European Competition Law Annual 2007: A Reformed Approach to Article 82 EC 9781472560353, 9781841138381

This is the twelfth in a series on EU Competition Law and Policy produced by the Robert Schuman Centre of the European U

225 71 4MB

English Pages [667] Year 2008

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

European Competition Law Annual 2007: A Reformed Approach to Article 82 EC
 9781472560353, 9781841138381

Citation preview

(A) Ehlermann 08 Prelims

10/11/08

13:01

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 Commerce Court West 199 Bay Street Toronto, Ontario Canada M5L 1A9 Tel.: +1 416 863 22 80 Fax: +1 416 863 26 53 E-mail: [email protected] Cleary, Gottlieb, Steen & Hamilton Contact: Prof. Mario Siragusa Rome Office Piazza di Spagna, 15 I-00187 Rome Italy Tel.: +39 06 695 221 Fax: +39 06 692 00 665 E-mail: [email protected] CRA International Contact: Lorenzo Coppi Vice President Orion House 5 Upper St Martin’s Lane London WC2H 9EA UK Tel.: +44 20 7438 8117 Fax: +44 20 7240 6136 E-mail: [email protected] Hengeler Müller Contact: Jochen Burrichter Benrater Strasse 18-20 D-40213 Düsseldorf, Germany Fax: +49 211 83 04 222 E-mail: [email protected]

(A) Ehlermann 08 Prelims

vi

10/11/08

13:01

Page vi

Sponsors

Howrey LLP Contact: James Rill Esq. 1299 Pennsylvania Ave., NW Washington, DC 20004 USA Tel.: +1 202 383 65 62 Fax: +1 202 383 66 10 E-mail: [email protected] LECG Contact: Atilano Jorge Padilla Managing Director Park Atrium 11 Rue des Colonies B-1000 Brussels, Belgium Tel.: +32 2 517 6070 Fax: +32 2 788 1211 E-mail: [email protected] Martínez Lage & Asociados Contact: Santiago Martínez Lage Claudio Coello 37 ES-28001 Madrid Spain Tel (34) 91 426 44 70 Fax (34) 91 577 37 74 E-mail: [email protected] RBB Economics Contact: Andrea Lofaro RBB Economics London The Connection 198 High Holborn London, WC1V 7BD UK Tel.: +44 20 7421 2414 Fax: +44 20 7421 2411 E-mail: [email protected]

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page vii

Sponsors vii Skadden, Arps, Slate, Meagher & Flom LLP Contact: James Venit Brussels Office 523 Avenue Louise B-1050 Brussels Belgium Tel.: +32 2 639 03 00 Fax: +32 2 639 03 39 E-mail: [email protected] White & Case LLP Contact: Prof. Ian S. Forrester Brussels Office 62, rue de la Loi B-1040 Brussels Belgium Tel.: +32 2 219 16 20 Fax: +32 2 219 16 26 E-mail: [email protected] WilmerHale LLP 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]

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page viii

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page ix

1 2 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: A REFORCED APPROACH TO ARTICLE 82 EC

Chairs Giuliano Amato, European University Institute—Florence John Fingleton, Office of Fair Trading—London Massimo Motta, European University Institute—Florence Heike Schweitzer, European University Institute—Florence

List of Participants Rafael Allendesalazar Corcho, Martínez Lage & Asociados—Madrid Simon Bishop, RBB Economics—London William Blumenthal, U.S. Federal Trade Commission—Washington, DC Jochen Burrichter, Hengeler Müller—Düsseldorf Andrea Coscelli, CRA International—London Claus-Dieter Ehlermann, WilmerHale—Brussels Einer Elhauge, Harvard Law School Amelia Fletcher, Office of Fair Trading—London Ian Forrester, White and Case—Brussels Eleanor Fox, New York University School of Law Damien Geradin, University of Tilburg; Howrey—Brussels David Gerber, Chicago-Kent College of Law—Chicago Cal Goldman, Blake, Cassels & Graydon—Toronto Martin Hellwig, Max Planck Institute for Research on Collective Goods— Bonn Judge Frédéric Jenny, Cour de Cassation and ESSEC Business School— Paris; OFT—London Philip Lowe, DG Competition—Brussels Mel Marquis, Robert Schuman Centre, EUI—Florence Doug Melamed, WilmerHale—Washington DC David Meyer, U.S. Department of Justice—Washington DC Damien Neven, DG Competition—Brussels

(A) Ehlermann 08 Prelims

x

10/11/08

13:01

Page x

List of Participants

Robert O’Donoghue, Brick Court Chambers—London & Brussels Atilano Jorge Padilla, LECG, Brussels—London and Madrid Mark Patterson, Fordham University School of Law—New York; EUI— Florence Emil Paulis, DG Competition—Brussels Pierre Régibeau, University of Essex Patrick Rey, University of Toulouse Lars-Hendrik Röller, European School of Management and Technology— Berlin Mario Siragusa, Cleary Gottlieb Steen & Hamilton—Rome Marc van der Woude, Stibbe—Brussels Jim Venit, Skadden, Arps, Slate, Meagher & Flom—Brussels Sir John Vickers, University of Oxford Daniel Zimmer, University of Bonn

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xv

TABLE OF CASES I. Judgments of the European Courts Joined Cases C-204/00 P et al., Aalborg Portland A/S and Others v Commission [2004] ECR I-123.................................................................. lxxix, 351 Case C-89/85, Ahlström Osakeyhtiö and Others [1993] ECR I-1307 ...................... 105 Case 66/86, Ahmed Saeed Flugreisen and SilverLine Reisebüro GmbH v Zentrale zur Bekämpfung unlauteren Wettbewerbs eV [1989] ECR 803 ................. lxxvii, 145 Case T-342/99, Airtours v Commission [2002] ECR II-2585 .................................. 111 Case C-109/01, Secretary of State for the Home Department v Akrich [2003] ECR I-9607 ............................................................................................................ xxxiv Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359 ....................................................... xxxv, lxiii, 70–1, 151–2, 154, 330–1, 338 Case C-67/96, Albany International BV v Stichting Bedriffspensioenfonds Textielindustrie [1999] ECR I-5751 ............................................................ xl, cxxvi Joined Cases T-191 and T-212/98 to T-214/98, Atlantic Container Line and Others v Commission [2003] ECR II-3275 ...................... xlii, 72, 141, 264, 322, 351 Case 41/96, Bayer AG v Commission [2000] ECR II-3383 ..................................... 696 Case 127/73, Belgische Radio en Televisie (BRT) v SV SABAM and NV Fonior [1974] ECR 313 ...................................................................................... lxxviii, 615 Case 77/77, Benzine en Petroleum Handelsmaatschappij BV and others v Commission [1978] ECR 1513 ........................................................................... 346 Case 30/87, Corinne Bodson v Pompes funèbres des régions libérées [1988] ECR 2479 ........................................................................... 145, 568, 624, 684, 688 Case T-65/89, BPB Industries and British Gypsum Ltd v Commission [1993] ECR II-389; upheld on appeal: Case C-310/93 P, [1995] ECR I-865 ...................................................................................... 174–5, 264, 322 Case C-427/93, Bristol-Myers Squibb and Others v Paranova [1996] ECR I-3457; Case C-267/95 Merck and Others v Primecrown and Others [1996] ECR I-6285 ...................................................................................................... 495 Case T-219/99, British Airways v Commission [2003] ECR II-5917; upheld on appeal: Case C-95/04 P, [2007] ECR I-2331 ................ lxxi, lxvii, lxxii, lxxvii, lxxxi, lxxxiii, 28, 57, 73, 82, 88, 103–5, 140–42, 176, 197, 242, 250, 259, 265–8 Case 226/84, British Leyland Plc v Commission [1986] ECR 3263 .... 74, 145, 568, 625 Case C-125/97, Eco Swiss China Time Ltd v Benetton International NV [1999] ECR I-3055 ............................................................................................ xliv Case C-7/97, Oscar Bronner GmbH & Co KG v Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co KG, Mediaprint Zeitungsvertriebsgesellschaft mbH & Co KG and Mediaprint Anzeigengesellschaft mbH & Co KG [1998] ECR I-7791 .................................................... xlvi–xlvii, 82, 156–9, 176, 338 Case C-258/98, Giovanni Carra and Others [2000] ECR I-4217 ............................. 356 Case 311/84, Centre belge d’études de marché/Télémarketing (CBEM) v SA Compagnie luxembourgeoise de télédiffusion (CLT) and Information publicité Benelux (IPB) [1985] ECR 3261 ................................................. 346, 356

(A) Ehlermann 08 Prelims

xvi

10/11/08

13:01

Page xvi

Table of Cases

Joined Cases C-94/04 and C-202/04, Cipolla and Macrino [2006] ECR I-11421 .................................................................................................. xxxiv, xcix Joined Cases T-24/93, T-25/93, T-26/93, and T-28/93, Compagnie Maritime Belge Transports SA and Others v Commission [1996] ECR II-1201, upheld: Joined Cases C-395/96 P and C-396/96 P, Compagnie Maritime Belge Transports SA and Others v Commission [2000] ECR I-1365 ............................. 330 Case 53/87, Consorzio italiano della componentistica di ricambio per autoveicoli and Maxicar v Régie nationale des usines Renault [1988] ECR 6039 .................. 356 Joined Cases 56 and 58/64, Consten and Grundig v Commission [1966] ECR 299 ..................................................................................................... 27, 214 Case 6/72, “Continental Can” [1973] ECR 215 .................................. xliv, xlv, 11, 19, 104, 123, 138–40, 142, 145, 161, 175, 461, 499, 515–6 Cases 40/73 to 48/73, 50/73, 54/73 to 56/73, 111/73, 113/73 and 114/73, Coöperatieve Vereniging “Suiker Unie” UA and Others v Commission [1975] ECR 1663 ..................................................................... 105, 175–6, 331, 356 Case C-18/93, Corsica Ferries Italia Srl v Corpo dei Piloti del Porto di Genova [1994] ECR I-1783................................................................................. 701 Case C-453/99, Courage Ltd v Bernard Crehan [2001] ECR I-6297 ...................................................... lxii, lxxi, 201, 243, 245–6, 252, 280 Case T-151/01, Der Grüne Punkt – Duales System Deutschland GmbH v Commission [2007] ECR II-1607, on appeal: Case C-385/07, not yet decided .... 643 Case T-229/94, Deutsche Bahn AG v Commission [1997] ECR II-1689 .................. 701 Case 78/70, Deutsche Grammophon Gesellschaft mbH v Metro-SB-Großmärkte GmbH & Co KG [1971] ECR 487 ............................................................... 346, 684 Case T-271/03, Deutsche Telekom AG v Commission, judgment of the Court of First Instance of 10 April 2008, not yet reported ..........................................lxxv Case C-63/93, Duff and Others v Minister for Agriculture and Food and Attorney General [1996] ECR I-569 ................................................................................. 358 Case T-340/03, France Télécom v Commission [2007] ECR II-107; on appeal: Case C-202/07, not yet decided ....................................... 8, 111, 151, 322, 324, 349 Case T-210/01, General Electric v Commission [2005] ECR II-5575 ...................... 182 Case 26/75, General Motors Continental NV v Commission [1975] ECR 1367 .......................................................................................... 146, 499, 643 Case C-313/99, Gerard Mulligan and Others v Agriculture and Food, Ireland and Attorney General [2002] ECR I-5719 ...................................................... 357–8 Case T-168/01, GlaxoSmithKline Services v Commission [2006] ECR II-2969; on appeal: Cases C-501/06, C-513/06, C-515/06 and C-519/06, not yet decided .................................................... xxxi, lxxix, 22, 25, 27–8, 63, 174, 242, 367, 381, 383, 421, 605–6, 720 Case C-250/92, Gøttrup-Klim Grovvareforeninger v Dansk Landbrugs Grovvareselskab AmbA [1994] ECR I-5641 ..................................... xxxix, 105, 585 Case C-242/95, GT-Link v DSB [1997] ECR I-4449 .............................................. 145 Case C-10/89, SA CNL-SUCAL NV v HAG GF AG (“ Hag II”) [1990] ECR I-3711 .................................................................................................... xxxiv Case T-30/89, Hilti AG v Commission [1991] ECR II-1439; upheld on appeal: Case C-53/92, [1994] ECR I-667 ....................................................................... 378 Case 85/76, Hoffmann-La Roche and Co AG v Commission [1979] ECR 461 .................... 15, 60, 70, 72, 140–41, 161, 174–6, 300, 331, 382, 676

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xvii

Table of Cases

xvii

Case C-41/90, Klaus Höfner and Fritz Elsner v Macrotron GmbH [1991] ECR I-1979 ............................................................................................ 356 Case T-209/01, Honeywell v Commission [2005] ECR II-5527 ......................... 34, 182 Case 22/78, Hugin Kassaregister AB v Commssion [1979] ECR 1869 .................... 111 Case T-184/01 R, IMS Health Inc. v Commission (interim measures) [2001] ECR II-3193, upheld: Case C-481/01 P(R), NDC Health Corp. [2002] ECR I-3401 ...................................................................................................... 705 Case C-418/01, IMS Health v NDC Health [2004] ECR I-5039 ....................................... xlvii, cxxiv, 159, 175, 521, 705, 707–13, 718–9, 721 Joined Cases 6/73 and 7/73, Istituto Chemioterapico Italiano and Commercial Solvents v Commission [1974] ECR 223 ............................................................. 176 Zoja/CSC–ICI (“Commercial Solvents”) [1972] OJ L299/51 ............................ 566–9 Case C-438/05, International Transport Workers’ Federation and Finnish Seaman’s Union v Viking Line ABP and OÜ Viking Line Eesti [2007] ECR I-10779 ................................................................................................ xl, xliv Case T-228/97, Irish Sugar v Commission [1999] ECR II-2969; upheld on appeal: Case C-497/99, [2001] ECR I-5333 .................................. 174–5, 265, 282, 322, 346 Case 13/63, Italy v Commission [1963] ECR 165 ................................................... 698 Case 41/83, Italy v Commission (British Telecommunications) [1985] ECR 873 ........................................................................................................... 356 Case T-111/96, ITT Promedia v Commission [1998] ECR II-2937 ................. 174, 357 Case C-55/96, Job Centre coop arl [1997] ECR I-7119 .......................................... 356 Case T-19/92, Groupement d’achat Edouard Leclerc v Commission [1996] ECR II-1851 ..................................................................................................... 585 Case T-88/92, Groupement d’achat Edouard Leclerc v Commission [1996] ECR II-1961 ..................................................................................................... 585 Case C-267/91, Keck and Mithouard [1993] ECR I-6097..................................... xxxiv Case C-341/05, Laval un Partneri LTD v Svenska Byggnadsarbetareföbundet and Others [2007] ECR I-11767 ........................................................................... xl Case 31/80, L’Oreal v PVBA De Nieuwe Amck [1980] ECR 3775 ......................... 174 Joined Cases 110/88, 241/88 and 242/88, François Lucazeau and others v SACEM and Others [1989] ECR 2811 .............. 145–6, 568, 624, 628, 639, 653, 684 Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni and Others [2006] ECR I-6619 ............ lxii, lxxi, 201, 243, 245–6, 252 Case T-203/01, Manufacture Française des Pneumatiques Michelin v Commission (“Michelin II”) [2003] ECR II-4071 ....... lxxii, lxxxiii, 8, 70, 72–3, 84, 86, 111, 151, 174–5, 197, 214, 269–70, 300, 385 Case C-344/98, Masterfoods and HB Ice Cream Ltd [2000] ECR I-11369 ............. 210 Case T-54/99, max.mobil Telekommunikation v Commission [2002] ECR II-313 ....................................................................................................... 175 Case 187/80, Merck & Co. Inc. v Stephar BV and Petrus Stephanus Exler [1981] ECR 2063 ............................................................................................... 495 Case C-127/08, Metock and Others v Minister for Justice, Equality and Law Reform, judgment of the ECJ of 25 July 2008, not yet reported ..................... xxxiv Case 26/76, Metro-SB-Gro‚märkte GmbH & Co. KG v Commission [1977] ECR 1875 ......................................................................................................... 585 Case75/84, Metro-SB-Gro‚märkte GmbH & Co. KG v Commission (“Metro II”) [1986] ECR 3021 ............................................................................................... 585

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xviii

xviii Table of Cases Case T-201/04, Microsoft v Commission [2007] ECR II-3601 ..... xxvii, xlii, xlvii, lxxv, lxxviii, c, cvi, 84, 159, 225, 242, 489, 567, 569 Case T-167/08, Microsoft v Commission, not yet decided ............................... cvi, 574 Case T-313/05, Microsoft v Commission, removed from the register on 27 November 2007 ............................................................................................ 572 Case 395/87, Ministère Public v Jean-Louis Tournier [1989] ECR 2521 ................. 146 Case T-89/98, National Association of Licensed Opencast Operators (NALOO) v Commission [2001] ECR II-515 .................................................... 684 Case 322/81, NV Nederlandsche Banden-Industrie Michelin v Commission (“Michelin I”) [1983] ECR 3461 .......................xxxv–xxxvi, 70–1, 111, 141, 174–6, 214, 264, 295, 319, 321 357, 550 Joined Cases T-213/01 and T-214/01, Österreichische Postsparkasse AG and Bank für Arbeit und Wirtschaft AG v Commission [2006] ECR II-1601 ............. xliv Case T-328/03, O2 (Germany) GmbH & Co. OHG v Commission [2006] ECR II-1231 ..................................................................................................... 382 Case 24/67, Parke, Davis and Co v Probel, Reese, Beintema-Interpharm and Centrafarm [1968] ECR 55 ........................................................................ 346, 622 Case C-163/99, Portugal v Commission [2001] ECR I-2613 ................................... 346 T-69/89, Radio Telefis Eireann (RTE) v Commission (“Magill”) [1991] ECR II-485; upheld on appeal: Joined Cases C-241/91 P and C-242/91 P, Radio Telefis Eireann (RTE) and Independent Television Publications Ltd (ITP) v Commission [1995] ECR I-743 ................................... xliv, cxxiii, 159, 282, 521, 705, 710, 712–13, 719 Case 42/84, Remia and Verenigde Bedrijven Nutricia v Commission [1985] 2545 ...................................................................................................... lxxix Cases T-310/01 and T-77/02, Schneider Electric SA v Commission [2002] ECR II-4201 ........................................................................................... 111 Cases C-468/06, etc., Sot. Lélos Kai Sia E.E. and Others v GlaxoSmithKline AEVE Farmakeftikon Proionton (Syfait II), not yet decided ............................... c Case 40/70, Sirena Srl v Eda Srl and others [1971] ECR 69 ........................ 346, 622–3 Case 56/65, Société Technique Minière v Maschinenbau Ulm GmbH [1966] ECR 235 ................................................................................................... 214, 382 Case 29/69, Stauder v City of Ulm [1969] ECR 419............................................. xxxiv Case C-53/03, Synetairismos Farmakopoion Aitolias & Akarnanias (Syfait) and Others v GlaxoSmithKline plc and GlaxoSmithKline AEVE [2005] ECR I-4609 ............................................................................. lv, lxxvii, cxxvi, 350 Cases T-5/02 and T-80/02, Tetra Laval BV v Commission [2002] ECR II-4381, upheld: Cases C-12/03 P and C-13/03 P, Commission v Tetra Laval [2005] ECR I-987 ......................................................................... 111, 290, 315–16 Case C-333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951 ................................................................................. 111, 151, 154, 717 Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755 ........................................................................................ 151, 346, 679 Case T-51/89, Tetra Pak Rausing v Commission [1990] ECR II-309 ... lxxxiii, 174, 182 Case T-504/93, Tiercé Ladbroke SA v Commission [1997] ECR II-923 .......... 706, 711 Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207 ............... cxvi, 70, 74, 145, 346, 479, 522, 535, 550–1, 568, 588, 599, 603, 623–4, 626–8, 637, 641, 643, 683–6, 688, 693, 695

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xix

Table of Cases

xix

Case T-65/98, Van den Bergh Foods v Commission [2003] ECR II-4653 ............................................................................... 174, 264, 322, 331 Cases 43/82 and 63/82, VBVB and VBBB v Commission [1984] ECR 19 ............. lxxix Case 26/62, Van Gend en Loos [1963] ECR 3 ........................................................ lxxi Case C-73/95 P, Viho Europe BV v commission [1994] ECR I-5457 .................... lxviii Case 238/87, AB Volvo v Erik Veng (UK) Ltd [1988] ECR 6211 ......................................................... xxxviii, xlvii, cxviii, 356, 591, 643–4, 713 Case C-309/99 Wouters and Others v Algemene Raad van de Nederlandse Orde van Advocaten [2002] ECR I-1577 .......................................................... cxxvi

II. Decisions of the European Commission Commission Decision, 98/153 of 11 June 1998, Alpha Flight Services/Aéroports de Paris [1998] OJ L230/10, upheld: Case T-128/98, Aéroports de Paris v Commission [2000] ECR II-3929, upheld on further appeal: Case C-82/01 P, Aéroports de Paris v Commission [2002] ECR I-9297 ........................................ 701 Commission Decision of 12 October 2005, Amer/Salomon, Case No COMP/M.3765 .................................................................................. 410 British Gypsum (‘Super Stockist Scheme’), Case No IV / 32.929, [1992] OJ C321/9 (Article 19(3) Notice) ...................................................................... 321 Commission Decision 84/379 of 2 July 1984, British Leyland [1984] OJ L207/11, upheld: Case 226/84, British Leyland Plc v Commission [1986] ECR 3263 .............................................. 145–6, 499, 521, 625, 643, 684, 688 Commission Decision 95/364 of 28 June 1995, Brussels National Airport [1995] OJ L216/8 ............................................................................................... 701 Commission Decision 93/82 of 23 December 1992, Cewal, Cowac and Ukwal [1993] OJ L34/20, upheld: Joined Cases T-24/93, T-25/93, T-26/93, and T-28/93, Compagnie Maritime Belge Transports SA and Others v Commission [1996] ECR II-1201, upheld on further appeal: Joined Cases C-395/96 P and C-396/96 P, Compagnie Maritime Belge Transports SA and Others v Commission [2000] ECR I-1365 ................................................................. 332, 345 Commission Decision 76/353 of 17 December 1975, Chiquita [1976] OJ L95/1; upheld in part: Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207 ............................. 145–6, 480, 653 Commission Decision of 22 June 2005, Coca-Cola, Case COMP/39.116 ...... 321, 380 Commission Press Release IP/90/7 of 9 January 1990, Coca-Cola Export ............ 321 Commission Decision 2001/892 of 25 July 2001, Deutsche Post AG (“Deutsche Post II”) [2001] OJ L331/40 ................................. 145–6, 154, 568, 627 Commission Decision 2003/707 of 21 May 2003, Deutsche Telekom AG [2003] OJ L263/9, upheld: Case T-271/03, Deutsche Telekom AG v Commission, judgment of the Court of First Instance of 10 April 2008, not yet reported ......................................................................................... 479, 520 Commission Decision 85/609 of 14 December 1985, ECS/AKZO [1985] OJ L374/1; upheld: Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359 ............................................................................... 71, 174, 717 Commission Decision of 25 March 2004, Euromax v IMAX, Case COMP/37.761 ............................................................................................... 625–6

(A) Ehlermann 08 Prelims

xx

10/11/08

13:01

Page xx

Table of Cases

Eurofix-Bauco v Hilti, 1988 OJ L65/19, upheld: Case T-30/89, Hilti AG v Commission [1991] ECR II-1439, upheld on further appeal: Case C-53/92, [1994] ECR I-667 .............................................................................................. 346 Commission Decision of 21 January 2004, GE/Amersham, Case No COMP/M.3304 ........................................................................................... 315 Commission Decision of 15 June 2005, Case No. COMP/37.507, Generics/AstraZeneca [2006] OJ L332/24; on appeal: Case T-321/05, AstraZeneca v Commission, not yet decided .............................................. 377, 380 Commission Decision 2004/134 of 3 July 2001 (Case No COMP/M.2220 – General Electric/Honeywell) [2001] OJ L48/1, appeal dismissed: Case T-210/01, General Electric v Commission [2005] ECR II-5575 .................... 34 Commission Decision 75/75 of 19 December 1974, General Motors [1975] OJ L29/14, upheld: Case 26/75, General Motors Continental NV v Commission [1975] ECR 1367 ................................ 74, 145, 568, 623, 625, 653, 684 Commission Decision 94/210 of 29 March 1994, HOV-SVZ/MCN [1994] OJ L104/34 ....................................................................................................... 701 Commission Decision 78/68 of 8 December 1977, Hugin/Liptons [1978] OJ L22/23 ..................................................................................................... 566–8 Commission Decision 1999/98 of 10 February 1999, Ilmailulaitos/Luftfarsverket [1999] OJ L69/24................................................................................................ 701 Commission Decision 97/624 of 14 May 1997, Irish Sugar plc [1997] OJ L258/1 .. 155 Commission Decision of 22 March 2006, Joint selling of the media rights to the FA Premier League, Case COMP/38.173 .................................................... 284 Commission Decision 89/205 of 21 December 1988, Magill TV Guide/ITP, BBC and RTE [1989] OJ L78/43, upheld: T-69/89, Radio Telefis Eireann (RTE) v Commission [1991] ECR II-485; upheld on further appeal: Joined Cases C-241/91 P and C-242/91 P, Radio Telefis Eireann (RTE) and Independent Television Publications Ltd (ITP) v Commission [1995] ECR I-743 ........... 356, 567 Commission Decision 90/363 of 26 June 1990, Metaleurop SA, Case IV/32.846 ... 499 Commission’s Decision of 24 March 2004, Microsoft, Case COMP/37.792 (summary published in [2007] OJ L32/23), upheld: Case T-201/04, Microsoft v Commission [2007] ECR II-3601 ....................... xxxii, 112, 113, 114, 159, 341, 354, 356, 362, 380–1, 385, 417, 419, 600, 614, 712, 720 Commission Decision 2002/165 of 3 July 2001, NDC Health/IMS Health: Interim Measures [2002] OJ L59/18, withdrawn: Commission Decision 2003/741 of 13 August 2003, NDC Health/IMS Health: Interim Measures, [2003] OJ L268/69 ..................................................................................... 567, 705 Commission Decision 1999/199 of 10 February 1999, Portuguese Airports [1999] OJ L69/31 ............................................................................................... 701 Commission Decision of 29 March 2006, Prokent/Tomra, Case COMP/38.113 ........................................................................................... 325, 332 Commission Decision of 23 July 2004, Scandlines Sverige v Port of Helsingborg, Case COMP/36.568 ......................................................................... 74, 625–6, 684 Commission Decision 2000/75 of 14 July 1999, Virgin/British Airways [2000] OJ L30/1, upheld: Case T-219/99, British Airways v Commission [2003] ECR II-5917; upheld on further appeal: Case C-95/04 P, [2007] ECR I-2331 .......................xxxviii–xxxix, xliv, lxvii, xcix, 8, 11, 21, 31, 57, 71, 111, 174–6, 187, 197, 203, 291, 296, 298–300, 308, 315, 321–2, 324, 332, 385

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xxi

Table of Cases

xxi

Commission Decision of 4 July 2007, Wanadoo España/Telefónica, Case COMP/38.784; on appeal: Case T-336/07, Telefónica SA and Telefónica de España, not yet decided ..................................................................... 479, 520–1 Wanadoo Interactive, Case No. COMP/38.233; upheld on appeal: Case T-340/03, France Télécom v Commission [2007] ECR II-107; on further appeal: Case C-202/07, not yet decided ............................. 8, 57, 84, 88, 242, 265–6 Commission Decision 2000/12 of 20 July 1999, 1998 Football World Cup, [2000] OJ L5/55 ................................................................................................. 533

III. US cases (courts and the Agencies) A.A. Poultry Farms, Inc. v. Rose Acre Farms, 881 F.2d 1396 (7th Cir. 1989) ........ 180 Appalachian Coals, Inc. v. United States, 288 U.S. 344 (1933) ................................ liii Aquatherm Industries, Inc, v. Florida Power & Light Co., 145 F.3d 1258 (11th Cir. 1998) ................................................................................................. 706 Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985) ....... 158, 335 Biovail Corp., No. C-4060 (F.T.C. 23 April 2002) ................................................ 578 Bristol-Myers Squibb Co., No. C-4076 (F.T.C., 18 April 2003) ..................... 492, 578 Broadcom Corp. v. Qualcomm Inc., No. 06-4292 (3d Cir. Sept. 4, 2007), 2007 U.S. App. LEXIS 21092 ....................................................................... 602–3 Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993) ........................................................................ xlvi, 149–50, 152–4, 213, 240 Brown Shoe Co. v. United States, 370 U.S. 294 (1962) ...................................... 261–2 Caribbean Broadcasting System, Ltd. v. Cable & Wireless plc, 148 F.3d 1080 (D.C. Cir. 1998) ........................................................................................ 715 Chevron Corp., File No. 051-0125 (F.T.C. 2 August 2005) ................................... 578 City of Lafayette v. Louisiana Power & Light Co., 435 U.S. 389 (1978) ................ 548 Continental T.V., Inc. v. GTE Sylvania, Inc., 433 U.S. 36 (1977) .......................... xcix DDAVP Direct Purchaser Antitrust Litigation, No. 06-5525-cv (2d Cir.), not yet decided ................................................................................................. 578 Fashion Originators’ Guild of America v. FTC, 312 U.S. 457 (1941) ...................... 719 Eastman Kodak v. Image Technical Services, 504 U.S. 451 (1992) ........... 113–14, 504 Illinois Tool Works Inc v. Independent Ink, Inc., 547 U.S. 28 (2006) ......................... c Intergraph Corp. v. Intel Corp., 195 F.3d 1346 (Fed. Cir. 1999) ...................................................................................... cviii, 706, 711, 714–6 International Salt Co. v. United States, 332 U.S. 392 (1947) ................................. 110 Jacobellis v. Ohio, 378 U.S. 184 (1964) ............................................................ 15, 523 Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2 (1984) ........................................................................................................ c, 451–2 Langenderfer, Inc. v. S.E. Johnson Co., 917 F.2d 1413 (6th Cir. 1990) ................. lxvii Leegin Creative Leather Products, Inc. v. PSKS, Inc., 127 S. Ct. 2705 (2007) ....... 110 LePage’s Inc. v. 3M, 324 F.3d 141 (3rd Cir. 2003) ................................. 206, 225, 262 Lorain Journal Co. v. United States, 342 U.S. 143 (1951) ..................................... 436 Matsushita Elec. Industrial Co. v. Zenith Radio Corp., 475 U.S. 584 (1986) ................................................................................................. 149, 152, 158 MCI Communications Corp. v. AT&T, 708 F.2d 1081 (7th Cir. 1983) ........... 645, 712 Miller v. California, 413 U.S. 15 (1973) .................................................................. 15

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xxii

xxii Table of Cases New York Mercantile Exchange, Inc. v. Intercontinental Exchange, Inc., 323 F. Supp. 2d 559 (S.D.N.Y. 2004) ...................................................................... 713 Otter Tail Power Co. v. United States, 410 U.S. 366 (1973) .................................. 648 Rambus, Inc., No. 9302 (F.T.C., 2 August 2006), on appeal: Rambus, Inc. v Federal Trade Commission, Docket Nos. 07-1086, 07-1124 (D.C. Circuit), not yet decided ............................................................................... 492, 578, 603–4 Rambus, Inc., No. 9302 (F.T.C., 5 February 2007) (opinion on remedy) .......................................................................................... 492, 578, 603–4 Rambus, Inc., v. Federal Trade Commission, 522 F.3d 456 (D.C. Cir. 2008) ................................................................................................ lxviii–lxix Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d 210 (D.C. Cir. 1986), cert. denied, 479 U.S. 1033 (1987) ................................................ 178–9 Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447 (1993) ............................... lxvii, 81 Standard Oil of New Jersey v. United States, 221 U.S. 1 (1911) ............................ 125 State Oil v. Khan, 522 U.S. 3, (1997) .................................................................... xcix United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940) ............................... liii United States v. United Shoe Machinery Corp., 110 F. Supp. 295 (D. Mass. 1953), affirmed per curiam, 347 U.S. 521 (1954) ..................................... 125–7, 579 United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir. 1945) ............................................................. 18, 109–10, 123, 125–6, 227, 373, 439 United States v AMR Corp., 140 F. Supp. 2d 1141 (D. Kan. 2001), affirmed, 335 F.3d 1109 (10th Cir. 2003) .............................................................. 152–3, 335 United States v. Colgate & Co., 250 U.S. 300 (1919) ............................................. 157 United States v. Dentsply Int’l, Inc., 399 F.3d 181 (3d Cir. 2005) ................... 225, 578 United States v. E.I. du Pont de Nemours & Co., 351 U.S. 337 (1956) .................... 258 United States v. Grinnell Corp., 384 U.S. 563 (1966) ..................................... 373, 439 United States v. Loew’s, Inc., 371 U.S. 38 (1962) .................................................. 394 United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001), cert. denied, 534 U.S. 952 (2001); on remand, United States v. Microsoft Corp., 231 F.Supp.2d 144 (D.D.C. 2002) .......................... 87–8, 158, 272, 341, 438, 716–8 United States v Microsoft Corp., Civ. Action No. 98-1232 (CKK) (D.D.C.) ........ 578 United States v. Monsanto Co., No. 1:07-cv-00992 (D.D.C. filed 31 May 2007) ..... 32 United States v. Oracle Corp., 331 F. Supp. 2d 1098 (N.D. Cal. 2004) ................. 111 United States v. Paramount Pictures, Inc., 334 U.S. 131 (1948) ............................ 394 United States v. Terminal Railroad Association, 224 U.S. 383 (1912) .................... 159 United States v. Topco Associates, 405 U.S. 596 (1972) ........................................ 100 United States v. Trans-Missouri Freight Ass., 166 U.S. 290 (1897) ........................ 144 United States v. Trenton Potteries Co., 273 U.S. 392 (1927) .................................. 144 United States v. Von’s Grocery, 384 U.S. 270 (1966) ............................................. 229 Union Oil Co. of Cal., No. 9305 (F.T.C., 4 March 2003) ............................... 492, 578 Utah Pie Co. v. Continental Baking Co., 386 U.S. 685 (1967) .......................... 149–50 Valassis Communications, Inc., No. C-4160 (F.T.C., 19 April 2006) .............. 492, 578 Virgin Atlantic Airways v. British Airways, 257 F.3d 256 (2d Cir. 2001) ............... 187 Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko LLP, 540 U.S. 398 (2004) .......................................................... xlvi, cii, 110, 144, 157–9, 225–6, 228, 335, 447, 518, 645, 681 Volvo Trucks North America, Inc. v. Reeder-Simco GMC, Inc., 546 U.S. 164 (2006) ................................................................................................. 699

(A) Ehlermann 08 Prelims

10/11/08

13:01

Page xxiii

Table of Cases

xxiii

Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., 549 U.S. 312 (2007) .................................................................... xlvi, 150, 152, 154, 225, 228

IV. Other national jurisdictions (courts and national competition authorities) Canada B-Filer Inc. v. The Bank of Nova Scotia, 2005 Comp. Trib. 31 .............................. 446 Canada (Commissioner of Competition) v. Canada Pipe, 2005 Comp. Trib. 3 ................................................................................................... 461–2, 464 Canada (Commissioner of Competition) v. United Grain Growers Ltd, 2002 Comp. Trib. 35 ......................................................................................... 449 Canada (Director of Investigation & Research) v. BBM Bureau of Measurement (1980), 60 C.P.R. (2d) 26 ........................................................................... 456, 459 Canada (Director of Investigation and Research) v. Bombardier Ltd. (1980), 53 C.P.R. (2d) 47 ............................................................................................... 455 Canada (Director of Investigation and Research) v.D&B Companies of Canada Ltd. (1996), 64 C.P.R. (3d) 216 (Comp. Trib.) ..................................... 455 Canada (Director of Investigation and Research) v. Laidlaw Waste Systems Ltd., (1992), 40 C.P.R. (3d) 289 (Comp. Trib.); ................................................. 455, 462 Canada (Director of Investigation and Research) v. NutraSweet Co. (1990), 32 C.P.R. (3d) 1 (Comp. Trib.) ......................... 374, 453, 455–7, 461, 462–4, 465–6 Canada (Director of Investigation and Research) v. Tele-Direct (Publications) Inc. (1997), 73 C.P.R. (3d) 1 (Comp. Trib.) ........................................ 452–3, 455–7 Commissioner of Competition v. Canada Pipe Ltd., 2006 FCA 233, [2007] 2 F.C.R. 3 ................................................................. 445–6, 457, 461, 464–6 Construx Engineering Corporation v. General Motors of Canada, 2005 Comp. Trib. 21 .......................................................................................... 449, 521 General Motors of Canada Ltd. v. City National Leasing Ltd., [1989] 1 S.C.R. 641 ...................................................................................................... 446 Hunter v. Southam, [1984] 2 S.C.R. 145 ................................................................ 448 Nordenfelt v. Maxim Nordenfelt Guns & Ammunition Co., [1894] A.C. 535 (U.K.H.L.) ........................................................................................ 446 Denmark Decision of the Danish Competition Council of 30 November 2005, Elsam A/S, upheld in part and annulled in part: judgment of the Danish Competition Appeal Tribunal of 14 November 2006, Elsam A/S ............................................... 590, 637 France Judgment of the Cour d’Appel de Paris (1ère chamber – Section H) of 23 January 2007, Société Pharma-Lab, S.A.R.L. v. Conseil de la Concurrence .................... 186 Germany Press Release of the Bundeskartellamt of 20 December 2006, RWE and E.ON .... 629 Judgment of the Reichsgericht of 1897, Sächsisches Holzstoffkartel (Saxon Wood Pulp Cartel) [1897] 38 RGZ 155 ....................................................................... 168 Press Release of the Bundeskartellamt of 17 April 2003, Stadwerke Mainz AG .... 589

(A) Ehlermann 08 Prelims

xxiv

10/11/08

13:01

Page xxiv

Table of Cases

Decision of the Bundeskartellamt of 21 March 2003, Stadwerke Mainz AG .................................................................................................. 635, 646 Judgment of the Bundesgerichtshof of 7 February 2006, Strom II plus, KZR 8/05 .................................................................................................. 635, 646 Italy AGCM Decision 16250 of 20 December 2006, Case A366 – Comportamenti restrittivi sulla borsa elettrica, Boll. 49/2006. .......................................... 638, 647–8 The Netherlands Press release 04-22 of the NMa of 30 September 2004, Essent .............................. 635 Judgment of 30 November 2006 of the Trade and Industry Appeals Tribunal (CBb), GTS v. NMa .......................................................................................... 631 Spain Judgment of the Audiencia Nacional of 26 January 2005, Laboratorios Farmacéuticos ................................................................................................... 186 Judgment of the Tribunal de Defensa de la Competencia of 7 July 2004, Empresas eléctricas ........................................................................................... 636 Resolution 602/05 of the Comisión Nacional de la Competencia of 28 December 2006, Viesgo Generación .................................................................. 636 United Kingdom Judgment of 23 June 2003 of the Competition Appeals Tribunal, Aberdeen Journals v Director General of Fair Trading, Case No. 1009/1/1/02, [2003] CAT 11 .................................................................................................. 300 Judgment of 18 December 2006 of the Competition Appeals Tribunal, Albion Water Limited v Water Services Regulation Authority, Case No. 1046/2/4/04, [2006] CAT 36 .................................................................................................. 534 Judgment of 2 February 2007 of the England and Wales Court of Appeal (Civil Division), Attheraces (UK) Limited v. The British Horseracing Board Limited, Case No. A3/2006/0126, [2007] EWCA Civ 38 ................................................. 536 Judgment of 11 March 2004 of the Competition Appeal Tribunal, Genzyme Ltd v. Office of Fair Trading, Case No. 1016/1/1/03, [2004] CAT 4 ............ 408, 418 OFT Decision of 11 May 2000, LINK Interchange Network Ltd, Case No. CP/0642/00 ................................................................................................ 540 OFT Decision of 30 March 2001, Napp Pharmaceutical Holdings Ltd and Subsidiaries (Napp), Case No. CA98/2/2001, upheld on appeal: Competition Appeals Tribunal, judgment of 15 January 2002, Napp Pharmaceuticals Holdings Limited and Subsidiaries and Director General of Fair Trading, Case No. 1001/1/1/01 ......................................... civ, 541, 588, 627–8, 629, 634, 640

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxv

Introduction, Summary, Remarks

Mel Marquis* This volume contains a collection of written contributions prepared for the 12th edition of the Annual EU Competition Law and Policy Workshop, held on 8–9 June 2007 at the Robert Schuman Centre for Advanced Studies of the European University Institute in Florence. It also includes edited transcripts of several rounds of debate held on that occasion. Finally, it features a series of conclusions drawn from the debates by Philip Lowe, Director General of the European Commission’s Directorate General for Competition.1 Background. The EU Competition Law and Policy Workshop is an ongoing program designed to explore topical policy and enforcement issues in the area of EU competition law and economics. In this connection, the increasingly global character of competition law necessarily also plays an important role. Each year, in the spirit of intellectual pluralism, the Workshop brings together a group of top-level EU and international policy makers, judges, legal practitioners, economic experts and scholars to take part in an intensive roundtable debate regarding specific competition-related issues in an informal and non-commercial environment. The 12th edition of the Workshop was devoted to A Reformed Approach to Article 82 EC.2 As most competition specialists around the world know, Article 82 of the EC Treaty (which would take on the new numerical identity of “Article 102 TFEU” if the Treaty of Lisbon were to be ratified by all EU Member States3) is a form of monopoly control. It prohibits an abuse by one * European University Institute, Florence and University of Verona. Claus Ehlermann, who shall remain blameless, kindly reviewed a draft of this chapter. For hospitality and kindness, I am also deeply grateful to everyone at the Istituto di Diritto Internazionale e dell’Unione Europea, University of Macerata. 1 See pp. 723 et seq. 2 For information concerning previous editions of the Workshop and the corresponding European Competition Law Annuals series, see the website of the Robert Schuman Centre: www.eui.eu/RSCAS/Research/Competition/. The written contributions prepared for recent editions of the Workshop may be downloaded freely from that website. 3 See the Treaty of Lisbon, signed on 13 December 2007, [2007] OJ C306/01. The consolidated versions of the Treaties, with the Lisbon amendments, were published in [2008] OJ C115/01. The text of Article 102 of Treaty on the Functioning of the European Union (“TFEU”) preserves the text of Article 82 EC nearly verbatim with one minor touch-up: the rather archaic reference to incompatibility with the “common market” would be reformed to make it clear that the concept is one of incompatibility with the “internal market”. This is an update of no substantive importance. By contrast, what cannot be predicted as of this writing is whether any legal significance will be

(AA) Ehlermann 08 Introduction

xxvi

12/11/08

10:25

Page xxvi

Introduction, Summary, Remarks

or more undertakings of a dominant position if the abuse is capable of affecting trade between the Member States. When one first encounters Article 82 and reads it in the light of Article 81 (the latter provision establishing, in principle, a prohibition against restrictive agreements capable of affecting trade between Member States), it is quite striking to see that, not only does Article 82 lack an exemption in the style of its sister provision (an asymmetry at the heart of academic debates concerning the concepts of “objective justification” and “objective economic justification”), but the very language it contains— an “abuse” as compared to the “prevention, restriction or distortion of competition”—seems to envisage a paradigm of conduct beyond redemption. To the extent that Article 82 creates that impression, its drafting is regrettable because it is not sufficiently agnostic. As repeatedly highlighted by many participants of the Workshop, unilateral exclusionary conduct is rarely devoid of at least some redeeming qualities; furthermore, the line between socially desirable competitive vigour and genuinely anticompetitive exclusion can invite subjective judgment calls and speculation about the fate of the relevant market. Therefore, while unilateral exclusionary conduct certainly has the potential to harm competition and consumers, and while it should accordingly be condemned in some cases, it is doubtful that such anticompetitive behaviour should draw unambiguous reproach comparable to the stigma more appropriately reserved for hard core cartels.4 This should be borne in mind even though the “abuse” tag is a deeply entrenched legal and political fact that will not change anytime soon. Looking back on the history of monopoly control in Europe, relatively little attention was given to the potentially benign aspects of aggressive commercial conduct pursued by a dominant firm. In this field, the trajectories determined by a series of landmark judgments of the European Court of Justice in the 1970s have led to a fairly wide gulf between the traditional legal practice and what some observers today would consider to be enlightened monopoly policy.5 By the early years of the present century, this dissatisfacattached to the fact that the long-cherished Article 3(1)(g) EC would be consigned to the hindquarters of the Treaties, essentially to create the illusion that a politically embarrassing “system ensuring that competition is no distorted” is not an essential condition for a properly functioning internal market. By means of legal triage and subterfuge (see Protocol on the Internal Market and Competition), the provision was rescued from the political tampering, but only time will tell whether any permanent damage has been sustained. For early speculations, see Alan Riley, “The EU Reform Treaty and the Competition Protocol: Undermining EC Competition Law”, 28 European Competition Law Review 703 (2007); Heike Schweitzer, “Competition Law and Public Policy: Reconsidering an Uneasy Relationship. The Example of Art. 81”, EUI Working Paper LAW 2007/30, available at http://hdl.handle.net/1814/7623, pp. 14–15. 4 The importance of distinguishing between cartel behaviour and unilateral exclusionary conduct was emphasized in particular by Cal Goldman in his oral remarks. See infra note 105. The distinction was also noted, e.g., by David Gerber in his written contribution. See pp. 48–49. 5 For a comprehensive analysis and a rich, original approach to the ECJ’s jurisprudence, see Ekaterina Rousseva, Exclusionary Abuses in EC Competition Law: Rethinking Article 82 of the EC Treaty, forthcoming, Hart Publishing, 2009.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxvii

Introduction, Summary, Remarks

xxvii

tion with the approach to Article 82 was obvious, all the more so because other areas of EU competition policy had either been substantially reformed or were due to be. The debate with respect to Article 82 began in earnest at the 2003 edition of this Workshop, which concerned the fundamental question: What is an Abuse of a Dominant Position?.6 Soon thereafter the Commission formally announced a comprehensive review of its policy in this area, beginning with exclusionary conduct. And since then, much intellectual capital has been spent on the enterprise, both within the Commission’s walls and in public debate. The ongoing reflection on a reformed approach to Article 82 has been quite valuable, as all interested observers have been forced to sharpen their understanding of the subject and to re-examine received wisdom with a keener critical eye. Nevertheless, while the process has led to an outpouring of fresh ideas and to a clearer understanding of the economics of commercial behaviour, at the present juncture Article 82 remains unfinished business.7 It was the conscious aim of the 2007 Workshop to help complete this round of reform. The debates were held, with the participation of key policy makers, in the absence of any official published guidance by the Commission on a new approach to Article 82, and indeed even without any draft guidelines.8 Consequently, the discussion was able to flow freely without reference to any confines that otherwise could potentially have been set down by words on paper. One reason commonly cited during 2007 for the need to wait before launching a draft set of guidelines on the application of Article 82 was the fact that a certain cause célèbre was pending before the Court of First Instance, and it was rightly anticipated that the CFI would have some important things to say in its judgment. Curiously, however, even after the Microsoft judgment was handed down on 17 September 2007,9 and even though no appeal was filed, as of this writing (spring 2008) no draft guidelines have emerged. In that light, the concerns and positions advocated by the lawyers and economists taking part in these debates may prove particularly useful because 6 See Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006. 7 These loose ends continue to draw sharp criticism. For example, according to Akman, “the EC Commission either has to demonstrate in its decisions unambiguously that it follows a welldefined ‘consumer welfare’ approach or it should unambiguously declare what approach it follows in practice. Currently, Article 82 EC remains an area of law with more questions than answers.” Pinar Akman, “‘Consumer Welfare’ and Article 82 EC: Practice and Rhetoric”, CCP Working Paper 08-25, available at http://www.ccp.uea.ac.uk/publicfiles/workingpapers/CCP0825.pdf, p. 26. 8 One important document that had already been released for public comment approximately eighteen months prior to the Workshop was DG Competition’s Discussion Paper on exclusionary conduct, available at http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005. pdf. Both at the Workshop and more generally, the Discussion Paper was helpful and stimulated debate on numerous open issues. 9 Case T-201/04, Microsoft v Commission, not yet reported.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxviii

xxviii Introduction, Summary, Remarks they shed a good deal of light on the modern, effects-based approach to Article 82 as regards both exclusionary conduct and exploitative conduct. While there was no total agreement among the participants regarding the future direction of law and policy under Article 82, and while alternative views should never be dismissed merely because they are unorthodox or perceived to be démodées, some degree of consensus or majority view did emerge on certain points. Most importantly, a majority of participants held the strong conviction that it is high time for the law to move on in this area. It is time, in other words, to focus on coherent theories of harm supported by concrete market analysis and (actual or likely) concrete effects.10 “Effects”, in turn, should be evaluated by reference to how consumers are likely to be affected, although the term consumer welfare is often ambiguous and requires further clarification since, for example, consumers are not a homogeneous class with homogeneous interests.11 It also appears that we are entering into an era in which dominant companies will increasingly seek to justify exclusionary behaviour on the basis of efficiencies, and this too seems consonant with an effects-based approach. Furthermore, while the debate on Article 82 in recent years has often dwelled on the concepts of abuse and efficiencies, it 10 Note that a form-based approach need not imply amnesia with respect to the role the competition rules play in the process of European economic integration. The legal significance of market integration is clear from the text of the EC Treaty and from the controlling jurisprudence. While it is sometimes suggested that the integration objective is incompatible with a modern conception of competition policy, and even that it is obsolete, the better view is that the gradual interpenetration national and regional markets is conducive to an effect-based approach to competition, that integration and competition are symbiotic and can yield important, welfareenhancing synergies, and that occasional conflicts between the two goals can be reconciled, at least in the broader perspective. See infra notes 56–57 and accompanying text. The political significance of market integration, though of less relevance here, is as plain as ever. See, e.g., “Facing the Challenge: the Lisbon strategy for growth and employment”, Report from the High Level Group (the so-called “Kok Report”), November 2004, available at http://ec.europa.eu/growthandjobs/pdf/kok_report_ en.pdf, p. 25. Finally, while a “pure competition approach” (the meaning of which, however, remains contested) is generally desirable in the pursuit of competition policy, it is doubtful that the goals enunciated in Article 2 EC—which (while omitting industrial policy) includes harmonious, balanced and sustainable development of economic activities, sustainable growth, competitiveness, etc.—could be simply disregarded (see infra p. lviii). 11 On the occasionally diverging needs and interests of consumers, see Francis McGowan, “State Monopoly Liberalization and the Consumer”, in Damien Geradin, ed., The Liberalization of State Monopolies in the European Union and Beyond, Kluwer, 1999, pp. 207 et seq. Indeed, even the interests of a single consumer may be contradictory, depending on the time horizon used. However, it is difficult to account for these kinds of (horizontal or longitudinal) heterogeneity in the abstract when defining competition policy. The EU opts for a more standard approach, whereby “consumers” are treated as a composite group that includes both intermediate traders and final consumers. See Guidelines on the application of Article 81(3), [2004] OJ C101/97, para. 84; Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings, available at http://ec.europa.eu/comm/competition/mergers/legislation/nonhorizontalguidelines.pdf, para. 16. The EC Merger Regulation specifies that the interests of both intermediate and final consumers must be taken into account by the Commission when evaluating whether a concentration is compatible with the common market. See Article 2(1)(b) of Council Regulation 139/2004 on the control of concentrations between undertakings, [2004] OJ L24/1.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxix

Introduction, Summary, Remarks

xxix

would be a mistake to neglect the need to work toward robust, sober assessments of dominance (defined as substantial, insulated, durable power over price).12 When the accuracy of a finding of dominance is genuinely in doubt, e.g., due to different plausible views concerning market definition or price constraints, it may be appropriate to modulate the analysis of the alleged abuse to reflect that uncertainty. Finally, an effects-based competition approach is relevant to enforcement priorities. For example, the performance of certain sectors can have rippling effects across the economy as a whole, and may thus require more urgent attention than others where the enforcer’s resources are constrained. At the same time, with regard to unilateral conduct by dominant firms, and for all the appeal of a sophisticated economic analysis of the kind that intrigues many competition specialists, several of the Workshop participants also stressed the need for an approach tempered by the requirements of clarity and workability. The challenge for enforcement policy in this area is to find the right mixture of precision and pragmatism. Structure of the Workshop. The debates held at the 2007 Workshop proceeded in three main sessions, namely: (1) A Broad Comparison of Monopolization under Section 2 of the Sherman Act and Abuse of a Dominant Position under Article 82; (2) A Reformed Approach to Exclusionary Conduct; and (3) Exploitative Conduct and the Remedies. In the remainder of this chapter, an overview is provided of the main points discussed, first in the written contributions and then in the course of the related roundtable debates. Reflections and critical remarks on certain specific points are also included from time to time. The chapter is not designed to be read in one sitting; it is meant to be a reference to assist the reader, and to serve as a companion to the individual papers and panel discussions. Note, however, that for reasons of space, individual oral presentations are not summarized in this chapter. The reader is referred to the various Panel discussions, where all presentations are reproduced verbatim.

12 In addition to discussions of dominance in the written contributions of Fox, Vickers and Bishop in this volume, see also the Unilateral Conduct Working Group (UCWG) of the International Competition Network, “Dominance/Substantial Market Power Analysis Pursuant to Unilateral Conduct Laws—Recommended Practices”, available at http://www. internationalcompetitionnetwork.org/media/library/unilateral_conduct/Unilateral_WG_1.pdf.

(AA) Ehlermann 08 Introduction

xxx

12/11/08

10:25

Page xxx

Introduction, Summary, Remarks

1. Session One: A Broad Comparison of Monopolization under Section 2 of the Sherman Act and Abuse of a Dominant Position under Article 82 1.1 Panel I (papers by Gerber, Ahlborn & Padilla, Zimmer, Fox, Schweitzer, Venit) A. Written contributions This first session focused above all on the objectives of competition policy, and in particular those animating Article 82. It should therefore be seen in some respects as a reprise of the 1997 Workshop, also dedicated to the objectives of competition law.13 As is well known, the objectives (purportedly) served by European competition policy have run a wide gamut, and they have evolved significantly over time.14 Even in the short span of the last ten years, much has changed in Europe in terms of emphasis and understanding.15 When considered in parallel with the evolution of US antitrust law since the 1970s, one finds mixed elements of convergence and mutual learning (though the East wind tends to dominate), mutual apprehension, a few misapprehensions and, occasionally, philosophical and institutional incommensurables. The EU and US systems are thus rich in raw materials for comparison, contrast and refraction in relation to both competition law’s objectives and, in the context of monopoly control, the range of attitudes and norms concerning market power. As will be seen, the papers and oral debate touched on a variety of related themes as well, sometimes but not necessarily seen from a comparative perspective. 1. The title of the first paper, contributed by David Gerber, is The Future of Article 82: Dissecting the Conflict. This paper, which focuses on the goals of Article 82 and its role in the context of the EC Treaty and the European construction, sounds a clear note of dissatisfaction with the state of the reform debate. According to Gerber, the debate has been surrounded by confusion and misconceived notions about what Article 82 stands for and how it should function. What is clear is that two contrasting visions have emerged: a more 13 See Claus-Dieter Ehlermann and Laraine Laudati, European Competition Law Annual 1997: The Objectives of Competition Policy, Hart Publishing, 1998. 14 See, e.g., Massimo Motta, Competition Policy: Theory and Practice, Cambridge University Press, 2004, pp. 17 et seq.; François Souty, Le droit de la concurrence de l’Union européenne, 2ème édition, Montchrestien, 1999, pp. 23 et seq. For comparisons with the US experience, see Stephen Martin, “The Goals of Antitrust and Competition Policy”, in Wayne Dale Collins, ed., Issues in Competition Law and Economics, Volume I, American Bar Association, 2008, pp. 19 et seq.; Giuliano Amato, Antitrust and the Bounds of Power, Hart Publishing, 1997. 15 This was noted during the 2007 proceedings by Claus Ehlermann.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxi

Introduction, Summary, Remarks

xxxi

traditional “competitive distortion” model; and the “economic turn” of the past decade, largely reflecting the US orthodoxy. One might loosely describe the competing visions as being driven, on the one hand, by market process theory and ordoliberal views,16 and on the other by a neoclassical conception of competition.17 An implicit question underlying the paper seems to be whether it is advisable, and whether the Commission has a legitimate mandate, to veer away from concepts established long ago in the European jurisprudence and to embrace a “consumer welfare” standard to guide the application of the EU competition rules.18 However, Gerber’s primary aim is not to champion a particular approach but to try to “dissect” the conflict and to peel away layers of what he believes to be serious misunderstandings about the goals of Article 82. Without a better understanding of those goals, he says, there can be no constructive dialogue. Gerber begins with the historical development of competition law in Europe, the narrative of which may forever bear his signature.19 He traces the origins of the competitive distortion model and then describes the architecture 16 The term “ordoliberalism” is imprecise and problematic, for at least two reasons. First, the concepts associated with it cannot be adequately conveyed in a few lines, and paper-thin generalizations abound. Unfortunately, the (quite necessary) quest for a more economically intelligent competition policy in Europe has been accompanied by a tendency in some of the literature to use the terms “ordoliberal” and “ordoliberalism” as epithets (e.g., “The judgment in such-andsuch a case is a throwback to ordoliberalism.”). It is not difficult to caricature concepts that are poorly understood. Second, although ordoliberalism is often discussed as if it were a static and homogeneous school of thought, in reality it has been more of a family of ideas, with important strands that have evolved over the last 75 years. It is a curious fact of European competition law history that what is often perceived today to represent ordoliberalism is a set of ideas that seem frozen in the period of 1933 to 1950 or 1957. Some evidence of the evolution beyond that golden age may be seen in Heike Schweitzer’s written contribution to this volume. 17 Gerber may be using the term “neoclassical”, which could have different meanings, as a kind of shorthand. The Commission’s economic thinking is said to be influenced by an “extended” structure-conduct-performance paradigm and by modern industrial economics. A fair amount of emphasis is placed on strategic behaviour and game-theoretic models. For details, see Luc Peeperkorn and Vincent Verouden, “The Economics of Competition”, in Jonathan Faull and Ali Nikpay, eds., The EC Law of Competition, 2nd edition, OUP, 2007, chapter 1. See also Damien Neven, “Competition Economics and Antirust in Europe”, in 21 Economic Policy 741 (2006); Lars-Hendrik Röller, “Economic Analysis and Competition Policy Enforcement in Europe”, in Peter van Bergeijk and Erik Kloosterhuis, Modeling European Mergers: Theory, Competition Policy and Case Studies, Edward Elgar, 2005, pp. 11 et seq. 18 After warming up to the idea over a number of years, the Court of First Instance has explicitly declared that the ultimate objective of EU competition law is the welfare of final consumers. Case T-168/01, GlaxoSmithKline Services Unlimited v Commission, [2006] ECR II-2969, on appeal: Cases C-501/06 et al., not yet decided. For details concerning the judgment and its significance, see, e.g., Richard Eccles, “Parallel Exports in the Pharmaceuticals Sector: Take Nothing for Granted”, 28 European Competition Law Review 134 (2007); Mel Marquis, “Tutela del consumatore e disciplina della concorrenza: in margine alla sentenza del Tribunale di prima istanza nel caso GlaxoSmithKline”, 12(2) Il Diritto dell’Unione Europea 369 (2007). It seems that the Court of Justice will have no choice but to pronounce on the question of how far EU competition law has swung toward a consumer welfare standard (the focus in this case being on longterm welfare and dynamic efficiency). 19 See David Gerber, Law and Competition in Twentieth Century Europe: Protecting Prometheus, OUP, 1998.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxii

xxxii Introduction, Summary, Remarks of the “goal structure” that accompanies this model. More clarity with respect to this goal structure, he contends, can help to make sense of a number of different interrelated but often underspecified legal tools and concepts of varying function such as “competition on the merits”, fairness, economic freedom20 and the imperative of the single market. Preservation of not just competition but the conditions of competition is central, in both the short and the long run. However, Gerber argues that the reform debate has been characterized by, and has even encouraged, misconceptions about the competitive distortion model. He identifies a number of such misconceptions and discusses how each of these may be dispelled. For example, he points to the claim that fairness and market integration are improper objectives under Article 82, because they do not contribute to improved economic performance. Such a claim disregards the fact that, historically, the competitive distortion model, and with it Article 82, was designed and has always been intended to yield desirable economic outcomes, particularly from the standpoint of Europe’s hobbled competitiveness following World War II.21 Another example is the claim that this provision protects competitors rather than consumers. Here Gerber points out that the competitive distortion model protects the competitive process (in the distinctive European sense 22), and this redounds to the benefit of consumers 20 Although there is a distinct European paradigm of economic freedom (which influenced, for example, the evolution of both jurisprudence and policy under Article 81), it is also interesting to compare and contrast certain (anti-Lochnerian) conceptions that emerged and had an impact in the US. See, e.g., Robert Hale, “Coercion and Distribution in a Supposedly NonCoercive State”, 38 Political Science Quarterly 470 (1923); Robert Hale, Freedom Through Law: Public Control of Private Governing Power, Columbia University Press, 1952. For a discussion of Hale’s work, including his interesting and provocative views on property rights (cf. Mark Patterson’s oral remarks concerning IP in Panel VI, pp. 600–601), see Barbara Fried, The Progressive Assault on Laissez-Faire: Robert Hale and the First Law and Economics Movement, Harvard University Press, 1998. 21 Apart from the obvious goal of harnessing Germany’s military capabilities and establishing durable peace, one of the most compelling aims behind the Treaty of Rome of 1957 was to establish a market large enough to permit major industrial firms to achieve greater productive efficiency and to try to stake out a claim on global markets which the US and the Soviet Union tended to dominate. A simple glance at the Spaak Report of 1956 and other historical evidence confirms the main preoccupations of the “framers”. See Pinar Akman, “Searching for the Long-Lost Soul of Article 82 EC”, CCP Working Paper 07-5 (March 2007). An updated version of the paper, dated December 2007, is available at http://www.ccp.uea.ac.uk/publicfiles/ workingpapers/CCP07-5.pdf. A version will also be published in the Oxford Journal of Legal Studies. See also Eleanor Fox, “Monopolization and Dominance in the United States and the European Community: Efficiency, Opportunity and Fairness”, 61 Notre Dame Law Review 981, 983–984 (1986) and references therein. 22 A version of this idea (sharing affinities with, though not quite identical to, what Gerber has in mind) is expressed by the Commission in its Decision of 24 March 2004, Case COMP/37.792, Microsoft, para. 969: “Under Community competition law an undistorted competition process constitutes a value in itself as it generates efficiencies and creates a climate conducive to innovation.” (emphasis added) It may seem unusual to say that the “competition process” is a value in itself if the reason is that it serves to promote (static and dynamic) efficiencies. But note that the word used is “value”. The Commission does not say that the competitive process is an end in itself. That seemingly small distinction may have some significance. Coming back to the model described by Gerber, the European view of the competitive process, with its emphasis on

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxiii

Introduction, Summary, Remarks

xxxiii

in the long run.23 While the maintenance of that process may sometimes also be favourable to a dominant firm’s rivals, that is merely incidental to the goal of preserving open markets in the long-term interests of consumers. The tensions between the past and the present with respect to European competition law and policy are often discussed and will surely continue to fuel a vibrant debate regardless of whether guidelines on Article 82 are ultimately adopted. What is perhaps given less explicit attention is another main argument set forth in Gerber’s paper, according to which economics may be better placed for providing factual and empirical analysis (a “fact-interpretive role”), and may not necessarily be suited for programmatic questions involving the fundamental values and direction of competition law. Gerber does not appear to reject the prospect of a normative role for economics,24 but he is concerned that the fact-interpretive role and the normative role are often conflated, and he points out that although normative tasks can be assigned to economics, such as casting the definition of “anticompetitive” in terms of consumer welfare, such an assignment requires a decision legitimated by the maintaining open markets over the long run, has been criticized precisely on the ground that it is seen as an end in itself. See Jeremy West, “Background Note”, OECD Policy Roundtable on Competition on the Merits (2005), available at http://www.oecd.org/dataoecd/7/13/35911017.pdf, p. 20. This criticism seems slightly misdirected because many commentators that are committed to a “competitive process” approach consider that process to be a necessary pre-condition for competition. For scholars holding to that view, it is competition (a facet of a functional market economy and free society) which is an end in itself, although the incidental economic benefits of competition are not ignored. See the oft-cited paper of Wernhard Möschel, “Competition Policy from an Ordo Point of View”, in Alan Peacock and Hans Wilgerodt, eds., German Neo-Liberals and the Social Market Economy, MacMillan, 1989, pp. 142 et seq. Competition, so the argument goes, generally cannot subsist where a monopolist or large oligopolists have a stranglehold on the market. To that extent, the European and US conceptions of the competitive process, though they are profoundly different, each appear to regard the competitive process as a means to an end. It seems that the more appropriate target for a critic would therefore be the claim that competition itself is an irreducible goal that cannot be regarded as subservient to, say, aggregate social welfare or consumer welfare. 23 A seldom acknowledged fact in the English-language legal literature is that ordoliberals often attached significant value to consumer benefits. See, e.g., Franz Böhm, “The Non-State (‘Natural’) Laws Inherent in a Competitive Economy”, in Stützel, et al., eds., Standard Texts on the Social Market Economy, Gustav Fischer, 1982, at pp. 107–113 (interpreting the transfer of wealth from profit-maximizing producers to consumers as a result of competitive pressures as an instance of Smith’s invisible hand at work). Indeed, consumer benefits were at the heart of the “performance competition” concept. See generally David Gerber, “Constitutionalizing the Economy: German Neo-liberalism, Competition Law and the ‘New’ Europe”, 42 American Journal of Comparative Law 25, 53 (1994). 24 The question of whether economics, even in its most austere and abstract forms, could be a neutral and objective “science”, is a separate matter on which Gerber does not seem to take a position in his paper. There are very different views concerning the Weberian distinction between positive and normative social science. Whereas some social scientists seem to take the distinction for granted, scholars from other traditions have argued vigorously that the notion of positive, value-free social science is illusory. See, e.g., Hilary Putnam, The Collapse of the Fact/value Dichotomy and Other Essays, Harvard University Press, 2002; Gunnar Myrdal, Objectivity in Social Research, Duckworth, 1970. Weber’s original elaboration (inspired by Hume) is set forth in The Methodology of the Social Sciences, Free Press, 1949 (originally published in German as Gesemmelte Aufsätze zur Wissenschaftslehre, Mohr, 1923).

