Commission, when it receives an application, will be obliged to enter into a settlement proceeding or whether this should be left to the Commission’s discretion. As the settlement system is primarily designed to reduce the workload of the Commission and to free up its resources, the latter seems more appropriate, as long as the relevant aspects on which the Commission bases its determination are made available to the public in advance, preferably in a Notice. Further, the Commission will need to clarify the extent to which third parties can get involved in the settlement process. In principle, these third parties should not be integrated into the negotiation procedure.27 The settlement should be struck solely between the Commission and the undertaking 27

Plea Agreement Guidelines, supra note no. 3, at B. (“Plea Agreement Negotiations”).


16/10/07

13:59

Page 623

“Plea Bargaining”/“Direct Settlement” Sytem in EC Competition Law 623 concerned. This would not violate the fundamental concept of a transparent and public proceeding, since—unlike criminal proceedings—in antitrust cases the relevant facts are regularly published after the decision has been rendered.28 However, this should not preclude the Commission from publishing an envisaged settlement on its website or in the Official Journal in order to invite comments from interested third parties similar to the procedure for “commitments decisions” established by Article 9(1) of Regulation 1/2003. In order to have an impact on the outcome of the settlement, these comments would have to be submitted before the settlement becomes final. During that time, third parties might reveal valuable information to the Commission regarding the nature and the extent of the alleged infringement. This could also reduce the potential for abuse by the undertakings concerned if they willingly admit to only parts of the anticompetitive conduct in order to prevent the Commission from discovering the full extent of the infringement.

3.5. Potential impediment to the cooperation of national competition agencies While especially on both sides of the Atlantic members of the respective competition authorities have reiterated the increasing importance of cooperation between the different national competition agencies in their fight against international cartels,29 the implementation of a “direct settlement” system might encroach upon this successful practice. While the first builds on the exchange of non-confidential information between the agencies, the latter may be used by the undertakings concerned to limit the information available to one national agency in order to prevent any other information from finding its way into another jurisdiction. This was already a concern for the Commission in 1996. In cases “where the US authorities have obtained a judgment on the basis of a plea bargain, only a severely limited amount of information will have been made publicly available. This restriction on the availability of information may limit the extent of cooperation.”30 Therefore, 28 Article 30 of Regulation 1/2003. Compare Kerse C. and Khan N. (2005), supra note no. 133, at para. 6-047. But see, for US criminal proceedings, Fed. R. Crim. Proc. Rule 11(b)(1). 29 See, e.g., Hammond S. (2006): “Charting New Waters in International Cartel Prosecutions”, speech delivered at The Twentieth Annual National Institute on White Collar Crime, 2 March 2006, text available at http://www.usdoj.gov/atr/public/speeches/214861.htm; Monti M. (2004): “International Antitrust—A Personal Perspective”, speech delivered at the Fordham Corporate Law Institute, New York, 7 October 2004, text available at http://europa.eu/ rapid/pressReleasesAction.do?reference=SPEECH/04/449&format=HTML&aged=0&language= EN&guiLanguage=en. 30 Commission Report to the Council and the European Parliament on the application of the Agreement between the European Communities and the Government of the United States of America regarding the application of their competition laws, COM(1996) 479 final, reprinted in European Commission (1997), Brussels, at p. 314.


624

16/10/07

13:59

Page 624


the implications of the proliferation of “plea bargaining” systems on international cooperation in antitrust matters will have to be taken into account when conceiving the “direct settlement” system in Europe. This may especially be an issue for the exercise of discretion in deciding whether a direct settlement procedure is appropriate.

D. Outlook Left with the impression that the implementation of a “direct settlement” system warrants careful analysis of core principles and might in fact involve a system change in the antitrust enforcement scheme in Europe, it is already safe to say that Commissioner Kroes has initiated an interesting debate not only on the Community level but also in the Member States. At this time, it seems that not only will the enforcement agencies profit from the accelerated procedure; the settlement will also prove beneficial for the undertakings concerned.

(Ze) Ehlermann 07 Session C Roth

12/10/07

13:14

Page 625

SESSION C


2 PANEL VII: ENSURING THAT EFFECTIVENESS OF ENFORCEMENT DOES NOT PREJUDICE LEGAL PROTECTION—RIGHTS OF DEFENCE

WORKING PAPERS


12/10/07

13:14

Page 626


12/10/07

13:14

Page 627

I Peter Roth QC 1 Ensuring that Effectiveness of Enforcement Does Not Prejudice Legal Protection. Rights of Defence. Fundamental Rights Concerns

A. Introduction It is unsurprising that the continuing endeavour to achieve more effective enforcement of competition law against cartels involves the introduction of enhanced means of investigation, the imposition of heavier penalties (including the possibility of sanctions on individuals) and, more recently, the encouragement of private actions to serve as an additional deterrent.2 But it is inevitable that such developments bring into sharper relief the protection of the fundamental rights of those subject to investigation. The same tension becomes manifest in the pursuit of illegal cartels as in the prosecution of other areas of wrongdoing: the desire on the one hand to have the most effective detection and punishment of conduct that does substantial harm to society; and the concern on the other hand to ensure the preservation of fundamental norms developed to protect the innocent and sustain a high standard of conduct by public authorities. The issues involved become more complex now that there are parallel regimes of enforcement by the Community institutions and the authorities of the Member States, applying the same substantive law but with differences in domestic procedure, levels of protection, and penalties. Among the areas in which fundamental rights may be engaged are the following: a) At the investigation stage: Evidence gathering: (i) documentary evidence: “dawn raids”, especially at private homes of directors or other employees; intrusive forms of surveillance; documents setting out legal advice; (ii) requests for information: questions directed at enterprises and at individuals. 1 Barrister, Monckton Chambers, Gray’s Inn, London; Visiting Professor of Law, King’s College, London 2 The deterrent effect of private actions is highlighted at the outset of the Commission’s Green Paper on Damages actions for breach of the EC antitrust rules, COM (2005) 672 final, which describes private enforcement as complementary to public enforcement (para. 1.1.).


628

12/10/07

13:14

Page 628


Exchange of information as between the Commission and National Competition Authorities (“NCAs”) and between NCAs directly; and with authorities outside the EU. b) In prosecution proceedings: (i) The rights of defence, including access to the file. (ii) The problem of parallel proceedings (ne bis in idem) c) In judicial control: in a system where the finding of infringement and sanction are determined by an administrative body d) As regards duration: in all the areas (a)–(c) above. This paper first looks at the relevant sources of fundamental rights in the EU and notes some of the particular features of the EC competition law that should inform the discussion. It then considers by way of illustration the issues raised by fundamental rights concerns in the investigation stage, i.e. (a) above.

B. Sources of rights The two main instruments enumerating fundamental rights within the EU regime are the European Convention on Human Rights (“ECHR”) and the EU Charter of Fundamental Rights (“the Charter”).

1. The ECHR As is well known, the ECHR is not directly binding on the EU institutions and there is no recourse to the European Court of Human Rights (“ECtHR”) against a decision or legal instrument by such an institution. However, the Community Courts have long recognized the special significance of the ECHR as a statement of fundamental rights recognized by all the Member States.3 Article 6(2) of the Treaty on European Union declares that the Union shall respect the fundamental rights protected by the Convention as they result from the constitutional traditions common to the Member States. And the Community Courts have often referred to the jurisprudence of the ECtHR when assessing fundamental rights under Community law. Indeed, the ECJ in PVC II acknowledged that the Community judicature must have regard to developments in the jurisprudence of the ECHR since a question 3

Opinion 2/94 on Accession by the Community to the ECHR [1996] ECR I-1759.


12/10/07

13:14

Page 629

Ensuring Enforcement Does Not Prejudice Legal Protection 629 was last considered under EC law.4 Therefore, in the application of these fundamental rights the Community, in effect, accords deference to another supranational court under a differently constituted legal regime. Moreover, although the acts of the EU institutions are not justiciable under the ECHR regime, that regime does apply to the acts of the individual Member States. The ECtHR has held that where a Member State is acting to comply with its obligations that flow from Community law, its actions are subject to the ECHR.5 Since almost all action by the EU institutions ultimately relies for enforcement on the relevant Member State, this constitutes a form of indirect control by the ECtHR. Significantly, the Strasbourg Court has held that the protection of fundamental rights afforded by the Community was sufficient to give rise to a presumption of Convention compliance by the State in implementing its obligations under Community law.

2. The Charter The Charter adopted in Nice in December 2000 is in the form of a “solemn proclamation” without express legal effect. It is incorporated in the EU Constitution, but the prospect of the Constitution now being adopted seems remote. Nonetheless, as the Charter is expressed to be a reaffirmation of the rights that result from the existing constitutional traditions and international obligations of the Member States (including the ECHR), it seems appropriate for the Community Courts to refer to it in that regard.6 The CFI and several Advocates General have notably prayed in aid various provisions of the Charter in a series of judgments7 and Opinions.8 The ECJ has yet to refer to the Charter,9 but significantly it has been referred to by the ECtHR as reflecting developments since the adoption of the ECHR.10 4

Cases C-238/99 P, etc., [2002] ECR I-8616, para. 274. Bosphorus v. Ireland, No. 45036/98, 30 June 2005. 6 In Case T-210/01 General Electric v. Commission [2005] ECR II-5575, para. 725, the CFI observed on this basis that the Charter rights there invoked are already protected as fundamental principles of EC law. 7 E.g.: Cases T-67/00 etc., JFE Engineering Corp v. Commission, [2004] ECR II-2501, para. 178; Case T-116/01 P&O European Ferries (Vizcaya) SA v. Commission [2003] II-2957, para. 208; Case T-54/99 Max-Mobil v. Commission [2002] ECR II-313, para. 48; Case T-38/02 Group Danone v. Commission, judgment of 25 October 2005, [2005] ECR II-4407, para. 216. 8 E.g.: Case C-94/04 Cipolla v. Fazari, Opinion of AG Poiares Maduro of 1 February 2006, para. 78; Case C-119/04 Commission v. Italy, Opinion of AG Poiares Maduro of 26 January 2006, para. 35; Case C-301/04 P Commission v. SGL Carbon, Opinion of AG Geelhoed of 19 January 2006, para. 64; Case C-105/04 P Nederlandse Federatieve Vereniging voor de Groothandel op Elektrotechnisch Gebied v. Commission, 8 December 2005, Opinion of AG Kokott, para. 107. 9 Indeed, the ECJ declined to do so even when the Charter was expressly the subject of a question by the referring national court: Case C-347/03 ERSA v. Ministero delle Politiche Agricole e Forestali [2005] ECR I-3785. 10 Goodwin v. United Kingdom, No 28957/95, 11 July 2002, para. 100 (referring to Art II-9). 5


630

12/10/07

13:14

Page 630


Whereas the Charter includes provisions that correspond to those in the ECHR, and in that respect is to be interpreted the same way,11 the difference in wording is sometimes significant and the Charter contains rights that go considerably beyond the ECHR. In the present context, Article II-41 of the Charter is potentially significant, setting out a “right to good administration”, which includes the right to have one’s affairs handled by the EU Institutions within a reasonable time, and is clearly distinct from the more specific right to a fair trial under Articles II-47/48 that corresponds to Article 6 ECHR. Furthermore, in the specific context of the enforcement of competition law, Recital (37) of Regulation 1/2003 specifically states: “This Regulation respects the fundamental rights and observes the principles recognised in particular by the Charter of Fundamental Rights of the European Union. Accordingly, this Regulation should be interpreted and applied with respect to those rights and principles.”

It is thus not surprising that in her Opinion in the Dutch FEG case, Advocate General Kokott observed that the Charter “must be taken into account in cartel proceedings”.12 However, although the Charter may therefore be regarded as binding on the EU institutions, it is not directly applicable against the NCAs in the courts of the Member States. Whether it will be applied judicially to NCAs insofar as they are acting pursuant to Regulation 1/2003 remains to be seen.13

C. Distinctive features of competition law in the EU A number of aspects of competition law should be borne in mind when considering particular issues in the application of fundamental rights.

1. Corporate and individual defendants Undertakings for the purpose of the competition rules may be legal or natural persons. However, in the context of cartels, Article 81 EC has been applied

11

Art 52(3). Case C-105/04P, supra note 8, at fn. 59. 13 Art II-51(1) states that the Charter applies to the Member States when they are implementing EU law, but that begs the question of the current legal status of application of the Charter in the first place. 12


12/10/07

13:14

Page 631

Ensuring Enforcement Does Not Prejudice Legal Protection 631 almost exclusively to legal persons.14 This is distinct from the position regarding decisions of associations of undertakings, where the undertakings in question may be individual professionals, but there too Article 81 EC prohibition has been applied to the association and not to its members.15 Yet if a participant in a cartel were a business that did not have distinct legal personality (e.g. a partnership in English law), the Article 81 EC prohibition would doubtless have to be applied to natural persons. The two provisions of fundamental rights that most frequently arise in the competition law context are those enshrined in Article 6 ECHR16 (the right to a fair trial) and Article 8 ECHR17 (the right to respect for private and family life). There is no reason why many of the fair trial principles should not apply with equal force to corporate as to individual defendants, but it is under Article 6 that the right against self-incrimination has been adumbrated and it may be thought appropriate in that regard to draw at least some degree of distinction between natural and legal persons. In the United States, the privilege against self-incrimination under the 5th Amendment to the US Constitution does not apply to corporate parties.18 It is under the scope of Article 8 that the ECHR applies to dawn raids, but there again it may be that a lesser degree of protection is appropriate for corporate than for domestic premises.

2. Commercial confidentiality It is axiomatic that the enforcement of competition law involves intervention in a market, and this will itself affect competition between undertakings. Almost any antitrust system must grapple with the problem of protecting sensitive commercial information from disclosure to competitors or customers whilst still allowing adequate investigation and scrutiny of suspected cartelists. The complainants to the administrative authority (or indeed claimants in a private action) will be either competitors or customers of the undertaking which will therefore have a particularly acute interest in maintaining the confidentiality of its commercial information vis-à-vis those parties. In this context, human rights arguments may pull both ways, with complainants wishing to rely on their Article 8 or equivalent right to confidentiality, and defendants wishing to rely on their Article 6 or equivalent right to know the case against 14 In Case T9/99 HFB v. Commission [2002] ECR II-1530, para. 105, the CFI noted that the Commission had intentionally not established liability with respect to the individual who controlled a group of companies in the Pre-Insulated Pipes cartel but had instead imposed fines on the companies. 15 E.g., Belgian Architects Association, OJ l 4 [2005] (Summary). 16 Under the Charter, Arts. II-47 and 48. 17 Under the Charter, Art. II-7. 18 Hale v. Henkel, 201 U.S. 43 (1906) (regarding the antitrust laws); Braswell v. United States, 487 U.S. (1988) (corporate officer).


632

12/10/07

13:14

Page 632


them. Accordingly, fundamental rights arguments are simply one factor amongst many. Hence in considering the confidentiality of evidence held by the Commission, the ECJ held that: “[. . .] as the Commission and the German and United Kingdom Governments have rightly observed, the Commission’s ability to guarantee the anonymity of certain of its sources of information is of crucial importance with a view to ensuring the effective prevention of prohibited anti-competitive practices.”19

3. Administrative authorities and criminal charges The question of whether or not policy is best served by combination of the roles of prosecutor and judge in a single competition authority is not within the scope of this paper. But it is a characteristic feature of EC competition law, both at Community level and in most Member States. Despite earlier rulings to the contrary by the ECJ,20 in the light of the recent case law of the ECtHR the Community Courts now acknowledge that an infringement of Article 81 constitutes a “criminal charge” in terms of fundamental rights protection and, specifically, Article 6 ECHR,21 irrespective of its status under Community or national law. It is beyond dispute that the Commission, and similarly a NCA of a Member State, does not constitute an independent judicial tribunal. The result is that the determination of a criminal charge is made in the first instance by a body that does not satisfy human rights standards. The accepted solution has been that the right of appeal to an independent and impartial tribunal can render the proceedings as a whole compatible with Article 6 ECHR;22 and the CFI is undoubtedly such a tribunal. This approach was indeed taken by the CFI in Enso Española v. Commission;23 the Court repeated that the Commission was not a tribunal for the purposes of Article 6, but observed that the judicial control which it exercised over the Commission and its power under Article 230 EC to examine whether the Commission’s conclusions were soundly based in fact and law satisfied the requirements of Article 6; moreover, under Article 229 it had unlimited jurisdiction over decisions imposing a fine. 19

Case C-94/00 Roquette Frères [2002] ECR I-9039, para.64. E.g., Cases 209/78 etc. Van Landewyck v. Commission [1980] ECR 3125. 21 Case C-199/92 P Hüls v. Commission [1999] ECR I-4287, paras. 149 and 150; Case C-235/92 P Montecatini v. Commission [1999] ECR I-4539, paras. 175 and 176. The admissibility decision of the ECtHR in OOO Neste St Petersburg v. Russia, 3 June 2004, does not affect this conclusion since it turned largely on the particular nature of the Russian legislation in question which did not involve penal or deterrent sanctions. 22 Le Compte et al v. Belgium, A/43 (1981), para. 51. And see Bryan v. UK, A/335-A (1996) 21 EHRR 342, paras. 32–47. 23 Case T-348/94, [1998] ECR II-1875, paras. 55–65. The point was not pursued on appeal: C-282/98 P Enso Española v. Commission [2000] I-9817. 20


12/10/07

13:14

Page 633

Ensuring Enforcement Does Not Prejudice Legal Protection 633 However, to satisfy human rights norms, this may be said to have involved an expansion of the scope of jurisdiction in the CFI when hearing appeals. Moreover, while it is generally assumed that this analysis applies equally to a criminal charge involving significant penalties as it does to the determination of civil rights and obligations, the former has not been established by the ECtHR.24 It is notable that under UK competition law, an appeal against a penalty for infringement imposed by the NCA brings an automatic suspension of the penalty, precisely because of such human rights concerns.25 Under EC law, there is no such suspension although in the practice the Commission does not demand payment of the penalty against provision of appropriate security.

D. Application in cartel investigations 1. Inspections at premises (“dawn raids”) Under Arts 20 and 21 of Regulation 1/2003, the Commission may carry out inspections at, respectively, premises of the undertakings concerned or premises of any director, manager or other member of staff of the undertaking concerned. These powers engage the protection of the home under Article 8 ECHR and the equivalent Article II-7 of the Charter. Although when first confronted with the argument the ECJ held that Article 8 did not apply to business premises,26 the subsequent case law of the ECtHR has established that business premises will constitute a “home” for this purpose, including the premises of an undertaking that is a legal person.27 As the ECtHR observed in Colas est when considering a dawn raid carried out by the French competition authority, there is no doubt that the enforcement of the law against cartels is “manifestly in the interest of both ‘the economic wellbeing of the country’ and ‘the prevention of crime’.” Such searches for documents therefore have a legitimate aim.28 The question is whether there is sufficient control to ensure that the search is proportionate and subject to sufficient safeguards. The ECJ addressed this in detail in 24 Schmautzer v. Austria, Series A, No 328-A [1995] 21 EHRR 511, para. 36; Umlauft v. Austria [1996] 22 EHRR 76, paras. 37–39; but cf. Findlay v. UK [1997] 24 EHRR 221, para. 79. 25 Competition Act 1998, sect. 46(4). The Act also makes clear that the appeal is a full appeal on the merits: Sched 8, para 3(1). 26 E.g., Cases 46 & 227/88 Hoechst v. Commission [1989] ECR 2859, para. 18. 27 Niemietz v. Germany, Ser A no. 251-B, judgment of 16 December 1992; Colas Est v. France, Applic. No. 37971/97, judgment of 16 April 2002, para 41. 28 Supra note 27, at para. 44.