(AA) Ehlermann 08 Introduction

xxxiv

12/11/08

10:25

Page xxxiv

Introduction, Summary, Remarks

legal-political system. He further amplifies his discussion by stressing that the adoption of a consumer welfare model would entail a fundamental rupture with the competitive distortion model, since the emphasis would be on outcomes rather than process. He also highlights the tension between a scientific, economics-based approach and the principle of stare decisis,25 although the adhesion of the Court of Justice to its past precedents is not absolute.26 This tension deserves careful thought, as does the legitimacy question, considering that, with respect to at least Article 81, the consumer welfare model already seems to have arrived.27 2. The contribution of Jorge Padilla and co-author Christian Ahlborn is entitled From Fairness to Welfare: Implications for the Assessment of Unilateral Conduct under EC Competition Law. This paper presents an extensive legal and economic discussion of Article 82 and proposes two alternative, elabor25 The EC Treaty does not indicate, and the Court of Justice has never acknowledged, that the Court is bound by the principle of stare decisis. Indeed, the Court has been known to deviate occasionally from its past practice, whether explicitly or more subtly (see following footnote). Broadly speaking, however, precedents tend to have rather strong de facto lock-in effects. The overall picture thus does not seem to differ substantially from traditions that adhere to an explicit stare decisis doctrine. 26 See the Opinion of Advocate General Maduro in Joined Cases C-94/04 and C-202/04, Cipolla and Macrino [2006] ECR I-11421, paras. 28–29 (quoted infra note 198). As the Advocate General notes, the ECJ has always been circumspect about reversing its own case law, but there are times when the Court does decide to change direction. Some of the famous older examples include Case 29/69, Stauder v City of Ulm; Case C-267/91, Keck and Mithouard, para. 14; and Case C-10/89, Hag II, para. 10. For a recent example, see Case C-127/08, Metock and Others v Minister for Justice, Equality and Law Reform, judgment of the ECJ of 25 July 2008, not yet reported, para. 58 (holding that it was necessary to “reconsider” the Court’s earlier judgment in Case C-109/01, Akrich, where the Court had indicated that third-country nationals married to EU citizens could not benefit from Regulation 1612/68 unless they had previously been legally resident in another Member State). Of course, many courts (the ECJ being no exception), often employ an array of procedural, remedial and other similar devices to erode established substantive standards with which they may disagree, thus avoiding a frontal assault. The antitrust jurisprudence of the US Supreme Court provides several examples. See, e.g., William Page, “Legal Realism and the Shaping of Modern Antitrust”, 44 Emory Law Journal 1, 51 (1995) (“Rather than overrule established precedents, the Court has reinterpreted those precedents in light of Chicago’s positive models . . . [The Court has] altered the doctrine primarily by formulating subsidiary decisional rules that govern the application of the rules of liability, that define the types of harm that are compensable in private suits, and that determine the sufficiency of evidence to go to the jury.”). Cf. William Reynolds and Spencer Weber Waller, “Legal Process and the Past of Antitrust”, 48 Southern Methodist University Law Review 1811 (1994). For more general discussions of the legal process school, see, e.g., William Eskridge and Philip Frickey, “The Making of the Legal Process”, 107 Harvard Law Review 2031 (1994); Neil Duxbury, “Faith in Reason: The Process Tradition in American Jurisprudence”, 15 Cardozo Law Review 601 (1993). 27 This is at least the position of the Commission and of the Court of First Instance, although there seem to be differences in how each of them understands that model (for the CFI’s view, see supra note 18). While there is clearly some value in achieving consistency in the manner in which Articles 81 and 82 are conceptualized and applied, consistency alone is not a sufficient reason for a paradigm shift with respect to Article 82. If Article 82 is to be transformed, this must be justified by other advantages more specifically linked to Article 82 itself. It is the task of reformminded actors and commentators to articulate those advantages.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxv

Introduction, Summary, Remarks

xxxv

ately constructed tests by which to evaluate unilateral conduct on the basis of a “radical” (consumer) welfare standard. The authors wear their utilitarian orientation on their sleeve, assigning value to legal norms according to their aggregate impact on welfare. It is thus clear from the beginning of their paper and from its title, not to mention from the authors’ previous work, that economic welfare assumes a rank of primacy vis-à-vis the separate value of fairness. Given the ambitious scope of the paper, only a selective résumé will be presented here. After reviewing the various objectives that may have influenced EU competition policy at various times, and having presented a reasonably literate account of early ordoliberalism,28 the paper considers the “ordoliberal heritage”, i.e., the lasting impact of ordoliberalism on European concepts of dominance and abuse. That heritage is evident, claim the authors, in the Court’s venerable judgments. The 50% market-share threshold established in AKZO as being sufficient (absent exceptional circumstances) for a presumption of dominance, it is argued, is a relic of Walter Eucken’s ideal-type of “complete competition”.29 The authors further claim that Michelin’s doctrine of the “special responsibility” of a dominant firm is a direct descendant of the notorious “as if” competition standard. Although the authors seem to treat this claim as self-evident, the presumed links between the two, and their allegedly kindred purpose, remain to be substantiated or falsified.30 This will 28 What is lacking, however—from the paper but also from the English language legal literature in general—is a satisfactory discussion of how ordoliberal ideas have evolved from the 1960s to the present. See supra note 16. 29 “Complete competition” (absence of coercive power), despite some resemblance to the traditional notion of “perfect competition” (essentially, infinite elasticity of demand, as it was understood in the 1930s by Joan Robinson and others), should not be equated with it. More importantly, however, ordoliberals came to recognize that the concept of “complete competition” did not fit very well with their broader philosophical program. As Vanberg explains: “Ordnungspolitik in the Freiburg sense is foremost competition policy, a policy that aims at securing a competitive process with desirable working properties, one that works to the benefit of consumer interests. It should be mentioned, though, that while the ordo-liberals were fairly clear about the general aim that they wanted competition policy to pursue, namely to realise to the largest extent possible consumer sovereignty, some ambiguity arose in their more specific recommendations for how such a policy should proceed. Ambiguity has been caused, in particular, by the fact that, in addition to the criterion of Leistungswettbewerb (performance competition), they referred to the concept of ‘complete competition’ (vollständige Konkurrenz), a concept that, because of its outcome-oriented focus, does not square well with the general [process-oriented] logic of the Freiburg paradigm”. Viktor Vanberg, “The Freiburg School: Walter Eucken and Ordoliberalism”, Freiburg Discussion Paper 04/11 (2004), at p. 13. See further ibid., footnote 28, and references therein. Vanberg’s paper is available at: http://ies.fsv.cuni.cz/storage/sylab/138_ vanberg2004.pdf. Furthermore, as ordoliberal scholars evolved toward Hayekian assumptions about the bounded and subjective knowledge not only of market actors but also of any observer seeking to understand their behaviour, conventional equilibrium theory came to be regarded as defective, inadequate, or useless. See Hayek, Individualism and Economic Order, University of Chicago Press, 1948, pp. 33 et seq. 30 Taking a step back, it can be argued that Article 82 as written does not bear the stamp of ordoliberalism as clearly as other Treaty rules do. (These norms include, but are not limited to, Article 81(1), the fourth condition under Article 81(3), Article 87(1) and Article 86(2), although the latter provision, it seems, was originally a Benelux idea that won the immediate support of

(AA) Ehlermann 08 Introduction

xxxvi

12/11/08

10:25

Page xxxvi

Introduction, Summary, Remarks

require further archaeological digging. In any event, what is often overlooked is that the “as if” standard for dominant firms, originally advocated by Leonard Miksch and then by Eucken, was repudiated in the ordo camp itself long before the Court handed down its judgment in Michelin in 1983.31 Padilla and Ahlborn also see a connection between the “as if” standard and the fact that Article 82(a) prohibits unfair selling prices. This connection is not obvious either, however, since dominant firms are clearly not required under Article 82 to act as though they were price takers.32 The authors themselves appear to recognize this in their discussion of “competition on the merits”, a concept the authors criticize, as others have, for its circularity. Having reviewed the traditional objectives behind Article 82, and finding those objectives and the per se or quasi-per se rules they inspire to be unsatisfactory, the authors proceed by setting out a framework with which to evaluate the costs of error associated with alternative legal norms. Summarizing, the Germans and the immediate resistance of the French until the famous compromise was agreed). As Schweitzer notes in her written paper, the original exponents of ordoliberalism did not have fully elaborated views on monopoly control. Nevertheless, it seems fair to say that if the “as if” concept had prevailed, it would have been part of a model generally free of dominant firms in the first place, other than “unavoidable” monopolies such as natural monopolies. The “as if”, hypothetical price-taker standard of behaviour would then probably have governed the few monopolies left standing. (Belief in the need to regulate natural monopolies obviously does not distinguish ordoliberalism from other approaches; the position as regards IPRs is less clear, and it is doubtful that Eucken had a satisfactory solution.) Therefore, if it is correct for Padilla and Ahlborn to claim that the “special responsibility” doctrine is the offspring of the “as if” competition standard, it could presumably only be a result of that standard being superimposed by the ECJ in Michelin on a provision that seems intrinsically incoherent with the “as if” idea. My own suspicion is that the ECJ in that case was charting its own misguided course. It would not be surprising if the primary source of inspiration for the idea that a dominant firm has a special duty not to impair undistorted competition was Article 3(1)(g) read together with Article 82. For further discussion of the “as if ” standard, see the following footnote. 31 See Schweitzer, this volume, p. 122 at footnote 17; p. 135 at footnote 79 and accompanying text; and p. 19 (oral presentation). See also Manfred Streit, “Economic Order, Private Law and Public Policy: The Freiburg School of Law and Economics in Perspective”, in 148 Journal of Institutional and Theoretical Economics 675, 689 n. 24 (1992). It is implicit from Streit’s remarks in the latter article that a principal reason why “as if ” competition was ultimately rejected by ordoliberals related to the information asymmetries that present challenges for economic regulation. More fundamentally, from a Hayekian point of view, “any attempt to make a firm act ‘as if’ competition existed is simply absurd, since a discovery procedure’s results cannot be anticipated and hence [cannot be] dictated”. Manfred Streit and Michael Wohlgemuth, “The Market Economy and the State”, Diskussionsbeitrag 06-97, paper prepared for the Fifth Annual SEEPConference on Economics, Ethics and Philosophy, Marienrode, 29 October—1 November 1997, p. 16 (citing Hayek, Law, Legislation and Liberty, Vol. III: The Political Order of a Free People, University of Chicago Press, 1979, pp. 70 et seq.). 32 See van der Woude, this volume. As van der Woude notes, the ban on unfair selling prices does not preclude normal profit maximization by monopolies; its application is limited to instances of what he calls “supramonopoly” pricing. It is also useful to recall that Article 82 was drafted by the Working Group on the Common Market in a context in which many of the negotiators were counting on a large, single market to enable firms to achieve minimum efficient scale and modernize production techniques. See supra note 21. It is unlikely that, in drafting Article 82(a), the negotiators had any general preference for low price-cost margins. See also Schweitzer, this volume, p. 136 (pointing out that Article 82(a) may be understood, in historical context, as deregulatory and liberating).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxvii

Introduction, Summary, Remarks

xxxvii

Type I errors undermine welfare, they are too common and too costly (largely due to externalities and unintended consequences). Furthermore, the cost of Type I errors is greater than that of Type II errors because if anticompetitive conduct goes undetected and/or unpunished, ultimately the market is likely to self-correct, whereas if neutral or efficient conduct is punished, the damage may be irreversible. In the context of the EU, these observations spell trouble, since the Article 82 jurisprudence has tended de facto to favour per se rules of illegality. The authors explain why such rules are formalistic, harmful and fraught with social costs. Several pages are then devoted to a critique of DG Comp’s Discussion Paper of December 2005. Details aside, that critique boils down to the claim that the Discussion Paper represents more continuity than change: the Commission is still too worried about rivals’ fortunes and too wary of efficiency claims.33 Padilla and Ahlborn then proceed to critique the EAGCP’s Report of July 2005.34 While they plainly support the view that the goal driving Article 82 ought to be long-term consumer welfare, they disagree with the proposal to discard the dominance filter. They also lay out several reasons why the EAGCP’s unstructured rule of reason is in their view unsatisfactory, arguing for example that unstructured balancing entails significant transaction costs. The proposal of Padilla and Ahlborn, which builds on a classic article by then-Professor Easterbrook,35 is presented as a “third way” between the ordoliberal approach and the unstructured rule of reason. This “third way” 33 Notwithstanding their critical comments, Padilla and Ahlborn approve of paragraph 54 of the Discussion Paper, according to which “[t]he essential objective of Article 82 when analysing exclusionary conduct is the protection of competition on the market as a means to enhancing consumer welfare and of ensuring an efficient allocation of resources”. The authors interpret this language as subordinating the (European) concept of the “competitive process” to the ultimate goal of consumer welfare as reflected in low prices, quality and innovation. However, from their overall assessment of the Discussion Paper they conclude that the move toward a consumer welfare standard under Article 82 remains incomplete. Paragraph 54 of the Discussion Paper is modelled on statements in the Commission’s guidelines in other areas, such as paragraph 13 of its Article 81(3) guidelines. See [2004] OJ C101/97. Significantly, in paragraph 13 of the Article 81(3) guidelines, the Commission further links the ends-means conception described in paragraph 54 of the Discussion Paper to the requirements of an open single market (see further supra note 10). With regard to allocative efficiency, which Padilla and Ahlborn consider to be overemphasized in the Discussion Paper, it is also worth recalling that pursuant to Article 98 EC the Community and the Member States, in conducting their economic policies, “shall act in accordance with the principle of an open market economy with free competition, favouring an efficient allocation of resources”. This mandate would be preserved nearly verbatim in Article 120 TFEU (the only change being the substitution of “Union” for “Community”) if the Treaty of Lisbon were ultimately ratified by all Member States. 34 Gual, Hellwig, Perrot, Polo, Rey, Schmidt and Stenbacka, “Report by the EAGCP: ‘An economic approach to Article 82’” (July 2005), available at http://ec.europa.eu/comm/competition/ publications/studies/eagcp_july_21_05.pdf. 35 Frank Easterbrook, “The Limits of Antitrust”, 63 Texas Law Review 1 (1984). Judge Easterbrook’s approach to antitrust law was shaped to some degree while working in the 1970s for the Department of Justice under then-Solicitor General Robert Bork.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxviii

xxxviii Introduction, Summary, Remarks in fact breaks down into a “third and fourth way”, as it consists of two alternative tests to evaluate the legality of unilateral conduct. The first of these is a structured rule of reason comprising, in addition to a dominance screen, three additional obstacles to a finding of liability: (i) the presence of certain structural conditions which might vary slightly, depending on the relevant facts, and which seem aimed at ensuring that the dominant firm has both the ability and the incentive to act abusively; (ii) a coherent and falsifiable theory of harm; and (iii) a balancing of harm and efficiencies, with appropriate discounting commensurate with the degree of speculation involved. According to the authors, this structured sequence would be superior to the unstructured rule of reason because the complex task of balancing would only have to be carried out in those cases where anticompetitive effects are likely. The second alternative proposed by Padilla and Ahlborn, driven (as both alternatives are) by the goal of minimizing Type I errors, is a system of qualified per se legality. The authors give an example that draws eclectically on Volvo/Veng’s implicit “exceptional circumstances” concept and certain elements of the more structured Magill/IMS Health extensions. The authors consider the relative merits of the structured rule of reason and the qualified per se legality approach without explicitly favouring one over the other. However, they point out that a system of qualified per se legality might perform better from the standpoint of Type I errors and administrative costs. In a concluding section which served as the basis for Padilla’s oral presentation,36 the authors re-emphasize the need for predictable, ex ante rules that avoid case-by-case trekking through the selva oscura of economic theory, and through whatever unsavoury portals to which that dark wood may lead. 3. The third contribution, On Fairness and Welfare: The Objectives of Competition Policy, by Daniel Zimmer, is a comment on the papers by Gerber and by Padilla and Ahlborn. But Zimmer’s paper is less focused on the future of Article 82; it seeks instead to identify where we are today. In that regard, Zimmer points out that in fact there has already been some convergence between the legal approach under Article 82, on the one hand, and the approach under Article 81 and the Merger Regulation on the other. Consistent with the traditional approach, the ECJ’s judgment in British Airways37 seems to indicate, as Zimmer notes, that a prima facie infringement under Article 82 may be established with a blind eye to the impact of the dominant firm’s conduct on consumer welfare. This first analytical step is thus concerned not with consumers as such but with the competitive process (again, in the European sense), or what Zimmer calls the “structural pre36

See pp. 7 et seq. of this volume. Case C-95/04 P, British Airways v Commission [2007] ECR I-2331. For a detailed discussion of the Court’s judgment, see the case note by Okeoghene Odudu, 44 Common Market Law Review 1781 (2007). 37

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xxxix

Introduction, Summary, Remarks

xxxix

requisites” of competition. However, British Airways opens up at least a theoretical vista in which efficiencies that produce consumer benefits may also be taken into account and may outweigh the conduct’s anticompetitive effects.38 Zimmer observes that such a two-stage analysis would only appear to make sense if Article 82 is guided by certain impulses not strictly linked to (shortterm) consumer welfare. Zimmer then refers to a horizontal joint purchasing arrangement to illustrate his point. As he notes, joint purchasing may well fall within the scope of Article 81(1),39 even though the lower prices imposed on suppliers might translate into lower prices for end customers buying from the “cartellists”. “A demand cartel”, he says, “is prohibited regardless of whether or not it leads to consumer harm”.40 His thesis is that it is not consumers that Article 81 protects here, but suppliers. In the oral debate, Zimmer transposed this reasoning to Article 82, envisaging a monopsonistic purchasing market. During the debate, he claimed that the “abuse of a monopsonistic position cannot be explained by reference to consumer welfare”.41 One could hardly contest the proposition that Article 82, de lege lata, protects suppliers as well as purchasers. However, if consumer welfare were adopted as the central objective of the competition rules, it does not seem to follow that either horizontal purchasing agreements or monopsonies would be exempt from them. For example, a monopsonist (or a group of purchasers competing on the selling market downstream) which exploits its buyer power upstream may force purchase prices down below the competitive level by purchasing fewer quantities.42 If there is inadequate elasticity of supply, such exploitation of buyer power can be expected to harm the welfare of consumers downstream. In effect, the monopsonist (or joint buyers) exploit both suppliers and customers. This is not to argue in favour of a consumer welfare goal that excludes all others. The point is merely to show that resistance to such an exclusive goal cannot convincingly base itself on the presumption that a modernized Article 82 would be indifferent to the abusive exploitation 38

See paras. 69 and 85–86 of the ECJ’s judgment. Normally, if the combined market shares of the parties on the purchasing market and on the downstream market are each less than 15%, the Commission would consider the arrangement either as falling outside of Article 81(1) or as qualifying for an exemption under Article 81(3). Joint purchasing that involves parties with a combined market share “significantly above 15%” where the market is concentrated is likely to fall within Article 81(1) and would have to be justified under Article 81(3). See Commission Notice—Guidelines on the applicability of Article 81 of the EC Treaty to horizontal cooperation agreements, [2001] O.J. C3/2, paras. 115 et seq. and especially paras. 130–131. Furthermore, as Zimmer notes, joint purchasing may also fall outside Article 81(1) where it serves to prevent a powerful supplier from exploiting its cusomers. See Case C-250/92, Gøttrup-Klim Grovvareforeninger v Dansk Landbrugs Grovvareselskab AmbA [1994] ECR I-5641. 40 See Zimmer, p. 105. 41 See Panel V, p. 508. 42 See, e.g., Ioannis Kokkoris, “Buyer Power Assessment in Competition Law: A Boon or a Menace?”, 29 World Competition 139, 143–144 (2006). For further discussion, see infra notes 223–225 and accompanying text. 39

(AA) Ehlermann 08 Introduction

xl

12/11/08

10:25

Page xl

Introduction, Summary, Remarks

of buyer power.43 What would change would be the rationale behind the application of the provision: instead of protecting suppliers as such without any reference to consequences for consumers, consumer harm would become a necessary condition for intervention.44 Leaving that minor objection behind, Zimmer makes several additional observations, particularly concerning the relationship between fairness and welfare. First, he notes that Kaplow and Shavell, while advocating a welfare standard, define welfare in a broad manner that encompasses any advantage with a positive impact on individuals’ well-being. Perhaps, Zimmer suggests, a broader notion of consumer welfare might help to ameliorate the “communication problem” that characterizes the dialectic between the “competitive distortion” model described by Gerber and the welfare model favoured by Padilla and Ahlborn. He adds that the current conception of Article 82 has much to recommend it. Insofar as it protects individuals against unjust exploitation, it serves the “prerequisites” of competition by promoting market participation and thereby promoting the aggregate well-being of individuals. In other words, fairness—far from being opposed to welfare—may be a component of it. In any case, Zimmer concludes, it is difficult to disregard the fact that a pure consumer welfare standard would not sit comfortably beside the plurality of objectives one finds in Articles 2, 3 and 4 of the EC Treaty.45 As the ECJ now routinely recognizes, for example, the Community has “not only an economic but also a social purpose”.46 4. Whereas Europe’s past, present and future are the dominant themes of the first three papers, Eleanor Fox turns more explicitly to a comparative overview of European and US attitudes in her paper on The Market Power 43 The question of when exploitation becomes abusive if of course important, but that is a separate issue. 44 Zimmer points out that, by protecting suppliers against dominant purchasers, Article 82 stimulates market entry at the supplier level. This may be true, but to the extent that the exercise of buyer power might be harmful to consumer welfare, suppliers would continue to enjoy indirect protection, whereas currently it is consumers who enjoy indirect protection. Moreover, if a dominant purchaser also has market power downstream on the selling market but if there are no high and persistent entry barriers, new entry should effectively bid up supply prices and should lead to increased output upstream, to the benefit of both suppliers and consumers. 45 Looking ahead (perhaps too optimistically), see also Article 3(3) of the post-Lisbon TEU. 46 Case C-438/05, International Transport Workers’ Federation and Finnish Seaman’s Union v Viking Line ABP and OÜ Viking Line Eesti, judgment of 11 December 2007, not yet reported, para. 79; Case C-341/05, Laval un Partneri LTD v Svenska Byggnadsarbetareförbundet and Others, judgment of 18 December 2007, not yet reported, para. 105. It is worth noting, however, that the Court, in both Viking and Laval, is still guided by the requirements of, respectively, Article 43 and Article 49 of the Treaty. The social purpose of the Community thus has to co-exist and be reconciled with the imperatives of the fundamental freedoms. Going back several years and turning to the field of antitrust, in its Albany judgment the ECJ had already famously interpreted Article 81 in light of the Community’s social policy objectives, as expressed in, inter alia, Articles 2, 3(1)(j) and 136–143 EC. See Case C-67/96, Albany International BV v Stichting Bedrijfspensioenfonds Textielindustrie, [1999] ECR I-5751, paras. 54 et seq.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xli

Introduction, Summary, Remarks

xli

Element of Abuse of Dominance. Since both jurisdictions share a roughly similar definition of market power that focuses on the durable ability to maintain supracompetitive prices or otherwise diminish consumer welfare,47 it is all the more interesting to consider differences in methodology. As Fox notes, these differences are linked to certain “background factors” including distinct historical paths and distinct degrees of trust in the resilience of the market. These factors seem particularly significant since, even within a single legal culture, a wide spectrum of values is possible, the US Supreme Court providing just one example. Fox’s discussion of the century-long evolution of US antitrust and of the comparatively compressed shifts in approach seen in Europe suggests that there is certainly scope for each tradition to learn from the other. Consistent with much of her other work, there is a hint that the East wind referred to earlier should take the trouble to blow West sometimes. The second part of Fox’s paper concerns matters of measurement and screens. She begins by asking what is “substantial market power”,48 and this leads her into a discussion of, among other things, the difficulty of assessing market power in cases of aftermarkets, network effects and Schumpeterian competition. She then considers the role of market shares and the utility of market share thresholds in establishing safe harbours or AKZO-style presumptions of dominance. Here, Fox makes the fundamental point that market definition can make or break the entire market power analysis. If the market is well defined, then a persistently high market share can indeed be a useful prima facie shortcut. Leaving open the delicate question of “how high is high”, she takes note of the more prudent approach in Europe, as reflected in the various market share “bands” described in DG Comp’s Discussion 47 Fox points out, however, that Europeans instinctively attach somewhat more importance to variety and pluralism, which reflects a “certaine idée” of the interests of both the supply side and the demand side. See generally Amato, Antitrust and the Bounds of Power, cited supra note 14. 48 This question highlights a regrettably loose use of language in Europe and a consequent ambiguity. The problem is that “significant market power” surely can only be a necessary and not a sufficient condition for dominance under Article 82. The more emphatic “substantial market power” is better suited (provided it is further qualified by “durable” or “persistent”, etc.) to be the economic equivalent of the legal concept of dominance, and DG Competition recognized this in particular when it issued its Discussion Paper on exclusionary abuses. In the Discussion Paper, “substantial market power” is effectively treated as a synonym for dominance (see, e.g., paras. 23 et seq.). By contrast, due to an unfortunate hangover in the telecoms field, when the threshold for legitimate ex ante regulation was raised in 2002 to the level of dominance, the expression “significant market power” (SMP) was maintained. See Commission, Guidelines on market analysis and SMP in electronic communications, [2002] OJ C165/6 (but see ibid para. 30). Significant market power may indeed serve in general as a pre-condition for antitrust and/or regulatory concerns, and it might entail, e.g., the inapplicability of a block exemption in a certain case. But the term “significant” is inappropriate in the context of Article 82 because, leaving aside the aberration in telecoms, a non-dominant firm may well have significant market power. This is made clear enough, for example, in paragraph 8 of the Commission’s Non-Horizontal Merger Guidelines, cited supra note 11 (“Non-horizontal mergers pose no threat to effective competition unless the merged entity has a significant degree of market power (which does not necessarily amount to dominance) in at least one of the markets concerned.”) (emphasis added).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xlii

xlii Introduction, Summary, Remarks Paper.49 In agreement with Janusz Ordover, Fox cautions against treating the Lerner index or market shares as if, ipso facto, they answer the question of whether a firm is dominant or enjoys “SMP”. The key, as in all things, is to use the tools without becoming their tool: “The problem is not with thresholds as such but with ‘excessive reliance on mechanistic criteria’.”50 The remainder of Fox’s paper is devoted primarily to a number of points emerging from the work of the ICN’s Working Group on Unilateral Conduct with respect to recommended practices in assessing dominance and substantial market power.51 Fox’s synthesis helps to illuminate the main issues and implicitly points to institutional factors which explain some of the differences in market power assessment in the two jurisdictions discussed. One point that seems to emerge is that, the more uncertain the market definition, the less casual one can afford to be in saying that a firm controls 90% of the relevant market, or that it is “superdominant”. By the same token, an uncertain market definition calls for particular care in establishing that the market as defined is incontestable and that the defendant’s position is indeed unshakable. If the law is going to cling to anachronisms such as, in particular, the “special responsibility” doctrine, scrupulous respect for such an enhanced duty of care is required.52 49 Here Fox quotes paragraph 31 of the Discussion Paper. That paragraph does not seem to be drafted very carefully. As John Vickers noted during the oral debate, it “hits the wrong note” and may invite unnecessary uncertainty. See Vickers, p. 198. 50 Fox, p. 115 (quoting Ordover). 51 See supra note 12. 52 Heike Schweitzer points out that, although the “special responsibility” doctrine has caused confusion, recent case law has clarified that it only means the same type of behaviour might be accepted if pursued by a non-dominant firm but condemned if adopted by a dominant firm. See Schweitzer, this volume, p. 141, footnote 100 and accompanying text (citing the CFI’s 2003 judgment in Atlantic Container Lines). This clarification is indeed helpful; however, it may also underscore the undesirability of retaining such a “doctrine” if it is not a doctrine at all. Otherwise, there is a risk that the concept will continue to induce mischief and errant conjecture. See, e.g., Stephen Wisking, Craig Pouncey and Jacques Buhart, “European Union: Competition—Refusal to Supply”, 30(1) European Intellectual Property Review N1, at N3 (2008) (suggesting that it was the “special responsibility” that drove the Commission to impose a particularly heavy fine on Microsoft). See also Dieter De Smet, “The Diametrically Opposed Principles of US and EU Antitrust Policy”, 29(6) European Competition Law Review 356, 358–359 (2008) (taking the “special responsibility” doctrine to be a manifestation of the principle of fair competition, and recalling the reference to fair competition in the Preamble of the EC Treaty; it is therefore claimed that the “special responsibility” doctrine provides protection for small and medium-size enterprises). According to De Smet, “[i]t was foreseeable that once Microsoft was found to have a dominant position . . . that its ‘special responsibility’— coupled with a likely foreclosure of the media player market—would lead to an abuse of dominance”. In short, the “special responsibility” doctrine is confusing, counterproductive and should be shelved. Failing that solution, the doctrine at least should not lead to deus ex machina dénouements in Article 82 cases. Fortunately, in the Microsoft case referred to above, the CFI did seem to keep its references to this doctrine to a minimum. See David Howarth and Kathryn McMahon, “‘Windows has performed an illegal operation’: The Court of First Instance’s judgment in Microsoft v Commission”, 29(2) European Competition Law Review 117, 130–131 (2008). Better yet, the expression was avoided entirely in British Airways, both by the Court and by the Advocate General. For further discussion, see Rafael Allendesalazar, this volume, pp. 319 et seq.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xliii

Introduction, Summary, Remarks

xliii

5. Like the paper submitted by Jorge Padilla and Christian Ahlborn, the contribution provided by Heike Schweitzer—The History, Interpretation and Underlying Principles of Section 2 Sherman Act and Article 82 EC—is exceptionally rich in scope and depth. As it is worth lingering for a while on her paper, this summary is followed by some additional reflections. In order to set the scene for her comparative investigation, Schweitzer first of all recites what she calls the “standard narrative”. In the most concise terms, it boils down to something like: Compared to EU antitrust, US antitrust is a more highly evolved life form. Beginning with a review of the substantial swing of the ideological pendulum over many decades in the US with respect to Section 2 of the Sherman Act, Schweitzer proceeds to expose what she has excavated from the Florence historical archives53 concerning the Treaty negotiations in 1956–1957 at Val Duchesse and elsewhere. While much digging remains to be done not only in Florence but also Brussels and in the national capitols of the Six, Schweitzer has uncovered some of the most impressive and useful findings in recent times.54 These efforts debunk certain lingering myths and refute the not uncommon assumption that the European competition rules, and more generally, the Treaty of Rome itself, were a recipe for an ordoliberal feast. Such a view is both exaggerated and simplistic in light of the historical record.55 53

See http://www.eui.eu/ECArchives/EN. For further insights, see also Pinar Akman, “Long-Lost Soul”, cited supra note 21. For an extensive historical analysis with rather controversial claims to the effect that efforts to embed a full ordoliberal program in the Treaty of Rome not only failed but in practice failed utterly, see Sigfrido M. Ramírez Perez, “Public policies, European integration and multinational corporations in the automobile sector: the French and Italian cases in a comparative perspective 1945–1973”, Ph.D dissertation, EUI, 2007. For further details concerning the reasons for this failure (including in particular the ability of France and of European multinationals to counter attempts to push the ordoliberal agenda at the European level), see Ramírez Pérez, “Anti-trust ou Anti-US? Industrie automobile européenne et les origines de la politique de la concurrence de la CEE”, in Eric Bussière, Michel Dumoulin, Sylvain Schirmann, eds., Europe organisée, Europe du libre-échange, Peter Lang, 2006, pp. 203 et seq.; Ramírez Pérez, “La politique de la concurrence de la Communauté Economique Européenne et l´industrie européenne : les accords sur la distribution automobile (1972–1985)”, 27(1) Histoire, Economie et Société 63 (2008). 55 For further discussion with respect to the competition rules, see Giuliano Marenco, “The Birth of Modern Competition Law in Europe”, in Armin von Bogdandy, Petros Mavroidis and Yves Mény, eds., European Integration and International Co-ordination: Studies in Transnational Economic Law in Honour of Claus-Dieter Ehlermann, Kluwer, 2003, pp. 279 et seq. (focusing on the impact of the German influence generally, and not specifically on ordoliberalism). As for the Treaty of Rome more generally, see Fritz Scharpf, “Negative and Positive Integration in the Political Economy of European Welfare States”, Jean Monnet Chair Working Papers No. 28/1995; Christian Joerges, “The Market Without a State? States Without a Market? Two Essays on the Law of the European Economy”, EUI Working Paper LAW No. 96/2, p. 6; Miguel Maduro, “Reforming the Market or the State? Article 30 and the European Constitution: Economic Freedom and Political Rights”, 3 European Law Journal 55, 65 (1997). The continuing vitality of ordoliberal ideas is contemplated, with varying conclusions, by, e.g., Christian Joerges, “What’s Left of the European Economic Constitution? A Melancholic Eulogy”, 30 European Law Review 461 (2005) and Armin Hatje, “The Economic Constitution”, in Armin von Bogdandy and Jürgen Bast, eds., Principles of European Constitutional Law, Hart Publishing, 2006 (implicitly much more bullish than Joerges’ eulogy in spite of the changes wrought by Maastricht, Amsterdam, Nice and—projecting forward, as Hatje was writing prior to its demise—the European Constitution). 54

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xliv

xliv Introduction, Summary, Remarks Schweitzer then proceeds to dispel a number of false perceptions of ordoliberalism while countering claims that, under the traditional approach, Article 82: privileges fairness over efficiency; protects competitors rather than competition; and serves as a tool to regulate dominant firms. In contrast to these claims, Schweitzer does confirm a close link between the competition rules and the market integration objective. But she points out that, conceptually, the goals of market integration and efficiency are far from incompatible.56 Indeed, it seems obvious that the Treaty is built on the foundational premise that the two goals go hand in hand, as the effects of one without the other would be substantially limited.57 Reviewing the interpretation of Article 82 by the ECJ in the 1970s and 1980s, Schweitzer takes the Mestmäcker-inspired Contintental Can judgment of 1973 not only as a seminal judgment but perhaps as the critical juncture in the law under Article 82.58 As Schweitzer explains, Continental Can demonstrates, among other things, that the goal of Article 82 is a competitive market structure,59 and it shows that although dominance is permitted, Article 82 nevertheless protects residual competition in dominated markets.60 After 56 For a short teleological and ahistorical attempt to elaborate the relationship, see Marquis, “O2 (Germany) v. Commission and the Exotic Mysteries of Article 81(1) EC”, 32 European Law Review 29 (2007), at 30–31, footnote 5. 57 Relying on the structure of the Treaty (and in particular Article 3(1)(g)), as opposed to the more functional logic suggested here, the Court of Justice has explicitly stated that “Article 81 EC (ex Article 85) constitutes a fundamental provision which is essential for the accomplishment of the tasks entrusted to the Community and, in particular, for the functioning of the internal market”. Case C-125/97, Eco Swiss China Time Ltd v Benetton International NV [1999] ECR I-3055, para. 36. More recently, see the discussion of Advocate General Maduro in Viking Line, cited supra note 46, paras. 32–34. At paragraph 33, the Advocate General notes that, “[w]ithout the rules on freedom of movement and competition, it would be impossible to achieve the Community’s fundamental aim of having a functioning common market”. 58 As an emeritus scholar, Mestmäcker continues to produce important works. See, most recently, A legal theory without law: Posner v. Hayek on economic analysis of law, Mohr Siebeck, 2007. 59 Cf. the ECJ’s judgment in Case C-95/04 P, British Airways plc v Commission, cited supra note 37, para. 106 (referring to Continental Can and stating that “Article 82 EC is aimed not only at practices which may cause prejudice to consumers directly, but also at those which are detrimental to them through their impact on an effective competition structure, such as is mentioned in Article 3(1)(g) EC”). 60 Schweitzer notes that the ECJ’s reference to protecting residual competition is inspired by the fourth condition of Article 81(3), which refers to undue elimination of competition. This may also be of some relevance concerning the debate over whether efficiencies claimed in an Article 82 case should be interpreted consistently with the conditions of Article 81(3). The CFI has attached particular significance to those conditions, taking them to reflect broader imperatives with respect to competition policy. See Joined Cases T-213/01 and T-214/01, Österreichische Postsparkasse AG and Bank für Arbeit und Wirtschaft AG v Commission [2006] ECR II-1601, para. 115 (“It should be pointed out . . . that the ultimate purpose of the rules that seek to ensure that competition is not distorted in the internal market is to increase the well-being of consumers. That purpose can be seen in particular from the wording of Article 81 EC. Whilst the prohibition laid down in Article 81(1) EC may be declared inapplicable in the case of cartels which contribute to improving the production or distribution of the goods in question or to promoting technical or economic progress, that possibility, for which provision is made in Article 81(3) EC, is inter alia subject to the condition that a fair share of the resulting benefit is allowed for users of those products.”).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xlv

Introduction, Summary, Remarks

xlv

touching on some of the other formative judgments of the ECJ, Schweitzer highlights some of the characteristics distinguishing the Article 82 jurisprudence from the corresponding case law in the US under Section 2. First, the conceptualization set forth in Continental Can has developed into the view that Article 82 is more process-oriented (and accordingly less reliant on neoclassical analysis), the assumption being that a competitive process will indirectly benefit consumers.61 Second, the ECJ’s approach seeks to preserve market access, and exclusionary conduct therefore often attracts close scrutiny. Third, Schweitzer explains that Article 82 seeks not only to promote efficient outcomes, but also to protect the individual rights of those who compete with dominant firms. She further indicates that if the values of efficiency and individual rights clash in a given case, the latter prevail: “The difference between the EC approach and the US approach is that EC competition law . . . assumes an individual right of each competitor not to be excluded by illegal acts, whether or not the exclusion results in a verifiable overall decrease of competition or efficiency in the marketplace.”62 This suggests that market participation rights in Europe are not contingent on the efficiency of an undertaking exercising those rights. Schweitzer chooses three substantive areas for her comparative analysis: exploitative conduct, predatory pricing and refusals to deal. Regarding the first of these, she begins with the formal observation that, whereas the Sherman Act does not establish any direct price control mechanism (largely due to an underlying faith in self-correcting markets63), such an instrument is stamped on the very face of Article 82. Beyond this formal difference, however, Schweitzer points out that Article 82(a) is rarely pressed into service to curb prices. Relying on the paper by Motta and de Streel from the 2003 Workshop,64 she indicates that most instances of price intervention under Article 82 can be linked either to market-partitioning concerns or to the liberalization of utilities (where there may be strategic reasons to ensure that the public does not pay too high a price for open markets). If those idiosyncratic cases are placed to one side, then given the Commission’s general aversion to patrolling prices, the contrast between the EU and the US seems slight. If there is some modest divergence, Schweitzer maintains, it relates more to institutional particularities than to genuine philosophical differences.65 61 This is of course a major element of the “competitive distortion” model to which Gerber refers (see above pp. xxx to xxxiv). 62 See page 143 (emphasis in original). 63 Cf. Adam Smith, Lectures on Jurisprudence (R.L. Meek, D.D. Raphael and P.G. Stein, eds.), Oxford University Press, 1978, p. 363 (“[I]f any trade is overprofitable all throng to it until they bring it to the natural price, that is, the maintenance of the person and the recompense of the risque he runs.”). 64 Massimo Motta and Alexandre de Streel, “Excessive Pricing and Price Squeeze under EU Law”, in Claus-Dieter Ehlermann and Isabela Atanasiu, European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 91 et seq. 65 For further discussion of Article 82(a), see Panels V and VI.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xlvi

xlvi Introduction, Summary, Remarks With respect to predatory pricing, Schweitzer notes that the approaches taken in the two jurisdictions diverge significantly. The well-known US attitude, evident in Brooke Group66 and more recently in Weyerhaeuser,67 emphasizes the costs for dominant undertakings and the benefits for consumers if a predatory strategy should fail. Accordingly, US law imposes an onus on plaintiffs to show a dangerous probability of recoupment by the defendant, a tall order to fill. On the basis of post-Chicago scholarship, Schweitzer questions the wisdom of that approach. She also points out that the Community Courts have specifically declined to adopt such a forbidding standard of proof, and have attached importance to the intent of the alleged predator. Nevertheless, she says, there do not seem to be reasons to suppose that predatory pricing law in Europe inhibits price competition. Furthermore, she notes that some scholars in the US explain the underinclusive legal test there by reference to the specificities of the US legal system, including for example treble damages in private litigation. Needless to say, Europe is different. However, Schweitzer also digs a bit deeper and explains that the distinct approaches in the two jurisdictions reflect “different views of the structure and purpose of competition law”.68 Article 82, she explains, provides legal protection to a competitor who is foreclosed due to below-cost pricing—even where a dominant firm gambles and loses to the advantage of consumers. This is not because Community law deliberately seeks to reward inefficient competitors; rather, as suggested above, it “reflects the understanding that competition is a process that results from the exercise of individual rights”,69 which the defect of inefficiency cannot forfeit. Schweitzer then turns to the question of a duty to deal in cases involving either a withdrawal of supplies or a refusal to grant access to “essential facilities” (i.e., a refusal to start supplying some allegedly indispensable input). Here one is obliged to distinguish, as Schweitzer does, between cases involving intellectual property rights and those involving other kinds of assets such as raw materials, “must-stock” products, or distribution systems. With respect to non-IP cases, Schweitzer takes the Bronner case in the EU and the Trinko case in the US as natural points of comparison, as each case may be seen as staking out Areedean limits to forced sharing.70 Despite the similarities in those cases, however, there does seem to be more scepticism of the essential facilities doctrine in the US. Indeed, Schweitzer says that in Europe the doctrine is “well accepted”, and that it has played a notable role in 66

Brooke Group Ltd. v. Brown and Williamson Tabacco Corp., 509 U.S. 209 (1993). Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., Inc., 549 U.S. 312 (2007). 68 Page 154 of this volume. 69 Ibid., at 155. 70 The Bronner judgment effectively laid down the criteria that must be satisfied where essential facility-type claims are advanced. On the facts, Mr Bronner had to go home empty-handed precisely because Mediaprint’s nation-wide home delivery system was not deemed an essential facility. 67

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xlvii

Introduction, Summary, Remarks

xlvii

connection with access cases concerning assets controlled by former national monopolies. Although Bronner seems to have raised the bar to intervention, the criteria it laid down may be satisfied in such cases, in particular where networks cannot be duplicated. A more striking difference, perhaps, lies in the reluctance of most US courts to force IP holders to share their property,71 whereas in Europe, Magill and IMS Health make it possible, de facto, to characterize some IPRs as essential facilities. Schweitzer’s views in this regard are elaborated in a separate work.72 My own impressionistic sense is that one might arguably discern two parallel but incongruous trends in Europe. On the one hand, in non-IP cases, Bronner introduces a “last resort” approach and appears to ratchet up the conditions that must be fulfilled before imposing a duty to deal (though it remains to be seen whether and how Bronner may affect future cases where, contrary to the situation involving Bronner and Mediaprint, there is a preexisting commercial relationship between the parties). On the other hand, Magill, IMS Health and now Microsoft may progressively lower the bar compared to earlier statements of principle in Volvo/Veng,73 especially since IMS Health revealed that the two-market requirement might be satisfied merely by imagining a market upstream for the desired intellectual property.74 The consequence of these two trends might be that, in Europe, intervention may in

71 Although the Circuits are split, even those that have not adopted a per se legality approach in IP cases share that reluctance, an attitude that can only be reinforced in light of Trinko and in light of the personnel changes on the Supreme Court since Trinko was decided. 72 See Heike Schweitzer, “Controlling the Unilateral Exercise of Intellectual Property Rights: A Multitude of Approaches but No Way Ahead? The Transatlantic Search for a New Approach”, EUI Working Paper LAW No. 2007/31, available at http://cadmus.eui.eu/dspace/ handle/1814/7625. Consistent with the paper contributed for this volume, in her working paper Schweitzer points to a number of characteristics which distinguish Europe from the US and which limit the scope of convergence toward the US “model”. She also proceeds to suggest a way forward for a European approach that respects the existing legal framework and recognizes that some rare intervention can be justified, in particular where there are persistently high entry barriers and where the market does not “tend towards competition” in the medium term. But within that framework she cautions against “overly regulatory tendencies” and stresses that “competition for the market” can also be an important form of competition. 73 Volvo/Veng is susceptible of more than one interpretation. One could read the judgment as a bright line rule shot full of holes, which Magill later had to patch up. Like all cases in this area, Veng is ambivalent, but nevertheless it seems that the Court was trying to mark out more stringent principles for compulsory licensing of IP than those that apply in non-IP duty to deal cases. 74 Admittedly, it appears that the ECJ in IMS Health was attempting to impose some limiting principles, in particular by making it clear that cloned products would attract no sympathy. And since the IMS Health test is “cumulative” (though it probably does not and should not exhaust the concept of exceptional circumstances—see Pierre Larouche discussing Microsoft in TILEC Discussion Paper DP 2008-021, pp. 13–16), a stringent “new product” requirement might help to balance out a flimsy two-market requirement. However, the “new product” requirement could potentially be manipulated, and its robustness will depend entirely on the way it is applied in concrete cases. For a much more carefully considered discussion of IMS Health than the sketchy notes here, see Mark Patterson’s written contribution to this volume, pp. 705 et seq.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xlviii

xlviii Introduction, Summary, Remarks some cases be more easily justified in IP cases than in non-IP cases.75 This may be linked to the fact that the seemingly formidable condition of “indispensability” is liable in many cases to be satisfied straightaway if the defendant holds unique IP, whereas a plaintiff may sometimes be in a position to build another bridge farther down the river, or develop one’s own substitute raw materials (though it may have to borrow funds or find a partner to do so). To the extent that this hypothesis of incongruous trends is valid, European law has been evolving in unusual directions.76 In a concluding section, Schweitzer collects her main arguments, restates them and underlines their mutual connections. The paper is clearly a tour de force that should not go unheeded. Many of the issues raised by Schweitzer could be developed in further research. Only a couple of possible vectors will be mentioned here. As widely recognized, modern industrial economics teaches that market power, for all of its dangers (high prices, deadweight loss, sloth and slack, rent-seeking, etc.), may also have redeeming virtues (above all in the form of productive and/or dynamic efficiencies) in certain industries under certain assumptions. Ordoliberalism is 75 A key element in this regard would seem to be indispensability. Unlike Mr Bronner, who had alternative (though inferior) means of distribution available him, in IP cases indispensability may frequently be a low hurdle. In such cases, the main battle may well be fought on the basis of whether the “new product” criterion is satisfied. 76 For a battery of arguments to the effect that antitrust law should be no less ready to impose a duty to deal in IP cases than in non-IP cases, see Cyril Ritter, “Refusal to Deal and ‘Essential Facilities’: Does Intellectual Property Require Special Deference Compared to Tangible Property?”, 28(3) World Competition 281 (2005). Gustavo Ghidini is concerned that IP protection may no longer strike the right balance, and that antitrust intervention might be justified in exceptional cases where, for example, there are network effects. This might suggest a sectoral approach. At the same time, while he is open to exceptional intervention to permit follow-on innovation, Ghidini would also insist on grantbacks of derivative technology to the original inventor. See Ghidini, Intellectual Property and Competition Law: The Innovation Nexus, Edward Elgar, 2006. For the somewhat more common view that special deference in IP cases is warranted to avoid reducing incentives to innovate, see, e.g., the EAGCP Report, supra note 34, p. 44. See also Hovenkamp, Janis and Lemley, “Unilateral Refusals to License in the U.S.”, John M. Olin Program in Law and Economics Working Paper No. 303, Stanford Law School Law (April 2005), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=703161, p. 17: “Intellectual property law generally permits owners to enforce their rights by means of an injunction, and does not compel them to use or license their rights to others. For antitrust law to reach a contrary conclusion would require it to make illegal precisely the same conduct that the intellectual property laws explicitly authorize.” The authors conclude that “the better view is that an intellectual property right itself cannot constitute an essential facility, and that the doctrine should not be applied to cases that seek access to an intellectual property right in any but the most unusual of circumstances.” Ibid., p. 23. For further discussion, see, inter alia, Richard Gilbert, “Holding Innovation to an Antitrust Standard”, 3(1) Competition Policy International 48 (2007); James Langenfeld, “Intellectual Property and Antitrust: Steps Toward Striking a Balance”, 52 Case Western Reserve Law Review 91 (2001). Also relevant to the debate, of course, is the literature that seeks to identify the degree of competition that is optimal for stimulating innovation. The theoretical work suggests that it is far too simple to say, in the abstract, that either a competitive or a monopolistic market structure is ideal. For a recent discussion, see Jonathan Baker, “Beyond Schumpeter vs. Arrow: How Antitrust Fosters Innovation” (June 2007), available at http://ssrn.com/abstract=962261.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xlix

Introduction, Summary, Remarks

xlix

often instinctively associated with “complete competition”, which would virtually exclude the possibility of concentrated market power and with it those potentially desirable qualities. Another instinctive association links ordoliberalism with the dreaded “as if” standard of competition for “unavoidable” monopolies. Yet complete competition and “as if” competition should be seen as temporary, unsuccessful “experiments” within the ordoliberal research program, and should not be taken as that tradition’s defining elements. As is clear from Schweitzer’s paper, ordoliberalism is, within limits, a heterogeneous and dynamic set of ideas, values and principles.77 If one leaves behind the caricatures of complete competition (which was too static for the ordoliberal paradigm and which was never really more than an ideal-type) and “as if” competition, one difficult dialectic seems to remain. Consistent with the liberal tradition, ordoliberalism is built first and foremost on individual rights (both economic and political). Conceptually, it is incompatible with any theory that justifies the sacrifice of individual rights for the sake of maximizing aggregate welfare. What might seem paradoxical, however, is that ordoliberalism has been described as not just liberal (rights-based and anti-constructivist, with decentralized economic decision-making) and democratic (e.g., because it seeks to prevent capture by special interests) but also utilitarian.78 How can that be? The principle of “the greatest good for the greatest number” need not be construed in such a way as to make individual rights expendable. A synthesis of rights-based and utilitarian approaches may be found, for example, in the influential work of Adam Smith. Contrary to popular perception, Smith did not regard “Man” as a purely egoistic actor seeking to preserve and cultivate his personal wealth.79 And his view of Man was too subtle to cast him as a

77 See also, e.g., Manfred Streit and Michael Wohlgemuth, “The Market Economy and the State”, cited supra note 31, at p. 3 and references therein. Even Eucken and Böhm themselves, though complementary, should not be seen as homogeneous thinkers. See ibid., p. 5 (noting certain affinities between Böhm and Hayek not shared between Hayek and Eucken). To appreciate why it may often be useful to speak of individual scholars rather than of an ordoliberal “school”, it suffices to recall Alfred Müller-Armack, whose ideas differed significantly from of those of, e.g., Eucken. See further Vanberg, cited supra note 29, p. 2. 78 See Cristoph Engel, “Imposed liberty and its limits: The EC treaty as an economic constitution for the Member States”, in Talia Einhorn (ed.), Spontaneous Order, Organization and the Law. Roads to a European Civil Society. Liber Amicorum Ernst-Joachim Mestmäcker, Asser Press, 2003, pp. 429 et seq., at 433. 79 In the proverbial “vast literature”, see, e.g., Ronald Coase, “Adam Smith’s View on Man”, in 19 Journal of Law and Economics 529, 529 (1976) (quoting The Theory of Moral Sentiments, originally published in 1759: “How selfish soever man may be supposed, there are evidently some principles of his nature, which interest him in the fortune of others, and render their happiness necessary to him though he derives nothing from it, except the pleasure of seeing it.”). Naturally, this should not be interpreted as a suggestion that men are wholly altruistic either. As Smith pointed out, our capacity for (deep) sympathy can at times be quite marginal when suffering individuals are distant from us. See ibid., pp. 530–531. Yet even this aspect of human frailty may potentially be overcome, not necessarily by genuine sympathy, but by a “love of what is honourable and noble”, and by a desire to rise above our lesser qualities. See ibid. at pp. 531–532.

(AA) Ehlermann 08 Introduction

l

12/11/08

10:25

Page l

Introduction, Summary, Remarks

“rational utility-maximizer”.80 Nevertheless, Smith did take the view that Man can never quite escape a tendency to pursue self-interest, and even his noble capacities for sympathy, benevolence and altruism are animated by that propensity.81 In short, Man is capable of great humanitarian enterprises, but ironically it is his frailty that drives him to undertake them, if it does not preclude him in the first place from doing so. This highlights the point for which Smith is perhaps best known: the connection between one’s folly, frailty and self-love, and his or her micro-scale contribution to broader patterns of productivity. The social order, as Smith perceived it, is hardly a perfect mechanism, but it is nevertheless an ingenious way to reconcile the value of individual liberty with more collective or utilitarian values. By protecting—or by constitutionally “imposing”—liberty,82 and thereby ensuring the possibility of voluntary, decentralized economic transactions within a frame of general rules (including property and contract law), benefits redound to the social order as a whole.83 It has been amply demonstrated that the ordoliberals were firmly rooted in this Smithean (and Mandevillean) tradition.84 More specifically, while it is well known that the ordos favoured individual economic liberty and opportunity, it sometimes passes unnoticed that they also valued consumer sovereignty,85 that is to say, they valued the natural aggregate fruits of Smith’s envisaged spontaneous system. Indeed, it would appear that the whole point of “performance competition” was to promote (short- and long-term) consumer welfare.86 80 See ibid., pp. 545–546 (Smith “thinks of man as he actually is—dominated, it is true, by selflove but not without some concern for others, able to reason but not necessarily in such a way as to reach the right conclusion, seeing the outcomes of his actions but through a veil of selfdelusion”). 81 Coase stresses the “natural selection”, “self-preservation” and “propagation of the species” features of Smith’s approach to the social order, which of course was formulated long before the theory and evidence of evolution could be fully developed. See ibid., at pp. 539 et seq. 82 See Engel, cited supra note 78. 83 Hayek saw the Smithean order, functioning spontaneously with the aid of general and abstract rules of just conduct and coordinated by means of the price system, as an ingenious solution to the “constitutional” limits of human (subjective) knowledge and to the radically dispersed state of economic information (e.g., cost or demand information, etc.) in a continually changing environment. Many of his works pivot around this theme. For just one example, see Hayek, “The Use of Knowledge in Society”, 35 American Economic Review 519, 526 (1945). For general discussion, see, e.g., Manfred Streit, “Cognition, Competition and Catallaxy”, 4(2) Constitutional Political Economy 223 (1993); John Vickers, “Concepts of Competition”, 47(1) Oxford Economic Papers 1, 12–16 (1995) (also discussing relevant developments in the economics of imperfect information). 84 See, e.g., Razeen Sally, Classical Liberalism and International Economic Order: Studies in Theory and Intellectual History, Routledge, 1998. 85 See Vanberg, cited supra note 29, at pp. 12–13 (citing the works of Böhm and Eucken). 86 Furthermore, as Engel (cited supra note 78) has noted, the ordos were perfectly aware that a free market would tend to produce both winners and losers. Although they left room for basic social welfare, their system was not supposed to guarantee side-payments to the losers. There does not seem to be any prescription according to which an undertaking must subsidize its competitors, at least if the “as if” competition standard is forgotten (see supra note 31).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page li

Introduction, Summary, Remarks

li

This may present interesting possibilities for future research aimed at the interactions between rights-based and utilitarian approaches, in particular the manner in which liberty can be protected without going down the “complete competition” road and enjoining, a priori, concentrated economic power.87 Despite the rather bad reputation assigned nowadays to the performance/impediment dichotomy, perhaps performance competition could be a key guiding principle. Finally, further research might usefully explore the nature and content of the individual rights to which Schweitzer refers thematically in her paper. Interestingly, the concept of “economic freedom” continues (on occasion) to appear even in the discourse of the Commission.88 If the competition rules do indeed enshrine individual rights, should they be interpreted as essentially an opportunity to participate in the market, and no more? This would primarily require vigilance to ensure that markets remain contestable, for example in cases of government failure where legal barriers to entry do not serve the public interest.89 Alternatively, should these rights imply something more, as Schweitzer seems to suggest? If undertakings have a right not to be foreclosed by dominant firms,90 this would seem to accord well with Kantian ethics,91 87 As Schweitzer points out (see, e.g., page 134, footnote 76 of her paper), Franz Böhm appears to have accepted that market power can drive markets forward by stimulating performance, competition and new entry. 88 See Neelie Kroes, “The Law and Economics of State aid control—a Commission Perspective”, SPEECH/07/601, Berlin, 8 October 2007. Addressing a largely German audience, Kroes remarked that “[c]ompetition policy—and that includes State aid policy—embodies a simple but healthy principle of our market economies—the basic idea of freedom. Freedom of initiative and entrepreneurship, freedom to compete, freedom to win or to lose on the market.” Ibid. at p. 3 (emphasis in original). For that matter, Philip Lowe in his oral presentation at this Workshop referred specifically to the concept of Wettbewerbsfreiheit (“freedom to compete”, a concept associated in particular with the work of Hayek’s successor at Freiburg, Erich Hoppmann). See page 609. 89 It would seem appropriate to promote contestability in relation to not just a dominant firm’s home market but also neighbouring markets. An auxiliary concern, therefore, would be to ensure that firms that have legitimately attained dominance on one market through skill, foresight and industry do not then resort to shortcuts (naked exclusion) to capture other markets, particularly those in their infancy. 90 As Schweitzer says, Article 82 “will ensure that the fate of each competitor will depend on skill, business acumen and luck, and not on the exclusionary exercise of market power by a dominant firm”. See page 155. 91 Kant’s work had a direct influence on the views of both Eucken and Böhm regarding private economic power. See Streit and Wohlgemuth, cited supra note 31, at p. 7, footnote 9. See also Schweitzer, this volume, page 162, footnote 211 and accompanying text. A further study (in English) would be useful to draw out the full implications of Kant’s ethics for the economic order and for competition policy in particular. A logical point of departure would appear to be the principle of “external freedom”, i.e., the external corollary (law) to internal moral conscience. A central and famous principle of external freedom, which seems consistent with classical liberalism, is that “everybody may pursue his happiness in the manner that seems best to him, provided he does not infringe on other people’s freedom to pursue similar ends”. Kant, On the Old Saw: That May Be Right in Theory But It Won’t Work in Practice 290 (E. Ashton, trans. 1974). Elsewhere, Kant exhorts individuals to act “in such a way that the free use of your will is compatible with the freedom of everyone according to a universal law”. Kant, The Metaphysical Elements of Justice 231 (J. Ladd, trans. 1965). Kant’s ethics are developed in the second part of

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lii

lii Introduction, Summary, Remarks but it might be incompatible with a common assumption broadly shared by the Workshop participants, namely that there is both bad foreclosure and good foreclosure (crudely put, eclipse and elimination of wasteful rivals whose efforts could be better spent elsewhere).92 It is conceivable that Franz Böhm may have recognized a distinction between good and bad foreclosure,93 although this would have to be verified by further study. In any event, such a distinction is arguably consistent with the liberal tradition, since a rule prohibiting an efficient dominant firm from foreclosing inefficient competition could arguably be seen as a form of indirect subsidies extracted from that firm in favour of its rivals.94 The competitive process might be equally protected if the emphasis is on ensuring contestable markets so that efficient potential competitors (and “pre-efficient” start-ups) are able challenge the dominant firm, within a reasonable time frame, if the dominant firm should settle into a complacent life.95 6. Jim Venit’s paper, Cooperation, Initiative and Regulation—a Cross Cultural Inquiry, digs into the history and psychology lying behind the different antitrust cultures in the US and in Europe. One of the questions guiding Venit’s reflections is why societal attitudes in 20th Century Europe, expressed in laws and institutions, tended to privilege industrial cooperation and collusion while disapproving of aggressive and individualistic commercial behaviour on the part of dominant firms. Inspired by the work of David The Doctrine of Virtue (M. Gregor, trans. 1964). Needless to say, the commentary on Kant’s ethics and approach to justice alone is endless. For a thoughtful drop in the ocean, see Mary Gregor, Laws of Freedom, Oxford, 1963. 92 See, e.g., John Vickers’ oral remarks, this volume, page 197. The distinction between good (or neutral) foreclosure and bad seems to have been an important premise underlying DG Comp’s Discussion Paper. Cf also the Commission’s Non-Horizontal Merger Guidelines, cited supra note 11, paras. 29, 31–32, 59, 94, and 111. It is clear from the Guidelines that (input or customer) foreclosure will only be regarded as anticompetitive if post-merger prices are likely to rise. 93 In a different context (i.e., while discussing the folly of State intervention), Vanberg refers to Böhm’s argument that “the essence of economic power lies in the ability of inferior suppliers to prevent customers from accessing more attractive alternatives”. Vanberg, cited supra note 29, at p. 14 (citing Böhm; emphasis added). 94 Schweitzer by no means advocates subsidies in favour of rivals, who merely “engage in a process in which they may lose and possibly perish”. See page 155. 95 This paragraph questions the wisdom of maintaining a policy that could protect inefficient competitors (apart from the “infanticide” scenario where a new entrant is driven out before it has a reasonable chance to achieve efficient scale). If a welfare perspective is adopted, it might also seem desirable to sanction exclusionary conduct where the presence of inefficient rivals results in a degree of additional output in the market (allocative efficiency) that more than offsets the cost in terms of reduced market-wide productive efficiency. See, e.g., Jeremy West, “Background Note”, OECD Policy Roundtable on Competition on the Merits, cited supra note 22, p. 29. (Since the welfare consequences of eliminating an inefficient rival appear a priori to be ambiguous, this perspective would not protect inefficient rivals on the ground, for example, that a less concentrated market structure is in itself desirable). However, while lower market prices resulting from the persistence of inefficient competitors may have some appeal and may provide theoretical support for intervention, such an approach also bears significant risks of its own. See ibid. and references therein.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page liii

Introduction, Summary, Remarks

liii

Gerber, Venit begins with the ascendancy and appeal of industry-wide cartels in certain European countries in the 19th and early 20th Centuries. The postWar period in Germany featured an unprecedented intellectual and political challenge to the captains of industry, as momentum in favour of a genuine anti-cartel law took hold. Yet not even the sustained efforts of Franz Böhm and his associates could ensure the adoption of a competition law free of loopholes and escape clauses. The European cartel culture, or to use a more benign expression, the solidarity culture, was strongly rooted in German society and across the Continent. Perhaps only the combination of leniency in cartel cases and tough sanctions was capable, in the 1990s, of prompting Europe to start scraping off the barnacles. In the US, by contrast, only in the clutches of the Great Depression did the sirens of economic stability attract the sympathy of the courts in cartel cases.96 So much for cartels, where ordoliberal and consumer-welfare approaches to competition share some common ground. As for single-firm conduct, as Venit points out, the two jurisdictions have followed quite different paths. The prevailing tendency in the US, at least since the 1970s, has of course been to encourage aggressive competition in the name of efficiency and welfare. The ordoliberal approach, with its emphasis (for both economic and constitutional reasons) on the interrelated goals of maintaining long-term openness in markets, suppressing “impediment” competition (i.e., exclusionary or predatory practices), and promoting economic liberty, recognize the value of efficiency but only as a function of, and only insofar as it coincided with, those normatively superior objectives.97 In parallel with concerns about the 20th Century tendency toward central planning, the risk to an open society was considered too great to give free rein to a large firm whose efficiency depended on its control over the market. While one should guard against conflating the constitutional and “market managing” tenets of ordoliberalism on the one hand and the path of Community competition law on the other, as Gerber, Schweitzer and Venit have all shown, the impact of the former on the latter has been profound and perhaps inestimable. For Venit, the mutually unique histories of the US and Europe have shaped different psychological understandings of competition itself, to say nothing of competition “on the merits”. Such psycho-linguistic differences may not be obvious; once uncovered, however, Venit points out that their consequences for policy preferences become so. The law on discounting practices by dominant firms is the example he chooses. While much of Venit’s paper is devoted to the search for explanatory insights, there is no mistaking his normative claim that a meaningful reform 96 See Appalachian Coals, Inc. v. United States 288 U.S. 344 (1933), upholding a price-fixing agreement in the coal sector. By 1940 the Supreme Court had restored the per se prohibition of hard core cartels. See United States v. Socony-Vacuum Oil Co. 310 U.S. 150 (1940). 97 See Venit, page 173 (quoting Giorgio Monti, “Article 81 EC and Public Policy”, 39 Common Market Law Review 1057 (2002)).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page liv

liv Introduction, Summary, Remarks of the traditional Community law approach to single-firm conduct is long overdue. Citing the work of Giuliano Amato, Venit stresses the “disquieting contradiction” between intolerance for hard-nosed competition by dominant firms and a cloying dependence on the State and collectivist (non-market conforming) solutions.98 Here it could be argued that Venit does not distinguish adequately between the strong ordoliberal aversion to public intervention in market processes by public authorities or by public or privileged undertakings (partly to protect economic freedom and partly to avoid capture and government failure) and the insistence on a strong constitutional framework within which an independent antitrust authority must maintain open markets. Whereas Venit seems to take the view that intervention by an antitrust authority is comparable to other (possibly anticompetitive) market interferences by public authorities, an ordoliberal might respond that a decision to refrain from intervening in cases where a large firm shuts down competition is equivalent to an affirmative decision to interfere with the market or to endorse ongoing distortions. In a sense, the ordoliberals sought to avoid the naturalistic error of assuming that the status quo reflects natural market dynamics. On the other hand, Venit is on very solid ground when he underlines the need for competition policy and enforcement to consider the reasons accounting for a firm’s dominant position. Perhaps this suggests, as posited earlier, that the concept of “performance-based” competition—which would have no sympathy for arbitrary State-granted exclusive rights—might be worth holding on to after all. As for firms with dominant positions resulting from the underlying economics of the market itself, a judicious application of the concept should not lead to the excessive intervention about which Venit expresses his concern. As competition authorities are hopefully coming to recognize in Europe and elsewhere, it is necessary to prioritize, to take enforcement costs into account, to guard against prosecutorial bias once an investigation is launched but also against the seductiveness of competitors, and to evaluate error costs where the effects of commercial practices are likely to be ambiguous. There is no perfect competition policy in the sky, so the watchwords should be: performance—with careful attention to the underlying economics of affected markets rather than a prejudice for certain ideal market structures;99 long-term consumer welfare; and, in the European context (again, for both economic and constitutional reasons), market integration.100 The “C” word, convergence of competition policies (let alone convergence of 98

See page 183. I recognize that Community law limits the extent to which Article 82 (and Article 81) can be interpreted in an a priori structure-free manner. Schweitzer makes this abundantly clear in her paper, for example at pp. 140–142. 100 With regard to market integration, I have noted elsewhere that, when that objective is liberated from a dogmatic understanding, the balanced pursuit of it may be largely consistent with and even (without denying its constitutional rank but recognizing its dynamic texture) instrumental to consumer welfare. See further supra notes 56–57 and accompanying text. 99

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lv

Introduction, Summary, Remarks

lv

the deeper values they reflect), if it means total convergence and homogeneity, clearly lies outside the realm of the possible.101 However, a greater emphasis on effects and a serious willingness to give due weight to productive and dynamic efficiencies and concrete market effects may provide a constructive response to Venit’s critique without implying the loss, in Europe, of the constitutional role of the competition rules.

B. Discussion102 The roundtable discussion following the first panel of speakers (Gerber, Padilla, Zimmer, Fox, Schweitzer, Venit) reflected both the cultural diversity highlighted in the papers discussed above and, to some extent, the diverse professional backgrounds of the participants. Two themes in particular—the proper objectives of competition law and the role of economists—served to guide the debate.103 The objectives of competition law John Fingleton launched the discussion with an inventory of seven candidate goals for competition law: consumer welfare, efficiency, market integration, protecting rivals, fairness, economic freedom and individual rights. The implicit task for the participants was to whittle the list down, but they quickly managed to expand it. In general, the discussion circled around the dialectic of market outcome versus market process. Some participants questioned whether the distance between the two poles was really as far as it seems. For example, John Vickers indicated that in his view a process-based approach to competition and an effects-based approach might in fact be close relatives if competition is understood in functional as opposed to abstract terms. Other participants suggested that an enlightened approach to antitrust enforcement might consist of a synthetic blend, or a “double helix”, to borrow a catch phrase from a well-known scholar. 101 By contrast, and as noted below with reference to the Panel I discussion, a “limited convergence”, perhaps not as to what the defining goals should be but at least as regards what they are not, is more plausible. Even the CFI’s judgment in GlaxoSmithKline (with its emphatic approval of a standard promoting the welfare of final consumers) would not imply convergence, since in the US consumer welfare means something else (see the oral remarks of Doug Melamed and David Meyer in Panel I; those remarks are also highlighted below in this chapter). Indeed, the CFI’s view does not even seem to correspond to that of the Commission. An extended discussion will not be attempted here. 102 As noted above, only the roundtable discussions, and not the presentations of the participants, are reviewed in this chapter. However, the presentations are reproduced verbatim in this volume (see Panel discussions). 103 A number of other issues were also addressed, such as, e.g., the usefulness of market shares for the purpose of assessing market power. However, the following account is by no means comprehensive. For a complete picture, the reader is referred, as always, to the actual transcripts contained in this volume.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lvi

lvi Introduction, Summary, Remarks Adding to Fingleton’s list of objectives, Jochen Burrichter referred to the ordoliberal parallel between control of public power through the rule of law and control of private power through competition policy, and in particular through the maintenance of open markets. Welfare, on this view, is a dividend of free competition. By contrast, and like Vickers (and Venit104), Patrick Rey questioned the proposition that competition should be regarded as an end in itself. Rey affirmed a consensus among economists that competition is indeed best conceived of as instrumental to the pursuit of other goals. This normative proposition also feeds into the question of the appropriate role of economics in competition policy (see infra page xliii). Another normative proposition that Rey seemed to favour (drawing on Venit’s presentation and later supported by Paulis) is the possibility that goals ensconced in the Treaty of Rome, and in particular the imperative of market integration, are capable of different interpretations at different points in time as a function of changed circumstances. Einer Elhauge returned to the question of whether the differences between US law and Community law are more apparent than real. In his view, the differences relate more to vocabulary than to substance. A related point was that the differences are sometimes deceptively magnified by caricatures—that of the strictly outcome-oriented US approach and that of the European “protect rivals at the expense of consumers” approach. In the view of Emil Paulis, the outcome-based and process-based approaches are rather more different than Elhauge suggests: conceptual differences matter. But Paulis posits the notion, referred to above, of a qualified synthesis. According to this view, if the market outcome model is applied in a manner that takes into account the likely long-term effects of commercial practices vis-à-vis consumers, that model can be reconciled to a certain extent with the process-based model. Yet an important difference remains, because as Paulis recognized, the “long-term consumer welfare” standard which he described, and which the Commission strives to apply, is in the final analysis an effects-based approach that implies a meticulous market analysis. It is not necessarily the same as protecting the competitive process on the basis of a belief that maintaining open markets and commercial opportunities for all— which might imply a more abstract analysis—is both desirable in itself and beneficial in the long run for consumers. On the other hand, although he supported an effects-based approach in general, Paulis also seemed to transpose the concept of “restrictions by object”, familiar from the Article 81 context, to the abuse of dominance context: prohibiting certain practices a priori, he says, may obviate a more searching analysis with respect to certain practices that are harmful by their very nature. A final point was that the legitimacy of efficiency-based justifications is by now recognized under both Article 81 and 104

See Venit’s paper in this volume.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lvii

Introduction, Summary, Remarks

lvii

Article 82 as well as in the merger context. In his view, the fact that producers can invoke efficiencies responds to some of the concerns expressed by Daniel Zimmer in his presentation while perhaps bridging to some extent (though not entirely) the distance between the consumer surplus and producer surplus standards. Mario Siragusa said he would welcome an approach that adopts a longterm perspective, and one that is not confined to consumer welfare. But in his view, competition authorities, especially at the national level, cannot pry themselves away from the traditional emphasis on immediate consumer benefits and thus cannot embrace a genuine process-oriented enforcement policy. Furthermore, he was concerned about the potential for mischief arising from Article 9 of Regulation 1/2003. There is too great a temptation, he suggested, that this instrument will be used as a means of extracting a quick payoff where allegations of dubious merit are investigated. Bill Blumenthal introduced Competition Law Objective Number 9, administrability, and he recalled the emphasis laid by Philip Areeda and Donald Turner on manageable rules as they helped to develop a post-Bain Harvard school in the 1970s. Recognizing a need to accept certain tradeoffs, Areeda and Turner sought to steer the law toward norms that would be based on more focused objectives and designed to achieve reasonable certainty without too much administrative cost. Cal Goldman considered the question of the aims of antitrust from a global enforcement perspective. As reflected by a survey conducted by ICN, he explained, competition law enforcers worldwide appear to be homing in on the twin goals of consumer welfare and economic efficiency. He pointed out that this reflected a relatively increased emphasis on consumer welfare, as in the 1990s an OECD survey suggested that consumer welfare was regarded as a secondary aim or incidental benefit. A tenth objective, proposed by Ian Forrester facetiously yet at the same time in a serious vein, was deterrence and punishment of “economic crimes”. Forrester was plainly concerned by the conscious or unconscious tendency to attach moral reproach and stigma to aggressive business practices, as reflected by the imposition of fines. Later in the day, Cal Goldman forcefully interjected the point that demonizing single-firm conduct is inappropriate policy.105 Doug Melamed’s remarks brought the grand total of antitrust law objectives to eleven. In the US, he explained, when one peers past the ubiquitous rhetoric of consumer welfare, it can be seen that the law guards against the “inefficient creation of market power for the benefit of trading partners, 105 It is worth quoting a part of Goldman’s remarks here: “I cannot equate unilateral conduct, under any circumstances, with concerted or cartel activity. They’re worlds apart. One has a behavioural component that is far more egregious and deserves sanctions of the highest order. The other is a unilateral course of business, albeit aggressive, but often efficiency-directed.” For the full passage, see the Panel IV discussion.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lviii

lviii Introduction, Summary, Remarks whether they are suppliers or purchasers”.106 This denotes, as he put it, a “trading partner welfare” standard. Melamed then critiqued the argument according to which a process-based, long-term perspective produces long-run consumer welfare as a dividend. The “dividend” argument, it is recalled, is linked both to the traditional “competitive distortion” model discussed by David Gerber and to the via media described as the Commission’s approach by Emil Paulis. On the one hand, Melamed said, ignorance of the future implies the risk of unreliable predictions of entry and innovation (something Paulis himself had earlier acknowledged implicitly). On the other hand, protecting the market process is unlikely to ensure a desirable market outcome in the long run if legal rules chill efficient behaviour. David Meyer strongly endorsed that critique. The “competitive process”, as Meyer understands it, relates to the safeguarding of the market so that it can function according to its own dynamics, recognizing that in a healthy, functioning market, some businesses may drop out of the race: “it is the competitive process that yields those outcomes, rather than the unilateral judgment of a monopolist, or the collective judgment of cartel members who’ve decided how much to make and how high to set their prices”.107 Meyer also underscored Melamed’s point about “trading partner” welfare by explaining that, from the standpoint of the DOJ, “consumer welfare” is by no means limited to the welfare of final consumers. Finally, coming back to the caricatures discussed by Elhauge, Meyer pointed out that the DOJ considers not only short-term market effects but also long-term probabilities with respect, in particular, to innovation. Short-term effects may be relevant but they are not always determinative. If short-term allocative efficiency dictated enforcement decisions, he suggested, this would imply undesirable market engineering with the aim of imposing conceptually attractive market structures. Marc van der Woude recalled that Article 82 and the Treaty’s other competition rules are normally read in the light of Article 3(1)(g) EC (or Protocol 27, if the Treaty of Lisbon survives the ratification process), which in turn should be read in the light of Article 2 EC (or Article 3(3) of what would be the amended TEU). The provision refers perhaps most importantly to the EU’s task of establishing the internal market, but it embraces a range of other aims too, such as competitiveness and economic and social cohesion. For good or ill, it may be legally problematic to conceive of European competition policy as being isolated from such other collateral objectives. It was not until near the end of the roundtable discussion that the tangleweed of objectives was finally uprooted by Eleanor Fox: “Nobody really thinks are ten objectives. . . . All those objectives could be shrunk down to something like the following formulation: competition law aims at helping to 106

Page 31. Page 32. On the surface, Meyer’s language is very similar to that used by Heike Schweitzer, in her paper, with respect to the unilateral judgment of dominant firms. See Schweitzer, p. 155. 107

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lix

Introduction, Summary, Remarks

lix

make markets work, and it does so through rules that promote the incentives to support the market rather than obstructing the market.”108 Implicit in this formulation is the efficient functioning of markets. But as Fox observed, it is precisely on the question of how to promote efficient markets that methodologies diverge, and again, methodologies can make a difference. Both the outcome-oriented and process-oriented models have their potential strengths. However, the line between cautious enforcement and under-enforcement is slender. Fox’s comments illustrate the fact that even the dichotomy between the short run and the long run is interpreted differently according to the model in question. From a traditional European perspective, uncertain positive future effects suggest the need to apply the “precautionary principle”. From a mainstream US perspective, uncertain negative future effects call for unassuming abstention by the authorities. The contrast reflects, once again, underlying cultural and hence cognitive differences.109 Two final remarks were added by Jorge Padilla. First, he questioned the wisdom and legitimacy of any approach that seeks to ensure a level playing field unless doing so can be tied to a further utilitarian end. The implicit point is that, while a process-oriented enforcement approach does not exclude utilitarian benefits, such benefits (at least in a case-by-case sense) do not appear to be a necessary condition for action by the enforcer. The second remark related to the short-term/long-term issue, which prompted Padilla to recall the theoretical and empirical work of economists studying innovation and growth. The results of that research are tentative, but they do suggest that the correspondence between market structure and innovation is more complex than is often supposed, and that consequently (as Meyer had suggested) a one-size-fits-all structural approach would be ill-conceived. The appropriate role of economics in competition law The looming presence of economics in competition law and policy is here to stay. And while in Europe the serene life of antitrust lawyers seems only recently to have been disrupted by an unruly class of outsiders using mystic symbols, the need for economic expertise in this field of law has long been recognized in the US. This is illustrated anecdotally by Judge Wyzanski’s decision to hire economist Carl Kaysen to work as a law clerk on the United Shoe Machinery case over a half-century ago.110 However, views as to the precise role economics should play, in a room shared by lawyers and economists, could hardly be expected to be uniform. This discussion of the role economics should play may be summarized as follows. 108

Page 34. Relevant to this point are all of the papers presented in Panel I, including in particular that of Venit. 110 See Carl Kaysen, United States v. United Shoe Machinery Corporation: An Economic Analysis of an Anti-Trust Case, Harvard University Press, 1956. 109

(AA) Ehlermann 08 Introduction

lx

12/11/08

10:25

Page lx

Introduction, Summary, Remarks

Adding to the two roles described by David Gerber in his paper and presentation (i.e., the normative role and the “fact-interpretive” role), John Vickers referred to a third, an “analytical role”. This seems to consist of a variety of activities carried out by economists, from empirical work designed to understand industries better to the scrutiny and falsification (or verification) of received wisdom. The analytical role thus may be, on the one hand, an important policy input (or corrective), and on the other, a source of information about the effects of particular practices in particular industries, which can then be used to piece together a framework for competitive analysis. The remarks of Patrick Rey indicate that it is not unusual for epistemic communities such as industrial organization specialists to develop normative propositions (e.g., “competition is a means to an end”). But then there are other times (perhaps where the distributional consequences are more overt) when such specialists may prefer to abstain as best they can from conscious advocacy and to limit themselves to discussing the advantages and disadvantages of different policy choices (e.g., a consumer surplus standard versus a total surplus standard). Mario Siragusa did not mince words: “of course economics should be used, but it should be used to support the legal rule—and it should not substitute itself for the legal rule”.111 And while economics has a clear value in the realm of antitrust law, Siragusa seemed to be concerned that a little knowledge can be dangerous, perhaps inducing an enforcer to justify intervention on the basis of speculative scenarios. Furthermore, he added, economics is difficult for an antitrust lawyer to grasp. If that is so, then odds are that non-specialist judges will fare no better. Marc van der Woude likewise pleaded for a balance between sophistication and simplicity. However, Bill Blumenthal challenged the premise of the “economics clutters the case too much” argument. It is implicit from Blumenthal’s remarks that a legal environment devoid of solid economics would make cases too messy and difficult to administer. Economics, he said, provides tools of greater rigour and precision than those found in other disciplines. The last word on this debate over roles was had by an economist, Jorge Padilla. In his view, economists contribute to competition law by developing rational and workable tests, and by lending their expertise to the analysis of concrete cases. According to Padilla, these tasks are quite necessary, proper and often fruitful.

111

Page 28.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxi

Introduction, Summary, Remarks

lxi

1.2 Panel II (papers by Meyer, Vickers and Burrichter) A. Written Contributions 1. David Meyer’s contribution is entitled The Development and Communication of Competition Law Standards Applicable to Single-Firm Exclusionary Conduct: A Perspective Based on U.S. Experience. The paper serves on the one hand to advocate the adoption of guidelines covering single-firm (exclusionary) conduct and on the other hand to subtly illustrate some of the differences that persist in the US and European enforcement cultures with respect to this area of the law. Following a brief review of the US approach (meant as a reminder that Section 2 of the Sherman Act does not apply to exploitative conduct and that aggressive competition on the merits by dominant firms, including aggressive price competition, is strongly encouraged), Meyer identifies several reasons why carefully crafted enforcement guidelines and other comparable authoritative statements can be a valuable interpretive tool for competition authorities, courts and, perhaps above all, undertakings and their advisors. Guidelines in respect of single-firm exclusionary conduct would respond to a particularly pressing need because uncertainties arising from the real or perceived possibility of Type I errors may well inhibit hard-nosed competition, to the detriment of consumers and the economy. Drawing on the past experience of the US antitrust agencies, Meyer concludes that clear rules of engagement in the form of guidelines would enhance legal certainty in both the US and Europe. Safe harbours, he says, would create significant added value. The first logical safe harbour in the single-firm context would relate to dominance: “to the extent that competition law is concerned with the ability of a single firm acting unilaterally to cause significant harm to the competitive process, it is hard to imagine a firm having that ability if its market share is under 50–60 percent”.112 Meyer also refers to a need for safe harbours to place certain kinds of conduct presumptively in the “competition on the merits” zone. As indicative examples, he refers to predatory pricing (in the paper) and rebates (during the roundtable discussion) as practices for which safe harbours would be feasible and desirable. Meyer’s paper thus clearly conveys the US side of the story. The contrast with not only the traditional European approach but also to relatively recent authoritative statements such as the judgment of the ECJ in British Airways is unmistakable. It remains to be seen whether future policy guidance in the US might have an appreciable indirect effect in Europe, but at least in the short term that does not seem terribly likely.

112

Meyer, page 234.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxii

lxii Introduction, Summary, Remarks 2. John Vickers provided a Speaking Note in which he draws on some of his earlier influential work. The first point he makes is that any guidelines relating to exclusionary abuses under Article 82 should accord with the Commission’s (now definitive) Non-Horizontal Merger Guidelines.113 This is not consistency for consistency’s sake; rather, it is a reflection of the parallel economics underlying the single-firm and non-horizontal merger scenarios, and their analogous competitive effects. Drawing on the non-horizontal merger context, Vickers identifies five elements that should be carried over to guidance under Article 82: an emphasis on consumer impact; the crucial distinction between pro-competitive foreclosure and anticompetitive foreclosure; recognition of efficiencies; clear theories of harm to competition and consumers; and scrupulous attention to facts and their compatibility with the putative theory of harm. With respect to the task of identifying where foreclosure is in fact the “bad kind” of foreclosure, Vickers reviews the sacrifice test, the as-efficient competitor test and the consumer harm test. None of these tests is ideal, he says. But since each is given expression in the law, in a range of variations and settings, there is a need for clarity as regards their application and a need to consider how to maximize their value, either by developing them further or at least by avoiding their pitfalls. The utility of guidelines in this area is thus underscored once again. Vickers also offers a number of remarks, not summarized here, concerning dominance and institutional differences between enforcement systems. 3. Jochen Burrichter’s paper, A Reformed Approach to Article 82: The Impact on Private Enforcement, is a timely contribution to the European debate on private antitrust litigation, which is now reaching full stride.114 While there has been a good deal of discussion of private actions in the wake of Courage and Manfredi, Burrichter approaches the subject from the relatively unexplored perspective of Article 82 litigation. He touches on a variety of themes, including: the implicit nullity sanction under Article 82; injunctive relief; stand-alone and follow-on actions; locus standi; causation; and the various methodologies that can be used to calculate damages, having regard to the nature of the abusive conduct. The assumption underlying Burrichter’s paper is that, in Article 82 cases, a potential plaintiff faces high hurdles. For example, the defendant may be privileged by asymmetric information regarding production costs, confidentially negotiated contracts, discounts not reflected in price lists, possible efficiency defences, and so on. Perhaps the highest hurdle of all, he says, is associated with the move toward an effects-based competition culture. To the extent that such a culture implies a need for more intense analysis to verify

113 114

Cited supra note 11. For details, see http://ec.europa.eu/comm/competition/antitrust/actionsdamages/index.html.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxiii

Introduction, Summary, Remarks

lxiii

whether a given business practice ultimately reduces or enhances welfare, a plaintiff’s life becomes more difficult. A defendant-oriented response might be that this is precisely the intended antidote to a traditional regime tainted by too many false convictions. From Burrichter’s point of view, however, the increased uncertainties, together with a somewhat broader scope for defendants to justify their behaviour, call for procedural devices to soften the blow. He envisages several situations in which, once a plaintiff adduces prima facie evidence to establish some element of his case, the (evidential) onus might fall to the defendant to refute it. For example, Burrichter favours an AKZO-style presumption of dominance once it is established that the defendant has a certain share of the relevant market.115 To rebut that presumption, the task of establishing significant competitive constraints or other reasons to doubt the allegations of dominance would then be the responsibility of the defendant. As for abuse, Burrichter extrapolates from Article 2 of Regulation 1/2003 that prima facie evidence of abuse should suffice to shift the burden to the defendant to substantiate any claims of countervailing efficiencies,116 a logical approach given that the defendant is more likely than other parties to have the information relevant to such claims.117 Here it is not clear whether Burrichter is referring to the legal burden of proof (i.e., the burden of persuasion) or rather the evidential burden of proof, but other commentators have

115 It is not clear from Burricher’s paper whether the solomonic 50% figure endorsed by the ECJ in AKZO should be retained as an appropriate prima facie marker. The AKZO rule is likely to be more meaningful in some markets than in others. If a defendant has held 50% of the market for a significant period of time and the rest of the market is highly fragmented, there may be good grounds for taking the market share as a prima facie indicator. But a market-blind application of the rule may be inappropriate. If the defendant’s market share is 50% but the market is, for example, a duopoly, this would be better interpreted as prima facie evidence of no singlefirm dominance. However, the Court’s judgment in AKZO may be flexible enough to account for such scenarios, as the 50% mark only triggers a presumption of dominance in the absence of “exceptional circumstances”. See Case 62/86, AKZO Chemie BV v Commission [1991] ECR I-3359, para. 60. 116 Consider also Recital 5 of the Regulation: “It should be for the undertaking or association of undertakings invoking the benefit of a defence against a finding of an infringement to demonstrate to the required legal standard that the conditions for applying such defence are satisfied.” The recital appears to be drawn with a broader stroke than Article 2 of the Regulation, which refers only to the exemption provided for by Article 81(3). It is difficult to know whether such broader wording was deliberate. If so, it is still not clear whether the recital could prevail over the impression given by Article 2 of the Regulation, to the effect that it applies only to the conditions of Article 81(3) whereas it is utterly silent as regards Article 82. This silence, while it appears to be quite deliberate, may be due to the fact that, until around mid-2003 (i.e., after the adoption of Regulation 1/2003 in December 2002), there was little discussion of opening up Article 82 cases to full-blown assessments of efficiency justifications (in contrast to discussions of the Community’s merger control regime, where the role of efficiencies was more widely debated). Yet in any event, the absence of reference to Article 82 in Article 2 of the Regulation seems to preclude allocating the legal burden of proof to the defendant in an Article 82 case. See infra note 118. 117 See also Ekaterina Rousseva, “The Concept of Objective Justification”, 2(2) Competition Law Review 27 (2006), available at http://www.clasf.org/CompLRev/Issues/Vol2Issue2Art2Rousseva. pdf, pp. 69–70; and see the oral remarks of Mark Patterson in Panel II, pp. 213–214.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxiv

lxiv Introduction, Summary, Remarks insisted that in an Article 82 case it must be the latter.118 In any event, Burrichter argues that a plaintiff should not be expected to prove that what appears in the first instance to be an abuse furthermore cannot be justified because it has not been necessary and proportionate to generate cognizable efficiencies. Burrichter also touches on other themes that later surfaced in the Commission’s White Paper on Damages actions, issued in April 2008,119 including the standing of indirect purchasers, questions of fault and good faith, causation and so forth. As Burrichter argues, the reform of Article 82 enforcement should not fail to take into account the collateral damage it might have on the incentives of plaintiffs to bring private antitrust actions. It does not seem inappropriate to consider whether unprecedented latitude for defendants to compete in the market aggressively and to take advantage of efficiencies should be matched by carefully selected procedural mechanisms aimed at promoting the truthseeking role of the judicial process. Above all, information asymmetries should be addressed, as they can only distort outcomes, to the detriment of the public. On the other hand, particularly with regard to substantive matters, one ought to be equally cautious about the risk of reintroducing a degree of formalistic criteria in contentious proceedings that the “re-thinking” of Article 82 is intended to overcome.