634

12/10/07

13:14

Page 634


Roquette Frères,29 and the drafting of Articles 20(7)–(8) and 21 of Regulation 1/2003 are clearly intended to reflect the ruling in that important judgment. Hence:30 a) the decision imposing an inspection must specify the subject matter and purpose of the inspection; b) for an inspection at the premises of the undertaking, where the Commission officials need assistance from the national authorities (as it seems would always be the case) prior judicial authorization is required when the national rules of the Member State in question so provide. But for inspection at the home of a director or other staff member of the undertaking (“private home”), prior judicial authorization is always required as a matter of EC law; c) for an inspection at a private home, the Commission must have a “reasonable suspicion” that there are kept at those premises books or other records “which may be relevant to prove a serious violation of Article 81 or Article 82”; and the decision must state the reasons that have led the Commission to conclude that such a suspicion exists. d) the control that may be exercised by the judicial authority is circumscribed. The wording used is slightly different as between Arts. 20(8) and 21(3), but in summary: a. the national court may ask for explanations of the grounds on which the Commission suspects an infringement, the seriousness of the infringement, and the nature of the involvement of the undertaking concerned; and in the case of inspection at a private home, of the reasonable likelihood of the books and records being kept at those premises; b. in the case of inspection at a private home, those matters are stated to be among the factors that the court can take into account to “control” that “the coercive measures envisaged are neither arbitrary nor excessive” and the court may ask the Commission for detailed explanations on those elements that are necessary to control the proportionality of the coercive measures envisaged; c. however, whether the authority is sought for an inspection at the premises of the undertaking or at a private home, the national court may not call into question the necessity for the inspection or demand that it be provided with information in the Commission’s file. d. the lawfulness of the Commission’s decision to make an inspection may be reviewed only by the Community Courts. 29

Supra note 19. For a thorough analysis of these provisions in Regulation 1/2003, see Wils W. P. J. (2006): “Powers of Investigation and Procedural Rights and Guarantees in EU Antitrust Enforcement: The Interplay between European and National Legislation and Case-law”, 29 World Competition 3. 30


12/10/07

13:14

Page 635

Ensuring Enforcement Does Not Prejudice Legal Protection 635 The distinction drawn between the level of control over inspections at the premises of undertakings and private homes seems to reflect the varying intensity that should apply to Article 8 protection, as was acknowledged, although not actually decided, by the ECtHR in Colas est.31 These provisions may be seen as reflecting the need to provide for ex ante control by the national court and the desire to avoid transmitting too much evidence from the Commission which may be highly confidential. As the ECJ stated in Roquette Frères: “Clearly, if the Commission were obliged to send to the various national competition authorities factual information and evidence revealing the identity of its sources of information, or enabling that identity to be deduced, that might well increase the risks to informants of disclosure of their identity to third parties, in view, particularly, of the procedural requirements of national law. Second, it must also be borne in mind that the physical transmission to the competent national authorities of the various items of factual information and evidence held in the Commission’s file could give rise to other risks as regards the effectiveness of the action taken by the Community, especially in cases involving parallel investigations to be carried out simultaneously in more than one Member State. Account must be taken of the uncertainties and delays that may affect such transmission and the different procedural rules with which they may have to comply under the legal systems of the Member States concerned, as well as the time which the authorities in question may need to consider potentially complex and voluminous documents.”32

This may be considered to be a sensible compromise between expediency and protection in the context of multi-national cartel investigation. Moreover, in the context of an application for judicial authority to conduct a dawn raid on business premises under national competition law, the English High Court was satisfied that the equivalent control mechanisms were sufficient to ensure proportionality and come within the justification of Article 8(2) ECHR.33 On the other hand, it may be thought that some problems remain: a) where the undertaking is a natural person and the dawn raid is to be conducted at his private home, he should receive the higher level of protection for a search of a private home, although it seems that only the lower level of judicial control under Article 20(8) would apply. b) Neither Colas est nor Roquette Frères was concerned with a search of a private home. Subsequently, the ECtHR has considered the compatibility of search and seizure measures at a private home with Article 8. While noting that a State may well consider it necessary to provide for such inspections to obtain evidence in a sphere where it is otherwise impossible to 31 32 33

Supra note 27, at para. 49 (this point was there conceded). Roquette Frères, supra note 19, paras. 65–66. Office of Fair Trading v. X, [2003] EWHC 1042, [2003] 2 All ER (Comm) 183.


636

12/10/07

13:14

Page 636


detect the guilty, the Court held that: “having regard to the severity of the interference with the right to respect for his home of a person affected by such measures, it must be clearly established that the proportionality principle has been adhered to.”34 It may be questioned how far a national court can satisfy itself as to the proportionality of an application before it without the right to see any of the evidence held by the authority making the application.

2. Attorney-client privilege Following the well-known AM&S case,35 under EC law only communications between an undertaking and an independent lawyer licensed to practice in the Community are protected from compulsory disclosure. This is different from the position in several Member States36 in two respects: a) the protection does not extend to advice from in-house Counsel; b) the protection does not extend to communication with members of the legal profession in a non-Member State. The first, but not the second, of these limitations is now under consideration before the CFI in the pending case of Akzo-Nobel, where the President of the CFI, in an interim ruling, noted that it raised an issue of principle “which merits very special attention”.37 For those who are understandably determined to ensure effective detection of cartels, protection of communications with in-house lawyers from scrutiny is regarded as the unwarranted shielding of potentially valuable evidence. The argument against is portrayed as founded on expediency, and it has been asked why companies should complain when there is a wide choice of independent lawyers available, particularly when those companies with the resources to have in-house counsel can well afford to take outside advice.38 As against this, it may be thought inappropriate, and imposing wholly unjustified costs, to compel a company to take external advice when it has a perfectly competent source of expertise on its staff, probably with greater knowledge of the circumstances of the business. However, the real issue concerns not practical convenience but fundamental rights. The privilege from disclosure of lawyer-client communications is 34

Buck v. Germany, Appl no 41604/98, judgment of 28 April 2005, para. 51. Case 155/79 AM & S v. Commission [1982] ECR 1575. 36 E.g., the United Kingdom and Ireland. Belgium also respects the confidentiality of communications with in-house counsel. 37 Cases T-125/03 R & 253/03 R Akzo-Nobel v. Commission [2003] ECR II-4771, para. 120; appeal allowed, Case C-7/04 P(R), [2004] ECR I-8739. 38 See Wils, supra note no. 30. 35


12/10/07

13:14

Page 637

Ensuring Enforcement Does Not Prejudice Legal Protection 637 not based on expediency but on the right of any person to obtain skilled advice about the law, which cannot be effectively obtained unless the client can place all facts before the adviser without fear that they may afterwards be disclosed and used against him.39 As regards common law jurisdictions, Lord Taylor CJ stated in the House of Lords in 1996:40 “[. . .] if a balancing exercise was ever required in the case of legal professional privilege, it was performed once and for all in the 16th century, and since then has applied across the board in every case, irrespective of the client’s individual merits.”

But the present-day rationale for the rule is not its antiquity but its function in promoting the rule of law. It is obvious that the privilege means that sometimes relevant probative material will be unavailable to the authorities. But the privilege recognises the public interest that in a complex world individuals and corporations need to seek advice from lawyers in order to arrange their affairs in an orderly fashion; and that it is necessary that the full facts are placed before their lawyers to achieve this result. That is something that can only be achieved in all cases if the client can be confident that he can speak with complete candour.41 The ECtHR has recognised that interference with the confidentiality of lawyer-client communications involves violation of Article 8 ECHR, which can be justified only in exceptional circumstances.42 As regards in-house counsel, the issue is complicated by the variation as between Member States of the position of in-house counsel in the context of the rules of their professional Bar. In some, such as the United Kingdom and Ireland, an employed solicitor or barrister remains subject to the disciplinary obligations of his or her professional body. The same situation pertains in some other but not all Member States. Should the rule on protection from disclosure be formulated to reflect such a criterion?43 As regards non-EU lawyers, the international nature of many businesses means that they are likely to receive legal advice from outside as well as within the EU. Moreover, it is a sign of the success of competition law that many jurisdictions throughout the world have enacted prohibitions equivalent to Article 81 EC. If a European company receives advice from its American lawyers about its arrangements from the perspective of US antitrust law, it would obviously be of great concern to the company if that advice could become subject to mandatory disclosure to the European Commission. Again, the concern of those seeking to uphold this limitation in the AM&S 39 R (Morgan Grenfell) v. Special Commissioner [2003] 1 AC 563, per Lord Hoffmann at [7]. This was a tax case where the tax authorities sought access to legal advice regarding a tax avoidance scheme. 40 R v. Derby Magistrates Court, Ex p B [1996] AC 487, 508. 41 See generally Lord Scott in Three Rivers v. Bank of England (No 6) [2005] 1 AC 610, at paras. 30–34. 42 Foley v. United Kingdom, Appl no. 33274/96, judgment of 20 June 2000, paras. 43–44. 43 CFI President Vesterdorf indicated this as a possible way forward in his interim ruling in Akzo Nobel, supra note 37, para. 125.


638

12/10/07

13:14

Page 638


ruling may have in mind some jurisdictions where the professional ethics of lawyers are considered less satisfactory. But as the EU itself increases in size, does it not become more invidious to draw such distinctions? The position is complicated further by the provision for exchange of information between NCAs under Article 12 of Regulation 1/2003. Hence, in the United Kingdom the Office of Fair Trading, in exercise of its powers under UK law,44 could not obtain communications between an undertaking and its in-house counsel or an American attorney; but if such documents were seized by an NCA in another Member State or by the Commission, they could be supplied to the OFT under Article 12. And the OFT has stated that in those circumstances it would consider itself entitled to use those documents in bringing proceedings.45 Such an eventuality would bring into sharp relief the potential contrast between EC law’s current approach to attorney-client privilege and the English law’s view of fundamental rights. It would then be for the English court to resolve that tension.

3. Self-incrimination The privilege against self-incrimination is well-established as a fundamental right and has been held by the ECtHR to be part of the right to a fair trial under Article 6 ECHR. Issues as to the scope of the privilege may arise in the context of documents that are sought by the investigating authority, whether in a dawn raid or pursuant to a request; and in the context of questions addressed to an undertaking or to individuals. The leading case in Community jurisprudence remains the Orkem judgment of 1989, where the ECJ held that a privilege against self-incrimination is not found consistently as regards legal persons (by contrast with individuals) in the case law of the then 12 Member States.46 The Court proceeded to hold that, having regard to the rights of the defence:47 “[. . .] whilst the Commission is entitled, in order to preserve the useful effect of Article 11(2 ) and (5) of Regulation No 17, to compel an undertaking to provide all necessary information concerning such facts as may be known to it and to disclose to it, if necessary, such documents relating thereto as are in its possession, even if the latter may be used to establish, against it or another undertaking, the existence of anti-competitive conduct, it may not, by means of a decision calling for information, undermine the rights of defence of the undertaking concerned.

44 45 46 47

Competition Act 1998, sect 30. OFT Guideline, Powers of Investigation (2004), para. 6.3. Case 374/87 Orkem v. Commission [1989] ECR 3283. Paras. 34–35.


12/10/07

13:14

Page 639

Ensuring Enforcement Does Not Prejudice Legal Protection 639 Thus, the Commission may not compel an undertaking to provide it with answers which might involve an admission on its part of the existence of an infringement which it is incumbent upon the Commission to prove.”

Since Orkem, the ECtHR has considered the scope of the privilege on a number of occasions. The various situations in which the question of selfincrimination can arise involve different considerations.

3.1. Documents It is difficult to discern a consistent line in the jurisprudence of the ECtHR. In Funke v. France,48 it was held that Mr. Funke’s rights under Article 6(1) ECHR were violated when he was required to produce bank statements and financial information by customs authorities investigating charges against him of infringement of exchange control legislation, and then convicted for refusing to supply those documents. However, in its subsequent judgment in Saunders,49 the ECtHR held that the use in criminal proceedings against Mr Saunders of answers obtained under compulsion from him during an investigation by inspectors under the UK Companies Act 1985 was a violation of his rights under Article 6; but observed that “the right not to incriminate oneself . . . does not extend to the use in criminal proceedings of material which may be obtained from the accused through the use of compulsory powers but which has an existence independent of the will of the suspect such as, inter alia, documents acquired pursuant to a warrant . . .”50 Although Saunders was widely regarded as in effect overturning the broad reach of Funke¸ this analysis was cast in doubt by JB v. Switzerland, where the Court then found a breach of the privilege against self-incrimination in the requirement for the production of documents about the applicant’s income as part of a tax investigation.51 Nonetheless, it may be thought that the Saunders distinction provides an appropriate approach to documentary evidence. It respects the rationale of the privilege in avoiding prosecutorial oppression in the context of the right of silence but does not place unreasonable hurdles in the way of detection and prosecution of unlawful activity. In his Opinion in SGL Carbon, Advocate General Geelhoed reaffirmed the Saunders approach as applicable in Community law;52 and in Office of Fair Trading v. X, the English High Court held that the privilege against self-incrimination is not absolute and applied 48

A/256-A, [1993] 16 EHRR 297. Saunders v. United Kingdom, App. No. 19187/91, (1997) 23 EHRR 313. See Riley A. (2000): “Saunders and the Power to Obtain Information in Community and United Kingdom Competition Law”, 25 European Law Review 264. 50 Para. 69. 51 No. 31827/96, judgment of 3 May 2001. 52 See supra note 8, paras. 65–66. 49


640

12/10/07

13:14

Page 640


the ruling in Saunders as a legitimate and proportionate limitation on such a right in the context of a cartel investigation.53

3.2. Questions to undertakings The Orkem distinction as between permissible and impermissible questioning has come in for much criticism. In Orkem itself, the ECJ held that questions as to who attended meetings, in what circumstances and in what capacity, and to seek factual clarification as to the measures taken at the meetings as disclosed by the documents were all allowed. But questions as to the purpose or objective pursued by such measures and for further details of the system whereby sales targets or quotas were attributed to the participants or of methods used to facilitate monitoring of compliance were impermissible. Aside from the confusion to which the Orkem principle can give rise in practice,54 it seems very doubtful that the principle can be reconciled with the subsequent jurisprudence of the ECtHR and, in particular, Saunders. So far, the Community Courts have adhered to the Orkem ruling. In PVC II¸ the ECJ recognised that the case law of the ECtHR had gone through new developments since Orkem of which account would have to be taken, but did not find it necessary to reach a final conclusion on the issue.55 In Mannesmann, the CFI effectively avoided the issue, applying Orkem and saying nothing about Saunders.56 Recital (23) to Regulation 1/2003 restates the Orkem principle as good law. And in his Opinion in SGL Carbon, Advocate General Geelhoed considered that there is no cogent reason to depart from Orkem, noting in particular that the decisions of the ECtHR concerned individuals and that level of protection conferred on legal persons does not need to be at the same level.57

3.3. Questions to individuals Aside from the particular case where an individual is an undertaking, the Commission can address questions to individuals only as regards an explanation of facts or documents relating to the subject matter and purpose of an

53 See supra note 33. However, in Case T-236/01 etc. Tokai Carbon v. Commission, judgment of 29 April 2004, the CFI applied Saunders to hold that certain kinds of documents could not be obtain by compulsory power (at para 408). The Commission has appealed against this ruling. 54 In Case C-60/92 Otto v. Postbank [1993] ECR I-5683, AG Gulmann said that the Orkem principle does not mean that there are certain matters about which questions may not be asked but only that there are certain types of questions which may not be asked (Opinion, para. 10). 55 Supra note 4, at paras. 273–292. 56 Case T-112/98 Mannesmannröhrsen-Werke v. Commission [2001] ECR II-729, paras. 65–67. 57 Supra note 8, at paras. 62–64.


12/10/07

13:14

Page 641

Ensuring Enforcement Does Not Prejudice Legal Protection 641 inspection at the premises of an undertaking.58 However, there is no provision for sanction on an individual for refusing to answer such questions. Nonetheless, effective detection of cartels may surely require greater power of questioning. In UK law, the OFT can not only compel individuals to give an explanation of particular documents, but can address to individuals general requests for information which they are required to answer under penal sanction.59 The protection against self-incrimination is provided by a provision precluding the use of any statement given in response as evidence against that individual in criminal proceedings for the UK criminal offence of making or implementing a cartel agreement.60 The result of such questioning in a cartel case can accordingly be communicated to another NCA under Article 12 of Regulation 1/2003. Article 12(3) seeks to provide protection of individuals as regards self-incrimination in that regard: “Information exchanged pursuant to paragraph 1 can only be used in evidence to impose sanctions on natural persons where: —the law of the transmitting authority foresees sanctions of a similar kind in relation to an infringement of Article 81 or Article 82 of the Treaty or, in the absence thereof, —the information has been collected in a way which respects the same level of protection of the rights of defence of natural persons as provided for under the national rules of the receiving authority. However, in this case, the information exchanged cannot be used by the receiving authority to impose custodial sanctions.”

Where the only sanction at issue is a penalty for supplying incorrect or misleading information, it seems doubtful that the privilege against selfincrimination arises at all. But a number of Member States have introduced criminal penalties for individuals concerned in the management of companies that are involved in a cartel. In those circumstances, concerns about selfincrimination are substantial, and they are scarcely alleviated by the protection against a custodial sanction if an individual may still face a large financial penalty and the stigma of a criminal conviction. Aside from the anxiety that any individual would feel as to how effective is the protection under Article 12(3), it is unclear how this would apply, for example, where the safeguard against self-incrimination is a provision that the information cannot be used by the transmitting authority for the purpose of a criminal prosecution under its domestic cartel law.61 58 Art 20(2)(e) of Regulation 1/2003. There is no equivalent power to seek an explanation of documents seized at a private home under Art. 21. 59 Competition Act 1998, sect. 26(1); compare sects. 26(6)(a)(ii), 27(5)(b)(ii) and 28(2)(e). 60 Competition Act 1998, sect. 30A. The cartel offence is created by sect. 188 of the Enterprise Act 2002. 61 See also the Commission’s Notice on cooperation within the Network of Competition Authorities, para. 28(c), which does not shed light on this issue.


642

12/10/07

13:14

Page 642


The difficulties of ensuring adequate protection in such circumstances highlight the problems that arise as cooperation between national authorities increases while, in order to further deterrence, sanctions on individuals involved in cartel activity are introduced under domestic laws of the cooperating states.

E. Conclusion In his Opinion in SGL Carbon, Advocate General Geelhoed observed that “the interplay between the fundamental rights of legal persons and competition enforcement remains a balancing exercise: at stake are the protection of fundamental rights versus effective enforcement of Community competition law.”62 Where the fundamental rights of natural persons are involved, the equilibrium may be different. Altogether, the above considerations suggest that the application of fundamental rights must be sensitive to the context. The ECHR is a living instrument and its interpretation adapts to the circumstances in which it is engaged. The specialist area of competition investigation and enforcement is far removed from the circumstances envisaged by the framers of the ECHR of 1950. The Charter, by contrast, is a recent document and may be regarded as reflecting current thinking. But the approach to both instruments should be one of critical interpretation in regard to the particular field in which they are applied, while remaining vigilant to protect the fundamental norms which they enshrine. As various initiatives are pursued to improve detection and deterrence of cartels, the implications for the protection of fundamental rights should be borne in mind so that the powers can be fashioned accordingly in a coherent and workable manner.

62

Supra note no. 8, para. 67.

(Zf) Ehlermann 07 Harding

12/10/07

13:15

Page 643

II Christopher Harding* Effectiveness of Enforcement and Legal Protection

A. Changing modalities of enforcement and evolving issues of legal protection It may be helpful at the outset of this discussion of the balance between effectiveness of enforcement and legal protection in the context of the control of business cartels, to note some main features of the legal process in this area of EC competition law. Over the last 30 years the regulation of serious infringements of the EC competition rules, and in particular of cartel arrangements which are contrary to Article 81 EC, has been the trigger for the development of both a strong regulatory mechanism at the supranational level and a concomitant corpus of protective rules for the defence of parties subject to this system of control. In other words, in addition to the intrinsic substantive significance of anti-cartel measures at this level, there has been a development of procedural law and basic rights protection of a highly distinctive (supranational) character and of significance in its own right. The role of enforcement in relation to major trans-European cartels has been and remains in the hands of the Commission, a supranational authority that has possessed, from the beginning, extensive powers of investigation and sanctioning. In this respect the Commission’s role has been largely novel and has for that reason given rise to some distinctive and significant legal issues and concerns. Notably, the concentration of a range of powers in the hands of a single authority and the establishment of a repressive system of enforcement without being able to formally characterise the latter as ‘criminal’ in character has led to a continuing debate concerning the extent and nature of the appropriate legal protection, or defence rights, for parties subject to this process. At the same time, the economic power and interests of those market actors typically subject to the process of control has provided both the motivation and resources for a determined legal challenge of the exercise of enforcement powers. The outcome has been a steady and considerable evolution of new law: big legal business in a practical sense and groundbreaking legal debate at a theoretical level. The appellate process (first to the Court of Justice (ECJ), then to the Court of First Instance (CFI) and the ECJ) has accounted for

* Department of Law, University of Wales—Aberystwyth, UK.


644

12/10/07

13:15

Page 644


thousands of pages of reported jurisprudence in the European Court Reports and spawned a body of supranational quasi-criminal law with a large element of basic rights protection. All of this remains part of competition policy and law, but also now has to a large extent its own character and dynamic. Thus any contemporary discussion of effectiveness of enforcement and legal protection takes us quickly from the broader ocean of competition policy into deeper and deeper waters of regulatory enforcement which need to be navigated by criminal lawyers and human rights lawyers as much as competition lawyers. Over the last ten years the important aspects of enforcement have included leniency programmes (suggestive of ‘plea-bargaining’ in the criminal justice context), appeals relating more and more to the justification for and quantum of fines (suggestive of criminal law sentencing principles and ‘tariffs’) and most recently the uncertain shadow of ‘real’ criminal law sanctions lurking around the corner at the national level. There is a strong dynamic in this sector of regulation that continues to take it forward into uncharted waters. Thus, while much has been achieved and clarified, new issues also spring up very quickly. Retrospectively, we may note the consolidation and firmer delineation of a number of defence rights in the jurisprudence of the European Courts,1 the way in which the CFI has emerged as a kind of trial court, effectively addressing separation of powers concerns about the Commission’s role,2 and the fleshing out of legal provision in both hard law and soft law instruments (Regulation 1 of 2003,3 and guideline notices on the setting of fines and non-implementation of sanctions4). But, as some areas of enforcement practice and their review become more settled, the modalities of enforcement continue to evolve and different issues come to the forefront. At the very least at the present time, the increasing importance of leniency programmes and the development of criminal liability and prospective greater use of civil claims at the Member State level,5 serve to complicate the wider landscape of enforcement and raise new due process questions in relation to the Commission’s enforcement activity. Bearing in mind therefore this significant, distinctive and dynamic character of the EC regulation of cartels, the discussion in this paper will focus upon a number of questions arising from such changing modalities of enforcement which have recently come to the forefront, and also anticipate some issues which may prove of significance in the near future. It is proposed to group these issues for discussion under three broad headings, respectively addressing 1 For a summary view, see Harding C. and Joshua J. (2003): Regulating Cartels in Europe: A Study of Legal Control of Corporate Delinquency, Oxford University Press, Oxford, Chapter 7. 2 Ibid., pp 184–90. 3 Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ L 1 [2003]. 4 Commission Guidelines on the method of setting fines pursuant to Article 15(2) of Regulation 17, OJ C 9 [1998]; Commission Notice on the non-imposition or reduction of fines, OJ C 45 [2002]. 5 See European Commission (2005): Green Paper on damages actions for breach of the EC antitrust rules, 19 December 2005, COM(2005) 672 final.


12/10/07

13:15

Page 645

Effectiveness of Enforcement and Legal Protection 645 three major themes in enforcement activity. First, there will be some consideration of the recent and continuing preoccupation with the application of fines and other sanctions and the related question of leniency. Much of the appellate business being taken through the CFI and ECJ is now concerned with the application of fines, and it may be anticipated that, as the operation of leniency measures produce more evidence, more legal questions concerning the way in which leniency has been applied will rise to the surface. Secondly, the discussion will confront the temporal dimension of cartel proceedings. A significant feature of this kind of legal process is the amount of time taken from the earlier stages of investigation through to final resolution of the appeal proceedings, which may amount to several years in major cases. This typical time-span has increasingly given rise to a number of issues in the process of judicial review and appeal, particularly in relation to the retrospective application of evolving enforcement policy and practice, and to the application of rules and sanctions in a context of changing corporate identities. Finally, there will be some consideration of issues emerging at the interface between European and national levels of enforcement. Member State authorities are acquiring wider responsibilities in relation to the enforcement of EC rules while national competition law enforcement is developing further in its own right. Thus, on the one hand there is the prospect of a more vigorous use of civil claims at the national level as a way of enhancing enforcement of the EC rules, and on the other hand there is the increasing provision of national criminal law sanctions against companies and cartelists infringing national criminal legislation dealing with cartels. The co-existence of European and national enforcement activity is likely to give rise to a number of legal protection issues. Two matters in particular may be identified as potential sites for legal debate. First there is the question of dual application of rules: the same cartels (and the same parties, but with the subsidiary question of how to identify these parties) being dealt with in both European and national jurisdictions. What issues of fair treatment and legal protection arise from this increasingly complex site of enforcement? In the second place, the prospect of enforcement and the consequent need for legal protection across two types of jurisdiction, raises questions concerning consistency of treatment (for instance as between a company being dealt with under EC law, and an individual being dealt with under national criminal law) and the potential conflict (or alignment) of legal protection norms within different jurisdictions (for instance, European Convention on Human Rights (ECHR) and EC law principles). In short, just as there is a need for liaison in enforcement (as set up under Regulation 1/2003) as between the European and national levels,6 so also there should be linkage between the two regarding questions of legal protection. 6 See: Commission Notice on cooperation within the Network of Competition Authorities, 27 April 2004, OJ C 101 [2004]; Joint Statement of the Council and Commission on the Functioning of the Network of Competition Authorities, 15435/02 ADD 1.


646

12/10/07

13:15

Page 646


B. The application of sanctions A survey of appeals against Commission cartel decisions in recent years will quickly demonstrate that the main objective of such appeals has become the removal or reduction of fines that have been imposed by the Commission on companies participating in cartels established to be in violation of Article 81 EC.7 Thus the legal focus of argument in such proceedings has shifted away from aspects of investigation8 to issues of penalty, often challenging the evidence used to justify the sanction or quantum of penalty, but also raising questions concerning the method of quantification used to arrive at the final amount of the penalty.

1. The transparency of penalty calculation Sometimes such arguments give rise to broader questions of legal principle, for instance, in relation to the retroactive application of the Commission’s Fining Guidelines,9 discussed further below. But many such appellate arguments involve more specific ‘tariff’ questions and in such cases the role of the CFI is similar to that of an appellate court dealing with sentencing decisions. Since the Commission adopted its Fining Guidelines in 1998,10 appellants have naturally wished to test decisions against the Guidelines, often arguing for greater predictability in the calculation of fines and against the exercise of much ‘sentencing discretion’ on the part of the Commission.11 In this way the underlining penal philosophy of the sanctioning system, based now on Article 23 of Regulation 1, has been brought into sharper relief. It is clear in principle that the system of fines is intended to have both a retributive and deterrent effect, and that for such purposes the Commission should be entitled to take into account the individual position and circumstances of each defending party. As the CFI stated in an appeal from the Welded Steel Mesh Cartel decision: 7 See for instance the recent discussion by Harding C. and Gibbs A. (2005): ‘Why Go to Court in Europe ? An Analysis of Cartel Appeals 1995–2004”, European Law Review 349. 8 But for a recent exception to this trend, see Joined Cases T-125/03 R and 253/03 R Akzo Nobel Chemicals v Commission, Order of the President of the CFI, 30 October 2003, in relation to the issue of legal professional privilege. 9 Joined Cases C-189/02 P etc. Dansk Rørindustri and Others v Commission [2005] ECR I-5425. 10 For a summary discussion of the content and role of the Guidelines, see : Harding C. and Joshua J., Regulating Cartels in Europe, supra note 1, pp. 242–8. 11 See, Cases C-189/02 P etc, supra note 9, the appeal by members of the Pre-Insulated Pipes Cartel against both the Commission’s decision in that case and the judgment of the CFI. The parties raised several arguments relating to the calculation of the fines, some of which are discussed below.


12/10/07

13:15

Page 647

Effectiveness of Enforcement and Legal Protection 647 “Fines constitute an instrument of the Commission’s competition policy. That is why it must be allowed a margin of discretion when fixing their amount, in order that it may channel the conduct of undertakings towards observance of the competition rules.”12

Yet the view has persisted in some quarters that a fair treatment of ‘convicted’ parties requires a high degree of transparency as to how the amount of fine is calculated, tantamount virtually to a visible application of a precise mathematical formula, so that the amount of likely fine could be predicted beforehand. For instance, some of the parties in the appeal relating to the PreInsulated Pipes Cartel argued (unsuccessfully) before the ECJ that: ‘Traders should therefore be able to assess the benefits of cooperating and should be in a position to calculate in advance the absolute amount of the fine payable, depending on whether or not they decided to cooperate.’13

In particular in that litigation the appellants were unhappy about a shift towards fixing the fine with regard to less predictable ‘gravity’ criteria rather than with reference to the turnover of the participating parties. There would be two consequences of any substantial move away from the use of discretionary criteria. First, the penalty would inevitably reflect damage to the market, as a predictably quantifiable element of the infringement, rather than ‘attitudinal’ factors (such as recidivism, delinquent corporate culture or role in maintaining the secrecy of the arrangement). This does not accord with the rhetoric of delinquency increasingly deployed by competition authorities14 or with the trend towards criminalisation in national legal systems. Secondly, a more mechanistic and transparent calculation of penalties would more easily enable the threat of penalty to be incorporated into the earlier estimation of business risk when deciding on participation in the cartel. In short, it would transform the game from one of penal consequence to one of business calculation, so undermining the deterrent and retributive thrust of Article 23 sanctions. As Spinks has commented, “full transparency would tend to . . . create an environment in which it is easier to manipulate and abuse the enforcement regime”.15 It may be argued therefore that the principle of transparency in calculation of penalties, if pushed to further limits, could compromise the enterprise of enforcement. This point was recognised by the ECJ in its judgment relating the Pre-Insulated Pipes Cartel appeal, when it concluded that: 12

Case T-150/89 Martinelli v Commission [1995] ECR II-1165, at p. 1186. Cases C-189/02 P, supra note 9, at para. 185 of the ECJ’s judgment. 14 For instance, as stated by Neelie Kroes, Commission member in charge of Competition Policy in a speech on 7 April 2005: “Cartels attack free markets at their very hearts. They don’t just mess up the grass on a level playing field. They blow great holes out of the surface. And it is consumers who are asked over and over again to pay the price of replacing the turf. The longterm eradication of cartels is therefore essential [. . .]”. 15 Spink P. (1999): “Recent Guidance on Fining Policy’, 20 European Competition Law Review 101, at p. 108. 13


648

12/10/07

13:15

Page 648


“ [. . .] in fixing the amount of fines, the Commission has a discretion which allows it to raise the general level of fines at any time, within the limits set out in Regulation No 17, if it is necessary to ensure the implementation of the Community competition policy [. . .]”.16

Nonetheless, there may yet be some shift in the direction of greater predictability in fining practice. The Fining Guidelines have been under review by the Commission and National Competition Authorities (NCAs) with the objective of adopting new Guidelines during 2006. There is a likelihood that greater importance will be attached in any new guidelines to the turnover affected by the infringement, which will naturally result in a higher degree of predictability for the level of fine to be set by the Commission. From the Commission’s perspective, greater predictability could be something of an advantage in providing a more certain bargaining chip in possible ‘plea bargaining’ negotiations—the idea of offering at an earlier stage a definite penalty in return for a quick plea of guilty,17 which would then save large resources usually committed to both proving the case and defending appeals. Ironically, the appeal of plea bargaining arises from the higher yield of fruit from the leniency programme: the growing number of applications for immunity has led to the situation in which the Commission is now opening more cases each year than it can conclude.18 As ever, and as in the case of national criminal justice, resources are the problem, and lack of resources leads the system towards the short cut of negotiated justice, compromising the retributive and perhaps also deterrent imperative of the enforcement process. The outcome, in such processes of negotiated justice or ‘direct settlement’ may well be a good, or higher, clear-up rate, but that may not necessarily discourage cartel activity in the first place since the risk of detection, and of a more certain lower penalty, may more easily be absorbed into cartel planning.

2. The ‘financial constraints’ discount A further aspect of penalty calculation is the concern that large, though retributively appropriate, fines may have a devastating effect on the company concerned and drive it towards insolvency. This is an argument which may be taken into account under the ‘ability to pay’ attenuating circumstance listed in paragraph 5(b) of the 1998 Guidelines. It may in some cases be an 16

Cases C-189/02 P, supra note 9, at para. 191. See the speech by Neelie Kroes, 7 April 2005, supra note 14, in its reference to ‘direct settlement’. 18 See Van Barlingen B. and Varennnes M. (2005): ‘The European Commission’s 2002 Leniency Notice in Practice’, 3 Competition Policy Newsletter 6. 17


12/10/07

13:15

Page 649

Effectiveness of Enforcement and Legal Protection 649 especially significant issue for the Commission when applying the EC rules to cartel members,19 since the Commission’s choice of penal sanction is limited to financial penalties as applied to corporate actors and in a broader global context the Commission may well be acting in the wake of penalties already imposed by the US Department of Justice (DoJ) in respect of infringements of the Sherman Act. The Commission may then be faced with the ‘bankruptcy card’: a company arguing that it has already been dealt with severely under US law and a further heavy fine under EC law would push it to the brink of insolvency. Moreover, it may be argued that the Commission, with a wider responsibility in terms of social and economic policy and a concern for competitive structure and minimising economic dislocation to economic infrastructure and unemployment, would be especially sensitive towards such argument. Indeed, the CFI has recently affirmed that, while the Commission is not obliged to take into account the financial position’ of an infringing firm, the financial constraints referred to in paragraph 5(b) of the Guidelines “[. . .] applies only in a specific social context consisting of the consequences which payment of the fine would have, in particular, by leading to an increase in unemployment or deterioration in the economic sectors upstream and downstream of the undertaking concerned.”20

A recent example of the application of the ‘financial constraints’ discount involved the case of SGL Carbon, a member of three contemporaneous cartels (Graphite Electrodes (dealt with by the Commission in 2001),21 Specialty Graphites (2002),22 and Carbon and Graphite (2003)).23 Having refused to grant a discount in relation to the first cartel, the Commission did so in the subsequent cases, taking into account the impact of the recent previous fines and going so far as to award a 33 per cent discount in its third decision. Moreover, this was despite the fact that company had acted as ringleader of the Specialty Graphites Cartel and did little to cooperate with the Commission’s investigations. This appears to have been a very successful playing of the bankruptcy card, to the detriment of both deterrent and retributive objectives. There has been some criticism of a loose application of the financial constraints discount, both under US and EC law. Taking again the example of SGL Carbon, it is significant that the company had already successfully negotiated a plea bargain with the DoJ, on the ground of its inability to pay 19 See in particular the discussion by Stephan A. (2006): The Bankruptcy Wildcard in Cartel Cases, CCP Working Paper 06-5, March 2006, University of East Anglia Centre for Competition Policy. 20 Joined Cases T-236/1 etc. Tokai Carbon and Others v. Commission [2004] ECR II-1181, at recital 371. 21 OJ L 100 [2002]. 22 Non-confidential version of the Commission’s decision of 17 December 2002 available at http://ec.europa.eu/comm/competition/antitrust/cases/decisions/37667/en.pdf. 23 OJ L 125 [2004].


650

12/10/07

13:15

Page 650


a high fine, resulting in a lower fine, agreement for the fine to be paid in instalments and a waiver of interest payments. Yet, as Stephan has pointed out,24 SGL Carbon’s American subsidiary subsequently filed for bankruptcy, but its petition was dismissed by the 3rd Circuit Court of Appeals. The Court stated that the firm was not in imminent danger of insolvency even though it was facing probable civil actions under the Sherman Act and there was no evidence that either criminal or civil actions had caused damage to the company’s business relations.25 As Stephan comments: “The real significance of these uncertain bankruptcy discounts in different jurisdictions is that international firms guilty of collusion, can take advantage of these and use them to play competition authorities against each other with the result that the fines incurred are further reduced [. . .] SGL was apparently successful in taking advantage of the discretion that exists in both jurisdictions to reduce fines on the grounds of financial difficulty with the result that it paid fines in the US that were below the mimimum required by the US guidelines and by receiving a 33% discount from the European Commission twice.”26

Stephan also argues that the Commission’s own application of the ‘financial constraints’ discount lacks consistency as between cases, at least to the extent of not providing sufficiently detailed and convincing reasoning such as to justify the different outcomes.27 This, then, is a rather different argument for greater transparency from that referred to above. But the whole issue of the ‘financial constraints’ discount illustrates well the increasingly complex process of balancing effective enforcement and fair treatment of companies in a context of international, multi-jurisdictional enforcement.

3. Legal protection and soft law principles Looking beyond specific questions of penalty calculation, both leniency and fining practice (at the present time) and plea-bargaining (perhaps in the nottoo-distant future), raise in themselves new procedural issues and questions of defence rights, as does the embodiment of such important procedures in soft-law instruments such as the Commission’s notices and guidelines. Although changes have been made recently to the main implementing regulation in the field of EC competition policy (when Regulation 17/62 was replaced by Regulation 1/2003), in the area of cartel control a great deal of significant practice is now based on soft law instruments, as noted above. These provisions have already been the subject of argument in litigation and 24 25 26 27

Stephan, supra note 19, p. 25 et seq. Re SGL Carbon Corporation, 200 F.3d 154 (3d Cir. 1999). Stephan, supra note 19, at p. 27. Ibid., p. 21 et seq.