B. Discussion The roundtable discussion following the second panel of speakers (Meyer, Vickers and Burrichter) concerned, first, the need for the Commission to adopt guidelines covering the application of Article 82 to exclusionary conduct; and second, the appropriate objectives, form and content of any such guidelines.

118 See Robert O’Donoghue, this volume, page 350: “It is true that Article 2 of Regulation 1/2003 places the burden of proving the benefit of the conditions of Article 81(3) on the defendant, but it clearly does not apply the same principle to Article 82 and objective justification. In these circumstances, it should be for a plaintiff or competition authority to show that the anticompetitive effects outweigh the efficiencies, since otherwise no abuse has been proven. The dominant firm should simply bear the initial burden of producing colourable evidence to substantiate the efficiencies claimed.” (emphasis added). O’Donoghue thus maintains that the ultimate legal burden of proof remains with the enforcer or claimant, although the evidential burden may shift. This view appears to be consistent with that of Renato Nazzini, “The Wood Began to Move: An Essay on Consumer Welfare, Evidence and Burden of Proof in Article 82 Cases”, 31 European Law Review 518 (2006). As Nazzini points out, Article 2 of Regulation 1/2003 places the burden of proving a breach of Article 82 on the Commission, without any qualification. The legal burden thus “can never shift”. Ibid., pp. 519 and 538. 119 See http://ec.europa.eu/comm/competition/antitrust/actionsdamages/files_white_paper/ whitepaper_en.pdf.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxv

Introduction, Summary, Remarks

lxv

The need for Article 82 guidelines With respect to the need for guidelines, there was no debate. Apart from one lone, lively soul, there was a strong consensus among the group, including the enforcers from both Europe and the US, that guidelines are not only needed but, in the words of John Vickers, badly needed. In the light of the consensus reached and recorded in June of 2007, it is puzzling and regrettable that the apparent momentum has been lost and that, thus far, no draft has surfaced. The lone dissenting voice alluded to above was that of Mario Siragusa. He would prefer to sink his teeth into a case with solid facts rather than wading through the muck of econojuridical guidelines jargon. The related but more serious charge is that, captivated by the intellectual gratification of refining legal and economic concepts in guidelines, the Commission is neglecting its fundamental law-enforcement responsibilities. Although no one seconded the proposition that guidelines serve no purpose, certain others (e.g., Marc van der Woude and John Vickers) did agree that cases are certainly valuable as an additional source of guidance. On the other hand, Bill Blumenthal noted the irony of Siragusa’s call to open new cases if the thrust of an effects-based approach is to correct a perceived problem of overdeterrence. Elhauge was one of those who disagreed with the proposition that case-bycase development should be the only focus. As he explained, where the law develops only by accretion, with individual decisions leading where they may, there might be negative externalities because system-wide coherence is not adequately taken into account. In particular, a pure case-by-case approach is unlikely to pay sufficient attention to error costs and related ripple effects. For his part, John Vickers noted several reasons why guidelines under Article 82 are necessary, including, for example, the relatively sparse jurisprudence in this area. The few judgments that have been handed down were marked by peculiar facts, yet these odd cases serve as the source of generalized principles. Furthermore, Vickers notes that a series of cases decided in the 1970s and early 1980s continue to determine the law today, even when contemporary antitrust thinking has taken several steps beyond the state of the art that once prevailed. After decades of path-dependence, an unfortunate consequence is what Vickers calls “huge uncertainty” with respect to the proper interpretation of a critically important field of law. A casuistic attempt to correct the deficiencies in the law would take far too long, and in the meantime too many avoidable costs would be incurred. Objectives, form and content Regarding the objectives to be served by such guidelines and their proper form and content, there was general agreement on some points but not all. One would think that the objectives of competition law guidelines in the European context would be relatively easy to articulate. In no particular

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxvi

lxvi Introduction, Summary, Remarks order, the preparation and adoption of guidelines would seem to be motivated by the following (interrelated) goals:120 • clarity with respect to policy (implying the inclusion of unambiguous statements of principle where appropriate); • transparency with respect to analytical methodology (which can also foster dialogue and theoretical research); • technology transfer from competition policy makers to other public authorities (in particular, non-specialist courts, although guidelines do not have autonomous binding effects on them), which may imply a steep learning curve in the short-term to achieve a more sophisticated competition culture in the longer term; • uniform application of competition rules across Europe, within and among enforcers and courts, whether national or European (not only for the coherence of the system but also to reduce transaction costs); • practical utility, which may imply the taming of abstruse theory by means of operational principles and the use of bright line rules (e.g., safe harbours or “safety zones” where they are supported by strong justifications); and • legal certainty, not only for its own sake but to facilitate investment (all the more imperative where the law is in a messy state). Those objectives, as such, may not be terribly controversial. For example, most participants appeared to agree that broad statements of principle are needed in guidelines to provide a sense of direction. However, some disagreement emerged in the roundtable concerning the extent to which it is then necessary to delve more deeply into potentially treacherous details, bearing in mind the risk of embedding today’s analytical methodologies which tomorrow may prove to be fallacious. The attractiveness and elegance of a 15-page text laying down broad principles must be evaluated in light of the real need of courts, NCAs, undertakings and their advisors to analyze concrete (or contemplated) and sometimes complex agreements and practices, and to make risky decisions. From that perspective, any danger of freezing misguided, ephemeral concepts in guidelines is perhaps addressed to some extent by the general nature of soft law itself:121 (i) while the Commission’s guidelines interpreting, e.g., Article 81 or Regulation 139/2004, are in some circumstances capable of having binding effects vis-à-vis the Commission itself,122 as noted above they have no autonomous binding force in other respects (though undertakings ignore them at their peril); (ii) other authoritative guidance 120 These aims were all mentioned implicitly or explicitly during the roundtable, but they were not necessarily formulated in the following way. 121 For discussion, see Håkon Cosma and Richard Whish, “Soft Law in the Field of EU Competition Policy”, 14 European Business Law Review 25 (2003), and references therein. More generally, see Linda Senden, Soft Law in European Community Law, Hart Publishing, 2004, chapter 6. 122 See Luis Ortiz Blanco et al., EC Competition Procedure, 2nd edition, Oxford University Press, 2006, pp. 32–34.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxvii

Introduction, Summary, Remarks

lxvii

(speeches, Annual Reports, etc.) can ameliorate errors or bring aspects of the guidelines up to date; and of course (iii) guidelines can be reviewed and amended through relatively quick procedures, although, for reasons of both (public and private) costs and legal certainty, that should not be done lightly. A number of other remarks from the roundtable discussion are also worth noting. Doug Melamed drew attention to one of Article 82’s notable idiosyncrasies, namely that it can only apply to action executed by a firm with a preestablished dominant position. The implicit point is that there is a gap in the law.123 To illustrate the issue it may be useful to recall the ECJ’s holding in 123 In Europe, the fact that, in contrast to US law, Article 82 does not cover the wilful acquisition of monopoly power in the absence of pre-existing dominance has been recognized at least since the 1960s. See, e.g., René Joliet, “Monopolisation et abus de position dominante”, 5 revue trimestrielle de droit européen 645 (1969) (going too far, however, and deducing, in an astoundingly misguided analysis at pp. 682–691, that: Article 82 does not cover exclusionary abuses even where a firm does have a dominant position; and that the provision establishes a dirigiste system of price supervision in respect of which any “agressive, coercitive ou déloyale” conduct is irrelevant). More recently, see, e.g., Michel Waelbroeck and Aldo Frignani, European Competition Law (trans. Françoise Gamet-Pol), Transnational Publishers, 1999, at p. 217. A gap under Article 82 with respect to concentrations creating as opposed to strengthening a dominant position (see Eleanor Fox, “Monopolization and Dominance in the United States and the European Community”, cited supra note 21, pp. 989–990) was partially (but not adequately) addressed by Article 81 jurisprudence, and then the matter became a historical footnote with the adoption of Regulation 4064/89 (later replaced by Regulation 139/2004). The gap to which Melamed refers has been criticized by Einer Elhauge, who opines that: “the U.S. approach reflects a much sounder policy. Illicit conduct that creates dominant market power leads to higher prices that are both avoidable and socially undesirable. It is thus important to condemn such conduct, and the failure of E.C. competition law to do so leaves an unsound gap in its regulation of anticompetitive behaviour.” Elhauge, “Defining Better Monopolization Standards”, 56 Stanford Law Review 253, 332 (2003). While there may be some validity to this view, the argument would have to be more carefully specified. First, in many if not most cases, unilateral attempts to exclude rivals by a firm lacking substantial market power are likely to meet with limited success. This is one reason why many of the Workshop participants (e.g., Bishop, Coscelli, Padilla, Vickers and others) prefer to maintain a preliminary dominance analysis to filter out cases of lesser merit. Furthermore, in cases of relational market power falling short of dominance, many national laws on unilateral conduct (though not all) can be applied to exclusionary behaviour. More fundamentally, one should guard against facile comparisons between Article 82 and Section 2 of the Sherman Act. In particular, the “gap” critique ignores a significant difference between those provisions in the level of market power required as a condition for intervention. Apart from relatively rare US cases where an attempt to monopolize claim is sustained (see Spectrum Sports, 506 U.S. 447 (1993)), the application of Section 2 normally tends to be limited to cases where the defendant commands at least 65–70% of the market (though some courts have entertained cases involving somewhat lesser degrees of market power—see, e.g., Langenderfer, 917 F.2d 1413 (6th Cir. 1990) (58% share of the market)). By contrast, it has not been uncommon for the Commission or other claimants to successfully pursue Article 82 cases against defendants controlling around half of the relevant market or even, in more dubious cases, against defendants with a share of less than half of the market (sometimes narrowly defined, at that) and falling further, at least where the defendant is an “obligatory trading partner” and where rivals, though expanding, are still much smaller. See Case T-219/99, British Airways v Commission, paras. 210 and 224 (upheld, Case C-95/04 P, supra note 37) and the Commission’s Decision of 14 July 1999 in the same case (paras. 36, 47, 88–94). In short, if the threshold for intervention in Europe is lower than it is in the US (as is also claimed, e.g., by Padilla and Venit in their written contributions), the “gap” critique would seem to be significantly overstated.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxviii

lxviii Introduction, Summary, Remarks the Viho case (concerning the distribution policy of the UK company Parker Pen), according to which the contested agreements could not trigger Article 81 despite the fact that they appreciably restricted intrabrand competition. While it was acknowledged that the agreements resulted in absolute territorial protection (insofar as Parker Pen required its subsidiaries in various Member States, when receiving inquiries from customers in other territories, to refer the customer to the home country distributor), they could not be qualified as “agreements between undertakings” within the meaning of Article 81 but were merely internal arrangements to allocate territories within a single economic unit.124 Hypothetically speaking, if that territorial partitioning were to result in the acquisition by Parker Pen of a dominant position (e.g., by denying rivals the possibility of reaching efficient scale), and if the restrictive covenants were thereupon immediately eliminated on a voluntary basis, there would appear to be no abuse under Article 82. In the imaginary situation just described, it might appear that this lacuna could only be filled by an amendment of the Treaty. However, from the debate later held in panel V, it seems that in such a scenario, if the hypothetically dominant Parker Pen proceeded to charge high prices, an excessive pricing case under Article 82(a) might be brought as an indirect means to address the underlying anticompetitive foreclosure. Another example of a gap case, cited specifically in Panel V,125 might be a “patent ambush” scenario, if a firm misleads a standard-setting organization, secures the inclusion of its IP in the adopted standard and thereby acquires a dominant position. The firm’s conduct is unilateral and thus escapes Article 81 because it involves no concurrence of wills; but Article 82 cannot touch it directly because the disagreeable act is carried out prior to the acquisition of dominance (and there is no “dangerous probability” doctrine under Community law). This kind of fact pattern may only arise on infrequent occasions, but on the heels of a well-known case in the US,126 similar behaviour is presently under investigation by the Commission.127 124

Case C-73/95 P, Viho Europe BV v Commission [1994] ECR I-5457, paras. 15–18. See the oral remarks of Emil Paulis, p. 477. 126 Rambus, Inc. v. Federal Trade Commission, 522 F.3d 456 (D.C. Cir. 2008) (FTC has petitioned the Court of Appeals to re-hear the case en banc: see http://www.ftc.gov/os/caselist/ 0110017/080606rambusrehearingpetition.pdf). In this case of alleged patent ambush, the Court of Appeals in its judgment of 22 April 2008 overturned the FTC’s finding of monopolization under Section 2 of the Sherman Act. According to the court, if deceit on the part of Rambus (i.e., duping JEDEC into standardizing its DRAM technology) merely enabled the company to charge higher prices for its technology than it otherwise could have charged (which was one of the alternative grounds for liability that had been advanced by the FTC), insofar as full disclosure would have caused the SSO to insist ex ante on a RAND commitment, this would be insufficient to sustain a monopolization claim. On that basis alone, the court remanded the case for further proceedings. According to the Court, “an otherwise lawful monopolist’s use of deception simply to obtain higher prices normally has no particular tendency to exclude rivals and thus to diminish competition”. Slip, p. 15. Moreover, in pointed dicta the court indicated its scepticism regarding the FTC’s evidence of deceit. The FTC’s case might have been stronger if the agency had established, on the basis of solid evidence (and if it had pinned down its theory of liability 125

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxix

Introduction, Summary, Remarks

lxix

Regarding the use of excessive pricing cases as a gapfiller to handle cases involving the improper acquisition of a dominant position through anticompetitive conduct, it seems likely that the standard of proof would be the same as in any other excessive pricing case. In other words, there is no free lunch since, even if it can be established that past misconduct resulted in the defendant’s acquisition of a dominant firm, this would not in itself support a claim under Article 82 unless all of the usual elements of an excessive pricing case (no reasonable relation to the economic value of the product; unfairness of the price) are established. To conclude otherwise would circumvent or at least weaken the dominance requirement, since artificial enforcement action in those circumstances would effectively punish unilateral conduct carried out by a pre-dominant company. One might therefore wonder whether the circumstances envisaged are really just those of a garden-variety excessive pricing case, and whether it is therefore claiming too much to suggest that any gap would be closed. However, what seems to emerge implicitly from the Workshop is that in this scenario the enforcer might actually bring the case, rather than abstaining as it might otherwise do (and letting the market evolve naturally) in situations where a firm, having acquired dominance through more tasteful means, proceeds to charge supracompetitive prices. A second point made by Melamed was that the principal objective of guidelines should be “to give useful guidance to businesses”, as opposed to providing a text designed to serve litigants and judges in court. However, there was a feeling among some other participants that the prevailing circumstances in Europe (including Regulation 1/2003, the incipient development of private enforcement, and the gradual absorption of new legal systems in the Balkans and elsewhere) justify guidelines that do address litigation issues. The dualobjective view, “they both matter”, was put forward succinctly by Bill Blumenthal. One could also imagine, if need be, two sets of guidelines for two partially overlapping but otherwise different audiences.

instead of arguing in the alternative), that the deceitful conduct had resulted in the standardization of Rambus’ technology instead of alternative technologies, or in other words, that such alternatives would have been preferred if there had been full disclosure. For discussion, see Herbert Hovenkamp, “Patent Continuations, Patent Deception, and Standard Setting: the Rambus and Broadcom decisions”, University of Iowa Legal Studies Research Paper 08-25 (June 2008), available at http://ssrn.com/abstract=1138002; Andrea Giannaccari, “La sentenza della Corte d’Appello nel caso Rambus”, forthcoming, Mercato Concorrenza Regole. 127 See Case COMP/38.636, Rambus, described in MEMO/07/330 of 23 August 2007. In its Memo, the Commission expresses the preliminary view that, if Rambus had not improperly secured a place for its DRAM technology in the relevant JEDEC standard by means of deliberate non-disclosure of its essential IP, the company “would not have been able to charge the royalty rates it currently does” (or indeed would have been unable to charge any royalties at all). The Commission’s preliminary analysis of the case thus appears to be based on a theory of exploitative abuse under Article 82(a), and the envisaged remedy would be compulsory RAND or FRAND-term licensing. See also Philip Lowe, “The Commission’s current thinking on Article 82”, speech delivered in Brussels, 31 January 2008, available at http://ec.europa.eu/comm/ competition/speeches/text/2008_01_en.pdf, pp. 7–8.

(AA) Ehlermann 08 Introduction

lxx

12/11/08

10:25

Page lxx

Introduction, Summary, Remarks

Melamed then added a third point concerning the continuum between general principles and detailed analytical tools. He stressed the need to avoid principles or safe harbours that distinguish between categories of conduct essentially because they create an inefficient straitjacket effect and have little utility in novel cases. This point struck two somewhat different chords among the participants. On the one hand, economists such as Patrick Rey, John Vickers and others strongly agreed that formal, artificial distinctions should be avoided where ostensibly different practices have equivalent effects. That is of course one of the rationales behind the turn to an effects-based approach. At the same time, it was also pointed out (see below) that in some cases it is useful to identify specific forms of conduct and to search for rules or safe harbours that have logical application given the nature and effects of the conduct. Damien Geradin noted that the modern European antitrust era is in some ways characterized by an attempt to make sure that Article 82 does not prohibit or chill efficient conduct, without letting go entirely of legal certainty. For this reason he favours what he considers to be the intermediate approach advocated in the paper by Ahlborn and Padilla. On the issue of broad principles and detailed tests and specific guidance, Geradin supports the first but is cautious regarding the second. As an example, he points out that the test proposed in DG Competition’s Discussion Paper for rebates granted by dominant firms was too complicated for most in-house counsel to apply. Geradin added a number of observations going beyond the question of guidelines and concerning institutional or process-related issues. These include case selection, inadequate staffing, excessively drawn-out investigations (and hence a need to stimulate early settlements), and the strategic use of public enforcement mechanisms to raise rivals’ costs. On the question of principles versus details, Jim Venit resolved the hesitation expressed by Geradin in favour of guidelines emphasizing broad, horizontal principles and avoiding excessive specificity. Like John Vickers, Venit regards the Non-Horizontal Merger Guidelines as a model to emulate. Venit finished by picking up on Geradin’s point about the gaming of competition law procedures, and he stressed the need for enforcers to be alert to the risk of being drawn into such intrigues. Emil Paulis considered that guidelines limited to general principles and a general framework would not go far enough. According to Paulis, it is necessary to flesh out how the Commission considers the principles should be interpreted (and applied), because this is what will help non-specialist judges in Europe. Paulis then used Geradin’s example of rebates to show that, even if the test proposed in the Discussion Paper is sophisticated, it is useful to articulate such a test because it highlights the inadequacies of the law on rebates as it stands (which may chill legitimate price competition) and it reveals the methodology by which the Commission aims to achieve a change of policy. The technical difficulty of a more nuanced rule, which may be misunderstood

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxi

Introduction, Summary, Remarks

lxxi

and misapplied in some instances, may be seen as part of the cost of shifting to an effects-based approach. Paulis then added a couple of points concerning Jochen Burrichter’s presentation on the implications of an effects-based approach for private litigation under Article 82. First, he rejected the seeming irony that while certain reforms of national procedures are supported by the Commission to ameliorate the “risk-reward” calculus for plaintiffs in antitrust litigation, the pending substantive reorientation of Article 82 enforcement might make life more difficult for plaintiffs. If the consequence of the private enforcement initiative is a reduction of Article 82 litigation than would otherwise be the case, then “so be it”, he says. The goal of the Commission’s present initiative to enhance the state of private enforcement in Europe is not to encourage more litigation as such, but rather to maximize effective overall compliance with competition law.128 On the other hand, Paulis was concerned about the information asymmetries between plaintiffs and defendants, which call for significant procedural (and cultural) reforms at the national level with a view to stimulating discovery and access to evidence. David Meyer added a gloss to Doug Melamed’s point about the need to avoid pigeonholing particular forms of conduct. For example, if a safe harbour is provided for above-cost pricing, this might indeed influence commercial decisions and might yield an outcome that the market alone would not have produced. Nevertheless, he says, the importance of such a bright-line freedom is sufficient to justify a rule that may in some cases distort choices. Meyer then seconded the comments of Paulis with respect to the need to articulate a safe harbour for rebates granted by a dominant firm. At a minimum (though he might go further), Meyer would support a safe harbour according to which, once the full amount of the rebate has been allocated to the contestable share of products supplied by the dominant firm, if the effective price is still above an appropriate measure of cost it should be presumed to be legal. Mark Patterson took the recoupment criterion under US predatory pricing law as an example of how rules on the burden of proof can introduce distortions. The need to demonstrate a likelihood of recoupment as a pre-condition for liability is tantamount to a presumption that a dominant price discounter acts irrationally.129 Implicitly, Patterson suggested that it might be better to expect a defendant to show why a predatory scheme would fail rather than to expect a plaintiff to show why it would succeed. This points to his fundamental critique of US law in this area, namely that in many cases defendants 128 Parallel to this utilitarian perspective, there is of course a rights-based rationale. However, in Courage and Manfredi, and indeed in Van Gend en Loos itself, the effectiveness side of the coin predominates. For a different and well-argued view, see Paolisa Nebbia, 33 EL Rev 23 (2008). 129 That presumption may be seen as essentially a legal fiction driven by a policy choice that attaches high value to aggressive price competition despite the fact that plaintiffs might thereby be disadvantaged. As in other contexts, the development of substantive liability standards under Section 2 might reflect an “equilibrating tendency” given the perceived severity of treble damages.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxii

lxxii Introduction, Summary, Remarks are liable to have superior information and evidence with respect to their commercial strategies. Instead of having blanket standard of proof rules, Patterson favours a more variegated approach that takes account of the information asymmetries mentioned above and tailors those rules accordingly. Bucking the trend apparent from criticism of British Airways, Michelin II and other cases, Patterson considers that the European approach to fidelity rebates, where defendants are expected to come forward with convincing justifications for their pricing systems, is more sensible than a system that expects the evidence to be adduced by the party least likely to have it. Sensing a need to dispel a common perception or suspicion, Emil Paulis stressed that the reform of Article 82 enforcement should not be misinterpreted as a “get out of jail free” pass for dominant companies engaging in exclusionary conduct. The reform is not about placating monopolists but about paying more attention to consumer interests. Not only is the reform not an endorsement of a “light touch” regulatory philosophy; Paulis also underlined the need for particular vigilance in liberalized network industries. In these industries, residual, de facto monopolies on the one hand might themselves represent a lack of effective competition on the dominated markets, and on the other hand they present leveraging opportunities in emerging markets, all of which tends to limit the benefits of liberalization. Coming back to the more general debate concerning what the content of guidelines should be under Article 82, Paulis insisted on clarity as regards not just what is permitted but also what is proscribed. Einer Elhauge agreed with Paulis that guidelines should pass beyond generalities, not least because they can engender conflicting interpretations. Amelia Fletcher too felt that guidelines have to be detailed and specific with respect to the relevant competitive analysis, bearing in mind the need for consistent treatment of practices with equivalent effects. She also made certain suggestions concerning the style in which guidelines might be written. Taking a cue from John Vickers, she noted that for purposes of tone it might be useful to begin with the reasons why certain practices can produce efficiencies, and only then discuss their darker side.130 Furthermore, Fletcher distinguishes between “hard” safe harbours and “soft” safe harbours, that is to say between those safe harbours that apply without exception and those qualified by words such as “Typically, the Authority will not intervene in cases where . . .” Fletcher prefers soft safe harbours because they maintain the possibility for the enforcer to take action at the margins. On this point, David Meyer noted that in fact guidelines in the US often do couch safe harbours in “soft” terms to avoid undue rigidity. Although Meyer would not deny the value of soft safe harbours, he did underline the tradeoff between the benefits of hedging to ensure an enforcer’s discretion and the corresponding costs in terms of legal certainty. 130 For example, this kind of approach was liberally applied in the Commission’s Technology Transfer Guidelines, [2004] OJ C101/2.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxiii

Introduction, Summary, Remarks

lxxiii

Doug Melamed acknowledged Einer Elhauge’s point that broad, lofty principles are susceptible of conflicting interpretations, but he emphasized that “general principles” need not mean “vacuous aspirations”. The pitfall to avoid, he stressed once again, was the tendency to pigeonhole certain kinds of conduct in a manner that artificially channels commercial behaviour. It can be inferred from his remarks that his emphasis on general principles would not exclude appropriate adaptations. For instance, if an as-efficient competitor test is adopted as a unifying approach to the analysis of exclusionary conduct (note that this test is used only as an example; it is not Melamed’s preferred test), the application of that test will have different nuances depending on whether the factual context involves fidelity rebates, predatory pricing, or margin squeeze. The existence of such nuances does not alter the fact that what the test seeks to achieve is consistent in each case. If that is a plausible interpretation of Melamed’s remarks, it would be roughly consistent with David Meyer’s observation that a framework test with broad application can be “followed through” and applied to particular kinds of cases. Jochen Burrichter emphasized the utility of guidelines as a means to facilitate national antitrust litigation, and noted the important role that other guidelines issued by the Commission (especially the Guidelines on Vertical Restraints) have played, in his experience, before national courts. Burrichter also defended his proposals to introduce certain procedural devices such as burden-shifting rules on the ground that plaintiffs simply do not have any hope, at least in the short term, that civil procedure rules in Continental Europe will be transformed by the introduction of adequate discovery mechanisms. John Fingleton made a number of thoughtful remarks to bring the roundtable to a close. Perhaps most importantly, he pointed out that what goes into guidelines should not be dictated by majority rule. Rather, the content should consist of an organic system that matures and emerges from the sharing of experiences and lessons, and from debate and reflection.

2. Session Two: A Reformed Approach to Exclusionary Conduct 2.1 Panel III (papers by Bishop, Allendesalazar and O’Donoghue) A. Written Contributions 1. Simon Bishop’s paper, Loyalty Rebates and “Merger Standards”: A roadmap for the practical assessment of Article 82 investigations, is motivated by a clear agenda: meaningful reform of Article 82 enforcement on the basis of

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxiv

lxxiv Introduction, Summary, Remarks meaningful effects-based analysis. Only a brief summary of this rich paper will be presented here. The first observation is that, like other commentators, Bishop seeks to wrest Article 82 away from old habits of mind which equated the marginalization of rivals with harm to competition and which sometimes harmed consumers by adhering (de facto) to over-deterrent rules of illegality.131 The guiding question, in Bishop’s view, is the following effects-oriented query: “are consumers made worse or better off by a given practice, taking into account the dynamic nature of competition?” However, Bishop recognizes that effects on consumers do not easily lend themselves to direct measurement. He therefore suggests a two-track way to get there: first, he says it’s necessary to ratchet up the empirical rigour so that the conduct in question may be evaluated in the light of the best available market evidence; second, single-firm conduct analysis should be brought into line with the standard assessment of non-horizontal mergers and of vertical restraints. Such an alignment, he explains, would transform the traditional tendency to presume exclusionary conduct by a dominant firm to be per se anticompetitive into a presumption of commercially efficient “share-shifting” that benefits consumers. Bishop then enters into an in-depth critique of the traditional treatment of loyalty rebates as a case study by which to drive home the foregoing general points, and to raise a few additional ones. Among other observations, Bishop criticizes the traditional approach for its undue emphasis on the customer’s last unit(s) of purchase. Bishop describes it with more sophistication, but the point seems to be that focusing on the last unit tends paternalistically to preempt a customer’s own ex ante judgment about whether to offer up its fidelity to the defendant supplier. He also calls attention to several potential efficiencies that may arise from loyalty rebate schemes. Furthermore, he calls on the Commission and all the other interested parties to take a hard look at dominance. He points to the troubling fact that, once the defendant has been found dominant, that finding has often determined the outcome of the abuse assessment—and yet sometimes there is an underlying uncertainty with regard to whether the defendant is indeed dominant, due in particular to ambiguous market definitions.132 Given its limitations, Bishop insists that dominance needs to be reined in and limited to its proper role, i.e., that of a screen by which to filter out bad cases. He is concerned by the relative lack of attention given thus far to the role of dominance in the context of the Article 82 policy review because the success of an effects-based assessment of abuse—the whole point of the exercise—might be frustrated if such issues are ignored. 2. Earlier it was suggested that, if the “special responsibility” doctrine is no more than an elegant statement of the truism that a firm is subject to abuse 131 Of course, per se rules (or quasi-per se rules) have not been applied under Article 82 to all forms of potential abuse (a clear example being refusals to start supplying third parties). 132 Cf. also the discussion of dominance by Eleanor Fox in her paper.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxv

Introduction, Summary, Remarks

lxxv

control under Article 82 once it enjoys a dominant position in the internal market or a substantial part of it, then the doctrine has a weak raison d’être and should be discarded due to its tendency to promote confusion and distorted theories (see supra note 52). Rafael Allendesalazar launches a direct attack against the doctrine in a paper largely summed up by its title, Can we finally say farewell to the “special responsibility” of dominant companies? The straightforward contention behind this paper is that the doctrine “neither helps to recognize when conduct will be considered abusive nor explains why it will be prohibited”.133 Allendesalazar adds, inter alia, that by removing certain conduct from the range of legitimate options open to a dominant company when the same behaviour could be adopted unobjectionably by a non-dominant firm, or by requiring a dominant company to abstain from granting certain rebates even where they are desired by clients, the “special responsibility” doctrine paradoxically requires the dominant firm to act “independently”, à la Hoffmann-La Roche. One might respond that this is of little moment, since “independence” (including the economic strength to impose prices above the competitive level) is merely the legal criterion for assessing dominance and cannot by itself trigger Article 82. Nevertheless, Allendesalazar’s point is an important one: if other undertakings in the relevant market routinely adhere to certain practices, or if clients seek certain pricing arrangements, then although such factors cannot immunize abusive conduct, they seem to merit more probative value than has thus far been attached to them. Allendesalazar also notes optimistically that the “special responsibility” doctrine seems to be on the decline in the relevant case law,134 and that it was justifiably ignored in DG Competition’s Discussion Paper. The rest of the paper exhorts the Commission to boldly blaze new trails across the virgin, effects-based land, even if this means leaving behind the comforts and amenities of the legal terra cognita. In addition (and as Bishop does in his paper), Allendesalazar calls for more systematic and prompt publication of Commission decisions that reject complaints (rather than the general practice of publishing only decisions finding infringements) in order to help illuminate the new landscape. 3. Robert O’Donoghue’s contribution, Verbalizing a General Test For Exclusionary Conduct Under Article 82 EC, provides a useful round-up of the various liability tests that have been proposed for exclusionary conduct under Article 82 and/or Section 2 of the Sherman Act, and it furthermore promotes a differently formulated test drawn from the language of Article 82 itself. 133

Page 320. The coup de grace has not yet landed. However, recent cases seem to give the doctrine only passing attention. See Case T-201/04, Microsoft v Commission, judgment of the CFI of 17 September 2007, not yet reported, paras. 229 and 775; Case T-271/03, Deutsche Telekom AG v Commission, judgment of the CFI of 10 April 2008, not yet reported, para. 122. 134

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxvi

lxxvi Introduction, Summary, Remarks Following a general discussion regarding the design of standards, in which O’Donoghue emphasizes the progressive development of the case law under Article 82 and concludes that there are now “virtually no practices that could be described as per se unlawful under Article 82”,135 he surveys the most commonly advocated liability tests: the profit sacrifice test and its close relative, the no economic sense test; the equally-efficient competitor test; and the consumer welfare test. O’Donoghue discusses not only the elements of these tests but also the various criticisms aimed at them. At least one horizontal lesson seems to emerge from the survey, namely, that it would be a mistake to suppose that any of these tests can be applied indiscriminately, without respect to the circumstances of the case at hand. Other problems with these tests include their practical utility in the daily decision-making of the marketplace. A pure rule of reason approach, for example, whereby the dispositive factor is net harm or benefit to consumers, is often regarded as placing unrealistic demands on undertakings because they lack sufficient information, ex ante, to be able to carry out the necessary balancing exercise. The shortcomings of the tests on offer lead O’Donoghue to the conclusion that the best alternative is to rely on the language of Article 82(b). That provision states that abusive behaviour might consist of: “liming production, markets or technical development to the prejudice of consumers”, provided there is an effect on trade between Member States. According to O’Donoghue, Article 82(b) “captures the key feature of exclusionary conduct, namely, that it makes competitors’ products or services less attractive or less available, rather than simply making the dominant company’s product better or more available”.136 Furthermore, such a standard would ensure that only conduct that restricts output would be caught, and then only if harm to consumers is established. Among its other advantages, O’Donoghue points out that the limiting production test is a flexible one that could in fact be corroborated by applying the other tests as and when needed. To illustrate, O’Donoghue runs through the application of the Article 82(b) test to strategic withdrawal of marketing authorizations in the pharmaceutical sector. O’Donoghue also takes the opportunity to record his disapproval of DG Competition’s proposal to read an “Article 82(3)” efficiency concept into the un-bifurcated Article 82. Presenting a number of arguments, O’Donoghue concludes that “the Commission would appear to have followed the wording of Article 81(3) for no good reason other than general notions of consistency”.137 The proposition that the dissimilarities between Article 81 and Article 82 decisively preclude the two-stage analysis of the former from being transposed to the latter seems consistent with the famous Opinion of

135 136 137

Page 332. Page 343. Page 352.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxvii

Introduction, Summary, Remarks

lxxvii

Advocate General Jacobs in Syfait.138 The Advocate General might have agreed that an Article 81(3)-style analysis of efficiencies is inappropriate under Article 82, although that is not altogether certain.139 However, one could take a different point of view on the ground that the first three conditions of Article 81(3) have a fairly strong rationale, and they might even be regarded as general principles,140 with consequences for Article 82 despite their obvious absence from that provision. Of course, that possibility may also be denied on the basis of the literal text.141 As for the fourth condition of Article 81(3), this condition has often been (mis)understood as precluding the 138 In taking the view that the very specific characteristics of the pharmaceutical industry suggest that a refusal to supply by a dominant firm might in some circumstances be consistent with the firm’s legitimate interests and hence might not constitute an abuse even where the refusal is intended to partition markets, the Advocate General explained: “the two-stage analysis suggested by the distinction between an abuse and its objective justification is to my mind somewhat artificial. Article 82 EC, by contrast with Article 81 EC, does not contain any explicit provision for the exemption of conduct otherwise falling within it. Indeed, the very fact that conduct is characterised as an ‘abuse’ suggests that a negative conclusion has already been reached, by contrast with the more neutral terminology of ‘prevention, restriction, or distortion of competition’ under Article 81 EC. In my view, it is therefore more accurate to say that certain types of conduct on the part of a dominant undertaking do not fall within the category of abuse at all.” Case C-53/03 Synetairismos Farmakopoion Aitolias & Akarnanias (Syfait) and Others v Glaxosmithkline AEVE [2005] ECR I-4609, para. 72 (emphasis supplied). 139 The “objective justification” in Syfait related essentially to factors that could be externally observed: the economics of the pharmaceutical industry, the effects on parallel trade and so on. Those kinds of factors seem amenable to a one-stage, or “integrated”, evaluation. By contrast, where the claimed efficiencies cannot reasonably be observed (which might be the case, for example, where the defendant expects to generate cost efficiencies), one could argue that Advocate General Jacobs’ reasoning cannot be easily transposed. What the Advocate General seemed to be saying was that, in light of the distortions and fragmentation already characterizing the market, the dominant firm’s policy of limiting parallel imports from Greece was unlikely to have significant additional market-dividing effects. Under those circumstances, the firm’s efforts to protect its commercial interests could be said to be objectively justified, which is equivalent to saying that they were beyond Article 82’s reach. One could conceivably distinguish this situation from one where there are established foreclosure effects which must be weighed against any efficiency gains generated by a dominant firm’s behaviour. 140 Support for that claim might be found in paragraph 86 of British Airways, cited supra note 37, where the ECJ plainly envisages the application of the concepts of pass-on and proportionality in the context of an “objective economic justification” analysis in Article 82 cases. 141 A literal interpretation may be persuasive, but probably not decisive. Note, for example, that Article 82 does not provide for a nullity sanction either, as Article 81(2) does, and yet the Court of Justice has recognized that, in national litigation, agreements contrary to that provision may be declared void and unenforceable. See Case 127/73, Belgische Radio en Televisie and Others (BRT) v SV SABAM and Fonior NV [1974] ECR 51, para. 15; Case 66/86, Ahmed Saeed Flugreisen and SilverLine Reisebüro GmbH v Zentrale zur Bekämpfung unlauteren Wettbewerbs eV [1989] ECR 803, paras. 32–33. Admittedly, the position is not quite the same as automatic nullity, subject to possible severance of illegal terms, as provided for by Article 81(2). Nevertheless, as noted by Whish, where an agreement amounts to an abuse of dominance, “it has been assumed that the prohibition of Article 82 means that the offending provisions are void”. Richard Whish, Competition Law, 5th edition, Oxford University Press, 2005, p. 296. See also Viven Rose and Peter Roth, eds., Bellamy and Child: European Community Law of Competition, 6th edition, Sweet & Maxwell, 2007, p. 1431, footnote 300 (interpreting cases such as BRT as implying that “agreements infringing Art 82 are unenforceable by virtue of the direct effect of that provision” and citing national jurisprudence to that effect); Jochen Burrichter, this volume, p. 243 (pointing out that most national laws provide for the nullity of any acts in breach of Article 82).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxviii

lxxviii Introduction, Summary, Remarks successful invocation by a dominant firm of an individual exemption under Article 81(3), and that understanding may have spilled over to the hypothetical discussion of the transposition of the four conditions to the Article 82 context. However, supposing that certain threshold questions were resolved, such as whether the “Article 82(3)” concept does too much violence to the text of Article 82, it would obviously defeat the purpose of such an approach if the “no substantial elimination of competition” criterion were incapable of being satisfied where the defendant is dominant. To that extent, just as in the context of Article 81(3), the residual competition criterion needs to be interpreted and applied with sensitivity to the facts of the case, bearing in mind the ultimate aim of that criterion, which is to preserve the “competitive process”. In that regard, it is useful to recall that there is good foreclosure and bad.142 It might be interesting to consider the residual competition criterion in its qualitative dimension as well as its quantitative dimension. In any case, even if it is strictly a quantitative question of how much competition is eliminated, one can imagine various scenarios in which efficient conduct by a dominant firm with incidental exclusionary effects does not result in substantial elimination of competition in respect of the products concerned. Notwithstanding those observations, at least two objections to the “Article 82(3)” approach remain.143 If Article 82 of the Treaty and Article 2 of Regulation 1/2003 are taken literally, then one might claim that such an approach would impermissibly shift the legal burden of proof to the dominant defendant. This would be inconsistent with the current “objective economic justification” analysis, whereby only the evidential burden is borne by the defendant144 while the ultimate legal burden (i.e., the burden of persuasion) remains always with the enforcer or complainant.145 There is no doubt that, under Article 82, the legal burden of proving countervailing efficiencies 142 This point was widely recognized among the participants, for example by John Vickers, Andrea Coscelli and others. 143 This paragraph was written following an exchange of thoughts with Richard Whish. His comments are greatly appreciated. If the text is blighted by errors, I alone am to blame. 144 See Case T-201/04, Microsoft v Commission, judgment of the CFI of 17 September 2007, not yet reported, para. 688 (and the identical para. 1144): “[A]lthough the burden of the existence of the circumstances that constitute an infringement of Article 82 EC is borne by the Commission, it is for the dominant undertaking concerned, and not for the Commission, before the end of the administrative procedure, to raise any plea of objective justification and to support it with arguments and evidence. It then falls to the Commission . . . to show that the arguments and evidence relied on by the undertaking cannot prevail and, accordingly, that the justification put forward cannot be accepted.” 145 For example, see Giorgio Monti’s comments on DG Comp’s Discussion Paper, 2006, at page 4: “I do not think that Article 82 admits defences in the way they have been set out [in Section 5.5 of the Discussion Paper]. The Article 82 ‘defences’ that have been established by the case law suggest that on certain facts there is no abuse. Thus, there is no shift in the burden of proof . . .” See Monti’s comments at http://ec.europa.eu/comm/competition/antitrust/art82/ 065.pdf. Similarly, see Albertina Albors-Llorens, “The Role of Objective Justification and Efficiencies in the Application of Article 82 EC”, 44 Common Market Law Review 1727, 1754–1755, 1760 (2007).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxix

Introduction, Summary, Remarks

lxxix

cannot shift to the defendant.146 It is also clear that neither Article 82 nor Article 2 of the Regulation contemplates any defence where an abuse is established. Yet the proposition that the efficiency analysis under Article 82 should be consistent with the familiar structure of Article 81(3) does not necessarily imply any shift of the legal burden of proof. A dominant firm could simply be required to come forward with evidence of the usual probative elements (quantitative or qualitative efficiencies, proportionality and consumer passon) in order to cast doubt on a prima facie inference of abuse. If that approach were adopted, the burden of persuasion would remain the responsibility of the enforcer or complainant. For that matter, despite the impression given by Article 2 of Regulation 1/2003 and the corresponding Recital 5, one may wonder whether Article 81(3) itself is different from that proposed model. That is to say, the ultimate burden of demonstrating an infringement of Article 81 as a whole is arguably borne by the enforcer or claimant, irrespective of the fact that it is for the defendant to substantiate its claim that the conditions for the exemption are satisfied. On that view, it would be erroneous to regard Article 81(3) as an affirmative defence.147 In short, whereas one might plausibly object, on Jacobs-like reasoning, that Article 82 does not admit a two-stage analysis, the objection that an “Article 82(3)” analysis is precluded because it shifts the legal burden of proof to the dominant defendant may not be tenable. It is also worth considering a second possible objection. As Pierre Larouche has suggested, any effort to promote efficiency-based justifications under Article 82 may be undermined if the efficiency analysis is postponed until

146

See Renato Nazzini, “The Wood Began to Move”, cited supra note 118. There is case law which might be interpreted consistently with the suggestion made here. See, e.g., Cases 43/82 and 63/82, VBVB and VBBB v Commission [1984] ECR 19, para. 52 (“[I]t is in the first place for the undertakings concerned to present to the Commission the evidence intended to establish the economic justification for an exemption and, if the Commission, has objections to raise, to submit alternatives to it.”); Case 42/84, Remia and Verenigde Bedrijven Nutricia v Commission [1985] 2545, para. 45. More recently, see Case T-168/01, GlaxoSmithKline Services Unlimited v Commission [2006] ECR II-2969, para. 236 (“As the Commission agrees in its written submissions,” where the parties come forward with arguments and evidence that the exemptions conditions are satisfied, the Commission “must refute those arguments and that evidence”). The CFI’s judgment has been appealed: see Cases C-501/06 etc., not yet decided. The burden of proof in Article 81 cases is discussed by Lars Kjølbye, Ali Nikpay and Jonathan Faull, “Article 81”, in Jonathan Faull and Ali Nikpay, The EC Law of Competition, 2nd edition, OUP, 2007, p. 292. The authors do not distinguish between the evidential burden of proof and the legal burden of proof; however, that distinction does seem to be implicit in their description of the procedure, according to which, if the Commission fails to refute the parties’ evidential showing, “it may be concluded that the burden of proof borne by the person who relies on Article 81(3) has been discharged”. Ibid. (echoing Joined Cases C-204/00, etc., Aalborg Portland, [2004] ECR I-00123, para. 79). In other words, the Commission in this case had failed to discharge its burden of substantiating the clamed infringement of Article 81 as a whole. Article 2 of the Regulation may not be drafted in a manner sufficiently clear to indicate that the enforcer or complainant bears the ultimate burden of proof as regards not only Article 81(1) but also Article 81 considered more broadly as an integrated norm. 147

(AA) Ehlermann 08 Introduction

lxxx

12/11/08

10:25

Page lxxx

Introduction, Summary, Remarks

the tail-end of the competition analysis.148 As he explains, “[o]nce the Commission is satisfied that the defendant holds a dominant position and that its conduct constituted an abuse, i.e., produced anti-competitive effects to the detriment of consumers, it is difficult to see how a defendant could turn the tide by arguing that in the end the course of conduct does create efficiencies which actually benefit consumers”.149 To that extent, any discussion of efficiencies becomes “practically pointless” or a “hopeless exercise”. This point inspires sympathy for what Damien Neven in his oral presentation called an “integrated” competition assessment.150 On the other hand, one may also question whether an efficiency justification styled after Article 81(3) is really doomed to fail, as it remains essentially untested. As with Article 81(3) itself, it may be that the effectiveness of such an approach depends ultimately on whether there is a genuine commitment to carrying out the analysis of efficiencies with an open mind and with a scrupulous effort to avoid confirmation bias (i.e., “anticompetitive effects likely; therefore, efficiencies unlikely”). The foregoing discussion indicates that further research and debate might be helpful to better evaluate the advantages and disadvantages of the notion of an “Article 82(3)”. Some authors have already given it a rather cool reception or at least expressed reservations.151 It remains to be seen whether the idea has any life left in it.

B. Discussion The roundtable discussion following the third panel of speakers (Bishop, Elhauge and Allendesalazar152) focused on two broad themes. First, there was the matter of what constitutes significant and thus cognizable foreclosure in the context of rebates, a question predicated on the distinction mentioned a moment ago between “good” and “bad” foreclosure. A number of sub148 Pierre Larouche, “The European Microsoft case at the crossroads of competition policy and innovation”, TILEC Discussion Paper DP 2008-021, available at http://arno.uvt.nl/show. cgi?fid=78001, pp. 13, 19–20 and 23. Larouche points out at p. 12 that Microsoft’s arguments based on the adverse impact that antitrust liability would have on its incentives to innovate “is not so much an ‘efficiency defence’ as a component of a proper competitive analysis”, in particular because there were no specific efficiency gains (here he cites the avoidance of free riding or amelioration of asymmetric information as examples) to be evaluated. 149 Ibid., p. 13. 150 See page 275. 151 See, e.g., Ekaterina Rousseva, “The Concept of Objective Justification”, cited supra note 117, pp. 31–32 and 55–72. However, Rousseva does recognize the shared objectives of Articles 81 and 82 and therefore recognizes the need to ensure, for example, that any approach under Article 82 designed to take account of efficiencies (which might be done, she proposes, by tailoring the definition of abuse or that of its flipside, “normal competition”/competition on the merits) must include proportionality and consumer pass-on elements. These elements, of course, are inspired by the second and third conditions under Article 81(3). 152 Original written contributions were received from Bishop and Allendesalazar. Their papers are included in this volume at pp. 287 and 319, respectively.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxi

Introduction, Summary, Remarks

lxxxi

sidiary points related to that issue were also discussed. The second principal theme concerned the broader question of how single-firm conduct standards can be designed and applied in a manner apt to cope with the complexities of exclusionary conduct in general. Identifying significant foreclosure in the context of rebates (and related issues) Eleanor Fox kicked off the discussion by questioning whether naked foreclosure might not be contrary to the European model even where it is not substantial.153 She also pointed out that conduct might be meritorious not only where it generates “efficiencies”, which is the usual locus of attention, but also where the conduct, though it tends to foreclose rivals, yields a greater degree of “competition”. A possible parallel to Fox’s “pro-competitive” justification might be, for example, joint ventures which open up new markets. The benefit to society is transaction-specific if none of the joint venture partners would have been able to make the necessary investments acting solo. Ian Forrester noted that “[e]veryone around the table agrees that the current [European] rules on rebates are an embarrassment”,154 and that reform in this area is long past due. As a matter of priority, therefore, the Commission should take a decision (whether it be a prohibition decision or a decision rejecting a complaint) and set forth its position as to what the law should be. A final judgment on appeal would then go some way toward providing badly needed legal certainty. Philip Lowe essentially agreed with Forrester that it is important for the Commission to take the lead as the trustee of the EU’s competition policy. However, he pointed out that a final judgment by the ECJ takes many years of litigation, and the Commission must also show policy leadership in the meantime, with the hope that the twain will meet. Andrea Coscelli addressed two different points. First, whereas Simon Bishop in his presentation had criticized British Airways for condemning the defendant’s behaviour as exclusionary even where rivals’ shares grew over time, Coscelli took issue with this, pointing out (as the jurisprudence has long done) that a rival’s (but-for) growth rate may indeed be held back by the dominant firm’s conduct. Second, Coscelli agreed with John Vickers (and others) that it is necessary to recognize that there is both good and bad foreclosure. However, competition authorities are not blind to this, he noted, and they may well routinely (and below the public’s radar) screen out cases involving only the former.

153 There is undoubtedly a quantitative difference between the words “significant” and “substantial”, although they are sometimes loosely treated as interchangeable; but Professor Fox adopted the word that had been used by Einer Elhauge in his presentation. 154 Page 267.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxii

lxxxii Introduction, Summary, Remarks Amelia Fletcher first of all agreed with Andrea Coscelli that a proper counterfactual may indeed indicate that competition has been harmed even where rivals are growing. Her second point related to a distinction made by economists between “deterred” entry and “blockaded” entry,155 the implication being that some economists might be more concerned where an incumbent acts deliberately to obstruct new entry (deterred entry) than where the incumbent is merely “doing what is does” to reap the greatest possible profits and the market is sealed off incidentally due to that profit-maximizing behaviour (blockaded entry).156 Since “abuse” under Article 82 is an objective concept (although intent does sometimes have a role to play, such as in predation cases), one might say that it applies equally to deterred entry and obtain entry. However, that does not seem to be a satisfactory position: a dominant firm might maximize its profits by offering superior products, or offering products at attractive prices. Or a dominant firm might innovate and obtain a patent with preclusive effects. In each of these cases, entry is blockaded, and in none of them (apart from an essential facilities scenario) could there be an actionable abuse. On the other hand, the distinction between deterred and blockaded entry might be just as blurry, at times, as the distinction between desirable aggressive competition and abusive competition that steps over the imaginary line. It is precisely in those situations in which blockaded and deterred entry appear indistinguishable that closer inquiry is appropriate. Fletcher raises the following question: “I wonder whether the special responsibility doctrine that we have in the EU is really about blockaded entry. A dominant firm certainly shouldn’t deter entry. But should it be held to a higher standard—should it think about not even maximizing its profits if this is necessary to avoid blockaded entry?”157 This question suggests yet again that the “special responsibility” doctrine promotes confusion. In those cases where blockaded entry clearly does not rise to the level of deterred entry—for example, where a dominant firm offers a superior product that cannot be

155 See Joe Bain, Barriers to New Competition, Harvard University Press, 1956, pp. 21–22. Bain also referred to “accommodated” entry, where the option of deterring entry would be too costly for the incumbent. 156 There seem to be parallels between this distinction and the “performance competition/impediment competition” dichotomy referred to earlier in this chapter, at least if impediment competition is defined functionally rather than as a category of specific forms of exclusionary or predatory conduct. 157 Page 268. In 2005, Fletcher indicated that insisting on such a higher standard would be illadvised: “[W]here entry is blockaded in this way [i.e., via mixed bundling to maximize short-run profits], the ‘special responsibility’ arguably forces firms to act against their own short-run selfinterest. This is risky, even when applied to dominant firms, not least because it provides little guidance about what firms are supposed to do instead. It can force firms to engage in inefficient pricing behaviour that is in fact bad for consumers, can give rise to an unacceptable degree of legal uncertainty, and is unlikely to promote EU competitiveness, efficiency or growth.” Amelia Fletcher, “The reform of Article 82: recommendations on key policy objectives”, speech at the Competition Law Forum, Brussels, 15 March 2005, p. 8. Later in the debate at the Workshop, Fletcher reiterated that she was not a “fan” of the special responsibility concept.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxiii

Introduction, Summary, Remarks

lxxxiii

matched by rivals (and hedging, again, for cases of essential facilities)—the special responsibility doctrine and Article 82 itself are both immaterial.158 Damien Geradin queried which of two characterizations best fit the jurisprudence on exclusionary rebates under Article 82. The first of these is that loyalty-inducing rebates granted by dominant firms have been treated almost as if they were “hard core” restrictions of competition, in that they have been subject to a nearly irrebuttable presumption of abuse. The ECJ took an important step forward by announcing in British Airways that exclusionary conduct might be justified by efficiencies, but it remains to be seen how far the door has been opened. According to the second, more benign reading of the relevant case law (espoused by Einer Elhauge and others), the outcome in Michelin II, for example, would have been different if the defendant had simply offered a reasonable business justification for its discounting practices. In other words, on this view, there has always been a “rule of reason” under Article 82, but it is a truncated one which procedurally must be activated by a plausible story articulated by the defendant. As a separate point, Geradin added his name to the list of those who would kill off the “special responsibility” doctrine. Patrick Rey addressed Geradin’s question by suggesting that any willingness by courts to entertain efficiencies in connection with rebates may be frustrated, ironically, by the fact that since efficiencies arise for exactly the same reason that these rebates are regarded as exclusionary—i.e., their fidelityenhancing nature—defendants are disinclined to acknowledge them, let alone showcase them. Elhauge’s simple solution to this is that defendants should be allowed to argue in the alternative, with no adverse inferences. Rafael Allendesalazar took the view that, while it might be legitimate to expect the defendant to offer a justification for its commercial behaviour, this cannot be legally required until after an initial stage in which the enforcer or plaintiff specifies how the defendant’s rebate program generates anticompetitive effects. Simon Bishop echoed this point and stressed the danger of too hastily shifting the burden, by which he may mean the danger of chilling legitimate commercial conduct. Robert O’Donoghue also stressed that Article 82

158 Put crudely: by maximizing its profits but doing no more than that (as opposed, say, to amassing excess capacity to dissuade challengers), and assuming its prices are not “excessive” (which should not simplistically be equated with a monopoly price—see van der Woude, this volume), a dominant firm is simply exercising its market power, a prerogative consistent with Article 82. Cf. Opinion of Advocate General Kirschner in Case T-51/89, Tetra Pak Rausing, [1990] ECR II-309, para. 68. If the market power of the dominant firm gives it an advantage in the sense that it can more easily introduce superior products due to, say, brand recognition, and if it thereby incidentally diverts market share to itself to the detriment of rivals, it is difficult to see how Article 82 could impose a “special responsibility” on the firm—even if the market structure is admittedly already “weakened”—to hold itself in check and behave according to a higher standard of civility. The area in which pure blockaded entry might warrant a closer look would be where the firm in question acquired dominance not through skill, foresight or industry but through the fruits of corporatist relations with national or local authorities.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxiv

lxxxiv Introduction, Summary, Remarks is a “system of prohibition” (the implication being that it is not a system of exception légale, as exemplified by Articles 81(1) and 81(3)), and the enforcer must therefore show an abuse—whereupon O’Donoghue would reject the notion of imposing a burden on the defendant to demonstrate such things as indispensability. Diverging somewhat from the foregoing views, as one might expect, Emil Paulis stressed that if a defendant does not have the obligation to rebut a showing of harm by specifying how the claimed efficiencies outweigh the adverse effects of the practice, the enforcer will not have sufficient information to carry out its assessment. Shifting gears from procedure to substantive policy, John Vickers described a hypothetical case of mixed bundling to present a stylized dilemma. According to this hypothetical, A and B are each more cost-efficient than incumbent C as regards the manufacture of their respective (and distinct) products. However, unlike A and B, incumbent C bundles the two products and thus achieves pro-consumer scope economies. This is a good example of the concept of “blockaded” entry (or rather, in this case, blockaded expansion) referred to by Amelia Fletcher (see supra page lxvi) because C exploits a commercial opportunity on the basis of its understanding of consumer demand. C’s mixed bundle unquestionably has an exclusionary/ marginalizing effect,159 but it does not appear to constitute deliberate obstruction of the market. Does (should) Article 82 prohibit C from its profitmaximizing behaviour in this case? One answer might be: “It depends; how much foreclosure is caused by the mixed bundle?” Other, more principled answers are possible depending on how one conceives of the proper objectives underlying Article 82. For example, drawing on the reflections above, one might say that if what C does is clearly a case not of deterred entry but of blockaded entry, then the enforcer should stand down. By doing so, as Vickers suggests (though one may object on “competitive process” grounds), consumers would appear to be better off. Patrick Rey reinforced the need for prudence in such circumstances, pointing out that prohibiting conduct that amounts only to blockaded entry might send Article 82 down a slippery slope. Rey also suggested a different way in which the “special responsibility” doctrine could be construed consistently with an effects-based approach. However, since there are already conflicting interpretations, it is not clear that adding yet another (albeit more progressive) one would do much to improve matters. On this point, Philip Lowe sensibly pointed out that, “if we are trying to go for an effects-based approach . . . then there are plenty of other legal and economic concepts that cogently explain what we are doing rather than referring to [the concept of special responsibility]”.160 159 Indeed, in Vickers’ example, the conduct marginalizes efficient competitors, at least from a market-by-market perspective. 160 Page 270.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxv

Introduction, Summary, Remarks

lxxxv

Einer Elhauge discussed the “profit sacrifice” test and stressed that a dominant firm can harm competition without sacrificing profits, and it can equally sacrifice profits and exclude rivals but produce efficiencies in the form of innovation. So a profit sacrifice standard would be both underinclusive and overinclusive. Elhauge also took the opportunity to recall that his “efficiency” test, whereby the analysis focuses on whether the conduct enhances the defendant’s efficiency or whether it impairs the efficiency of competitors irrespective of its own efficiency gains, is intended to be a final screen (a “backstop”) that supplements more specific tests. It is a test designed to promote administrability, in the sense that it enables the decision-maker to avoid more cumbersome balancing exercises at that stage of the analysis. Simon Bishop’s rebuttal of the points raised by Andrea Coscelli and Amelia Fletcher regarding the possibility of a rival growing and yet being marginalized at the same time was as follows: “with a loyalty rebate scheme, the counterfactual is always the same: the rival’s market share would always be higher in the absence of the rebate. But so what? This doesn’t tell us whether the rival has been marginalized to a sufficient extent.”161 If the rival is growing, Bishop says, then there is good reason to expect the enforcer or the plaintiff to explain why the rival’s success is failing to exert an effective competitive constraint on the dominant firm. Coping with the complexities of exclusionary conduct Whereas Damien Neven in his presentation had suggested that large firms and their lawyers protest too much about Type I errors,162 David Meyer pointed out that when the objective of minimizing Type I and Type II errors is discussed, a third type of error is often neglected. By this he meant an error (which is partly but not only a spillover caused by perceived or actual Type I errors) when a firm makes a decision about its future conduct. That decision will be made under numerous constraints, but one of these (assuming a rule that minimizes Type I and Type II errors but whose outcome is unpredictable) will be uncertainty as to whether that conduct will conform, if analyzed ex post, to the requirements of antitrust law. Insofar as that uncertainty dampens entrepreneurship, the aggregate cost of foregone pro-competitive behaviour is also partly borne by consumers. Meyer therefore stressed the importance of taking into account the cost of this uncertainty (though estimating the magnitude is plainly problematic) when standards are designed. In response to Meyer’s concerns, Philip Lowe emphasized that the objective, from the Commission’s perspective, is to identify and adhere to economically 161

Page 273. For a comparable view from the US, though perhaps more strident, see Andrew Gavil, “Exclusionary Distribution Strategies by Dominant Firms: Striking a Better Balance”, 72 Antitrust Law Journal 3, 17 (2004). 162

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxvi

lxxxvi Introduction, Summary, Remarks sound ex ante rules that promote predictability, although Lowe doubts that even predictable rules can exclude uncertainty costs. “If we go for a sound economic and legal approach,” Lowe says, “it has to make sense comprehensively, both when we establish the rules and in the analysis of the cases.”163 Emil Paulis sought clarification regarding the concept of an “integrated assessment”. That was a concept underlying the Commission’s (then-draft) Non-Horizontal Merger Guidelines,164 but it was also referred to by Philip Lowe in his oral remarks at the Workshop, and then again by Damien Neven in his presentation.165 Paulis pointed out that even if the positive and negative effects of a practice tend to be tangled together, each must still be identified and given some relative weight. Neven effectively agreed with that statement, the point of the “integrated evaluation” being that the assessment cannot be a simplistic and mechanical matter of placing harm and efficiencies on a scale and observing which side tips. John Vickers noted (as Robert O’Donoghue does in his written contribution) that, unlike the merger control context, in exclusionary abuse cases the counterfactual can be quite elusive, as a range of possibilities might present themselves. And if the imaginary benchmark is overestimated or underestimated, this can materially distort the assessment of the effects of the conduct that was in fact executed. Neven agreed that a proper formulation of the counterfactual is crucial. However, he pointed out that although in some cases this might indeed be a difficult task, in other cases there are guideposts, possibly including industry custom and usage. Eleanor Fox drew attention to certain legal doctrines in the US which serve as a guarantee against what O’Donoghue in his presentation had called “economic literalism”. O’Donoghue had suggested, for example, that it is theoretically irrational for a firm to refuse to supply anyone ready to pay more than marginal cost, and yet clearly in Europe—absent additional elements such as a prior course of dealing or essential facilities—the right to choose one’s trading partners is fundamental. Economic literalism, therefore, might lead to absurd outcomes. According to Fox, US courts recognize certain basic commercial freedoms that lie beyond even the long reach of economic science, namely (unilateral) refusals to deal, pure (non-rebate) low pricing, and innovation. She therefore intimates that European courts might develop their own techniques to avoid outcomes that defy common sense.

163 164

Page 284. In the definitive version of the Guidelines (supra note 11), see paragraphs 20–21, 28, 52 and

92. 165 Neven had explained: “We have deliberately chosen that language so as not to use a language of balancing”. Page 275. Balancing can be misleading, according to Neven, because “[m]ost of the time, the efficiencies and the anticompetitive effects associated with a given practice are closely intertwined”. Ibid. Thus, at paragraphs 21 and 52 of the Commission’s NonHorizontal Guidelines, the question is expressed as whether efficiencies would “counteract” any negative effects, as opposed to, e.g., outweighing them.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxvii

Introduction, Summary, Remarks

lxxxvii

To round out the session, O’Donoghue stressed, as he does in his paper, that Article 82, due to its distinctive features, cannot or at least should not be likened to Article 81 or to merger control for purposes of evaluating efficiencies.

2.2 Panel IV (papers by Coscelli & Edwards, Rey, Melamed, Goldman) A. Written contributions 1. The title of the paper submitted by Andrea Coscelli and co-author Geoff Edwards, Efficiency and Anticompetitive Effects of Tying, nicely captures both the content of the paper and the complexity of analyzing the impact of tying and bundling practices. On the positive side of the ledger, selling two or more products together may potentially be consistent with a wide assortment of benign motives, such as: • transaction cost savings (e.g., a reduction of search costs or sorting costs, or, paradoxically, a reduction of product permutations); • savings of production/distribution costs; • “compatibility” cost savings (which may obviate the need for standardization); • protection of IPRs (and thus protection of dynamic incentives); • quality assurance; • responding to some forms of systems competition (durables/consumables); • eliminating double marginalization; or • efficient price discrimination. Having summarized the lighter side of tying, the authors review certain economic models in which tying may have exclusionary effects. The starting point is the classic Chicago “single monopoly” thesis, although it was quickly realized that this approach to tying embraced an overly sanguine assumption as to the intensity of competition in the market for the tied product. The fragility of the Chicago model has been revealed by several economists over the years, either because its conditions often do not hold or because, even where they do, the model is too static to detect a number of dynamic leveraging or entrenchment tactics. Suffice it to cite one important example here. Under a series of assumptions (including, notably, market power in the tying market; a distinct, oligopolistic market for the tied product; strong customer appeal with respect to the tying product; and the ability of the tying firm to offer a quality product in the tied market), the tying firm may have both the incentive and the ability to exclude or marginalize rivals in the tied market. This is because: (i) the tied market presents an opportunity to seize rents over

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxviii

lxxxviii Introduction, Summary, Remarks and above those extracted in the primary market; and (ii) scale economies in the tied market imply that competition in the tied market between the firm tying the products together and other suppliers of the tied product will be a zero-sum game166 and the rivals’ margins may be jeopardized.167 Given that the potential welfare-enhancing and welfare-detrimental effects of tying can often be tangled together, Coscelli and Edwards propose a methodical rule of reason analysis under Article 82 where equally efficient competitors cannot match a dominant firm’s tied offer and a significant share of the tied market is foreclosed. Emphasizing the need to take credible efficiencies into account, the authors lay further stress on concrete market characteristics: “There is no set of necessary and sufficient conditions that can be derived from the economic analysis to allow a decision maker to easily adjudicate on a tying claim, so we believe a detailed empirical analysis under a (structured) rule of reason remains the most appropriate approach”.168 Indirectly, this statement seems to reflect the significant cost of errors in such cases. Resonating with some of the views expressed during the roundtable debate following Panel III (see above pp. lxxxvi et seq.), Coscelli and Edwards favour a two-stage analysis,169 beginning with the question of whether the tie excludes or disciplines efficient competitors or is dangerously likely to do so. If the answer to this “horizontal” question is yes, the vertical question is whether that foreclosure causes, or is likely to cause, long-run consumer harm. The first step requires the enforcer or plaintiff to advance a theory and corroborating facts to support its story of exclusion. To determine whether that story is sufficiently grounded, a series of different factors on both the supply side (rivals’ marginal costs and fixed costs) and the demand side (e.g., scope for product differentiation, network effects, etc.) should be evaluated, as should any realistic counter-strategies rivals might have at their disposal. For purposes of assessing the story of exclusion it might be necessary to consider not only foreclosure as regards the tied market but also the possibility, as suggested above, that the tie is being used as a strategic device either to nip competition in the bud where a new market is emerging or to entrench the defendant’s position in the primary market. The second step of Coscelli and Edwards’ proposed rule of reason aims to assess the risk of long-run consumer harm, again in light of theory and empirical evidence. As has been noted several times in this narrative, the authors 166

It is assumed here that the size of the market for the tied product is not wildly expanding. As Coscelli and Edwards point out, if the tie has the effect of evicting a competitor, there would be efficiencies in the form of avoided fixed costs, and for this reason the overall welfare implications of the tie are a priori ambiguous. 168 Page 412. 169 Both of the following conditions must be satisfied, which means that in a given case the second stage might never be reached. Note also that at this point of the inquiry it is already assumed that the firm is dominant. The rule of reason proposed by the authors might therefore in this sense be regarded as a three-stage analysis. 167

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page lxxxix

Introduction, Summary, Remarks

lxxxix

recognize that there can be good foreclosure and bad. According to the authors, “to establish anticompetitive tying conduct it is necessary for the relevant Authority to demonstrate both a theory of exclusion and a theory of harm, and to support each theory with empirical evidence wherever possible”.170 It seems clear from their description, however, that as a preliminary matter the defendant would be obliged to come forward with a justification based on efficiencies or other non-strategic reasons, failing which the exclusionary behaviour would be condemned. If the defendant adduces evidence of plausible efficiencies or other non-strategic motives, the decision maker would proceed to balance the harm and the benefits. This immediately raises the question of how such balancing might be executed. Many commentators, including some of the participants of the Workshop, seem ill-at-ease with the notion that positive and negative effects can be “weighed up”, not least because the outcome of such an exercise is difficult to anticipate in “real time” when businesses adopt or avoid a given commercial strategy. On the other hand, a rule of reason-style balancing test need not be a haphazard or mechanical computation. It is interesting to see that Coscelli and Edwards, though they are not lawyers, are drawn to a framework for evaluating the benefits of an exclusionary tie that is grounded in, or at least inspired by, the first three conditions of Article 81(3).171 2. The paper submitted by Patrick Rey, On the right test for exclusive dealing, begins with a capsule statement of Rey’s views regarding the re-think of Article 82 in general.172 As a backdrop, the “competitive process” is defined in the following terms: “Competition is a means to an end, which is the satisfaction of consumer needs. It is a process that forces firms to be responsive to consumers’ needs with respect to price, quality, variety, and so forth. Over time, it also acts as a selection mechanism, with more efficient firms replacing less efficient ones. The application of competition rules should therefore contribute to fostering this process and, in particular, it should place consumers as the heart of the analysis when agreements and conduct are assessed.”

This passage is quite plainly a reflection of the zeitgeist prevailing among many competition specialists in Europe. It is certainly also a contested vision of the way competition should be regulated,173 and it is likely to remain so, but there are good reasons why it has attracted support from many quarters, including many enforcers. As many Workshop participants noted, and as 170

Page 417. See section 4.2.3 of their paper. 172 By no accident, there is a strong kinship between the views described below and those that appeared in the EAGCP Report of July 2005 (supra note 34), of which Rey was a co-author. 173 For a survey of various conceptions and a plea for pluralism, see Oliver Budzinski, “Monoculture versus Diversity in Competition Economics”, 32(2) Cambridge Journal of Economics 295 (2008). 171

(AA) Ehlermann 08 Introduction

xc

12/11/08

10:25

Page xc

Introduction, Summary, Remarks

Rey stresses in his paper, it is a vision that seems to have been given some stamp of authority in the fall of 2006 by the Court of First Instance in GlaxoSmithKline, a case presently on appeal before the ECJ.174 His reading of the CFI’s judgment leads Rey to conclude that the court “is urging the Commission to take proper account of the actual or likely effects of those [supplier-imposed limitations of parallel trade] on the markets”.175 The same effects-oriented principle, according to Rey, should also guide the analysis of exclusive dealing under Article 82. Exclusive dealing, like other forms of single-firm conduct, is characterized by an “intrinsic ambiguity”, in that it may either promote or impair consumer welfare. In that light, an effects-based decision framework is superior to formalistic criteria: such a framework reduces error costs, and it allows for a consistent approach that traverses patterns of conduct of distinct form but equivalent competitive effects. To that extent, it can preclude “category shopping”. Rey also challenges a common a priori tendency to take anticompetitive effects more seriously than efficiencies, and he calls for a more balanced approach. An effects-based approach would also open the door to concentrating more on the impact of the conduct and less on the legal criterion of dominance,176 although Rey acknowledges that an assessment of dominance may still have value as a screen. Finally, as noted earlier in the summary of the Panel III discussion, Rey does not see the “special responsibility” doctrine as an obstacle to reform because an effects-based approach could essentially co-opt it. With reference to exclusive dealing, Rey recalls the Chicago critique of applying antitrust law to this kind of exclusivity: why would any reasonable, non-masochistic buyer agree to a one-sided contract in favour of a manufacturer (i.e., one that would reduce the buyer’s choice of supplier without something to sweeten the deal)? Of course, that approach has in its turn been critiqued (often on the basis of its strong assumptions, as noted above in connection with the paper by Coscelli and Edwards), but it usefully demonstrates the need to establish that the defendant has both the ability and the incentive to foreclose rivals by means of the supply agreement. These elements are part of the “story” that must be told as a condition sine qua non before the defendant can be expected to come forward with any rebuttal. Without necessarily singling out exclusive dealing, Rey also discusses an assortment of scenarios and variations in which (vertical) foreclosure may or may not be likely, depending on various contingent factors such as cost structure, credible commitments, links between the upstream and downstream markets, counter-strategies, coordination problems and so on. The implicit theme running through these scenarios is, once again, the dominant firm’s 174 Quite apart from the fact that the ECJ will have the final word in this matter, it is difficult to determine the precise significance of the case for sectors other than pharmaceuticals. 175 Page 422. 176 See also the EAGCP Report, supra note 34, pp. 14–15.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xci

Introduction, Summary, Remarks

xci

ability and incentives assessed in light of the underlying economics of the relevant market. If a clear and consistent picture of harm can be drawn, it is then necessary to consider the possibility of countervailing efficiencies, of which Rey provides several examples. The remainder of the paper is devoted to a critique of the approach to loyalty rebates proposed by DG Competition in its Discussion Paper of December 2005. 3. The pithy title of the paper submitted by Douglas Melamed is Thoughts about Exclusive Dealing. Melamed’s thoughts, in this context, consist of a succinct defence of the “no economic sense” test, a liability standard cursed with an ammunition-dump name that seems to place a lit fuse in the hands of detractors.177 Part of the problem is that the test has been understood and specified in different ways with varying degrees of success.178 Melamed’s version seems carefully considered, and in his paper he describes how such a test would work in cases of exclusive dealing.179 Beginning with a reminder that exclusive dealing often yields pro-competitive benefits (in particular, the alignment of incentives along the vertical chain and the protection of relationship-specific investments), and with the caveat that where exclusive dealing is liable to secure considerable incremental gains in market power there may be less restrictive alternative means to achieve those benefits, Melamed proceeds to review and reject a number of other possible approaches that could be applied to exclusive dealing. Much of his critique centres on the risk of overdeterrence, or the risk of category shopping where there are no clear and reliable horizontal principles. One of the approaches Melamed scrutinizes is welfare balancing. Is this exercise realistic beyond the animated PowerPoint slides of economists? Not for Melamed, nor in his view for anyone. Above all, Melamed sees a gulf between the rigorous requirements of a welfare balancing test and the need of businesses throughout the economy to know what the law demands right now—the need to know, in other words, how the law tomorrow is likely to evaluate behaviour undertaken today. This leads Melamed away from a welfare-based approach to a more “introspective” analysis of a firm’s conduct (though it is not a mere test of

177 The expression “sacrifice test”, another term Melamed and others have used, raises its own difficulties. It is not easy to find a better substitute term for the test. An unduly cumbersome expression such as a “counterfactual opportunity cost assessment” might convey a better sense of the nature of the inquiry, but it would still not be appropriate for all cases because, as Melamed explains, there may be situations under the NES test (including, most simply, where a dominant defendant fails to come forward with any plausible business justification) in which there is no need to undertake a comparative assessment of costs and benefits. 178 See John Vickers, “Abuse of Market Power”, 115 The Economic Journal F244, F256 (citing Aaron Edlin and Joe Farrell, “The American Airlines case: a chance to clarify predation policy”, in John Kwoka and Lawrence White, eds., The Antitrust Revolution: Economics, Competition and Policy, 4th ed., Oxford University Press, 2003, pp. 502 et seq.). 179 See also Melamed, “Exclusive Dealing Agreements and Other Exclusionary Conduct— Are There Unifying Principles?”, 73 Antitrust Law Journal 375, 403–411 (2006).