12/10/07

13:15

Page 651

Effectiveness of Enforcement and Legal Protection 651 as such a number of principles implicit in or underlying these measures have been reaffirmed or become embedded in the jurisprudence of the CFI and ECJ.28 It is therefore increasingly relevant to enquire how such instruments— not legally binding in the strictest formal sense but certainly of legal consequence—fit into the corpus of legal protection rules. There is of course some good reason for the casting of these instruments in a soft law format, since to a large extent they provide a clarification of the Commission’s exercise of discretion in the application of penalties. But it may perhaps be asked at what point the clarification of discretionary practice may achieve a normative hardening, such as to convert a loose expectation into an entitlement. The ECJ has responded to this question in its judgment in 2005 on the appeal from the CFI’s judgment relating to the Pre-Insulated Pipes Cartel. The issue was triggered by the appellants’ argument that the 1998 Fining Guidelines amounted to a significant change in fining practice, which if applied retrospectively to the date of those companies’ infringement, would violate the principle against retroactive application of sanctions, as laid down in both Article 7 of the European Convention on Human Rights29 and in Community law.30 One aspect of the argument concerned the question whether the Guidelines could be categorised as a legally consequential measure such as to clearly fall within the scope of the principle against retroactive measures. While the CFI assumed that the Guidelines did not form part of the legal framework that determined the amount of fines, the ECJ was of the view that the Guidelines did produce legal effects. The Court argued: “In adopting such rules of conduct and announcing by publishing them that they will henceforth apply to the cases to which they relate, the institution in question imposes a limit on the exercise of its discretion and cannot depart from those rules

28 For instance, we might refer to the underlying substantive retributive principle of gravity, enunciated clearly but briefly in Article 23 of Regulation 1, and very much evident in the Fining Guidelines, under which the basic amount of a fine is increased to take into account aggravating circumstances and reduced to take into account attenuating circumstances. The same principle more specifically justifies the leniency device, since the provision of information, as a form of cooperation, qualifies as an attenuating circumstance. At the same time, instruments such as the Guidelines are clearly to be applied in accordance with procedural principles such as nondiscrimination and proportionality. Thus, the preamble to the Guidelines states that the substantive principles outlined in that instrument: “[. . .] should ensure the transparency and impartiality of the Commission’s decisions . . . while upholding the discretion which the Commission is granted. . . . This discretion must, however, follow a coherent and non-discriminatory policy which is consistent with the objectives pursued in penalising infringements [. . .]”. 29 Article 7(1) provides : “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.” It is the second part of the provision which is relevant here. 30 Both the CFI and ECJ confirmed in the Pre-Insulated Pipes Cartel appeals that the principle of non-retroactivity of criminal laws, as enshrined in Article 7 ECHR, was also a general principle of Community law which must be observed when fines are imposed for infringement of the competition rules (see para. 202 of the ECJ judgment).


652

12/10/07

13:15

Page 652


under pain of being found, where appropriate, to be in breach of the general principles of law, such as equal treatment or the protection of legitimate expectations. It cannot therefore be excluded that, on certain conditions and depending on their conduct, such rules of conduct, which are of general application, may produce legal effects . . . although the Guidelines do not constitute the legal basis of the contested decision, they determine, generally and abstractly, the method which the Commission has bound itself to use in assessing the fines imposed by the decision [. . .] 31

This part of the judgment therefore clarifies in a significant way the legal status, impact and reviewability of such ‘soft law’ measures as guidelines and notices, as ‘rules of conduct’: what is important is to consider whether such ‘rules’ lay down, in a general and abstract manner, a method of proceeding to which the Commission has bound itself. Or, it might be said more simply, in such a case the institution in question may be objectively considered to have bound itself to follow its own guidelines, in effect converting some of its ‘soft law’ into ‘hard law’. The important consequence then in the context of the present discussion, is that the matter is no longer one of discretionary decision-making, which is not subject to appeal, but one of application of legal rules, which could be open to challenge on appeal. The subsequent outcome in the Pre-Insulated Pipes appeal is discussed in the following section.

C. The temporal dimension of cartel proceedings The time-scale of an EC cartel case, from the opening of a formal investigation to the final hearing of all the appeals, can be and often is considerable: possibly a matter of several years. One consequence of this aspect of cartel proceedings is that, during that time, a case may be overtaken by changing legal and factual circumstances. The impact of both types of changing circumstances may be illustrated with reference to some argument raised in recent appeals.

1. Legal rules or legal practice change after the event It is clear enough that an alleged infringement should be legally assessed according to the law applicable at the time the infringement took place. As stated above, the CFI and ECJ have confirmed the application by analogy of the principle underlying Article 7 of the ECHR, relating to non-retroactivity 31

Judgment in Cases C-189/02 etc, supra note 9, at paras. 211 and 213.


12/10/07

13:15

Page 653

Effectiveness of Enforcement and Legal Protection 653 of criminal law and sanctions, to EC competition proceedings. The issue arising in the appeal from the Pre-Insulated Pipes Cartel32 concerned an alleged change in the principles for imposing penalties brought about by the Commission’s 1998 Fining Guidelines. The CFI considered that the method of calculating fines set out in the Guidelines did not alter the legal framework of the penalties, which continued to be defined solely by Regulation 17, and that the Commission’s decision-making practice formed part of its exercise of discretion and was not part of the legal framework. We have seen, however, that the ECJ stated that an instrument such as the Fining Guidelines might embody rules of conduct that produce legal effects and as such would then be subject to general principles of equal treatment and protection of legitimate expectations. But the Court went on to argue that what is more precisely prohibited from retroactive change is not a gradual clarification of rules of criminal liability, but a new interpretation. It would seem, therefore, that by this reasoning the former would not alter the legal framework, whereas the latter would do so. For the Court, then, the crucial test was whether: “[. . .] according to [the Commission’s case-law] a judicial interpretation produces a result which was not reasonably foreseeable at the time when the offence was committed, especially in the light of the interpretation put on the provision in the case-law at the material time.”33

In the case of the Guidelines the Court considered that any changes to the method of calculating fines—by taking into account certain factors of gravity other than turnover and so increasing the general level of the quantum of fines—was indeed foreseeable at the time when the members of the cartel committed the infringement. In short, it was predictable that the Commission might, for both retributive and deterrent reasons, decide to impose more severe penalties, within the range permitted by Regulation 17, and that this would be a valid exercise of their discretion. Such a prediction might depend upon legal advice, but that would not be unreasonable in the circumstances and would be consistent with the approach taken by the European Court of Human Rights.34 The ECJ stated: “A law may still satisfy the requirement of foreseeability even if the person concerned has to take appropriate legal advice to assess, to a degree that is reasonable in the circumstances, the consequences which a given action may entail. This is particularly true in relation to persons carrying on a professional activity, who are used to having to proceed with a high degree of caution when pursuing their occupation. They can on this account be expected to take special care in assessing the risks that such an activity entails.”35 32

Cases C-189/02 P etc., supra note 9. Ibid., at para. 218 of the judgment. 34 The Court referred to the case of Cantoni v France (17862/91, (1996) ECHR 52, judgment of 15 November 1996). See generally the discussion by Cassesse A. (2003): International Criminal Law, Oxford University Press, Oxford, at p. 147 et seq. 35 Cases C-189/02 P, supra note 9, at para. 219 of the judgment. 33


654

12/10/07

13:15

Page 654


From all of this it would seem that the test of retroactive application comprises an enquiry into whether there has been a change in the legal framework for imposing liability, either by formal revision of the law or a new interpretation of the law, and in relation to the latter, a test of foreseeability should be employed. It is in this way that companies subject to cartel proceedings should attempt to judge whether they have been wrong-footed by the Commission’s evolving policy and practice in relation to sanctions.

2. Issues of corporate identity and succession Another complicating issue is that, over time, corporate identity may change: companies may cease to exist, may merge or be taken over, so raising the question of how and to whom rules of liability should be imposed at the later date when the proceeding is completed.36 Some of the legal issues arising from this kind of situation are illustrated by the arguments raised in the appeals from the Commission’s decision on the Alloys Surcharge case (or Stainless Steel Cartel). During the 1990s and early years of this century the Commission and the ECJ and CFI dealt with two major cartels in the steel market, first in relation to steel beams,37 and then stainless steel.38 Two of the companies involved, Thyssen Steel and Krupp Hoechst merged to form Krupp Thyssen in 1995. Krupp Thyssen and the Italian company Acciai speciali Terni (AST) were members of the Stainless Steel Cartel between 1994 and 1998, but Krupp Thyssen acquired all the shares in AST in 1996, although the latter remained a separate company. One argument related to the imposition of separate fines on Krupp Thyssen and AST after the acquisition of AST by Krupp Thyssen. It was argued by the latter that there was, by the date of the imposition of fines, a single economic entity and therefore only one fine was justified. However, the ECJ confirmed the CFI’s judgment that two fines were justified since it was not denied that, during the time of the cartel’s operation, the two companies were distinct legal persons who behaved autonomously; moreover the two companies retained separate legal identities after Krupp Thyssen had acquired AST.39 To hold otherwise would clearly undermine the retributive and deterrent effect of the sanctions, enabling a deft evasion of a sanction through the sub36 For a brief comment on the legal complications arising from corporate restructuring in cartel cases, see Harding and Gibbs, supra note 7, at pp. 355–6. See also the discussion by Hoegh K. (2004): “Succession of Liability for Competition Law Infringements—The Cement Judgment”, 25 European Competition Law Review 534. 37 Commission decision 1994; CFI judgment, 11 March 1999; ECJ judgment, 2 October 2003. 38 Commission decision 1998; CFI judgment, 13 December 2001; ECJ judgment, 14 July 2005, in Case C-57/02 P Acerinox [2001] ECR II-3859, and Joined Cases C-65/02 P and 73/02 P Thyssen Krupp and Thyssen Krupp Acciai Speciali Terni v Commission [2001] ECR II-3757. 39 Joined Cases C-65 and 73/02 P, supra note 38, at para. 68 of the Court’s judgment.


12/10/07

13:15

Page 655

Effectiveness of Enforcement and Legal Protection 655 sequent happenstance of corporate restructuring. The ECJ distinguished the situation40 which had arisen in relation to Stora Kopparbergs, a member of the Cartonboard Cartel, and which was not held responsible for the cartel activities of subsidiary companies engaged in before they had been acquired by Stora.41 The guiding temporal principle42 is that the sanctions should relate to and reflect the corporate identities at the time when the infringements took place, and not be affected by subsequent restructuring. However, as an exception, the outcome would be different in the case of a subsequent merger, since then the identity of the original offending company would have been lost for purposes of imposing a sanction; thus, for the sanction to have any meaning, it must attach to the merged actor. The same principle underscores another ruling within the ECJ’s judgment in the Alloy Surcharges appeal—that procedural rights should attach strictly to the actions being judged and therefore the responsibility for such actions. Although Thyssen Krupp (the merged company) had declared itself responsible for Thyssen Steel’s earlier infringements,43 it had not been given the opportunity to be heard by the Commission in relation to those activities. The CFI and ECJ both affirmed that an acknowledgment of responsibility should not imply a waiver of the right to be heard. As the ECJ explained, in so far as such a statement admitting responsibility for the acts of what is technically a separate legal person: “[. . .] constitutes an exception to the principle that natural or legal persons may be penalised only for acts imputed to them individually, it must be interpreted strictly. The Court of First Instance correctly inferred therefrom that, unless he gives some indication to the contrary, the person making such a statement cannot be presumed to have waived the right to exercise his rights of defence.”44

Thus, two underlying, well-established and inter-related principles may be seen to inform such outcomes: first, the intertemporal principle that requires actions to be judged in the context of the time at which they took place; and secondly, the substantive principle of liability based upon individual or personal responsibility.45 But in practice application of these principles may in some cases become a less straightforward matter of tracing the legal survival of the original offending party. This may prove a complicated exercise, depending on the more precise nature of corporate restructuring and its legal 40

Ibid. at para. 69 of the Court’s judgment. Case T-352/94 Mo Och Domsjo v Commission [1998] ECR II-1989. This might be referred to as the ‘innocent purchaser’ principle. 42 Sometimes referred to as the ‘Anic Rule’: see Case C-49/82 P Commission v Anic Partecipazioni [1999] ECR 4125. 43 As the Commission argued : Thyssen Krupp was the beneficiary and economic successor of Thyssen Steel, and the conduct of the two undertakings was characterised by a clear unity of action in relation to the part of Thyssen’s business taken over by Thyssen Krupp. 44 Joined Cases C-65 and 73/02 P, supra note 38, at para. 82 of the Court’s judgment. 45 This is sometimes described rather vaguely as the ‘principle of guilt’. An important application of this principle is the argument against any form of vicarious or collective liability. 41


656

12/10/07

13:15

Page 656


forms. While it is clear on the one hand that responsibility follows a subsidiary company which maintains its own legal personality46 or is ‘lost’ to a merged company as a new legal actor, there may on the other hand be less clear-cut intermediate situations. For instance, in dealing with the Cement Cartel the Community Courts47 were confronted with a transfer of assets from an original offending cartel member (Aalborg Portland-Cement Fabrik A/S) to a new subsidiary company (Aalborg Portland A/S) between the date of offending and that when fines were imposed. Both the CFI and ECJ upheld the imposition of a fine on the new subsidiary,48 taking the view that the transfer of assets constituted a reorganisation within the one and the same legal entity. Hoegh has criticized this approach as a misunderstanding of the facts and concluded that: “[. . .] it may become much more difficult to restructure the company to ensure its ongoing existence after having received a substantial fine that—in conjunction with a claim for damages—threatens the continued existence of the undertaking. A potential new co-owner of the assets and activities of the company will thus—as is well-founded on the background of the Cement judgment—fear to become liable for the infringements of the past.”49

However, such an argument may be somewhat ingenuous. In a case such as that of Aalborg Portland, it would at least seem justifiable to attempt to penetrate the reasons for such corporate restructuring and probe possible strategies aimed at diluting the effect of sanctions. This issue is also linked, of course, to the discussion above relating to claims of financial constraint.

D. The interface of European and national levels of enforcement Finally, and more prospectively, we may reflect on some emergent issues of enforcement and legal protection at the interface of the European and Member State legal orders. There is now an increasing possibility of the same cartels, and their corporate and human participants, being dealt with under both EC and national law. Companies may be subject to legal proceedings (investigation, prosecution and enforcement-related measures such as leniency programmes) and corporate sanctions (usually financial penalties) 46 For instance, in relation to the Cement Cartel (infra note 47),Ciment francais SA had its fine reduced on appeal since there had been no structural links with the offending Belgian cartel member, later its subsidiary, at the time of the cartel. 47 Joined Cases T-25/95 etc. Cimenteries CBR v Commission [2000] ECR II-491 (CFI); Case 204/00 P Aalborg Portland v Commission [2004] ECR I-123. 48 But, in the ECJ, Advocate General Ruiz-Jarabo Colomer argued to the contrary, that on a strict application of the principle of personal and individual responsibility, the fine should not be imposed on the new subsidiary. 49 Hoegh, supra note 36, at p. 536.


12/10/07

13:15

Page 657

Effectiveness of Enforcement and Legal Protection 657 under both kinds of legal order. It has already, for example, been discussed above how companies might exploit this situation for purposes of securing a ‘financial constraints discount’. There is also the possibility of both criminal proceedings (especially in relation to individuals) and civil proceedings (claims for damages on the part of those injured by cartel activities) being used as a second rank of enforcement in the wake of condemnation by competition authorities. Such parallel enforcement raises two aspects of legal protection in particular: the characterisation of dual enforcement as a form of double jeopardy; and the need for some consistency of legal protection across national and European contexts of enforcement.

1. Double jeopardy In a sense, double jeopardy considerations feed into the financial constraints argument referred to above, not so much as a matter of fair treatment for those subject to the process, but rather as a factual issue: that successive prosecution and punishment renders the defendant company vulnerable to bankruptcy and unable to pay the later penalties. But as a matter of fair treatment, the argument might be put that being dealt with under more than one kind of legal process in relation to the same cartel infringement constitutes an infringement of the principle ne bis in idem.50 There are a number of possible responses to such an argument. First, it may be countered that the use of different procedures (a public law process of control and sanctioning, such as that employed by the Commission; criminal proceedings; civil claims) represent the resolution of distinct disputes involving different interests, and as such there is no recycling of process and repetition of sanctions. After all, national legal systems commonly allow both criminal prosecution and civil claims in relation to the same offending conduct. This is a persuasive response, but it would still be helpful, in the cumulative deployment of these processes across legal orders, to be clear about the way in which different interests are being protected by different types of procedure and to identify those interests. While there may not be an objection in principle to the cumulative resort to such processes, it may nonetheless be sensible for account to be taken within each of the use of the other procedures, for both effectiveness and fairness in enforcement. In practical terms, there may indeed be limits to the quantum of sanction which can be applied to a single party, and some attention should be paid to the 50 For earlier discussion and resolution of double jeopardy claims, see Case 14/68 Walt Wilhelm v. Bundeskartellamt [1969] ECR 1; Case 45/69 Boehringer Mannheim v. Commission [1970] ECR 769.


658

12/10/07

13:15

Page 658


possibly distorting impact of the simultaneous or successive application of sanctions across a number of legal orders. This has already proven to be an issue, for instance, in relation to the availability of evidence and the operation of leniency programmes. Secondly, in a related way, it might be argued that different offences or infringements are being addressed in these different proceedings, and similarly that different persons are targeted by such procedures and sanctions. For instance, it is possible to conceive of a ‘corporate’ offence of anticompetitive behaviour (such as price fixing in itself) and an ‘individual’ offence of dishonest planning or conspiracy in relation to such an outcome. This is the position in relation to the present UK regulation, through Section 36 of the Competition Act 1998 (which lays down a kind of administrative offence committed by companies) and Section 188 of the Enterprise Act 2002 (which lays down a criminal ‘cartel ‘offence’ committed by individuals). This response to double jeopardy objections is something of a ‘lawyer’s solution’, convincing more in terms of legal distinction than philosophical basis. For the problem remains that, on another view, the identity of the company and its individual employees or executives overlap—the latter, in their activity of planning and implementation may still be convincingly regarded as the alter ego of the corporate person. As noted already by Harding and Joshua: “such questions probe the justification for parallel ‘administrative and ‘criminal’ sanctions”.51

2. Consistency of legal protection This issue arises from the increasing complexity of global regulation, which employs legal process, and concomitant standards, under a number of different types of legal order: public international law; EC law; and a number of national jurisdictions. The EC Commission is a major enforcement agent for purposes of the regulation of cartels, and acts as such under rules of EC law. National competition authorities within the EU apply both EC and national rules in their role of enforcement. National courts, in jurisdictions within and beyond the EU, apply rules of national law relating to both criminal and civil liability. These enforcement activities are subject to scrutiny, in order to ensure their fair and reasonable application and the basic rights of those persons subject to enforcement. But the relevant legal protection rules themselves emanate from distinct legal orders: EC law (in relation to the Commission), national constitutional law, and international instruments such as the ECHR, in relation to national agencies and courts. In practical 51 Harding and Joshua, supra note 1, p. 264. The issue is further explored in Harding C., (2007): Individuals, Organisations and Criminal Responsibility, Willan Publishing, Cullompton.


12/10/07

13:15

Page 659

Effectiveness of Enforcement and Legal Protection 659 terms, therefore, in relation to the same cartel, while the Commission’s enforcement role is regulated by legal protection principles under EC law, that of a criminal court within an EU Member State is regulated by national rules and those of the ECHR (but also indirectly by EC law, via the principle of supremacy of EC law and the requirements of Article 10 EC). But, in so far as the same cartel and the same corporate and individual actors are being dealt with across these jurisdictions, consistency of treatment and of legal protection would be clearly desirable. There is little doubt in practice that courts in these different jurisdictions strive towards this goal of consistent treatment. The issue is rather that of how that objective may be achieved and ensuring that it is effectively gained. Just as there has evolved a European Competition Network for purposes of ensuring an effectively co-ordinated pan-European system of enforcement, a similar mechanism is necessary for purposes of ensuring a consistent application of standards of legal protection. In practice, and despite occasional claims to the contrary,52 there appears to evidence of real convergence, though it is often accomplished with little fanfare. A good example is provided by the ECJ’s adoption of the jurisprudence of the European Court of Human Rights in relation to the issue of retroactive application of new law and practice in the Pre-Insulated Pipes Cartel appeal, discussed above.53 The ECJ’s judgment firstly, at paragraph 215, cross-refers to a number of decisions and judgments of the Human Rights Court, and then in paragraph 219 embodies verbatim the test of foreseeability of a judicial interpretation, including the elements of necessary legal advice and the position of people carrying on a professional activity, which was laid down by the Human Rights Court in the case of Cantoni v. France.54 On the evidence of such examples, the issue is not so much that courts in different European jurisdictions may be singing from different hymn-sheets, but that the subjects of enforcement may be assured or convinced that there is a consistent approach to legal protection from one legal order to another.

E. Concluding comments: effectiveness and legal protection It would be trite to observe that the main task in this area is to balance effective enforcement of competition policy with a due respect for the rights of those subject to enforcement procedures. Yet it may be well to remind ourselves that a key aspect of such discussion—that of effectiveness—is not 52 See for instance, Ameye E. (2004): “The Interplay between Human Rights and Competition Law in the EU”, 25 European Competition Law Review 332. 53 Joined cases C-189/02 P, supra note 9. 54 See supra note 34.


660

12/10/07

13:15

Page 660


always as clearly analyzed as it might be. Some basic questions regarding the meaning of ‘effective enforcement’ are insufficiently addressed. The objectives of competition policy may be clear enough, but these do not necessarily supply the objectives of the enforcement of that policy. And to a large extent the determination of defence rights depends upon the requirements of enforcement rather than of the underlying policy, since it is the process of enforcement which actually impinges upon such rights. In particular, it may be helpful to identify three distinct goals of enforcement activity: the clearing up of cases; deterrence; and retributive justice. Clear-up is the most easily achieved, since it entails detection followed by a sanction; the case is then closed and in that sense finally dealt with. But that outcome need not achieve deterrence, and indeed the rate of recidivism among companies and cartelists suggests that deterrent efficacy is a very elusive goal in this context. Retributive efficacy may also prove problematical since it requires a restoration of the balance between wrongdoing gain and compliance loss, which in turn is a matter of assessing the impact of sanctions such as financial penalties, awards of compensation, or even prison terms. Thus, if the underlying principle is one of matching the restriction of basic freedoms and the needs of enforcement, a more precise understanding of the goals of enforcement should inform this exercise. For instance, the right to a transparent reading of the calculation of sanctions must be balanced against the need to have effective sanctions, but the latter is itself dependant upon an understanding of the different senses in which a sanction may be considered effective. If, then, a sanction is effective in clearing up a case, but not as a long-term deterrent (the point about direct settlement made above), that will have some bearing on the whether there has been effective enforcement, and in turn upon the question of the extent of legal entitlement for those subject to the enforcement process, since deterrence justifies less transparency than clear-up. But the interrelationship of and any ranking between different objectives of enforcement remains to be more fully worked out.

(Zg) Ehlermann 07 Forwood

12/10/07

13:15

Page 661

III Nicholas Forwood* Effective Enforcement and Legal Protection— Friends or Enemies?

It is perhaps useful to start with a number of remarks, most of which will have already been discussed at earlier editions of this Workshop: 1. The primary measure of the effectiveness of laws prohibiting cartels, like most other laws, should be the extent to which the terms of those laws, when taken with the likely manner of their enforcement in the event of their breach, combine to deter companies from entering into cartels. Post hoc punishment and/or compensation are therefore, as a matter of principle, second-best solutions. 2. What distinguishes unlawful cartels from many other types of unlawful or criminal conduct is that, unlike most crimes of violence or even crimes against property, most potential cartelists are likely to act with a higher than normal degree of rationality in deciding whether or not to enter into a cartel. 3. Punishment and compensation will also usually be sub-optimal “solutions” for correcting the consequences of a detected cartel. Fines for cartel behaviour produce no direct benefit to those (competitors, customers or final consumers) directly injured by cartels. Civil court compensation is also in many cases a very imperfect mechanism—potentially high cost, unsuitable for many types of cases (atomized losses, passing-on, etc.). 4. While fines or prison sentences, if high enough (i.e. many times the potential gains, discounted for the risk of non-detection), may be effective deterrents, there are real difficulties where the risk of detection is perceived as small (NB: Instinctively, it is probable that most cartelists believe that they will not be caught, and this may be one area where they will tend to act irrationally by underestimating the likelihood of detection). While an economist might justify a system in which fines were automatically adjusted for the perceived risk of detection, such an approach would present difficulties that are both theoretical (e.g. traditional ideas of proportionality) and practical (how can the court/enforcement authority realistically determine the perceived risk). 5. In seeking to optimize the deterrent effect of cartel laws and enforcement practice, it therefore appears sound in principle to focus on measures that * Judge at the Court of First Instance of the EC—Luxembourg.


662

12/10/07

13:15

Page 662


are likely to increase the objective likelihood, and a fortiori the perceived likelihood, of detection. 6. It is often said that one of the most common characteristics of cartels is that cartelists “cheat” on each other. (Hence, the most effective cartels are usually those with highly developed monitoring and reporting mechanisms—even though these mechanisms in turn often lead to an increased risk of detection). It is therefore inherently probable that cartelists will also be likely to “cheat” on one another by acting as whistleblower, at least if the (perceived) advantages of doing so are high. The situation is thus a variant of the well-known “prisoners’ dilemma”.1 7. The objective likelihood of any one cartelist acting as whistleblower will be significantly increased by two inter-related factors: first, the extent of the benefits potentially on offer from the enforcement authorities, and second, the a priori degree of certainty he feels (normally after taking skilled professional advice) that those benefits will in fact be accorded to him if he does decide to “blow the whistle”. This latter factor is particularly important since it will normally be difficult, if not impossible, to get a firm indication of the benefits (leniency/immunity) that will be granted in a given case without first “letting the cat out of the bag”. These considerations are, I believe, helpful in trying to find the right answers to the questions we have been asked to address in this last session of the Workshop.

A. Decision-making: by an administrative agency or by the courts? I must confess at the outset to being someone who, for the last twenty-five years, has practised exclusively as barrister, then as judge, in an environment in which fines for cartel infringements are set by an administrative agency, but are subject to review by a court with “unlimited jurisdiction”. I apologize in advance if, therefore, I under- or over-estimate the respective disadvantages of other systems. In my view, the strong preponderance of arguments still points towards the relative merits of the present system as applied in the EU, where determinations of culpability and the amount of fines are set by the competent administrative agency, but where that decision is subject to a full review by a court covering the facts, law, economic appreciation (the latter being subject to a 1 For those not familiar with this basic concept of game theory, applied to business, an elegant description is given by Avinash Dixit and Barry Nalebuff in The Concise Encyclopedia of Economics, http://www.econlib.org/LIBRARY/Enc/PrisonersDilemma.html.


12/10/07

13:15

Page 663

Effective Enforcement and Legal Protection—Friends or Enemies? 663 manifest error control) and the appropriateness and amount of fine. It goes without saying that the administrative decision should be taken at the culmination of an administrative procedure that ensures effective exercise of the rights of defence. These benefits are, in my view, particularly clear when it comes to questions both of fining policy, and of the formulation and application of any leniency programme. First, even though the Court of First Instance is accorded by the Treaties and the relevant regulations “unlimited jurisdiction” with regard to fines, I am among those who consider that this provision is not in itself a sufficient or proper basis to justify the Court setting its own policy with regard to fines, or as to a leniency programme, which might run counter to any such policy set by the Commission. There are several reasons, both of principle and of practicality. First, as a matter of principle, the Treaties entrust the power to formulate and apply competition policy to the Commission and not to the Court. The formulation of a general policy with regard to the level of fines, of the type set out in the Guidelines, which of necessity is announced in advance of any future specific implementation of that policy in a given case, does not fall within the formal remit of the Court. Even if the Court has unlimited jurisdiction to control the level of fines imposed in a given case, that does not, in my view, justify it making ex cathedra statements in its judgments as to the appropriate level of fines in cases other than those immediately before it. In practice, moreover, it would be extremely difficult, if not wholly impracticable, for the Court itself to formulate such a policy. First, there is the difficulty of gathering the necessary information on the basis of which any such policy should properly be based. While DG Comp has day-to-day contact with cartels, and has a much fuller picture of the reality of conditions on the market place, a picture that is moreover reinforced by its regular contacts with NCAs, the Court of First Instance sees only the tip of the iceberg, in the form of the cases that come by way of challenge to a fining decision. Then there is the issue of whether the Court itself could agree on such a policy. Again, there is the very real problem of whether, by formulating and promulgating such a policy, in advance of a given case, the Court would be liable to be criticized for having “pre-judged” the level of fine, or at least the starting point for its calculation, in a way that arguably compromises its status as an independent Tribunal. This list is by no means exhaustive. Similar considerations apply to the formulation of any leniency policy. The tailoring of such a policy is clearly a matter for which the experience of leniency programmes, and the practical difficulties that may arise in their implementation, is crucial. No less important is the extremely problematic situation for those who, as in my own country, are extremely averse to judicial involvement in “plea bargaining” situations. That is not to say, however, that the Court will not exercise any control over the Commission’s formulation of a fining or leniency policy, as distinct from the application of that policy in a given case. Thus, it would be open to


664

12/10/07

13:15

Page 664


a company, in the context of a challenge to the amount of a fine set in accordance with the current guidelines, to contend that the guidelines were themselves inappropriate as being in breach of certain principles of Community law, such as proportionality or non-discrimination. Such an argument, if successful, would not of course lead to the annulment of the guidelines as such, but would necessitate their de facto revision and application in the instant and any future cases. The nature of any such control would probably be that appropriate for other areas of policy setting, namely that of manifest error. However, I would like to add a word of caution. To be most effective, guidelines on fines need to be treated as precisely that: guidelines and not invariable rules. That requires, I believe, a balanced and considered approach by the Community Courts to the application of principles such as legitimate expectation and legal certainty. There is a danger, which may arguably be illustrated in some of the judgments of the CFI and the Court of Justice, that an over-rigid or formalistic application of the guidelines will be counterproductive. In this connection, it is not without relevance that few cartel members would, I believe, contend that they only entered into the cartel after a finely balanced judgment as to the level of fines that they might receive should the cartel be discovered. Nor, if such contentions were advanced, do I think that they would be received with a great deal of sympathy, or credibility.

B. Leniency programmes—soft or hard law? By contrast, I do consider that there is a strong case for a “hard law” approach to leniency programmes, if by this it is suggested that the Court should exercise a close control of whether the conditions of a leniency programme have been correctly applied to a cartel member who has come forward to denounce a cartel, or who has contributed materially in assisting the Commission to establish its case. Here, it is highly probable that the decision to come forward will have been made after a very careful consideration, probably with a lawyer, of the consequences of approaching the Commission. In this situation, the greater the degree of certainty that the benefits identified in the leniency notice will in fact be accorded to the company, the greater the likelihood of the denunciation being made. And, to repeat the point I made at the outset, the greater the likelihood that cartels will be denounced, the less the likelihood that companies will enter into them in the first place.

(Zh) Ehlermann 07 Biblio

16/10/07

14:00

Page 665

BIBLIOGRAPHY

A. Publications: Abrantes-Metz R., Froeb L., Geweke J. and Taylor C. (2005): A Variance Screen for Collusion, FTC Bureau of Economics Working Paper No. 275. Abreu D. (1986): “Extremal Equilibria of Oligopolistic Supergames”, 39 Journal of Economic Theory 191. —— Pearce D. and Stacchetti E. (1986): “Optimal Cartel Equilibria with Imperfect Monitoring”, 39 Journal of Economic Theory 251. Abril P. and Morales Olazabal A. (2006): “The Locus of Corporate Scienter”, 1 Columbia Business Law Review 81. Adenaes J. (1971): “The Moral or Educative Influence of Criminal Law”, 27 Journal of Social Issues 17. —— (1975): “General Prevention Revisited: Research and Policy Implications”, 66 Journal of Criminal Law & Criminology 338. Arbault F. and Peiro F. (2002): “The Commission’s New Notice on Immunity and Reduction of Fines in Cartel Cases: Building on Success”, 2 Competition Policy Newsletter 15. Areeda P. (1986): Antitrust Law, Little Brown, Boston. Armstrong M. (1998): “Network Interconnection in Telecommunications”, 108 The Economic Journal 545. Arp J. and Swaak C. (2003): “A Tempting Offer: Immunity From Fines for Cartel Conduct Under the European Commission’s New Leniency Notice”, 23 European Competition Law Review 9. Assaf D. and Lefebvre E. (2003): “The Investigatory Tools under Sections 11 and 15 of the Competition Act: Selected Issues of Law and Practice”, paper presented at the Canadian Bar Association Competition Section, Annual Fall Conference, October 2, 2003. Athey S. and Bagwell K. (2001): “Optimal Collusion with Private Information”, 32 RAND Journal of Economics 428. —— —— (2004): Collusion with Persistent Cost Shocks, Stanford and Columbia Universities, mimeo. —— —— and Sanchirico C. (2004): “Collusion and Price Rigidity”, 71 Review of Economic Studies 317. Aubert C., Kovacic W. and Rey P. (2004): “The Impact of Leniency and Whistleblowing Programs on Cartels”, 21 International Journal of Industrial Organization 347. Bagwell K. and Staiger R. (1997): “Collusion over the Business Cycle”, 28 RAND Journal of Economics 82. Bain J. (1948): “Output Quotas in Imperfect Cartels”, 62 Quarterly Journal of Economics 617.


666

16/10/07

14:00

Page 666

Bibliography

Bain J. (1956): Barriers to New Competition, Harvard University Press, Cambridge, Massachussets. Bajari P. and Summers G. (2002): “Detecting Collusion in Procurement Auctions”, 70 Antitrust Law Journal 143. Baker D. (2001): “The Use of Criminal Law Remedies to Deter and Punish Cartels and Bid-Rigging”, 69 George Washington Law Review 693. —— (2004): “Revisiting History—What Have We Learned About Private Antitrust Enforcement That We Would Recommend To Others?”, 16 Loyola Consumer Law Review 379. —— and Reeves B. (1977): “The Paper Label Sentences: Critique”, 86 The Yale Law Journal 619. Barnett T. O. (2004): “Antitrust Enforcement Priorities: A Year in Review”, address before the Fall Forum of the ABA Section of Antitrust Law 2–3, November 19, 2004, available at http://www.usdoj.gov/atr/public/speeches/206455.pdf. Bauer J. (2004): “Reflections on the Manifold Means of Enforcing the Antitrust Laws: Too Much, Too Little, or Just Right?”, 16 Loyola Consumer Law Review 303. Becker G. (1968): “Crime and Punishment: An Economic Approach”, 76 Journal of Political Economy 169. —— and Stigler G. (1975): “Law Enforcement, Malfeasance, and Compensation of Enforcers”, 3 Journal of Legal Studies 1. Benoît J.–P. and Krishna V. (1987): “Dynamic Duopoly: Prices and Quantities”, 54 Review of Economic Studies 23. Berghahn V. (1986): The Americanization of West German Industry 1945–1973, Berg, New York. Bériault Y. (2004): “Negotiating a Plea of a Corporate Defendant in Canada”, paper presented at the ABA International Cartel Workshop, New York, February 2004. Bernheim B. and Whinston M. (1990): “Is Everything Neutral?”, 96 Journal of Political Economy 308. Bernheim D. and Whinston M. (1990): “Multimarket Contact and Collusive Behaviour”, 21 RAND Journal of Economics 1. Blair R. (1985): “A Suggestion for Improved Antitrust Enforcement”, Antitrust Bulletin 433. Blake S. and Schnichels D. (2004): “Leniency Following Modernisation: Safeguarding Europe’s Leniency Programmes”, 2 Competition Policy Newsletter 7. Bloom M. (2005): “Benefits and Challenges of Leniency Programmes in the Context of EC Modernization”, in Rafffaelli E., ed., Antitrust between EC Law and National Law, Bruylant, Treviso, 183–. —— (2005): “Should Foreign Purchasers Have Access to US Antitrust Damages Remedies? A Post-Empagran Perspective from Europe”, 61 New York University Annual Survey of American Law 433. —— (2005): “The Great Reformer: Mario Monti’s Legacy in Article 81 and Cartel Policy”, 1 Competition Policy International 55. —— (2005): “What Is the Future for International Antitrust?”, paper presented at the ABA Antitrust Section Fall Forum, November 15–16, 2005. Blume A. and Heidhues P. (2003): Modeling Tacit Collusion in Auctions, University of Pittsburgh and WZB, Berlin.


16/10/07

14:00

Page 667

Bibliography

667

Böge U. and Ost K. (2006): “Up and Running, Or Is It? Private Enforcement—The Situation in Germany and Policy Perspectives”, 27 European Competition Law Review 197. Bolotova Y., Connor J. and Miller D. (2005): The Impact of Collusion on Price Behavior: Empirical Results from Two Recent Cases, Department of Agricultural Economics, Purdue University. Bork R. (1978): The Antitrust Paradox, Basic Books, New York. Breit W. and Elzinga K. (1986): Antitrust Penalty Reform: An Economic Analysis, American Enterprise Institute for Public Policy Research, Washington DC. Brickman L. (2005): “An Analysis of the Financial Impact of S. 852: The Fairness in Asbestos Injury Resolution Act of 2005”, 27 Cardozo Law Review 991. Brock W. and Scheinkman J. (1985): “Price Setting Supergames with Capacity Constraints”, 52 Review of Economic Studies 371. Brown S. and Warner J. (1980): “Measuring Security Price Performance”, 8 Journal of Financial Economics 205. —— —— (1985): “Using Daily Stock Returns: The Case of Event Studies”, 14 Journal of Financial Economics 31. Bryant P. G. and Eckhard E. W. (1991): “Price Fixing: The Probability of Getting Caught” Review of Economics and Statistics 531. Buccirossi P. and Spagnolo G. (2005): “Optimal Fines in the Era of Whistleblowers. Should Price Fixers Still Go To Prison?”, LEAR Research Paper 05-01. Bush D., Connor J., Flynn J., Ghosh S., Grimes W., Harrington J., Hawker N., Lande R, Shepherd W. and Weinstein N. (1989): “Optimistic Biases About Personal Risks”, 246 Science 1232. Cahill D. and Cooke J., eds (2004): The Modernisation of EU Competition Law Enforcement—FIDE 2004 National Reports, Cambridge University Press, Cambridge. Calkins S. (1986): “Summary Judgment, Motions to Dismiss, and Other Examples of Equilibrating Tendencies in the Antitrust System”, 74 Georgia Law Journal 1065. —— (1988): “Equilibrating Tendencies in the Antitrust System, With Special Attention to Summary Judgment and Motions to Dismiss”, in White L., ed., Private Antitrust Litigation, MIT Press, Massachusetts, 185–. —— (1997): “Corporate Compliance and the Antitrust Agencies’ Bi-Modal Penalties”, 60 Law & Contemporary Problems 127. —— (1998): “In Praise of Antitrust Litigation: The Second Annual Bernstein Lecture”, 72 St. John’s University Law Review 1. —— (2000): “California Dental Association: Not the Quick Look But Not the Full Monty”, 67 Antitrust Law Journal 495. —— (2003): “Perspectives on State and Federal Antitrust Enforcement”, 53 Duke Law Journal 673. —— (2006): “Government Civil Remedies”, forthcoming in University of San Francisco Law Review. Calvani T. (2004): “Competition Penalties and Damages in a Cartel Context: Criminalization and the Case for Custodial Sentences”, paper presented at the ACLE workshop on Remedies and Sanctions in Competition Policy, Amsterdam, February 2005, available from http://www.acle.nl. —— (2004): “Enforcement of Cartel Law in Ireland”, in Hawk B., ed., 2003 Annual Proceedings of the Fordham Corporate Law Institute International Antitrust Law & Policy, Juris Publishing, New York, 1-.