(AA) Ehlermann 08 Introduction

xcii

12/11/08

10:25

Page xcii

Introduction, Summary, Remarks

exclusionary intent), namely in the form of the NES test.180 As applied in this context, the NES test implies that “exclusive dealing is permissible, regardless of whether it injures competition, if it would have been profitable for the Firm even if it had not injured competition, after taking into account available alternatives such as non-exclusive or less exclusive arrangements with the [buyers] or exclusive arrangements with fewer [buyers]”.181 Under this approach, the costs incurred by the supplier (or the buyer, in a scenario where the exclusivity runs upstream to the detriment of the buyer’s rivals) are taken as a proxy for the sum of the costs borne by the buyer(s) and the costs borne by the supplier’s marginalized/excluded competitors. Assuming that the defendant supplier claims that exclusive dealing is efficient, a comparison is then made between the resulting benefits (in particular in terms of greater sales,182 which are calculated on the basis of a competitive price, and not on the basis of an unconstrained price made possible by the exclusion) and the corresponding costs, i.e., the additional compensation necessary to induce the buyer to accept the exclusive arrangement. A net positive figure is interpreted as arising from efficient, lawful commercial conduct whereas, if the costs exceed the benefits, and if the exclusive dealing preserves or reinforces the supplier’s market power, the conduct would be condemned. In this light, it can be seen that the NES test is doomed to neither circularity183 nor obscurity,184 as is sometimes suggested. The remainder of Melamed’s paper is devoted to rebutting the claim that the proxy (i.e., the supplier’s costs) understates the real opportunity costs caused by the exclusive dealing, because a dominant supplier can impose exclusivity obligations on its trading partners at no cost to it. On that basis, critics have maintained that the NES test is underinclusive. Melamed acknowledges this possibility, but he shows that exclusive dealing can only be costless to the supplier if a set of restrictive conditions are satisfied. From a normative perspective, Melamed makes a solid case supporting his preferred test—which, as he recognizes in his writings, trades away theoretical purity to meet the needs of pragmatism and self-assessment. It seems plausible that substantial transaction costs could indeed be avoided if businesses could focus less on the uncertainties of an ex post legal assessment that takes 180

In further detail, see ibid. pp. 389 et seq. Page 440. 182 If the size of the market is held constant, greater sales for the supplier would obviously imply fewer sales for its competitors, but from the perspective of the NES test this is merely a reflection of competition on the merits. 183 By contrast, if the focus is on exclusion then it begs the question to say, merely, that conduct is exclusionary if it would not make business sense but for its tendency to exclude competitors. See Vickers, cited supra note 178, at F253 and F255. 184 See Jeremy West, “Background Note”, OECD Policy Roundtable on Competition on the Merits cited supra note 22, p. 28 (“If all the test does is ask why, without giving any guidance about what to do with the answer, then it leaves us in a kind of shadowy place that is on the road to a definite standard but not quite there yet.”). 181

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xciii

Introduction, Summary, Remarks

xciii

account of factors that only later come to light (but is it too much to expect undertakings to have some in-depth knowledge of the markets in which they operate?), and if they could focus instead on the profitability/efficiency of the contemplated course of behaviour (i.e., profitability that does not depend on post-exclusion rent gains). Provocatively, one might even query whether the NES test operationalizes the task of distinguishing performance competition (competition on the basis of efficiency) from impediment competition. The NES test by no means provides a magic solution: there would be plenty of room for disputes over which figures should be used for purposes of comparing the benefits and costs, and this implies some tension with administrability.185 Furthermore, the NES test has been criticized, for example on the ground that it is possible to identify instances in which the test might be overinclusive.186 However, some of this criticism appears to overlook the fact that the test is not meant to be the exclusive tool of a decision-maker; rather, it is intended to be applied as a complement to safe harbours which rule out antitrust intervention where single-firm conduct plainly involves no more than competition on the merits, as in the case of investments in new or improved technologies.187 The greatest challenge for the NES test in Europe, however, might simply be that, de lege lata, it does not square well with the European experience. Indeed, the implications of the test would in fact overturn an article of faith in Europe that (effects-based or not) competition law protects consumer interests, in particular by guaranteeing a partial transfer to consumers of benefits that derive from commercial agreements and practices.188 There is a rebuttal to this argument, but the issue is at least as much a matter of culture 185

Similarly, see the oral remarks of Amelia Fletcher in Panel IV. See Robert O’Donoghue, this volume (suggesting that, in cases involving network effects, it only makes economic sense for a firm to invest in the facility if the market can tip in favour of the investor and if the alternative networks of competitors are thereby prevented from reaching or maintaining efficient scale; such a scenario might perversely be condemned under the test). This is precisely why one of the promoters of the test has suggested that the test might be ill-suited in a “case in which the inevitable outcome of the competitive process would be a single surviving competitor”. Greg Werden, “The ‘No Economic Sense’ Test for Exclusionary Conduct”, 31 Journal of Corporate Law 293, 303 (2006). Werden makes a similar point in “Identifying Exclusionary Conduct Under Section 2: The ‘No Economic Sense’ Test”, 73 Antitrust Law Journal 413, 421 (2006). In passing, there may be a way to reconcile the NES test with scenarios involving powerful network effects. If the network itself is the product, then tipping the market and leaving other alternative networks behind tends to increase the value of the product to consumers. From this perspective, the dominant firm’s investments in growing the network arguably pass the NES test because the excluded rivals were eliminated only as an incident to the enhancement of the value of the product. 187 See, e.g., Werden, “Identifying Exclusionary Conduct Under Section 2”, cited previous footnote, at 419 n. 25 and accompanying text. I am grateful to Greg Werden for providing thoughtful comments on this point. 188 Emil Paulis, in particular, recognized how alien the NES test would be to the European legal framework, and how contrary it would be to the Commission’s sense of direction, in his oral remarks in Panel IV (see pp. 382–383). 186

(AA) Ehlermann 08 Introduction

xciv

12/11/08

10:25

Page xciv

Introduction, Summary, Remarks

and embedded tradition as it is of logic. Suffice it to say that the NES test is neither in its ascendancy in Europe nor on the point of being so. 4. Cal Goldman and co-authors Navin Joneja and Elisabeth Yuh delivered a paper entitled A Canadian Perspective on Tied Selling and Exclusive Dealing. Beginning with an overview of Canada’s competition law system, including certain aspects of both public and private enforcement, the authors present the lay of the land in their jurisdiction with respect to tying and exclusive dealing. The Canadian approach to these practices is notable for several reasons, including the fact that the relevant statutory provisions, though they were absorbed by Canada’s Competition Act of 1986, actually pre-date the Act by ten years. This may provide at least a partial explanation for some of the distinct features of those provisions, such as the differences in language, emphasis and temporal scope that seem to set them apart from Canada’s general prohibition against the abuse of dominance.189 One difference, for example, is that dominance is not a strict prerequisite for a finding of anticompetitive tying or exclusive dealing, if the defendant is a “major supplier” or if the practice in question is widely employed across the market. Such differences have been taken seriously by, in particular, the Federal Court of Appeal. On the other hand, while the provisions on tying and exclusive dealing have distinct characteristics and could be treated as stand-alone theories of liability, the authors report that the Canadian Competition Bureau tends to focus primarily on the general abuse of dominance provision, and to apply the more specific provisions in parallel with it. Another feature of the Competition Act is that, unlike the general abuse of dominance provision, the sections on tying and exclusive dealing make no explicit provision for efficiency justifications on the basis of “superior competitive performance”. As in the case of Article 82 in the European context, therefore, resort must be had to indirect support for such justifications. In that regard, the Competition Act is somewhat more generous than the EC Treaty, as the first section of the Act sets forth several underlying objectives which include an efficient and adaptable economy, competitive prices and consumer choice. Both the Canadian courts and the Competition Bureau seem to be searching for ways to allow for the consideration of efficiencies, but the analytical approach is still under development. Also under development is the precise relationship between the abuse provision and the more specific provisions and the potential of the latter to strike out on autonomous paths, particularly in the context of private litigation. In light of a general theme of the Workshop, namely the risks of having a variegated approach to different practices

189 Tying and exclusive dealing are covered by Section 77 of the Act, while the prohibition against abuse of dominance appears in Section 79. Both provisions are included in Part VIII of the Act. For details, see the Appendix to the paper.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xcv

Introduction, Summary, Remarks

xcv

that have similar effects, one may doubt the wisdom of going down the road of further fragmentation.

B. Discussion The roundtable discussion following the fourth panel of speakers (Coscelli, Jenny, Rey, Melamed, Goldman) concerned the difficult question of what to do about tying and exclusive dealing. Giuliano Amato sensed in advance that the views around the table would be rather divided. One fundamental problem in Europe is that several of the controlling judgments in these areas were decided at a time when it was not quite fully accepted that tying and exclusive dealing could frequently be efficient business practices. A sense of the evolution of the prevailing mentality in this respect can clearly be seen in the Article 81 context by comparing, for example, the rules on tying and exclusivity contained in the old Patent and Know-How Licensing Block Exemptions190 with the Commission’s radically more nuanced approach in the Technology Transfer Guidelines of 2004.191 However, given the relatively greater risks where a firm has substantial and sustainable market power, and given the rigidities of the old case law on abuse of dominance, it is no surprise that proposals to relax the traditional presumptions under Article 82 have been greeted with more caution. The summary below begins with a debate concerning the “no economic sense” test for single-firm exclusionary conduct, and then turns to a number of more miscellaneous points that were raised. As will be seen, one of the incidental themes of the discussion was the need to find some balance between the desirability of simple rules that judges can apply without excessive difficulty, and the usefulness of more sophisticated rules which to some extent may be inescapable in complex business litigation. The “no economic sense” test Einer Elhauge posed some questions to test the robustness (i.e., the error costs) of the NES test advocated by Doug Melamed. Elhauge’s first hypothetical scenario was where a dominant firm maximizes its profits by tying products together, thereby price discriminating and capturing surplus, while at the same time foreclosing a substantial part of the market and raising rivals’ costs. But in this hypothetical case the tying/price discrimination does not generate any efficiencies. The NES test would appear to allow the practice, since the benefits of the price discrimination arise from the extra surplus and not from post-exclusion additional rents. Melamed seemed sceptical of the plausibility of the scenario, since tying often does produce efficiencies, but if that were the case, he pointed out that attempts to force inefficient tied sales would fail the test. 190 191

Regulation 2349/84, [1984] OJ L219/15 and Regulation 556/89, [1989] OJ L61/1. Cited supra note 130.

(AA) Ehlermann 08 Introduction

xcvi

12/11/08

10:25

Page xcvi

Introduction, Summary, Remarks

A second scenario posited by Elhauge relates to innovation, in this case risky investment in the development of a new drug. The decisive motivation for risking the money is the expectation of future monopoly profits (beyond mere recoupment). Implicitly, then, Elhauge’s question is how the NES test copes with the concept of patent protection, or perhaps the concept of “creative destruction” in general. If innovation is driven by the will to exclude all competition, is it meaningful to dissociate the business rationale from the winner-takeall reward?192 It has already been observed that this line of critique appears to be a kind of category error, since plain competition on the merits apparently obviates the need to apply the NES test.193 Curiously, however, during the debate Melamed’s argument seems to have been that the NES test can be applied: if a new product excludes rivals, then the question would be whether a competitive price (obviously problematic in this scenario) would be insufficient for the innovator to recover its investment in the new product. If so, and if recoupment thus depended on the post-eviction “payoff”, then the conduct would be predatory and it would fail the test. However, if such an analysis proved necessary, and if it led to a finding of predation, then the “innovation” would seem to be nothing but a sham. The answer to Elhauge’s question, which assumes legitimate innovation, may be that the behaviour would fall within an “on the merits” safe harbour and the NES test would play no role. A further query raised by Elhauge was whether, in the end, the NES test was equivalent to the prosaic question of whether or not the dominant defendant can advance some pro-competitive justification for its exclusionary behaviour. As Melamed pointed out, there is more to the test than that. Elhauge was not the only sceptic, however. Amelia Fletcher pointed out that the test is often very difficult to apply. To the extent that it can reveal whether a trading partner’s compensation is offset by the efficiencies of the exclusive dealing or whether post-exclusion rents are necessary to pick up the slack, Fletcher recognizes that the information would be useful but she fears that the necessary calculations might entail “pointy-headed” economic abstractions that could overwhelm judges that do not speak that language. Responding to this point, Melamed stressed the utility of the test for purposes of providing routine risk assessment for businesses. He acknowledged that in court it might be harder to apply, but “compared to what?” One could add that, rather than limiting the complexity of rules on the ground that they will exceed judges’ capacity, it might be better to focus on providing judges with more sophisticated economic training.194 192 The suspicion that the NES test might be overinclusive for this reason is also expressed by Robert O’Donoghue in his paper. Similar suggestions were made by Jeremy West in his Background Note to the OECD’s Policy Roundtable on Competition on the Merits, cited supra note 22, at pp. 26 (profit sacrifice test) and 28 (NES test, same argument). 193 See the discussion relating to Melamed’s paper, supra note 187 and accompanying text. 194 Cf. Ernest Gellhorn, William Kovacic and Stephen Calkins, Antitrust Law and Economics in a Nutshell, 5th edition, Thomson West, 2004, at 202 (“That judges may have trouble evaluating

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xcvii

Introduction, Summary, Remarks

xcvii

The next challenger was Emil Paulis, who correctly identified the essential friction between the NES test and European competition law. On the one hand, if consumer welfare is to be a touchstone, as suggested for example by Article 81(3), then the generation of efficiencies between the parties is a necessary but not a sufficient condition to justify exclusion of rivals. On the other hand, the alleged simplicity of the NES test could only be achieved by sacrificing a more systematic and wide-ranging analysis of market effects,195 a doctrinal approach in which the Commission has invested heavily. In that regard, Paulis points out that balancing positive and negative effects is what the Commission already does in relation to Article 81 and merger control. Melamed challenged the relevance of merger cases (i.e., horizontal ones), but as for Paulis’ essential points there is little to say. Other points relating to tying and exclusive dealing Among other remarks that were made during the discussion, only a few will be noted here. Mario Siragusa was particularly concerned at the prospect of deviating from or indeed abandoning the ECJ’s jurisprudence in these areas. Unlike Emil Paulis, Siragusa opposes a balancing of effects in Article 82 cases, and he observes that an important underlying part of the debate is about the legal presumptions developed in the early judgments, which amount to an “abusive unless objectively justified” approach. As a separate criticism of the Commission, Siragusa considers that the modernization of Article 82, like much of the modernization of European competition law in general, serves as an excuse for the Commission to take little real action. Ian Forrester voiced similar concerns, adding that more appealable decisions are needed to develop the law and establish precedents. Philip Lowe was less convinced of the Commission’s inactivity. Furthermore, he highlighted a certain change of investigatory culture within DG Competition—less of an apodictic style, as it were, and greater emphasis evidence is little excuse for summary prohibition where conduct often yields mixed competitive effects. It merely makes the case more pressing for competent tribunals with expertise in law and economics.”). In Europe, efforts are indeed underway to enhance judges’ ability to understand and apply commonly accepted economic principles in the field of competition law. Between 2002 and 2007, around 3,500 judges in Europe participated in courses co-financed by the Commission. In September 2007, the Parliament and the Council established a new legal base for such efforts. See Decision No. 1149/2007/EC of the European Parliament and of the Council of 25 September 2007 establishing for the period 2007–2013 the Specific Programme ‘Civil Justice’ as part of the General Programme ‘Fundamental Rights and Justice’, [2007] OJ L257/16. Pursuant to that Decision, a Call for Proposals was launched and closed on 31 March 2008. See http://ec.europa. eu/dgs/competition/proposals2/20040316_call/call_en.pdf. However, it is difficult to estimate the impact of initiatives such as those just described. 195 A similar point is noted by Barry Hawk in his contribution to the OECD’s Competition on the Merits, cited supra note 22, at p. 256 (citing Andrew Gavil, “Exclusionary Distribution Strategies by Dominant Firms”, cited supra note 162).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xcviii

xcviii Introduction, Summary, Remarks nowadays on building a coherent theory of harm. According to Lowe, this activity on the ground supports and marches in step with DG Comp’s broader effort to consolidate the case law of the European Courts, to build on it, and—using a carefully chosen yet ambiguous word—to develop it. Massimo Motta, perhaps feeling that his trade had come under threat in light of Mario Siragusa’s remarks, pointed to the case of exclusive dealing in order to illustrate the important role of economic theory and evidence in modern competition law. While it is customary for some to think of longerterm exclusivity as being objectionable on foreclosure grounds,196 it is often overlooked that exclusivity can potentially be an important mechanism to protect and hence encourage investment. Siragusa clarified that he was not rejecting the need for sound economic analysis. Indeed, in his view, economic effects have always been taken into account, including in cases where longterm exclusivity (up to 20 years, “never contested” by the Commission) was necessary to protect investments. Patrick Rey did not seem persuaded by the claim that economics has always enjoyed the warm hospitality of competition law in Europe. It is useful to reflect on Siragusa’s suspicion that the modernization exercise is overtly or covertly subverting the ECJ’s time-honoured presumptions as to what constitutes an abuse. This raises the question of the elasticity of judgemade law. Can the antitrust law pronouncements of a supreme court be challenged, legitimately, by the deliberate reforms of an administrative authority? Yes, in my view, provided that: (i) the underlying assumptions on which the original jurisprudence was based are later understood to be anachronistic in light of well-established economic theory and evidence; (ii) conditions are such that the reforms do not impinge intolerably on legal certainty and legitimate expectations, for example, by way of overnight changes with no prior public debate; and (iii) the progressive interpretation and application of that jurisprudence by the authority remains subject to the control of the courts.197 That is 196 It is remarkable how certain numbers—five, for example—are sometimes seemingly plucked from the air as a procrustean yardstick by which contemplated business arrangements are measured. A sliding scale might be appropriate, in light of tied market share, but undue a priori emphasis on specific numbers—necessarily blind to the nature of individual commercial investments—seems liable to have a straitjacket effect. 197 My argument here is obviously a matter of controversy. For example, Giorgio Monti expresses a more orthodox view: “[C]hange to the Article 82 doctrine of abuse and to the concept of dominance can only come about incrementally as the case law develops. The Commission cannot ‘legislate away’ the case law of the Court with a soft law instrument [i.e., a set of guidelines].” Giorgio Monti, “Comments on DG Comp’s Discussion Paper”, cited supra note 145, p. 7. It is also worth pausing to consider the following memorable paragraph by Advocate General Kokott, referring to BA’s reliance (in argumentation before the ECJ in late 2005) on what the company claimed to be future reforms in Commission policy concerning the grant of conditional rebates by a dominant firm:

“[I]t is immaterial how the Commission intends to define its competition policy with regard to Article 82 EC for the future. Any reorientation in the application of Article 82 EC can be of relevance only for future decisions of the Commission, not for the legal assessment of a decision already taken. Moreover, even if its administrative practice were to change,

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page xcix

Introduction, Summary, Remarks

xcix

not to suggest that precedents should not be respected but to recognize that in some fields judge-made law may potentially have a shorter life-cycle than in others.198 To belabour the point a bit, one might recall the old, specious the Commission would still have to act within the framework prescribed for it by Article 82 EC as interpreted by the Court of Justice.” Opinion of the Advocate General in British Airways, cited supra note 37, para. 28 (footnote omitted; emphasis in original). The foregoing passage reveals that, in the view of the Advocate General, the Commission may have some room to maneuver but clearly there are boundaries. The insistence that the Commission must “act within framework prescribed for it by Article 82 EC as interpreted by the Court of Justice” is, in a sense, unassailable. Yet at the same time it overlooks the simple argument I put forward in the main text, namely that the ECJ itself has the discretion (within the limits of the law—whose contours are of course the subject of the broader debate), whether it is impelled to do so by its own wisdom (as in a case such as, say, Nungesser) or whether it does so at the invitation of the Commission. It does not at all seem improper for an administrative agency and a court to whose authority it is subject to engage in an ongoing dialogue, as long as certain criteria are respected, such as those elaborated in the main text. 198 Cf. the majority opinion of Justice Sandra Day O’Connor in State Oil v. Khan, 522 U.S. 3, 20–21 (1997): “In the area of antitrust law, there is [an interest competing with the repose of stare decisis], well-represented in this Court’s decisions, in recognizing and adapting to changed circumstances and the lessons of accumulated experience. Thus, the general presumption that legislative changes should be left to Congress has less force with respect to the Sherman Act in light of the accepted view that Congress ‘expected the courts to give shape to the statute’s broad mandate by drawing on common-law tradition.’ National Soc. of Professional Engineers v. United States, 435 U.S. 679, 688 (1978). As we have explained, the term ‘restraint of trade,’ as used in §1, also ‘invokes the common law itself, and not merely the static content that the common law had assigned to the term in 1890.’ Business Electronics , 485 U.S., at 732; see also GTE Sylvania , 433 U.S., at 53, n. 21; McNally v. United States, 483 U.S. 350, 372–373 (1987) (Stevens, J., dissenting). Accordingly, this Court has reconsidered its decisions construing the Sherman Act when the theoretical underpinnings of those decisions are called into serious question. See, e.g., Copperweld Corp., supra, at 777; GTE Sylvania, supra, at 47–49; Tigner v. Texas, 310 U.S. 141, 147 (1940).” As for the permanence of European jurisprudence, the Court of Justice has no formal doctrine of stare decisis (see supra notes 25–26). Nevertheless, where the interpretation of the EC Treaty is concerned, it is the prerogative of the Court to dilute or deviate from its past pronouncements. The ECJ’s cautious practice in this regard is explained by Advocate General Maduro in Case C-94/04, Cipolla and Macrino [2006] ECR I-11421, paras. 28–29: “The Court has always shown itself to be circumspect with regard to reversing an interpretation of the law given in earlier judgments. Without determining whether those judgments constituted legal precedents the Court has always shown deference to a line of well-established case-law. . . . Even though the Court is not formally bound by its own judgments, by the deference it shows them it recognises the importance of the stability of its case-law for its interpretative authority and helps to protect uniformity, cohesion and legal certainty within the Community legal system. “It is true that stability is not and should not be an absolute value. The Court has also recognised the importance of adapting its case-law in order to take account of changes that have taken place in other areas of the legal system or in the social context in which the rules apply. It has also accepted that the appearance of new factors may justify adaptation or even review of its case-law. The Court has none the less agreed only cautiously to depart from its earlier judgments . . .” (citations omitted) For cases in which the ECJ has indeed departed from its earlier jurisprudence, see supra note 26.

(AA) Ehlermann 08 Introduction

c

12/11/08

10:25

Page c

Introduction, Summary, Remarks

presumption that a patent conferred substantial market power, or a dominant position. A mere 25 years ago, that was the law of the land in the United States.199 However, the US antitrust agencies rejected that presumption in their 1995 IP licensing guidelines, recognizing that exclusive IP rights and market power do not necessarily coincide. This view quickly gained ground in the US, Europe and elsewhere, and indeed it has since been vindicated in the Supreme Court’s more recent jurisprudence.200 Damien Geradin offered some remarks relating specifically to tying. In a world where tying cases are no longer simply bricks plus mortar but gadgets plus an innumerable variety of functionalities, tying cases are bound to become increasingly complex. Geradin doubts whether antitrust has a role to play in bundled functionalities cases even where exclusion can be shown, and he refers implicitly to Microsoft as an example of the futility of mandatory unbundling remedies in such cases. Who is the more appropriate judge, the market or the regulator? Geradin’s point is a good one in the sense that it calls for circumspection in tying cases, where issues such as separate demand and the crafting of effective remedies can present severe difficulties. However, it is also fair to add that the potential risks of leveraging market power can be heightened where the tied market is in its infancy. A final selected point to be noted here is Ian Forrester’s call for a rejection of the moral stigma attached to abuses under Article 82. This point was hammered home by Cal Goldman: “I cannot equate unilateral conduct, under any circumstances, with concerted or cartel activity. They’re worlds apart. One has a behavioural component that is far more egregious and deserves sanctions of the highest order. The other is a unilateral course of business, albeit aggressive, but often efficiency-directed.”201 In Europe, it is apparently only now beginning to dawn on some that, at least as a general matter, per se rules of illegality, whether de jure or de facto, have no place under Article 82.202

199

See Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2 (1984). See Illinois Tool Works Inc. v. Independent Ink, Inc., 547 U.S. 28 (2006). For discussion of the Court’s decision to discard the traditional presumption, see William Kovacic, “Antitrust in the O’Connor-Rehnquist Era: A View from Inside the Supreme Court”, 20 Antitrust 21, 25 (Summer 2006) (also noting the influence of Justice O’Connor’s concurring opinion in Jefferson Parish and on changing preferences in Congress). 201 Page 387. 202 See the Opinion of Advocate General Colomer of 1 April 2008 in Cases C-468/06 etc., Sot. Lélos Kai Sia E.E. and Others v GlaxoSmithKline AEVE Farmakeftikon Proïonton (Syfait II), not yet decided (judgment of the Grand Chamber of the ECJ due 16 September 2008), paras. 62–76. The Advocate General’s reasoning seems doubtful on several points, but he at least recognizes that one ground for the “inadéquation” of per se illegality under Article 82 is that single-firm conduct often has both desirable effects and undesirable side-effects. See ibid., para. 73. 200

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page ci

Introduction, Summary, Remarks

ci

3. Session Three: Exploitative Conduct and the Remedies 3.1 Panel V (papers by Paulis, Röller, Fletcher & Jardine, Forrester, Blumenthal) A. Written contributions 1. Emil Paulis submitted a paper entitled Article 82 EC and exploitative conduct. This contribution focuses on the classic form of exploitative conduct under Article 82, namely, the application of an unfairly high price by a dominant firm. Broadly speaking, the paper concerns: first, the role of the competition authority in the control of excessive pricing if the applicable rules forbid a dominant firm to charge unduly high prices; and second, the conditions which would have to be satisfied before a given price could be characterized as excessive. The first aspect is linked to the well-known aversion, though Paulis does not put it in these terms, to any system in which the competition authority, acting as a latter-day central planner, maintains control over market prices. A system in which price controls were the rule rather than the exception would clearly be incompatible with the concept of undistorted competition envisaged by the Treaty, and Article 82(a) must be read in that light. However, Paulis focuses on more down-to-earth points, such as the fact that the analysis of whether a price is “excessive” is fraught with mathematical peril, and the fact that price interventions sometimes breed still more interventions further down the road as market conditions change. Those considerations lead Paulis to conclude, as a general matter, that the Commission should be quite wary about enforcing the prohibition, which is consistent with the Commission’s long-standing general policy, and with the view of most competition law specialists. But when, then, should the Commission (or a national competition authority) be prepared to intervene? To answer this question, Paulis reviews some of the literature on the subject, referring in particular to works by Motta and de Streel, Evans and Padilla and O’Donoghue and Padilla. The common theme running through those works, as Paulis points out, is the insistence that a necessary (but not sufficient) condition for Article 82(a) to apply should be the presence of very high and long-lasting entry barriers. Paulis agrees, although he adds barriers to expansion to that formula.203 By contrast, whereas other factors indicated in 203 Barriers to expansion may arise, for example, where rivals face capacity constraints, or in some cases where there is significant product differentiation. See, e.g., Commission, Horizontal Merger Guidelines, [2004] OJ C31/03, paras. 30 et seq.

(AA) Ehlermann 08 Introduction

cii

12/11/08

10:25

Page cii

Introduction, Summary, Remarks

the literature may be relevant—such as the grant of exclusive rights by a public authority, or the importance of investment and innovation in a market— he does not see any reason to adopt them as conditions that must be satisfied before an authority can act. Some final thoughts are given to the question: how high is too high? According to the philosophy underlying the Trinko judgment of the US Supreme Court, monopoly prices provide a key stimulus for innovation and growth. From that perspective, paradoxically, what may seem like rent extraction or consumer exploitation from one angle is in fact part of a healthy process that benefits consumers in a wider sense. If one holds stoically to that ethic, no price is too high. Paulis is unwilling to go down that road, as it would ultimately be incompatible with Article 82(a). However, he recognizes that supracompetitive prices are not in and of themselves objectionable. Only “very large deviations from competitive conditions” could be indicative of an abuse, though Paulis does not consider it possible, at the present stage, to define what “very large” might mean. This seems wise: any attempt to do so may be doomed in light of the infinite variety of markets across the economy and across time. Finally, Paulis notes a need for particular caution in markets characterized by substantial investment, innovation, and (in some industries) solidarity between successful R&D projects and their less fortunate counterparts. 2. Lars-Hendrik Röller’s paper, Exploitative Abuses, is addressed, first, to the economic effects of enforcement of a prohibition of exploitative abuse (in particular, excessive pricing), and second, to the identification of scenarios in which the vigilance of the enforcer might be desirable. As a preliminary observation, Röller recalls that the drive to increase market power, and hence to extract rents, is the “essence of pro-competitive behaviour”. He then moves to his first main point, the effects of intervention. This concerns, above all, the relationship between prevailing prices, the likelihood of intervention to control them, and the likelihood of market entry. This in turn largely depends on whether there are entry barriers, and how significant they are. Where entry barriers are low, a significant threat of intervention to “correct” high prices might deter entry and investment, whereas matters would be different where entry barriers are high. Since the market might in fact be working as it should in the first scenario, notwithstanding the high (but transitory) prices, the appropriateness of intervention should only be evaluated in the second. The need for circumspection is further reinforced, irrespective of the magnitude of entry barriers, by the complexity of determining when a price is “excessive”, and by the attendant risk of getting it wrong. The difficult issues that may arise when conducting such an analysis are illustrated by reference to data from the European airline industry. The second part of the paper concerns particular kinds of cases which, from an economic perspective, can potentially justify some degree of control. A first scenario draws attention to a certain inconsistency between what makes

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page ciii

Introduction, Summary, Remarks

ciii

economic sense, on the one hand, and the way in which Article 82 was designed, on the other. In particular, Article 82 fails to reach anticompetitive, “monopolization”-type conduct if the undertaking in question does not (or did not, at the material time) occupy a dominant position.204 From the perspective of Article 82, therefore, a firm is permitted to flout the principle of “competition on the merits” as it fights its way to the top, so long as it respects the prohibition on abuse once it dominates the market. According to Röller, “[t]his is an enforcement ‘gap’, since it is precisely the way in which dominance is acquired that matters in terms of economic effects”.205 Implicitly calling for a legal standard that would discard a long-held article of faith in Europe—“no abuse without pre-existing dominance”—he explains that the gap could be closed if European antitrust law prevented the acquisition of a dominant position by means of anticompetitive exclusionary conduct. Perhaps no Treaty amendment would be required, although that seems highly debatable, but at any rate a significant re-conceptualization of the concept of abuse surely would be. Another scenario highlighted by Röller is where the enforcer fails to catch or prevent an exclusionary abuse; in those circumstances, the possibility to control post-exclusion rents gives the enforcer a second bite at the apple. However, caution is advised here, since a “two shot” policy may reduce false acquittals but it may also lead to more false convictions. On the basis of those reflections, and referring also to some of the other work in this area, Röller would mark a very narrowly defined (ideal) terrain on which an antitrust enforcer might want to take action to address high prices. In particular: (i) there must be significant entry barriers; (ii) the market must be unlikely to self-correct;206 (iii) there must be no (structural) remedy available, by which Röller seems to mean there must be no easy means of eradicating the entry barriers in question, e.g., through prompt legislative action; (iv) there must generally be no sector-specific regulator, or at least no effective one; and (v) in line with the discussion above, the facts must present either a “gap” case,207 or—possibly (though Röller has his doubts)—a case in which there has already been a failure to catch or prevent an exclusionary abuse by a dominant firm. 3. Amelia Fletcher and co-author Alina Jardine submitted a paper entitled Towards an Appropriate Policy for Excessive Pricing. Their paper presents the arguments both for and against a “hands off” antitrust enforcement policy with respect to excessive pricing, reaching conclusions broadly consistent 204

For further discussion, see supra note 123. Page 529. 206 The first condition could probably be folded into the second one. 207 According to Röller, a “gap” case would include not only the scenario in which a firm acquires a dominant position through anticompetitive exclusion, but also one in which a firm is endowed with a legal monopoly by the State and thus is spared from having to compete for it. 205

(AA) Ehlermann 08 Introduction

civ

12/11/08

10:25

Page civ

Introduction, Summary, Remarks

with those in the papers of Emil Paulis and Lars-Hendrik Röller, and with a similar emphasis on the causes of persistent high prices rather than on symptoms. The arguments in favour of abstention are based on the following observations: (i) high prices may stimulate the market in a pro-competitive manner; (ii) in a multi-market scenario, the adoption of high prices for one product may not be a reliable indicator, since the resulting margins might then be “given away” to the same consumers when they buy the firm’s other products at competitive prices; (iii) determining whether a price is “excessive” is difficult; (iv) price intervention can be dangerous because, inter alia, it can inhibit entry and expansion by competitors, and discourage productive and dynamic efficiency. In light of such unintended consequences, Fletcher and Jardine emphasize a need to take account of the particular risks of distortion whenever an enforcer intervenes. Presumably, this would include market-wide (or economy-wide) distortions, as opposed to focusing only on the incentives of the firm under investigation. Furthermore, in order to minimize the risk that a price intervention regime will have some degree of distortive effects on pricing decisions (perhaps small from a limited perspective but significant in the aggregate), the authors provocatively suggest that infringers in such cases should be shielded from fines and from liability for damages in private suits. Remedies in such cases would therefore be limited to forward-looking, ex nunc relief. The other side of the debate is supported by quite different considerations: (i) competition policy objectives are well served by intervention against excessive prices because it yields direct consumer benefits; (ii) prices are already to some extent subject to constraints under other principles of competition law and consumer protection law; (iii) concerns regarding the difficulties of assessing excessive prices are overgeneralized, as illustrated by the OFT’s sophisticated benchmarking exercise in the Napp case; (iv) the putative distortions resulting from interventions against excessive prices are exaggerated; and (v) by focusing on causes rather than symptoms, excessive pricing cases can be pursued and addressed without imposing price controls and hence without the risk of perverse consequences. Fletcher and Jardine take this last point as an opportunity to stress the importance of demand-side entry barriers, including in particular search and switching costs, information costs and information asymmetries, all of which render customers less price-responsive and more susceptible to exploitation. Indicating the usefulness of a competition approach bearing some affinities with consumer protection principles, the authors conclude their paper with a round-up of their policy recommendations: • intervention should be avoided if high prices can be expected to attract timely new entry; • multi-market scenarios require a sophisticated and holistic market analysis;

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cv

Introduction, Summary, Remarks

cv

• offenders should be shielded from fines and damage awards in order to avoid a generalized chilling effect on pricing decisions; • intervention should be avoided in cases where a dominant firm charges high prices for its patented products, provided the patent remains valid and provided the patent system is not itself defective (e.g., due to the risk of overbroad protection); • more generally, competition authorities should always take account of how ex post action might impact ex ante investment incentives; and • the authorities should rely as much as possible on alternatives to price controls in order to address the root cause of the excessive prices, in particular by correcting demand-side problems.208 4. Though it may be asked, what’s in a title?, the one chosen by Ian Forrester will delight Bard-lovers: Sector-specific price regulation or antitrust regulation—A plague on both your houses? This paper concerns the ancient struggle between faith in market forces and the urge to harness them. One cannot mistake the basic message: where high prices are concerned, the seductive sirens by the rocks all too easily lead man into folly. To illustrate the point, Forrester cites several case studies. As a primary example, he reviews the peculiar story of the EU’s Roaming Regulation. This Regulation imposes functional price caps on international roaming charges—contrary, in Forrester’s view, to the spirit of the 2002 telecoms regulatory framework. This story raises the question of whether Europe’s future holds other public relations-driven price regulation in other sectors.209 Such regulation could conceivably be challenged as “unconstitutional” by the standards of the EC Treaty (or of the TFEU if it should enter into force). But whereas the ECJ has at times held Member States accountable where they adopt anticompetitive regulation, in this hypothetical case the Court may not be so intrepid, and it may not subject the Union legislator to the same duty of care. A more encouraging story of regulatory intervention, according to Forrester, involves the railroad industry in the United Kingdom. Here he shows how, in contrast to the Roaming Regulation, well-designed ex ante access charges determined according to objective and transparent criteria can be preferable to the instruments of competition law, with its drawn-out investigations and legal wars.210 Forrester then turns to a summary of pricing remedies adopted in past refusal to deal cases and proceeds with a recap of excessive pricing cases. A more detailed account is then given of the impasse 208

Fletcher and Jardine list a series of possible demand-side interventions. As for the telecoms sector itself, Commissioner Reding (undoubtedly encouraged by the Roaming Regulation) has also called for regulation of charges for SMS messages sent and received while travelling abroad. See Commission Press Release IP/08/1144 of 15 July 2008. 210 In this case, the question is how best to intervene; the prior question of whether price controls are necessary in the first place is not particularly controversial in a natural monopoly-type segment such as railway system operation. 209

(AA) Ehlermann 08 Introduction

cvi

12/11/08

10:25

Page cvi

Introduction, Summary, Remarks

that faced the Commission and Microsoft as regards the compensation the company may require for divulging interoperability information.211 Forrester highlights this aspect of the European Microsoft case as an instructive example of how antitrust intervention may bleed into price regulation more easily than common rhetoric suggests. On the basis of such musings, Forrester essentially concludes that, where it is ascertained that a troubled market cannot self-correct, and where a “manmade” solution is required to address high prices, sector-specific regulation may be a lesser evil than the vagaries of an antitrust investigation. 5. Bill Blumenthal’s contribution, Discussant Comments on Exploitative Abuses under Article 82 EC, should be read in the light of the papers of Emil Paulis, Amelia Fletcher and Ian Forrester. Blumenthal notes that these authors uniformly favour a circumspect application of Article 82 to excessive pricing cases, and he extracts from their papers a number of points supporting that approach; but he takes a step further and asks, why not just remove exploitative abuses altogether from the purview of Article 82? While one might think it predictable for an official representing a US antitrust agency to make such a suggestion, Blumenthal explains that his position has little to do with US law; rather, it is inspired by the observation that where prices are too high, competition authorities are often ill-suited for the task of correcting them. Underlying this argument is the premise that interventions to keep prices down tend to embroil competition authorities in a recurring exercise of control, as opposed to other kinds of price interventions (e.g., an injunction against predatory pricing) which might involve complex price-cost calculations but which are less likely to raise the risk of requiring permanent oversight. “A remedy that entails ongoing regulation of prices and profits by courts or competition authorities”, says Blumenthal, “is almost certain to fail”.212 This leads him to a conclusion similar to that of Forrester: if price regulation there must be, the more advisable instrument is sector-specific regulation.

B. Discussion The discussion following the fifth panel of speakers (Paulis, Röller, Fletcher, Forrester, Blumenthal) concentrated first of all on the (restrictive) preconditions that should be satisfied before it might be acceptable to apply 211 That impasse ripened into a penalty payment imposed on Microsoft pursuant to Article 24(2) of Regulation 1/2003. See Commission Press Release IP/08/318 of 27 February 2008. The Commission’s Decision, which concerned Microsoft’s licence fees up to 22 October 2007, has been appealed by the company before the CFI. See Case T-167/08, Microsoft v Commission, not yet decided. 212 Page 580.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cvii

Introduction, Summary, Remarks

cvii

Article 82 in cases of excessive pricing, and whether there are alternative remedies available. A second main theme concerned how well-placed competition authorities are—relative to sector-specific regulators with superior technical knowledge of particular industries—for the task of policing prices. Pre-conditions for applying Article 82 to control high prices The Workshop as a whole reveals that, as regards exclusionary conduct by a dominant firm, there remain significant differences among the experts. To begin with, a general transatlantic divide may be detected. But even within Europe there are different orientations. These range from a rather legalistic approach, to an approach informed by both transaction cost economics and post-Chicago scholarship. At times, at least among some of the economists, even the Chicago school continues to resonate.213 In contrast to that rather heterogeneous range of views, the Workshop participants reached a strong quasi-consensus with respect to basic principles in the area of exploitative conduct. All who spoke appeared to agree that competition authorities need to think twice before interfering with markets where prices are high (and indeed where they are likely to attract the attention of politicians, the media and the public in general). As will be seen, the differences expressed essentially concerned the precise circumstances which might rouse the authorities and trigger an investigation. Furthermore, and almost without exception, the participants favoured the adoption of guidelines. In light of the widely recognized need for prudence in this area, the primary purpose of such guidelines would be to confine interventions to a narrow category of excessive pricing cases.214 As Patrick Rey put it, the answer to the question of when a competition authority should take action to correct high prices is somewhere between “later” and “much later”. Thus, for example, it was generally agreed that, at a minimum, before any intervention is contemplated it must first be established that barriers to entry (or expansion) are high and persistent. In the absence of such conditions, the preferred solution is to let the market work. However, enlightened self-restraint in this area of enforcement is not easy. As Massimo Motta and several others pointed out, bringing prices down through prosecution, or through the threat of prosecution, is visible and comprehensible to the public, and it promises a substantial and instant political gratification. Similarly, expedience might be a potent factor: a glaring case in

213 In the United States, for many lawyers and economists alike, Chicago-school antitrust remains the object of significant interest. For recent reflections, see George Priest, “The Abiding Influence of The Antitrust Paradox”, 31 Harvard Journal of Law and Public Policy 455 (2008). 214 Lars-Hendrik Röller preferred not to take a position on the desirability of guidelines. However, it is clear from his paper and from his oral remarks that he agreed with the need for competition authorities to be very cautious about applying antitrust law to control high prices.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cviii

cviii Introduction, Summary, Remarks point, as noted by Eleanor Fox, is the pressure that might be exerted on antitrust authorities to do something about oil prices. On the other hand, Frédéric Jenny explained that it is not a simple matter of whether the competition authority, for the greater good of the economy, can resist the temptation of forcing down prices and winning public acclaim. Advocacy by a competition authority against intervention where prices appear to be excessive, he says, is a “public relations disaster”. And as John Fingleton and Philip Lowe noted, bad public relations can cause negative feedback for competition authorities as regards their other enforcement priorities. Many of the participants seemed to recognize this delicate point, although neither Jenny nor anyone else suggested that public relations considerations could justify unwarranted enforcement action. For Jenny, it is largely a matter of reflexivity, tone and sensitive “packaging”: for example, he suggests that a message couched in terms of “We will intervene as soon as . . .” might be preferable to language such as “We will not intervene unless . . .” Coming back to the minimum set of conditions for intervention, Massimo Motta suggested that in fact it is underspecified and overinclusive to say that high and persistent entry or expansion barriers suffice to justify intervention, because not all barriers are alike. An enforcer should be particularly prudent in cases of endogenous entry barriers, that is, where the barriers result from investments in innovation, as in the case of an industry with network effects. In such cases, the opportunity to charge prices considerably above cost was obviously the primary ex ante inducement to invest.215 The need to scrutinize the entry barriers in question before taking action, and the need to consider ex ante incentives, was also supported by Damien Neven. Rafael Allendesalazar did not question the need for competition authorities to be cautious with regard to enforcement in this area. However, he did point out the irony of a cautious approach: where prices are high, competition authorities are unlikely to step in; but where a dominant firm lowers its effective price through rebates, there is a risk of intervention. Furthermore, since Article 82 could be used later to bring prices down if an exclusionary strategy proves successful, this indicates, in Allendesalazar’s view, that a competition authority in Europe can afford to be more passive in the area of exclusionary conduct. John Fingleton drew attention to a need for vigilance in scenarios involving system pricing where a dominant firm packages products together. As an example, he cited certain banking services where consumers do not understand what they are being charged for. While single-firm dominance may be unusual in banking, the broader point is an interesting one because it implicitly links opaque prices to the concept of “unfair” prices under Article 82. Of course, in some cases, dubious pricing for packaged products 215 By analogy, this observation seems to hearken back to the distinction between deterred entry and blockaded entry, discussed in the context of exclusionary conduct. See supra notes 155–158 and accompanying text.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cix

Introduction, Summary, Remarks

cix

may also involve an exclusion-based theory of harm. In cases that escape the competition rules because there is neither dominance nor any agreement (but not only in those cases), consumer protection remedies may be necessary to ameliorate information asymmetries. Reflecting on the presentations of Emil Paulis and Lars-Hendrik Röller, each of whom had referred to “gap” cases,216 Einer Elhauge said that, on hearing the idea of using excessive pricing actions as a way to close the gap, this was the first time that the prohibition on unfair prices made any sense to him.217 He further wondered why intervention in excessive pricing cases should not be limited to gap cases only. Intervention would thus only be possible where it is shown that the defendant’s dominant position was acquired through anticompetitive behaviour.218 In addition, as a way of avoiding the risk of ongoing price regulation by the antitrust authority, Elhauge proposed the use of (horizontal) structural remedies in such cases. Of course, the Commission’s discretion to impose such remedies is limited by the principle or proportionality219 and a “less intrusive means” check, in accordance with Article 7 of Regulation 1/2003. Such an approach would therefore face legal obstacles as far as any break-up remedy is concerned. With respect to Elhauge’s substantive proposal to limit intervention exclusively to gap cases, Damien Neven did not want to go that far because in his view it could lead paradoxically to an undue expansion of the scope for intervention, in particular in cases involving the type of endogenous or strategic entry barriers discussed above. Jorge Padilla was also sceptical, in part because he would prefer to limit scrutiny of exclusionary conduct to scenarios of dominance rather than “pre-dominance”. Amelia Fletcher added that Elhauge’s suggestion 216 As explained earlier in relation to Röller’s paper (see supra p. ciii), a “gap” case in this context is a case where Article 82 might be applied to control exploitative conduct by a dominant firm to compensate for the fact that the provision could not be triggered to prevent “predominant” unilateral conduct. 217 Following the Workshop, and recalling his criticism of Article 82 for not covering anticompetitive, monopolization-type behaviour where the dominance threshold has not been crossed (see supra note 123), Elhauge has continued to reflect on such gap cases. Referring to the prohibition on unfair prices, he says that “perhaps we should understand it as a claim designed to deal with the gap that otherwise would be left because EC law prohibits the abuse ‘of’ a dominant position, but does not prohibit anticompetitive conduct that is used to obtain that dominant position in the first place”. Elhauge, “Disgorgement as an Antitrust Remedy”, Harvard John M. Olin Discussion Paper No. 613 (May 2008), p. 12. 218 As noted above, Röller had in fact contemplated this approach in his paper. However, Röller added an important qualification, namely that “gap cases” should be construed as including not just the anticompetitive acquisition of a dominant position but also situations where the dominant position results from State-granted exclusive rights. This qualification would be of considerable significance in Europe’s post-liberalization industries (see main text below noting Eleanor Fox’s agreement on this point). By recognizing each of these two kinds of gaps to which Article 82 does not apply, Röller seeks (as does Elhauge) to ensure that a monopolist who fought a fair fight to capture the market would never be turned on and punished for charging high prices. 219 In this context, the proportionality principle is given specific shape by Recital 12 of Regulation 1/2003, according to which a mandatory restructuring of the defendant should only be contemplated where there is a substantial risk of a lasting or repeated infringement that derives from the defendant’s very structure.

(AA) Ehlermann 08 Introduction

cx

12/11/08

10:25

Page cx

Introduction, Summary, Remarks

would effectively deprive the antitrust authority of the ability to correct an erroneous decision not to challenge exclusionary conduct where it was clear that the firm was dominant. In contrast to Fletcher, however, both Neven and Andrea Coscelli thought it unwise in such “non-gap” cases to explore the use of Article 82 at a later stage to correct false acquittals in relation to exclusionary conduct. The priority, as Neven said, should be to fix the way Article 82 is applied to exclusionary conduct in the first place. Eleanor Fox echoed Röller’s point that one factor an enforcer should consider when evaluating whether to take action against an exploitative abuse is whether the defendant acquired its dominant position by competing on the merits or whether it was granted market privileges by a Member State. It appears that, for both Röller and Fox, there is little difference between the wilful acquisition of monopoly power by anticompetitive means, and a shortcircuit of the competitive process by the grant of a legal monopoly. Although the need for intervention must always be assessed with sober judgment based on the facts of a given case, their proposition is quite compelling. Robert O’Donoghue, Patrick Rey and others endorsed a point raised by Amelia Fletcher in her presentation, namely that alternatives to price regulation (such as measures bringing down entry barriers or reducing consumer information costs and promoting price transparency) should be considered whenever possible. As positive examples from the UK experience, O’Donoghue cited cases concerning the use of industry standards and information disclosure requirements. On a separate point, he reinforced Emil Paulis’ observation that the question of “how high is too high a price” cannot be answered with a one-size-fits-all answer, and that it is dangerous to make any statements to the effect that a particular percentage above a given benchmark is presumptive evidence of abusive pricing. To the contrary, the message should be that if a particular margin of excess leads to a finding of abuse in one case, pains should be taken to indicate that this margin does not prejudge other cases, and certainly not cases in other industries. Jorge Padilla introduced a distinction between: (i) cases involving exploitation of final consumers; and (ii) cases where the buyer is somewhere in the middle of the supply chain, and where the high margins cannot be passed on to the buyer’s own customers (presumably due to the buyer’s lack of market power downstream). According to Padilla, antitrust enforcers should ignore the second type of case because, if the high price is not passed on and borne by final consumers, then there will be no negative consequences for allocative efficiency. All that happens in such a case is that surplus is transferred from a downstream producer to an upstream producer of inputs.220 On the question 220 If Article 82(a) is read in the light of the other examples given of an abuse, arguments could be made both in support of and contrary to Padilla’s proposition. In any event, his suggestion relates to enforcement priorities rather than what the correct interpretation of the law might be.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxi

Introduction, Summary, Remarks

cxi

of alternative means of addressing high prices, Padilla agreed with Amelia Fletcher that the advocacy role of competition authorities is vital. Making their case to legislators, competition authorities can help to prevent the adoption of (or secure the repeal of) regulation that protects incumbents or is otherwise defective, for example from the standpoint of market design. Daniel Zimmer intervened to ask how Article 82 should be interpreted where producers are exploited by a powerful buyer which occupies a dominant position on the purchasing market and forces down the price of the input it buys. Zimmer’s argument is that the essential objective of Article 82 must not be to protect consumer welfare, because in this scenario it is producers who are harmed, not consumers.221 Apparently taking a cue from Padilla’s remarks (see above), Philip Lowe suggested that a simple transfer of surplus of this kind might not a good candidate for enforcement action at Community level.222 However, both Zimmer and Lowe may have overlooked the possibility that the undue exercise of buyer power can indeed harm consumer welfare, for example if it leads to restricted output.223 The Commission appears to have recognized this in its 1991 guidelines on the application of the competition rules in the telecoms sector.224 In more benign cases, of course, buyer power might translate into lower retail prices. That possibility, in turn, is likely to depend on whether the firm that is dominant on the purchasing market faces competitive pressure downstream when it acts as a seller.225 Where the use of buyer power enhances consumer welfare, it does not seem problematic to suggest that Article 82 can sit idly by. A final point to be flagged in this non-exhaustive summary was raised by Frédéric Jenny and seconded by Ian Forrester and Bill Blumenthal. They reminded the group that, whereas modern economies characterized in the main by competitive industries can afford to maintain a “light touch” enforcement approach in the area of exploitative conduct, smaller transition economies where the necessities of life may be overpriced might have very different requirements.

221

See supra notes 40–44 and accompanying text. Lowe pointed out, however, that exploitation of a supplier by a purchaser might be caught, in some jurisdictions, by national laws prohibiting the abuse of a position of economic dependence. 223 Other conceivable theories of consumer harm in monopsonistic settings are drawing increasing interest. See, for example, Paul Dobson and Roman Inderst, “Differential Buyer Power and the Waterbed Effect: Do Strong Buyers Benefit or Harm Consumers?” (2007), available at http://personal.lse.ac.uk/inderst/Dobson-Inderst%20-%20waterbed_eclr_final_ version%207-4-07.pdf, and references therein. 224 [1991] OJ L233, paras. 116–120. 225 See, e.g., Paul Dobson and Michael Waterson, “Countervailing Power and Consumer Prices”, 107 Economic Journal 418 (1997). 222

(AA) Ehlermann 08 Introduction

cxii

12/11/08

10:25

Page cxii

Introduction, Summary, Remarks

Institutional issues There seemed to be broad agreement in the room that price regulation should not generally be the occupation of antitrust enforcers, and that in specific industries typified, for example, by natural monopoly features, sector-specific regulation has a role to play. However, the interplay between competition law and regulation is complex, not least because it raises difficult institutional choices.226 What Bill Blumenthal called the “boundary conditions” therefore remain a matter of some debate. One pertinent scenario is where a sector-specific regulator either positively errs or fails to act. Emil Paulis and Rafael Allendesalazar expressed two opposing views as to how an antitrust authority should proceed in such a case. Whereas Paulis had maintained that corrective action under Article 82 might be necessary, Allendesalazar considered that the Commission should seek to address the underlying cause of the regulatory failure (which might be government intervention, for example) rather than using Article 82 to condemn the undertaking in question. Sensing that the latter view might entail some kind of presumption against enforcement of the competition rules where the undertaking is subject to the oversight of a regulator, Blumenthal objected that any generalized deference in favour of decisions by a regulator would be inappropriate. However, Blumenthal also considered that further reflection was needed to identify appropriate conditions that could keep antitrust authorities from trampling on regulators’ turf willy nilly. It was clear from his presentation that Blumenthal did not consider antitrust authorities to be well-equipped to manage prices. In unusual sectors with recurrent market failure, he said, sector-specific regulators, for all their potential weaknesses, are probably better suited for the task. Possibly reading too much affinity for regulation into Blumenthal’s remarks,227 John Fingleton reeled off a number of reasons why more regulation would be a bad idea, and why competition authorities in fact can handle price regulation chores. For example, a competition authority seems more apt to look for market-conforming solutions than a regulator might. Furthermore, the substance of a pricing case may involve issues similar to those a competition authority deals with routinely, such as yield management.228 Competition

226 For a recent thoughtful discussion, see Giorgio Monti, “Managing the Intersection of Utilities Regulation and EC Competition Law”, 4(2) Competition Law Review 123 (2008), available at http://www.clasf.org/CompLRev/Issues/Vol4Iss2Art2Monti.pdf. 227 Blumenthal had said that, if an industry is unregulated but the market mechanism isn’t working, it may be necessary to consider a regulatory solution. However, it seems clear from his remarks and from his paper that Blumenthal was referring to marginal cases, and that he is not unaware of regulation’s dark side. 228 Yield management (a.k.a. revenue management) involves the commercial quest to allocate the right capacity to the right customer at the right time at the right price so as to maximize revenues. Price discrimination in the airline industry is a typically cited example.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxiii

Introduction, Summary, Remarks

cxiii

authorities may also have a better understanding of alternative remedies. Regulatory agencies, moreover, have often become bloated pufferfish, fed liberally by their “clients” and guided by their permanent instinct for selfpreservation. In addition, while Fingleton accepts that competition authorities are also capable of doing harm, he points out that creative institutional solutions can minimize that danger. For example, in one of the models he refers to, one competition agency has jurisdiction to review the price regulation efforts of another and to roll back or otherwise modify the regulation as appropriate in light of changing market conditions. Like Fingleton, Jorge Padilla was sceptical that sectoral regulators can be more effective than competition authorities in handling excessive pricing cases. At the same time, however, Padilla returned to the point that intervention in excessive pricing cases should not be a primary occupation of antitrust enforcers. Padilla’s remarks serve as a reminder of the basic observation that price regulation, at least if it is widespread, is ultimately antithetical to the concept of a market economy. Fingleton’s preference for letting competition authorities administer regulation may well be justified in some industries, but in others (as Fingleton seemed to recognize) there may also be contraindications. Martin Hellwig told the cautionary tale of entrusting the Bundeskartellamt and the regional cartel authorities in Germany with responsibility to regulate access to energy networks ex post on the basis of competition law.229 This failed experiment was finally abandoned in parallel with the adoption at Community level of sector-specific directives in 2003 that required each Member State to appoint (or maintain) an energy regulator. More generally, Hellwig expressed reservations about sector specialists and competition specialists being housed under the same roof due to the risk of a social constructivist-type impact that a regulatory culture might have on a competition culture. A diluted competition culture, in turn, might compromise the effectiveness of the competition authority as it pursues its basic mission. Blumenthal later picked up and underlined the same point. Philip Lowe agreed with those who felt that, in some cases, regulation might be a necessary evil. A competition authority, however valiant it may be, is limited in what it can accomplish by its resources, case docket, the broader policy environment, time constraints, and so on. However, Lowe added that the same principle applies in the opposite direction: where regulation is an inappropriate instrument to address problems in the market, and where an 229 Under the first generation of European energy directives (Directive 96/92 and Directive 98/30), Germany maintained a negotiated access regime, whereby disputes were settled either by arbitrators or, in a heavier procedure, by the Bundeskartellamt or regional cartel offices. Under that regime, formal disputes were sometimes costly and time-consuming as the appeals process lumbered along. For details, see Commission, first benchmarking report on the implementation of the internal electricity and gas market (December 2001), available at http://ec.europa.eu/ energy/electricity/benchmarking/doc/1/report-amended_en.pdf, pp. 15–16.

(AA) Ehlermann 08 Introduction

cxiv

12/11/08

10:25

Page cxiv

Introduction, Summary, Remarks

industry is better served by avoiding regulation and letting antitrust do the job, it is incumbent on the competition authority to make its case before the legislator and the public. The final word fell to Emil Paulis, who highlighted the EU’s telecoms regulatory framework as a success story with regard to the coordination of national regulators’ activities and the oversight role of the Commission. As is well known, the procedure laid down by Article 7 of Directive 2002/21/EC (the “Framework Directive”) requires national authorities to carry out market assessments and to notify their findings and proposed measures to the Commission, the goal being to peel off layers of regulation where effective competition breaks out and to maintain or impose ex ante obligations on dominant firms where it hasn’t. According to Paulis, the Article 7 procedure has produced a fruitful interaction among institutions.230 This constant communication, which apparently avoids institutional commingling, has been beneficial for both sides, since the antitrust enforcer gains a deeper understanding of the industry while regulators learn more about competition-based approaches to the functioning of the market.231 However, the claim that the 2002 regulatory framework has achieved real deregulatory effects has been questioned.232

3.2 Panel VI (papers by van der Woude, Siragusa, Régibeau, Geradin, Patterson) A. Written contributions 1. Marc van der Woude’s paper, Unfair and Excessive Prices in the Energy Sector, is a timely one, considering that the dysfunctional energy sector is in 230 For a useful FAQ regarding this procedure, see Commission, MEMO/07/457 of 13 November 2007. See also Commission, Report on Competition Policy 2007, point 46. 231 For the argument that the Article 7 model should be extended to the energy sector, see Francesco Maria Salerno, “The Competition Law-ization of Enforcement: The Way Forward for Making the Energy Market Work?”, EUI Working Paper RSCAS 2008/07 (FSR), available at http://www.eui.eu/RSCAS/WP-Texts/08_07.pdf. It should be borne in mind, however, that regulatory models are not very easily transposed from industry to industry. 232 The Commission seems to be doing what it can. In December of 2007, the Commission recommended a reduction in the number of markets in which ex ante regulation might be contemplated from 18 to 7. See Commission Recommendation of 17 December 2007 on relevant product and service markets within the electronic communications sector, [2007] OJ L344/65. On the other hand, Alexandre de Streel has observed that, despite the intention to roll back regulation (a raison d’être of the 2002 framework), in practice the trend has been just the opposite. See de Streel, “Remedies in the Electronic Communications Sector: Too much regulation? Too little ambition?”, slide presentation at the EUI Competition Day, Florence, 4 April 2008, especially slides 7 and 8. The presentation is available at http://www.iue.it/ECO//Conferences/ CompetitionDay/Papers/De%20Streel_presentation.pdf. For further details, see de Streel, “The Relationship Between Competition Law and Sector Specific Regulation: The case of electronic communications”, 47 Reflets et Perspectives de la vie économique 55 (2008).

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxv

Introduction, Summary, Remarks

cxv

the spotlight these days more than ever before. It is clear from the paper that van der Woude shares many of the opinions expressed in Panel V, in particular the view that Article 82(a) should be used only sparingly, and only after seeking first to address the causes of high prices. However, whereas many participants in Panel V reached that view on grounds of sound economics or policy, van der Woude bases his conclusions on what he considers to be the logic of Article 82. He also tries to identify how the concept of “fairness” fits into that logic, since the example given in Article 82(a) is unfair prices and other unfair trading conditions. Van der Woude seems to suggest that fairness is not normally relevant under Article 82 but that it becomes so when a dominant firm no longer faces competitive constraints, or as he puts it, once the competitive process ceases to play its “welfare-creating role”. Furthermore, he stresses that Article 82 allows a dominant firm to determine its price in accordance with its profit-maximizing output.233 It is only when such a firm systematically charges prices above that level—supramonopoly prices, as he calls them—that an abuse is possible.234 Following a review of the pertinent case law and practice on excessive pricing, the discussion turns to energy. Van der Woude observes that most of the competition problems attracting the attention of the Commission in its Sector Inquiry involve not exploitation but rather exclusionary effects, for example by means of discriminatory practices or long-term exclusive supply contracts. However, he discusses two issues which have arisen at national level and which conceivably could draw scrutiny under Article 82(a). The first scenario relates to access charges applied by a natural monopolist for the use of transportation or distribution networks. However, this problem seems largely theoretical since Community legislation imposes a model of regulated third party access (“regulated TPA”) whereby the regulator approves tariffs, or at least tariff methodologies, ex ante.235 As van der Woude points out, tariff regulation 233 This elementary point about Article 82 is sometimes misunderstood. Elhauge, for example, describes Article 82(a) as “puzzling because, on its face, it seems to condemn the possession of dominant market power itself, for any firm with such power could be said to be pricing excessively”. Elhauge, cited supra note 217, p. 11. Such an interpretation would immediately clash with the fact that dominance as such is not prohibited. It is difficult to see the logic behind a regime that accepts dominant market power and imposes a ban on unfair pricing, but then imposes a blanket rule whereby profit-maximizing prices charged by a dominant firm are ipso facto unfair. See also Amelia Fletcher, this volume, p. 535 (pointing to the absurd implications of a hypothetical Article 82 that forbids above-cost pricing: a monopoly would be precluded from such behaviour but in a non-collusive oligopoly situation the oligopolists, facing no such prohibition, would be free to price well above cost). 234 For further remarks in light of the historical circumstances surrounding Article 82, see supra note 32. 235 See the “Second Energy Package” Directives: Articles 20(1) and 23(2) et seq. of Directive 54/2003 (Second Electricity Directive), [2003] OJ L176/37; and Articles 18(1) and 25(2) et seq. of Directive 55/2003 (Second Gas Directive), [2003] OJ L176/57. In some cases (but generally not in cases where dominance is established or reinforced), there may be no mandatory access and hence no tariff at all, due to a statutory exemption from the TPA regime. Such an exemption is provided for, where strict conditions are satisfied, in order to induce investments in new gas or

(AA) Ehlermann 08 Introduction

cxvi

12/11/08

10:25

Page cxvi

Introduction, Summary, Remarks

appears to be stricter than Article 82(a) insofar as regulation essentially imposes cost-reflective pricing (which may include efficiency-driving components).236 This makes sense: by definition, a natural monopoly means that the market cannot support competition and cannot self-correct. Tariff regulation is justified under those circumstances as a second-best solution.237 The second scenario concerns wholesale prices for electricity.238 In that regard, electricity presents a number of unique features that must be taken into account. First of all, the wholesale price is determined by reference to the marginal cost of the last plant used. Consequently, high profits may accrue to the owners of more cheaply run plants (e.g., coal or nuclear-based plants) contributing the base load of the power supplied. Could the fact that producers benefit from prices set by marginal plant which do not reflect base load costs justify intervention under Article 82(a)? Or, in other cases, monopolists and oligopolists have been known to “game” the wholesale market by withdrawing available capacity, thereby pushing the price up, or simply by price gouging (scarcity rents where a firm capitalizes on the indispensability of its capacity in peak demand periods). Again, the question is whether Article 82(a) is, in those circumstances, an appropriate instrument. In the first scenario described above, van der Woude notes that the application of Article 82(a) would be problematic because owners of the base load plants are not necessarily the price setters, and if they are, the high profits may be justified to cover high fixed costs and the costs of new investment. Even if it is accepted that market power may be inferred where a firm frequently sets the price with its marginal plants and where it profitably sells large volumes at that price, van der Woude is sceptical that such prices are “unfair” within the meaning of Article 82(a). On the one hand, the producer is pricing according to its marginal costs; on the other, there are the fixed cost and investmentrecovery arguments. Turning to the pricing issues arising in the second scenario, van der Woude points out that policing against the tactical withdrawal of capacity is a task for Article 82(b) rather than Article 82(a). In the case of price gouging, where the producer capitalizes on its understanding of the demand curve by raising its price above marginal cost, it does not appear electricity infrastructure. See Article 22 of the Second Gas Directive and Article 7 of Regulation 1228/2003 on conditions for access to the network for cross-border exchanges in electricity, [2003] OJ L176/1. For further details, see DG TREN’s interpretative note on exemptions from the TPA regime, available at http://ec.europa.eu/energy/electricity/legislation/doc/notes_for_ implementation_2004/exemptions_tpa_en.pdf. 236 Costs might well be relevant to an examination under Article 82(a), but as van der Woude notes, under United Brands they are not necessarily determinative. Other factors may be more probative, depending on the facts of the case. 237 Article 82 is different. It accepts that, in the competitive game, monopolies exhibiting skill, foresight and industry are born, grow old, and sooner or later die. If there are no high and persistent entry barriers to insulate a monopolist from competitive pressure, then market failure seems unlikely and Article 82 can rest unperturbed. 238 Gas is not particularly relevant here, as price formation in that sector is comparable to the norm in other commodity industries.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxvii

Introduction, Summary, Remarks

cxvii

that van der Woude would object to the application of Article 82(a), provided that the strategy is systematic and could not be defeated by new entry. The rest of the paper reviews and critiques a number of pricing cases in Spain, Germany, Denmark and Italy. On reading this survey one has the general impression that national authorities are more willing than the Commission to take action under Article 82(a). This appears once again to underscore the need for published guidelines setting forth clear criteria for the application of Article 82, and for the abstention therefrom, in cases of exploitation. 2. The “unorthodox thoughts” in Mario Siragusa’s Excessive Prices in Energy Markets: Some Unorthodox Thoughts refer to Siragusa’s plea for the bolder use of structural antitrust remedies as a means of coping with high energy prices. At first glance, Siragusa might seem hawkish as far as structural remedies are concerned.239 However, the claims made in the paper are confined in the sense that they relate chiefly to the energy sector, and then only after the stringent criteria advocated some of the other participants (Paulis, van der Woude, etc.) have been met. In the energy sector, structural separation of related business units is obviously a hot topic.240 However, the separation relevant to that discussion is vertical unbundling to address foreclosure and investment problems; by contrast, the remedies Siragusa presumably has in mind include horizontal divestitures as a frontal attack against incumbents’ durable market power. It is also clear from his paper that he regards virtual power plant (VPP) and gas release remedies as structural remedies.241 The reason why Siragusa favours structural remedies is a good 239 In particular, Siragusa notes that, in Volvo/Veng, the ECJ would have allowed an interference with the design rights at issue if, hypothetically, the right holder had imposed excessive royalties. In Siragusa’s view, the judgment should be construed as authorizing structual remedies in excessive pricing cases. Strictly speaking, a compulsory IP licence is not a structural remedy but a hybrid remedy involving both a behavioural element and an interference with property rights. Further discussion of structural remedies follows in the main text and in the following two footnotes. 240 As is well known, the hottest part of the debate concerns ownership unbundling, which has been implemented in some Member States but which has been resisted by other (influential) Member States at Community level. Politically, the campaign to move forward with ownership unbundling in the gas sector has been lost for the time being. The outcome as regards electricity transmission may prove to be somewhat more fruitful. If future Community legislation does (in principle) require ownership unbundling with respect to the electricity grid, it is plausible that some Member States might challenge it, or that it might be called into question in a preliminary reference procedure, on the theory that it disproportionately interferes with property rights. There are good reasons why that theory should fail, but no attempt will be made to elaborate them here. For the argument that such a claim should succeed, see Florian Becker, “Market Regulation and the ‘Right to Property’ in the European Economic Constitution’, 26 Yearbook of European Law 255, 288–296 (2007). 241 This is not an uncommon view. See, e.g., Herbert Ungerer, “European Commission’s Energy Sector Inquiry—Results and Prospects”, slide presentation made in Dublin, 21 February 2007, available at http://www.iiea.com/images/managed/events_attachments/HERBERT% 20UNGERER.ppt, slide 27. At most, however, such remedies may be called “structural” in the colloquial sense that they affect the structure of the market. To the extent that structural

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxviii

cxviii Introduction, Summary, Remarks one: they go to the heart of the matter by directly expropriating some of the defendant’s market power.242 Furthermore, whereas most ordinary industries might call for the utmost prudence when considering whether to break up a company, in the energy sector the monopoly power of the defendant might have been created by legislative or royal decree, and not necessarily as a reward for superior efficiency. The reasons why horizontal divestiture might appear justified are reinforced by the pragmatic consideration that breaking up a company, though it involves significant short-term transaction costs, can avoid ongoing regulatory monitoring by the antitrust enforcer. However, as suggested earlier,243 certain legal criteria must be satisfied before the Commission is permitted to order structural remedies.244 Siragusa appears to consider that structural remedies are legally permitted in the case of high energy prices because a cease-and-desist order would be either less effective or more onerous for the defendant.245 This may be true, but one might want to see further evidence of this before resorting to a break-up. Considering that sector-specific regulations routinely deal with such pricing issues, it is not obvious that a cease-anddesist order would be inferior to a structural remedy.246 Indeed, the regulator remedies refer instead to remedies that result in a reorganization of the undertaking concerned, VPP and gas release remedies may be regarded as either hybrid remedies (involving no divestiture but reducing the incentive and ability to exercise market power) or as “access remedies”. For the “hybrid remedy” classification, see Natalia Fabra, “Antitrust Remedies in the Energy Sectors”, slide presentation, 2nd EUI Competition Day, Florence, 4 April 2008, available at http://www.iue.it/ECO// Conferences/CompetitionDay/Papers/Fabra_presentation.pdf , slide 5. For the “access remedy” classification, see the draft revised Commission Notice on remedies acceptable under the Merger Regulation, available at http://ec.europa.eu/comm/competition/ mergers/legislation/draft_remedies_notice.pdf, paras. 62–63. The question of how to qualify remedies such as VPP and gas release programs goes beyond mere semantics because, although the Commission is always bound by the principle of proportionality regardless of the remedy it may seek to impose, structural remedies are subject to a specific regime under Regulation 1/2003. If it opts for structural remedies (a term that remains undefined under the Regulation), Article 7 and Recital 12 require the Commission to make a more compelling case. See supra note 219 and accompanying text. 242 A recent high-profile example in the UK concerns the envisaged forced divestiture by the British Airports Authority (BAA) of some of its UK airports. The Competition Commission is currently consulting on the proposed remedies. See News Release 24/08 of 20 August 2008 (“CC May Require BAA to Sell Three Airports”), available at http://www.competitioncommission.org.uk/press_rel/2008/aug/pdf/24-08.pdf. 243 See supra note 219 and accompanying text. 244 The relevant legal threshold may be somewhat lower under national law in some Member States. 245 Fines are another possible sanction. However, in the absence of competitive constraints, it is difficult to see why the brunt of any fines could not be passed on to consumers in a variety of ways. In the (presumably narrow) set of cases where excessive prices were charged in a period before competition broke out, and pass-on thereafter becomes unlikely, fines might offer a less restrictive alternative. 246 The following argument depends on the effectiveness and independence of the regulator (vis-à-vis the national government; independence from industry is already required by Community sectoral legislation), which is expected to be enhanced (in those Member States where national rules are unsatisfactory) after the Third Energy Package is adopted and enters into force.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxix

Introduction, Summary, Remarks

cxix

might be in a position to tip off the Commission in the event of failure to respect the order, thus relieving the Commission of monitoring duties. However, if this kind of strategy proved unavailing, the Commission could still impose a structural remedy. A cease-and-desist order in the “shadow” of the Commission’s break-up powers might yield the desired result while enabling the Commission to respect the proportionality principle expressed (with unusual specificity) in Recital 12 of Regulation 1/2003.247 3. Pierre Régibeau’s paper, The (Complex?) Relationship between Art 82(a) and Intellectual Property Rights, concentrates on two basic questions. First, are excessive pricing actions appropriate in settings involving patent (or copyright) protection? And if so, would it be justified to adopt an enforcement policy that accords special deference to industries with high rates of R&D spending (in particular, by accepting R&D expenditures as part of a defendant’s relevant costs)? Along the way, Régibeau offers a series of observations regarding some of the problems that arise at the famous competition/IP intersection.248 The starting point is the temptation to resort to excessive pricing actions under Article 82(a) in the IP context, since exclusive rights tend to boost price/cost margins.249 However, a high price/cost ratio is precisely the point of the patent system, as it tends to stimulate innovation and thus yields social benefits (against some difficult-to-measure opportunity costs). A zealous application of Article 82(a) would undermine or defeat the purpose of the exclusive rights granted to the inventor. In addition, Régibeau points out, there are significant or indeed intractable difficulties associated with most excessive pricing cases, such as the identification of proper benchmarks, allocation of costs, and design of remedies. He would not go so far as to impose a US-style ban on excessive pricing cases. Article 82(a), he says, provides at least some residual value. This is partly because, as recognized by

247 Note that such an approach is not the same as ex ante price regulation. In the scenario envisaged, it has already been ascertained ex post that the defendant systematically charged prices that were excessive and insulated against erosion by persistent entry barriers. It can be argued, therefore, that this kind of price control is not inconsistent with the goal of liberalizing the energy sector. Normally, the problem with price regulation is that it creates an entry barrier and protects the incumbent, whose losses resulting from a price cap are more than offset by the additional rents accruing from its protected position. But here the cause for the lack of entry is not the price control but pre-existing entry barriers. 248 Refusal by a dominant firm to license its IP is not discussed in Régibeau’s paper, which, as its title suggests is concerned with exploitative but not exclusionary conduct. Even a constructive refusal to deal, whereby the dominant firm applies a price too high for a would-be licensee to pay, should not be characterized as an exploitative abuse, since no transaction takes place and hence the refusal does not directly cause a transfer of surplus. For a discussion of the European approach to refusals to license IP, see Patterson, this volume, pp. 705 et seq. 249 Unlike van der Woude, Régibeau does not discuss “fairness” as a distinct concept; rather, he interprets the provision as a prohibition of “unacceptable price-cost margins”. In most cases, this probably boils down to a semantic difference. However, as van der Woude notes in his paper, a dominant firm can incur losses and yet still be condemned for excessive prices.