668

16/10/07

14:00

Page 668

Bibliography

Camilli E. L. (2005): “Optimal and Actual Fines in Cartel Cases: The European Challenge”, paper prepared for the conference “Remedies and sanctions in competition policy”, Amsterdam University, available at http://www.cms.uva.nl/ acle/object.cfm/objectid=F07DE744-C1D1-4F2E-876EEB31F7FA5B9F. Campbell J., Andrew W. and MacKinlay C. (1997): The Econometrics of Financial Markets, Princeton University Press, Princeton. Carle J., Lindeborg S. and Segenmark P. (2002): “The New Leniency Notice”, 23 European Competition Law Review 265. Cassese A. (2003): International Criminal Law, Oxford University Press, Oxford. Cavanagh E. (1987): “Detrebling Antitrust Damages: An Idea Whose Time Has Come?”, 61 Tulane Law Review 777. Chandler H. (2004): “Criminal Investigations: Process and Procedure”, paper presented at the Canadian Bar Association Competition Law Annual Conference, 24 September 1998. Chang M. (1991): “The Effects of Product Differentiation on Collusive Pricing”, 9 International Journal of Industrial Organization 453. Chemtob S. M. (2000): “Antitrust Deterrence in the United States and Japan”, remarks at a Conference on Competition Policy in the Global Trading System, Washington DC, 23 June 2000, available at http://www.usdoj.gov/atr/public/ speeches/5076.pdf. Chen J. and Harrington J. (2007), “The Impact of the Corporate Leniency Program on Cartel Formation and the Cartel Price Path”, in Ghosal V. and Sennek J., eds., Political Economy of Antitrust, Elsevier, Oxford. Chen Z. and Rey P., “Optimal Design of Leniency Mechanisms”, mimeo, University of Toulouse. Christie W. and Schultz P. (1999): “The Initiation and Withdrawal of Odd-Eighth Quotes Among Nasdaq Stocks: An Empirical Analysis”, 71 Journal of Financial Economics 317. Clarke R. (1983): “Collusion and The Incentives for Risk–Sharing”, 14 Bell Journal of Economics 383. Clifford J. and Roode J. (1999): “Practical Issues in Responding to Bureau Investigations and Ensuring Competition Law Compliance”, paper presented at the “Insight Competition Law Practices for Canadian Companies” conference, December 1999. Clifford J. and Rowley B. (1995): “Tackling Cross-Border Conspiracy,” in International Corporate Law, May 1995. Coffee J. (1981): “‘No Soul to Damn: No Body to Kick’: An Unscandalized Inquiry into the Problem of Corporate Punishment”, 79 Michigan Law Review 387. —— (1992): “Paradigms Lost: The Blurring of the Criminal and Civil Law Models— And What Can Be Done About It”, 101 The Yale Law Journal 1875. —— (2006): Reforming the Securities Class Action: An Essay on Deterrence and Its Implementation, Columbia Law and Economics Working Paper No. 293. Compte O. (1998): “Communication in Repeated Games with Imperfect Private Monitoring”, 66 Econometrica 597. —— (2000): Secret Price Cutting and Strategic Buyers: An Illustration of the Divide and Conquer Tactic, CERAS, Ecole Nationale des Ponts et Chausees, Paris, mimeo. —— and Jehiel P. (2002): Multi-Party Negotiations, CERAS, Ecole des Ponts et Chausées, Paris, mimeo.


16/10/07

14:00

Page 669

Bibliography

669

—— Jenny F. and Rey P. (2002): “Capacity Constraints, Mergers and Collusion”, 46 European Economic Review 1. Connor J. (1998): What Can We Learn from the ADM Global Price Conspiracies? Purdue University, Department of Agricultural Economics, Staff Paper #98-14. —— (2001): “Global Price Fixing: Our Customers Are Our Enemies—The Lysine Cartel of 1992–1995”, 18 Review of Industrial Organization 5. —— (2004): “Global Antitrust Prosecutions of Modern International Cartels”, 4 Journal of Industry, Competition and Trade 239. —— (2005): Price-Fixing Overcharges: Legal and Economic Evidence, American Antitrust Institute (AAI) Working Paper No. 04-05, available at http://www.antitrustinstitute.org/recent2/355.pdf. Cooper T. (1986): “Most Favoured Customer Pricing and Tacit Collusion”, 17 RAND Journal of Economics 377. Cooter R. (1984): “Prices and Sanctions”, 84 Columbia Law Review 1523. Cramton P. and Palfrey T. (1990): “Cartel Enforcement with Uncertainty About Costs”, 31 International Economic Review 17. Craycraft C., Craycraft J. and Gallo J. (1997): “Antitrust Sanctions and a Firm’s Ability to Pay”, 12 Review of Industrial Organization 171. Cseres K., Schinkel M. and Vogelaar F., eds. (2006): Remedies and Sanctions in Competition Policy: Economic and Legal Implications of the Tendency to Criminalize Antitrust Enforcement in the EU Member States, Edward Elgar, London. Dau-Schmidt K. (1990): “An Economic Analysis of the Criminal Law as a PreferenceShaping Policy”, 1 Duke Law Journal 14. —— (1993): “Sentencing Antitrust Offenders: Reconciling Economic Theory with Legal Theory”, 9 William Mitchell Law Review 75. —— (1998): “Preference Shaping by the Law”, in Newman P., ed., The New Palgrave Dictionary of Economics and the Law, Macmillan, London. Davidow J. (2004): “Recent US Antitrust Developments of International Relevance”, 27 World Competition 407. Davidson C. and Deneckere R. (1984): “Horizontal Mergers and Collusive Behavior”, 2 International Journal of Industrial Organization 117. —— —— (1990): “Excess Capacity and Collusion”, 31 International Economic Review 521. Dekeyser K. (2006): “Almost 2 Years of Cooperation Within the ECN”, paper presented at he IBC Conference on Advanced EC Competition Law, London, 5 May 2006. —— and Gauer C. (2004): “The New Enforcement System for Articles 81 and 82 EC and the Rights of Defence”, in Hawk B., ed. (2005): 2004 Annual Proceedings of the Fordham Corporate Law Institute—International Antitrust Law and Policy, Juris Publishing, New York, 537–. DeMott D. (1997): “Organizational Incentives to Care About the Law”, 60 Law and Contemporary Problems 39. Diemer C. (2006): “The Green Paper on Damages Actions for Breach of the EC Antitrust Rules”, 27 European Competition Law Review 309. Duso T., Gugler K. and Yurtoglu B. (2006): “How Effective Is European Merger Control?”, mimeo. Duso T., Neven D. and Röller L.-H. (2006): “The Political Economy of European Merger Control: Evidence Using Stock Market Data”, WZB Working Paper no. FS IV 02—34r, available at http://skylla.wz-berlin.de/pdf/2002/iv02-34r.pdf.


670

16/10/07

14:00

Page 670

Bibliography

Easterbrook F. (1985): “Detrebling Antitrust Damages”, 28 Journal of Law and Economics 445. Edwards C. (1955): “Conglomerate Bigness as a Source of Power”, Business Concentration and Price Policy 331, NBER Report, Princeton, NJ. Ehlermann C.-D. (1994): “Developments in Community Competition Law Procedures”, EC Competition Policy Newsletter 2. —— and Atanasiu I., eds. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland, Oregon. Eichenwald K. (2000): The Informant, Broadway Books, New York. —— (2005): Conspiracy of Fools, Broadway Books, New York. Ellison G. (1994): “Theories of Cartel Stability and the Joint Executive Committee”, 25 RANd Journal of Economics 37. Eswaran M. (1997): “Cartel Unity over the Business Cycle”, 30 Canadian Journal of Economics 644. Evans W. and Kessides N. (1994) “Living by the ‘Golden Rule’: Multimarket Contact in the U.S. Airline Industry”, 109 Quarterly Journal of Economics 341. Evenett S., Levenstein M. and Suslow V. (2001): “International Cartel Enforcement: Lessons from the 1990s”, 24 World Economy 1221. Faull J. and Nikpay A. (1999): The EC Law of Competition, Oxford University Press, Oxford. First H. (2001): “The Vitamins Case: Cartel Prosecutions and the Coming of International Competition Law”, 68 Antitrust Law Journal 711. 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 173. —— (2004): “Modernisation: An Extension of the Powers of the Commission?”, in Geradin D., ed., Modernisation and Enlargement: Two Major Challenges for EC Competition Law, Intersentia, Brussels, 83–. Forrester I., MacLennan K. and Komninos A. (2005): “EC Competition Law 2003–2004”, in 2005 Oxford Yearbook of European Law, 557–. Franzosi M. (1988): “Oligopoly and the Prisoner’s Dilemma: Concerted Practices and ‘As If’ Behaviour”, 9 European Competition Law Review 385. Frezal S. (2006): “On Optimal Cartel Deterrence Policies”, 24 International Journal of Industrial Organization 1231. Friedman J. (1971): “A Noncooperative Equilibrium for Supergames”, 28 Review of Economic Studies 1. Gallo J., Dau-Schmidt K., Craycraft J. and Parker C. (1994): “Criminal Penalties Under the Sherman Act: A Study of Law and Economics”, 16 Research in Law and Economics 25. —— —— —— —— (2000): “Department of Justice Antitrust Enforcement, 1955–1997: An Empirical Study”, 17 Review of Industrial Organization 75. Gauer C. (2005): “Due Process in the Face of Divergent National Procedures and Sanctions”, paper for the “Antitrust Reform in Europe” Conference, Brussels, 9–11 March 2005. Gavil A., Kovacic W. and Baker J. (2002): Antitrust Law in Perspective: Cases, Concepts and Problems in Competition Policy, Thomson West, American Casebook Series, New York.


16/10/07

14:00

Page 671

Bibliography

671

Genesove D. and Mullin W. (2001): “Rules, Communication and Collusion: Narrative Evidence from the Sugar Institute Case”, 91 American Economic Review 379. Geradin D. and Henry D. (2005): The EC Fining Policy for Violations of Competition Law: An Empirical Review of the Commission Decisional Practice and the Community Courts’ Judgements, Global Competition Law Centre Working Paper 03/05. Goldman C, Katz M. and Facey B. (2001): “Cross-Border Cooperation and Information Sharing in International Antitrust Cases: The Need for Balance,” paper presented at the Southwest Legal Foundation Annual Symposium on Private Investments Abroad, Dallas, Texas, June 19–20, 2001. Goldman C. and Kissack M. (1998): “US/Canada Antitrust Cooperation and CrossBorder Corporate Searches”, paper presented at the ABA 1998 Annual Meeting, Toronto, Canada, August 3, 1998. Goldman C., Witterick C. and Kissack M. (1997): “Cross-Border Computer Searches”, paper presented at the Grocery Manufacturers of America 1997 Annual Legal Conference, Washington, DC, October 7, 1997. Goldstein H. (2005): “Corporate Crime: Government Prosecutors Are Going to Extremes”, 234 New York Law Journal 5. Green E. and Porter R. (1984): “Noncooperative Collusion under Imperfect Price Information”, 52 Econometrica 87. Green N. (2004): “The Road to Conviction—The Criminalisation of Cartel Law”, in Hawk B., ed., 2003 Annual Proceedings of the Fordham Corporate Law Institute, International Antitrust Law & Policy, Juris Publishing, New York, 13–. Grossman P., ed. (2004): How Cartels Endure and How They Fail: Studies of Industry Collusion, Edward Elgar, Cheltenham. Grout P. and Sonderegger P. (2005): Predicting Cartels, Office of Fair Trading, Economic discussion paper of March 2005, text available at http://www.oft.gov.uk. Guersent O. (2002): “Cartels and Leniency”, speech delivered at the IBC Annual Conference, Brussels, 21–22 November 2002. —— (2004): “The Fight Against Secret Horizontal Agreement in the EC Competition Policy”, in Hawk B., ed., International Antitrust Law and Policy, Fordham Corporate Law 2003, Juris Publishing, New York, 43–. Hackner J. (1994): “Collusive Pricing in Markets for Vertically Differentiated Products,” 12 International Journal of Industrial Organization 255. Haltiwanger J. and Harrington J. (1991): “The Impact of Cyclical Demand Movements on Collusive Behaviour”, 22 RAND Journal of Economics 89. Hammond S. (2006): “Measuring the Value of Second-in Cooperation in Corporate Plea Negotiations”, address before the 54th Annual Spring Meeting of the American Bar Association, March 29, 2006, available at http://www.usdoj.gov/ atr/public/speeches/215514.pdf. —— (2000): “Detecting and Deterring Cartel Activity through an Effective Leniency Program”, paper presented at the Workshop on International Cartels, Brighton, 21–22 November 2000. —— (2001): “When Calculating the Costs and Benefits of Applying for Corporate Amnesty, How Do You Put a Price Tag on an Individual’s Freedom?”, presentation at National Institute on White Collar Crime, Mach. 8, 2001, available at http://www.usdoj.gov/atr/public/speeches/7647.htm.


672

16/10/07

14:00

Page 672

Bibliography

Hammond S. (2002): “A Review of Recent Cases and Developments in the Antitrust Division’s Criminal Enforcement Program”, The Conference Board, 2002 Antitrust Conference, New York, 7 March 2002. —— (2004): “Cornerstones of an Effective Leniency Program”, address before the ICN Workshop on Leniency Programs, Sydney, November 23, 2004, available at http://www.usdoj.gov/atr/public/speeches/206611.pdf. —— (2005): “An Update of the Antitrust Division’s Criminal Enforcement Program”, Antitrust Division, US Department of Justice, Washington DC, November 16, 2005, available from http://www.usdoj.gov/atr/public/ speeches/213247.htm. —— (2005): “An Overview Of Recent Developments In The Antitrust Division’s Criminal Enforcement Program”, address before the American Bar Association Midwinter Leadership Meeting, Kona, Hawaii, 10 January 2005. —— (2006): “Charting New Waters in International Cartel Prosecutions”, remarks presented to the Twentieth Annual National Institute on White Collar Crime, March 2, 2006, available at http://www.usdoj.gov/atr/public/speeches/214861.htm. —— (2006): “Measuring the Value of Second-in Cooperation in Corporate Plea Negotiations”, Antitrust Division, US Department of Justice, Washington, DC, March 29, 2006, available from http://www.usdoj.gov/atr/public/speeches/ 215514.htm. Harding C. (2002): “Business Cartels as a Criminal Activity: Reconciling North American and European Models of Regulation”, 4 Maastricht Journal of European and Comparative Law 393. —— and Gibbs A. (2005): ‘Why Go to Court in Europe? An Analysis of Cartel Appeals 1995–2004”, 30 European Law Review 349. —— and Joshua J. (2003): Regulating Cartels in Europe: A Study of Legal Control of Corporate Delinquency, Oxford University Press, Oxford. Harrington J. (1989): “Collusion Among Asymmetric Firms: The Case of Different Discount Factors”, 7 International Journal of Industrial Organization 289. —— (2005): “Optimal Corporate Leniency Programs,” Economics Working Paper Archive 527, Johns Hopkins University, Baltimore. —— (2006): “Corporate Leniency Programs and the Role of the Antitrust Authority in Detecting Collusion”, Competition Policy Research Center Discussion Paper, CPDP-18-E Johns Hopkins University, Baltimore. —— (2006): “Detecting Cartels”, in Buccirossi P., ed., Handbook in Antitrust Economics, MIT Press, Massachusetts. —— (2006): How Do Cartels Operate?, Foundation and Trends in Microeconomics Series, Johns Hopkins University, Baltimore. —— and Chen J. (2006): “Cartel Pricing Dynamics with Cost Variability and Endogenous Buyer Detection”, 21 International Journal of Industrial Organization 347. —— and Skrzypacz A. (2006): Collusion with Monitoring of Sales, Economics Working Paper Archive 509, The John Hopkins University, Baltimore. Hasnas J. (2005): “Ethics and the Problem of White Collar Crime”, 54 American University Law Review 579. Hay G. and Kelley D. (1974): “An Empirical Survey of Price Fixing Conspiracies”, 17 Journal of Law and Economics 13.


16/10/07

14:00

Page 673

Bibliography

673

Henry D. (2005): “Leniency Programmes: An Anaemic Carrot for Cartels in France, Germany and the UK?”, 26 European Competition Law Review 13. Hoegh K. (2004): “Succession of Liability for Competition Law Infringements—The Cement Judgment”, 25 European Competition Law Review 534. Holmes M. and Davey L., eds. (2004): A Practical Guide to National Competition Rules Across Europe, Kluwer Law International, London. Jephcott M. (2002): “The European Commission’s New Leniency Notice—Whistling the Right Tune?”, 23 European Competition Law Review 378. Jones C. (2004): “Exporting Antitrust Courtrooms to the World: Private Enforcement in a Global Market”, 16 Loyola Consumer Law Review 409. —— (2004): “Private Antitrust Enforcement in Europe: A Policy Analysis and Reality Check”, 24 World Competition 13. Joshua J. (1983): “The Element of Surprise”, 8 European Law Review 3. —— (2000): “Leniency in U.S. and EU Cartel Cases”, 14 Antitrust 19. —— (2004): “Brave New World: A North Atlantic Triple Entente Against Cartels”, 26 Competition Law Insight 36. —— (2004): “EC Fining Policy Against Cartels After the Lysine Rulings: The Subtle Secrets of X”, Global Competition Review 5. —— (2004): “Oral Statements in EC Competition Proceedings: A Due Process ShortCut?”, 26 Competition Law Insight 1. —— (2005): The European Cartel Enforcement Regime Post-Modernisation: Is It Working?, Lawyers’ Europe 24. —— and Camesasca P. (2004): “EC Fining Policy Against Cartels After the Lysine Rulings: The Subtle Secrets of X”, European Antitrust Review 5. —— —— (2005): “Cartels and Leniency: Hold or Fold? The Commission’s New Leniency Policy”, GCR European Antitrust Review 1. —— —— (2005): “Where Angels Fear to Tread: the Commission’s ‘New’ Leniency Policy Revisited”, GCR European Antitrust Review 10. Joshua J. and Klawiter D. (2002): “International Cartels Investigations: Practical Lessons in Multi-Jurisdictional Coordination”, GCR Antitrust Review of the Americas 7. —— and von Hinten R. (2001): “Rethinking Leniency at the European Commission”, GCR European Antitrust Review 15. —— Low D. and Spratling G. (2001): “International Cartels Investigations: How to Contain the Pain”, GCR Cartel Regulation 3. Jullien B. and Rey P. (2002): Resale Price Maintenance and Collusion, Working Paper, University of Toulouse. Jung H. and Nitschmann K. (2004): “Das Bekenntnis zum Schuldbekenntnis—zur Einführung des plaider coupable im französischen Strafprozess”, Zeitschrift für die gesamte Strafrechtswissenschaft 116. Kahan D. (1997): “Social Influence, Social Meaning, and Deterrence”, 83 Virginia Law Review 349. —— (1998): “Social Meaning and the Economic Analysis of Crime”, 27 Journal of Legal Studies 609. Kandori M. and Matsushima H. (1998): “Private Observation, Communication and Collusion”, 66 Econometrica 627. Katyal N. (1997): “Deterrence’s Difficulty”, 95 Michigan Law Review 2385.