(AA) Ehlermann 08 Introduction

cxx

12/11/08

10:25

Page cxx

Introduction, Summary, Remarks

some of the other participants (see Panel V), enforcement in cases of exclusionary behaviour (which in principle is the preferred approach because it aims to address causes, not symptoms) is imperfect. Still, he does demonstrate that the complexity of the exercise is exacerbated in the IP context by issues such as definition of the relevant market(s), assessment of dominance, and the appraisal of the “price” charged by a licensor (since, for example, the value of bargained-for ancillary covenants is uncertain). In section V of the paper, Régibeau reflects on the narrow category of situations in which the foregoing obstacles have been overcome—situations of “fairly extreme dominance”. It is within this narrow band of cases that the question of a latent conflict between Article 82(a) and IP protection might require attention. In this context, Régibeau reviews and comments on the literature regarding optimal patent design, since the breadth and duration of protection has a profound impact on profit margins,250 and he then examines the “directedness” of innovation, i.e., the degree to which R&D proceeds along a pre-determined track with targeted results.251 Where R&D is directed toward specific fields of research, he says, “it is important for the combined impact of patent policy and Article 82(a) to be neutral across sectors of innovative activities so that the relative private returns to R&D projects accurately reflect their relative social value”.252 However, empirically speaking it appears that certain R&D-intensive industries (pharmaceuticals, chemicals) sometimes benefit de facto from greater breadth of patent protection than other industries, due in part to differences in the way the protected claims are written and their enforceability. Under those circumstances, the “reasonable” application of Article 82(a) (that is to say, applying it judiciously and avoiding zealotry) could help to correct for a misallocation of resources favouring better protected and hence more profitable sectors. Conversely, letting Article 82(a) languish might aggravate the distortion. Finally, as an additional argument for the neutral application of Article 82(a) across sectors, Régibeau recalls that a policy which rewards R&D expenditures by reducing the relevant price/cost margin might simply encourage a surplus of costs. 4. The title of Damien Geradin’s contribution is Abusive pricing in an IP licensing context: An EC competition law analysis. The overarching argument 250 The thrust of the literature is that, from a social welfare perspective, the ideal is a long-life patent that confers finely tuned protection and leaves room for significant sequential and collateral innovation. As Régibeau points out, if patent systems lived up to that ideal, there would be a greater degree of substitute technologies and a lesser likelihood of dominance; consequently, any tension between Article 82(a) and patent protection would ordinarily be minimal. However, this is essentially a theoretical point because in practice there is significant breadth of protection in certain sectors (see infra, main text). 251 In general, the more that research is targeted, the greater the possibility that certain enforcement decisions might trigger a misallocation of resources by diverting investment to less highly lucrative fields to avoid the risk of intervention. 252 Page 664.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxxi

Introduction, Summary, Remarks

cxxi

is that “competition authorities and courts should proceed with extreme caution when facing claims that an IP licensee charges excessive royalties or abusively discriminates between its licensees”.253 Section II of the paper concerns market definition and dominance in technology markets, with a particular emphasis on market power where certain IP rights are essential to an industrial standard. The basic economics of market definition in this context are described, and Geradin shows how certain competitive constraints254 may mean that, even where the IPRs are essential to implementing a standard, the right holder should not be regarded a priori as being dominant. With considerable enthusiasm, he adds that standard-setting and licensing may well occur in a dynamic setting, with fatal winds gusting about that could either knock a giant on his hindquarters or at least force him to go foraging elsewhere for his grub. To be sure, there are real-life markets exhibiting genuine dynamic competition and transient monopolies. Generally, however, this should not be taken to mean that entire sectors are driven by such competition, or that that the possibility of sustainable dominance in such sectors is excluded. Dominance remains an empirical and fact-sensitive matter to be tested in individual cases. The next section of the paper discusses the relevance of Article 82(a) for the grant of IP licences against royalty payments, and it notes that this issue may become more salient in the age of the knowledge economy. Geradin begins with a series of observations stressing that price regulation is inappropriate in innovation-oriented sectors due to the need to preserve ex ante incentives to invest in R&D as well as incentives to stimulate market entry and to license IP out to third parties. This point of view does not seem controversial, as the discussion in Europe is not so much about whether systematic price controls should be imposed but rather about whether there is a narrow niche of cases in which Article 82(a) could do any good.255 Geradin then critiques the jurisprudence in this area and demonstrates the lack of consistency in the cases as regards the proper methodology for identifying an excessive price within the meaning of that provision. Implicitly, this would seem to be one more reason to abandon the exercise, at least in high-tech sectors. If that is a correct understanding, it is not clear whether the conclusion follows from the premise: one might also interpret the case law as confirmation that different markets, with different underlying economics, call for different approaches. 253

See page 702. These constraints may include: (i) competition between rival standards or downstream between the finished product and other products, regardless of whether the latter incorporate other standards; (ii) a tendency among holders of complementary technologies in a standard to take account of the level of royalties charged by the other contributors to the standard, and to adjust their own relative royalties downward, à la Cournot and his famous brass example; and (iii) ongoing competition in respect of later generations of the standard in a (perhaps rapidly) changing environment, also taking into account the strategic incentive not to be bypassed or otherwise punished in later rounds for unduly selfish behaviour. 255 This seems clear from the papers and discussions in Panel V and here in Panel VI. 254

(AA) Ehlermann 08 Introduction

cxxii

12/11/08

10:25

Page cxxii

Introduction, Summary, Remarks

However, on the specific question of whether there is any satisfactory methodology for determining whether royalties in a complex and bargainedfor licence are “unfair” (assuming there was some degree of real bargaining256), the argument seems more persuasive.257 It is undoubtedly true that taking on an excessive royalty case is likely to lead an enforcer or court into a quagmire of cost measurement and allocation dilemmas, only to reach another quagmire if it must then identify relevant price benchmarks. Geradin proceeds with a discussion of how the picture looks in the standard-setting context. He challenges the view that the adoption of a standard necessarily increases the market power of an essential IPR holder and facilitates ex post opportunism, although he does seem to recognize that the adoption of the standard can create market power if it effectively eliminates competition between the “winning” IP and close substitutes. In circumstances where the standard does imply additional market power for the essential IP holder, there is no reason, according to Geradin, why he should not be entitled to some of the benefits of standardization.258 Geradin is also sceptical about the claim made by some scholars that individual contributors to a standard essentially stack their individually reasonable royalties so high that in the aggregate they become unreasonable. In other works, Geradin and collaborators have challenged the royalty-stacking argument, largely on the following grounds: the theory finds little empirical support; it can only be valid (but is not automatically so) if a series of conditions are met; and mechanisms exist to cope with the externalities problem.259 Finally, drawing on his earlier work with Nicolas Petit, Geradin turns to the argument on discriminatory royalties. The basic conclusion here is that the 256 Two points should be noted here. First, if a licensee has significant bargaining power, this may preclude a finding of dominance in the first place. Second, if there is no meaningful bargaining between a dominant licensor and its licensee and if the licence is thus more akin to an adhesion contract, then retaining the possibility of intervening under Article 82(a), provided strict conditions are satisfied on the basis of firm and consistent evidence, might be more defensible. In the context of standards, FRAND commitments may in some cases help to alleviate asymmetric bargaining power. 257 Cf. Pierre Régibeau’s paper, this volume (taking a similar view, but nevertheless concluding that Article 82(a) should not be removed from the enforcer’s toolkit). Incidentally, it is recalled that the context here concerns exploitative behaviour. The analysis might differ in a scenario where a dominant licensor competes with its licensee(s) downstream and where the royalties are so high that they effectively result in margin squeeze-type exclusion. 258 Compare this scenario to the grant of monopoly rights by public authorities, which was singled out by Lars-Hendrik Röller (see supra notes 204–207 and accompanying text) as a kind of “gap” case for which Article 82(a) might be suitable because, under those circumstances, dominance is not acquired through competition on the merits. The additional market power conferred by a standard, if any, might be distinguishable if institutional arrangements serve to ensure that the IP incorporated in the standard is chosen according to objective performance criteria. Furthermore, as Geradin points out, essential IP in the context of standards may be disciplined by certain competitive pressures (see supra note 223) with which, by contrast, a legal monopolist need not be concerned. 259 See Damien Geradin, Anne Layne-Farrar and Jorge Padilla, “The Complements Problem within Standard Setting: Assessing the Evidence on Royalty Stacking” (2008), at http://papers. ssrn.com/sol3/papers.cfm?abstract_id=949599.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxxiii

Introduction, Summary, Remarks

cxxiii

two requisite elements of Article 82(c)—dissimilar prices or trading conditions imposed in equivalent transactions, and competitive disadvantage downstream—are rarely met in licensing scenarios. For example, IP licences are often heterogeneous in terms of their material, geographic and temporal scope, not to mention their payment or cross-licensing terms.260 Furthermore, discriminating against licensees in such a way as to prejudice their position downstream (secondary line injury) will frequently run counter to the interests of the licensor if the latter is not itself active downstream. By contrast, where a dominant licensor is vertically integrated and competes with its licensees, which takes us back to potentially exclusionary behaviour, Geradin favours some degree of scrutiny to guard against vertical foreclosure. In that regard, following Swanson and Baumol, he endorses an efficient component-pricing approach. 5. Bringing up the rear in this 26-paper march is Mark Patterson’s The Peculiar “New Product” Requirement in European Refusal to License Cases: A US Perspective. More than any of the other contributions in Panels V or VI, Patterson’s paper focuses on exclusionary conduct, in particular in the welltrodden field of refusals to license. While commentaries on cases such as Magill and IMS Health abound, Patterson takes a rather different tack. To understand his argument it is perhaps useful to recall that the “new product” requirement discussed in Magill seemed to be driven by consumer interests, that is to say, it was driven by a concern that the suppression of Magill’s 260 By way of digression, one question that might conceivably arise in relation to licence payments is whether “intensity of use” royalties discriminate among licensees, all else being equal. See Michel Waelbroeck, “Price discrimination and rebate policies under EU competition law”, Fordham Corporate Law Institute, chapter 10 (1995). If licensees X and Y each take an otherwise similar licence, but X pays the licensor greater royalties due to the greater use it makes of the IP in its manufacturing process, then the licensor’s profit margin will differ, assuming the licensor incurs equal marginal costs in licensing its IP to each licensee. On the face of it, this appears to meet Posner’s definition of price discrimination, quoted at footnote 95 of Geradin’s paper. From the perspective of X, the royalty methodology might be considered especially egregious because it “penalizes” greater output, and perhaps greater efficiency, as compared to Y’s production. That argument should be rejected, as it fallaciously supposes that the licences described above are homogeneous. Indeed, it is highly doubtful that the two transactions are equivalent: the benchmark should not be the total receipts from X and Y but rather whether the price they pay for each unit of use is equivalent. In a sense, X has purchased a greater number of units of use than Y. To focus instead on the licensor’s marginal costs and to insist on uniform compensation from X and Y despite their different intensities of use could only create a windfall for X at the expense of Y. Second, even if the transactions were deemed equivalent, one could not credibly maintain that X, by using the IP more intensively and presumably increasing its own sales, has been placed at a disadvantage relative to Y within the meaning of Article 82(c). Third, opening the door to discrimination claims where the royalty is based on intensity of use could implicitly call into question ad valorem royalties, since, there again, the dominant licensor’s profit margin across licences will vary if licensees’ sales differ. This would lead to an odd outcome since ad valorem royalties, a ubiquitous device for securing a return on licensed IP, are not normally regarded as appreciably restricting competition (at least where the licensed IP is incorporated in the product sold). See Commission, Technology Transfer Guidelines, cited supra note 130, paras. 156 and 160.

(AA) Ehlermann 08 Introduction

cxxiv

12/11/08

10:25

Page cxxiv

Introduction, Summary, Remarks

consolidated program listings would frustrate consumer demand. From there it might seem a small step to conclude, as the ECJ intimated in Magill and then confirmed in IMS Health, that if the dominant firm is already offering the product on the secondary market, this should defeat claims of a duty to license. Patterson essentially asks: is that really the right approach? To make his case, Patterson examines some of the possible constructions of the “new product” requirement. For example, on one reading it might call for distinct product markets; or it might merely contemplate some degree of differentiation between the claimant’s product and that of the dominant IP holder. Depending on the proper construction of the test, therefore, it appears that where there is competition between the products of the two parties (first reading) or where the products are close substitutes (second reading), the claim for a licence is likely to fail. Building on that point, Patterson posits a hypothetical scenario according to which the products of the parties are in fact identical, yet in this case the coveted IP represents a tiny (though essential) component of the homogeneous product. Is the new product criterion robust under those circumstances, or is it underinclusive? In some cases the issue of whether the claimant merely seeks to market a clone will be an unsatisfactory linchpin. Patterson proceeds to critique a US case from 1999, Intergraph v. Intel. In that case, the Federal Circuit took an approach that seems to diverge from that of the ECJ in the sense that the Federal Circuit made access to the facility conditional on competition between the two parties on the market downstream of the facility in question. Intergraph leads Patterson to consider the scenario in which the claimant and the defendant are competitors, but on some other market—not the one for which the IP is essential. This might leave the claimant with the hard choice of either staying put on that other market and suffering the consequences, or dropping out of that market as a quid pro quo for the licence it seeks.261 For quite different reasons, neither the Intergraph approach nor the IMS Health approach appears to condemn a defendant for playing that kind of hard ball. For Patterson, however, “it makes little sense to give monopolists carte blanche to do harm whenever they can find a way to link two markets in which they face competition”,262 all the less so when the coercive instrument is IP unrelated to the market in respect of which the defendant seeks to make life difficult for the claimant. In light of the drawbacks to the two tests described above, Patterson suggests an alternative approach undoubtedly influenced by his experience as a patent lawyer. Rather than focusing on whether the dominant IP holder is supplying the same product downstream in competition with the party seeking access, he would focus instead on the subject matter covered by the IP. To 261 Patterson also notes the likeness between this kind of retaliation scenario and some famous European cases. 262 Page 717.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxxv

Introduction, Summary, Remarks

cxxv

neglect the scope of the IP misses the point, in his view, because IP law is designed precisely to create boundaries between legal rights of exclusion and the public domain. If the activity the claimant seeks to pursue is not something that could be precluded in an infringement action, then a rule that permits the dominant IP holder to refuse access simply by operating downstream effectively re-draws those boundaries. From an IP law perspective, the question would be: why should a firm be allowed to monopolize a (downstream) market in which it has not produced any innovation worthy of protection? Therefore, as an alternative to the “new product” test as it is presently formulated, Patterson suggests an inquiry as to whether the downstream product falls within the scope of the claims of the patent (or the expression covered by the copyright), including any protection the IP holder also enjoys downstream. The paper also sets forth a number of reasons why this proposed scope inquiry would not unduly relax the test for liability in cases of refusal. It seems that, if such a test were adopted in Europe under Article 82, the claimant would have to show that the IP is indispensable to produce the downstream products, and that the product would not fall within the scope of the dominant firm’s IP. If those elements were established, the dominant firm would then have an opportunity to demonstrate why its refusal to license the IP is objectively justified, possibly on the basis that a compulsory licence would preclude it from deriving an adequate return on its creative efforts.263 Clearly, the law is not about to mutate toward such an IP law-based approach in the near future, but the idea does seem to merit further reflection.

B. Discussion The roundtable discussion following the sixth panel of speakers (Hellwig, van der Woude, Siragusa, Régibeau, Geradin, Patterson) was brief due to time constraints. Although Hellwig, van der Woude and Siragusa had addressed issues in the energy sector, the discussion focused more on issues arising in the intellectual property context. For example, Bill Blumenthal noted that, in the US, considerable deference is accorded to IP rights. The US position, he says, is not unlike that expressed by Damien Geradin in his presentation, especially as concerns the importance of dynamic competition.264 From a US perspective, any intervention to reduce royalties on grounds of excessive pricing would be contrary to the 263 Patterson suggests that this “monopoly reward” argument could be considered either in the context of objective justification or as part of the question of whether granting a licence is feasible, an element of the essential facilities doctrine. 264 See, e.g., Thomas Barnett, “The Gales of Creative Destruction: The Need for Clear and Objective Standards for Enforcing Section 2 of the Sherman Act”, opening remarks at the Section 2 hearings of the DOJ and FTC, Washington, D.C., 20 June 2006, available at http://www.justice.gov/atr/public/speeches/216738.pdf.

(AA) Ehlermann 08 Introduction

cxxvi

12/11/08

10:25

Page cxxvi

Introduction, Summary, Remarks

principle that IP carries with it the right to choose not to license it to others. This is equivalent to charging a royalty that the would-be licensee cannot pay, including an infinitely high royalty.265 David Meyer agreed with the notion that undermining IP rights is no task for antitrust authorities. Meyer stressed that IP is no different, in that respect, from any other kind of property, such as land. Daniel Zimmer was interested in the tension between: a (hypothetical) aggressive application of Article 82(a) where a dominant firm holds IP; and one of the driving purposes behinds IPRs, which is to create a time-limited opportunity to apply above-cost pricing. Zimmer suggested that a similar conflict between national IP systems and the application of Article 81 may lie behind the judgment of the CFI in GlaxoSmithKline (currently on appeal), where the court rejected the notion that GSK’s dual pricing scheme, openly designed to suffocate parallel trade in certain pharmaceutical products, should be regarded as a restriction “by object” under Article 81(1). With respect to both Article 81 and Article 82, Zimmer wonders whether one might not deduce a rule of reason,266 whereby competition law stands down so as not to do injury to national systems of IP protection. If this is a correct interpretation of GlaxoSmithKline, it would certainly seem a significant departure from the ECJ’s traditional distinction between the existence of IP rights and their exercise.267 In any case, it is equally plausible that the CFI was seeking not to create a safety zone for IPRs but (inspired, one would think, by Advocate General Jacobs’ opinion in Case C-53/03, Syfait) to introduce more rigour in the analysis of when a practice may be said to restrict competition. With regard to the patent system, Damien Geradin was concerned that defensive patenting, strategic patenting, and in general patenting en masse was getting out of hand.268 He also called attention to the fact that in the standard-setting context, a single standard might include thousands of 265 In the extremely narrow class of cases where US law might step in and constrain an IP holder to grant a licence, the infinite royalty view must of course give way, as the compensation exchanged in those circumstances must necessarily be a price that the licensee can afford. 266 Note that Zimmer is not referring here to a US-style rule of reason, in any of its forms. The allusion is rather to the quite different Cassis de Dijon-style rule of reason, also seen occasionally in the competition law context (above all in Case C-309/99, Wouters, or in Case C-67/96, Albany). According to this European rule of reason (an unfortunate appellation), a fundamental rule of the EC Treaty (normally a free movement rule—Articles 28, 49, etc.) is effectively disapplied as a means of avoiding undue interference with a Member State’s regulatory prerogatives, provided the national regulation is suitable and proportionate to the objectives it aims to achieve. Examples of overriding national interests include, for example, environmental protection, consumer protection, certain cultural activities, fiscal supervision requirements, the sound administration of justice (in the Wouters case), and many others. 267 Valentine Korah has often been critical of the existence/exercise divide. See, e.g., Korah, “The interface between intellectual property and antitrust: the European experience”, 69 Antitrust Law Journal 801, 805 (2001). 268 See, e.g., Robert Cowan et al., “Policy Options for the Improvement of the European Patent System” (May 2007), at http://www.tekno.dk/pdf/projekter/patent-system-STOA/p07_ STOA_Background_document_and_report.pdf, pp. 29 et seq.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxxvii

Introduction, Summary, Remarks

cxxvii

essential patents belonging to numerous members of the SSO. Since these tendencies can at least potentially put additional strain on the antitrust system, Geradin considers that institutional reforms are needed in both settings. Martin Hellwig picked up on Geradin’s point regarding some the dysfunctional features of the patent system, citing Dietmar Harhoff, who has worked in this area.269 Hellwig observed that, due to the advent of over-patenting, the margin for new innovators to operate is growing ever slimmer. He also notes that, although competition specialists sometimes consider this to be a patent law problem, at least some intellectual property experts have called for greater vigilance on the antitrust side.270 Claus-Dieter Ehlermann was fittingly the last to intervene before Philip Lowe proceeded with his closing address. Ehlermann’s point was an institutional one, linked to the unique features of the European Union. Unlike national systems, he noted, legislation in the EU is often partial and fragmentary. This applies in particular to intellectual property, where there is a Community Trade Mark, for example, but nothing comparable to a “federal” patent or copyright system. The EU is also distinct from national systems in that (leaving aside the unusual case of the European Central Bank) it has no full-fledged regulatory bodies with decision-making powers. The only “regulator” at Community level is therefore the Commission when it exercises its competition powers. To that extent, it is possible that some people might expect DG Competition to handle problems for which neither an antirust enforcer nor antitrust law is necessarily well-suited.

269 See, e.g., Dietmar Harhoff, “Strategic Patenting and Patent Policy”, slide presentation, Brussels, 19 September 2007, available at http://ec.europa.eu/dgs/policy_advisers/docs/strategic_ patenting_Harhoff.pdf. 270 See, e.g., Ghidini, The Innovation Nexus, cited supra note 76.

(AA) Ehlermann 08 Introduction

12/11/08

10:25

Page cxxviii

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 37

I David J. Gerber* The Future of Article 82: Dissecting the Conflict August 2007

Underlying the recurring debates over the future of Article 82 EC are competing images of what its goals are and should be. Such debates about the interpretation and application of Article 82 are not new, and they are also not likely to end, because the legal concept of “abuse” is sufficiently abstract and capacious to allow multiple conceptions of its goals.1 Where goals become contested and controversial, however, debates can lead to confusion and uncertainty rather than progress in thinking about the issues, and this threatens to occur in the context of discussions of Article 82 and its future. Clashing images of the goals of that provision have yielded much uncertainty about the future of the law in this area. They have also distorted images of both the existing law and of newer alternatives to it. This impedes the capacity of European judges and administrators to apply the law consistently and effectively. It should be valuable, therefore, to identify and assess the lines and contours of these debates and the images of law, economics and European integration that swirl within them. A careful look at the discussions surrounding the future of Article 82 reveals two basic conceptions of its goals. One centers on the idea of “competitive distortion”. It has been given shape and content over decades by the European Courts, by the European Commission, and by academic analysts. A competing conception was first given substance by DG Competition in 2005.2 It is based on the “more economic approach” to competition law that the Commission has pursued since the late 1990s, which is similar to current antitrust orthodoxy in the United States. This new version of the goals of Article 82 uses “consumer welfare”, as defined by neo-classical economics, as

* Distinguished Professor of Law, Chicago-Kent College of Law. 1 These debates and discussions are interwoven with the story of the development of competition law in Europe. For an in-depth treatment of that development, see David J. Gerber, Law and Competition in Twentieth Century Europe: Protecting Prometheus, Oxford University Press, 1998 (paperback 2001). For discussion of the state of thinking about the goals of European competition law in the late 1990s, see Claus-Dieter Ehlermann and Laraine L. Laudati, eds., European Competition Law Annual 1997: Objectives of Competition Policy, Hart Publishing, 1998. 2 See DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf.

(C) Ehlermann 08 Gerber

38

6/11/08

11:46

Page 38

Session One: The Sherman Act And Abuse Of Position Under Art 82

the standard for determining whether conduct constitutes “abuse” for purposes of Article 82. The discussions and debates have produced valuable insights, but also potentially significant misunderstandings. Misconceptions and confusion regarding each of the two main conceptions of Article 82’s goals have become a serious impediment to constructive dialogue about the future of this area of European competition law. Moreover, there has been curiously little analysis of the structure of the conflict and of the specific differences between the law as it has been developed and proposals to change it. Dissecting the conflict should thus be useful in moving the debate forward and providing a firmer basis not only for policy decisions, but also for the decisions that European decision makers, including national authorities and national judges, will have to take in individual cases in the future. Accordingly, this paper will examine several basic questions: What kinds of differences and changes are involved in this debate? How far apart are the two main conceptions that are in conflict? And how can the differences best be understood? The paper has four main objectives. One is to analyze the debate over the goals of Article 82. There has been no lack of discussion of why one approach is better than another, but that is not the goal here. My focus here is on the debate itself, the positions and claims that appear in it, and the factors that influence the way it is framed and interpreted. A second objective is to identify the similarities and differences between the current conception of the goals of Article 82 and the new approach being considered by the Commission. As the Commission and the courts seek to develop and improve the application of Article 82, there will be a continuing tension between current and past conceptions of the provision’s goals, on the one hand, and a newer conception of its goals, on the other. It is critically important, therefore, to understand just what those differences and tensions are. A third objective is to identify misconceptions and confusions that have become part of the thinking and discussions in the area. Such confusions and misconceptions can impair effective discussion and undermine the value of the policy decisions that derive from it. A fourth and more tentative objective is to identify ways in which this analysis can be beneficial for the future of thinking about this area of European competition law. The analysis leads to the following basic claims. First, the goals underlying Article 82 and pursued by European institutions are often misrepresented. Second, the failure to distinguish between two fundamentally different roles for economics leads to confusion and misconceptions regarding the reform agenda, in particular, and the use of a “more economic approach” in the application of Article 82, in general. Third, when these misunderstandings and confusions are identified, the conflict turns out to be narrower than often thought, but also perhaps more fundamental. Finally, this analysis suggests a path forward in thinking about and implementing Article 82, one that avoids the pitfalls of casting aside experience gained, but at the same time it incorporates

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 39

The Future of Article 82: Dissecting the Conflict

39

important advantages of increasing the use of economics in the application of Article 82.

I. The Competitive Distortion Model The basic idea behind the abuse concept is simple. It starts with the proposition that some firms have sufficient power to distort and thereby to harm the competitive process. Given that such distortions reduce the capacity of competition to produce wealth and benefit consumers, Article 82 should accordingly be used to deter such deleterious conduct. European institutions have built on the concept of “competitive distortion” as the basis for an entire conception of the role of Article 82. I will refer to this conception as the “competitive distortion” model.

A. Evolution of the Model The evolution of the basic concept is often misunderstood. The idea originally took shape in 1920s Europe, where there was concern that economically powerful enterprises could distort the competitive process and thus prevent Europe from overcoming its severe economic problems.3 As such, the competitive distortion model was conceived as a tool for improving economic performance. In that role, it was embodied in several European national laws, and it acquired international standing at the World Economic Conference in Geneva in 1927.4 Economic depression and war halted its development until after the Second World War, when it began to find its way back into European law and policy. The so-called “ordoliberal” school of law and economics contributed intellectual contours for this idea, and it played prominent roles in the early development of competition law in Europe, at both the national and Community levels.5 The term “ordoliberal” was originally applied to a small group of primarily German lawyers and economists who began to develop a new conception of the relationships among law, economics and economic policy. Their central claim was that law should be used to protect the 3

Gerber, supra note 1, at 115–164. For discussion of the conference, see Carl H. Pegg, The Evolution of the European Idea, 1914–1932, University of North Carolina Press, 1983, at 76–87. For its treatment of cartel problems and further references, see Gerber, supra note 1, at 159–162. 5 For detailed discussion, see David J. Gerber, “Constitutionalizing the Economy: German Neo-liberalism, Competition Law, and the ‘New Europe’”, 42 American Journal of Comparative Law 25 (1994). 4

(C) Ehlermann 08 Gerber

40

6/11/08

11:46

Page 40

Session One: The Sherman Act And Abuse Of Position Under Art 82

competitive process from interference by both the state and private actors. Legal institutions could, in effect, support and embed market competition as a specific form of economic organization by combating interference with its operations. During the Nazi period, ordoliberals often operated underground, but during the post-War occupation many were given prominent positions in the economic administration in Germany, and this enabled them to play major roles in the development of the “social market economy” and the German “economic miracle” associated with it.6 Competition law was a central component in the ordoliberal program, because it was the tool that could be most directly deployed to combat interference with a competitive economy. The ordoliberal conception envisioned a normative framework that combated agreements that distorted competition as well as single-firm (or “unilateral”) conduct of dominant firms that had the same or similar effects. It foresaw an independent and highly trained administrative body that would have authority to intervene directly against conduct it considered harmful to the competitive process. Although the focus of attention was generally on cartel conduct, the concept of abuse of dominance was developed to apply to single-firm conduct that had these harmful effects. These ideas combined with experience drawn from the decartelization laws enacted by the US during the occupation period and with lessons drawn from US antitrust experience to form the foundation for the German Law Against Restraints on Competition (GWB), which entered into effect in 1958 after a long and highly visible debate. The law created a comprehensive competition law that prohibited both contractual restraints on competition and the abuse of a dominant position (merger provisions were added almost two decades later). Of particular importance for our purposes is the implementation mechanism. This statute set up an independent agency, a Federal Cartel Office (FCO) that had primary and often sole responsibility for applying the national competition law provisions, and it subjected the FCO’s decisions to full, substantive review by the regular courts. The regular courts thus play an important role in the system, because they review the decisions of the FCO not only for procedural defects, but also for substantive error. This requires that the FCO’s decision-making process approximate judicial decision making, thereby assuring rigorous legal scrutiny of the competition law concepts that are applied under the GWB and engendering careful juridical development of those concepts. The concept of abuse of dominance was developed and applied by the FCO and the courts using this mechanism. Accordingly, and contrary to some assertions, the abuse provision was not conceived as an authorization for the exercise of discretion by the administrative authorities. The concept had to be developed in ways that satisfied the stringent requirements of the German judiciary, and those courts assured a high level of juridical develop6

I have described this development in detail elsewhere. See Gerber, supra note 1, at 257–265.

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 41

The Future of Article 82: Dissecting the Conflict

41

ment of the concept. Moreover, this process has been accompanied by a highly developed scholarly discussion of abuse issues. The German contributions to developing the abuse concept have been highly influential throughout Europe.7 For our purposes here, the focus is on the development of the concept of abuse by institutions of the European Union—in particular, the Commission and the European Courts.8 The evolution of the concept at this level has been widely studied, and thus I underscore here only two basic points that are particularly relevant to current discussions of the law in the area. One is that the law relating to the abuse of dominance has been developed almost exclusively in cases, including the decisions of both the Commission and the European Courts. This is important, because case law development that involves relatively abstract goals tends to generate inconsistencies and uncertainties over time. For example, the development of case law under Section 2 of the Sherman Act is at least as unclear and inconsistent as is the case law under Article 82.9 The second point is that the development has not been based on discretionary administrative action, but on rigorous examination of concepts and their practical application produced by courts, administrators and scholars, and based on experience at both the national and European levels. Although critics of the current law often paint the area as hopelessly uncertain, there was no major outcry along these lines until the reform process began in the mid-1990s. To be sure, there were complaints about uncertainty in the application of Article 82, but the situation was generally viewed as a natural result of the complexities of the issues and the early stage of development of the legal principles. Concern with this uncertainty has grown dramatically since the Commission’s reform project started in the mid-1990s, and much of it has come from those espousing an alternative model, which employs, as we shall see, a very different perspective (i.e., economic analysis) which until recently had had relatively little influence on competition law development in Europe.

B. Key to Analysis: Recognizing the Structure of Goals Article 82 does not have one simple and easily defined goal, but refers to a structure of goals in which individual goal elements are related to the central 7

See ibid. at 306–16. For a recent discussion of the case law under Article 82, see Thomas Eilmansberger, “How to Distinguish Good from Bad Competition under Article 82 EC: In Search of Clearer and More Coherent Standards for Anti-Competitive Abuses”, 42 Common Market Law Review 129 (2005). For detailed analysis of the case law from an economics perspective, see Robert O’Donoghue and A. Jorge Padilla, The Law and Economics of Article 82 EC, Hart Publishing, 2006. 9 See the in-depth analysis of generally analogous US antitrust law provisions in Einer Elhauge, “Defining Better Monopolization Standards”, 56 Stanford Law Review 253 (2004). 8

(C) Ehlermann 08 Gerber

42

6/11/08

11:46

Page 42

Session One: The Sherman Act And Abuse Of Position Under Art 82

idea of competitive distortion. This central objective gives content to each of the subsidiary concepts and relates each one to the others. Identifying this structure is thus essential for understanding what Article 82 has meant and how it has been used. In this conception of the abuse provisions, a central tool for identifying abusive conduct is the concept of “competition on the merits”.10 Where a dominant firm’s conduct is not “on the merits”, it is presumed to be possible only because the firm has a dominant position that allows it to operate with a degree of freedom from competitive constraints. Where such dominant firm conduct significantly alters the operation of a market by reducing the capacity of other firms to compete, the conduct is presumed to constitute an abuse of dominance. The objective is to allow the market to develop without being subject to distortions caused by a dominant firm’s “non-competitive” conduct. Some critics claim that this protects the structure of the market and thus inhibits competitive adaptation to change. This misconstrues the objective that has driven this conception of the law, which is to create a legal framework within which markets can develop on the basis of competitive conduct. It thus protects the conditions of operation on the market from distortions in order to allow the market to develop competitively. Note that application of this test requires use of a time dimension in which harm to the competitive process can be effectively assessed. The breadth of the time perspective depends on the characteristics of the market as well as the potential effects of particular conduct. Cases under Article 82 mention other goals for the abuse provisions, and interpreting these references has been a major source of misconceptions in this area. A key factor in resolving these misconceptions is to recognize that all such other goals relate to the central objective of combating distortion of competition and that the other goals have been developed in pursuing that objective. Among the most important of these are the concepts of fairness, European integration, and economic freedom. In each case, the concept is given meaning and understood in relation to the goal of protecting the process of competition against distortion. Often, however, the cases do not clearly explain the relationship between these goals and the central goal of preventing competitive distortion, and this has meant that over time the structure of goals has become less clear even to those applying the law. The concepts of fairness and economic freedom have been used, for example, to provide means of identifying distortions of the competitive process and assessing the extent of such distortions. If conduct causes harm 10 This concept has been subject to strong criticism from economists for its lack of precision, and it is unquestionably less precise than economic concepts. See, e.g., O’Donoghue and Padilla, supra note 8, at 176–8. Viewed from the broader perspective that has heretofore been used in assessing European competition law, however, it has generally been seen as a practicable analytical tool, at least until recently. Differences in views of this concept provide a revealing example of the extent to which a specifically economic perspective can alter assessments of case law.

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 43

The Future of Article 82: Dissecting the Conflict

43

to a competitor using methods that are considered “unfair” in the marketplace, this may indicate that the conduct is not “competition on the merits” and, therefore, that it could have a distorting effect on the competitive process. In this sense, the concept is not a “test” for actually determining that conduct distorts competition; it is rather an indicator of potential distortion. This function can be understood only if it is seen as part of a framework of goals as opposed to an independent goal capable of standing alone in determining whether competitive distortion has occurred. The same is true for the concept of “economic freedom”. This notion has been used to approximate the limits beyond which a dominant firm’s conduct becomes abusive. From this perspective, for example, it has been said that a dominant firm’s economic freedom cannot extend beyond the point where it impedes the effective rights of others to compete. This concept is also insufficiently precise on its own terms to identify conduct as abusive, but it can indicate situations in which further analysis must be used to determine the extent of harm which the conduct poses for the competitive process. The goal of European integration has been developed to counteract distortions of the competitive process associated with the existence of political borders within Europe. Where legal impediments such as intellectual property rights impede competition across borders within the European Union, the abuse provision has been used to assert the unity of the European market. The capacity of a dominant firm to use its market power to prevent competition across borders is seen as a potentially serious distortion of the competitive process, especially because it involves political borders and thus may implicate the enforcement powers of the state. The main point is that this goal derives from and applies the concept of competitive distortion, but here the goal is further defined by the specific context of the process of European integration.

C. Misconceptions about Article 82’s goals The lack of clarity in the case law about the objectives of Article 82 opens the way not only for divergent interpretations of the abuse concept but also for misconceptions of it. The reform controversy of recent years may tend to encourage such misconceptions by creating unconscious “confirmation biases” that can lead commentators to portray opposing conceptions negatively. I identify below four misconceptions that have been prominent in recent discussions and then apply the above discussion in responding to them. One misconception claims that Article 82 contains a mélange of unrelated goals and that it cannot, therefore, provide a basis for effective decision making. If an observer looks only at selected cases, this may appear to be an accurate description of the goals of Article 82, because, as noted, the body of

(C) Ehlermann 08 Gerber

44

6/11/08

11:46

Page 44

Session One: The Sherman Act And Abuse Of Position Under Art 82

cases as a whole contains references to several goals, and the decisions do not always or even often explain how those goals relate to each other. Where the goal references in cases are viewed in the context discussed above, however, they no longer appear random. Recognizing the structure of goals allows the analyst to see how seemingly disparate and unrelated goals relate to each other and to the underlying concept of competitive distortion. The lack of attention to the structure of goals in competition law decisions is unfortunate, but it is not surprising, given that the task of those who make such decisions is to decide a specific case and explain the reasons for it, rather than to clarify the development of the law in the area. There are, however, ways of improving the situation. For example, guidelines could be used to clarify the underlying goal structure and provide more accessible and integrated guidance for decision makers. This has been done in other areas of EC competition law, but very little effort has been made to do this in the context of Article 82, at least since the early years of its development. As we shall see, the first serious attempt to create such guidelines is very recent, and it has been shaped by a different conception of the goals of Article 82. A second misconception claims that existing law under Article 82 does not seek to improve economic outcomes but is concerned with issues—such as fairness and European integration—which do not contribute to improving economic performance. This claim is then a basis for the conclusion that the law in the area should be changed, because competition law should pursue only economic goals. Given the importance attached in Europe to the goal of increasing European competitiveness, this argument takes on particular importance and calls for special attention.11 If it were true, it would be a serious criticism of Article 82’s goals. The claim turns out, however, to be inaccurate and potentially harmful. The abuse concept has always been propelled by the idea that reducing distortions of the competitive process will improve economic outcomes. It was developed under circumstances in which improved economic performance was the highest priority for all economic policy making. In the early years of the development of the abuse concept, European competitiveness, particularly in relation to the US, was seen as a major raison d’etre for the process of European integration,12 and this objective has been made even more explicit since the mid-1990s. The goal of reducing competitive distortion was specifically intended to improve the capacity of the competitive process to generate economic benefits, and the abuse concept has been a significant part of this economic policy thrust. Throughout its history, it has been developed by institutions whose central concern has been the improvement of economic 11 See, e.g., the Community’s ambitious “Lisbon Program”, which was launched in 2000 and which has sought dramatic improvements in competitiveness within the EU. 12 For discussion of the goals of European competition law in the context of the broader goals of European integration, see Gerber, supra note 1, at 334–371.

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 45

The Future of Article 82: Dissecting the Conflict

45

performance in Europe. As noted above, the goals of fairness and European integration have supported this goal. Third on our brief list of misconceptions is the much overused and vacuous claim that Article 82 is about “protecting competitors rather than consumers”. As other commentators have pointed out,13 this slogan has no analytical meaning. It may be well-suited for serving rhetorical purposes, but it sheds little, if any, light on the issues involved, and is more likely to confuse than to clarify the issues. The competitive distortion concept has been developed in order to protect the competitive process and thereby ultimately to benefit consumers. Discouraging distortions of the competitive process has been understood as a means of improving the effectiveness of markets, and consumers are the most direct beneficiaries of such improved effectiveness. In applying Article 82, the long-run goal of protecting consumers may also provide incidental and temporary benefits to particular market participants, but this does not alter the goal itself. Finally, there is a claim that usually takes the following form: “European integration and fairness are separate political goals that are unrelated to the economic issues of competition law and thus should be disregarded.” Here the claim relates not to the confusion of goals mentioned above, but to the specific goals themselves. It is an important claim, because if it were accurate that these two goals are merely political, then there would be much to the argument that political circumstances have changed and that they are therefore no longer relevant and should be eliminated. However, the claim is misleading. As noted above, both goals have been conceived and applied in the service of protecting the process of competition from distortion, a conception of the role of competition law that is anchored in the Treaty of Rome.14 Both derive their roles and their legitimacy from that central objective. When viewed in this context, they are therefore not political but economic goals. Again, the importance of analyzing the goals in the context of their overall structure is critical to understanding them. This brief review of some of the more important misconceptions in the descriptions of the current understanding of Article 82’s goals indicates just how badly deformed discussion of this issue has become. Eliminating these types of confusion and misconceptions from discussions of the current goal model should be of significant value in improving the quality of the scientific and policy debates in this area.

13 See, e.g., Edward T. Swaine, “‘Competition, Not Competitors,’ Nor Canards: Ways of Criticizing the Commission”, 23 University of Pennsylvania Journal of International Economic Law 597 (2002). 14 According to the current Article 3(1)(g) of the EC Treaty, one of the goals of the EC is to achieve “a system assuring that competition in the internal market is not distorted”.

(C) Ehlermann 08 Gerber

46

6/11/08

11:46

Page 46

Session One: The Sherman Act And Abuse Of Position Under Art 82

II. Economic Science and the Consumer Welfare Model The Commission has recently moved toward a different conception of the objectives of Article 82. In it, the central idea is that dominant firm conduct is only harmful when and to the extent that it reduces “consumer welfare”, as defined by mainstream economic science—i.e., neoclassical economics.15 According to this standard, the basic issue in applying Article 82 is whether the conduct of a dominant firm leads to an increase of price above a competitive price.16 If it does, it is designated as harm to consumers, and this makes the conduct “anticompetitive” for purposes of the law. This conception of the goals of Article 82 is scientific.17 It rests squarely on economic science and absorbs its terminology, and it depends on economic concepts to give meaning to the abuse provision. In it, “abuse” takes on a conceptually more precise and definable form. This scientific approach also means that the analysis must rest on quantitative methods. The degree of harm can be quantified, and it is subject to formal (i.e., mathematical) economic modeling techniques. An economist can analyze market data by reference to a market model that has been accepted in the economics literature or she can develop a model specifically for the situation at hand. Such models can be evaluated by other economists according to accepted standards within the economics profession. Ideally, therefore, the process of assessing conduct is rigorous, scientific and objective. At the conceptual level, therefore, it is a much simpler and clearer conception of the goals of Article 82. Application of this methodology has many implications for the analysis of unilateral conduct under Article 82.18 For example, it calls for (or at least privileges) the use of a shorter time frame in assessing harm to competition. Economic analysis of price effects is most effectively used at a specific point 15 In the context of competition law, the neoclassical model of economics remains the central and often the sole reference point. Newer developments in economic theory that deviate from that model have so far found little traction. For general discussion of many of the recent trends, see, e.g., Diane Coyle, The Soulful Science: What Economists Do and Why it Matters, Princeton University Press, 2007. 16 This description is, of course, an oversimplification. There are many controversies among economists about the exact measure to be used. For our purposes here, however, the point is that whatever the exact standard, it must be precisely articulated in economic terms, defensible according to the conventional standards of the economics profession, and quantifiable. For further discussion, see Massimo Motta, Competition Policy: Theory and Practice, Cambridge University Press, 2004, at 17–26. 17 For full treatment by a group of leading economists advising the Commission, see Report of the EAGCP (Economics Advisory Group on Competition Policy), “An Economic Approach to Article 82”, available at http://ec.europa.eu/comm/competition/publications/studies/eagcp_ july_21_05.pdf. 18 For discussion of the embeddedness of economics in the institutions of European competition policy, see David J. Gerber, “Competition Law and the Institutional Embeddedness of Economics”, in Josef Drexl, Laurence Idot and Joël Monéger, eds., The Crisis in Competition Law: A Comparative Perspective (approximate title), forthcoming 2008.

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 47

The Future of Article 82: Dissecting the Conflict

47

in time. The longer the time frame, the less confidence economists are likely to have in the assessment of specific price effects.19 Game-theoretic modifications can extend the time frame, but this reduces the clarity of the analysis and may undermine confidence in its predictive capacities. The focus therefore remains on a narrow analytical time frame. In promoting this conception of the goals of Article 82, the Commission is applying to Article 82 an analytical framework that it has recently begun to apply in other areas of competition law. It is sometimes called the “US antitrust model”, because it was developed in the United States and is most closely associated with the US antitrust system.20 In the late 1970s it began to take hold in the US courts and enforcement agencies, and by the end of the 1980s it had achieved orthodoxy. It was given impetus in Europe by experience in the UK, where economic analysis began to be pursued with much success around the turn of the new century. The Commission began moving toward this model in the late 1990s, and by 2004 it had established its “more economic approach” in many areas of competition law.21 Article 82 represents the one area of competition law where its success has been limited.

A. Misunderstanding the Consumer Welfare Model Unfortunately, there has also been significant misunderstanding of this conception of the goals of Article 82. One source of misunderstanding is the terminology that is often used to refer to it. The term “more economic approach” has become the code word for this conception of European competition law, and it is frequently used in referring to the Commission’s proposals under Article 82. Although in some areas of competition law, the use and meaning of the term have caused little difficulty, its application to Article 82 has proven more problematic. An underlying problem is that the term “more economic approach” is ambiguous and in this context potentially misleading. For some, it implies that the new approach involves nothing more than a change of emphasis within an 19 Measurement is not limited to short-term effects, but this is the central reference point. It is the point at which the economic modeling that is central to the economic approach can best be performed, and it is the point, therefore, where economists can have the most confidence in its predictions. For discussion, see, e.g., Lawrence A. Boland, The Methodology of Economic Model Building, Routledge, 1989. 20 For discussions of this “model” in comparative perspective, see David J. Gerber, “Competition Law”, in Peter Cane and Mark Tushnet, eds., Oxford Handbook of Legal Studies, Oxford University Press, 2004, pp. 510 et seq.; David J. Gerber, “Comparative Antitrust Law”, in Mathias Reimann and Reinhard Zimmermann, eds., Oxford Handbook of Comparative Law, Oxford University Press, 2007, pp. 1193 et seq. 21 For discussion of the more economic approach and its development, see, e.g., Lars-Hendrik Röller, “Economic Analysis and Competition Policy Enforcement in Europe”, in Peter A.G. van Bergeijk and Erik Kloosterhuis, Modeling European Mergers: Theory, Competition Policy and Case Studies, Edward Elgar, 2005, pp. 11 et seq.

(C) Ehlermann 08 Gerber

48

6/11/08

11:46

Page 48

Session One: The Sherman Act And Abuse Of Position Under Art 82

existing “approach”. In this interpretation, it merely represents more use of economics in ways that economics has previously been used, but not fundamental changes in the content of the law. For others, however, the term “more economic approach” refers to a new approach, one that is defined by new uses of economics. It thus involves much more significant changes in the law under Article 82. In its Discussion Paper on the reform of Article 82 (cited above), the Commission referred to both continuity and change, but it did not clarify the nature and extent of each. In exploring new directions, it recognized that it is legally bound to follow interpretations by the European Courts, but it sought to integrate existing case law interpretations with its proposed new directions. This led to uncertainty about what the Commission actually intended to achieve in its new program and the extent of the changes it sought make, especially insofar as such changes might deviate from existing bases of authority. This uncertainty and skepticism may have further nourished misunderstandings and encouraged both intellectual and political resistance to the changes.

B. The Analytical Key: Distinguishing Two Roles of Economics In order to reduce misunderstanding of the consumer welfare model, it is important, therefore, to analyze the precise roles that economic science is intended to play in the “more economic approach” under Article 82. Analysis of the debates over the future of Article 82 reveals that they frequently conflate two distinct roles for economics. Distinguishing these two roles provides a key to analyzing the debate and to understanding the confusion and misunderstandings that have attended it. One role uses economics to improve the capacity of enforcement agencies and courts to ascertain the facts—i.e., to understand what has happened or is likely to happen. More precisely, economics in this sense is used to identify correlations and potentially causal relationships between certain kinds of conduct of dominant firms and changes in market conditions that reduce “consumer welfare”. I call this the “fact-interpretive” role of economics. Competition law norms refer to the effects of particular conduct, and economic science can be used to assess such effects more precisely, more effectively and with greater methodological stability than is otherwise possible. The value of increasing the use of economics to improve factual analysis is well established. Important decisions of the European Courts that have triggered dissatisfaction with existing law refer primarily or exclusively to this role.22 It takes on particular importance in the context of unilateral conduct because the effects on competition of single-firm conduct are highly contex22 For further discussion of the important merger cases that are relevant in this regard, see David J. Gerber, “Courts as Experts in European Merger Law”, in Barry Hawk, ed., Fordham Corporate Law Institute 2003: International Antitrust Law and Policy, Juris Publishing, 2004, at 475 et seq.

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 49

The Future of Article 82: Dissecting the Conflict

49

tual. Whereas the harmful consequences to consumers from cartel conduct are well established and largely uncontroverted, the difficulties of distinguishing the pro-competitive and anticompetitive effects of unilateral conduct are much greater. This heightens the need for more extensive economic analysis of the factual contexts in which those effects must be assessed. There is, however, a second role for economics imbedded in the discussions. Here the role of economics is normative, which is to say that economic science provides the norms of conduct under Article 82. In much of the recent literature on the reform of Article 82, there is an implicit assumption that the issue of whether there has been a violation of Article 82 is to be measured by whether the relevant conduct reduces “consumer welfare” (as defined by neoclassical economics) or can reasonably be expected to lead to such a reduction. In this literature, it is simply assumed that economic science itself provides the relevant criterion for assessing the legality of dominant firm conduct. From this perspective, prior decisions by courts or administrative authorities have little direct relevance for determining the norms of conduct. Economics provides the norms and dictates the analysis of whether particular conduct violates the norms. This assumption raises significant policy issues. Economics is a “positive” science—i.e., it seeks to describe or explain events or conditions. Its role is to answer “what” questions—for example, “what happened in the past or what can be expected to happen in the future?” In this capacity, it cannot in and of itself answer “should” (or “normative”) questions. It can be assigned normative tasks, whereby, for example, it maps the concept of consumer welfare onto the concept “anticompetitive”; but this requires a decision authorized by the legal-political system. If such an assignment is made, it necessarily alters the framework within which economists work as well as the roles they perform, and it may introduce elements into economic analysis that would normally not be present when the analysis is performed within the community of economists for their descriptive-analytical purposes. Distinguishing these two roles helps to clarify and perhaps recast the issues. It becomes clear that a key issue should be whether the norm-setting organs of the EU—the Courts and the Commission—should assign specific normative roles to economics and, if they do, what those normative roles should be. Lack of transparency on this issue may underlie many of the concerns expressed about the reform process.

III. Identifying the Differences Distinguishing between these two roles of economics also allows us to see more clearly the similarities and differences between the current model of

(C) Ehlermann 08 Gerber

50

6/11/08

11:46

Page 50

Session One: The Sherman Act And Abuse Of Position Under Art 82

Article 82’s goals and the conception espoused by the Commission. If we look at one of these two roles, the fact-interpretive role, the new conception does not represent significant change. If we look at the other role, the normative role, there is a change, and it is fundamental to the entire normative enterprise of Article 82. It involves more than just a change of emphasis within the existing “approach”. The distinction between the two roles is therefore central to the entire discussion of the future of Article 82. Without it, the discussion is likely to remain unclear and confused. The “fact-interpretive” role of economics does not involve fundamental change. It merely calls for more and better use of the tools of economic analysis in interpreting facts. When we identify and isolate this role, there is little basis for resistance to change. Economics is used to provide more and better data for use in assessing the consequences of conduct and to improve the capacity of European competition law institutions to perform this function. In this role it can provide potentially very significant positive value for both the institutions themselves and for firms assessing the legality of their conduct. There are, of course, limits to how extensively the tools of economics can be employed. Their use imposes costs on enforcement institutions and courts. Economists are expensive, and unlike the situation in the US, much of the cost of increased use of economics is borne by those institutions.23 Moreover, the potential value of increased use of economics requires firms to employ economists for the assessment of their potential competition law liabilities, and this cost can again be substantial. These costs may therefore impose some constraints on the use of economics, but they are the price that has to be paid for better analysis of causality in economic relationships. Moreover, the issue is how the costs of using economics compare with the costs of the system without the use of economics, and it is by no means clear in that comparison that reasonable use of economics would significantly increase costs. The locus of differences between the competitive distortion conception of Article 82’s roles and the consumer welfare approach is the second or “normative” role being assigned to economics. Here the differences are fundamental. Whereas the competitive distortion model examines competition as a process and seeks to identify conduct that impairs or might impair that process, the consumer welfare model measures certain results produced by that process. It assesses conduct by reference to a specific and quantifiable measure of the effects of such conduct. It does not seek to evaluate conduct by reference to the process itself—either its structure or its dynamics. The 23 In US private antitrust litigation, in contrast, the costs of using economists are borne almost exclusively by the parties themselves, because each party has the responsibility for adducing its own evidentiary material. In continental European procedure, these responsibilities and their attendant costs fall primarily on the judges rather than the parties. For discussion of these procedural differences, see David J. Gerber, “Private Antitrust Enforcement in the U.S. and Europe: A Comparative Perspective”, in Thomas Möllers and Andreas Heinemann, eds., The Enforcement of Competition Law in Europe, Cambridge University Press, 2007.

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 51

The Future of Article 82: Dissecting the Conflict

51

focus of inquiry is thus fundamentally different, and would represent a fundamental change in the goals of Article 82. The dimensions and implications of this central difference are numerous. I note here some of the more important ones.

a) Dimensions of the legal inquiry In place of a goal structure that accommodates and relates several subsidiary or derivative goals and takes into account a variety of factors in assessing anticompetitive effects, the consumer welfare model calls for a single criterion or norm for applying the provision. If “anticompetitive” in the context of Article 82 is equated with “harm to consumer welfare” in the specific economics-based sense used here, there is no room for a decision maker explicitly to consider other issues. The conduct either does or does not meet the economic criterion applied. The inquiry ends there (economists may disagree, of course, about the application of the relevant economic theory to the conduct, but that is a different issue).

b) Authority issues The two models also use different sources as guides for decision-making and as authority for decisions reached. Application of the competitive distortion model relies primarily for these functions on legal analysis of prior cases and on a loose set of ideas about injury to the competitive process. Case decisions by courts and administrative authorities provide the main reference points for shaping decisions about the application of Article 82’s norms. Although guided in its early years by a framework of ideas propounded by a group of European legal and economic scholars,24 development of the abuse concept since at least the late 1970s has relied heavily on case-based experimentation. In this context, there is no clearly-articulated theoretical framework to which an analyst can refer to justify her decisions. In contrast, where the consumer welfare model is used as the normative basis for decisions, the decision maker does have a theoretical reference framework on which she can or, as the case may be, must rely. Legal sources (e.g., cases) that are inconsistent with that framework have little or no direct role in the decision-making process. The decision maker may refer to such cases for political reasons or for insights into factual scenarios, but if the consumer welfare model controls decision-making, she has no obligation to do so. Again, the only question is whether under appropriate application of economic science, there is harm to “consumer welfare,” as determined by economics. What prior cases have said is simply not relevant. This means that 24

For discussion, see Gerber, supra note 1, at 356–358.

(C) Ehlermann 08 Gerber

52

6/11/08

11:46

Page 52

Session One: The Sherman Act And Abuse Of Position Under Art 82

there is little, if any, room for combining the two perspectives in a consistent and meaningful way. One can imagine decision-making protocols intended to bridge the conceptual gap between these two approaches, but it would remain a fundamental inconsistency.

c) Expertise and interests Not surprisingly, the two conceptions of Article 82’s goals also implicate different forms of expertise and the interests that are associated with those possessing the relevant forms of expertise. The competitive distortion model relies for its application on legal expertise. The central issue is how the language and reasons that are accorded status in the legal system can be applied to new factual material, and here the capacity of legal professionals to grasp, apply and also manipulate language is the primary source of expertise. Where the consumer welfare model is used, legal language has significantly less importance. In applying this model, economists and their professional expertise play the central role. Economists alone will in most cases have the necessary tools to apply the model in any kind of rigorous way. They may be able to provide heuristics or other forms of guidance to legal decision makers that allow them in some cases to approximate outcomes that would be obtained through a more complete economic analysis. Administrative officials and judges might then perform some of the tasks of applying the consumer welfare model, and it may be possible to develop “rules” or “principles” for using the consumer welfare standard that can effectively guide them in taking some kinds of decisions without the aid of economists, but economists remain the ultimate source of authority.

d) Time frames Finally, the two models use different time frames. The competitive distortion model uses a temporal perspective that is necessarily developmental and may vary significantly, depending on the type of harm being evaluated. The characteristics of the particular type of restraint involved will have differing temporal horizons, and the analyst can apply whatever time frame seems appropriate for assessing those effects. The consumer welfare model, on the other hand, relies primarily on price-output calculations, often as expressed in mathematical models. These are, as noted above, typically framed within a much narrower time span, because confidence in the models diminishes rapidly where longer time periods are used. *

*

*

These comparisons highlight differences between the two models and thus reveal some of the implications of moving from the current conception of

(C) Ehlermann 08 Gerber

6/11/08

11:46

Page 53

The Future of Article 82: Dissecting the Conflict

53

Article 82’s goals to one in which an economic science model provides the content for Article 82. The two models shape analysis in fundamentally different ways, and they may lead to different outcomes. Identifying the differences and recognizing some of their implications clarifies the dimensions of choice about the future of Article 82. Resistance to the Commission’s vision of the future of Article 82 may stem, in part at least, from the perception that the process of reform of Article 82 has not adequately identified the nature of the changes and the legal and political basis for those changes, and thus clarifying those factors may help to prepare the way for more constructive discussion of these issues.

IV. Concluding Comments Misconceptions, misrepresentations and distortions have clouded debates and discussions regarding the future of Article 82. Such misconceptions relate to both the existing conception of Article 82’s goals and the conception that has been advanced to replace it. Misconceptions relating to the current conception of goals often arise from insufficient awareness of the objectives that have been pursued in developing the law in this area and inadequate appreciation of the relationships among the goals that are articulated in individual decisional contexts—i.e., inadequate awareness of the structure of goals. Misunderstandings relating to the proposed alternative often result from confusion of the roles that might be played by economists and the science of economics in this alternative conception. In particular, we have seen that reform discussions typically fail to distinguish between two distinct roles for economics, and thus drawing that distinction and focusing specific attention on the normative role of economic science in Article 82 and on its potential implications can contribute to a clearer understanding of those discussions. This makes visible the nature and degree of the changes being proposed, and highlights the points on which debate can usefully focus. These misconceptions and confusions have potentially harmful effects, because they increase the probability that important policy decisions will be based on an unnecessarily confused framing of the issues. Moreover, when these misconceptions and distortions enter the discussion, they tend to replicate themselves and to become durable components of thinking about the issues involved. This makes it all the more important to be wary of the potential for such distortions and to regularly revisit and re-examine the discussions in order to identify and minimize them. If we reduce or eliminate these two sets of misunderstandings, the conflict turns out to be narrower than sometimes thought, but also more fundamental.

(C) Ehlermann 08 Gerber

54

6/11/08

11:46

Page 54

Session One: The Sherman Act And Abuse Of Position Under Art 82

When the potential for such distortions and misconceptions is kept in mind and when the very distinct roles played by economics in a competition law system are kept in focus, we can begin to discern opportunities for fashioning a more firmly grounded pathway for developing the law relating to unilateral conduct. In this vision, the experience gained from applying the competitive distortion model could serve as the presumptive base for further development of Article 82, but normative uses of economics could be continually tested in individual contexts and case decisions. Where norms based on economics generate outcomes that are more convincing for legal decision makers than the competitive distortion model, those decision makers could either apply the economics-based norm, where they have authority to do so, or seek authority to do so where they do not have it. This opens a way of moving forward over time, but this scenario is very different from moving immediately to an economic science model as the controlling source of norms for all decision-making under Article 82. When John Vickers refers to “economicsbased rules”,25 he appears to envision a pathway at least somewhat similar to this, and, in my view, it holds great promise.

25 See, e.g., John Vickers, “Abuse of Market Power”, 115 Economic Journal 244 (2005). I am grateful to Professor Vickers for sharing his insights into these issues with me.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 55

II Christian Ahlborn and A. Jorge Padilla1 From Fairness to Welfare: Implications for the Assessment of Unilateral Conduct under EC Competition Law September 2007

I. Introduction U.S. antitrust policy had transitioned from the pursuit of fairness to a consumer welfare standard by the time Judge Robert H. Bork published his Antitrust Paradox in the 1970s.2 This transition is taking considerably more time on this side of the Atlantic, but judging from Commissioner Kroes’ 2005 Fordham speech and DG Competition’s Discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (the “Discussion Paper”), it is well under way.3 This paper explains why the move from fairness to welfare in European antitrust should be welcomed on normative grounds. Following a long philosophical tradition,4 we take the view that legal rules and institutions should be assessed taking exclusively into consideration their impact on the well-being of individuals. Thus, a legal rule will be regarded as superior to another if the aggregate impact of the former on the welfare of individuals exceeds the aggregate welfare impact of the latter. Under this normative approach, therefore, a fairness-based antitrust rule will necessarily be pronounced inferior to a welfare-driven rule, unless both coincide. Furthermore, it will also be considered relatively less efficient as a matter of economics. This

1 Christian Ahlborn is a Partner at Linklaters. A. Jorge Padilla is Managing Director at LECG Consulting and Research Fellow at CEMFI and CEPR. This paper was presented at the 12th Annual Competition Law and Policy Workshop held in Florence on 8–9 June 2007. It draws in part on work with David S. Evans. We wish to thank Mel Marquis for his detailed comments and suggestions. The usual caveats apply. Comments should be sent to [email protected] and [email protected]. 2 Robert H. Bork, The Antitrust Paradox, Free Press, 1978. 3 Commissioner Neelie Kroes, “Preliminary Thoughts on Policy Review of Article 82”, speech delivered at the Fordham University School of Law, 23 September 2005, available at http://europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/05/537&format=HTML &aged=0&language=EN&guiLanguage=en. DG Comp’s Discussion Paper, dated December 2005, is available at http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf. 4 See the discussion of “utilitarianism” in Will Kymlicka, Contemporary Political Philosophy, 2nd ed., Oxford University Press, 2002, chapter 2.

(D) Ehlermann 08 Ahlborn & Padilla

56

6/11/08

11:48

Page 56

Session One: The Sherman Act and Abuse of Position Under Art 82

is because, for economists, a legal rule is “efficient” when there is no other legal rule with a greater positive effect on aggregate welfare. As explained by Kaplow and Shavell (2002), “invocations of efficiency should thus be understood to entail a concern for individuals’ well-being rather than obeisance to some technical or accounting notion”.5 As explained in Section II below, current Article 82 policy is characterized by a mixed normative approach where the notions of welfare and fairness are each given positive weight. Not surprisingly, European competition policy exhibits many different goals. Welfare is one of them. But others, such as the protection of small competitors, can only be understood in terms of fairness. This hybrid approach is the result of many factors, but certainly among these the influence of the Ordoliberal school of legal and economic thought features very prominently. The impact of ordoliberalism on EC competition law is explained in detail in Sections III and IV. Ordoliberal concepts, such as the notion of “complete competition”, which gave rise to the key (traditional) principles of EC competition law, have lost their intellectual appeal. However, the principles that were derived from them, such as the “special responsibility of dominant firms”, are still part of the backbone of European antitrust. Our analysis in Section V shows that current Article 82 policy is inefficient, as it fails to deliver from a welfare perspective. Current Article 82 policy is likely to cause too many and too costly false positives (or “type I” errors, in the jargon of statistical decision theory). We show that these errors are particularly costly due to their nefarious impact on the incentives to innovate and thus on economic growth and prosperity. These errors are the result of a combination of factors: (i) the adoption of a vacuous standard based on ill-defined notions such as “competition on the merits”, “normal competition” and “objective justification”; (ii) the refusal to accept the consensus view among economists that the behaviour of the competitive fringe generally constitutes a valid benchmark by which to assess the pro-competitive or anticompetitive nature of the unilateral conduct of dominant firms; (iii) the tendency to neglect the beneficial effects of certain unilateral business practices on short-term consumer welfare, while its potentially adverse effects on rivals and, it is argued, on future consumer welfare, is over-emphasized; and (iv) the de facto or de iure adoption of per se rules that ignore the fact that most unilateral actions have both pro-competitive and anticompetitive effects. As noted above, EC competition policy is transitioning to a consumer welfare standard. As Commissioner Kroes recently stated, “the objective of Article 82 is the protection of competition on the market as a means of enhancing consumer welfare and ensuring an efficient allocation of resources”.6 A similar 5 Louis Kaplow and Steven Shavell, Fairness Versus Welfare, Harvard University Press, 2002, p. 37. 6 Kroes, supra note 3.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 57

From Fairness to Welfare 57 statement can be found in DG Comp’s Discussion Paper.7 This transition is welcome, and it will contribute to improving the efficiency of Article 82 policy. But as we argue in Section VI, it is not enough. Adopting a consumer welfare standard is nominally a positive first step, but the implementation of a welfare standard would require abandoning the special responsibility of dominant firms, recognizing the ubiquitous role of efficiencies, and taking into consideration the actual effects of a practice when assessing the likelihood of its anticompetitive effects. Our reading of the Discussion Paper and the recent judgments in British Airways 8 and Wanadoo 9 suggests that the Community institutions may not be ready, at least not for the time being, for such a change. While there is considerable agreement among economic and legal scholars and practitioners that the goal of competition policy should be the protection of consumer welfare, there is no consensus, or a very limited one, on how to achieve that goal. Indeed, there is an intensive debate concerning the design of legal rules and tests to identify anticompetitive practices. In Sections VII and VIII we present our views on these issues. We first consider the proposal of DG Comp’s Economic Advisory Group on Competition Policy (EAGCP),10 which mirrors to a large extent the views of Professor Steven Salop in the United States.11 In a nutshell, the EAGCP proposes to assess directly the impact of a given practice on consumer welfare through an unstructured rule of reason standard, whereby the anticompetitive and procompetitive effects of the practice are balanced according to the facts of the case. They also propose to eliminate the dominance filter. In our view, this proposal may lead to too many false positives and may be costly to enforce and self-enforce. That is why we discuss two alternative approaches that build on Judge Easterbrook’s structured rule of reason:12 a three-pronged structured rule of reason and a qualified per se legality rule (according to which a practice is considered legal in all but “exceptional circumstances”). We argue that the structured rule of reason is likely to be superior to the unstructured rule of reason in error cost terms, since it limits the complex balancing of procompetitive and anticompetitive effects to those practices that are likely to have anticompetitive effects. A qualified per se legality rule is cheaper to enforce and makes it easier for dominant firms to assess their own practices. 7

See supra note 3, para. 54. Case T-219/99, British Airways v Commission [2003] ECR II-5917, upheld on appeal, Case C-95/04 P, [2007] ECR I-2331. 9 Commission Decision COMP/38.233, Wanadoo Interactive, available at http://ec.europa. eu/comm./competition/antitrust/cases/decisions/38233/en.pdf, upheld on appeal: Case T-340/03, France Télécom v Commission [2007] ECR II-107; on further appeal: Case C-202/07, not yet decided as of this writing. 10 EAGCP, An Economic Approach to Article 82, July 2005, available at: http://ec.europa.eu/comm/competition/publications/studies/eagcp_july_21_05.pdf. 11 Steven C. Salop, “Exclusionary Conduct, Effect on Consumers, and the Flawed ProfitSacrifice Standard”, 73 Antitrust Law Journal 311 (2006). 12 Frank H. Easterbrook, “The Limits of Antitrust”, 42 Texas Law Review 1 (1984). 8

(D) Ehlermann 08 Ahlborn & Padilla

58

6/11/08

11:48

Page 58

Session One: The Sherman Act and Abuse of Position Under Art 82

Whether it is superior to the rule of reason and its variants depends on the cost of the type I and type II errors inherently associated with the assessment of unilateral business practices. Section IX concludes.

II. EC competition law objectives The precise objectives of Article 82 have never been articulated in any formal Community document or decision. However, the Commission has made it clear that it is inconceivable to apply EC competition policy without reference to the priorities fixed by the Community and the Community Courts have consistently insisted on the broad unity of purpose between the competition provisions of the Treaty. Consequently, the basic role of Article 82 is to promote the wider objectives of the EC Treaty by restricting or conditioning the strategies of companies with substantial market power.

A. Broad range of objectives Among the wider objectives of the EC Treaty, two in particular stand out in the context of EC competition policy: single market integration and, inevitably, the protection of competition, which, in the EU, has taken various different forms.

1. The objective of single market integration EC competition policy has to be understood in the context of the central task (and even raison d’être) of the Community, namely the creation of a single European market. As Gerber (1998) puts it: “This ‘unification imperative’ has shaped institutional structures and competences within the system, supplied much of its legitimacy and generated the conceptual framework for the development and application of its substantive norms.”13 Concern about market integration explains the high degree of attention given to vertical restraints affecting trade across borders in EC competition law—clearly disproportionate for those trained in the Chicago tradition.14 It also explains the Commission’s policy towards any unilateral action 13 David Gerber, Law and Competition in Twentieth Century Europe: Protecting Prometheus, Oxford University Press, 1998. 14 Aaron Director is viewed by most observers as the intellectual father of the Chicago School approach to antitrust. See Richard A. Posner, “The Chicago School of Antitrust Analysis”, 127 University of Pennsylvania Law Review 925 (1979).

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 59

From Fairness to Welfare 59 (e.g., excessive pricing or refusals to deal) or agreement which can affect the economic viability and scope of parallel trade within the European Union.

2. The objective of protecting competition One of the central objectives of the EC Treaty is to ensure a system of free competition. Article 4 of the EC Treaty states that the activities of the Community and its Member States should be “conducted in accordance with the principle of an open market economy with free competition”. Article 98 also requires the Community and its Member States to “act in accordance with the principle of an open economy with free competition, favouring an efficient allocation of resources”. Competition policy in Europe, and in particular Article 82, must serve the goal of protecting competition. Yet, as noted by Gerber (1998), there are at least four basic conceptions of what it means to protect competition in EC competition law: (a) protecting efficiency and welfare, (b) protecting economic freedom, (c) promoting economic change, and (d) constraining economic power. a. Protecting rivalry and the competitive process The constraints imposed on the actions of large corporations have most often been directed at protecting and enhancing rivalry, which is regarded as the central characteristic of the competitive process. Some authors have incorrectly stated that EC competition law seeks to protect competitors rather competition. That is certainly not the case. European competition authorities do not attribute more weight to the profits of the dominant firm’s competitors than to the profits of the dominant firm itself. They do not seem to value either too much. However, the Commission and the Community Courts firmly believe that there is no effective competition where there are no competitors. Decisions and judgments which seem geared to protect competitors are actually intended to protect the competitive process and, sometimes, to consolidate the liberalization efforts of the Commission and the Member States. Not surprisingly, competition policy intervention and especially intervention under Article 82 in the last few years has concentrated on recently liberalized sectors (especially telecoms) as well as on sectors characterized by network effects (where the risk of self-perpetuating monopoly is high. The protection of rivalry and the competitive process also explains the position of the Commission and the Community Courts in connection with refusals to deal. EC case law has been and still is more inclined to mandating access than US antitrust law. For example, terminating an existing commercial relationship without objective justification constitutes an abuse of a dominant position, even when the product to which access is granted is not indispensable to compete.

(D) Ehlermann 08 Ahlborn & Padilla

60

6/11/08

11:48

Page 60

Session One: The Sherman Act and Abuse of Position Under Art 82

b. The protection of economic freedom and the limitation of market power Competition has also been regarded as an instrument to prevent the creation and misuse of private economic power. The protection of competition is, according to the EC tradition, necessary to ensure economic freedom and, most importantly, to create a prosperous, free and equitable society. Article 82 has been considered a tool to curb the economic and political power of large corporations within the EU. Companies which for different reasons, both legitimate15 and illegitimate,16 enjoy significant prerogatives in their Member States, may have the incentive and ability to “undermine competitive (and sometime political or social) structures”.17 Somewhat paradoxically, EC competition law is meant to protect economic freedom by constraining, in a non-discretionary fashion, the decisionmaking freedom of dominant firms. However, intervention is justified as follows. First, the restrictions imposed on dominant firms seek to protect and promote the “common good”. Second, dominant firms are encouraged to compete vigorously provided that they compete “on the merits”; they are only prohibited from “recourse to methods different from those governing normal competition”.18 Third, the application of Article 82 is subject to the general principles of Community law: proportionality, equality, legal certainty and the rights of defence. The protection of human rights is guaranteed within the context of EC competition law. c. Fairness In addition, as stated by Gerber (1998: 2), “[t]he goal of protecting competition has often been interwoven with the idea of achieving social justice”. The Commission and the Community Courts have indeed condemned prices and contractual practices found to be “unfair”. Fairness considerations are also behind the idea that large, well-resourced firms should not unduly hamper the commercial activities of small or medium-size firms. Companies with market power must act “as if” they did not possess that power. Their actions must not deviate too far from the actions of companies that operate in competitive markets. Otherwise, their actions could be regarded as unfair and abusive even if they could prove to be welfare-enhancing.

15 They are too big to fail, as it were, due to the massive employment and/or financial implications for the local economy if the company downsizes its operations. 16 That is, plain corruption or more benign forms of regulatory capture. 17 Gerber, supra note 13, p. xi. 18 Case 85/76, Hoffmann-La Roche & Co AG v Commission [1979] ECR 461, para. 6.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 61

From Fairness to Welfare 61 d. Welfare and efficiency For most economists the main, if not the only, goal of competition law should be to prevent any unilateral or collective practice that could harm consumers and reduce the efficiency of markets. Protecting competition in this sense means protecting and promoting efficiency. Economists have long debated whether the goal of competition law should be the protection of consumer welfare or aggregate welfare (i.e., the sum of consumer welfare and industry profits). They have disagreed on whether competition law should focus on short-term consumer welfare or whether instead it should also take into account the impact of certain commercial practices on future generations of consumers. And they have also disputed whether competition policy should concentrate on allocative efficiency (i.e., on prices, quality and output) or whether dynamic efficiency (i.e., innovation) should be considered the main goal of an economically sound antitrust law. But there is a generalized consensus amongst economists, at least those who regard themselves as orthodox or mainstream, that companies’ actions should be deemed illegitimate only when their impact on welfare is adverse. EC competition law, in general, and Article 82, in particular, also seeks to protect and enhance consumer welfare and market efficiency.

3. Types of goals and consistency In a debate about the goals that should guide competition policy, it is important to distinguish between direct goals and indirect goals. Almost everyone would agree that the ultimate goal of competition policy is to increase economic welfare: the question, however, is whether it should be a direct goal, i.e. a goal which ultimately affects and guides competition policy. A comparison may help to make this distinction clear: for ordoliberals, the direct goal of competition policy was the protection of economic freedom and the preservation of “complete” competition. The underlying assumption was that the preservation of economic freedom would ultimately lead to the maximization of economic welfare, but it was an assumption which was not tested. This contrasts, for example, with the Chicago school approach, which focuses on economic welfare as the only direct goal. Here, economic welfare is not an untested assumption but the benchmark against which competition policy is measured. In the remainder of this paper, we will classify the various competition policy goals according to categories: • Fairness goals: which include fairness, the protection of economic freedom, the protection of rivalry and the competitive process and the protection of small and medium-size firms. • Welfare and efficiency goals: which cover both of the principally discussed welfare objectives, i.e., the goal of consumer welfare and that of total welfare.

(D) Ehlermann 08 Ahlborn & Padilla

62

6/11/08

11:48

Page 62

Session One: The Sherman Act and Abuse of Position Under Art 82

• Market integration goals: which deal with the goal of a single European market and the reduction of obstacles to cross-border trade. While the goals within each category may lead to somewhat differing policies, they are broadly consistent: for example, the protection of economic freedom will go hand in hand with the control of dominant firms and the protection of small and medium-size firms, and fairness considerations will quite naturally come into play. Again, the consumer welfare and total welfare goals may lead to somewhat different rules, but these differences are unlikely to be significant, at least in practice. By contrast, across the three categories, it is less likely that the goals encompassed by each will be consistent.

B. The dynamics of EC competition law goals Competition law objectives change over time. Sometimes this shift is fairly obvious, sometimes it is slow and gradual. In the US, exponents of the Chicago school, such as Robert Bork or Richard Posner, triggered a debate about the goals of antitrust in the 1960s and 1970s, at a time when the US authorities pursued a range of different goals not unlike the current fairness goals in EC competition policy, including the preservation of rivalry and the protection of small firms. The Chicago School advocated a single objective of economic welfare on the basis that a policy which followed a number of inconsistent objectives could not be predictable. In the US, the Chicago school has clearly won this debate and the US enforcement authorities look most frequently at what is best for consumers. The Chicago school victory was described recently by Ken Heyer as follows: “Over the past several decades, there has emerged a rough consensus among professional antitrust practitioners . . . that the “competition” referred to in [the US] antitrust statutes is not to be interpreted simply as pre-merger rivalry among entities. Rather it is best viewed as a process, the outcome of which is welfare, with welfare—not rivalry—being [the] object of interest.”19

In the EU, the development has been markedly different. There has been no debate about the goals of competition policy comparable to that in the US, and the various objectives under EC competition policy have been, and continue to be, co-existent. Arguably, however, there has been a shift of emphasis: fairness goals and integration goals seem to have lost some of their previous importance and economic welfare seems to have gained greater prominence. 19 Ken Heyer, “Welfare Standards and Merger Analysis: Why Not the Best?”, Economic Analysis Group Discussion Paper, EAG 06-8, March 2006, available at: http://www.abanet.org/ antitrust/at-source/pdf/references/heyer-ken-06-8.pdf. A version of this paper is available in 2(2) Competition Policy International (2006).

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 63

From Fairness to Welfare 63

1. Market integration and economic welfare As a result of the Community objective of market integration, competition law has consistently intervened where private measures were being used as barriers to cross-border trade. The Commission was clearly concerned that state barriers would be replaced by business measures to partition national markets. This has led to an almost per se prohibition on absolute territorial protection, independently of any efficiency considerations. In other words, the market integration objective trumped the efficiency consideration. The Commission generally avoided addressing any conflict on the basis of the assumption that competition and market integration both serve the ends of consumer welfare since the creation and preservation of an open single market promotes an efficient allocation of resources. More recently, however, the Court of First Instance seems to have reversed the order of the market integration and welfare objective, removing the per se illegality of territorial protections.20

2. Fairness goals and economic welfare The shift from fairness goals to economic welfare has been more gradual and has generally occurred at times when a particular area of competition law was “modernized” (as happened, for example, with the assessment of vertical restraints between 1999 and 2001). Nevertheless, many cases even today still show a strong emphasis on fairness considerations and equality (“equal opportunity competition”). This is particularly true for Article 82. In parallel with a greater reliance on economic welfare objectives in “modernized” areas of competition law, there has also been a noticeable emphasis on economic welfare objectives in statements by Commission officials, whereas previous statements had often highlighted fairness objectives. For example, in the 1980s the Commission considered that “[e]ffective competition preserves the freedom and right of initiative of the individual economic operators and fosters the spirit of enterprise”.21 Where reference was made to the consumer welfare objective, as for example in the Commission’s Guidelines on Vertical Restraints,22 care was taken to emphasize that fairness goals and market integration goals were generally consistent with the consumer welfare objective. 20 See Case T-168/01, GlaxoSmithKline Services Unlimited v Commission [2006] ECR II-2969, on appeal: Cases C-501/06, C-513/06, C-515/06 and C-519/06, not yet decided. 21 Commission, Fifteenth Annual Report on Competition Policy (1986), p. 11. 22 Commission Notice, Guidelines on Vertical Restraints, 2000 OJ C291/1, para. 7 (“The protection of competition is the primary objective of EC competition policy, as this enhances consumer welfare and creates an efficient allocation of resources.”) (emphasis added).