674

16/10/07

14:00

Page 674

Bibliography

Kauper M. (1993)” “The Use of Consent Decrees in US Antitrust Law”, in Slot P. J. and McDonnell A., eds., Procedure and Enforcement in E.C. and U.S. Competition Law, Sweet & Maxwell, London, 105–. Kay K. (2001): “Fundamentals of Criminal Anti-Competitive Conduct”, in Houston D., ed., Papers of the Canadian Bar Association Annual Fall Conference on Competition Law—2000, Juris Publishing, Inc., New York, 231–. Kennealy K. and Breen K. (2005): “White Collar Crime”, 29 Champion 44. Kerse C. and Khan N. (2005): EC Antitrust Procedure, 5th edition, Thomson—Sweet & Maxwell, London. Kihlstrom R. and Vives X. (1992): “Collusion by Asymmetrically Informed Firms”, 1 Journal of Economics and Management Strategy 271. Klawiter D. (2000): “Corporate Leniency in the Age of International Cartels: The American Experience”, 14 Antitrust Magazine 13. —— (2005): “Antitrust Compliance in 2005: It’s Not The Numbers, It’s The Results”, GCR Antitrust Review of the Americas 14. Kolasky W. (2002): “DOJ Antitrust Compliance Programs: The Government Perspective”, speech at the PLI Corporate Compliance 2002 Conference, July 12, 2002, available at http://www.usdoj.gov/atr/public/speeches/11534.pdf. —— (2004): “Criminalizing Cartel Activity: Lessons from the U.S. Experience”, 5th lunch talk of the Global Competition Law Centre, Brussels, 29 September 2004, accessible at http://www.gclc.coleurop.be/lunchtalk20040929.htm. Komninos A. (2002): “New Prospects for Private Enforcement of EC Competition Law: Courage v. Crehan and the Community Right to Damages”, 39 Common Market Law Review 447. Korobkin R. and Ulen T. (2000): “Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics”, 88 California Law Review 1051. Kovacic W. (1992): “The Influence of Economics on Antitrust Law”, 30 Economic Inquiry 294. —— (1993): “The Identification and Proof of Horizontal Agreements Under the Antitrust Laws”, 38 Antitrust Bulletin 81. —— (1996): “Whistleblower Bounty Lawsuits as Monitoring Devices in Government Contracting”, 29 Loyola of Los Angeles Law Review 1799. —— (1997): “Antitrust Policy and Horizontal Collusion in the 21st Century”, 9 Loyola Consumer Law Reporter 97. —— (1998): “The Civil False Claims Act as a Deterrent to Participation in Government Procurement Markets,” 6 Supreme Court Economic Review 201. —— (2001): “Private Monitoring and Antitrust Enforcement: Paying Informants to Reveal Cartels”, 69 George Washington Law Review 766. —— (2003): “Private Participation in the Enforcement of Public Competition Laws”, London, 15 May 2003, available from http://www.ftc.gov/speeches/other/ 030514biicl.htm. —— (2003): “The Modern Evolution of U.S. Competition Policy Enforcement Norms”, 71 Antitrust Law Journal 377. —— (2006): “Competition Policy and Cartels: The Design of Remedies”, in Cseres K., Schinkel M. P. and Vogelaar F., eds., Criminalization of Competition Law Enforcement: Economic and Legal Implications for the EU, Edward Elgar, Northampton.


16/10/07

14:00

Page 675

Bibliography

675

—— Marshall R., Marx L. and Raiff M. (2006): “Bidding Rings and the Design of Anti-Collusion Measures for Auctions and Procurements”, in Dimitri N., Piga G. and Spagnolo G., eds., Handbook of Procurement, Cambridge University Press, Cambridge. Kraakman R. (1984): “Corporate Liability Strategies and the Costs of Legal Controls”, 93 The Yale Law Journal 857. Kroes N. (2005): “Damages Actions for Breaches of EU Competition Rules: Realities and Potentials”, opening speech at the conference ‘La reparation du prejudice cause par une pratique anti-concurrentielle en France et à l’étranger: bilan et perspectives’, Cour de Cassation, Paris, 17 October 2005. —— (2005): “Effective Competition Policy—A Key Tool for Delivering the Lisbon Strategy”, Brussels, 3 February 2005, available from http://europa.eu.int/rapid/ pressReleasesAction.do?reference=SPEECH/05/73&format=HTML&aged=0%E2 %8C%A9uage=EN&guiLanguage=en. —— (2005): “Taking Competition Seriously—Antitrust Reform in Europe”, IBA/Commission Conference, Antitrust Reform in Europe: A Year in Practice, Brussels 10–11 March 2005. —— (2005): “The First Hundred Days”, speech delivered at the 40th Anniversary of the Studienvereinigung Kartellrecht 1965–2005, International Forum on European Competition Law, Brussels, 7 April 2005, available from http://europa.eu.int/rapid/ pressReleasesAction.do?reference=SPEECH/05/205&format=HTML&aged=0&l anguage=EN&guiLanguage=en. —— (2006): “Delivering on the Crackdown: Recent Developments in the European Commission’s Campaign Against Cartels”, speech delivered at the 10th Annual Competition Conference at the European University Institute, Fiesole, 13 October 2006, text available at http://europa.eu/rapid/pressReleasesAction.do?reference= SPEECH/06/595&format=HTML&aged=0&language=EN&guiLanguage=en. —— (2006): “More Private Antitrust Enforcement Through Better Access to Damages: An Invitation for an Open Debate”, speech delivered at the Conference on private enforcement in EC competition law: the Green Paper on damages actions (SPEECH/06/158), Brussels, 9 March 2006, text available from http://europa. eu.int/rapid/pressReleasesAction.do?reference=SPEECH/06/158& format= HTML&aged=0&language=EN&guiLanguage=en. Kühn K. (2001): “Fighting Collusion by Regulating Communication between Firms”, 32 Economic Policy 169. Kunzlik P. (2003): “Globalization and Hybridization in Antitrust Enforcement: European ‘Borrowings’ from the U.S. Approach”, 48 Antitrust Bulletin 319. Laffont J., Rey P. and Tirole J. (1998): “Network Competition: I. Overview and Nondiscriminatory Pricing”, 29 RAND Journal of Economics 1. Lambson V. (1987): “Optimal Penal Codes in Price–Setting Supergames with Capacity Constraints”, 54 Review of Economic Studies 385. —— (1995): “Optimal Penal Codes in Nearly Symmetric Bertrand Supergames with Capacity Constraints”, 14 Journal of Mathematical Economics 1. —— (1996): “Some Results on Optimal Penal Codes in Asymmetric Bertrand Supergames”, 62 Journal of Economic Theory 444. Landes W. and Posner R. (1975): “The Private Enforcement of Law”, 4 Journal of Legal Studies 1.


676

16/10/07

14:00

Page 676

Bibliography

Langlais E. (2006): “Corporate Criminality: Should We Rely on Gatekeepers’ Shortcomings to Deter Dishonest Principals”, mimeo. Langus G. and Motta M. (2006): The Effect of EU Antitrust Investigations and Fines on the Firm’s Valuation, CEPR Discussion Paper 6176, available from http://ideas.repec.org/p/cpr/ceprdp/6176.html. Laufer W. (1999): “Corporate Liability, Risk Shifting, and the Paradox of Compliance”, 52 Vanderbilt Law Review 1343. Lenaerts K. and Maselis I. (2001): “Procedural Rights and Issues in the Enforcement of Articles 81 and 82 of the EC Treaty”, 24 Fordham International Law Journal 1615. —— —— (2004): What Determines Cartel Success?, Working Paper, University of Michigan. —— —— and Oswald L. (2004): “Contemporary International Cartels and Developing Countries: Economic Effects and Implications for Competition Policy”, 71 Antitrust Law Journal 801. Levy N. and O’Donoghue R. (2004): “The EU Leniency Programme Comes of Age”, 27 World Competition 75. Liman A. (1977): “The Paper Label Sentences: Critique”, 86 The Yale Law Journal 619. Lowe P. (2003): “What’s the Future for Cartel Enforcement?”, DG Competition, Brussels, 11 February 2003, text available from http://europa.eu.int/comm/ competition/speeches/text/sp2003_044_en.pdf. —— (2006): “Actions for Damages in Competition Law”, speech delivered at the Conference of the Joint Working Party of the Bars and Law Societies of the United Kingdom on Competition Law, April 7, 2006. Lynch G. (1997): “The Role of Criminal Law in Policing Corporate Misconduct”, 60 Law and Contemporary Problems 23. Malueg D. (1992): “Collusive Behaviour and Partial Ownership of Rivals”, 10 International Journal of Industrial Organization 27. Mansfield P., Baer W. and Spratling G. (2004): “The EU Leniency Program: The Need for Procedural Reform”, paper presented at the ABA International Cartel Workshop, New York, 5–6 February 2004. Marenco G. (2001): “Does a Legal Exception System Require an Amendment of the Treaty?”, in Ehlermann C.-D. and Atanasiu I., eds., European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, London and Portland, Oregon, 145–. Marenco G. (2002): “The Birth of Modern Competition Law in Europe”, in von Bogdandy A., Mavroidis P. and Mény Y., eds., European Integration and International Co-ordination—Studies in Transnational Economic Law in Honour of Claus-Dieter Ehlermann, Kluwer Law International, London, 279–. Marshall R. and Marx L. (2005): Bidder Collusion, Penn State University, mimeo. —— —— and Raiff M. (2005): Cartel Price Announcements: The Vitamins Industry, Duke University Working Paper. Marshall R., Meurer M. and Richard J.-F. (1991): “The Private Attorney General Meets Public Contract Law: Procurement Oversight by Protest”, 20 Hofstra Law Review 1. Martin S. (1993): “Endogenous Firm Efficiency in a Cournot Principal-Agent Model”, 59 Journal of Economic Theory 445.


16/10/07

14:00

Page 677

Bibliography

677

—— (1995): “R&D Joint Ventures and Tacit Product Market Collusion”, 24 European Journal of Political Economy 357. Maskin E. (1999): “Nash Implementation and Welfare Optimality”, 66 Review of Economic Studies 23. Mason C., Phillips O. and Nowell C. (1992): “Duopoly Behavior in Asymmetric Markets: An Experimental Evaluation”, 74 Review of Economics and Statistics 662. McCubbins M. and Schwartz T. (1984): “Congressional Oversight Overlooked: Police Patrols vs Fire Alarms”, 28 American Journal of Political Science 165. McCutcheon B. (1997): “Do Meetings in Smoke–Filled Rooms Facilitate Collusion?”, 105 Journal of Political Economy 330. McElwee D. (2004): “Should the European Commission Adopt ‘Amnesty Plus’ in its Fight Against Hard-Core Cartels”, 25 European Compertition Law Review 558. Mobley S. and Arakistain M. (2000): “Seeking Leniency Worldwide—A Global Strategy”, Global Competition Review 16. Møllgaard P. (2002): Must Trust Bust?, Copenhagen Business School, Department of Economics Working Papers 02/2002. Monnet J. (1976): Mémoires, Fayard, Paris. Monti M. (2000): “Why and How? Why Should We Be Concerned With Cartels and Collusive Behaviour?”, speech delivered at the 3rd Nordic Competition Policy Conference Stockholm, 11–12 September 2000. —— (2004): “Competition for Consumers’ Benefit”, speech delivered at The Netherlands European Competition Day, Amsterdam, 22 October 2004, available at http://europa.eu.int/comm/competition/speeches/ text/sp2004_016_en.pdf. —— (2004): “International Antitrust—A Personal Perspective”, speech delivered at the Fordham Corporate Law Institute, New York, 7 October 2004, text available at http://europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/04/ 449&format=HTML&aged=0&language=EN&guiLanguage=en. —— (2004): “Private Litigation as a Key Complement to Public Enforcement of Competition Rules and the First Conclusions on the Implementation of the New Merger Regulation,” IBA—8th Annual Competition Conference, 17 September 2004. Moore J. (1992): “Implementation, Contracts, and Renegotiation in Environments with Complete Information”, in Laffont J., ed., Advances in Economic Theory, Sixth World Congress, Econometric Society Monographs, Vol. 1, 182–. Motta M. (2004): Competition Policy: Theory and Practice, Cambridge University Press, Cambridge. —— and Polo G. (2003): “Leniency Programs and Cartel Prosecution”, 21 International Journal of Industrial Organization 347. Mukasey M. (1997): “Comment on Lynch”, 60 Law and Contemporary Problems 71. Nevo A. (2000): “Mergers with Differentiated Products: The Case of the Ready-toEat Cereal Industry”, 31 RAND Journal of Economics 395. —— (2001): “Measuring Market Power in the Ready-to-Eat Cereal Industry”, 69 Econometrica 307. Norberg S. (2005): “Some Elements to Enhance Damages Actions for Breach of the Competition Rules in Articles 81 and 82 EC”, paper presented at the 32nd Annual International Antitrust Law & Policy Conference, Fordham, New York, 14 September 2005, available at http://ec.europa.eu/comm/competition/speeches/text/ sp2005_019_en.pdf.


678

16/10/07

14:00

Page 678

Bibliography

Nordlander: K. (2004): “Discovering Discovery—US discovery of EC Leniency Statements”, 24 European Competition Law Review 646. Nörr K. (1995): “Law and Market Organization: The Historical Experience in Germany From 1900 to the Law Against Restraints of Competition (1957)”, 151 Journal of Institutional and Theoretical Economics / Zeitschrift für die gesamte Staatswissenschaft 5. Osborne M. and Pitchik C. (1983): “Price Competition in a Capacity-Constrained Duopoly”, 38 Journal of Economic Theory 238. —— —— (1987): “Cartels, Profits and Excess Capacity”, 28 International Economic Review 413. Page W. (1989): “The Chicago School and the Evolution of Antitrust: Characterization, Antitrust Injury, and Evidentiary Sufficiency”, 75 Virginia Law Review 1221. Parker W. and Röller L.-H. (1997) “Collusive Conduct in Duopolies: Multimarket Contact and Cross-Ownership in the Mobile Telephone Industry,” 28 RAND Journal of Economics 304. Pascual y Vicente J. (2000): “Las conductas prohibidas en la reforma de la Ley de Defensa de la Competencia”, 205 Gaceta Jurídica de la Unión Europea y de la Competencia 13. Pate R. H. (2004): “Securing the Benefits of Global Competition”, address before the Tokyo America Center 4–6 (September 10, 2004), at http://www.usdoj.gov/atr/ public/speeches/205389.pdf. Pinkse J. and Slade M. (2004): “Mergers, Brand Competition, and the Price of a Pint”, 48 European Economic Review 617. Pitofsky R., Goldschmid H. and Wood D. (2003): Cases and Materials on Trade Regulation, 5th edition, Foundation Press, New York. Polinsky A. (1980): “Private Versus Public Enforcement of Fines”, 9 Journal of Legal Studies 105. Porter R. (1985): “On the Incidence and Duration of Price Wars”, 23 Journal of Industrial Economics 415. —— (2005): “Detecting Collusion”, 26 Review of Industrial Organization 147. —— and Zona J. (1999): “Ohio School Milk Markets: An Analysis of Bidding”, 30 RAND Journal of Economics 263. —— (1970): “A Statistical Study of Antitrust Enforcement”, 13 Journal of Law and Economics 365. Posner R. (2001): Antitrust Law, 2nd edition, The University of Chicago Press, Chicago. Raith M. (1996): Product Differentiation, Uncertainty and the Stability of Collusion, London School of Economics–STICERD Discussion Paper Series EI/16:49. Reich R. (1980): “The Antitrust Industry”, 68 Georgetown Law Journal 1035. Rey P. (2003): “Toward a Theory of Competition Policy”, in Dewatripont M., Hansen L. and Turnovsky L., eds., Advances in Economics and Econometrics: Theory and Applications, Cambridge University Press, Cambridge. Reynolds M. and Anderson D. (2006): “Immunity and Leniency in EU Cartel Cases: Current Issues”, 27 European Competition Law Review 82. Riley A. (1991): “European Competition Law: The Client v. the Public Interest”, 12 European Competition Law Review 39. —— (2000): “Saunders and the Power to Obtain Information in Community and United Kingdom Competition Law”, 25 European Law Review 264.


16/10/07

14:00

Page 679

Bibliography

679

—— (2003): “EC Antitrust Modernisation: The Commission Does Very Nicely— Thank You! Part One: Regulation 1 and the Notification Burden”, 24 European Competition Law Review 657. Rill J. and Goldman C. (1995): “A US and Canadian Perspective on International Antitrust Enforcement Post IAEAA, the 1995 International Guidelines and the Canada-US MLAT”, remarks at the Trade Policy Forum VIII, PECC, April 20–21, 1995, Taiwan. Ritter L. and Braun W. D. (2005): European Competition Law: A Practitioner’s Guide, 3rd edition, Kluwer Law International, The Hague. Roberts K. (1985): “Cartel Behaviour and Adverse Selection”, 33 Journal of Industrial Economics 401. Rosochowicz P. (2004): “The Appropriateness of Criminal Sanctions in the Enforcement of Competition Law”, 25 European Competition Law Review 752. Ross T. (1992): “Cartel Stability and Product Differentiation”, 10 International Journal of Industrial Organization 1. Rotemberg J. and Saloner G. (1986): “A Supergame–Theoretic Model of Business Cycles and Price Wars During Booms”, 76 American Economic Review 390. —— —— (1989): “The Cyclical Behaviour of Strategic Inventories”, 104 Quarterly Journal of Economics 73. Schmalensee R. (1987): “Competitive Advantage and Collusive Optima”, 5 International Journal of Industrial Organization 351. Schnichels D. (2004): “The Network of Competition Authorities: How will it work in practice?”, in Geradin D., ed., Modernisation and Enlargement: Two Major Challenges for EC Competition Law, Intersentia, Antwerp, 99–. Schröter H. (1996): “Cartelization and Decartelization in Europe, 1870–1995: Rise and Decline of an Economic Institution”, 25 Journal of European Economic History 129. Shughart W. (1990) “Private Antitrust Enforcement: Compensation, Deterrence, or Extortion?”, 13 Regulation 53. Skrzypacz A. and Hopenhayn A. (2004): “Tacit Collusion in Repeated Auctions”, 114 Journal of Economic Theory 153. Snyder C. (1996): “A Dynamic Theory of Countervailing Power”, 27 RAND Journal of Economics 747. Snyder E. and Kauper T. (1991): “Misuse of the Antitrust Laws: The Competitor Plaintiff”, 90 Michigan Law Review 551. Spagnolo G. (2007): “Fines, Leniency and Whistleblowers in Antitrust”, in Buccirossi P, ed., Handbook of Antitrust Economics, MIT Press, Massachussets. Spierenburg D. and Poidevin R. (1994): The History of the High Authority of the European Coal and Steel Community, Weidenfeld and Nicholson, London. Spink P. (1999): “Recent Guidance on Fining Policy’, 20 European Competition Law Review 101. Spratling G. (1998): “The Corporate Leniency Policy: Answers to Recurring Questions,” paper presented at the ABA Antitrust Section Spring Meeting, Washington, DC, April 1, 1998. —— (1999): “International Cartel Prosecutions: Antitrust Division Policies Relating to Plea Agreements in International Cases,” presented at the ABA Criminal Justice Section’s Thirteenth Annual National Institute on White Collar Crime, San Francisco, California, March 4, 1999.