(D) Ehlermann 08 Ahlborn & Padilla

64

6/11/08

11:48

Page 64

Session One: The Sherman Act and Abuse of Position Under Art 82

It is only recently that Commission officials have unequivocally been backing a single, or at least an overriding consumer welfare objective for EC competition policy. As Commissioner Neelie Kroes has stated: “Consumer welfare is now well established as the standard the Commission applies when assessing mergers and infringements of the Treaty rules on cartels and monopolies. Our aim is simple: to protect competition as a means of enhancing consumer welfare and ensuring an efficient allocation of resources.”23

III. The origins of fairness in EC competition policy: ordoliberalism The “fairness goals” were imported into EC competition law (via German competition law and German officials) from ordoliberalism. This is not to say that, without ordoliberalism, EC competition law would not have formulated “fairness goals” (indeed, as we have seen, US antitrust in the 1950s and 1960s were driven by very similar policy considerations without having been exposed to ordoliberal ideas). But without an understanding of ordoliberalism, it is hard to grasp how “fairness goals” shaped EC competition policy.

A. Competition as an unstable equilibrium Ordoliberalism, which was developed in Germany in the 1930s and 1940s, saw itself as the “third way” between the centrally planned economy and the unregulated market by laissez-faire liberalism. From the experience of the Weimar Republic, the Great Depression, and the ascendance of totalitarianism, ordoliberals drew two fundamental conclusions. First, the centrally planned economy and the market economy (or “transaction economy”) were fundamentally incompatible. Second, ordoliberals regarded the competitive process of the market economy as inherently fragile, threatened both internally and externally.

1. Direct instability Internally, the ordoliberals saw the competitive process endangered by an inherent, self-destructive aspect, which Walter Eucken described as follows:

23 Neelie Kroes, “European Competition Policy—Delivering Better Markets and Better Choices”, SPEECH 05/512, London, 15 September 2005.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 65

From Fairness to Welfare 65 “The supplier and the customer always—wherever possible—seek to avoid competition and to acquire or assert monopolistic positions. There is an omnipresent, strong and irrepressible urge to eliminate competition and to acquire a monopolistic position. Everyone espies possibilities of becoming a monopolist. Why should three bakers in a 13th century town compete with one another? They could simply come to an agreement and create a monopoly. This was the situation then and the same applies today.”24

In particular, according to Eucken, once a market becomes concentrated (i.e. when it moves away from a polyopoly), “this oligopoly—or part oligopoly—situation often passes by rapidly, and soon leads to the creation of a cartel, i.e., to a collective monopoly or an individual monopoly, by overpowering the opponent”, although sometimes “this unstable condition of the oligopoly or part oligopoly exists for many years or decades”.25

2. Indirect instability Of equal if not greater concern is what may be termed “indirect instability”, namely that the creation of economic power enables firms to influence decision makers, which in turn allows them to restrict competition through political intervention (for example through the creation of barriers to entry). Competition, to Eucken and other ordoliberals, thus has not only an economic but also a very important political dimension: “Competition is by no means only an incentive mechanism but, first of all an instrument for the deprivation of power (Entmachtungsinstrument). . . . The most magnificent and most ingenious instrument of deprivation of power in history.”

B. The ordoliberal answer For the ordoliberals, the answer to the problem of the inherent instability of the competitive process was twofold: the competitive order of “complete competition” as the guiding principle; and strong protective measures in accordance with what ordoliberals call “Ordnungspolitik”.

1. The competitive order of “complete competition” For ordoliberals it was clear that the market economy was the key to a prosperous and humane society. Competition and only competition would

24 Walter Eucken, “The Competitive Order and its Implementation”, 1949, reprinted in 2(2) Competition Policy International 222 (2006). 25 Eucken, cited previous footnote, at p. 244.

(D) Ehlermann 08 Ahlborn & Padilla

66

6/11/08

11:48

Page 66

Session One: The Sherman Act and Abuse of Position Under Art 82

achieve sustained economic development. However, according to ordoliberalism, not every form of competition could be expected to produce this beneficial outcome, but only the form of “complete competition”. Complete competition is competition in a market where no firm has the power to coerce conduct of other firms. Complete competition can be identified by two indicators. First, “[t]he price is not forced upon the market by way of a market strategy but is taken from the market”,26 in other words, all market participants are price takers. The second indicator is the absence of “certain measures [which would] clearly indicate that competition does not exist because these measures cannot be implemented under complete competition: for example, obstructions to purchasers or suppliers that have dealings with competitors, or loyalty rebates or predatory pricing or dumping or destruction of stocks”.27 In other words, complete competition is the market form which ensures “performance competition” (Leistungswettbewerb). Despite the fact that the ordoliberals do not link complete competition to a particular market structure (and despite the ordoliberals’ criticism of the neoclassical model of perfect competition), the concept of complete competition does have an underlying structural assumption of a polyopoly and can best be understood as the adaptation of the model of perfect competition to the real world. In summary, ordoliberals see a virtuous circle perpetuated by, on the one hand, a de-concentrated market structure where players have no (significant) market power, and competition on the merits on the other. De-concentrated markets prevent “hindrance competition” and ensure that competition takes place on the merits, and competition on the merits in turn preserves deconcentrated markets.

2. Protective measures of “Ordnungspolitik” Ordoliberalism saw the need for strong protection through a political and legal framework which would safeguard the efficient functioning of competition and which would protect it from any self-destructive tendencies. Here, ordoliberals foresaw a clear separation of roles for the state and the private sector: “The policy of competitive order does not leave the choice of the market forms and monetary systems to the economy itself because the experience of the laissez-faire policy speaks for itself. The development of the framework in which business and households can plan and act freely is governed by the economic policy under which the framework is supervised. Businesses are free to choose what they produce, what technology they use, what raw materials they purchase and what markets they wish to sell on [. . .] Freedom of the consumer exists, but not the freedom to choose how

26 27

Eucken, supra note 24, at p. 230 Ibid.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 67

From Fairness to Welfare 67 to define the rules of the game or the forms which the economic process takes. This particularly falls within the field of Ordnungspolitik (order-based policy).”28

For ordoliberals, every measure of economic policy was ultimately competition policy, in the sense that it was intended to safeguard and enhance complete competition. It included the primacy of monetary policy (against the experience of hyper-inflation in the Weimar Republic), the protection of open markets against state measures and private measures, consistency and predictability of economic policy, private ownership and competition policy (in the strict sense).

C. Key principles of ordoliberal competition policy Ordoliberal competition policy is characterized by the following four fundamental principles:

1. Complete competition as the predominant market form According to the ordoliberals, competition policy, and in particular the control of monopolies, would ultimately fail, not least for political reasons, if large parts of the industry were highly concentrated. By contrast, the situation would be different where the legal and political framework of Ordnungspolitik operate to ensure de-concentrated markets and complete competition: “The creation of companies with monopolistic power is prevented. Not only by prohibitions of cartels but also—and far more importantly—by an economic and legal policy which breaks through the strong forces of competition, as exists in a modern economy . . .”29 Thus, under ordoliberal economic policy, complete competition would be the predominant form of competition: monopolies and oligopolies would be exceptions.

2. Regulation of monopolies In an ordoliberal system, competition policy would be in the hands of an independent competition authority (monopoly office) whose task would be to 28 Eucken, supra note 24, at p. 227. See further Eli F. Hekscher, Der Merkantilismus, volume 1, Gustav Fischer, 1932, pp. 448 et seq. With respect to the economic liberalism in England in the early 19th Century, as opposed to mercantilism, Hekscher writes: “The old method would have been an attempt to create a barrier to fundamental changes (Umwälzungen). The new victorious method allowed them to take a free course. Therefore, they were enforced with power unparalleled in mankind’s ancient economic history. The third alternative would have been neither to intervene in the course of events nor to regulate its course but to direct it in ordered ways. This concept has never been tried.” 29 Eucken, supra note 24, at p. 239.

(D) Ehlermann 08 Ahlborn & Padilla

68

6/11/08

11:48

Page 68

Session One: The Sherman Act and Abuse of Position Under Art 82

break up “avoidable monopolies” and to regulate “unavoidable monopolies”. The basic principle for the regulation of monopolies was the principle of “behaviour analogous to competition”, reflected in the “as if ” approach: “The aim of monopoly legislation and monopoly supervision is to ensure that the bearers of economic power behave as if complete competition prevailed. The behaviour of the monopolists should be ‘analogous to competition’.”30 Eucken described the practicalities of this approach as follows: “Every form of impediment competition by embargoes, loyalty rebates, predatory pricing etc is prohibited [. . .] This creates a condition which would automatically arise in a situation of complete competition, where impediment competition would be pointless. Admittedly, in order to achieve a result analogous to competition, it is necessary to introduce an obligation to contract, as here coercion is necessary to achieve the same result as would automatically arise under complete competition. As is generally known, under complete competition the same prices will become established for the same goods and services. Supply monopolies for example, whilst striving for the highest profit, have a tendency to demand differentiated prices for the same goods or services from individual segments of demand. Price differentiation should be prohibited under the competitive order. What is most difficult is to implement the fundamental principle within the scope of determining price levels. The price is to be fixed in such a way that supply and demand are in equilibrium at this price and, at the same time, the marginal costs are just covered.”31

3. Competition policy towards oligopolies Ordoliberals regarded oligopolies as a transient market form: “This situation of oligopoly—or part oligopoly—often passes by rapidly, and soon leads to the creation of a cartel, i.e. to a collective monopoly or an individual monopoly, by overpowering the opponent”.32 However, given that sometimes “the unstable condition of the oligopoly or part oligopoly exists for many years or decades”,33 ordoliberals still saw the need to address the issue. There were two conflicting views among ordoliberals. Some of them proposed a special regulatory regime for oligopolies under state supervision. Others regarded this as too burdensome and, for the following reasons, unnecessary: “With a decisive monopoly supervision, the oligopolists have no reason to destroy each other by aggressive means or to attain a position of monopoly of their own. This is because [the last survivor] comes up against a rigorous monopoly control. Furthermore, the oligopolists themselves will attempt to behave as if complete

30 31 32 33

Eucken, supra note 24, at p. 241. Ibid., at p. 243. Ibid., at p. 244. Ibid.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 69

From Fairness to Welfare 69 competition prevailed, as they will otherwise come to the individual attention of the monopoly office.”34

4. Limited discretion of the competition authority According to ordoliberals, it is within the exclusive sphere of the state to define the “rules of the game”: “Freedom of the consumer exists, but not the freedom to choose how to define the rules of the game or the forms which the economic process takes. This particularly falls within the field of Ordnungspolitik (orderbased policy)”.35 Economic policy, and in particular competition policy in accordance with the Ordnungspolitik, is subject to guiding principles which tightly limit the discretion of the policy maker. There is first and foremost the principle of complete competition as the guiding star. However, Eucken pointed out that this is not sufficient for economic policy, and he highlighted the need for “symptoms” and rules of thumbs to make competition policy workable: “Economic policy needs indications, symptoms, by which to implement an economic policy of competitive order. It needs a rule of thumb. Does such a thing exist? The answer is yes. Two methods exist. The most direct method is to find out from companies themselves whether or not plans developed under competition. [. . .] [A second,] less direct, method would be as follows: certain measures exist from the outside, for example from the opposite end of the market, [which] clearly indicate that complete competition does not exist because these measures cannot be implemented under complete competition: for example, restraints to purchasers or suppliers that have dealing with competitors, or loyalty rebates or predatory pricing or dumping [or] destruction of stocks.”36

IV. Article 82 EC: the ordoliberal heritage Ordoliberalism had a huge impact on EC competition policy during the formative years (and still continues to influence German competition policy). As a result, many of the ordoliberal concepts have been hard-wired into the EC system, even when at times (or indeed frequently) the link to ordoliberalism has been obscured or forgotten. Nowhere can this be demonstrated more clearly than in the context of Article 82, where EC competition policy has remained virtually unchanged 34 35 36

Ibid., at p. 245. Ibid., at p. 227. Ibid., at p. 230.

(D) Ehlermann 08 Ahlborn & Padilla

70

6/11/08

11:48

Page 70

Session One: The Sherman Act and Abuse of Position Under Art 82

for the last 30-odd years: for example, the approach of the Commission in Michelin II (2001) is arguably indistinguishable from that in Michelin I (1981) and can be traced back to cases of the 1970s. In other words, much of today’s policy on the control of unilateral conduct by dominant firms operates within an intellectual framework shaped by ordoliberalism and pursues policy goals of fairness, such as the protection of rivalry and economic freedom, with an emphasis on SMEs. During the modernization debate in relation to Article 82, many of the ordoliberal concepts which are still today at the heart of EC dominance policy have proved to be in conflict with the increasingly prevailing consumer welfare approach and have frequently been the focus of the debate. Examples of this conflict are in particular: • • • •

The dominance threshold; The “special responsibility” of dominant firms; The formalistic approach towards many types of unilateral behaviour; and A wide range of “fairness abuses”.

A. The dominance screen According to the standard definition established by the Court of Justice in United Brands, dominance “relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately of its consumers”.37 This test was cited with approval in Hoffmann-La Roche, although the Court of Justice added the caveat that “such a position does not preclude some competition . . . but enables the undertaking . . . if not to determine, at least to have an appreciable influence on the conditions under which that competition will develop, and in any case to act largely in disregard of it so long as such conduct does not operate to its detriment”.38 In practice, the Commission has relied heavily on market shares as an important element in the assessment of dominance. This is by no means unusual. What distinguishes the EC approach from that of some of the other regimes, however, is the intervention threshold. Under EC competition law, and in particular in accordance with AKZO, the (rebuttable) presumption of dominance—save in “exceptional circumstances”—kicks in where market shares are in excess of 50 percent.39 This contrasts with the assessment of

37 Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207, para. 65. 38 Supra note 18, para. 39. See also United Brands, cited previous footnote, para. 113. 39 Case C-62/86, AKZO v Commission [1991] ECR I-3359, para 60.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 71

From Fairness to Welfare 71 market power under US law, where a market share of 70 to 75 percent for at least five years is required to lead to a presumption of monopoly power.40 The relatively low threshold and the absence, in the AKZO standard, of reliance on persistence of market shares, can be traced directly back to ordoliberalism with its ideal of complete competition (i.e., as explained above, a market where all participants are price takers) and its wariness of even short-term market power. From an ordoliberal perspective, a market share with 10 suppliers of 10 percent each seems to be inherently better (both from an economic and political perspective) than a market in which one player has 40 percent of the market and 9 players each have 6 or 7 percent. This view seems to be reflected in the British Airways case, where the Commission stated: “Despite the exclusionary commission schemes, competitors of BA have been able to gain market share from BA since the liberalisation of the United Kingdom air transport markets. This cannot indicate that these schemes have had no effect. It can only be assumed that competitors would have had more success in the absence of these abusive commission schemes.”41

B. The special responsibility of dominant firms Article 82 does not condemn the mere possession of a dominant position; it does not condemn the creation of a dominant position either. Under EC competition law, the statement that it is not an offence for a firm to have a dominant position comes invariably with the qualification that the law does impose on such firms “a special responsibility not to allow its conduct to impair undistorted competition on the common market”.42 The scope of this special responsibility is not entirely clear. In a narrow sense, it can be interpreted as saying no more than that Article 82 EC imposes obligations on dominant firms which are not imposed on non-dominant firms. In AKZO, the Commission made clear that a dominant firm is entitled to compete “on the merits”.43 The Commission added that aggressive, legitimate competition was to be encouraged and that it did not suggest that “large producers should be under an obligation to refrain from competing vigorously 40 Phillip E. Areeda and Herbert Hovenkamp, Fundamentals of Antitrust Law, Aspen Publishing, 2003. 41 Commission Decision 2000/75 of 14 July 1999, Virgin/British Airways [2000] OJ L30/1, para 107. The Commission’s prohibition decision withstood the subsequent appeals (see citations supra note 8). 42 Case 322/81, Nederlandsche Banden-Industrie Michelin v Commission [1983] ECR 3461, para. 57. 43 Commission Decision 85/609 of 14 December 1999, ECS/AKZO [1985] OJ L374/1, para. 81; upheld on appeal in AKZO, supra note 39.

(D) Ehlermann 08 Ahlborn & Padilla

72

6/11/08

11:48

Page 72

Session One: The Sherman Act and Abuse of Position Under Art 82

with smaller competitors or new entrants”.44 However, certain types of conduct that raise no issues when carried out by firms that are not dominant may be abusive under Article 82 when carried out by a dominant firm.45 This is because, it is argued, in the absence of market power such conduct is either irrational or unlikely to have a sufficient adverse effect on competition to justify intervention. A wider interpretation would suggest that a dominant firm must refrain from any action which would increase its market power and harm competitors, even where the behaviour is efficiency-based. The origin of the “special responsibility” doctrine can be traced back to the ordoliberal “as if” principle, according to which firms which are not price takers, i.e., which possess (significant) market power have not only have a negative obligation (i.e., not to commit certain harmful acts) but also a positive obligation (i.e., to behave as if they do not have any market power).

C. The notion of abuse—hindrance competition v. competition on the merits A dominant firm infringes Article 82 when it does not compete “on the merits”. That is, when the dominant firm, “through recourse to methods different from those which condition normal competition in products or services on the basis of the transactions of commercial operators, [adopts behaviour that] has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition”.46 An abuse is therefore a commercial practice pursued by a dominant firm which cannot be regarded as normal competition based on quality and price and which has the effect of restricting competition. Or, according to the formulation put forward in Michelin II by the CFI, an abuse may consist of exclusionary conduct that lacks “objective economic justification”.47 The assessment of abuse under EC competition law has equally been shaped by ordoliberalism. As described above, for Eucken and other ordoliberals, certain types of unilateral behaviour, such as price discrimination, loyalty rebates and tying, were inherently abusive. From an ordoliberal point of view, these would be clear examples of impediment competition with no redeeming features.

44

Ibid. See Joined Cases T-191/98, T-212/98 to T-214/98, Atlantic Container Lines AB and Others v Commission [2003] ECR II-3275, para. 1460. 46 Hoffmann-La Roche, supra note 38, at para. 91. 47 Case T-203/01, Manufacture française des pneumatiques Michelin v Commission [2003] ECR II-4071, paras. 107 and 110. 45

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 73

From Fairness to Welfare 73 The view that certain types of unilateral behaviour are per se harmful and therefore do not require any effects analysis has been reiterated by the Community Courts at regular intervals, most recently in Michelin II 48 and British Airways.49 In Michelin II, the CFI took the view that, in order to establish abuse, it was sufficient for the Commission to show that the conduct “tends to restrict competition”, i.e., that it “is capable of having, or likely to have, such an effect”.50 In British Airways, the CFI further stated that, “where an undertaking in a dominant position actually puts into operation a practice generating the effect of ousting its competitors, the fact that this hoped-for result is not achieved is not sufficient to prevent a finding of abuse”.51

D. “Fairness” abuses A number of types of abuse have a particular focus on fairness considerations, in particular: discriminatory abuses, exploitative abuses and abuses which protect existing commercial relationships. The latter two are discussed below.

1. Abusive exploitation The control of exploitative abuses reflects the ordoliberal principle of forcing dominant firms to behave “as if” they were subject to complete competition.52 Indeed, the wording of Article 82 (“Such an abuse may, in particular, consist in: . . . (a) directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions . . .”) echoes Eucken’s list of prohibited monopoly practices in his chapter on the “monopoly problem in the competitive order”:53 general terms and conditions in the contract to the disadvantage of the trading partner of the dominant firm,54 and prices which are in excess of the equilibrium price, i.e., prices in excess of marginal costs.55 In an attempt to make Eucken’s “as if” principle operational, methodologies for determining excessive prices were developed in the early years of 48

Ibid. British Airways, supra note 8. 50 Michelin II, supra note 47. 51 British Airways, supra note 8, para. 297. As indicated earlier (supra note 8), the judgment of the CFI was later upheld by the ECJ. 52 See Alexandre de Streel and Massimo Motta, “Excessive Pricing and Price Squeeze under EU Law”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 91 et seq.; David Evans and Jorge Padilla, “Excessive Prices: From Economic Theory to Administrable Rules”, 1 Journal of Competition Law and Economics 97 (2005). 53 Eucken, supra note 24, at pp. 240 et seq. 54 Ibid. 55 Ibid., at p. 241. 49

(D) Ehlermann 08 Ahlborn & Padilla

74

6/11/08

11:48

Page 74

Session One: The Sherman Act and Abuse of Position Under Art 82

German and EC competition policy, for example, in United Brands, which compared actual costs and prices and the prices of the dominant firm with that of its competitors.56 The practical problems of price control which Eucken recognized (but arguably nevertheless underestimated) were reflected in the fact that only a small number of exploitative abuses were pursued under German and EC law.57 However, EC and German law do contrast with the US policy under Section 2 of the Sherman Act: in the US, firms which have lawfully acquired monopoly power are entitled to exploit it; the concept of an exploitative abuse does not exist.

2. Harm to existing commercial relationships The Commission has treated refusal to supply an existing customer more severely than the refusal to supply a company with whom the supplier has no relationship. This is reflected in DG Comp’s Discussion Paper, according to which the termination of an existing supply relationship by a dominant firm is presumed to be abusive whenever it has a negative effect on competition (subject only to an objective justification), whereas refusal to supply in the absence of a pre-existing relationship is subject to the stringent Bronner (or IMS Health) test, which requires that the product or service to which access is demanded be “indispensable” to compete.58 The Commission seeks to justify this differential treatment within the context of its consumer welfare approach by arguing that the existence of an ongoing relationship allows the presumption that such a relationship is efficient, and points out further that the customer of the dominant firm may have invested in such a relationship. Neither argument is particularly convincing: in any sector, commercial relationships are constantly re-created and terminated, reflecting the changing market conditions and changing circumstances of the parties. Termination of an existing relationship therefore does not justify a (rebuttable) presumption of anticompetitive behaviour. As far as investments of the customer which are based on the ongoing relationship are concerned, such investments must be relationship-specific for customers to be locked. More to the point, it is also not clear why the customer in such circumstances would not have been able to obtain contractual protection and

56 Robert O’Donoghue and A. Jorge Padilla, The Law and Economics of Article 82 EC, Hart Publishing, 2006, chapter 11. 57 The most prominent cases under EC law are: Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207; Case 26/75, General Motors [1976] ECR 1367; Case 226/84, British Leyland [1986] ECR 3263. See also Commission Decision of 23 July 2004, Case COMP/36.568, Scandlines Sverige v Port of Helsingborg, available at: http://ec.europa.eu/comm/competition/antitrust/cases/decisions/ 36568/ reject_en.pdf. 58 See Discussion Paper, supra note 3, at paras. 228 et seq.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 75

From Fairness to Welfare 75 why competition policy should offer the protection that the customer itself could have procured for itself through, for example, contractual means. A more plausible explanation for the different treatment of the termination of existing relationships (compared to the refusal to enter into new ones) lies in fairness considerations: competition policy is used to achieve an “equitable” sharing of risks between the dominant firm, on the one hand, and its contractual partner on the other.

V. What’s wrong with a fairness-based Article 82? Although there have been relatively few Article 82 cases since the adoption of the Treaty of Rome and although those cases have concerned only a few dominant companies,59 there seems to be a wide consensus among practitioners, academics and some competition officials in the EU (with the exception of Germany) that current Article 82 policy may be chilling aggressive competition to the ultimate detriment of consumers.60 Numerous commentators have criticized recent Commission decisions and judgments of the Community Courts in Article 82 matters on the grounds that perfectly legitimate behaviour was condemned as abusive. These voices maintain that such errors are due to the per se or quasi-per se approach to Article 82 adopted by the Commission and the Community Courts—an approach which, following the ordoliberal tradition, neglects the potential efficiencies resulting from the adoption of certain practices when undertaken by dominant companies (dominant firms have a special responsibility to maintain and promote competition) and sets a very low and mainly formalistic bar for authorities or claimants to show likely anticompetitive effects (any behaviour that deviates too far from the benchmark of competition on the merits under complete competition is taken to distort competition, irrespective of its actual effects). The critical voices mentioned above have repeatedly asked the Commission for a reform of the criteria determining the application of Article 82. As explained in this Section, their demands are, in our view, justified.

59

Robert O’Donoghue and A. Jorge Padilla, supra note 56. This emerges clearly from the Comments on the public consultation on discussion paper on the application of Article 82 to exclusionary abuses, March 2006. These comments are available at: http://ec.europa.eu/comm/competition/antitrust/art82/contributions.html. 60

(D) Ehlermann 08 Ahlborn & Padilla

76

6/11/08

11:48

Page 76

Session One: The Sherman Act and Abuse of Position Under Art 82

A. Assessing fairness through the lenses of social welfare Following Kaplow and Shavell (2002), we submit that “legal rules should be selected entirely with respect to their effects on the well-being of individuals in society and that notions of fairness . . . should receive no independent weight in the assessment of legal rules”.61 Under this approach to the assessment of rules, a concept of fairness, such as the protection of small and medium-size firms, is only a proxy to help guide the design of antitrust rules that foster the well-being of individuals; fairness cannot become a goal in itself. Antitrust rules which seek the pursuit of fairness will be regarded as optimal only when they coincide with rules that take into consideration the consequences that their implementation produces on individuals’ well-being. Consider, for example, a policy which makes every consumer worse off because it redistributes output from efficient, low-priced competitors towards inefficient, high priced ones. Could such a policy be justified because it redistributes market shares from large companies to small companies? Likewise, consider a policy that forces innovators to give away the results of their efforts in order to level the playing field with their competitors. Such a policy may eliminate the incentives to innovate, hence hurting consumers. Could it be justified because it leads to increased rivalry? These questions have a simple answer: no. The social value of policies aimed at preserving rivalry and ensuring a competitive level playing field is given by the impact of such policies on aggregate social welfare. Protecting rivalry is not an end in itself: it only makes sense if it helps to increase consumer welfare. Or, in other words, social welfare is the meta-objective that justifies objectives such as the promotion of competition and the protection of the competitive process. It follows that any rule that is non-consequentialist or, which being consequentialist, does not focus exclusively on the consequences that bear on individuals’ welfare will be automatically regarded as undesirable. As noted in Section III above, the rules advocated by the ordoliberals of the 20th Century were clearly consequentialist and welfaristic: they argued in favour of fairness as a means to protect and foster long-term welfare. However, as shown in what follows, the per se rules emerging from the case law that was inspired by ordoliberal principles are not consequentialist: they emphasize form and neglect effect. The case law appears to regard fairness as a goal in itself and not just an instrument in the pursuit of well-being. Not surprisingly, we will conclude that these are inappropriate because they are inefficient, i.e., welfare-reducing.

61

Kaplow and Shavell, supra note 5.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 77

From Fairness to Welfare 77

B. An error-cost framework for the assessment of legal rules Under a welfare standard, antitrust rules should determine the conditions in which unilateral business practices undertaken by dominant firms are likely to be found welfare-reducing. Unfortunately, given the vagaries of any judicial or administrative process and the absence of proper economic guidance, there will always be mistakes.62 Distinguishing pro-competitive from anticompetitive actions with certainty is impossible. In some cases the practice will be considered abusive when it is not, while in others it will be regarded as legitimate when it should have been considered abusive. The assessment of any set of rules must take into account the risk and cost of those mistakes. Socially desirable antitrust rules—i.e., those promoting welfare-enhancing behaviour and deterring anticompetitive actions—minimize the expected cost of error, while maintaining a degree of predictability for businesses (legal certainty) and administrative ease for courts (administrability). Table 1 shows the standard error matrix, with the shaded boxes reflecting the two possible errors that enforcement agencies and the judicial system can make: falsely condemning competitive practices (“false convictions”, or type I errors) and falsely absolving anticompetitive practices (“false acquittals”, or type II errors).63 The likelihood of type I errors is given by the percentage of cases falsely condemning legitimate practices (the dark shadow box), while the likelihood or incidence of type II errors is given by the percentage of cases falsely absolving legitimate practices (the light shadow box). Table 1. Possible Errors in the Antitrust Assessment of Business Practices Illegal

Legal

Harmful to competition

Percent of cases correctly condemning anticompetitive practices

Percent of cases falsely condemning legitimate practices

Not harmful to competition

Percent of cases falsely absolving legitimate practices

Percent of cases correctly absolving legitimate practices

The expected costs of those two kinds of errors are a function of their welfare implications and the likelihood with which they occur. The latter 62 See David Evans and Jorge Padilla, “Designing Antitrust Rules for the Assessment of Unilateral Practices: A Neo-Chicago Approach”, 72 University of Chicago Law Review 73, 80–81 (2005). 63 Ibid.

(D) Ehlermann 08 Ahlborn & Padilla

78

6/11/08

11:48

Page 78

Session One: The Sherman Act and Abuse of Position Under Art 82

depends, in turn, on a number of factors, including most importantly the legal rule itself. A weaker rule or standard makes it easier to establish that a practice is anticompetitive and therefore reduces the likelihood of type II errors (false acquittals) while increasing the probability of the type I errors (false convictions). That is, type I errors will tend to be most frequent under a per se illegality rule and least frequent under a per se legality approach; the opposite will hold true for type II errors. The cost of a type I error when assessing unilateral practices is typically given by a reduction in the firms’ incentives to invest and innovate. This is because a tight antitrust policy regarding the practices undertaken by firms with market power is akin to introducing an upper bound on profits. Evidence of false convictions is bound to reduce the incentives to invest by reducing the expected rate of return on successful innovations. In welfare terms, therefore, the cost of a type I error is equal to the loss in welfare resulting from the lack of introduction of valuable goods and services for which there is potential demand. (That is given by area CS + ⌸ in Figure 1.) The cost of a type II error is typically equal to the sum of (1) the loss of consumer welfare that results from supra-competitive prices resulting either from direct exploitation or from the market power acquired by exclusion, and (2) the potential reduction in consumer welfare resulting from the fact that some new products and services may not see the light of day if the dominant firm succeeds in excluding rivals. The first term, in turn, is given by two effects: (a) some consumers pay more than in an otherwise competitive market to obtain the good or service they wish (area ⌸ in Figure 1), and (b) other consumers see themselves excluded from consumption despite their relatively high valuations (area L).

Figure 1: Consumer Welfare, Profits and the Deadweight Loss of Monopoly

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 79

From Fairness to Welfare 79

C. Too many and too costly false positives Current Article 82 policy is likely to cause too many and too costly type I errors. There may be type II errors as well but those are likely to be much less likely, as we explain below. This is problematic. We expect the costs of type I errors in antitrust decisions involving unilateral practices to be significantly larger than those of type II errors. First, as noted by Judge Easterbrook, “the economic system corrects monopoly more readily than it corrects judicial errors. There is no automatic way to expunge mistaken decisions of the Supreme Court. A practice once condemned is likely to stay condemned, no matter its benefits. A monopolistic practice wrongly excused will eventually yield to competition though, as the monopolist’s higher prices attract rivalry.”64 That is, if an anticompetitive business practice is mistakenly permitted, the monopoly profits flowing from that practice attract aggressive competition and new entrants. Market forces play no such general corrective role for pro-competitive business practices found anticompetitive.65 Second, as shown by Evans and Padilla (2005)66 and Ahlborn, Evans and Padilla (2005),67 area CS in Figure 1 is likely to be larger than area L. In other words, the welfare cost of a reduction in the incentives to innovate is likely to outweigh the increase in welfare caused by more short-term competition. The reasons why the ordoliberal approach to Article 82 is likely to produce too many false positives in both absolute and relative terms are as follows.

1. A vacuous standard: “competition on the merits” Under current Article 82 policy, a dominant firm abuses its market position when it does not compete on the merits. (See Section IV.) But nowhere is “competition on the merits” defined. Economists would generally agree that the behaviour of companies without market power should be taken as an indication of what competition on the merits is. However, the case law makes it clear that the behaviour of the competitive fringe does not necessarily constitute a valid benchmark of pro-competitive behaviour: sometimes it is, but sometimes it is not, and there is no way to know a priori which case is which. Firms with market power are supposed to behave almost “as if” they did not have market power, but the behaviour of firms without market power is not always regarded as a valid benchmark. This leads to circular and conclusory 64

Easterbrook, supra note 12, p. 15. Firms will likely be reluctant to implement alternative business practices that replicate the one found to be anticompetitive, as such practices are also likely to be found anticompetitive. 66 Evans and Padilla, supra note 62. 67 Christian Ahlborn, David Evans and Jorge Padilla, “The Logic and Limits of the ‘Exceptional Circumstances’ Test in Magill and IMS Health”, 28 Fordham International Law Journal 1109 (2005). 65

(D) Ehlermann 08 Ahlborn & Padilla

80

6/11/08

11:48

Page 80

Session One: The Sherman Act and Abuse of Position Under Art 82

tests: competition on the merits is every practice that the Community institutions find to be legitimate competition. Some authors have argued that current Article 82 policy is not vacuous. The goal, according to these commentators, is unambiguous: the protection of rivals. This is incorrect. There are business practices, such as investing in superior quality, which the Commission would regard as competition on the merits despite its adverse effect on competitors’ welfare. So, even though the protection of rivalry constitutes the cornerstone of the system, some practices which may lead to increased concentration and market power may still be regarded as legitimate. When? When the Commission and the Community Courts acknowledge that the practice in question is objectively justified or when they conclude that the practice represents aggressive competition. Again, competition on the merits is every unilateral practice that the Community institutions find to be legitimate. This is a vacuous standard.

2. The link between the protection of rivalry and long-term welfare The ordoliberal approach to Article 82 is intellectually grounded on the assumption that there is a clear-cut link between rivalry and long-term welfare. As explained in Section II above, protecting rivalry is seen as a necessary condition for long-term economic freedom and thus for long-term welfare. If rivalry disappears, or is seriously weakened, mutually-reinforcing economic and political monopolies will emerge. Democracy will be overturned and the market will no longer ensure the efficient allocation of resources. The rents unfairly extracted from consumers will be used to consolidate a totalitarian regime, which will further entrench the monopolies’ grip on the economy. (See Section III.) This doomsday theory is speculative and incorrect. The triumph of the totalitarian regimes in the first half of the 20th Century had more to do with the inability of some states to cope with unemployment and poverty than with the existence of a flamboyant plutocracy. Furthermore, what matters for long-term welfare is not the number of competitors currently operating in the market, but rather the extent to which the economy is able to grow dynamically.68 In modern economies the rate of growth is largely determined by the ability to innovate by introducing new, more efficient processes and new, more desirable products. In contrast to what is often stated, as a matter of economics, there is no unambiguous relationship between innovation and the number of actual competitors at any point in time. As recently stated by Katz and Shelanski (2007), “enforcement authorities cannot confidently presume as a matter of economic theory or experience that more competitors are 68 See, e.g., Benjamin H. Friedman, The Moral Consequences of Economic Growth, Knopf, 2005; William J. Baumol, The Free-Market Innovation Machine: Analyzing the Growth Miracle of Capitalism, Princeton University Press, 2004.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 81

From Fairness to Welfare 81 beneficial or that market power is detrimental for R&D, except in the limited case of merger to monopoly where the evidence supports a moderate presumption of harm”.69 The impact of the ordoliberal doomsday theory has been, on the one hand, the neglect of beneficial effects arising from certain unilateral business practices on short-term consumer welfare and, on the other hand, an overemphasis on the potentially adverse effects of such practices on rivals. The tradeoff between the certain, positive effects on short-term consumer welfare of a given practice and its negative though (highly) uncertain adverse effects on long-term welfare has typically been resolved in favour of the latter. In most Article 82 cases, the consequences of the alleged abuse for short-term consumer welfare was neglected and the protection of rivalry acquired, de facto, the status of ultimate goal.

3. Per se rules and type I errors The case law on Article 82 has given rise to a series of per se or quasi-per se rules which seek to operationalize in practice vacuous notions such as competition on the merits, normal competition and objective justification.70 Some business practices are considered abusive because they (a) cannot be characterized as competition on the merits, (b) are likely to produce anticompetitive effects (i.e., to harm rivalry and hence long-term welfare), and (c) lack objective justification (even when similar practices are commonly undertaken by companies without market power). However, other practices, which produce similar effects on rivals and consumers and which are no more ubiquitous, are accepted as competition on the merits and are therefore outside the scope of Article 82 (or per se legal). In short, unilateral business practices are classified as per se illegal or per se legal based on formal criteria rather than on case-bycase analyses of effects or likely effects. Consider, for example, the following two scenarios. In the first scenario, a dominant company decides to increase the length of its product portfolio by adding a new product for which there is positive consumer demand and for which there are no current substitutes in the market. Suppose further that consumers prefer to concentrate their purchases on a single supplier (one-stop shopping). Due to shopping costs, the introduction of the new product by the dominant company will increase its sales of other products in its portfolio. 69 Michael Katz and Howard Shelanski, “Mergers and Innovation”, 74 Antitrust Law Journal 1 (2007). 70 These are “quasi-per se rules” because, whereas substantial market power must be established prior to any finding of an infringement under Article 82, liability under Section 2 of the Sherman Act can be established in cases of attempt to monopolize, even in the absence of clear market power, if certain restrictive conditions are satisfied (predatory or exclusionary conduct; a specific intent to monopolize the relevant market; and a dangerous probability of success). See Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447 (1993).

(D) Ehlermann 08 Ahlborn & Padilla

82

6/11/08

11:48

Page 82

Session One: The Sherman Act and Abuse of Position Under Art 82

Consumers of other products will switch to the dominant firm to save the costs of shopping around. Competitors will be harmed and some may even be forced to leave the market. Consumer welfare is unambiguously increased in the short-term (consumers have one more product to enjoy), but the effect on long-term welfare is ambiguous if competitors are forced to exit. Yet we would expect the introduction of a new product for which there are no substitutes but for which there is consumer demand to be regarded as competition on the merits. In the second scenario, a dominant firm decides to integrate two previously existing products into a single product. Suppose further that the company decides not to sell the components of the integrated product as stand-alone products. Consumers are willing to pay more for the integrated product than they were willing to pay for the existing products because the new product is easier to use and produces more satisfaction than the previous combination. As a matter of economics, the welfare evaluation of this second practice is practically identical to the welfare evaluation of the practice described in the first scenario. Indeed, both practices are isomorphic when the cost of shopping around in the first scenario is sufficiently high. When that is the case, the only way of buying the new product in the first scenario is to buy the existing products from the same supplier. So the new product is de facto integrated with the existing ones. And yet, the integration of two existing products into one by a dominant firm and its refusal to sell the components separately is very likely to be regarded as abusive because it allegedly: (a) limits the number of options available to consumers, (b) hurts competitors selling only one or the other component, and (c) cannot be defended on efficiency grounds because, if the integrated product is really superior, its impact on the structure of the market will be extensive—rivals will be forced out of the market and long-term welfare will be negatively affected as a result of the monopolization of the market. As noted above, most economists would treat the two practices in the same way. However, according to the case law, one is likely to be considered per se legal and the other per se illegal. In principle, under a per se illegality standard there are type I errors but no type II errors, since the practice would always be considered abusive. The opposite is true under a per se legality standard: no type I errors but possibly a large number of type II errors. However, a careful review of the existing case law indicates that, while those practices that have been regarded as per se legal under EC competition law (such as abovecost pricing and investment in superior quality) are likely to be welfareincreasing in most circumstances, some of the practices that have been considered per se illegal (such as some individualized rebates) are unlikely to be welfare-reducing in many circumstances. In other words, the per se legality rules in the case law involve no (or only limited) type II errors, whereas the per se illegality rules do cause type I errors. So the case law is characterized by more type I errors, which are likely to be, for the reasons explained above,

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 83

From Fairness to Welfare 83 more costly. Once again, the reason for this asymmetry is found in the ordoliberal tradition and, in particular, in the so-called “special responsibility” of dominant firms, which in practice justifies the neglect of legitimate efficiency arguments.

4. Lack of legal certainty Current Article 82 policy is likely to chill innovative business practices that have a positive effect on individuals’ welfare. Companies with market power will refrain from adopting new business practices for fear that they may not be regarded as competition on the merits or objectively justified and, as a result, that they may be found per se illegal. This entails great social costs. Antitrust rules are more a system of deterrence than a system of regulation. They should be designed so that they deter conduct that is anticompetitive and welfare-reducing without discouraging pro-competitive, welfareenhancing competition. And their outcomes should be predictable, because otherwise they may fail to deter the business practices that harm social welfare while at the same time chilling socially desirable business practices. In other words, a well-functioning system of deterrence requires legal certainty. A vacuous standard implemented through a series of per se rules may appear to involve a high degree of legal certainty. But that is a false impression. Existing Article 82 per se rules yield predictable outcomes only when the business practice under scrutiny fits seamlessly into the rule. But more often than not, the facts of the case will not match the rule, and then the company will have to anticipate what per se rule will apply in its particular case: legality or illegality. And to do so it will have to consider whether the practice in question constitutes “competition on the merits” and (no less challenging) whether it is “objectively justified”. It will also have to anticipate whether or not competitors may be marginalized and the impact of all that in terms of long-term consumer welfare. But since the notions of “competition on the merits” and “objective justification” are empty, and since assessing the effect of a given practice on long-term consumer welfare is exceedingly difficult, commercial decisions will have to be adopted in the midst of an impenetrable regulatory fog. Some of these practices may be regarded as per se legal ex post, i.e., when the practice has been implemented, but that can by no means be anticipated ex ante, i.e., when the strategic decision is made. Fear of a finding of infringement under Article 82 will do the rest, and consumers will suffer the cost of inaction.

(D) Ehlermann 08 Ahlborn & Padilla

84

6/11/08

11:48

Page 84

Session One: The Sherman Act and Abuse of Position Under Art 82

VI. The Consumer welfare rationalization of the “modernized” enforcement policy under Article 82 Unlike in the case of mergers or Article 81, the Commission has not yet suffered a major court setback in connection with Article 82.71 The Commission faces no pressure from the Community Courts to “modernize” current Article 82 policy; on the contrary, its policy has been endorsed by the Courts time and again. The pressure this time comes from the public. The Commission has been severely criticized for its decisions in Michelin II, British Airways, Wanadoo and Microsoft, even though the first three decisions have been upheld by the Courts.72 The Commission’s reaction to these criticisms has been to open an internal and external debate about its policy towards unilateral business practices. The first output of this evaluation exercise was the publication in December 2005 of the Discussion Paper.

A. DG Comp’s Discussion Paper The Discussion Paper lays out “possible principles for the Commission’s application of Article 82 of the Treaty to exclusionary abuses”73 and describes “the analytical approach that could be used by the Commission”74 in assessing unilateral business practices. Most importantly for our current purposes, paragraph 54 of the Discussion Paper states that “[t]he essential objective of Article 82 when analysing exclusionary conduct is the protection of competition on the market as a means to enhancing consumer welfare and of ensuring an efficient allocation of resources”. The emphasis on the “protection of competition on the market” reflects the traditional ordoliberal view that competitive rivalry should be protected.75 However, the Discussion Paper clarifies that the protection of the competitive process is not an end in itself but rather is meant to enhance consumer welfare, which is traditionally served by low prices, and by quality and innovation. This seems positive: a clear move from fairness to welfare. However, a careful reading of the Discussion Paper suggests that the ordoliberal tradition still exerts a strong influence on the Commission’s views about Article 82 and that the move to a consumer welfare standard is not complete. We identify below 71 See Damien Neven, “Competition economics and antitrust in Europe”, 48 Economic Policy 741 (2006). 72 As of this writing, the judgment of the CFI in Microsoft is still pending. 73 Discussion Paper, supra note 3, para. 1. 74 Ibid., para. 2. 75 See generally Gerber, supra note 13.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 85

From Fairness to Welfare 85 three problems with the Discussion Paper’s framework for the analysis of Article 82.

1. Excessive emphasis on allocative efficiency The Discussion Paper adopts a “static view” of consumer welfare: the objective of Article 82 is to ensure “an efficient allocation of resources”. Economic theory shows that, under fairly general conditions, increased rivalry tends to improve the allocation of resources. Industry fragmentation brings prices closer to (incremental) costs and enhances short-term consumer welfare. However, in the long term, consumer welfare is not only influenced by the relative efficiency with which the scarce resources of an economy are allocated; it is also positively affected by the emergence of new products, technologies and services. In industries where firms compete by launching new products and by developing new technologies, having numerous firms with no market power, and hence low market prices, is not necessarily ideal. In fact, the opposite may well be true: the prospect of lower prices may reduce the incentives to invest and innovate. As noted above, there is no support in economic theory or evidence for the oft-stated proposition that product market fragmentation promotes innovation.76

2. A broad notion of anticompetitive effects The Discussion Paper states at paragraph 55 that “Article 82 prohibits exclusionary conduct which produces actual or likely anticompetitive effects in the market and which can harm consumers in a direct or indirect way”. This paragraph is problematic both for what it says and for what it fails to say. Unilateral actions by dominant firms may be found abusive even if there is no evidence of “actual” effects if they could “likely” have an adverse effect on consumer welfare. However, it is not clear how likely those adverse effects would have to be for the practice to qualify as abusive. Nor is it clear how significant the effect on competition should be for unilateral conduct to be regarded as abusive. The Discussion Paper does not clarify what is the standard of proof in Article 82 cases. Is it a function of: the degree of market power of the dominant company, the existence of countervailing efficiencies, or the degree of certainty of its potential anticompetitive effects? This is a matter of concern because an excessively broad or abstract interpretation of the notion of anticompetitive effects in specific cases could lead to the prohibition of conduct that is beneficial to consumer welfare. The Discussion Paper states at paragraph 58 that “[f]oreclosure thus can be found even if the foreclosed rivals are not forced to exit the market: it is 76

See supra note 69 and accompanying text.

(D) Ehlermann 08 Ahlborn & Padilla

86

6/11/08

11:48

Page 86

Session One: The Sherman Act and Abuse of Position Under Art 82

sufficient that the rivals are disadvantaged and consequently led to compete less aggressively . . . Foreclosure is said to be market distorting if it likely hinders the maintenance of the degree of competition still existing in the market or the growth of that competition and [it] thus [will] have as a likely effect that prices will increase or remain at the supra-competitive level.” This notion of foreclosure is excessively broad. Competitors may be foreclosed, according to the definition in the Discussion Paper, even if they are not forced to exit or cornered into a market niche. All that the Commission will have to show under this standard is that the actions of the dominant firm constrained the growth of its competitors by placing them at a competitive disadvantage.77 This makes any claim of market foreclosure very hard, if not impossible, to rebut. If foreclosure means exit, then showing no exit is a rebuttal. If it means marginalization, then showing that rivals are experiencing growth is a rebuttal. However, if foreclosure means a reduction in the rate of growth of competitors, there is no possible rebuttal, since there is no counterfactual available. What would be the rate of growth of competitors absent the alleged abuse? This is precisely the problem experienced by Michelin, British Airways, France Télécom and, more recently, Telefónica, when defending their practices. The fact that they had lost market share at the expense of their allegedly foreclosed competitors was considered irrelevant by the Commission and by the Community Courts.

3. Impossible defences The second step in the analytical framework proposed in the Discussion Paper requires a consideration of the possible objective justifications and/or efficiency defences. Paragraph 77 states: “Exclusionary conduct may escape the prohibition of Article 82 [if] the dominant undertaking can provide an objective justification for its behaviour or [if] it can demonstrate that its conduct produces efficiencies which outweigh the negative effect on competition”. Paragraph 84 stipulates that, in order for this “efficiency defence” to be accepted, “the dominant company must demonstrate that the following conditions are fulfilled: (i) that efficiencies are realised or likely to be realised as a result of the conduct concerned; (ii) that the conduct concerned is indispensable to realise these efficiencies; (iii) that the efficiencies benefit consumers; (iv) that competition in respect of a substantial part of the products concerned is not eliminated”. While the Commission’s acceptance of an efficiency defence is a step forward, the conditions that need to be fulfilled to accept efficiency as a defence 77 In Michelin II, cited supra note 47, the CFI upheld the Commission’s conclusion that Michelin’s declining market share did not constitute proof of lack of effect. The Commission argued, and the court accepted, that competitors would have done better if Michelin had behaved differently.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 87

From Fairness to Welfare 87 are very strict and difficult to prove. The fourth condition has an unambiguous ordoliberal flavour. As stated in paragraph 91: “Ultimately the protection of rivalry and the competitive process is given priority over possible pro-competitive efficiency gains.” Ordoliberals understood that increased concentration could be efficient due to economies of scale and scope but were concerned that those efficiencies would be exploited to monopolize markets and dominate society. However, from a welfare standpoint, the absolute priority given to the protection of rivalry over possible pro-competitive efficiency gains could lead to the maintenance of inefficient market structures. A dominant company seeking to justify its business practices on efficiency grounds will have to demonstrate that the pro-competitive effects of those practices more than offset its potential anticompetitive effects. The burden of proof in this complex balancing exercise is placed on the defendant. This is in contrast with the way these analyses are conducted across the Atlantic. In the US, if the monopolist asserts a pro-competitive justification that stands unrebutted, then the plaintiff must demonstrate that the anticompetitive harm of the conduct outweighs the pro-competitive benefits.78 The allocation of the burden of proof is likely to have a decisive role in Article 82 cases. While the likely anticompetitive effects of a given practice, if any, may be easy to measure, its pro-competitive effects, while certainly important, will be difficult to measure with precision. Thus, the balancing test will be hard to carry out and the party with the burden of proof is likely to lose the case.79

B. Nihil novum sub sole The tone and content of the Discussion Paper indicate that the Commission sees no reason to turn its current Article 82 enforcement policy upside down. While the Discussion Paper suggests some limited reforms in areas such as the treatment of efficiencies and the adoption of an as-efficient competitor test, its main proposition is that the case law on Article 82 is fully consistent with the consumer welfare standard. According to the Discussion Paper, the case law emphasis on the protection of rivalry, or more euphemistically the “competitive process”, can be justified within the consumer welfare standard: protecting rivalry is the means to ensure that market outcomes maximize long-term consumer welfare. However, the Commission’s rationalization of its current policy in terms of welfare does not resolve any of the problems discussed above in Section IV. What constitutes competition on the merits according to the Commission? 78

See U.S. v. Microsoft, 253 F.3d 34, 95–97 (D.C. Cir. 2001). See Damien Geradin, “Limiting the scope of the Article 82 EC: What can the EU learn from the U.S. Supreme Court’s judgment in Trinko in the wake of Microsoft, IMS, and Deutsche Telekom?”, in 41 Common Market Law Review 1519, 1542 (2004). 79

(D) Ehlermann 08 Ahlborn & Padilla

88

6/11/08

11:48

Page 88

Session One: The Sherman Act and Abuse of Position Under Art 82

We are told that behaviour that is unlikely to foreclose competitors and is objectively justified constitutes competition on the merits. But given how broadly foreclosure is defined and the hurdles imposed on any efficiency defence, such advice does not clarify much. Are dominant firms still bound by the special responsibility principle? Though the Discussion Paper makes no explicit reference to this principle, only the application of this principle can explain the wide notion of foreclosure adopted in the text—one that requires dominant players to consider the impact of their actions on the rate of growth of their rivals. Many passages suggest that we have not moved much. Some of the recent cases in which we have been involved, including most importantly Microsoft, confirm that suspicion. The application of Article 82 is still likely to produce too many type I errors and, if anything, the degree of legal certainty in Europe is for the time being as low as ever. What will happen next? Recent court decisions in British Airways and Wanadoo may reinforce the position of those who think that change is unnecessary. On the other hand, the growing importance of economics and economists within the Commission may lead to a more radical reform. Apparently, all options remain open. It remains to be seen which side will prevail. We certainly do not have the knowledge required to predict with a minimum degree of rigour how a complex administrative and political body such as the Commission is likely to move. But we can discuss the normative question: what should happen next? Several proposals have been advanced. We discuss them in turn.

VII. A radical consumer welfare approach to Article 82 The shortcomings of the ordoliberal approach have led some economists, mostly academics, to propose a radical reform of Article 82. This proposal can be found in the report written by DG Comp’s Economic Advisory Group on Competition Policy (EAGCP) and published in July of 2005.80 According to the key features of this proposal: (a) the goal of competition policy should exclusively be the protection of long-term consumer welfare, (b) the dominance hurdle should be removed, and (c) unilateral conduct should be assessed according to the so-called “unstructured” rule of reason. We support (a). In what follows, we discuss (b) and (c) in some detail.

80

EAGCP, supra note 10.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 89

From Fairness to Welfare 89

A. Down with dominance According to the EAGCP, “there is no need to establish a preliminary and separate assessment of dominance”.81 First, such an evidentiary hurdle is not needed. Competition authorities and courts using an economic approach to Article 82 should consider directly whether a given practice is abusive given the facts of the case. Given that firms can only commit an abuse if the firm has a dominant position, “no further separate demonstration of dominance is needed”.82 Furthermore, traditional methods for the assessment of dominance are rather imperfect and often unable to properly identify actual situations of market power.

B. The unstructured rule of reason The EAGCP considers that the assessment of unilateral conduct is likely to produce both false convictions and false acquittals, and it therefore advocates the application of an unstructured rule of reason under Article 82. Under this rule, courts and agencies would balance the costs and benefits of intervention given the specific facts of the case and would intervene only when the benefits outweigh the costs of intervention. This balancing exercise would be conducted on a case-by-case basis. First, a theory of harm must be specified. Then, if there are alternative interpretations of the effects of the conduct in question, and if the conduct can be regarded as legitimate under one or more of those theories, the facts of the case must be thoroughly scrutinized to discriminate among the competing interpretations. The rule of reason proposed by the EAGCP is unstructured because it introduces no filter prior to the balancing of pro-competitive and anticompetitive effects, other than the identification of those effects. Nonetheless, the EAGCP’s proposal demands that a full-fledged theory of harm be specified. It also suggests that both “the standard and the process adopted” should reflect the fact that one type of error may be more costly than the other. Thus, for example, “if the cost of false negatives is expected to be higher, then the balance should be tilted towards plaintiffs and against dominant firms”.83 We are not sure how this recommendation would be implemented in practice.

81 82 83

Ibid., p. 4. Ibid., p. 14. Ibid., p. 7.

(D) Ehlermann 08 Ahlborn & Padilla

90

6/11/08

11:48

Page 90

Session One: The Sherman Act and Abuse of Position Under Art 82

C. Is this the end of Article 82 “history”? This question has a simple answer: no. Both limbs of the EAGCP proposal have problems. On dominance, we believe that market power is a fundamental screen. A dominance threshold simplifies administration and self-assessment. Enforcement can be better targeted and the risk of over-deterrence is minimized with a clearly-specified dominance threshold. If anything, in our view, optimality would require a higher, less ambiguous dominance threshold. Judging from the responses given during the public consultation to the Discussion Paper, we do not walk alone in this regard.84 As regards the unstructured rule of reason, we have argued elsewhere that this too is problematic.85 In short, we believe that an unstructured rule of reason standard is likely to underestimate the expected costs of false convictions (type I errors) and hence will tend to over-enforce. We base this conclusion on several reasons. First, firms generally engage in unilateral practices because it is efficient for them to do so. This is true whether they enjoy market power or not. However, it is widely recognized that economists do not understand these sources of efficiencies well and that in any event they are difficult to document, let alone quantify, persuasively.86 It is also conventional wisdom that businesspeople have difficulty documenting and sometimes even articulating these efficiencies. Consequently, unless competition authorities and courts show some deference to efficiency explanations offered for unilateral practices, especially those that are common in competitive markets, efficiency rationales will be wrongly disregarded, which will cause many legitimate practices to be mistakenly regarded as abusive. An unstructured rule of reason standard shows no such deference; instead, it treats each candidate explanation as equally likely. Second, an unstructured rule of reason test ignores that the principal implication of several decades of economic investigation of the competitive effects of unilateral practices is that there should be no presumption that these practices are anticompetitive, even when undertaken by firms with monopoly power.87 Firms with the ability to cause consumer harm do not often have an incentive to do so, and firms with the incentive to act anticompetitively do not

84

See supra note 60. See, e.g., Christian Ahlborn, David Evans and Jorge Padilla, “The Antitrust Economics of Tying: A Farewell to Per Se Illegality”, 49 Antitrust Bulletin 287 (2003). 86 See David S. Evans and Michael A. Salinger, “Why Do Firms Bundle and Tie? Evidence from Competitive Markets and Implications for Tying Law”, 22 Yale Journal on Regulation, 37 (2005). 87 See, inter alia, Michael D. Whinston, “Exclusivity and Tying in U.S. v. Microsoft: What We Know and What We Don’t Know”, 15 Journal of Economic Perspectives 63 (2001); Patrick Rey, Paul Seabright, and Jean Tirole, “The Activities of a Monopoly Firm in Adjacent Competitive Markets: Economic Consequences and Implications for Competition Policy”, IDEI mimeo, 2001, http://idei.fr/doc/wp/2001/activities_abs.pdf. 85

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 91

From Fairness to Welfare 91 often have the ability to do so. Unilateral practices may be proven to be anticompetitive in particular cases, but no unilateral practice should be regarded as anticompetitive per se. Third, comparing the efficiency effects and the anticompetitive effects of a unilateral practice is necessarily an extremely complex exercise. As noted by Geradin, “balancing ex ante vs. ex post efficiencies is obviously a very difficult process, which even the most sophisticated economists may find daunting”.88 The risk of mistaken decisions is therefore high. On the one hand, as we have just mentioned, measuring the benefits of that practice may prove difficult. In addition, the game-theoretic models that show the possibility of anticompetitive effects do not provide necessary and sufficient conditions susceptible of empirical testing that could be used in a rule of reason inquiry.89 Due to this complexity, the outcome of an unstructured rule of reason is bound to be highly unpredictable, which is highly undesirable given that, as noted above, Article 82 creates mainly a system of deterrence, and effective deterrence requires dominant firms to be able to carry out meaningful self-assessment. One of the problems with the unstructured rule of reason is that, due to its complexity, it will lead to very high levels of error for firms self-assessing their behaviour. Fourth, a recent study by Nalebuff and Majerus (2003) provides some insights into the relative likelihood of type I and type II errors in the assessment of unilateral practices. These authors reviewed 11 tying cases in the US and Europe and assessed whether the reviewing authority reached the right judgment.90 Table 2 summarizes our understanding of their conclusions. There are no instances in which the courts and administrative agencies erroneously found that a tying practice considered by the authors to be anticompetitive was legal—that is, there were no false acquittals. By contrast, in three of the 11 cases the decision bodies wrongly found that a tying arrangement considered by the authors to be pro-competitive was illegal. The rate of false convictions was therefore 27 percent. In other words, Nalebuff and Majerus found no evidence of type II errors but they did find a relatively high percentage of type I errors.

88 Damien Geradin, “Limiting the Scope of Article 82 of the EC Treaty: what can the EU learn from the US Supreme Court’s Judgment in Trinko, in the wake of Microsoft, IMS, and Deutsche Telekom?”, mimeo, Global Competition Law Centre, College of Europe, Bruges, at p. 19. 89 See Dennis Carlton and Michael Waldman, “The Strategic Use of Tying to Preserve and Create Market Power in Evolving Industries”, 33 RAND Journal of Economics 194, 215 (2002). 90 Barry Nalebuff and David Majerus, “Bundling, Tying and Portfolio Effects: Part 2-Case Studies”, DTI Economics Paper, February 2003. While the report nominally contains thirteen separate cases, two are about different aspects of the GE/Honeywell merger. Also, since one of the cases had not been decided when the report was completed, we exclude it, leaving 11 cases. See also David S. Evans, A. Jorge Padilla and Michael A. Salinger, “A Pragmatic Approach to Identifying and Analyzing Legitimate Tying Cases”, in Ehlermann and Atanasiu, supra note 52, at pp. 557 et seq.

(D) Ehlermann 08 Ahlborn & Padilla

92

6/11/08

11:48

Page 92

Session One: The Sherman Act and Abuse of Position Under Art 82

Table 2. Possible Errors in the Antitrust Assessment of Business Practices Illegal

Legal

Harmful to competition

Four (36%)

None (0%)

Not harmful to competition

Three (27%)

Four (36%)

Fifth, an unstructured rule of reason may yield absurd outcomes. Consider again the case of a dominant company which decides to enlarge its product portfolio by adding a new product for which there is positive consumer demand and for which there are no current substitutes in the market in an industry where consumers prefer to concentrate their purchases with a single supplier (one-stop shopping). This decision is likely to increase short-term welfare but, with some probability, it may lower long-term welfare by throwing some competitors out of the market. Is this a business practice that should be subject to the costly rule of reason, or would it be preferable to treat it as per se legal? We believe that per se legality is unambiguously preferable in this case. Sixth, an unstructured rule of reason is also likely to make it difficult for appeals courts to effectively perform their role as an error correction mechanism. Balancing pro-competitive and anticompetitive effects requires detailed knowledge of facts but, very often, it also requires the use of complex economic reasoning and the application of sophisticated quantitative techniques. Appeals courts may be ill-equipped to deal with those tools and theories. Even if they did have the necessary capacity, they may be subject to legal rules requiring them to show deference to the economic analysis conducted by the competition agencies or lower courts. Seventh, but not least importantly, the fact that the Commission acts both as prosecutor and judge in competition policy matters implies a risk of “prosecutorial bias”—i.e., officials may tend to concentrate on evidence that confirms their own judgment.91 Prosecutors operating in an inquisitorial system like the European antitrust enforcement system have an incentive to focus on one side of the argument. They are afraid that, by looking for evidence on both sides, their evidence will not be conclusive.92 As concluded by 91 Kai-Uwe Kühn, “Reforming European merger review: targeting problem areas in policy outcomes”, 2(4) Journal of Industry, Competition and Trade 311 (2002); Wouter W. Wils, “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 (2004). 92 Mathias Dewatripont and Jean Tirole, “Advocates”, 99 Journal of Political Economy 1 (1999).

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 93

From Fairness to Welfare 93 Neven (2006): “Overall, it would thus appear that the self-confirming biases that may be induced by the prosecutorial role that the Commission assumes cannot be dismissed as insignificant.”93 It is our firm belief that the risk of prosecutorial bias is enhanced under legal rules which leave a broad margin of discretion to prosecutors, and that it is minimized when the hands of prosecutors are tied by clear-cut rules. This is one additional reason of concern regarding the unstructured rule of reason. As Benjamin Franklin once said, “[s]o convenient a thing is to be a reasonable creature since it enables one to find or make a reason for everything one has a mind to do”.94

VIII. A Third Way There is a middle ground between the ordoliberal approach and the unstructured rule of reason. On the one hand, this “third way” takes consumer welfare seriously. No ambiguous concepts such as competition on the merits, objective justification, or the competitive process are adopted. The number of competitors, especially the number of small competitors, means nothing per se under this approach. On the other hand, our approach recognizes that “administrability” and “legal certainty” are key desirable characteristics in competition law. So this approach minimizes the extent of discretion. The antitrust rules advocated under this approach seek to maximize longterm consumer welfare, taking into account (i) the risk and cost of both type I and type II errors, and (ii) the transaction costs of carrying out a full-fledged balancing exercise. The rules that emerge from this constrained maximization exercise are described below.

1. A structured rule of reason Under this rule, a unilateral practice is considered legal unless it can be shown to survive a sequence of screens aimed at establishing the likelihood of anticompetitive effects. If that likelihood is established, courts and agencies would then still have to balance the anticompetitive and pro-competitive effects of the practice in a final stage prior to concluding that it is anticompetitive overall. While the game-theoretic models developed by industrial organization economists over the last few decades do not provide a universally valid set of conditions that could be used by competition authorities and courts to distinguish between pro-competitive and anticompetitive practices with certainty, these 93 94

Neven, supra note 71, at p. 773. Benjamin Franklin, Writings, Library of America, 1987.

(D) Ehlermann 08 Ahlborn & Padilla

94

6/11/08

11:48

Page 94

Session One: The Sherman Act and Abuse of Position Under Art 82

models do provide a series of screens for determining which practices antitrust authorities should investigate and ultimately condemn as abusive. First, economic theory shows that a unilateral business practice can plausibly have anticompetitive effects only when certain structural conditions are met. For example, in the case of tying, a review of the economic literature by Ahlborn, Evans and Padilla (2003) leads these authors to identify the following structural conditions for tying to be anticompetitive: (1) market power in the tying market; (2) imperfect competition in the tied market; (3) significant economies of scale and scope; (4) barriers to entry and re-entry into the tying and the tied markets; and (5) absence of buyer power. These structural conditions are not too difficult to analyze empirically. Unilateral business practices taking place in markets where these structural conditions are met would then need to be subjected to a second screen—a further analysis to determine whether the practices at issue are likely to have anticompetitive effects.95 Second, economic theory provides models that can be used to develop a theory of harm. We agree with the EAGCP that every Article 82 case should rely on a well-specified theory of harm with a clear link to established economic theory. One does not need to have a fully specified mathematical theory published in an economic journal. But one needs to have a theory that can be confirmed or falsified by testing it against facts prior to finding an abuse. Only when those assumptions hold true given the facts of the case can one conclude that the practice is likely to produce anticompetitive effects. Practices that are likely to have anticompetitive effects (i.e., practices that survive these two screens) would finally need to be subjected to a third screen to determine whether there are offsetting efficiencies. Third, this final screen requires a determination of whether the business practice under scrutiny generates efficiencies that can only be achieved through that practice, and whether these efficiencies are greater than its potentially anticompetitive effects. In conducting this analysis one would need to consider dynamics and uncertainty. The anticompetitive effects need time to materialize—market foreclosure leads to exit, which then leads to higher prices. Those effects need to be discounted to reflect the fact that they occur in the future and are uncertain. As noted above when discussing the appropriateness of the unstructured rule of reason, this balancing exercise is an empirically demanding task. As Carlton and Waldman (2002) have recently explained: “We would like to caution that trying to turn the theoretical possibility for harm . . . into a prescriptive theory of antitrust enforcement is a difficult task. For example, the courts would have to weigh any potential efficiencies from the tie with possible losses due to foreclosure, which by itself is challenging due to the difficulty of measuring both the relevant efficiencies and the relevant losses.”96 95 96

Ahlborn, Evans and Padilla, supra note 85. Carlton and Waldman, supra note 89, p. 212.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 95

From Fairness to Welfare 95 The structured rule of reason is likely to be superior in error cost terms to the unstructured rule of reason because it limits the complex balancing of procompetitive and anticompetitive effects to those practices that are likely to have anticompetitive effects. Indeed, a proper inquiry into efficiencies can be time-consuming, and accordingly it is best done only for those practices that pass through the earlier screens. For example, if a defendant lacks significant market power, economic theory says that it lacks the ability to exclude competitors and cause consumer harm, so we should end the inquiry there. Compared to an evaluation of efficiencies, analyzing market power is much more amenable to the standard tools available to economists. Both the structured and unstructured rule of reason may produce non-trivial type I and type II errors in those cases where the business practice in question is likely to produce anticompetitive effects as well as significant efficiencies.

2. Qualified per se legality Under this rule, a unilateral practice is presumed legal unless one or more well-defined, and empirically testable conditions are met—i.e., the practice is considered legal but for “exceptional circumstances”. A qualified per se legality rule will produce less type II errors but possibly more type I errors than a pure per se legality rule. Likewise, it will produce less type I errors but possibly more type II errors than a pure per se illegality rule. From an error cost perspective, given that false convictions involve a greater expected cost, it appears that a qualified per se legality rule should outperform a per se illegality rule in welfare terms, as the former minimizes the incidence of the type of error that is most costly. Furthermore, a qualified per se legality rule should necessarily be superior to a per se legality rule, provided that the set of exceptional circumstances which triggers intervention is properly designed. Ahlborn, Evans and Padilla (2005) illustrate how to construct an exceptional circumstances test (i.e., a qualified per se legality rule) in the case of refusals to license intellectual property (IP).97 They show that, unless the refusal to license IP prevents the introduction of new products for which there is substantial demand, the cost of a type I error in those cases is greater than the cost of a type II error. The optimal legal standard for the antitrust assessment of refusals to license IP by dominant companies takes the form of a modified per se legality rule, whereby compulsory licensing is required only when: (a) the requested IP is indispensable to compete; (b) the refusal to license causes the exclusion of all competition from the downstream market; (c) the refusal prevents the emergence of markets for new products for which there is substantial demand; and (d) the products to be developed by the 97

Ahlborn, Evans and Padilla, supra note 67.

(D) Ehlermann 08 Ahlborn & Padilla

96

6/11/08

11:48

Page 96

Session One: The Sherman Act and Abuse of Position Under Art 82

licensees are sufficiently differentiated from those of the IPR holder, e.g., because they satisfy needs that the existing products fail to address. In their article, Ahlborn, Evans and Padilla rely on economic theory and experience to identify the circumstances where: (1) the pro-competitive effects (or ex post efficiencies) of compulsory licensing are greatest, and (2) the disincentive effects (or ex ante inefficiencies) of the obligation to license are minimal or non-existent. Conditions (a) and (b) in the foregoing paragraph ensure that the short-term welfare loss resulting from a refusal to license is maximal. Condition (c) implies that the refusal has a long-run cost as well as a short-term cost. And condition (d) aims to ensure that the long-run cost of compulsory licensing—the reduction in the incentives to innovate—is low. This method thus restricts intervention to scenarios where it is certain to create more benefits than costs.

3. Comparing alternative structured antitrust rules Our choice of legal standard boils down to choosing between the structured rule of reason standard and the qualified per se legality rule. One important factor in choosing between a rule of reason standard and a per se rule is the administrative and enforcement costs of implementing the legal standard. A qualified per se rule is arguably easier and cheaper to administer and enforce.98 But the key fact in this comparison is obviously the assessment of the likelihood and cost of error under each of the two approaches. Both rules may produce false acquittals and false convictions. And neither of these two rules is a priori superior to the other in error terms because the expected cost of error of each will depend on their fine details: the set of “exceptional circumstances” that define the qualified per se legality and the sequence of screens that characterizes the structured rule of reason. One might expect, however, that the qualified per se legality rule would result in relatively more false acquittals, while the structured rule of reason might cause relatively more false convictions. The structured rule of reason would also involve the additional administrative costs of having to proceed through a series of complex screens.

4. Sacrifice test, no economic sense test, and as-efficient competitor test Some other rules have been proposed in the last few years, in particular the sacrifice test, the no economic sense test and the as-efficient competitor test.99 98 See Keith N. Hylton, Antitrust Law, Economic Theory and Common Law Evolution, Cambridge University Press, 2003. 99 See John Vickers, “Abuse of Market Power”, 15 Economic Journal 244 (2005), and references therein.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 97

From Fairness to Welfare 97 What all of these rules have in common is that, although their ultimate goal is consumer welfare, they do not regard just any practice that harms consumers as being anticompetitive, but only those which do so by distorting the competitive process. For example, the as-efficient competitor test marks a practice as abusive only when it eliminates a competitor whose efficiency is equal to or greater than that of the dominant company. The no economic sense test identifies a unilateral practice undertaken by a dominant company as abusive when the only possible explanation for that practice is the exclusion of competitors, i.e., the practice in question does not make economic sense but for the exclusion of competitors. In our opinion, these rules are best treated as potential filters within a structured rule of reason analysis, rather than as autonomous antitrust rules. Otherwise, as shown by Vickers (2005)100 and Salop (2007),101 they are likely to produce both type I and type II errors from a consumer welfare viewpoint.

IX. Whither Article 82? The debate about Article 82 and its reform has lasted for several years now. As far as we can recall, the idea of clarifying the Commission’s policy on Article 82 was first launched by Commissioner Mario Monti at the 8th Annual EU Competition Law and Policy Workshop organized by ClausDieter Ehlermann and Isabela Atanasiu in Florence in June 2003.102 That was approximately four years ago. Since then, Commissioner Kroes delivered her celebrated Fordham 2005 speech,103 DG Comp published its Discussion Paper on exclusionary abuses,104 and many scholars and practitioners have published contributions on the subject. This debate appears to have led to what we shall denote as the “Brussels consensus”. This consensus involves two key propositions. First, the goal of Article 82—or more generally, the goal of EC competition law—should be aggregate consumer welfare. Second, the existing form-based approach to Article 82 needs to be replaced by an effects-based or economics-based approach to the assessment of unilateral conduct. The consensus we have just described is based on the following observations:

100

Ibid. Supra note 11. 102 Mario Monti, “Introductory Statement”, in Ehlermann and Atanasiu, supra note 52, at pp. 3 et seq. 103 Supra note 3. 104 Supra, note 3. As noted earlier, the comments in response to the Discussion Paper were also made public. See supra note 60. 101

(D) Ehlermann 08 Ahlborn & Padilla

98

6/11/08

11:48

Page 98

Session One: The Sherman Act and Abuse of Position Under Art 82

• A policy which reduces aggregate welfare, i.e., a policy with an adverse impact on individual welfare, cannot be legitimately justified by reference to vague notions of fairness. As noted by Professors Kaplow and Shavell in their remarkable book Fairness versus Welfare,105 deontological and/or non-consequentialist rules, like those which characterize the ordoliberal approach to Article 82, cannot be justified unless they produce the same effects on consumer welfare that would obtain under an effects-based or consequentialist approach, which has consumer welfare as its only goal. • Competition policy should not be concerned with distributional issues. First, those issues can be addressed by other means. Second, as recently argued by Professors Alesina and Tabellini, income distribution is not the sort of policy that should be left to bureaucracies and/or independent agencies.106 • Unilateral conduct that is harmful to competitors may generate efficiencies and may have a net positive effect on aggregate welfare. Distinguishing between exclusionary acts, which reduce social welfare, and competitive acts, which increase it, is very difficult. Courts and competition regulators using a simplistic form-based approach are bound to make mistakes: some practices will be found legal when they are welfare-reducing and vice versa. This is because two practices or transactions which share the same form— e.g., two exclusive dealing contracts—can have very different welfare effects depending on the structure of the market, while two practices which appear to be different from a formalistic point of view may have the same adverse effects on consumer welfare. • Existing form-based rules are likely to cause too many type I errors (false positives) because they are not properly grounded on economic theory and experience, and because they largely neglect the potential efficiencies generated by conduct which under purely formalistic criteria superficially appears to be anticompetitive. Consensus does not mean unanimity, however. Some disagree with the two commandments of the Brussels consensus (i.e., aggregate consumer welfare and an effects-based approach). Some are still attached to the ordoliberal principles that underlie customary Article 82 policy. In their opinion, there is no evidence that existing policy has produced inefficient outcomes—i.e., that it has chilled competition. They further note that the traditional policy has been endorsed time and again by the Community Courts. At the other extreme, some other scholars and practitioners, mainly from the other side of the Atlantic, maintain that competition policy is not a very useful tool, especially as regards the assessment of unilateral conduct. These laissez-faire

105

Kaplow and Shavell, supra note 5. Alberto Alesina and Guido Tabellini, “Bureaucrats or Politicians? A Single Policy Task, Part I”, 97 American Economic Review 169 (2007). 106

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 99

From Fairness to Welfare 99 advocates consider that society would be better off without Article 82. Needless to say, we do not share the views of either the ordoliberal “originalists” or the non-interventionist libertarians. Instead, we want to briefly discuss the fundamental issue on which there is still disagreement among those that endorse the Brussels consensus. What is that fundamental issue? It is the degree of discretion that society should allow courts and competition agencies for purposes of balancing the pro-competitive and anticompetitive effects of unilateral conduct. For that matter, the Brussels consensus on Article 82 remains incomplete. There is still disagreement as regards: (i) the weight to be given to evidence (or lack thereof) of actual effects as opposed to likely effects; (ii) the allocation of the legal burden of proof (as opposed to the evidentiary burden of proof); (iii) the role to be given to economic and econometric analysis in all of this; and (iv), as illustrated in the preceding sections of this paper, the optimal design of antitrust rules. The “rules versus discretion” controversy is a debate about objectives as well as about implementation. Is the objective of Article 82 consumer welfare ex post or is it instead long-term consumer welfare ex ante, i.e., taking into account the likelihood and cost of error? Or, in other words, is Article 82 a system of enforcement or a system of optimal deterrence? On one side of the debating table, those that favour rules over discretion argue that balancing pro-competitive and anticompetitive effects is hard to do and that economists, lawyers and courts are all fallible. The outcomes of unconstrained balancing exercises—in other words, the outcomes resulting from an unstructured application of the rule of reason—will be uncertain ex ante (at the time strategic business decisions need to be made) and often incorrect ex post (when commercial behaviour is assessed). Companies will not be able to predict the outcome of exercises of such complexity when taking their day-to-day commercial decisions (ex ante). And competition authorities and courts may (with bona fide intentions) condemn practices that are legitimate, and vice versa, after those practices have been implemented (ex post). Both shortcomings should be a matter of concern, but the first problem—uncertainty ex ante or, in other words, the lack of legal certainty—is likely to be the more costly of the two. This is because, as noted before, Article 82 establishes a system of deterrence that can only work properly if companies can self-assess their practices, which in turn requires legal certainty. This is why uncertainty in the context of Article 82 may lead to over-deterrence and may have a chilling effect on competition. And this problem would be particularly severe if companies perceived that the policy is not only uncertain but also tends to produce too many false positives through ex post enforcement. As noted by Director General Philip Lowe, “[c]lear and simple rules provide not only enforcers but also companies with legal certainty. They [signal] where to draw the line between legal and illegal behaviour. They also have

(D) Ehlermann 08 Ahlborn & Padilla

100

6/11/08

11:48

Page 100

Session One: The Sherman Act and Abuse of Position Under Art 82

greater deterrent effect and generate lower enforcement costs”.107 As Greg Werden, an economist from the DOJ said recently, “[a]lthough legal uncertainty cannot be eliminated in any practical manner, it can be mitigated significantly, and the simplest way to do so is through the adoption of clear and simple rules limiting the domain of Section 2 and Article 82”.108 We fully agree. There are at least two different ways to structure such rules: the structured rule of reason and the qualified per se legality rule. Relying on available economic theory and evidence, a number of papers have shown that, “on average”, these rules maximize long-term consumer welfare. They do so by minimizing the expected cost of error in the application of Article 82. (See Section VIII above.) Both rules are superior to the unstructured rule of reason because: (i) they limit the complex balancing of pro-competitive and anticompetitive effects to those practices that are likely to have serious anticompetitive effects, and to situations where market power (dominance, if you like) is real and not just marginal; (ii) paraphrasing Professor Hovenkamp, they operate with “tolerable accuracy, in particular without excessive . . . errors”;109 and (iii) they are easier to administer. Regarding administrability, we are convinced that these structured rules will facilitate judicial review and that they will be more robust against manipulation or, if you will, against the so-called “prosecutorial bias” (see Section VII). On the other side of the debate, you have those that consider that there is no reason why courts and competition regulators should not be “free to ramble through the wilds of economic theory in order to administer a flexible approach”—to use the language of U.S. v. Topco.110 We strongly believe that economics may be more useful in designing “workable rules” than in balancing efficiencies against anticompetitive effects on a case-by-case basis. In our opinion, for an unstructured rule of reason to perform satisfactorily, our courts and competition authorities would have to be populated by enlightened economists born and bred in the arcane business of balancing pro- and anticompetitive effects—a sort of Philosopher King, like those whom Plato would have ruling his Republic. We do not believe in such utopias. Since our goal is the welfare of individuals, and since that is in our view a treasure too precious to play with, we prefer a more pragmatic approach to economic policy in general and to competition policy in particular. The structured rule of reason and the qualified per se legality rule are 107 Philip Lowe, “Enforcement authorities”, in Barry Hawk, ed., International Antitrust Law & Policy: Fordham Competition Law 2005, Juris Publishing, 2006, p. 44. 108 Gregory J. Werden, “Identifying single-firm exclusionary conduct: from vague concepts to administrable rules” in Barry Hawk, ed., International Antitrust Law & Policy: Fordham Competition Law Institute 2005, Juris Publishing, 2006, p. 601. 109 Herbert J. Hovenkamp, The Antitrust Enterprise: Principle and Execution, Harvard University Press, 2006. 110 United States v. Topco Associates, 405 U.S. 596, 610 (1972).

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 101

From Fairness to Welfare 101 pragmatic approaches with desirable welfare properties. We concur with the following observations of Professor Vickers: “To say that the law on abuse of dominance should develop a stronger economic foundation is not to say that rules of law should be replaced by discretionary decision making based on whatever is thought to be desirable in economic terms case by case. There must be rules of law in this area of competition policy, not least for reasons of predictability and accountability. So the issue is not rules versus discretion, but how well the rules are grounded in economics. To that end there is great scope for economic analysis and research to contribute to the development of the law on abuse of dominance. To be effective, however, economics must contribute in a way that competition agencies, and ultimately the courts, find practicable in deciding cases.”111

Though we are very confident that the legal and economic pragmatism of Hovenkamp, Vickers, Werden and others will prevail, we are concerned about the most recent developments. DG Comp’s Article 82 initiative appears to be stalled. The Discussion Paper has not led to Draft Guidelines or any other public document. This is problematic. It creates a vacuum that reactionary forces will fill. It is important for DG Comp not to lose interest in its Discussion Paper, and a revised paper is needed in the near future. The availability of guidelines would be extremely useful not only for practitioners and companies, but most importantly for its own internal use. That would increase consistency and would show commitment to what we have called the Brussels consensus. A delay in providing the necessary guidance may produce too many casualties (i.e., false convictions) and in legal matters the weight given to precedents makes the corpses of today the zombies of tomorrow, and we all know that killing zombies is not easy.

111

Vickers, supra note 99.

(D) Ehlermann 08 Ahlborn & Padilla

6/11/08

11:48

Page 102

(E) Ehlermann 08 Zimmer

6/11/08

11:48

Page 103

III Daniel Zimmer* On Fairness and Welfare: The Objectives of Competition Policy A comment on papers by David J. Gerber and by Christian Ahlborn & A. Jorge Padilla August 2007

Both papers on the objectives of competition law are very clear in their analyses. Both provide a solid basis for discussion. David Gerber makes a clear-cut distinction between two different sets of questions where economics can come into play: first, economics play an important role with respect to the analysis of facts and empirical data; second, economics may serve as a policy input when deciding on the normative question of which goals competition policy should serve. The paper reflects, in a clear manner, the state of the law: European competition law, in particular the core provisions within Articles 81 and 82, in the courts’ interpretations, are process-oriented, and not, at an initial stage of analysis, directly consumer-oriented. Christian Ahlborn and Jorge Padilla very thoroughly analyze the different reform proposals and develop their own suggestions. Deciding in favour of consumer welfare as the single goal of competition policy, they discuss a very sophisticated rule of reason and a qualified per se legality approach and evaluate both under cost-of-error aspects. What remains is the question at the policy level. Should the Community’s competition policy, as restated by David Gerber, be changed in a fundamental sense—possibly in the sense of a direct and sole consumer-welfare orientation as advocated by Christian Ahlborn and Jorge Padilla?

Competition means more than just consumer welfare After the judgment of the Court of Justice of 15 March 2007 in British Airways,1 the overall structure of competition law seems more transparent than ever before. At first examination, European competition law follows an open, structure- and process-oriented approach, meaning that it is unnecessary— * University of Bonn. 1 Case C-95/04 P, British Airways v Commission [2007] ECR I-2331.

(E) Ehlermann 08 Zimmer

104

6/11/08

11:48

Page 104

Session One: The Sherman Act And Abuse Of Position Under Art 82

even for an application of the prohibition rules—to prove that, in a specific case, the conduct is detrimental to consumers. Competition law is about competition. Thus, starting with a semantic analysis, it seems obvious that competition means more than just consumer welfare. In British Airways, the ECJ repeated the statement it had already made in its 1973 Continental Can judgment,2 to the effect that there are structural prerequisites to competition. According to the ECJ, the Court of First Instance had not erred in law by foregoing an examination of whether the conduct in question had caused prejudice to consumers and by limiting itself to an analysis of whether British Airways’ bonus scheme had a restrictive effect on competition.3 However, in light of the ECJ’s judgment in British Airways, we know that there may be a second stage of analysis at which efficiency and consumer welfare are directly introduced. At paragraph 86 of the judgment, the Court indicates that an efficiency justification is also permissible under Article 82 EC.4 Hence, today it seems clear that European competition law requires, in all three areas—Article 81, Article 82 and Merger Control—a two-stage analysis. At the first stage it follows a more open approach and does not directly concern itself with consumer benefits. Why not? According to the law as it stands in Europe, competition seems to serve more purposes than solely consumer welfare. Of course, consumer benefit is at the core of the concept, but by virtue of its open approach, EC competition law is able to cover more. It has always been an effect of competition law to protect individuals—who are not necessarily final consumers—against an abuse of market power. There are some test cases by which this may be demonstrated. One of these is demand power, or demand cartelization. If, for example, large car manufacturers form a demand cartel in order to purchase parts more cheaply, this is a cartel under Article 81(1) EC. One might argue that this cartel benefits consumers, at least in the case where, corresponding to lower purchase costs, the car manufacturers will lower car sales prices. Under the legal system of Article 81, however, this justification is only admissible within the narrow limits of paragraph 3—which entails an inquiry both as to the indispensability of the

2 Case 6/72, Europemballage Corp and Continental Can Co Inc v Commission [1973] ECR 495, paragraphs 25 et seq. 3 British Airways, supra note 1, para. 107. 4 To quote the paragraph in full: “Assessment of the economic justification for a system of discounts or bonuses established by an undertaking in a dominant position is to be made on the basis of the whole of the circumstances of the case (see, to that effect, Michelin, paragraph 73). It has to be determined whether the exclusionary effect arising from such a system, which is disadvantageous for competition, may be counterbalanced, or outweighed, by advantages in terms of efficiency which also benefit the consumer. If the exclusionary effect of that system bears no relation to advantages for the market and consumers, or if it goes beyond what is necessary in order to attain those advantages, that system must be regarded as an abuse.”

(E) Ehlermann 08 Zimmer

6/11/08

11:48

Page 105

On Fairness and Welfare

105

restraint for the attainment of the objective and as to the subsistence of residual competition in the market following the transaction. Under the current system of Article 81, demand cartels are—unless the conditions of paragraph 3 are met—prohibited. As the ECJ has repeatedly stated, competitors must not substitute “practical cooperation” for the “risks of competition”.5 The cartel prohibition applies not only downstream, but upstream as well: as long as there is no need to create a counter-weight to a powerful supplier, a purchasing cartel falls within the prohibition of Article 81(1).6 Thus, by securing the independence of decision-making among competitors, the law protects certain prerequisites, or “living conditions” of the competitive process. This “behavioural” component is, together with the “structural” conditions discussed above in the context of the British Airways judgment, another important element of the ECJ’s concept of the competitive process. An effect of this system is that suppliers are protected against demand cartelization. The example of a demand cartel clearly shows that the law does not currently confine itself to the goal of maximizing the welfare of final consumers. A demand cartel is prohibited regardless of whether or not it leads to consumer harm.

The relationship between fairness and welfare The relationship between fairness and welfare requires, first, some clarification with respect to this terminology. Louis Kaplow and Steven Shavell have advanced the thesis that social policies should be assessed entirely on the basis of their effects on individuals’ well-being.7 However, they advocate a wide notion of welfare, opening the concept up to accommodate any advantage which could be considered to further the well-being of individuals. Such a concept of welfare, which is able to integrate the value judgments of a legislator, might help to overcome the existing communication problem between proponents of a “competitive distortion” model and a “welfare” model of competition policy. The EC Treaty’s decision to protect competition as such is based on a fundamental confidence in the benefits of a market economy. As has been demonstrated above, this concept does not allow for a reduction to just one single aim such as consumer welfare.8 Competition advances a number of 5 Case C-89/85, Ahlström Osakeyhtiö and others [1993] ECR I-1307, para. 63. See also Case 40/73, Suiker Unie [1975] ECR 1663, paras. 173 et seq. 6 Case C-250/92, Gøttrup-Klim and others v DLG [1994] ECR I-5641, paras. 28 et seq. 7 Louis Kaplow and Steven Shavell, “Fairness versus Welfare: Notes on the Pareto Principle, Preferences, and Distributive Injustice”, 32 The Journal of Legal Studies 331 (2003). 8 For an economics-based argument in favour of a consumer welfare orientation, see Damien Neven and Lars-Hendrik Röller, “Consumer surplus vs. welfare standard in a political economy model of merger control”, 23 International Journal of Industrial Organization 829 (2005).

(E) Ehlermann 08 Zimmer

106

6/11/08

11:48

Page 106

Session One: The Sherman Act And Abuse Of Position Under Art 82

aims which cannot be determined in an exclusive manner. Individual freedom of action, the protection of market participants against the abuse of market power by others, an interest of consumers in a cheap supply of the goods desired, and a collective interest in the promotion of technical and scientific progress have all been regarded as figuring among these aims. The application of competition law is not dependent on positive evidence that just one of these ends is achieved. The law protects competition as such because it is deemed, by way of presumption, to have favourable effects of one sort or another. A deviation from the principle of competition is admissible only in exceptional cases. It requires a demonstration that a restriction of competition, such as a cartel agreement, is necessary for the improvement of the production or the distribution of goods or for the promotion of technical or economic progress, whilst allowing consumers a fair share of the resulting benefit. This two-stage analysis (beginning with the principal prohibition of restraints on competition, and only then allowing for exceptional justifications on grounds of efficiency) applies, as noted earlier, in all three areas of competition law: cartel prohibition, abuse of dominance and merger control. There is no need, nor is there any reason, for a change of the described policy. As shown above, to concentrate on a single aim, namely that of consumer welfare, would reduce the scope of competition law significantly. The current law protects not only consumers but also producers. For example, suppliers of goods are, as has been demonstrated, protected against purchasing cartels. Suppliers may also invoke the prohibition provision of Article 82, which prohibits an abuse of demand power by dominant purchasers as well. Why does the law protect producers and not only consumers? Producers have the same rights as anyone else in a market economy. In particular, they enjoy property rights. If we wish, and this seems to be the most important point, to encourage people to commit themselves to an activity on a market, we ought to protect them from expropriation. This includes protection against expropriation by way of cartelization or abuse of dominance. If we took the view that it is acceptable for a purchasing cartel to deny a manufacturer the possibility of earning revenues on a previous investment, we would create a disincentive to market participation. This reasoning refers to the infrastructural function of law in general and of competition law in particular: in order to create incentives, not disincentives to participate in the market, the legal system must provide a set of rules guaranteeing that the legitimate expectations of the players are realized. This includes not only the possibility to enforce contractual obligations, but also a protection against unjust exploitation. One may call this—if one wishes to do so—a “fairness” consideration. In this sense, competition law serves, inter alia, a goal of fairness. It is important to see that this consideration is not a goal in itself, but rather serves an economic function as well: we have good reason to assume that “fair” and “equitable” rules promote the willingness of firms and individuals to engage in market operations, thereby furthering the

(E) Ehlermann 08 Zimmer

6/11/08

11:48

Page 107

On Fairness and Welfare

107

aggregate well-being of individuals. This argument demonstrates that welfare requires more than the avoidance of deadweight loss and the acceptance of “efficient” restraints on competition. The use of modern economic techniques when applying competition law cannot be praised highly enough. The contribution of economic theory in the context of the assessment of facts—such as market definition and the prognosis of the effects of particular conduct or of structural changes—is indispensable.9 However, as David Gerber points out in his paper, economic theory cannot provide an answer to normative questions. In a democracy with a division of powers, it is for the legislator to decide upon normative issues. The EC Treaty, in Articles 2 to 4, lays down a multitude of objectives, including a common policy in the areas of agriculture, fisheries and transport, the coordination of the Member States’ employment policy, social and environmental policy, the promotion of research and technological progress, the encouragement of the development of trans-European networks and measures to achieve a high level of health protection. In the particular context of competition policy, the Treaty follows—as has been demonstrated—an open, process-oriented approach. A reduction to just one single aim—such as consumer welfare—is inconsistent with this normative decision.

9 For an analysis of the use of modern economic tools in European and German merger control, see Ulrich Schwalbe and Daniel Zimmer, Kartellrecht und Ökonomie—Moderne ökonomische Ansätze in der Europäischen und Deutschen Zusammenschlusskontrolle, Verlag Recht und Wirtschaft, 2006.

(E) Ehlermann 08 Zimmer

6/11/08

11:48

Page 108

(F) Ehlermann 08 Fox

6/11/08

11:48

Page 109

IV Eleanor M. Fox* The Market Power Element of Abuse of Dominance— Parallels and Differences in Attitudes—US and EU July 2007

Introduction US law prohibits monopolization and attempts and conspiracies to monopolize. EU law prohibits the abuse of a dominant position. In the 21st Century, these concepts are largely substitutes for one another. Both are meant to proscribe anticompetitive conduct by firms with dominance or monopoly power. On both sides of the ocean, mere possession of monopoly power is not in itself an abuse or violation, although in earlier times US courts and agencies flirted with the notion that possession of monopoly power was an actionable offense, and in the famous Alcoa case1 the Second Circuit came close to so holding. This essay reflects principally on contemporary parallels and differences between US and EC law. On both sides of the ocean, certain threads have long since been woven into the law, and therefore the essay begins with history. Second, the essay reflects on issues of proof, thresholds and presumptions. Third, the essay describes and assesses a draft document by an ICN working group concerning suggested best practices for identifying dominance/substantial market power. Finally, it concludes with a note on the integral nature of the market power/market effect analysis.

I. History: What threads of history inform contemporary thinking? Whether a firm has substantial market power would seem to be an empirical question, and therefore technical and not normative. However, there are several factors at play that give the inquiry a normative gloss. These may be enumerated as follows. * Walter J. Derenberg Professor of Trade Regulation, New York University School of Law. 1 United States v. Aluminum Corp. of America, 148 F.2d 416 (2d Cir. 1945).

(F) Ehlermann 08 Fox

110

6/11/08

11:48

Page 110

Session One: The Sherman Act And Abuse Of Position Under Art 82

1) Law must be workable and enforceable; therefore, economics must be generalized.2 Generalization requires proxies and presumptions. They move the law in one direction or another, and that direction is informed by objectives and values. 2) The US law was passed and the EC Treaty provisions were adopted each in response to perceived problems, and each in a particular socio-politicaleconomic context which may be the source of those objectives and values. 3) Competition law may be more or less aggressive, given its specific goals (e.g., consumer welfare or efficiency). The law may aspire to move society closer to the goal when the market falls short, or it may reflect a preference against intervention. Therefore, the degree of trust that antitrust intervention will do better than an imperfect market is a variable that matters. In this section I will briefly consider history and trust as background factors that may explain certain differences between the EU and US approaches to identifying market power.3 For the United States, the history may be merely that, for we have moved away from and rejected paradigms that held sway for 90 years. The US antitrust laws were a response to concentrated economic power which, Americans feared, would deliver the country to communism or fascism, destroy its democracy, and undermine personal autonomy. With such a perception underlying our laws (particularly influencing the law in the years of World War II), the courts easily presumed market power and monopoly from large relative size and from other possible proxies such as patents, as documented in Alcoa 4 and International Salt.5 Beginning in the late 1970s, however, the US antitrust jurisprudence embraced a more technical microeconomic perspective.6 In the 1980s, the jurisprudence began to dismantle the pro-diversity, anti-power decisions of earlier times,7 and in the last few years Supreme Court case law introduced the notion that antitrust intervention against conduct other than cartels is usually an evil to be avoided.8 In 2007, the Supreme Court virtually completed the project of dismantling anti-power antitrust.9 The contemporary impulse of US jurists and enforcers is to help make American business strong and competitive in the new world economy by withdrawing the “heavy hand” of government. The

2 See Justice Breyer, dissenting in Leegin Creative Leather Products, Inc. v. PSKS, Inc., 127 S. Ct. 2705 (2007). Law is not economics, and even the views among economists differ, although economics greatly influences the law. 3 There is much more commonality than divergence, but this paper concerns divergence. 4 Supra note 1. 5 International Salt Co. v. United States, 332 U.S. 392 (1947). 6 See Eleanor Fox, “The Modernization of Antitrust—A New Equilibrium”, 66 Cornell Law Review 1140 (1981). 7 Ibid. 8 See, as to refusals to deal, Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004). 9 See, e.g., Leegin, supra note 2. Compare the majority opinion with Justice Breyer’s dissent.

(F) Ehlermann 08 Fox

6/11/08

11:48

Page 111

The Market Power Element of Abuse of Dominance 111 older structural analysis has not died; but it has become heavily elaborated by facts, hypotheses, and free-market presumptions.10 In Europe in 1957, the driving force behind the Treaty of Rome was the quest for peace among hostile post-war countries through the vehicle of market integration. Market dominance was not seen by the Treaty negotiators as a problem in itself. But it was anticipated that dominant firms would seek to abuse their power, to the detriment of all market players, be it competitors, customers, or suppliers. There was concern that dominant firms would tilt the playing field in their own direction, disadvantaging other market players.11 In familiar Treaty language, it was held that dominant firms have the special responsibility “not to distort competition”.12 Just as the US had done, Europe adopted many formalistic rules. Much later, in the wake of growing global competitive pressures, DG Competition embarked on a bold and courageous program of “modernization,” which entails among other things the adoption of a realistic economic approach. The substantive reforms associated with modernization pervade the work of the European Commission. Moreover, the Court of First Instance has at times in fact been out in front of the Commission, demanding careful economic analysis and sufficient evidence of facts to fit the theories.13 However, the reforms have not necessarily reached the European Court of Justice, which has thus far adhered to a jurisprudence based on a combination of legal formalism and the rights of excluded actors.14 Apart from the tension between the virtues of legal formalism and those of economic analysis, there are two cross-Atlantic differences that may be lasting. First, values. Article 82 was designed to protect market actors from abuses. A dominant firm’s use of power to secure market advantages (e.g., its use of leverage to shift market share to itself), even if it does not thereby gain more economic power, may continue to violate Article 82.15 Moreover, concern with preserving forces of rivalry (the competitive structure of the 10

See United States v. Oracle Corp., 331 F. Supp. 2d 1098 (N.D. Cal. 2004). Article 82 (along with Article 86) was seen as especially important as a means to rein in state-owned and state-privileged enterprises. Furthermore, relational market power has played a role in certain European cases, as it has in the antitrust law of a number of other countries, including Germany and Japan. For an example in the European jurisprudence, see Case 22/78, Hugin Kassaregister AB v Commssion [1979] ECR 1869. 12 Case 322/81, NV Nederlandsche Banden-Industrie Michelin v Commission [1983] ECR 3461; Case T-203/01, Manufacture Française des Pneumatiques Michelin v Commission [2003] ECR II-4071. 13 See, e.g., Case T-342/99, Airtours v Commission [2002] ECR II-2585; Cases T-5/02 and T-80/02, Tetra Laval BV v Commission [2002] ECR II-4381, affirmed, Cases C-12/03 P and C-13/03 P, Commission v Tetra Laval [2005] ECR I-987; Cases T-310/01 and T-77/02, Schneider Electric SA v Commission [2002] ECR II-4201. 14 See Case C-95/04 P, British Airways v Commission [2007] ECR I-2331; Case T-340/03, France Télécom SA v Commission (Wanadoo), judgment of the CFI of 30 January 2007, not yet reported. However, the British Airways and Wanadoo judgments upheld Commission decisions that were adopted before the “economics revolution”. 15 See Case C-333/94 P, Tetra Pak International SA v. Commission [1996] ECR I-5951. 11

(F) Ehlermann 08 Fox

112

6/11/08

11:48

Page 112

Session One: The Sherman Act And Abuse Of Position Under Art 82

market), avoiding exclusion, and facilitating easier access to markets is likely to continue to be a part of the European economic model, “in line with Europe’s Lisbon agenda”,16 even if another economic model may be equally consistent with efficiency.17 Second, European competition law seems to be endowed with a more aggressive stance than its US counterpart. Americans often claim that US antitrust law is meant to maximize consumer welfare or efficiency; but that is a euphemism. US law is not so interventionist. The antitrust agencies do not proclaim—even when it is true—“I know how to increase consumer welfare”, and then set about to (try to) do so. The European Commission is more likely than the United States to take opportunities to open closed markets (as it did when approving cross-border telecoms alliances) and to apply EC principles imposing affirmative duties on dominant firms, especially multi-functional dominant firms that have incentives to suppress the competitive forces that challenge them.18

II. Proof, thresholds and presumptions 1. Substantiality: What is substantial market power? For dominance or monopoly power to exist, the undertaking must at least have—according to the law of most nations—“substantial market power”. But what is “substantial market power”? Market power is usually defined as the power to increase price significantly over competitive price and to hold it at the high level for a significant period of time. DG Competition’s Discussion Paper on exclusionary abuses under Article 82 states: “Market power is the power to influence market prices, output, innovation, the variety or quality of goods and services, or other parameters of competition on the market for a significant period of time.”19 It states further that the power to “increase

16 Report of the Economic Advisory Group for Competition Policy, advising DG Competition (“EAGCP Report”), p. 17, available at: http://ec.europa.eu/dgs/competition/eagcp.htm. 17 See Trinko, supra note 8. See also Eleanor Fox, “What is Harm to Competition?— Exclusionary Practices and Anticompetitive Effect”, 70 Antitrust Law Journal 371 (2002); Eleanor Fox, “Is There Life in Aspen After Trinko?: The Silent Revolution of Section 2 of the Sherman Act”, 73 Antitrust Law Journal 153 (2005). 18 Microsoft, Case Comp/C-3/37.792, Commission decision of 24 March 2004, available at: http://ec.europa.eu/comm/competition/antitrust/cases/decisions/37792/en.pdf. On appeal: Case T-201/04, not yet decided as of this writing. 19 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), available at: http://ec.europa.eu/comm/competition/antitrust/art82/ discpaper2005.pdf.

(F) Ehlermann 08 Fox

6/11/08

11:48

Page 113

The Market Power Element of Abuse of Dominance 113 prices” is often used as shorthand for all of the foregoing.20 The United States tends to focus more narrowly on price and output than on variety. Both EU and US authorities draw inferences of power over price from structural conditions, and sometimes from behavior that would otherwise not make sense. However, European administrative decisions and court judgments might more readily draw inferences from structural conditions, especially large market shares. Substantiality is a fact-based inquiry. It means, at least, significant; not trivial and not fleeting. Sometimes, especially in the case of a firm that has a very high share in a well-defined, high-barrier market, you may (think you) know it when you see it. Other times, the question is one of fair debate. This was the case in Eastman Kodak Co. v. Image Technical Services.21 Kodak made micrographic equipment. Plaintiffs were independent service organizations (ISOs) dedicated to servicing Kodak equipment. On entering the original equipment market, Kodak had encouraged entry of ISOs dedicated to servicing its equipment. It supplied them with the parts necessary to do so. When the service and repair business proved lucrative, Kodak decided to do the servicing itself, then cut off the ISOs from the parts they needed, and raised the price of service to its equipment customers. Kodak’s service at the higher price was sometimes inferior to the ISOs’ service at the lower price. The ISOs sued Kodak for tying and monopolization of the Kodak equipment service aftermarket. They did not allege that the equipment market was non-competitive. When the case came before the Supreme Court, the sole question to be decided was whether the case could go to trial or whether it was fatally deficient because Kodak did not have market power in a Kodak service aftermarket. (The case did go to trial, and plaintiffs won.) The Supreme Court, in an opinion by Justice Blackmun, held that summary dismissal was improper because the facts (taken in the best light for plaintiffs at this stage) showed that Kodak had the market power to exploit its locked-in customers. Justice Scalia, joined by Justices O’Connor and Thomas, dissented, arguing that competition in the original equipment market would prevent supra-competitive pricing in the aftermarket, and that a manufacturer’s power over its own brand equipment is not “the sort of ‘monopoly power’ sufficient to bring the sledgehammer of § 2 into play”. Given the recent changes in the constituency of the U.S. Supreme Court, one can expect that the dissenters’ assessment of the substantiality factor would command a majority of the justices today. The result was a matter of perspective. The US Microsoft case also presented a substantiality problem. Microsoft occupied more than 90% of the market for Intel-based PC operating systems. 20 21

Ibid., para. 24. 504 U.S. 451 (1992).

(F) Ehlermann 08 Fox

114

6/11/08

11:48

Page 114

Session One: The Sherman Act And Abuse Of Position Under Art 82

Microsoft argued that market forces constantly constrained it to behave competitively, lest it be destroyed by a new technology that would effect a paradigm change. Microsoft lost this battle. In both cases, the market power issue was linked with the market definition issue; in addition, the question of market power was argued separately from the market definition issue. In Kodak, if the market had been drawn to include original equipment sales and thus interbrand competitors, the market would have been broader and the market power claim would have been foreclosed. So too in Microsoft, if the market had been drawn to include Apple, UNIX, and hand-held devices. But even with the market narrowly defined, the defendant had, as it were, a second bite at the Apple. In each case the defendant highlighted all of the constraints that could limit its power: in Kodak, the possibility of switching to other original equipment providers, and concern that it would lose original equipment sales; in Microsoft (also), concern that dissatisfied customers would shift (not much of a concern in view of network effects), and its paranoid fear of paradigm change in this new-economy, fast moving market. As these examples illustrate, substantiality and insubstantiality come in many forms: the robustness or contestability of the market definition, the size of market share, the probable durability of the share, the options of customers, the distance of the next nearest price constraint, and the immediacy of the threat of paradigm change. 2. What is the role of market share in the assessment of dominance/market power, and should thresholds be specified for presumptions and safe harbors? “Market share” assumes a market definition. If the market is well defined and barriers to entry and expansion are high, a persistently high share is a useful proxy for market power. It would normally support a shift of the burden of proof/disproof of market power to the putative dominant firm. But market definition comes first. If it is flawed, or if it simply reflects a reasonable but hard choice, the foundation for using a high market share as a proxy for substantial market power has not been built. Nonetheless, and even assuming a solid, uncontroversial market definition, should market shares of a certain large size presumptively indicate a dominant position and should market shares of a certain small dimension place the firm in a safe harbor? The Commission’s Discussion Paper22 states: “It is very likely that very high markets shares, which have been held for some time, indicate a dominant position. This would be the case where an undertaking holds 50% or more of the market, provided that rivals hold a much smaller share of the market. In the case of lower market shares, dominance is more likely to be found in the market share range of 40% to 50% than below 40%, although also undertakings 22

Cited supra note 19.

(F) Ehlermann 08 Fox

6/11/08

11:48

Page 115

The Market Power Element of Abuse of Dominance 115 with market shares below 40% could be considered to be in a dominant position. However, undertakings with market shares of no more than 25% are not likely to enjoy a (single) dominant position on the market concerned.”23

United States law is more likely to use two-thirds of the market (or more) as a rule of thumb for presuming monopoly power. Further, a market share under 40% to 50% is likely to fall within a safe harbor. Professor Janusz Ordover has reflected as follows on the use of market share proxies and thresholds: “• Both with respect to L[erner] and market share proxy, [there is] no clear prescription from economics at what levels market power converts into DP [dominant position] or SMP [significant market power]. • EU market share thresholds seem unreasonably low but there is no empirical evidence behind the various dicta by the US courts promulgating much higher thresholds.”24

Nonetheless, even in the face of factual ambiguity, agencies have an interest in administrative rules that allow them to act with reasonable expedition and without unreasonable expense, and businesses have an interest in clear guidance.25 The problem is not with thresholds as such but with “excessive reliance on mechanistic criteria”.26

III. The ICN Working Group’s draft best practices on the meaning of dominance and substantial market power The ICN formed a Working Group on Unilateral Conduct at the annual meeting in Cape Town in 2006. The Working Group project includes a segment on “Dominance/Substantial Market Power Analysis”. In 2006/07, the Working Group produced a Discussion Draft for possible suggested best practices.27 The draft includes the following preliminary ideas: “1. As to the legal definition of dominance or significant market power (SMP) [the definition] should take into consideration the complexity of dominance/SMP assessment. Therefore, it should not be based exclusively on a market share threshold, as

23

Ibid., para. 31 (footnotes omitted). Janusz Ordover, “Dominant Position and Monopoly Power: A View from the Cathedral of Economics”, ABA Antitrust Section Spring Meeting (19 April 2007), slide 10. 25 Ibid.; John Vickers, “Market Power in Competition Cases”, 2 European Competition Journal 3, special issue (2006). 26 Ordover, supra note 24. See also Vickers, cited in previous footnote. 27 Editor’s note: The ICN substantially adopted the draft. See http://www.international competitionnetwork.org/media/library/unilateral_conduct/Unilateral_WG_1.pdf. See also the Working Group’s “Report on the Objectives of Unilateral Conduct Laws, Assessment of Dominance/Substantial Market Power, and State-Created Monopolies” (May 2007), available at http://www.internationalcompetitionnetwork.org/index.php/en/library/working-group/18. 24

(F) Ehlermann 08 Fox

116

6/11/08

11:48

Page 116

Session One: The Sherman Act And Abuse Of Position Under Art 82

this may misguide the assessment. The definition should rather focus on a firm’s relative freedom from competitive constraints or ability to act in ways that a competitive firm could not. . . . 2. As to the assessment of dominance/SMP [the assessment] should entail a comprehensive consideration of factors that may be relevant for the competitive conditions in the market under investigation. —At a minimum, market shares and barriers to entry and expansion/durability of market power should be assessed: • Market share analysis (market shares of the company under investigation and position of its existing competitors including their development over the past years) can be used as an initial indicator or starting point for the market power analysis. However, an assessment of dominance/substantial market power based exclusively on a market share threshold is not appropriate. • Many competition authorities find it beneficial to use market share thresholds as dominance/substantial market power presumptions and/or safe harbor presumptions, which in practice provide some form of relief of the agency’s or the plaintiff’s burden of proof. In designing and using such thresholds, the benefits, in particular increased enforcement efficiency and legal certainty especially for businesses, should be weighed against the risks, in particular potential overemphasis on market shares. To address this risk dominance/SMP presumptions and safe harbor presumptions should usually be rebuttable, i.e., that the presumption can be disproved by other evidence. • Entry analysis provides information on the significance of potential competitors for the market concerned and thus about the durability of market power, the constraints faced by the company under investigation and its likely incentives (to compete or engage in anti-competitive conduct). • Beyond these factors, the relevance of further factors such as behavior of competitors as well as buyer power and other aspects should at least be considered. —Competition authorities should make their dominance/substantial market power assessment principles transparent by e.g. publishing their enforcement decisions and/or market power assessment guidelines.”

My reflections in the foregoing pages accord generally with the ICN Working Group’s draft suggested practices. Such an approach would leave to the agency the decision regarding what presumptions to use; it would place emphasis on the assessment of factors such as entry conditions likely to constrain any anticompetitive conduct, and it would encourage the formulation and publication of a framework for analysis that allows the putative dominant firm to tell its story.

(F) Ehlermann 08 Fox

6/11/08

11:48

Page 117

The Market Power Element of Abuse of Dominance 117

IV. A Note in Conclusion In the usual course of analysis, one proceeds from identifying dominance to identifying anticompetitive acts. However, the two inquiries may be part of an integrated assessment. As both the EAGCP Report28 and the FTC/DOJ Commentary on the US Merger Guidelines29 suggest, the enforcer should ask an initial question: What is the story of anticompetitive harm? If there is no credible story, the investigation should be closed. If there is a credible story that the undertaking is using exclusionary strategies to obtain or preserve market power, that observation may imply that market constraints are insufficient to discipline the firm so that it behaves competitively. In short, the credible story may imply that the putative violator has significant market power.

28

See supra note 16. Available at http://www.ftc.gov/os/2006/03/CommentaryontheHorizontalMergerGuidelines March2006.pdf. The Commentary applies only to mergers. The observation would be applicable to dominance only if the conduct identified could not occur without significant market power. See also Vickers, supra note 25, p. 14. 29

(F) Ehlermann 08 Fox

6/11/08

11:48

Page 118

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 119

V Heike Schweitzer* The History, Interpretation and Underlying Principles of Section 2 Sherman Act and Article 82 EC September 2007

I. Introduction European competition lawyers habitually look across the Atlantic for inspiration and guidance when engaging in policy debates and deciding on reforms. As far as rules regarding market power are concerned, this look reveals substantial divergence. In the US, tests for identifying anticompetitive single-firm conduct under Section 2 Sherman Act are frequently more narrowly construed than the tests applied in the EU under Article 82 EC. As far as the enforcement activity of the relevant public enforcement agencies is concerned, cases concerning anticompetitive single-firm conduct are relatively rare both in the US and in the EU, but European agencies nonetheless appear to be substantially more active than their American counterparts.1 These differences seem to indicate a fundamental divergence in attitude, the source of which is unclear. There is furthermore a suspicion of a certain backwardness of EC competition law in the air: is Community law repeating the early mistakes of US antitrust law by protecting competitors instead of competition? Are the somewhat mysterious theories of ordoliberalism to blame? Is EC competition law too formalistic and too slow to absorb insights from modern economic theory? The scope of divergence between US and EC law and the underlying reasons have been explored repeatedly during the 50 years of coexistence of the two regimes. In a monograph published in 1970—i.e., at a time when no Article 82 case had yet been decided by the ECJ—René Joliet compared the wording and possible meaning of Section 2 Sherman Act and Article 82 and argued that Article 82 was restricted to the pursuit of exploitative abuses, and hence did not extend to exclusionary conduct.2 He explicitly rejected the findings of a similar comparative study by Ernst-Joachim Mestmäcker—an * Professor of Law, European University Institute, Florence. 1 In the US, Section 2 Sherman Act claims are more common in private antitrust litigation. 2 René Joliet, Monopolization and Abuse of Dominant Position. A Comparative Study of the American and European Approaches to the Control of Economic Power, Martinus Nijhoff, 1970, p. 131.

(G) Ehlermann 08 Schweitzer

120

6/11/08

11:48

Page 120

Session One: The Sherman Act And Abuse Of Position Under Art 82

author affiliated with the ordoliberal school—who had argued that Article 82, like Section 2 Sherman Act, was primarily directed against the further restriction by dominant firms of residual competition.3 In 1986, and based on the first important judgments of the European Court of Justice applying Article 82, which had clearly established its applicability to exclusionary abuses, Eleanor Fox compared Article 82 and Section 2 Sherman Act conceptually and found a certain bias in the application of EC competition law which tended to protect the interests of those who deal with dominant firms, rather than protecting the freedom of action of the dominant firms themselves.4 Giuliano Amato identified the requirement in US antitrust law that consumer welfare must be reduced in order to find an abuse as one of the main differences in comparison with EC competition law, which imposes a “special responsibility” on firms in dominant positions to protect small competitors.5 More recent studies frequently focus on specific categories of exclusionary conduct, such as refusals to deal6 and predatory pricing.7 From a quick survey of this more recent literature, a relatively standard story of the essence of the transatlantic divergence and its reasons unfolds. According to this standard narrative, which pervades current scholarly writing on Article 82, the interpretation of Section 2 Sherman Act—initially influenced by political goals, strong popular hostility towards bigness as such and a desire to protect small and medium-sized businesses, of which the Robinson-Patman Act is evidence—has been revolutionized and brought into line with economic theory under the influence of Chicago school scholarship.8 By contrast, Article 82—having repeated mistakes made in the “old” Section 2 jurisprudence in some respects, and having committed its own idiosyncratic mistakes 3 See ibid., at 248 et seq.; and Mestmäcker, “Unternehmenszusammenschlüsse nach Artikel 86 des EWG-Vertrages“, in Ernst von Caemmerer, Hans-Jürgen Schlochauer and Ernst Steindorff, eds., Festschrift für Walter Hallstein, 1966, pp. 322 et seq. 4 Eleanor M. Fox, “Monopolization and Dominance in the United States and the European Community: Efficiency, Opportunity, and Fairness”, 61 Notre Dame Law Review 981, 1017–1018 (1986). 5 Giuliano Amato, Antitrust and the Bounds of Power, Hart Publishing, 1997, pp. 70–71. 6 See, e.g., Eleanor Fox, “Monopolization, Abuse of Dominance, and the Indeterminacy of Economics. The US/EU Divide”, 2006 Utah Law Review 799, 804 et seq. (2006); Christophe Hume and Cyril Ritter, “Refusal to Deal”, College of Europe Working Paper, 6 July 2005; Alexandros Stratakis, “Comparative Analysis of the US and EU Approach and Enforcement of the Essential Facilities Doctrine”, 27 European Competition Law Review 434 (2006). 7 See, e.g., William J. Kolasky, “What is Competition? A Comparison of U.S. and EU Perspectives”, 49 Antitrust Bulletin 29, 46 et seq. (2004); Cyril Ritter, “Does the Law of Predatory Pricing and Cross-Subsidisation Need a Radical Rethink?”, 27(4) World Competition 613 (2004); Brian A. Facey and Roger Ware, “Predatory Pricing in Canada, the United States and Europe: Crouching Tiger or Hidden Dragon?”, 26(4) World Competition 625 (2003). For a general comparative study, see also Reza Dibadj, “Article 82: Gestalt, Myths, Questions”, 23 Santa Clara Computer and High Tech Journal 615 (2007). 8 There is broad recognition that the modern “Harvard school” has likewise influenced the development of US antitrust law. See William E. Kovacic, “The Intellectual DNA of Modern U.S. Competition Law for Dominant Firm Conduct: The Chicago/Harvard Double Helix”, 2007(1) Columbia Business Law Review 1 (2007).

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 121

Principles of Section 2 Sherman Act and Article 82 EC 121 in others—has yet to take that step.9 Like “old” Section 2 Sherman Act jurisprudence, the ECJ has allegedly interpreted Article 82 in such a way as to “protect competitors, not competition”.10 The Court has also allegedly pursued fairness or equity goals instead of efficiency goals.11 In other respects, the case law under Article 82 allegedly displays “typically European” deficiencies, namely a regulatory or interventionist bias.12 The coverage of exploitative abuses and the European approach towards refusals to deal are frequently cited as evidence.13 As to the underlying reasons for the aberrations of European competition law, different explanations are given. The regulatory tendencies of Article 82 are frequently traced back to its wording: instead of adopting a prohibition of monopolization analogous to Section 2 Sherman Act, the drafters of the EC Treaty opted for a mere prohibition of abuse of dominance, and thus generally approved of dominance, but established a regime controlling its exercise.14 There is also a notion that Europe continues to be captivated by a deeply rooted pro-regulatory philosophy which underestimates the ability of markets to self-correct, puts excessive trust in the ability of competition law enforcement institutions to correct market failures, and is concerned more with avoiding “false negatives” than “false positives”. Another reason given for the deficiencies associated with Article 82 is the German ordoliberal influence15—“an approach that has ignored the need for sound economic

9 For the proposition that the interpretation of Article 82 EC is out of line with economic theory see, inter alia, John Temple Lang and Robert O’Donoghue, “Defining Legitimate Competition: How to Clarify Pricing Abuses Under Article 82 EC”, 26 Fordham International Law Journal 84, 84–85 (2002). 10 Richard Whish, Competition Law, 5th ed., Butterworths, 2003, Chap. 1, 2(C), pp. 19–20; Ian S. Forrester, “Article 82: Remedies in Search of Theories?”, 28 Fordham International Law Journal 919, 920 (2005); Stratakis, supra note 6, at 434; Ekaterina Rousseva, “Modernizing by Eradicating: How the Commission’s New Approach to Article 81 EC Dispenses with the Need to Apply Article 82 EC to Vertical Restraints”, 42 Common Market Law Review 587, 592 (2005); Duncan Sinclair, “Abuse of Dominance at a Crossroads—Potential Effect, Object and Appreciability under Article 82”, 25 European Competition Law Review 491, 494 (2004). 11 See, inter alia, Ian S. Forrester, “Article 82: Remedies in Search of Theories?”, cited previous footnote; Michal Gal, “Monopoly pricing as an antitrust offense in the U.S. and the EC: Two systems of belief about monopoly?”, 49 Antitrust Bulletin 343, 363 et seq. (2004). 12 See, e.g., Fox, supra note 4, at 983; Keith N. Hylton, “Section 2 and Article 82: A Comparison of American and European Approaches to Monopolization Law”, Boston University Law Working Paper No. 06-11, available at: http://papers.ssrn.com/sol3/papers. cfm?abstract_id=902655, p. 2. 13 See, e.g., Gal, supra note 11, at 346 (with regard to exploitative abuses). 14 For far-reaching claims regarding the practical relevance of the difference in language between Article 82 and Section 2 Sherman Act, see, inter alia, Joliet, supra note 2, at 9 and 11–12. 15 Whish, supra note 10, at 19–20; James S. Venit, “Article 82: The Last Frontier—Fighting Fire with Fire?”, 28 Fordham International Law Journal 1157 (2005); Stratakis, supra note 6. For the proposition that a special concern with “fair competition” is a characteristic of German ordoliberalism, see, e.g., Pinar Akman, “Searching for the Long-Lost Soul of Art. 82”, CCP Working Paper No. 07-5, available at http://papers.ssrn.com/sol3/papers.cfm?abstract_ id=977221, pp. 20–21.

(G) Ehlermann 08 Schweitzer

122

6/11/08

11:48

Page 122

Session One: The Sherman Act And Abuse Of Position Under Art 82

analysis”16 and which is sometimes also associated with a regulatory attitude towards competition law.17 The “economic freedom” paradigm, rightly attributed to ordoliberal theory, is found to have led European competition law away from the goals of consumer welfare and efficiency.18 The same is said about EU law’s “obsession” with the market integration goal. This paper strives to find out whether this story adequately describes the different attitudes towards rules regarding market power and the underlying reasons. The interest in capturing the “true” roots of divergence is not merely an academic one. The understanding of the reasons of divergence informs the debate on the reform of Article 82. The perceived need to bring Article 82 “in line with sound economics”, to introduce a “more economic” or an “effectsbased” approach,19 as well as the widespread perception that Article 82 is in need of a re-conceptualization similar to the one that the Chicago school introduced in the 1970s and 1980s into US antitrust law are based on the understanding that the current divergence in the application of Article 82 and of Section 2 Sherman Act is due to the use of old-fashioned competition theory in the EU. Implicit is the suggestion that a rational competition policy can only be achieved once the “economic freedom” paradigm is replaced with the consumer welfare goal. Before advocating such far-reaching steps, it appears opportune to ascertain the principles that have guided the Article 82 jurisprudence so far, their legal anchorage, their factual assumptions and the economic thinking they reflect. In order to achieve such an understanding, the paper will proceed in two steps. The first part of the paper analyzes the provisions in a historical perspective (II). The second part deals with the present interpretation of Article 82 and of Section 2 Sherman Act. It will compare some of the doctrines in which the two jurisdictions diverge in an attempt to gain insights into the reasons for the different attitudes (III). The last part summarizes the findings and draws some conclusions which may be relevant to the current reform debate in the EU (IV).

16 Venit, cited previous footnote, at 1158. See also ibid. at 1163: “It is clear from the foregoing that the basic tenets of ordoliberal doctrine do not cite to, nor rely on, any empirical economic evidence or micro-economic theory. Instead, they appear to be based on a philosophy of political or social economy. Nor does ordoliberalism embrace the twin goals of consumer welfare and efficiency that are widely accepted as the prevailing competition law standards.” 17 Frequently with reference to David J. Gerber, Law and Competition in Twentieth Century Europe. Protecting Prometheus, 1998, p. 252, where Gerber describes the concept of “as if” competition—a concept which some members of the ordoliberal school defended, but which was later overwhelmingly rejected even by those associated with the ordoliberal school—see infra note 79 and accompanying text. 18 Venit, supra note 15, at 1163 19 See Neelie Kroes, “Tackling Exclusionary Practices to Avoid Exploitation of Market Power: Some Preliminary Thoughts on the Policy Review of Art. 82”, speech at the Fordham University School of Law, New York, 23 September 2005.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 123

Principles of Section 2 Sherman Act and Article 82 EC 123

II. Section 2 Sherman Act and Article 82 in historical perspective—intellectual roots and shifts The history of Section 2 Sherman Act is marked by changing political attitudes towards market power and corresponding shifts in how the provision is interpreted. In a radical reaction towards an antitrust jurisprudence that had become dominated by political concerns (and in particular by the preoccupation with maintaining a market structure in which small and medium-sized enterprises could survive) and had thus distanced itself from protecting the functioning of competition according to economic principles, Chicago school scholarship led to a revamp of Section 2 doctrine and jurisprudence that is completely oriented towards consumer welfare and efficiency. The interpretation of Article 82 so far has not undergone equally distinctive shifts. Different political attitudes towards market power were present at the drafting stage; however, the final wording of Article 82 clearly reflected the “abuse” theory. This foreclosed moves in the direction of a “no fault” prohibition of monopolization along the lines of the judgment of the Second Circuit in Alcoa.20 The ECJ’s interpretation of Article 82 has, ever since the early judgments, been driven not by anti-bigness concerns, nor by “pure” efficiency concerns, but by its function within the broader system of the EC Treaty as articulated in Article 3(1)(g):21 Article 82 is one of the pillars of a “system ensuring that competition in the internal market is not distorted”, and from which efficiency, consumer welfare and economic progress is expected to result. Hence, the provision has been interpreted in light of the goal of market integration with the aim of ensuring the effectiveness of the fundamental freedoms (especially the free circulation of goods and services across borders within the Community) against the exercise of private power to preclude market access or to eliminate competitors. Article 82 has been, and continues to be, indissolubly intertwined with the EC’s internal market project—a primary feature distinguishing it from Section 2 Sherman Act. The fact that the evolution of the jurisprudence in relation to Article 82 has, until recently, been comparatively consistent as to its fundamental orientations is not to say that there are no normative uncertainties. Developing an adequate test for distinguishing between pro- and anticompetitive conduct on

20

United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir. 1945). See Case 6/72, Europemballage and Continental Can v Commission, [1973] ECR 215, para. 23. For the importance of this functional interpretation, see Thomas Eilmansberger, “How to distinguish good from bad competition under Article 82 EC: In search of clearer and more coherent standards for anticompetitive abuses”, 42 Common Market Law Review 129, 132 (2005). 21

(G) Ehlermann 08 Schweitzer

124

6/11/08

11:48

Page 124

Session One: The Sherman Act And Abuse Of Position Under Art 82

the part of dominant firms is a particularly difficult task,22 and EU law and US law share the same uncertainties in this respect. The evolution of Article 82 and Section 2 cannot be traced in detail here. However, a short comparative look at the antitrust history of the US and of the EU is necessary for an understanding of the different rules and attitudes. The history of Section 2 and its interpretation is well researched23 and will be summarized only briefly here (point 1 below). The history of Article 82, on the other hand, deserves a closer look. Strictly speaking, the drafting history of the EC Treaty is not relevant to the application of this provision. The signing of the Treaty of Rome and the establishment of the EC as a supranational entity was accompanied by a deliberate decision of the Member States to commit to a functional interpretation, and not to revert to “original intent”. It was thought that the EC should be developed dynamically with a view to the implementation of its own goals, and not with a view to the positions taken by the Member States during the negotiations. Indeed, in order to avoid interpretations based on “original intent”, the official records of the drafting process were not published.24 Faithful to this fundamental decision, the ECJ has never referred to the drafting history in interpreting the EC Treaty. Nevertheless, it is interesting to examine the drafting history with a view to testing the frequent claims being made about the fairness concerns, the regulatory tendencies, and the aim to protect competitors instead of competition allegedly enshrined in Article 82 (point 2 below). Following this backwards glance, the discussion will turn to a question that is legally more relevant, and also more telling with respect to the attitudes that underlie the rules on market power today, namely, the interpretation of Article 82 by the ECJ during the formative years of EC competition law (point 3).

1. Section 2 Sherman Act in historical perspective Section 2 Sherman Act prohibits monopolization and attempted monopolization, i.e., acts that change or entrench the structure of the market in a way that is undesirable from a competition law point of view. No attempt is made 22 See David Evans and Jorge Padilla, “Designing Antitrust Rules for Assessing Unilateral Practices: A Neo-Chicago Approach”, 72 University of Chicago Law Review 73 (2005): “. . . the welfare effects of unilateral practices are inherently difficult to assess”. See also Antitrust Modernization Commission, Report and Recommendations, April 2007, available at: http://www.amc.gov/report_recommendation/amc_final_report.pdf, at p. 81. In search of a test for the US, see, inter alia, Einer Elhauge, “Defining better monopolization standards”, 56 Stanford Law Review 254 (2003). For a recent critical appraisal of EU law see, inter alia, Eilmansberger, cited previous footnote, at 129 et seq. 23 See, e.g., Rudolph J.R. Peritz, Competition Policy in America. History, Rhetoric, Law, Oxford University Press, 1996. For the early US antitrust history, see also Hans Thorelli, The Federal Antitrust Policy: Origination of an American Tradition, Johns Hopkins Press, 1954. 24 See Gerber, supra note 17, at 343.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 125

Principles of Section 2 Sherman Act and Article 82 EC 125 to control the mere exercise of power vis-à-vis consumers, i.e., to address exploitative abuses. Despite the structural focus of Section 2, the provision does not prohibit the existence of monopolies per se. Rather, it prohibits certain types of conduct that create or threaten to create monopoly. The early history of Section 2 jurisprudence revolves to a significant extent around the different meaning and weight given to the structural and the conduct element at different times.25 As has frequently been recounted, the enactment of the Sherman Act was motivated by the economic conditions and sentiments of the times.26 In the words of Chief Justice White in Standard Oil, these were: “the vast accumulation of wealth in the hands of corporations and individuals, the enormous development of corporate organization, the facility for combination which such organizations afforded, the fact that the facility was being used, and that combinations known as trusts were being multiplied, and the widespread impression that this power had been and would be exerted to oppress individuals and injure the public”.27

The accumulation of wealth in the hands of a few was perceived as a threat not only to the economic order, but also to democracy. With both the political dimension and the economic implications of antitrust law in mind, courts struggled with the interpretation of Section 2, oscillating between more structure- and more conduct-oriented approaches (“abuse theories”). Bemoaning the uncertainties surrounding Section 2, Levi wrote in 1947: “We are not sure whether we are against monopolies or the abuse of monopoly . . . We do not know whether we are opposed to size or merely to unreasonable high prices.”28 The famous Alcoa case mentioned earlier marks an apex of the structuralist approach in the interpretation of Section 2. In this case, the Second Circuit found an offence mainly based on the fact that Alcoa held, and had managed to maintain, an overwhelming market share. Although the Alcoa judgment 25 In United States v. United Shoe Machinery Corp., 110 F. Supp. 295, 342 (D. Mass. 1953), Judge Wyzanski distinguished three different approaches towards Section 2: (1) an enterprise has monopolized if it has acquired or maintained a power to exclude others as a result of using an unreasonable “restraint of trade” in violation of Section 1; (2) a monopolization offence is committed where an undertaking with effective market control uses this control, or plans to use it, to engage in exclusionary practices, even if these are not technically “restraints of trade”; (3) the acquisition of an overwhelming market share is a monopolization under Section 2 even if there is no showing of any exclusionary conduct. But the defendant escapes liability if it can show that he owes his monopoly power to legitimate causes (superior skill, business acumen, etc.). The Chicago school added a fourth approach, based on the assumption that “exclusionary conduct” is normally not viable. 26 See, e.g., Lawrence A. Sullivan and Warren S. Grimes, The Law of Antitrust: An Integrated Handbook, 2nd ed., Thomson West, 2006, pp. 4–7. 27 Standard Oil of New Jersey v. United States, 221 U.S. 1, 50 (1911). 28 Edward Levi, “The Antitrust Laws and Monopoly”, 14 University of Chicago Law Review 153 (1947).

(G) Ehlermann 08 Schweitzer

126

6/11/08

11:48

Page 126

Session One: The Sherman Act And Abuse Of Position Under Art 82

stopped short of establishing a per se prohibition of monopoly power, it did not require much in terms of exclusionary conduct or of specific intent to monopolize.29 The simple pursuit of normal business practices without a predatory tendency but with the effect of defending the dominant firm’s superior market position could apparently suffice for finding a violation of Section 2. Based on the assumption that the Sherman Act’s aim was “to perpetuate and preserve, for its own sake and in spite of possible cost, an organization of industry in small units which can effectively compete with each other”,30 and that it would be absurd to condemn price fixing contracts in Section 1 without extending this condemnation to monopolies who necessarily fix the market price when they sell, Judge Learned Hand essentially held that the active seeking of monopoly power, even if by means of perfectly legitimate business conduct, could be qualified as illegal monopolization under Section 2.31 This far-reaching interpretation of Section 2 was critically discussed in the US antitrust community. According to a widespread opinion, it took the structural elements of the prohibition too far. It was considered that a more intense inquiry into intent and into the competitive legitimacy of the methods employed to acquire or maintain a monopoly should be required to establish illegal monopolization.32 The debate about the need to distinguish legitimate from illegal practices, and thus to flesh out the features of illicit exclusionary conduct, intensified. When Chicago school scholars undertook their comprehensive reconceptualization of US antitrust law, the Alcoa position towards Section 2 was thus already on a slow retreat, although it was still good law. The case became a favourite target of attack for the Chicago school. According to 29 Alcoa, supra note 20, at 432: “[I]n order to fall within Section 2 the monopolist must have both the power to monopolize, and the intent to monopolize”. But to require a more specific intent would cripple the Act “for no monopolist is unconscious of what he is doing”. 30 Ibid. 31 Judge Hand does stress that, “[s]ince the [Sherman] Act makes ‘monopolization’ a crime, as well as a civil wrong, it would be not only unfair, but presumably contrary to the intent of Congress” to include instances in which a firm has become a monopolist by force of accident. The decision also states that “the Act does not mean to condemn the resultant of those very forces which it is its prime object to foster: finis opus coronat. The successful competitor having been urged to compete, must not be turned upon when he wins”. On the other hand, Alcoa was found to have violated Section 2 on the following grounds: “It was not inevitable that [Alcoa] should always anticipate increases in the demand for ingot and be prepared to supply them. Nothing compelled it to keep doubling and redoubling its capacity before others entered the field. It insists that it never excluded competitors; but we can think of no more effective exclusion than progressively to embrace each new opportunity as it opened, and to face every newcomer with new capacity already geared into a great organization, having the advantage of experience, trade connections and the elite of personnel. Only in case we interpret ‘exclusion’ as limited to manoeuvres not honestly industrial, but actuated solely by a desire to prevent competition, can such a course, indefatigably pursued, be deemed not ‘exclusionary’. So to limit it would in our judgment emasculate the Act”. It therefore seemed that a defence according to which a monopoly was “thrust upon” a firm would be limited to cases where the achievement of power was not deliberate, but due to circumstances outside the free choice of the firm. 32 See, e.g., United Shoe Machinery, supra note 25.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 127

Principles of Section 2 Sherman Act and Article 82 EC 127 Robert H. Bork, Alcoa stood for the proposition that “monopoly . . . is illegal unless the monopolist could not avoid it. Superior efficiency is not only no excuse, it is an ‘abuse’.”33 Indeed, with Alcoa, Section 2 jurisprudence appeared to have reached a point that left firms with monopoly power little room to compete. This, as well as the Robinson-Patman Act, which made protecting competitors an explicit goal of the law,34 were obvious targets of critique. However, the Chicago school’s reform project had more fundamental ambitions: consumer welfare was to be established as the only ultimate goal of antitrust law,35 and price theory was to be the method by which to predict consumer welfare effects.36 Applied to unilateral conduct, this translated into a highly permissive approach:37 firstly, most unilateral practices would, according to Chicago school scholars, typically create efficiencies. Secondly, firms with monopoly power would lack incentives to engage in welfarereducing practices. Most unilateral practices should therefore be lawful per se. Although Chicago school scholarship was never adopted into antitrust doctrine wholesale,38 it did gain significant influence in the courts, particularly during the 1970s and 1980s. The empirical claim that, due to a lack of rational incentives or of ability, instances of (successful) exclusionary conduct will be rare have led to the development of rather stringent tests for finding an infringement of Section 2. No less importantly, the claim that consumer welfare is the only goal, or at least the primary goal of antitrust law, and that it should directly guide the legal appraisal of single-firm conduct, has had strong resonance in the academic community as well as in the courts. During the last decade or so, Chicago school thinking has lost some of its influence. “Post-Chicago” scholarship, while subscribing to efficiency as the ultimate objective of antitrust law, has started to challenge the overly simplifying assumptions on which Chicago school theory was based39 and has uncovered conditions under which unilateral conduct can indeed have anti-

33 Robert H. Bork, The Antitrust Paradox: A Policy at War with Itself, Basic Books. 1978, p. 170. 34 For a critical appraisal of the Robinson-Patman Act, see Mestmäcker, Der verwaltete Wettbewerb, Mohr Siebeck, 1984, pp. 45–47. 35 For the broad agreement in today’s US antitrust scholarship that consumer welfare is antitrust’s ultimate purpose, see Herbert Hovenkamp, The Antitrust Enterprise: Principle and Execution, Harvard University Press, 2005, p. 31. 36 See Richard A. Posner, “The Chicago School of Antitrust Analysis”, 127 University of Pennsylvania Law Review 925, 928 (1979). 37 See ibid., summarizing the conclusions of Chicago School scholarship on unilateral conduct as follows: “firms cannot in general obtain or enhance monopoly power by unilateral action— unless, of course, they are irrationally willing to trade profits for position. Consequently, the focus of antitrust laws should not be on unilateral action . . .”. 38 See Hovenkamp, supra note 35, at 37 and Kovacic, supra note 8, both emphasizing the continuing influence of Harvard school scholarship. 39 For a summary of the criticism, see Hovenkamp, supra note 35, at 34–35.

(G) Ehlermann 08 Schweitzer

128

6/11/08

11:48

Page 128

Session One: The Sherman Act And Abuse Of Position Under Art 82

competitive effects.40 The opportunities and incentives for strategic behaviour and exclusionary conduct, largely denied by traditional Chicago school scholarship, are generally acknowledged today.41 This evolution coincides with a widespread recognition that many of the tests currently applied under Section 2 to identify illegal monopolization tend to be under-inclusive conceptually42 and can create a significant number of “false negatives”. Nonetheless, these tests continue to enjoy widespread support by leading scholars of various schools, including the modern Harvard school. Taking account of certain distinctive features of the US system of private antitrust enforcement such as, inter alia, treble damages and the involvement of potentially error-prone juries,43 these scholars argue that broader and less underinclusive tests could create excessive incentives to sue.

2. The history of Article 82 EC a) The drafting history A confrontation with the diversity of ideas and projections, the utter uncertainty about the future role of competition and competition rules in the EU, and the “veil of ignorance” under which the negotiations of the Treaty of Rome took place is fascinating for anyone used to dealing with today’s so well-established system of rules. After plans for the political integration of continental Europe had failed in 1954, the creation of a common market was at the centre of the European project: the common market was to foster economic growth and stability, raise living standards, and most of all to ensure harmonious and peaceful relations between the Member States.44 From the outset, the negotiating parties agreed that the common market was to be based on free movement rules to prevent the Member States from impeding or distorting cross-border trade; but these free movement rules would need to be backed up by competition rules so that the state barriers to trade would not be replaced by private restraints. 40 See, e.g., Thomas G. Krattenmaker and Steven C. Salop, “Anticompetitive Exclusion: Raising Rivals’ Costs to Achieve Power over Price”, 96 Yale Law Journal 209 (1986); Jonathan B. Baker, “Recent Developments in Economics that Challenge Chicago School Views”, 58 Antitrust Law Journal 645 (1989). For an overview, see Hovenkamp, “Antitrust Policy after Chicago”, 84 Michigan Law Review 213 (1985); Hovenkamp, “Post-Chicago Antitrust: A Review and Critique”, 2001(2) Columbia Business Law Review 257 (2001). 41 See Richard Posner, Antitrust Law, 2nd ed., University of Chicago Press, 2001, p. 194. 42 See, e.g., Hovenkamp, supra note 35, at 45 et seq.; Kovacic, supra note 8, at 53, 63–64 and 74 et seq. 43 For discussion, see Kovacic, supra note 8, at 51 et seq.; Hovenkamp, supra note 35, at 47. 44 See Intergovernmental Committee of the Messina Conference, Report by the Heads of Delegations to the Foreign Ministers (“Spaak Report”), 21st April 1956, Title II Chap. 1— Competition Rules, p. 57.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 129

Principles of Section 2 Sherman Act and Article 82 EC 129 This idea was fully developed already in the so-called Spaak Report of 1956.45 It addressed both the problem of cartels and of positions of market power in a section entitled “monopolies”. The common market, according to the Report, would create opportunities for companies to grow and realize economies of scale without endangering competition.46 However, for the potential gains to be realized, the future Treaty would need to contain provisions ensuring that existing monopolies or abusive practices would not frustrate the common market goal.47 While the Spaak Report did not propose an exact formulation of the future competition rules, it did anticipate much of their content and structure.48 The competition rules were to be directly applicable in the Member States, and would enjoy primacy over national law. They should be interpreted by the Commission and be further developed over time by a European court.49 The Spaak Report by no means summarized a consensus, but it was the basis on which the future Member States entered into the negotiations of the Treaty of Rome.50 For the drafting of the competition rules it was an important reference point51—possibly more important than the ECSC Treaty’s competition rules.52 The design of the Treaty of Rome’s competition rules was initially controversial among the negotiating parties,53 as was the economic order of the future European Community in general. The German and the French positions marked the two poles of the debate. The German delegation envisaged a common market based on principles constitutive of a market economy.54 The French delegation, while subscribing to the idea of a market economy of some 45 See Spaak Report, Title II Chap. 1 b)—Monopolies, pp. 59–60. The Spaak Report was prepared by Paul-Henri Spaak (presiding), Carl Friedrich Ophüls (Germany), Felix Gaillard (France), Ludovico Benvenuti (Italy), Baron Snoy (Belgium), Lindhorst Homann (Netherlands) and Lambert Schaus (Luxemburg), i.e., the heads of the national delegations to the Messina conference, who were, however, in preparing this report, not bound by instructions of the Member States. A draft of the Report was prepared by Pierre Uri and Hans von der Groeben. For an insider’s perspective on the drafting of the Spaak Report, see Hans von der Groeben, Deutschland und Europa in einem unruhigen Jahrhundert, Nomos, 1995, pp. 269 et seq., and particularly pp. 276–279. 46 See Spaak Report, supra note 44, at 8. 47 Ibid., pp. 59–60 48 Apart from rules on cartels and undertakings with market power, the Spaak Report also envisaged a system of merger control. See ibid. at 60. 49 Ibid. 50 See von der Groeben, supra note 45, at 279. 51 For the relevance of the Spaak Report to the drafting of the EC Treaty’s competition rules, see Dokument 53, Vermerk über die Sitzung der Arbeitsgruppe “Gemeinsamer Markt” in Brüssel vom 3.9.-5.9.1956, 7.9.1956, in Reiner Schulze and Thomas Hoeren, eds., Dokumente zum Europäischen Recht, Band 3: Kartellrecht (bis 1957), Springer, 2000, pp. 163–164. 52 The ECSC Treaty encompassed a prohibition of cartels, an abuse control for undertakings with market power, and a system of merger control. 53 For a report on the negotiations, see von der Groeben, supra note 45, at 280–289. 54 A number of members of the German delegation were to some extent affiliated with the ordoliberal school. As mentioned earlier (supra note 45), von der Groeben was involved in the drafting of the Spaak Report. He also presided over the “Common Market” Working Group during the negotiations of the Treaty of Rome itself. Alfred Müller-Armack—Secretary of State

(G) Ehlermann 08 Schweitzer

130

6/11/08

11:48

Page 130

Session One: The Sherman Act And Abuse Of Position Under Art 82

sort, favoured a more dirigiste approach, with vaster potential for intervention and economic planning by the Member States.55 These positions were reflected in the delegations’ propositions for how to frame and enforce the Treaty’s competition rules. The French proposal56 started out with a broad and non-specific prohibition of discrimination: all firms should be required to treat competing buyers or sellers equally with regard to both price and conditions of trade.57 Secondly, it featured a general prohibition of cartels, monopolies, and abusive practices directed towards, or potentially resulting in, an impediment to competition. Examples listed included price fixing, the restriction or control of production, technological development or investment, the partitioning of markets, and the enabling of a total or partial domination of markets for certain products by a firm or a group of firms. No less general than the prohibition was the exception foreseen: it took the form of a broad efficiency defence. Practices should be exempted from the prohibition where they contributed to the improvement of production or distribution or to fostering technological and economic progress. These rules were not meant to be directly applicable, but were to be implemented by the Member States. Where trade between two or more Member States was affected, Member States could request the involvement of the Commission in a consultation procedure. Ultimately, the Council would decide. According to this proposal, state monopolies and services publics should not be subject to these general rules, but rather should be governed by special rules which the proposal did not specify.58 A Belgian-Dutch proposal was similar to the French proposal in that it treated anticompetitive agreements and dominant positions under one single rule, but it was envisaged that both would be subject to an abuse control, no more. On the other hand, the Commission—and not the Member States—should be in charge of this control.59 The German proposal differed significantly: it contained separate rules for anticompetitive agreements and market dominance. Cartel agreements should generally be prohibited, subject to a narrow exception. With regard to in the cabinet of Economics Minister Ludwig Erhard and author of the term “social market economy”—was a member of the German delegation that negotiated the Common Market rules. 55 See Hanns Jürgen Küsters, Die Gründung der Europäischen Wirtschaftsgemeinschaft, Nomos, 1982, p. 364 with further references. 56 See Document 51 in Schulze and Hoeren, supra note 51, Vorschlag der französischen Delegation, 4.9.1956, p. 158. 57 See ibid., p. 157, Article X: “Innerhalb des gemeinsamen Marktes sind verboten: Preiserhöhungen und—senkungen, sowie Änderungen von Verkaufsbedingungen für vergleichbare Geschäfte gegenüber Käufern oder Verkäufern, die miteinander im Wettbewerb stehen.” The German delegation proposed to abandon this general non-discrimination principle and to include a more limited prohibition in the rules on abuse of dominant positions. See Dokument 53, in Schulze and Hoeren, supra note 51, at 163. 58 See Document 57, Entwurf eines Protokolls über die Sitzungen der Arbeitsgruppe vom 3.5.9.1956 in Brüssel, 10.9.1956, in Schulze and Hoeren, supra note 51, at 172. 59 See Document 59, Synoptische Darstellung der Artikelentwürfe über die Wettbewerbsregeln für Unternehmen, in Schulze and Hoeren, supra note 51.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 131

Principles of Section 2 Sherman Act and Article 82 EC 131 monopolies and oligopolies, the German delegation proposed to prohibit merely the abuse of dominance. According to the proposal, this prohibition should be applicable to private and state undertakings and state monopolies alike.60 No rules on the enforcement of the EC competition rules were proposed. Such rules were to be established in a separate Treaty to be concluded between the Member States within a period of 2 years. If no agreement were reached within this time period, the Commission should enact an enforcement regulation with the approval of 2/3 of the Council. The controversial question of whether primary responsibility for enforcement of the competition rules should reside with the Commission or with the Member States,61 as well as the question of direct applicability, was thus to be decided at a later point of time. The negotiations mainly revolved around: the non-discrimination principle which the French delegation had proposed; the question of whether the competition rules should apply only to private undertakings or also to state undertakings and services publics; and whether to have one single rule or separate rules for cartels and dominant firms. The broad non-discrimination principle was highly controversial. For the French delegation, it was an essential pre-condition to the implementation of the competition rules.62 This idea was a variation on a more general notion in France that the competition rules should apply only subject to the prior establishment of a “level playing field”, where all “distortions” of competition, such as different working conditions, wages, social burdens, tax systems, etc., had previously been equalized.63 This notion had in fact already been discussed and rejected in the Spaak Report.64 In the negotiations leading to the Treaty of Rome, the idea of a broadly construed prohibition of discrimination as part of the competition rules was most strongly opposed by the 60

See Document 57, in Schulze and Hoeren, supra note 51, at 172. The German and the French delegations generally favoured a competence of the Member States for the enforcement of the competition rules. The Italian delegation favoured a Community competence. See ibid. at 170 and 173. 62 See Document 57, Entwurf eines Protokolls über die Sitzungen der Arbeitsgruppe vom 3.-5.9.1956 in Brüssel, 10.9.1956, in Schulze and Hoeren, supra note 51, at 171. 63 For a reaction to this position, see Eberhard Günther, “Das Kartellproblem in internationaler Beleuchtung”, 19.7.1956, in Schulze and Hoeren, supra note 51, Document 50, p. 155: “Dabei wird verkannt, dass es nicht Aufgabe der Mitglieder eines Gemeinsamen Marktes sein kann, die unterschiedlichen Produktionskosten der Betriebe ganz oder auch nur zum Teil einander anzugleichen, und dass es eine Verfälschung des Zieles der Herstellung eines Gemeinsamen Marktes bedeuten würde, wenn man die Produktionskosten oder Teile von ihnen durch staatliche Maßnahmen angleichen wollte. Es soll gerade der Sinn des Gemeinsamen Marktes sein, unterschiedliche Produktionskosten von Unternehmen im Qualitäts- und Preiswettbewerb dem Verbraucher zugute kommen zu lassen.” 64 Spaak Report, Chapter 2 Section 1: Distortions, pp. 64–65: “Es herrscht vielfach die Auffassung, ein wirklicher Wettbewerb sei erst dann möglich, wenn die Hauptfaktoren der Gestehungskosten überall einander angenähert worden sind. Gerade auf der Grundlage gewisser Unterschiede kann sich aber ein Gleichgewicht bilden und der Handel entwickeln. Dies gilt z.B. für die Unterschiede im Lohnniveau, wenn sie Unterschieden in der Produktivität entsprechen. . . .”. 61

(G) Ehlermann 08 Schweitzer

132

6/11/08

11:48

Page 132

Session One: The Sherman Act And Abuse Of Position Under Art 82

German delegation. Alfred Müller-Armack stressed the important function of price discrimination in competition and warned against incorporating regulatory tendencies into the future competition rules.65 Ultimately, the idea of a general non-discrimination principle was abandoned. Instead, a general prohibition of discrimination on the basis of nationality was made part of the introductory Treaty norms. This left open the question whether it would apply only to Member States or also to private firms. A compromise was also found on the question of whether the competition rules should apply to private undertakings only, or whether they should also apply to state undertakings and services publics: the competition rules were formulated in a broad and general way. Article 86(1) EC (Article 90(1) EEC) clarified that they would apply to state undertakings as well. However, Article 86(2) EC (Article 90(2) EEC) provided for a possible exception where an undertaking has been entrusted by public authorities with “services of general economic interest”—an exception the scope of which was anything but clear at the time of the signing of the Treaty of Rome. On the last question, i.e., whether to have a single, comprehensive prohibition for cartels and abuses of dominant positions with a generally applicable exception to both or rather two separate rules, the German delegation ultimately prevailed. In fact, the competition rules as ultimately drafted, and particularly Article 82 EC (Article 86 EEC), came closest to the original German proposals, although those proposals by no means survived unchanged. The German influence on the shape of the EC competition rules likely resulted from the particular importance which the German delegation attached to them—not only, and maybe not even primarily with a view to the impact they would have in shaping the future common market, but rather against the background of a parallel, internal German debate on a national competition law (the GWB) which was finally adopted in July of 1957, several months after the Treaty of Rome was signed.66 Throughout the negotiations on the Treaty, Ludwig Erhard, the German Minister of Economic Affairs, was concerned that if the European competition rules deviated too much from the German competition rules he intended to enact, they would torpedo his attempt to secure the adoption of an effective German competition law 65 See Document 55, Darlegungen des Sprechers der deutschen Delegation zu den Entwürfen der Artikel 40–43, 8.9.1957, in Schulze and Hoeren, supra note 51, p. 167, no. 3: “Die sogen. Diskriminierung oder die unterschiedliche Preisstellung gehört zu den legalen Formen des Wettbewerbs. Würde man die unterschiedliche Behandlung von Abnehmern oder Lieferanten verbieten, so würde damit eine Regel zur Behinderung und nicht zur Förderung des Wettbewerbs geschaffen. Auf vielen Märkten, namentlich den oligopolistisch strukturierten, ist die unterschiedliche Behandlung häufig die Vorstufe einer allgemeinen Preisänderung; sie hält auf diese Weise die Preisstellung elastisch. Das Verbot der unterschiedlichen Behandlung kommt einer staatlichen Aufforderung zur Kartellbildung sehr nahe. Zu untersagen wäre lediglich die mit der Schädigungsabsicht verbundene unterschiedliche Behandlung von Käufern oder Lieferanten durch Kartelle oder marktbeherrschende Unternehmen.” 66 For the relevance of this internal German debate to the negotiations in Brussels, see Documents 62 and 66 in Schulze and Hoeren, supra note 51, at 195 and 204–205.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 133

Principles of Section 2 Sherman Act and Article 82 EC 133 against the intense opposition of German industry. No other delegation appears to have given similar weight to the exact shape of the EC competition rules—particularly since the question of their enforcement was left open. Against the background of a generally prevailing pro-competition attitude67 in the Working Group for the Common Market, which was entrusted with the drafting of the competition rules, the proposal ultimately presented by the group’s chairman, Hans von der Groeben,68 was in the end approved. There were, however, a number of points on which the German delegation did not prevail. Article 81(3) EC (Article 85(3) EEC) provided for much broader exceptions to the general prohibition of cartels than the German side had proposed. According to Müller-Armack, this turned Article 81 into a hybrid between a prohibition principle and an abuse principle.69 Also, the EC Treaty did not contain rules on merger control—an important component of a full-fledged system of competition rules in the eyes of proponents of the ordoliberal school. Does the German influence on the drafting of the competition rules, and particularly on Article 82 EC, support the claim that Article 82 EC is a creature of ordoliberal theory? In fact, the degree of congruence between Article 82 EC and ordoliberal positions is difficult to determine. No fully developed ordoliberal position on the treatment of market dominance existed at the time the EC Treaty was negotiated. Against the background of a heavily cartelized German industry, the main concern of German ordoliberals had been how to deal with cartels. Some thought had been given to the problem of dominance, of course. In the context of the German competition law debate, economist Walter Eucken had proposed that, wherever possible, monopolies should be prohibited per se.70 Those monopolies that were technologically or economically 67 The French delegation was somewhat divided, but important members of the delegation, namely (future Commissioner) Robert Marjolin, Jacques Donnedieu de Vabres and JeanFrançois Deniau, were generally favourable to strengthening competition as a means to increase the performance of French industry. Similarly, leading economic circles in Italy at the time favoured a liberal market regime with strong competition rules at EC level and saw the Treaty as an opportunity to create a level competitive playing field in Europe. See Küsters, supra note 55, at 364–366. 68 The competition provisions as they appear today in the EC Treaty were based on a draft presented by H. Thiesing. See Document 56, Entwurf zu den Artikeln 42, 42a-d, vorgelegt von H. Thiesing, 10.9.1957, in Schulze and Hoeren, supra note 51, at 168 et seq. The draft still contained a separate prohibition of discrimination of buyers or sellers which are in competition with one another based on their nationality (see Article 42 of the draft). However, the cartel prohibition (Article 42a) and the prohibition of abuses of a dominant position (Article 42b) come close to the final version that was finally adopted in the Treaty. 69 The German delegation, and in particular Müller-Armack, was of the opinion that an excessive number of exceptions had been integrated into Article 81(3) (Article 40(2) of the draft), and that this in effect led to a mixing of the “prohibition” principle and the “abuse” system. See Document 73 in Schulze and Hoeren, supra note 51, at 228. 70 Walter Eucken, “Überlegungen zum Monopolproblem“, in Wirtschaftsmacht und Wirtschaftsordnung, Walter Eucken Archiv (LIT), 2001, pp. 79 and 83. This idea was influential in the ordoliberal group during and immediately after the 2nd World War, under the impression of the significant contributions of dominant German firms to the rise of the Nazi regime and the

(G) Ehlermann 08 Schweitzer

134

6/11/08

11:48

Page 134

Session One: The Sherman Act And Abuse Of Position Under Art 82

unavoidable, i.e., natural monopolies, were to be placed under regulatory supervision and required to act “as if” there were competition in the market. Eucken thus adopted this (in)famous concept of “as if” competition for the narrow case of regulating infrastructure monopolies,71 borrowing the concept from Leonard Miksch,72 who advocated its application on a broader scale. Yet “as if” competition, which many today associate with ordoliberalism generally,73 was not a proposition uniformly accepted by ordoliberals. It was one of the concepts discussed in ordoliberal circles after the 2nd World War, and it did appear in the “Josten” draft of 1949,74 which launched a long legislative procedure leading ultimately to the GWB (see above). But ”as if” competition was already abandoned in a draft of the GWB that was presented to the German Parliament by Franz Böhm and others in 1953.75 This draft proposed to place dominant firms under “supervision” (§ 10);76 however, contrary to first impression, the draft did not envisage a full-fledged regulatory scheme but rather a selective control of specific abuses of dominant positions. Although the 1953 war economy. See, e.g., the “Entwurf eines Gesetzes zur Sicherung des Leistungswettbewerbs” (the so-called “Josten draft”) of 5.7.1949, in the preparation of which Franz Böhm, one of the leading ordoliberals, had participated. This draft envisaged the elimination of all positions of dominance, wherever possible, if necessary by way of breaking firms up (see § 15 of the draft). Undertakings which had achieved their dominant position by way of competition on the merits were, however, exempted from this rule—see Begründung zu § 3, p. 38 of the draft: “Auch echter Leistungswettbewerb kann für Spitzenunternehmen zu Sonderstellungen im Markt führen. Sie sind jedoch dadurch von Machtstellungen im Sinne des Gesetzes unterschieden, daß sie ihrer Natur nach nur vorübergehender Art sind und gegenüber anstürmendem Wettbewerb täglich neu erworben werden müssen. Es ist geradezu der Sinn des Leistungswettbewerbs, dem technischen Fortschritt und der Gütesteigerung zu dienen, den auf diesem Gebiet erfolgreichen Unternehmen die Möglichkeit zu einer wirtschaftlichen Besserstellung zu geben und auf diese Weise die unternehmerische Initaitive anzuregen. Es entspricht daher nur dem Zweck des Gesetzes, den Wettbewerb anzuregen und den Fortschritt zu fördern, wenn Sonderstellungen dieser Art, wie sie aus Pionierleistungen, einem Leistungsvorsprung anderer Art und Liebhaberleistungen erwachsen, von den Vorschriften dieses Gesetzes ausdrücklich freigestellt werden.” 71 Walter Eucken, Grundsätze der Wirtschaftspolitik, Mohr Siebeck, 7. unveränderte Aufl. 2004, p. 295: “Ziel der Monopolgesetzgebung und der Monopolaufsicht ist es, die Träger wirtschaftlicher Macht zu einem Verhalten zu veranlassen, als ob vollständige Konkurrenz bestünde. Das Verhalten der Monopolisten hat ‚wettbewerbsanalog’ zu sein.” 72 Leonard Miksch, Wettbewerb als Aufgabe. Grundsätze einer Wettbewerbsordnung, 2. Aufl. 1947, pp. 98–99. 73 See, e.g., Gerber, supra note 17, pp. 252–253. 74 Section 22 “Josten-Entwurf“ (“Verhalten im Markt”): “Inhaber wirtschaftlicher Macht sollen sich im Geschäftsverkehr so verhalten, wie sie sich verhalten würden, wenn sie einem wirksamen Wettbewerb ausgesetzt wären.” However, as pointed out above (supra note 70), those who had achieved a position of dominance by competition on the merits were exempted from this regulatory scheme. 75 Antrag der Abgeordneten Dr. Böhm, Dr. Dresbach, Ruf und Genossen: Entwurf eines Gesetzes gegen Wettbewerbsbeschränkungen, BT-Drs. II/1269. 76 Unlike Article 82, the 1953 draft did not entail a general prohibition of abuses of dominance, but merely placed dominant firms under the supervision of the cartel authority. The idea of a per se prohibition of dominance had been abandoned. Böhm and others acknowledged the important incentive function that competition for a superior market position can have. Also, the idea of breaking up dominant positions was politically unacceptable in Germany at the time, and perceived to be a continuation of Siegerjustiz (“victor’s justice”).