680

16/10/07

14:00

Page 680

Bibliography

Spratling G. (1999): “Negotiating the Waters of International Cartel Prosecutions— Antitrust Division Policies Relating to Plea Agreements in International Cases”, speech before the ABA Criminal Justice Section’s Thirteenth Annual National Institute on White Collar Crime, March 4, 1999, available at http://www.usdoj.gov/ atr/public/speeches/2275.htm. —— (19999): “Making Companies an Offer They Shouldn’t Refuse”, paper presented at the Bar Association of the District of Columbia’s 35th Annual Symposium on Associations and Antitrust, Washington, DC, February 16, 1999, available at http://www.usdoj.gov/atr/public/speeches/2247.pdf. —— Jarrett D. and Shephard A. (2006): “Making the Decision: What to Do When Faced with International Cartel Exposure. Developments Impacting the Decision in 2006”, paper for American Bar Association Section of Antitrust Law 54th Antitrust Law Spring Meeting, Washington DC. Staiger R. and Wolak F. (1992): “Collusive Pricing with Capacity Constraints in the Presence of Demand Externalities”, 23 RAND Journal of Economics 203. Stark C. (2000): “Improving Bilateral Antitrust Cooperation,” Competition Policy in the Global Trading System, Washington, DC, June 23, 2000. Stephan A. (2006): The Bankruptcy Wildcard in Cartel Cases, CCP Working Paper 06-5, March 2006, University of East Anglia, Centre for Competition Policy. Stigler G. (1964): “A Theory of Oligopoly”, 72 The Journal of Political Economy 44. Stone C. (1980): “The Place of Enterprise Liability in the Control of Corporate Conduct”, 90 The Yale Law Journal 1. —— (1991): “Sentencing the Corporation”, 71 Boston University Law Review 383. Sullivan M. and Filion J. (2004): “The Basics of International Cartel Enforcement in Canada”, presented at the Osgoode Hall Continuing Legal Education Program, “Canada’s Competition Regime: Thinking Strategically A Practical Guide for Business”, January 14, 2004. Sunstein C. (1996): “On the Expressive Function of the Law”, 144 University of Pennsylvania Law Review 2021. Sutton J. (1996): “Investigations under the Competition Act: Recent Issues”, in The Canadian Institute on Competition, ed., Important Changes in Competition Law and Competitive Business Practices. Swaak C. and Mollica M. (2005): “Leniency Applicants to Face Modernisation to EC Competition Law”, 26 European Competition Law Review 507. Symeonidis G. (2002): “Cartel Stability and Multiproduct Firms”, 20 International Journal of Industrial Organization 339. —— (2003): “In Which Industries is Collusion More Likely? Evidence from the UK”, 48 Journal of Industrial Economics 1. Telser L. (1960): “Why Should Manufacturers Want Fair Trade?”, 3 Journal of Law and Economics 86. Thalm J. (2005): “Optimal Collusion under Cost Asymmetry”, University of Toulouse, mimeo. Thompson D. (2003): “Principles of Federal Prosecution of Business Organizations”, US Department of Justice, January 20, 2003, available at http://www.usdoj.gov/ dag/cftf/corporate_guidelines.htm. Tirole J. (1988): The Theory of Industrial Organisation, MIT Press, Massachusetts.


16/10/07

14:00

Page 681

Bibliography

681

Turner V. (2005): “International Cartel Enforcement: Multijurisdictional Defence Strategies”, paper presented at the ABA Section of International Law Annual Conference, Brussels 26–29 October 2005. Tyler T. (1990): Why People Obey the Law, Yale University Press, New Haven. Ulen T. (1997): “The Economic Case for Corporate Criminal Sanctioning”, in Lofquist W., Cohen M. and Rabe G., eds., Debating Corporate Crime, Anderson, New York, 117–. Valentine D. (2000): “Cross-Border/US Cooperation in Investigations and Enforcement Actions”, paper presented at the conference organized by the Canada/United States Law Institute, Case Western Reserve University School of Law, April 15, 2000. Van Bael I. (1995): “The Lottery of EU Competition Law”, 4 European Competition Law Review 237. Van Barlingen B. (2003): “The European Commission’s 2002 Leniency Notice After One Year of Operation”, 2 Competition Policy Newsletter 18. —— and Barennes M. (2005): “The European Commission’s 2002 Leniency Notice in Practice”, 3 Competition Policy Newsletter 10. Venit J. (2003): ‘‘Brave New World: The Modernization and Decentralization of Enforcement Under Arts 81 and 82 of the EC Treaty’’, 40 Common Market Law Review 545. —— and Louko T, “The Commission’s New Power to Question and its Implications on Human Rights”, in Hawk B., ed., 2004 Annual Proceedings of the Fordham Corporate Law Institute—International Antitrust Law and Policy, Juris Publishing, New York. Verboven F. (1998): “Localized Competition, Multimarket Operation and Collusive Behaviour”, 39 International Economic Review 371. Von Finckenstein K. (2000): “Opening Remarks”, delivered at the American Bar Association, Section of Antitrust Law Panel on Global Warming?: International Reaction to the ICPAC Report, New York, July 11, 2000. —— (2001): “International Antitrust Cooperation: Bilateralism or Multilateralism”, presentation to a joint meeting of the American Bar Association Section of Antitrust Law and the Canadian Bar Association National Competition Law Section, May 31, 2001, Vancouver, Canada. Waelbroeck D. (2006): “Actions for Damages in Competition Law”, paper presented at the Conference of the Joint Working Party of the Bars and Law Societies of the United Kingdom on Competition Law, London, 7 April 2006. Waelbroeck M. and Frignani A. (1997): Commentaire Megret. Vol. IV: Concurrence, ULB Brussels. Waller S. (2003): “The Incoherence of Punishment in Antitrust”, 78 Chicago-Kent Law Review 207. Wehmhörner N. (2005): “Optimal Fining Policies”, paper prepared for the conference “Remedies and Sanctions in Competition policy”, Amsterdam University, text available at http://www.cms.uva.nl/acle/object.cfm/objectid=F07DE744-C1D14F2E-876EEB31F7FA5B9F. Werden G. (2004): “Economic Evidence on the Existence of Collusion: Reconciling Antitrust Law with Oligopoly Theory”, 71 Antitrust Law Journal 719. —— and Simon M. (1987): “Why Price Fixers Should Go to Prison”, The Antitrust Bulletin 917.


682

16/10/07

14:00

Page 682

Bibliography

Whish R. (2003): Competition Law, 5th edition, Butterworths, London. White S. (2005): “Harmonisation of Criminal Law under the First Pillar”, 31 European Law Review 81. Williamson O. E. (1969): “Economies as an Antitrust Defense: The Welfare Tradeoffs”, 58 American Economic Review 18. Wils W. P. J. (2002): The Optimal Enforcement of EC Antitrust Law, Kluwer Law International, London. —— (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–. —— (2003): “Should Private Antitrust Enforcement Be Encouraged in Europe?”, 26 World Competition 473. —— (2003): “Self-incrimination in EC Antitrust Enforcement: A Legal and Economic Analysis”, 26 World Competition 567. —— (2004): “The Combination of the Investigative and Prosecutorial Function and the Adjudicative Function in EC Antitrust Enforcement: A Legal and Economic Analysis”, 27 World Competition 201. —— (2005): “Is Criminalization of EU Competition Law the Answer?”, 28 World Competition 117. —— (2005): Principles of European Antitrust Enforcement, Hart Publishing, Oxford and Portland, Oregon. —— (2006): “Optimal Antitrust Fines: Theory and Practice”, 28 World Competition 117. —— (2006): “Powers of Investigation and Procedural Rights and Guarantees in EU Antitrust Enforcement: The Interplay between European and National Legislation and Case-law”, 29 World Competition 3. Zane P. (2003): “The Price Fixer’s Dilemma: Applying Game Theory to the Decision of Whether to Plead Guilty to Antitrust Crimes”, 48 The Antitrust Bulletin 1. Zuleeg M. (2003): “Criminal Sanctions to Be Imposed on Individuals As Enforcement Instruments in European Competition Law”, in Ehlermann C.-D. and Atanasiu I., eds. (2003): European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, Oxford and Portland, Oregon, 453–.

II. Legislation, preparatory acts etc.: Agreement between the Government of Canada and the Government of the United States of America Regarding the Application of their Competition and Deceptive Marketing Practices Laws, August 3, 1995. Antitrust Division, US Department of Justice (1993): Corporate Leniency Policy, available at http://www.usdoj.gov/atr/public/guidelines/0091.htm. Antitrust Division, US Department of Justice (1994): Leniency Policy for Individuals, available at http://www.usdoj.gov/atr/public/guidelines/0092.htm. Bundeskartellamt (2006): Notice no. 9/2006 of the Bundeskartellamt on the immunity from and reduction of fines in cartel cases, Bonn, 7 March 2006, available from http://www.bundeskartellamt.de/wEnglisch/download/pdf/06_Bonusregelung_e.pdf.


16/10/07

14:00

Page 683

Bibliography

683

Canadian Competition Bureau (2000): Draft Immunity Program Under the Competition Act, available at http://strategis.ic.gc.ca/SSG/ct01705e.html. Commission Communication on the implications of the Court’s judgment of 13 September 2005 (Case C-176/03 Commission v. Council), COM(2005) 583 final. Commission Guidelines on the method of setting fines imposed pursuant to Article 23(2)(a) of Regulation No 1/2003, OJ C 210 [2006]. Commission Guidelines on the method of setting fines pursuant to Article 15(2) of Regulation 17, OJ C 9 [1998]. Commission Notice on cooperation within the network of competition authorities, OJ C 101 [2004]. Commission Notice on immunity from fines and reduction of fines in cartel cases, OJ C 45 [2002]. Commission Notice on immunity from fines and reduction of fines in cartel cases, OJ C 298 [2006]. Commission Notice on the co-operation between the Commission and the courts of the EU Member States in the application of Articles 81 and 82 EC, OJ C 101 [2004]. Commission Notice on the non-imposition or reduction of fines, OJ C 45 [2002]. Commission Notice on the non-imposition or reduction on fines in cartel cases, OJ C 207 [1996]. Commission Notice on the rules for access to the Commission file in cases pursuant to Articles 81 and 82 of the EC Treaty, Articles 53, 54 and 57 of the EEA Agreement, and Council Regulation (EC) No 139/2004, OJ C 325 [2005]. Commission Proposal for a Regulation of the European Parliament and of the Council establishing a European Small Claims Procedure, COM(2005)87 of 15 March 2005. Commission Proposal for a Regulation of the European Parliament and the Council on the law applicable to non-contractual obligations (“Rome II”), COM(2006)83 of 21 February 2006. Commission Report to the Council and the European Parliament on the application of the Agreement between the European Communities and the Government of the United States of America regarding the application of their competition laws, COM(1996) 479 final. 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 [2003]. Council Regulation 44/2001 of 22 December 2000 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters, OJ L 12 [2001]. Draft amendment of the 2002 Commission Notice on Immunity from fines and reduction of fines in cartel cases, 22 February 2006, available on http://ec.europa.eu/comm/ competition/antitrust/legislation/leniency2_en.pdf. Irish Competition Authority (2001): Cartel Immunity Programme, available from http://www.tca.ie. Green Paper on damages actions for breach of the EC antitrust rules, COM (2005) 672. Joint Statement of the Council and Commission on the Functioning of the Network of Competition Authorities, 15435/02 ADD 1. Office of Fair Trading (1998): Guidance as to the Appropriate Amount of a Penalty, available at www.oft.gov.uk/shared_oft/business_leaflets/ca98_guidelines/oft423.pdf. Office of Fair Trading (2003): Competition Disqualification Orders—Guidance, available at http://www.oft.gov.uk/Business/Legal+Powers/Enterprise+Act/publications.htm.


684

16/10/07

14:00

Page 684

Bibliography

Office of Fair Trading (2003): The Cartel Offence—Guidance on the Issue of No-action Letters for Individuals, available at http://www.oft.gov.uk/Business/Legal+Powers/ Enterprise+Act/publications.htm. Office of Fair Trading (2005): Leniency and No-action, available from http://www.oft. gov.uk/NR/rdonlyres/FE4F98BF-D8C5-48FB-9F7E-6133F907538/0/oft803.pdf. Regulation 1049/2001 of the European Parliament and of the Council of 30 May 2001 regarding public access to European Parliament, Council and Commission documents, OJ L 145 [2001]. Regulation 773/2004 relating to the conduct of proceedings by the Commission pursuant to Articles 81 and 82 of the EC Treaty, OJ L 123 [2004]. Treaty between the Government of Canada and the Government of the United States on Mutual Assistance in Criminal Matters, Canada Gazette, Part I, 953(1990) [MLAT]. Treaty of Extradition United States-Canada, of March 22, 1976, 1976 Can. T.S., No. 3. US Department of Justice, Antitrust Division (1991): Antitrust Grand Jury Practice Manual, Chapter 9 (1st ed. 1991), available at http://www.usdoj.gov/atr/public/ guidelines/207144.htm. US Department of Justice, Antitrust Division (1993): Corporate Leniency Policy, reprinted in 4 Trade Reg. Rep. (CCH) 13,113.

III. Official reports, studies etc.: European Commission (2002): XXXIInd Report on Competition Policy, available from http://europa.eu.int/comm/competition/annual_reports/2002/en.pdf. European Commission (2004): Commitment decisions (Article 9 of Council Regulation 1/2003 providing for a modernised framework for antitrust scrutiny of company behaviour), MEMO/04/217, 17 September 2004. Federal Trade Commission (2003): Bureau of Economics Contributions to Law Enforcement, Reseach, and Economic Knowledge and Policy (2003), transcript of a roundtable with former Directors of the Bureau of Economics, http://www.ftc.gov/ be/workshops/directorsconference/index.htm. IDEI (2003): The Economics of Tacit Collusion, report for DG Competition of the European Commission, available at http://ec.europa.eu/comm/competition/ mergers/review/ the_economics_of_tacit_collusion_en.pdf. International Competition Network (2006): “Drafting and Implementing an Effective Leniency Program”, in Anti-Cartel Enforcement Manual, available from http://www.internationalcompetitionnetwork.org/capetown2006/FINALFormatte dChapter2-modres.pdf. OECD (1998): Recommendation of the Council Concerning Effective Action Against Hard Core Cartels, OECD C(98)35 (1998), available at http://www.oecd.org/daf/ clp/Recommendations/Rec9com.htm. OECD (2002): Report on the Nature and Impact of Hard Core Cartels and Sanctions Against Cartels Under National Competition Laws, DAFFE/COMP(2002)7 (9 April 2002), available at http://www.oecd.org/dataoecd/16/20/2081831.pdf. OECD (2003): Hard Core Cartels—Recent Progress and Challenges Ahead, available from http://www1.oecd.org/publications/e-book/2403011E.PDF.


16/10/07

14:00

Page 685

Bibliography

685

OECD (2004): Cartels: Sanctions Against Individuals, DAFFE/COMP(2004) 39. OECD (2005): Hard Core Cartels: Third Report on the Implementation of the 1998 Recommendation, available at http://www.oecd.org/dataoecd/58/1/35863307.pdf. OECD (2006): Competition Law and Policy in the European Union, available from http://www.oecd.org/dataoecd/7/41/35908641.pdf. OECD Competition Committee (2005): Ten Strategies for Winning the Fight Against Hardcore Cartels, Paris, France. OECD: Glossary of Industrial Organisation Economic and Competition Law, available at http://www.oecd.org/dataoecd/8/61/2376087.pdf. US Department of Justice and FTC (1999): Note for the Mini-Roundtable on Oligopoly, OECD Paris, DAFFE/CLP/WD(99)13, available at http://www.ftc.gov/bc/ international/docs/compcomm/1999—Mini-roundtable%20 on%20Oligopoly.pdf. Waelbroeck D., Slater D. and Even-Shoshan G. (2004): Study on the Conditions of Claims for Damages in Case of Infringement of EC Competition Rules (“the Ashurst study”) , 31 August 2004, text available at http://ec.europa.eu/comm/competition/ antitrust/others/actions_for_damages/comparative_report_clean_en.pdf#search= %22waelbroeck%20damages%20conditions%22.


16/10/07

14:00

Page 686

European Competition Law Annual 2006: Enforcement of Prohibition of Cartels

European Competition Law Annual 2002 (European Competition Law Annual)

European Competition Law Annual 1997: Objectives of Competition Policy

European Competition Law Annual 2005: The Interaction between Competition Law and Intellectual Property Law (European Competition Law Annual)

European Competition Law Annual 1998: Regulating Communications Markets (European Competition Law Annual)

European Competition Law Annual, 2004: The Relationship Between Competition Law and the (Liberal) Professions (European Competition Law Annual)

Economics and the Enforcement of European Competition Law

European Competition Law Annual 2000: The Modernisation of EU Competition Law

The Enforcement of Competition Law in Europe (The Common Core of European Private Law)

The Evolution of European Competition Law: Whose Regulation, Which Competition?

Principles of European Antitrust Enforcement

A Principled Approach to Abuse of Dominance in European Competition Law (Antitrust and Competition Law)

Select Proceedings of the European Society of International Law: 2006

European Competition Law: A Practitioner's Guide

Law Without Enforcement

International Competition Enforcement Law Between Cooperation and Convergence

Law enforcement career starter

Law and Law Enforcement Issues

Encyclopedia of Law Enforcement (3 Vol Set)

The Economics of Prohibition

Law Enforcement Jiu-Jitsu

NBER Macroeconomics Annual 2006

Annual Review of Psychology Vol 57, 2006

Annual Review of Immunology Volume 24 2006

India Macroeconomics Annual 2006

The Europeanisation of Law (European Law Series)

Abingdon Preaching Annual 2006

Principles of European Constitutional Law (Modern Studies in European Law)

The prohibition of torture

Law Enforcement Career Guides: California

European Competition Law Annual 2006: Enforcement of Prohibition of Cartels

European Competition Law Annual 2002 (European Competition Law Annual)

European Competition Law Annual 1997: Objectives of Competition Policy

European Competition Law Annual 2005: The Interaction between Competition Law and Intellectual Property Law (European Competition Law Annual)

European Competition Law Annual 1998: Regulating Communications Markets (European Competition Law Annual)

European Competition Law Annual, 2004: The Relationship Between Competition Law and the (Liberal) Professions (European Competition Law Annual)

Economics and the Enforcement of European Competition Law

European Competition Law Annual 2000: The Modernisation of EU Competition Law

The Enforcement of Competition Law in Europe (The Common Core of European Private Law)

The Evolution of European Competition Law: Whose Regulation, Which Competition?

Principles of European Antitrust Enforcement

A Principled Approach to Abuse of Dominance in European Competition Law (Antitrust and Competition Law)

Select Proceedings of the European Society of International Law: 2006

European Competition Law: A Practitioner's Guide

Law Without Enforcement

International Competition Enforcement Law Between Cooperation and Convergence

Law enforcement career starter

Law and Law Enforcement Issues

Encyclopedia of Law Enforcement (3 Vol Set)

The Economics of Prohibition

Law Enforcement Jiu-Jitsu

NBER Macroeconomics Annual 2006

Annual Review of Psychology Vol 57, 2006

Annual Review of Immunology Volume 24 2006

India Macroeconomics Annual 2006

The Europeanisation of Law (European Law Series)

Abingdon Preaching Annual 2006

Principles of European Constitutional Law (Modern Studies in European Law)

The prohibition of torture

Law Enforcement Career Guides: California

Recommend Documents