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 135

Principles of Section 2 Sherman Act and Article 82 EC 135 draft addressed exploitative abuses,77 the focus was clearly on preventing exclusionary abuses. Tying practices, predatory pricing, abusive discrimination and refusals to deal in essential facility settings were specifically addressed. Like the 1953 draft, the official draft for a GWB presented by the government also distanced itself explicitly from the concept of “as if” competition.78 In the competition policy debate, it continued for a while to be advocated by some. However, certain scholars associated with the ordoliberal school, and in particular Ernst-Joachim Mestmäcker, were among the most outspoken critics of the concept of “as if” competition,79 and they were influential in ensuring that it never became part of German competition law. All in all, the question of how best to deal with positions of market power was generally an open one as the negotiations on the Treaty of Rome took place. There were no hard and fast answers, and no clear “role model”. The debate which preceded and accompanied the drafting of Article 82 during the negotiation of the Treaty of Rome is evidence of this general uncertainty. The negotiating parties were certainly aware of the existence of Section 2 Sherman Act, but a prohibition of monopolization analogous to the US model was not explicitly discussed as a potential option for the EC Treaty. The possibility of a per se prohibition of monopoly positions appeared in the Spaak Report and was—subject to broad exceptions—taken up by the French delegation; but it was unacceptable to the German delegation. The general perception in Germany was that market dominance could legitimately be obtained by competition on the merits, that competition for a superior market position could have important incentive effects, and that dominant positions should therefore not be generally outlawed or dismantled. The Spaak Report had already suggested the idea that, if firms were kept from abusing their power, the opening up of national markets would dismantle monopoly positions while at the same time allowing firms to grow. It was probably against this background that the German proposal of a prohibition of the abuse of market dominance was able to gather support. A further, unspoken reason for the German opposition to a per se prohibition of dominance may have been the political unacceptability of a competition law that would threaten to break up leading German companies. There is no evidence in the documents regarding the negotiations that the drafters, by framing the provision as a prohibition of abuse, envisaged a regulatory scheme. Members of the German delegation, particularly Müller77

Section 11, No. 2 of the draft. See Mestmäcker, “Verpflichtet § 22 GWB die Kartellbehörde, marktbeherrschenden Unternehmen ein Verhalten aufzuerlegen, als ob Wettbewerb bestünde?”, DB, 1968, 1800, at p. 1804. 79 See Mestmäcker, “Die Beurteilung von Unternehmenszusammenschlüssen nach Artikel 86 des EWG-Vertrags”, reprinted in Mestmäcker, Wirtschaft und Verfassung in der Europäischen Union, 2. Aufl. 2006, pp. 597 et seq., at p. 607; Mestmäcker, cited previous footnote, particularly at 1803–1806. 78

(G) Ehlermann 08 Schweitzer

136

6/11/08

11:48

Page 136

Session One: The Sherman Act And Abuse Of Position Under Art 82

Armack, on various occasions strongly warned against introducing any regulatory tendencies into the competition rules. The prohibition of abuses simply appeared to be the only clear-cut alternative to a per se prohibition of dominance at that time. The fact that the list of examples of possible abuses included exploitative abuses (see Article 82(a)) is no evidence to the contrary. Indeed, the idea of implementing competition rules which would cover exploitative abuses may even be seen as a “deregulatory” move. Previously, price controls had been an accepted (but utterly ineffective) instrument to fight inflation, which had repeatedly hit European economies hard in the recent past, to the point of threatening political and economic stability. In the period in which the EC Treaty was drafted, price levels were still an important concern. The concept of exploitative abuses preserved an instrument of intervention, but—and this was new—linked it to the concepts of competition policy. The way in which this new concept would work in practice was neither much discussed nor clearly foreseen. However, the change from anti-inflationary state price controls to controlling excessive pricing policies by dominant firms in the framework of a competition law regime implied a rejection of the idea of ongoing price controls. Only clear excesses should be controlled. In its utter vagueness, the concept of exploitative abuses was apparently immediately acceptable to all delegations. No debate ensued between the more liberal delegates and those that tended more towards a planning approach. Even to the liberal, market-oriented delegates, a limited control of excessive pricing or other exploitative abuses may have appeared acceptable in a market environment in which many of the dominant firms had achieved their position of dominance with the help of governments, e.g., through government privilege and protection, as was characteristic of the post-War period. It should also be emphasized that the fact that the examples in Article 82 clearly cover exploitative abuses does not imply that the drafters were concerned only with exploitative abuses. The drafters were aware of the relevance of exclusionary abuses and of the need to prevent the acquisition of positions of dominance by means other than competition on the merits. Again, the available documentation of the negotiations in the Common Market Working Group does not give evidence of much debate. It is obvious, however, that the examples in Article 82 are intended to represent both exploitative and exclusionary abuses. Even Article 82(a), which prohibits “unfair prices”, was directed against not only unfairly high prices but also unfairly low prices, i.e., predatory pricing schemes. The parallel debate in Germany concerning abuses of dominant positions under German competition law was clearly focused on exclusionary abuses80—not on exploitative abuses—an important 80 In Germany, the main concern was with boycotts and similar exclusionary techniques— the most pervasive types of abuses during the 1920s. Contrary to exploitative abuses, which were merely placed under the supervision of the German Cartel Authority, such exclusionary behaviour was generally prohibited by the GWB (specifically by § 26(2) of the statute), and this prohibition was made directly applicable.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 137

Principles of Section 2 Sherman Act and Article 82 EC 137 fact shedding light on the intent of the drafters of Article 82, given the German influence. The list of examples in Article 82 was not meant to be comprehensive. Rather, the examples stand for those types of conduct which were perceived to present the most pressing problems at the initial stage of the common market. Especially Müller-Armack cautioned against an attempt to provide a full list of possible abuses, advocating instead a more open approach that would leave room for evolution in the light of growing experience. Summing up, the drafting history confirms few of the standard claims about the attitudes and philosophy underlying Article 82. One claim that this history does confirm is the close link between the competition rules and the market integration goal. Indeed, creating rules that would constrain the ability of dominant firms to impede foreign entry into “their” national markets was a driving concern. This goal was by no means found to contradict efficiency concerns. Rather, the national boundaries were believed to be artificial and inefficient, and their opening was expected to create new potential for realizing efficiencies. Similarly, the debate about the competition rules themselves reflects efficiency concerns. In the discussions surrounding Article 82, the tension between allowing dominant firms to compete and restraining their capacity to exclude competitors was clearly identified, as was the danger that a provision on the abuse of dominance could become regulatory if it were applied in such a way as to micro-manage firms’ behaviour. Such a regulatory approach was clearly rejected, particularly by the German delegation. Nor does the drafting history support the proposition that the drafters were more concerned with equity than with efficiency. The proposals presented by the French delegation probably came closest to such a stance, with a very broad and strong general anti-discrimination principle for the whole of the economy, dissociated from the requirement of market power, and with the idea that a paramount challenge of the common market would be to create a level playing field in all respects. This position was, however, rejected. All in all, the discussions in the Working Group for the Common Market give the impression that the negotiating parties strived to create a system of competition rules that would allow the common market to become reality. This was the clear and primary concern. As for the enforcement regime within which the competition rules were to operate, the drafting history provides only limited insight into its design. It is important to recall that, when the competition rules entered into force in 1958, they were a mere potentiality whose actual functioning was not predictable. The Treaty of Rome left open the all-important question of how and by whom the competition rules should eventually be enforced after an initial period in which enforcement would be left to the Member States. This fundamental question was decided only later with the adoption of Regulation 17/62.81 The direct applicability of the 81 [1962] OJ 13, p. 204/62, as subsequently amended. Regulation 17/62 has of course been repealed and replaced by Regulation 1/2003, [2003] OJ L1/1.

(G) Ehlermann 08 Schweitzer

138

6/11/08

11:48

Page 138

Session One: The Sherman Act And Abuse Of Position Under Art 82

competition rules in the Treaty was not foreseen at the time the EC Treaty entered into force. Müller-Armack had stressed on various occasions that the competition rules at EC level should state general principles and guidelines, no more.82 Whether these rules would gain practical relevance, and to what extent, was completely unclear. When the first Commission took up work and divided competences among its members, neither France nor any other Member State was particularly interested in the idea of a designated Commissioner responsible for the free movement and competition rules. Ultimately, however, competence over these policy areas was assigned to a somewhat disappointed von der Groeben—the last one to choose.83

b) The interpretation of Article 82 EC—the formative period The Treaty of Rome entered into force in 1958. For the first decade of the Treaty’s existence, Article 82 was not applied. The first case that reached the ECJ was Continental Can.84 The Court’s judgment in this case continues to be of fundamental importance for the understanding and interpretation of Article 82. Before Continental Can, commentators’ ideas about the meaning and relevance of Article 82 sharply diverged. Representative of the uncertainties surrounding Article 82 is the monograph by René Joliet on monopolization and abuse of dominance, published in 1970.85 Contrasting Article 82 with the (at that time) strongly structural approach towards monopoly power under Section 2 Sherman Act, Joliet hypothesized that Article 82 lacked any structural component and took a purely behavioural stance. According to Joliet, the application of Article 82 was limited to controlling exploitative abuses, and would not extend to exclusionary abuses.86 In concentrating on exploitation, the Treaty of Rome exhibited a purely regulatory character: “The EEC Treaty [. . .] tends to curb only the abuses of power and thus to regulate the market behavior of dominant firms. The approach taken by Article 86 [Article 82 EC] is based upon an attitude of neutrality toward the existence of market dominant positions. It does not try to break up monopolistic positions, but instead, is confined to supervising the conduct and performance of dominant firms. Remedies are thus behavioral rather than structural. In cases of abuses, the enforcement 82 See Document 55, Darlegungen des Sprechers der deutschen Delegation zu den Entwürfen der Artikel 40–43, 8.9.1957, in: Schulze and Hoeren, supra note 51, at 167: “Die Erfahrungen aus der Kartellgesetzgebung sprechen dafür, in den Vertrag über den Gemeinsamen Markt nur einige präzise Leitsätze aufzunehmen. Andernfalls müßte eine weitgehende Kasuistik vorgesehen werden, die wohl nur in der Form des unmittelbar anwendbaren supranationalen Rechts geschaffen werden könnte. Die Leitsätze hingegen sollen noch kein unmittelbar anwendbares Recht darstellen.” 83 See von der Groeben, supra note 45, at 300–301. 84 Cited supra note 21. 85 Supra note 2. 86 See ibid. at 131.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 139

Principles of Section 2 Sherman Act and Article 82 EC 139 agency could go as far as to set prices at which dominant firms can sell or to fix the quantities which they must produce. The EEC approach amounts to a kind of public utility regulation”.87

Joliet concluded that the main preoccupation of the Treaty of Rome was not the maintenance of a competitive system. Rather, “the major objective of Article 86 is to ensure that dominant firms do not use their power to the detriment of utilizers and consumers”.88 Mestmäcker—at that time special advisor to DG IV (now DG Competition) and an influential voice in the development of EC competition law—took a radically different view. In preparing the Commission’s position in the Continental Can case, he started with the assertion that Article 82 had to be interpreted with a view to the overriding purpose of the competition rules, i.e., the aim of protecting a system of undistorted competition in the common market against distortions.89 Actions of dominant firms that are objectively incompatible with a system of undistorted competition must therefore be prohibited by Article 82. However, abuses of dominance cannot be defined based on the effects of a dominant firm’s actions on third parties alone. Article 82 prohibits a certain type of market conduct, not a certain type of market structure as such.90 Yet an abuse of a dominant position can lie in the restriction of (residual) competition, in defending a dominant position against current or potential competition, especially by hampering market entry, or in expanding a dominant position into adjacent markets. The fact that Article 82 does not oppose the formation of dominant firms does not preclude a finding of abuse in the case of a further strengthening of market dominance. Rather, by covering the maintenance and strengthening of dominance (other than by means of performance), Article 82 covers the most widespread, typical and dangerous exclusionary acts. Mestmäcker went on to establish certain guiding principles for the interpretation of Article 82. First of all, he stressed the close links between competition policy and the protection of open markets within the Community. The competition which the competition rules protect results from the opening up of the markets of the Member States. EC competition law must hence take particular care to ensure that dominant firms will not use their power to impede the entry into markets which the elimination of state barriers to trade has made possible.91 Secondly, he proposed the following criterion for determining whether an abuse has been committed under Article 82: third parties must be protected against harm to which they would not be exposed in the presence of effective competition. The dominant firm must not engage in acts which it could not carry out in a competitive 87

Ibid. at 127–128 Ibid., at 131. 89 Mestmäcker, “Die Beurteilung von Unternehmenszusammenschlüssen nach Artikel 86 des EWG-Vertrags”, supra note 79, at p. 603. 90 Ibid. at 604. 91 Ibid. at 606. 88

(G) Ehlermann 08 Schweitzer

140

6/11/08

11:48

Page 140

Session One: The Sherman Act And Abuse Of Position Under Art 82

environment.92 Thirdly, he stressed that the finding of an abuse should not depend on a finding that the elimination of a competitor has had a negative market effect. Competition law does not merely protect a certain degree of market efficiency, but it protects individual liberties against types of conduct that endanger competition if generalized. The protection of individual liberties is, at the same time, closely linked to the protection of competition as an institution, and to competition law’s economic rationale: Article 82 must, in the medium and long term, protect the possibility that positions of dominance will be corrected by the market. This presupposes the protection of those elements of competition that still persist.93 In Continental Can, the ECJ followed the principal lines of Mestmäcker’s arguments. In interpreting Article 82, the Court relied heavily on a functional approach. The general objective of establishing a system of undistorted competition in the common market, as articulated in Article 3(1)(g) of the Treaty,94 was found to be directly relevant for the interpretation of Article 82.95 The non-exhaustive list of abusive practices in Article 82 clearly covered both exploitative and exclusionary abuses, i.e., “practices which may cause damage to consumers directly, but also [. . .] those which are detrimental to them through their impact on an effective competition structure, such as is mentioned in Art. 3(f) [now Article 3(1)(g) EC] of the Treaty” (para. 26). In light of the Treaty’s fundamental decision to establish a common market with real or potential competition, and to effectively protect residual competition—a concept taken from Article 81(3)(b)—the ECJ concluded that the prohibition of abuse of dominance in Article 82 extends to a merger that would lead to a significant strengthening of dominance.96 Beyond Continental Can’s importance for the development of merger control (culminating finally with the Merger Regulation in 1989), two enduring messages flow from the Court’s judgment. Firstly, the goal of Article 82 is to protect a competitive market structure, i.e., one that does not render any serious chance of competition practically impossible (para. 25); and secondly, although Article 82 does not prohibit dominance, it does protect residual competition, i.e., the competition that remains in spite of existing dominance (Restwettbewerb). These far-reaching determinations have clarified that the main focus of Article 82 is not on exploitative, but on exclusionary abuses. In a series of decisions of the 1970s and 1980s, the ECJ further developed the contours of Article 82. In Hoffmann-La Roche, the ECJ defined the concept of abuse as: 92 Mestmäcker, “Die Beurteilung von Unternehmenszusammenschlüssen nach Artikel 86 des EWG-Vertrags”, supra note 79, at 607–608. 93 Ibid. at 608. 94 At that time, the provision was Article 3(f) EEC. 95 The ECJ rejected the applicant’s allegation that Article 3(1)(g) merely outlined a general programme, devoid of legal effect. See Continental Can, supra note 21, para. 23. 96 Ibid., para. 26. More recently, see Case C-95/04 P, British Airways plc v Commission [2007] ECR I-2331, para. 57 (with further references).

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 141

Principles of Section 2 Sherman Act and Article 82 EC 141 “an objective concept relating to the behaviour of an undertaking in a dominant position which is such as to influence the structure of the market where, as a result of the very presence of the undertaking in question, the degree of competition is weakened and which, through recourse to methods different from those which condition normal competition in products or services on the basis of the transactions of commercial operators, has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition”.97

The Court thus adopted the distinction between “competition on the merits” (Leistungswettbewerb), in which every undertaking, dominant or not, may engage, and illegal exclusionary conduct—a distinction which it has maintained ever since. While it highlighted the important principle that, under EC competition law, dominant firms are entitled to compete vigorously and aggressively, Hoffmann-La Roche failed to establish a clear test that would help to specify exactly where the line between legitimate competition and exclusionary conduct is to be drawn.98 An indication that has created some controversy and confusion is the phrase first used by the ECJ in Michelin I, according to which a dominant firm has a “special responsibility not to allow its conduct to impair undistorted competition on the common market”.99 While some have associated the concept of “special responsibility” with a tendency of EC competition law to protect smaller and less efficient competitors, this is, as has meanwhile been clarified, not what the Court meant. What the concept of “special responsibility” does entail is the generally uncontroversial observation that conduct engaged in by a dominant firm may be abusive, even when the same conduct carried out by a non-dominant firm is perfectly legitimate.100 Methods of competition which are, in principle, legitimate, can lead to the maintenance and extension of market power when they are applied by dominant firms. 101 While this is not the place to attempt a comprehensive summary of the ECJ’s Article 82 jurisprudence, some of the important features which distinguish it from the current Section 2 Sherman Act jurisprudence in the US will be highlighted here. 97 Case C-85/76, Hoffmann-La Roche v Commission [1979] ECR 461, para. 91. See also British Airways, cited previous footnote, at para. 66. 98 See Robert O’Donoghue and A. Jorge Padilla, The Law and Economics of Article 82 EC, Hart Publishing, p. 176. 99 Case 322/81, NV Nederlandsche Banden-Industrie Michelin v Commission [1983] ECR 3461, para. 57. 100 See joined cases T-191/98, T-212/98 to T-214/98, Atlantic Container Lines AB and Others v Commission [2003] ECR II-3275, para. 1460: “special responsibility means only that a dominant undertaking may be prohibited from conduct which is legitimate where it is carried out by nondominant undertakings”. 101 See Mestmäcker, Das Marktbeherrschende Unternehmen im Recht der Wettbewerbsbeschränkungen, Tübingen, 1959, p. 6: “. . . die rechtliche Behandlung von Marktmacht ist vor allem deshalb so schwierig, weil Inhaber von Macht befähigt sind, die Institute des Privatrechts und die unter den Voraussetzungen freier Konkurrenz legitimen Mittel des Wettbewerbs in den Dienst der Marktbeherrschung zu stellen”.

(G) Ehlermann 08 Schweitzer

142

6/11/08

11:48

Page 142

Session One: The Sherman Act And Abuse Of Position Under Art 82

First of all, Article 82 is “not only aimed at practices which may cause damage to consumers directly, but also at those which are detrimental to them through their impact on an effective competition structure, such as is mentioned in Article [3(1)(g) EC]”.102 While this phrase was originally used by the ECJ to confirm the applicability of Article 82 not only to exploitative, but also to exclusionary abuses, it has today come to stand for the more farreaching claim that the EC competition rules protect the competitive process, the degree of residual competition that persists in the market as such, and that they do not require a finding of direct consumer harm.103 The protection of consumer interests is mediated through the protection of competition—from which consumer welfare is generally thought to result. This approach is based on the assumption that competition will typically result in more innovation and efficiency than monopoly. It is preferable to let the market enforce efficiency and innovation, rather than relying on the announced efficiency goals of private monopolists, and rather than relying on appraisals of likely efficiencies by competition authorities and courts. Secondly, Article 82 jurisprudence has maintained a focus on protecting market access for competitors.104 Actions of dominant firms that produce an “exclusionary effect” are not necessarily abusive per se, but the ECJ will have a closer look at whether they are economically justified.105 Generally, the ECJ appears to favour a balancing approach: an exclusionary effect disadvantageous to competition may be counterbalanced or outweighed by advantages in terms of efficiency. But if the exclusionary effect bears no relation to the advantage for the market and consumers, or if it goes beyond what is necessary to attain those advantages, it will be regarded as an abuse.106 Thirdly, and in parallel with the foregoing remarks, the ECJ’s jurisprudence has maintained its fundamental understanding that Article 82 is not only about protecting outcome efficiency but also aims to protect the individual rights of competitors. According to some, this demonstrates that EC competition law is about protecting competitors instead of competition. However, this is an “empty slogan”.107 The challenge, in both EC and US antitrust law, is to distinguish those acts with exclusionary effects that result from legitimate competition on the merits from other exclusionary acts which cannot be justified as normal acts of competition but which, to the contrary, exploit the special power that a dominant firm possesses so as to entrench the 102 Continental Can, supra note 21, para. 26. For a recent confirmation, see British Airways, supra note 97, para. 106. 103 See British Airways, supra note 97, para. 107. 104 Ibid., para. 67. 105 Ibid., para. 69. 106 Ibid., para. 86. 107 Wolfgang Wurmnest, “The Reform of Article 82 EC in the Light of the ‘Economic Approach’”, in Josef Drexl, Beatriz Conde Gallego, Stefan Enchelmaier and Mark-Oliver Mackenrodt, eds., Art. 82 EC: New Interpretation, New Enforcement Mechanisms?, Springer, forthcoming (Working Paper of October 2006 on file with the author; quoted expression at p. 14).

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 143

Principles of Section 2 Sherman Act and Article 82 EC 143 firm’s position in the marketplace. The difference between the EC approach and the US approach is that EC competition law, based on this distinction, assumes an individual right of each competitor not to be excluded by illegal acts, whether or not the exclusion results in a verifiable overall decrease of competition or efficiency in the marketplace. Antitrust law in the US, on the other hand, which takes consumer welfare to be the only goal of antitrust, tends to require a showing of verifiable effect in the marketplace, and thus negates the notion that the competition which is protected by competition law is constituted by the exercise of individual liberties.

II. Exploitative and exclusionary abuses: regulatory tendencies in Article 82? The insights we can gain into the different attitudes underlying Article 82 and Section 2 Sherman Act by looking at the drafting history of Article 82 (see point II.2(a) above) are by necessity limited. This section of the paper will therefore look at three areas which are frequently said to stand for a divergence of interpretation and philosophy as between Article 82 and Section 2, namely: exploitative abuses (1); predatory pricing (2) and refusals to deal / essential facilities (3).

1. Exploitation of monopoly power under Section 2 Sherman Act and Article 82 EC One difference between Section 2 Sherman Act and Article 82 that is frequently held out to be indicative of fundamentally different attitudes towards rules regarding market power is the fact that only Article 82 addresses exploitative abuses.108 Pursuing exploitative abuses implies making judgments about a dominant firm’s price and output decisions, and this necessarily comes into the vicinity of regulatory supervision. The fact that Article 82 covers exploitative abuses is thus taken as proof of the regulatory approach embraced by EC competition law,109 and as proof of an overriding concern with fairness rather than efficiency.110 Frequently, the decision to address exploitative abuses is traced back to the alleged ordoliberal influence on the

108 109 110

See, e.g., Fox, supra note 4, at 993–994. Joliet, supra note 2, at 127–128 and 131. See also Fox, supra note 4, at 992–994. For that claim see, e.g., Gal, supra note 11, at 363 et seq. See also Fox, supra note 4, at 985.

(G) Ehlermann 08 Schweitzer

144

6/11/08

11:48

Page 144

Session One: The Sherman Act And Abuse Of Position Under Art 82

formulation and interpretation of Article 82,111 and particularly to the supposedly ordoliberal “as if” competition approach.112 According to this view, competition authorities are, under Article 82(a) and (b), required to ensure that dominant firms set output and price as if they operated in competitive markets.113 Section 2, on the other hand, does not control the exercise of monopoly power,114 but only its acquisition or maintenance. It protects the openness and competitive structure of the market to which the determination of price and output level are then left. The decade before the enactment of the Sherman Act had been one of rapid economic growth and declining prices, even in those industries dominated by trusts.115 Congress was therefore not concerned with implementing controls against excessive prices. Price and output controls would furthermore have contravened the dominant “freedom of contract” philosophy of the times. In one of its early decisions, the US Supreme Court acknowledged the inherent difficulties that antitrust authorities and courts face when required to oversee output and price decisions.116 But the control of exploitative abuses is not only rejected on these pragmatic grounds. As a matter of principle, monopoly power and monopoly pricing are viewed as a part of the competitive process and as an important driving force. According to the US Supreme Court’s opinion in the Trinko case: “The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the freemarket system. The opportunity to charge monopoly prices—at least for a short period—is what attracts ‘business acumen’ in the first place; it induces risk taking that produces innovation and economic growth.”117

The underlying assumption is that the monopoly position will be a transitory one. Absent significant entry barriers, monopoly prices can be expected to invite new market entry which will eventually drive prices down. From the standpoint of US law, in those specific cases where there is a durable monopoly position, it is not for competition authorities and antitrust courts to intervene but for Congress to establish special regulatory oversight. The difference between Article 82 and Section 2 with regard to the coverage of exploitative abuses is indeed remarkable. In this regard, during the first decade of the Community’s existence there was significant speculation as to 111

See Gal, supra note 11, at 364 et seq. O’Donoghue and Padilla, supra note 98, at 604. 113 Ibid. 114 Emphasized in Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407–408 (2004). 115 Herbert Hovenkamp, Federal Antitrust Policy: The Law of Competition and its Practice, 3rd ed., 2005, p. 51. 116 See United States v. Trans-Missouri Freight Ass., 166 U.S. 290 (1897). See also United States v. Trenton Potteries Co., 273 U.S. 392 (1927): “The reasonable price of today may through economic and business changes become the unreasonable price of tomorrow.” 117 Trinko supra note 114, at 407. 112

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 145

Principles of Section 2 Sherman Act and Article 82 EC 145 the fundamental role and function of Article 82. However, the view that Article 82 was intended to cover only exploitative abuses, defended, inter alia, by Joliet,118 became obsolete with the ECJ’s judgment in Continental Can. The functional interpretation of Article 82, guided by the goal of establishing a system of undistorted competition and of promoting the integration of formerly barricaded national markets, has led to a competition policy and jurisprudence which is focused almost exclusively on controlling exclusionary abuses. Exploitative abuses, meanwhile, have suffered (or profited) from “benign neglect”.119 Throughout the years, the Commission has adopted only four formal decisions condemning excessive prices.120 In a series of cases, the ECJ has confirmed—mostly in preliminary rulings—that Article 82 can apply to exploitative abuses; but the Court has established a high threshold for finding that a certain price level is excessive. In only one case has it actually found an abuse.121 The relevant test was first established in United Brands. Here, the ECJ accepted that “[c]harging a price which is excessive because it has no reasonable relation to the economic value of the product supplied [is] . . . an abuse”.122 A two-stage test must be applied, however, to determine whether a price is reasonably related to the “economic value of the product”. Firstly, the difference between the cost actually incurred and the price actually charged needs to be determined; and if this difference is excessive, it must further be determined “whether a price has been imposed which is either unfair in itself or when compared to competing products” (United Brands, at 118 Joliet, supra note 2, pp. 11 and 131. Under Article 82, the offence lies “mainly in abusive market exploitation through unreasonably high prices or monopolistic restriction of output. As is shown by the examples listed in Article [82], the main preoccupation of the Treaty is not the maintenance of a competitive system. All the examples relate to cases of practices and policies through which a firm exploits its market dominant power. None of them concerns means by which market dominant power can be achieved or maintained. Large size is considered as an economic necessity, the basic assumption underlying Article [82] being that monopolistic structure does not lead inevitably to monopolistic performance. The monopolist’s performance may be in harmony with the public interest. The EEC monopoly policy has adopted an attitude of neutrality toward market dominant power. A dominant position implies a power to fix unilaterally unfairly high prices. The Treaty assumes however that this power will not be systematically utilized. This is why monopoly power as such is not condemned. Monopoly is not in itself an evil. Only the unilateral fixing of unfair prices is in violation of the law. It is the exercise of monopoly power which can be subject to regulation.” More recently, Pinar Akman has claimed that Article 82 was intended to cover exploitative abuses only. See supra note 15, at 4 and 37–38. 119 This is recognized by O’Donoghue and Padilla, supra note 98, at 608. 120 See Commission Decision 75/75 of 19 December 1974, General Motors [1975] OJ L29/14; Commission Decision 76/353 of 17 December 1975, Chiquita (United Brands) [1976] OJ L95/1; Commission Decision 84/379 of 2 July 1984, British Leyland [1984] OJ L207/11; Commission Decision 2001/892 of 25 July 2001, Deutsche Post II [2001] OJ L331/40, paras. 159–167. 121 See Case 116/84, British Leyland Plc v Commission [1986] ECR 3263. In a limited number of other preliminary rulings, the ECJ has acknowledged the possibility of an abuse. See Case 110/88, Lucazeau et al. v SACEM et al. [1989] ECR 2811; Case 30/87, Bodson v Pompes funèbres [1988] ECR 2479; Case C-66/86, Ahmed Saeed Flugreisen et al. v Zentrale zur Bekämpfung unlauteren Wettbewerbs [1989] ECR 803; Case C-242/95, GT-Link v. DSB [1997] ECR I-4449. 122 Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207, para. 250.

(G) Ehlermann 08 Schweitzer

146

6/11/08

11:48

Page 146

Session One: The Sherman Act And Abuse Of Position Under Art 82

para. 252). In full realization of the practical difficulties associated with determining the costs of production,123 the ECJ charged the Commission with the burden of proving the excessiveness of a price.124 In the case at issue, the Commission had failed to establish that the prices charged by United Brands were unrelated to the economic value of the product, and the Court therefore annulled the relevant part of the Commission’s Decision. Motta and de Streel have demonstrated in a careful study that, based on the ECJ’s case law, excessive pricing cases have been pursued successfully only in the presence of special circumstances.125 In these rare cases, either: (i) the dominant undertaking at issue enjoyed a de facto monopoly (see, for example, SACEM 126), in which case the ECJ has tended to lower the preconditions for finding a violation of Article 82(a) as well as the standard of proof the Commission had to meet,127 and the abuse furthermore created serious impediments to the internal market and included concerns about price discrimination and artificial barriers to parallel trade (General Motors, British Leyland);128 or (ii) the dominant undertaking was active in markets recently opened to competition (Deutsche Post II,129 telecommunications cases130), and any pricing abuses could have weakened the political momentum for Europe’s liberalization programme. Where such special circumstances are absent, the Commission has declared its general unwillingness to act as a price regulator vis-à-vis dominant firms.131 The Commission’s reluctance to pursue 123 Difficulties in quantifying costs can result, inter alia, if the firm: has made long-term investments; has taken particular risks in developing a product; is diversified with a multi-product structure; or has intellectual property rights. 124 “[H]owever unreliable the particulars supplied by [the dominant company] . . ., the fact remains that it is for the Commission to prove that [the dominant company] charged [excessive] prices” (United Brands, para. 264). 125 Massimo Motta and Alexandre de Streel, “Excessive Pricing and Price Squeeze under EU Law”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, p. 91, at 107, with further references. 126 Lucazeau v SACEM, supra note 121. See also Case 395/87, Ministère Public v Tournier [1989] ECR 2521. In the SACEM cases, the operators of French discotheques complained that SACEM, the French Copyright collecting society, was charging higher fees for licences of performing rights than those charged by similar collecting societies located in other Member States. In preliminary rulings, the ECJ found that the fees charged by SACEM qualified as “unfair trading conditions” if the rates were manifestly higher than those applied by identical copyright societies in other Member States. 127 On the lowering of the preconditions for finding excessive pricing in SACEM as compared to United Brands, but wrongly generalizing the SACEM test, see Gal, supra note 11, at 370 et seq. 128 See ibid. at 375 et seq. According to Motta and de Streel, “[t]he Commission was more concerned with the freedom of circulation than with the anticompetitive exploitation of end users and the associated allocative inefficiencies”. Supra note 125, at 107. 129 Deutsche Post II, supra note 120, paras. 159–167. 130 On the handling of excessive pricing in the telecoms sector, see Commission, Notice on the application of competition rules to access agreements in the telecommunications sector, [1998] OJ C265/2, paras. 105–109. 131 Commission, XXIVth Report on Competition Policy (1994), para. 207: “. . . the existence of a dominant position is not in itself against the rules of competition. Consumers can suffer from

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 147

Principles of Section 2 Sherman Act and Article 82 EC 147 exploitative abuses is based on the same reasons underlying the inapplicability of Section 2 Sherman Act to exploitative abuses: the fact that it is extremely difficult to establish conditions that provide predictability132 as to when a price should be viewed as “excessive”;133 the spectre of continuous price regulation that would go along with a more forceful attempt to push Article 82(a) any further;134 and the fact that if there were a serious threat that gaining significant market power would expose firms to a regime of price control, this would negatively affect successful companies’ incentives to innovate and invest.135 Based on these findings, the US antitrust and EU competition law perspectives on exploitative abuses do not appear to be far apart. There is broad consensus that competition law should not intervene where the market can be expected to self-correct exploitative practices in the short or medium term.136 On the other hand, both jurisdictions accept that an economic rationale for price regulation can exist where high non-transitory barriers to entry, like government monopolies, exclude competition in the longer term.137 The difference between the two systems boils down to a difference in the allocation of decision competences. In the US, it is for Congress to decide whether a regulatory scheme is necessary. In the EU, while regulatory instruments are a dominant company exploiting this position, the most likely way being through prices higher than would be found if the market were subject to effective competition. The Commission in its decision-making practice does not normally control or condemn the high level of prices as such. Rather it examines the behaviour of the dominant company designed to preserve its dominance, usually directly against competitors or new entrants who would normally bring about effective competition and the price level associated with it.” This policy approach is reaffirmed in, e.g., Commission, XXVIIth Report on Competition Policy (1997), point 77. The same is true for other exploitative abuses. 132 For the relevance of the problem of legal certainty, see O’Donoghue and Padilla, supra note 98, at 622. 133 See Motta and de Streel, supra note 125, at 109: “Indeed, in many situations even computing the relevant measures of costs would be a complex exercise: How does one allocate common costs to different products (long-run incremental costs, stand-alone costs)? How does one choose between different accounting methods (historic costs, current costs)? Which measure of costs should be adopted to measure profits in industries where there are important fixed costs? All these difficulties are underlined by the fact that a competition authority may not have as deep a knowledge of the sector being investigated as an industry regulator”. See also O’Donoghue and Padilla, supra note 98, at 621: “. . . no generally accepted criterion exists in the decisional practice and case law to determine when prices are ‘excessive’. Further, even if a criterion, or series of criteria, could be agreed upon as a benchmark, determining an excessive price in practice is extremely complex and subject to a number of difficulties.” O’Donoghue and Padilla also point to the potentially high cost of error when competition authorities or courts attempt to identify excessive prices. 134 Motta and de Streel, supra note 125, at 109. 135 O’Donoghue and Padilla, supra note 98, at 621: “. . . prices above marginal cost are common and necessary in many industries where high profits are necessary to recover large up-front capital and other fixed costs”. See also Fox, supra note 4, at 993 (noting the relevance of investments made in securing IPRs). 136 O’Donoghue and Padilla, supra note 98, at 605. 137 Ibid. at 638 (referring to the OFT Draft Competition Law Guidelines for Consultation, Assessment of Conduct, April 2004, para. 2.6). See also Gal, supra note 11, at 383.

(G) Ehlermann 08 Schweitzer

148

6/11/08

11:48

Page 148

Session One: The Sherman Act And Abuse Of Position Under Art 82

employed to some extent in certain sectors, competition law authorities can also intervene. The actual exercise of price regulation remains difficult under both regimes. Regulatory agencies may, in the end, be better placed to engage in price regulation. Against the background of the EU’s limited legislative competence in this area, competition rules may, however, be a useful safeguard and substitute. In this perspective, the coverage of exploitative abuses by EC competition law is no evidence of a fundamental divergence in “antitrust philosophy”, but reacts to different legislative capacities at the federal US or EU level respectively. If this overall picture is true, more fascinating than the actual difference between EU and US competition law itself is the narrative that has been constructed around it. According to O’Donoghue and Padilla, the objectives of Article 82(a) “lie at the core of EC competition law: to prevent the exploitation of consumers by firms with significant market power”.138 Michal Gal, while acknowledging the lack of practical relevance of exploitative abuses in EC competition law,139 claims nonetheless that their coverage reflects important “ideological goals”,140 particularly a concern with social and redistributive goals and “fairness”, which are then attributed to German ordoliberal influence.141 Such allegations erect a strawman. They blame German ordoliberalism for tenets it never defended, and assume a meaning or tendency of Article 82 for which there is no evidence in the case law of the last 50 years. They allege an opposition between EC and US antitrust law which does not exist. Regulatory “aspirations” on the part of EC competition law, and its instrumentalization for non-economic goals, are claimed where none can be found. Over the years, the Commission as well as the ECJ have subscribed to all of the same reservations against controlling exploitative abuses that are typical of US antitrust law. A revival of exploitative abuses in EC competition law beyond the narrow setting of a non-transitory (mostly state-protected) monopoly is both unlikely and undesirable.

138

O’Donoghue and Padilla, supra note 98, at 637. Gal, supra note 11, at 360 and 374 –375. 140 See ibid. at 346: “The regulation of excessive pricing encapsulates issues such as the goals and underpinnings of EC and U.S. antitrust systems; the equilibrium point which was adopted to balance between the forces of Darwinian capitalism and those of social justice; the role of government regulation; the balance between practical problems and theoretical principles; and the assumptions regarding the relative administrability of various types of regulation. Monopoly pricing regulation is thus, in many ways, a microcosm of competition policy.” Furthermore, the prohibition of excessive prices allegedly stands for an opening of EC competition policy to the wider set of EC Treaty goals set out in Article 2 EC, namely a harmonious development of economic activities, a continuous and balanced expansion, enhanced stability, an accelerated raising of the standard of living, and closer relations between the Member States. Ibid. at 361–362. 141 Ibid. at 364. 139

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 149

Principles of Section 2 Sherman Act and Article 82 EC 149

2. Predatory pricing in EC competition law and US antitrust law While there does not appear to be a fundamental gap between EC competition and US antitrust “philosophy” with regard to exploitative abuses, predatory pricing is one of the areas in which the two jurisdictions do diverge. There is agreement on the general description of the phenomenon: predatory pricing schemes involve low pricing strategies—typically pricing below some measure of cost142—in an effort to eliminate competitors or to deter entry by potential competitors. If the plan succeeds, the reduction of actual and/or potential competition will allow the predator to raise prices to a supracompetitive level in the longer run. However, there is no transatlantic consensus on the legal test to be applied. In the US, the law on predatory pricing has been strongly influenced by the Chicago school. In an early and highly influential article by John S. McGee on the predatory pricing allegations in Standard Oil, and later in a variety of broader studies,143 Chicago scholars have long maintained that predatory pricing schemes are generally irrational, and therefore unlikely under all but very exceptional circumstances. According to these scholars, predatory pricing is a highly speculative scheme: a predator must incur losses now in the mere hope that he will be able to recover them in the future. The prospect of actual recovery is furthermore slim, since competitors can and will re-enter once the predatory pricing scheme is abandoned. As a consequence, Chicago scholars proposed to abolish the doctrine: “It seems unwise . . . to construct rules about a phenomenon that probably does not exist or which, should it exist in very rare cases, the courts would have grave difficulty distinguishing from competitive price behaviour. It is almost certain that attempts to apply such rules would do much more harm than good.”144 The issue reached the US Supreme Court in 1993 in Brooke Group.145 The Supreme Court took the opportunity to distance itself from prior case law on predatory pricing, namely from Utah Pie,146 which had, on the basis of the 142 The US Supreme Court limits illicit price predation to pricing below some measure of cost (see below). In the EU, illegal exclusionary conduct has sometimes been found in cases in which prices remained above both average variable cost and average total cost. Economists are divided on the question of whether such above-cost pricing should constitute an antitrust violation, but they acknowledge that it can, under some conditions, have an exclusionary effect. See Temple Lang and O’Donoghue, supra note 9, at 121–122. 143 See John S. McGee, “Predatory Price Cutting: The Standard Oil (N.J.) Case”, 1 Journal of Law and Economics 137 (1958), alleging that Standard Oil had never used predatory price discrimination to drive out competing refiners. See also Bork, supra note 33, at 144 et seq. 144 Robert H. Bork, supra note 33, at 154. 145 Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209. See also Matsushita Elec. Industrial Co. v. Zenith Radio Corp., 475 U.S. 584 (1986). 146 Utah Pie Co. v. Continental Baking Co., 386 U.S. 685 (1967). The judgment was not directly overruled in Brooke Group. Rather, as Richard Posner observes, “Brooke Group distinguishes Utah Pie to death”. Posner, supra note 41, at 223.

(G) Ehlermann 08 Schweitzer

150

6/11/08

11:48

Page 150

Session One: The Sherman Act And Abuse Of Position Under Art 82

Robinson-Patman Act, practically inferred predation from proof of price discrimination plus exclusionary intent.147 In Brooke Group, the Supreme Court introduced a new and very narrow cost-based test that eliminated the criterion of anticompetitive intent and instead required proof of market effect, or a dangerous probability of market effect. To substantiate a predatory pricing claim, a plaintiff now has to prove that: (i) the alleged predatory prices are below an appropriate measure of the defendant’s costs;148 and (ii) there is a “dangerous probability” that the defendant would be able to recoup its investment in below-cost prices,149 i.e., that the defendant would eventually be able to raise price above a competitive level to an extent sufficient to compensate for the amounts expended on the predation, including the time value of the money invested in it.150 It is generally acknowledged that recoupment is extremely difficult to prove.151 The prohibition of predatory pricing is, as a consequence, rarely enforced.152 Proof of recoupment along the lines developed in Brooke Group and subsequent case law not only poses practical problems of information and prediction, it also tends to ignore the more complex strategic reasons for which predatory pricing schemes may be pursued.153 Based on the test established in Brooke Group, a significant number of relevant predatory pricing schemes may thus not be caught.154 Nevertheless, Brooke Group was fully confirmed recently in a unanimous decision by the Supreme Court in Weyerhaeuser,155 a predatory bidding case. 147 Posner, supra note 41, at 221. For the proposition that Utah Pie went far in the direction of condemning hard price competition itself, see Hovenkamp, supra note 115, at 365–366. 148 Generally, courts apply the Areeda and Turner test which was developed in the 1970s. According to this test, prices below the average variable cost of a product are presumed to be predatory. For the broader debate surrounding this test, see, inter alia, F.M. Scherer, “Predatory Pricing and the Sherman Act: A Comment”, 89 Harvard Law Review 869 (1976); Phillip Areeda and Donald F. Turner, “Scherer on Predatory Pricing: A Reply”, 89 Harvard Law Review 891 (1976); Joseph F. Brodley and George A. Hay, “Predatory Pricing: Competing Economic Theories and the Evolution of Legal Standards”, 66 Cornell Law Review 738 (1981). 149 Brooke Group, supra note 145, at 224. 150 Ibid. at 225. 151 For a summary of what is required to prove a dangerous likelihood of recoupment, see Temple Lang and O’Donoghue, supra note 9, at 142. The authors acknowledge that this analysis constitutes “a considerable barrier to plaintiffs trying to establish a predatory pricing claim”. For the difficulty of proving recoupment, see also Hovenkamp, supra note 115, at 370. 152 According to Evans and Padilla, there have been no successful prosecutions of predatory pricing claims in the US since Brooke Group. Evans and Padilla, supra note 22, at 88 (2005). 153 For example, predatory pricing in one market can be a strategy to deter entry or effective competition in other markets in which the dominant firm is engaged. Even in the absence of barriers to entry, signalling the willingness to engage in predatory pricing schemes can have important deterrence effects. For further explanation, see Posner, supra note 41, at 211. 154 Hovenkamp emphasizes the inability of the Brooke Group case law to deal with oligopolistic settings. See supra note 115, at 370. See also Temple Lang and O’Donoghue, supra note 9, at 144–145, acknowledging that a recoupment criterion would imply a lacuna in cases of predatory pricing where a competitor, although not forced out of the market during the period of low pricing, is disciplined in the sense that it later raises its prices near to the level preferred by the dominant firm for fear that the latter would retaliate if it actively competed on price. 155 Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., 549 U.S. 312 (2007), 127 S.Ct. 1069.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 151

Principles of Section 2 Sherman Act and Article 82 EC 151 Community competition law has taken a very different approach towards predatory pricing. In AKZO Chemie BV 156 and Tetra Pak International,157 the ECJ and CFI distinguished between two relevant situations. First, the courts made clear that it is abusive per se for an undertaking in a dominant position to sell at prices below average variable cost. Predatory intent is presumed, because “the only interest which the undertaking may have in applying such prices is that of eliminating competitors”.158 Second, prices above average variable cost but below average total cost are abusive “if they are determined as part of a plan for eliminating a competitor”. In this case, “sound and consistent evidence”159 must be provided to show an intent and a strategy to pre-empt the market.160 In contrast to the US Supreme Court, the ECJ has explicitly rejected a requirement to prove a likely market effect in a predatory pricing case, i.e., a requirement to show that the dominant firm had a realistic chance of recouping its losses.161 In Tetra Pak, the ECJ stated:162 “[I]t would not be appropriate, in the circumstances of the present case, to require in addition proof that Tetra Pak had a realistic chance of recouping its losses. It must be possible to penalise predatory pricing whenever there is a risk that competitors will be eliminated. [. . .] The aim pursued, which is to maintain undistorted competition, rules out waiting until such a strategy leads to the actual elimination of competitors.”

Where intent has been shown, the conduct is assumed to be liable to have the desired exclusionary effect.163 And “where an undertaking in a dominant position actually implements a practice whose object is to oust a competitor, the fact that the result hoped for is not achieved is not sufficient to prevent that being an abuse of a dominant position within the meaning of Article 82”.164 The divergence between the EU approach and the US approach to predatory pricing is thus remarkable. Both jurisdictions start by looking at the 156

Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359. Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755, upheld on appeal: Case C-333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951. 158 AKZO, supra note 156, paras. 71–72; Case T-340/03, France Télécom S.A. v Commission (“Wanadoo”), [2007] ECR II-107, paras. 195–196 and 224; on appeal: Case C-202/07, not yet decided. 159 Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755, para. 151; Wanadoo, cited previous footnote, para. 197. 160 Case C-333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951, paras. 41 et seq.; Wanadoo, supra note 158, paras. 196 and 198. 161 Wanadoo, supra note 158, para. 195; Case T-203/01, Manufacture Française des Pneumatiques Michelin v Commission (“Michelin II”) [2003] ECR II-4071, paras. 241–242. 162 Case C-333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951, para. 44. 163 Wanadoo, supra note 158, para. 195: “If it is shown that the object pursued by the conduct of an undertaking in a dominant position is to restrict competition, that conduct will also be liable to have such an effect.” 164 Ibid., para. 196, with further references. 157

(G) Ehlermann 08 Schweitzer

152

6/11/08

11:48

Page 152

Session One: The Sherman Act And Abuse Of Position Under Art 82

differential between cost and price, but here the similarities end. In the US, the law relies fully on proof of below-cost pricing and market effect, or consumer harm, and dismisses the intent criterion. In the EU, the prohibition against predatory pricing may sometimes even extend to above-cost pricing,165 in which case intent will be the main criterion. A requirement to prove a likely market effect has been rejected just recently again by the CFI in the Wanadoo case.166 What explains such a serious divergence in this context between the EU and the US? To some extent, there may be, underlying the different legal tests, a different appraisal of the likelihood that such strategies will in fact be pursued, and of the likelihood that they will succeed.167 One of the reasons the US Supreme Court has given for its admittedly restrictive predatory pricing test is its acceptance of the Chicago school’s factual allegation that predatory pricing schemes are normally too speculative to be a rational business strategy, and that they will therefore rarely occur.168 The EC case law, by contrast, is based on the factual presumption that predatory pricing can indeed be a rational strategy for a dominant firm to eliminate competitors,169 and that it actually occurs in practice. It is of course possible that under the given market conditions in the EU and the US, predatory pricing is indeed a strategy that is more frequently, and potentially more successfully, applied on this side of the ocean. In this respect, only empirical research can ultimately provide certainty.170 However, it is unlikely that the difference in the factual setting is sufficient to explain the difference between the two legal regimes. Antitrust scholarship today generally recognizes that predatory pricing strategies are more plausible than early Chicago scholarship had maintained. In the 2nd edition of his monograph on “Antitrust Law”, Judge Posner finds that “predatory pricing cannot be dismissed as inevitably an irrational practice”, and he devotes substantial attention to it.171 Likewise, in US v. AMR (2003), the Tenth Circuit observes: 165 The US Supreme Court has refused to extend the purview of predatory pricing to such situations because it would be “beyond the practical ability of a judicial tribunal to control without courting intolerable risks of chilling legitimate” pro-competitive conduct. See Brooke Group, supra note 145, at 223; Weyerhaeuser, supra note 155, 127 S.Ct. at 1078. 166 Wanadoo, supra note 158, paras. 195–196, with further references. 167 For a very sceptical view that predatory pricing strategies are economically feasible and relevant, see Bork, supra note 33, at 145. 168 See Matsushita, supra note 145, at 589: “[P]redatory pricing schemes are rarely tried, and even more rarely successful”. See also Weyerhaeuser, supra note 155, 127 S.Ct. at 1077: “Predatory pricing requires a firm to suffer certain losses in the short term on the chance of reaping supracompetitive profits in the future. . . . A rational business will rarely make this sacrifice.” 169 See also Temple Lang and O’Donoghue, supra note 9, at 122: “. . . there is agreement that predatory pricing may be profitable and anticompetitive . . .”. 170 According to some scholars, the number of predatory pricing schemes that are actually implemented in the US is non-trivial. See Posner, supra note 41, at 214; Patrick Bolton, Joseph Brodley and Michael Riordan, “Predatory Pricing: Strategic Theory and Legal Policy”, 88 Georgetown Law Journal 2239, 2244–2247 (2000). 171 Posner, supra note 41, at 213 (emphasis in original).

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 153

Principles of Section 2 Sherman Act and Article 82 EC 153 “Recent scholarship has challenged the notion that predatory pricing schemes are implausible and irrational. [. . .] Post-Chicago economists have theorized that price predation is not only plausible, but profitable, especially in a multi-market context where predation can occur in one market and recoupment can occur rapidly in other markets.”172 Where predatory pricing can admittedly be a rational exclusionary strategy and constitutes a real risk, the broader test under EC law cannot be easily dismissed as resulting from “unsound economics”.173 Renowned antitrust scholars concede that the Brooke Group test for predatory pricing is in significant respects underinclusive,174 and that underdeterrence may be the result.175 The US Supreme Court itself has implicitly acknowledged the potential underinclusiveness of the Brooke Group test, but it has emphasized the high costs that “false positives” could potentially have:176 “The mechanism by which a firm engages in predatory pricing—lowering prices—is the same mechanism by which a firm stimulates competition.” If a broader test were applied, firms might start to fear predatory pricing allegations and might become reluctant to cut prices aggressively, to the detriment of consumers.177 “[M]istaken findings of liability would chill the very conduct the antitrust laws are designed to protect.”178 With a view to the importance of price competition, the US Supreme Court thus expresses a clear preference for underdeterrence as compared to a broader test that might have overdeterrent effects. Some have submitted that the broader approach towards predatory pricing in EC law stands for a different view of the comparative costs of type I versus type II errors, i.e., the costs of “false positives” versus “false 172

United States v. AMR Corp., 353 F.3d 1109 (10th Cir. 2003), at section III in fine. For the economic soundness of predatory pricing law under EC law, see also Temple Lang and O’Donoghue, supra note 9, at 87. Interestingly, EC law on predatory pricing is not so far from Posner’s proposal to prohibit pricing below short-run marginal cost per se (see Posner, supra note 41, at 215, stating that “[t]here is no reason consistent with an interest in efficiency for selling a good at a price lower than the cost that the seller incurs by the sale”) and to prohibit selling below long-run marginal cost where an intent to exclude a competitor can be shown. Ibid., pp. 215–216. 174 See Evans and Padilla, supra note 22, at 87, who nonetheless recommend the Brooke Group test: “This test fails to identify all possible price predation practices but follows from the view that it is better to err by allowing some predatory pricing than to condemn some competitive pricing. The Supreme Court has properly moved to a stricter standard for showing predation because (a) setting prices low is a hallmark of competition (so that the cost of falsely condemning legitimate price cutting is high) and (b) successful predation is rare (so that the likelihood of false acquittals is low)”. 175 See Kovacic, supra note 8, at 51 and 53. See also Antitrust Modernization Commission, supra note 22, at 87: “Particularly in the context of Section 2 predatory pricing enforcement— where overdeterrence may deprive consumers of the benefits of aggressive competition—courts have been increasingly willing to adopt potentially underinclusive, but simple and objective costbased legal rules”. 176 Brooke Group, supra note 145, at 226. 177 See Weyerhaeuser, supra note 155, 127 S.Ct. at 1074, where the Supreme Court explains that a broader test of predatory pricing, namely one that would include cases of above-cost price cutting, “could, perversely, chill legitimate price cutting, which directly benefits consumers”. 178 Ibid. at 1075 (citing Brooke Group, supra note 145, at 226). 173

(G) Ehlermann 08 Schweitzer

154

6/11/08

11:48

Page 154

Session One: The Sherman Act And Abuse Of Position Under Art 82

negatives”.179 Despite the much broader test for predatory pricing uncer EC law, it would be difficult to argue that this test is likely to severely chill price competition or to overdeter. The number of cases in which the Community Courts have actually found predatory pricing schemes is small.180 However, this does not mean that the same broad tests on predatory pricing, if transposed into US antitrust law, would work similarly. In assessing the risk and costs of overdeterrence, the institutional setting of antitrust enforcement must be taken into account. Indeed, US antitrust scholars have recently defended the narrow and underinclusive predatory pricing test by reference to the deterrence effects of private enforcement in the US. Given that a violation of Section 2 leads to treble damages, and given that lay juries decide questions of intent and that even a threat to sue may (due to the costs of litigation) may have a deterrence effect, the pressure for a narrow approach to predatory pricing may be strong.181 From that perspective, the current enforcement environment in the EU differs significantly. Despite efforts to strengthen private enforcement, incidents of independent private enforcement are still comparatively rare. Public enforcement dominates. Where private enforcement takes place, no juries are involved, and courts do not impose treble damages. Furthermore, a losing party will ultimately bear the costs of litigation, which diminishes the incentives to sue. While the difference in enforcement conditions goes some way toward explaining the difference between the EU and the US in this context, it may not exhaust the reasons for diverging attitudes in the field of predatory pricing. The fact that the US requires proof of actual or likely consumer harm, whereas under EU law a showing of intent to eliminate a competitor will suffice, is not adequately explained by different needs in the US and in the EU to lower or raise the hurdles faced by potential plaintiffs. More fundamentally, it reveals different views of the structure and purpose of competition law. The US predatory pricing test faithfully reflects the view that the protection of consumer welfare is the ultimate and only relevant goal. Where a competitor is harmed by below-cost pricing by a dominant firm (and even if driven by anticompetitive intent) but the objective likelihood of recoupment and thus of consumer harm appears to be small, the competitor will enjoy no protection.182 Community competition law takes a fundamentally different stance. It focuses not on the protection of a particular market outcome, but on the protection of the competitive process and of the competitors who participate in it. The latter are protected against exclusions that result not from competition on the merits, but 179

See Fox, supra note 6, at 803. See Temple Lang and O’Donoghue, supra note 9, at 125. At that time, the authors counted only three cases (AKZO, Tetra Pak II, and Deutsche Post). 181 Kovacic, supra note 8, at 53–54. 182 See Weyerhaeuser, supra note 155, at 1077 (citing Brooke Group, supra note 145, at 224). According to the Supreme Court, without successful recoupment, “predatory pricing produces lower aggregate prices in the market, and consumer welfare is enhanced”. 180

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 155

Principles of Section 2 Sherman Act and Article 82 EC 155 from the unilateral exercise of power by a dominant firm.183 Such protection of competitors is not in opposition to protection of competition, as the muchcited slogan “protecting competitors versus protecting competition” suggests. But it reflects the understanding that competition is a process that results from the exercise of individual rights. Competitors, in their exercise of economic freedom, engage in a process in which they may lose and possibly perish. However, competition law will ensure that the fate of each competitor will depend on skill, business acumen and luck, and not on the exclusionary exercise of market power by a dominant firm.184 The efficiency effects of such a concept of competition law will sometimes differ from the effects of a concept that looks directly to consumer welfare effects, as US antitrust law tends to do. Still, the concept is not necessarily less rational economically. It may somewhat weaken the incentives of firms to compete for dominance; but it will strengthen the incentives to enter markets and compete. The focus on market entry is implicit in the system of the EC Treaty and, in view of the actual market environment in Europe, it is sufficiently justified. Having sketched some “good” reasons for the differences between the US and the EC approach, it remains to refute other claims that cannot be sustained. First of all, the US and the EU do not diverge in their concern for protecting price competition. Community competition law does not strive to protect inefficient competitors. And both jurisdictions generally rely on the same economic theories in their attempt to distinguish illicit exclusion from legitimate competition. However, differences can be observed with respect to the translation of economic insights into legal rules, and the process of translation is not a technicality. It has to take into account: the relevant market conditions which may determine the likelihood that predatory pricing schemes may occur; the institutional framework of antitrust enforcement which may be relevant for assessing the risk of false positives and false negatives; and, not least, the normative structure of the law.

3. Refusal to deal and the essential facilities doctrine Another relevant example for the divergence between Article 82 and Section 2 Sherman Act is the case law on refusals to deal, and particularly the 183 See Commission Decision 97/624/EC, [1997] OJ L258/1—Irish Sugar plc, para. 34: “The maintenance of a system of effective competition does, however, require that competition from undertakings . . . be protected against behaviour by the dominant undertaking designed to exclude them from the market not by virtue of greater efficiency or superior performance but by an abuse of market power.” See also Eilmansberger, supra note 21, at 133, who explains that the purpose of Article 82 is “to ensure that the exercise of market power does not impair competitors’ possibilities to succeed or prevail on the market on the basis of superior business performance”. 184 See Mestmäcker, “Die Interdependenz von Recht und Ökonomie in der Wettbewerbspolitik”, in Monopolkommission, ed., Zukunftsperspektiven der Wettbewerbspolitik, Nomos, 2005, pp. 19 and 34–35.

(G) Ehlermann 08 Schweitzer

156

6/11/08

11:48

Page 156

Session One: The Sherman Act And Abuse Of Position Under Art 82

so-called “essential facilities” doctrine. Ian Forrester has recently emphasized the contrast between a “more liberal or minimalist approach” in the US and a “more formalistic or maximalist approach” taken by the Commission. According to Forrester, “[t]he Commission attributes comparatively lower weight to a dominant player’s freedom to run its own business, and comparatively more weight to the protection of competitors than U.S. courts”.185 Indeed, the Commission has, when faced with market access concerns, sometimes pro-actively pursued open-access policies. The Commission’s Discussion Paper on exclusionary abuses under Article 82186 proposes a general balancing approach towards refusal to deal cases which cannot be described as formalistic but which would certainly give the Commission significant discretion to implement rather broad open-access policies in innovative industries. The ECJ’s case law, on the other hand, does not confirm sweeping claims about a maximalist approach towards duties to deal.187 The limits of the “refusal to deal” doctrine under Article 82 were set out in the Bronner case.188 In Bronner, the ECJ was confronted with questions presented by the Higher Regional Court of Vienna regarding the legality of a press undertaking’s refusal to grant Oscar Bronner, the publisher of rival newspapers, access to its nationwide newspaper home-delivery scheme—the only nationwide home-delivery scheme that existed in Austria at the time. The press undertaking—Mediaprint—held a very large share of the daily newspaper market in Austria, while rival newspapers had a small circulation and were for that reason unable to put into place a competing home-delivery scheme. Nonetheless, the ECJ rejected claims of an abuse. According to the Court’s judgment, in order for a dominant company’s refusal to deal to constitute an abuse, a number of narrow preconditions must be fulfilled: access to the facility must be indispensable to carrying on the rival’s business, i.e., there must not be any actual or potential substitute for it; a duplication of the facility must be practically impossible; the refusal to deal must be likely to eliminate all competition on the part of the undertaking requesting the service (para. 38); and the refusal to deal must be incapable of being objectively justified (para. 41). In Bronner, these preconditions were not met: although there was only one nationwide home-delivery scheme, newspapers could be distributed by other means, even if they were less advantageous ones (para. 44). Furthermore, there were no technical, legal or economic obstacles to establishing a rival home-delivery scheme, and access to the facility was therefore not indispensable. In attempting to show that access to the facility was indispensable, it was not enough to argue that the establishment of a rival 185

Forrester, supra note 10, at 920. DG Competition, Discussion paper on the application of Article 82 EC to exclusionary abuses (December 2005), available at http://ec.europa.eu/comm/competition/index_en.html. 187 This is readily admitted by Forrester, supra note 10, at 920, in light of Case C-7/97, Oscar Bronner GmbH & Co. v Mediaprint [1998] ECR I-7791. 188 Cited previous footnote. 186

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 157

Principles of Section 2 Sherman Act and Article 82 EC 157 home-delivery scheme was not economically viable due to the small size of the rival newspaper (para. 45). Rather, for access to a facility to be regarded as indispensable it would be necessary “at the very least” to establish that it would not be economically viable for a competitor of equal size to duplicate the facility (para. 46). In other words, the fact that a dominant firm benefits from economies of scale in creating its own facilities is in itself no justification for obliging it to open such facilities to competitors. With this judgment, the ECJ effectively curbed certain expansionary tendencies which had previously been latent in the ECJ’s case law and in the Commission’s decision practice. Advocate General Jacobs has explained the underlying rationale for establishing a rigorous legal test. Firstly, the “right to choose one’s trading partners and freely to dispose of one’s property” is of fundamental and even constitutional value in the Member States (para. 56). Secondly, allowing a company to retain its facilities for its own use will generally be pro-competitive. If dominant undertakings could too easily be required to share their facilities with competitors, their incentive to invest in efficient facilities would be reduced, and competitors would have no incentives to develop competing facilities (para. 57). Thirdly, the purpose of Article 82 is “to prevent distortions of competition—and in particular to safeguard the interests of consumers—rather than to protect the position of particular competitors” (para. 58). It would therefore be unsatisfactory to decide refusal to deal cases only by looking at the dominant firm’s market power in the upstream market and to conclude that it is automatically an abuse for the dominant firm to reserve to itself a downstream market: “Such conduct will not have an adverse impact on consumers unless the dominant undertaking’s final product is sufficiently insulated from competition to give it market power.” (ibid.) Taking the Bronner case as an authoritative expression of modern European refusal to deal doctrine, the intellectual “rift” between EC and US law on this issue is not wide. The fundamental competition law principles and concerns expressed in Bronner are identical to the principles governing US antitrust law. These principles were summarized by the US Supreme Court in the much-debated Trinko decision in 2004.189 Trinko confirmed the so-called Colgate doctrine according to which the Sherman Act “does not restrict the long recognized right of [a] trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal”.190 Duties to share are in “some tension with the underlying purpose of antitrust law”, since they may “lessen the incentive for the monopolist, the rival, or both to invest in those economically beneficial facilities. Enforced sharing also requires antitrust courts to act as central planners, identifying the proper price, quantity, and other terms of dealing— 189 190

Supra note 114. United States v. Colgate & Co., 250 U.S. 300, 307 (1919).

(G) Ehlermann 08 Schweitzer

158

6/11/08

11:48

Page 158

Session One: The Sherman Act And Abuse Of Position Under Art 82

a role for which they are ill-suited. Moreover, compelling negotiation between competitors may facilitate the supreme evil of antitrust: collusion.”191 It follows that antitrust limitations on the right to refuse to deal must be narrowly construed. The Supreme Court took Aspen Skiing192 to be the leading US case on refusals to deal, found it to lie “at or near the outer boundary of § 2 liability” and interpreted it narrowly.193 According to the Court’s reading of Aspen Skiing, Section 2 liability will only lie where a voluntary—and thus presumably profitable—course of dealing is terminated unilaterally under circumstances that suggest a willingness to forsake short-term profits to achieve an anticompetitive end.194 In cases that fall outside the unilateral termination of a voluntary course of dealing scenario, Section 2 liability could be established only based on the essential facilities doctrine. In Trinko, the Supreme Court found it unnecessary to either recognize or repudiate the “essential facilities” doctrine, since it would not have been applicable in any case.195 However, the tone of the judgment suggests a sceptical attitude. With regard to any extension of Section 2 liability beyond the Aspen Skiing precedent, the Supreme Court underlines the need to perform a cost-benefit analysis: “Against the . . . benefits of antitrust intervention . . ., we must weigh a realistic assessment of its costs.” The costs, the Supreme Court finds, will be significant: in applying Section 2, courts confront significant difficulties because “the means of illicit exclusion, like the means of legitimate competition, are myriad”.196 Any mistaken inferences and the resulting false condemnations will be “especially costly, because they chill the very conduct the antitrust laws are designed to protect”.197 Furthermore, remedying refusals to deal may require continuous supervision by courts, a task which—depending on the concrete case—may be beyond their practical ability.198 These high costs, according to the Court, militate against any “undue” expansion of Section 2 liability. The comparison between Bronner and Trinko shows the many similarities between the EU and US approaches—and some differences. An obvious dif191

Trinko, supra note 114, at 407–408. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 601 (1985). 193 Trinko, supra note 114, at 408–410 (2004). 194 Ibid. at 409. In Aspen Skiing, the defendant had been unwilling to renew a joint skiing ticket cooperation, even if compensated at retail price. This indicated anticompetitive intent. 195 See Trinko, supra note 114, at 410–411 (citing Phillip Areeda and Herbert Hovenkamp, Antitrust Law, Aspen Publishing, p. 150 # 773e (2003 Supp.): “essential facility claims should . . . be denied where a state or federal agency has effective power to compel sharing and to regulate its scope and terms”). 196 Ibid. at 414 (citing United States v. Microsoft Corp., 253 F.3d 34, 58 (D.C. Cir. 2001) (per curiam)). 197 Ibid. (citing Matsushita, supra note 145, at 594). 198 Ibid. (citing Phillip Areeda, “Essential Facilities: An Epithet in Need of Limiting Principles”, 58 Antitrust Law Journal 841, 853 (1989) (“No court should impose a duty to deal that it cannot explain or adequately and reasonably supervise. The problem should be deemed irremediable by antitrust law when compulsory access requires the court to assume the day-today controls characteristic of a regulatory agency.”). 192

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 159

Principles of Section 2 Sherman Act and Article 82 EC 159 ference relates to the acceptance of an essential facilities doctrine: whereas Trinko left open the question whether Section 2 liability for refusals to deal may exist in cases where no prior voluntary course of dealing can be established, the possibility of such liability is well-accepted in the case law on Article 82. Indeed, this doctrine has played a formidable role in liberalizing European state monopolies. The greater concern under EC competition law with state monopolies may be one of the reasons why the essential facilities doctrine has resonated more strongly in Europe, although it is technically a legal import from US antitrust law.199 Other differences are more subtle. Both Bronner and Trinko emphasize the importance of protecting the freedom to deal or not to deal with a view to the negative effects any duty to share will have on long-term incentives to innovate and invest, and thus on incentives to compete. While Advocate General Jacobs in his Conclusions in Bronner also emphasizes the constitutional importance of “the right to choose one’s trading partner”, such a reference to freedom of contract is absent in Trinko. Instead, the Supreme Court stresses and elaborates the high costs of “false positives”—as it did in the predatory pricing cases. But as in those cases, it fails to specify the potential costs of “false negatives”. A significant divergence between the EU and the US may be seen with regard to the application of Article 82 and Section 2 to intellectual property rights. In Europe, Magill and IMS Health stand for an essential facilities approach which has never been accepted in IP cases by US courts. Both in the EU and in the US, the relevant case law on access to IP rights is currently highly controversial. On both sides of the Atlantic, it remains an evolving area of law.200 Finally, it must be pointed out that the approach towards refusal to deal cases proposed in the Commission’s Discussion Paper on Article 82 201 deviates substantially from existing ECJ case law. The general balancing test recommended in the Discussion Paper would give broad discretion to the Commission to establish open-access policies under Article 82. It would generalize the test applied by the Commission in the Microsoft case.202 New insights on the degree of convergence or divergence in this area of law may therefore follow from the CFI’s Microsoft decision which is expected this fall. In summary, the picture regarding convergence and divergence in the EU and US attitudes toward refusals to deal is mixed. The ECJ and the US 199

See United States v. Terminal Railroad Association, 224 U.S. 383 (1912). For a more detailed analysis, see Heike Schweitzer, “Controlling the unilateral exercise of intellectual property rights: a multitude of approaches but no way ahead?”, EUI Working Paper LAW 2007/31, available at http://cadmus.iue.it/dspace/handle/1814/7625. 201 Supra note 186. 202 Commission Decision of 24 March 2004, Microsoft, http://ec.europa.eu/comm/competition/ index_en.html, on appeal: Case T-201/04, Microsoft v Commission, not yet decided as of this writing. 200

(G) Ehlermann 08 Schweitzer

160

6/11/08

11:48

Page 160

Session One: The Sherman Act And Abuse Of Position Under Art 82

Supreme Court entertain similar reservations against finding antitrust liability in such cases; but the ECJ has nonetheless been somewhat more pro-active than the US Supreme Court. The Commission has frequently tended towards an even broader “open access” approach, at least in those industries in which a market-access problem has previously been identified. While this, as well as the application of Article 82 to IP cases, can certainly be criticized, the more general and harsh criticism which EU competition law’s refusal to deal case law has at times faced appears to be unfounded in view of the current state of the law: Community law as applied to refusals to deal does not stand for a general concern with fairness instead of efficiency, it does not tend to protect competitors instead of competition, and it is well aware of the regulatory dangers involved in the imposition of broad duties to deal.

IV. Conclusions: Comparative insights As can be seen from the foregoing discussion, a comparative look at the history and current application of Section 2 Sherman Act and Article 82 reveals important commonalities, but also important differences in the attitudes towards rules on market power. The picture that emerges is much more nuanced than the standard story suggests. In fact, the standard story is in many respects wrong or at least misleading. Firstly, the difference in language to which the divergence between EU and US antitrust rules is sometimes attributed appears to be less relevant for practical purposes than is often claimed. In contrast to Article 82, Section 2 Sherman Act prohibits incidents of monopolization or attempted monopolization by a firm irrespective of its current market position—but instances of monopolization without a prior position of market power are rare. In contrast to Section 2, Article 82 covers exploitative abuses, but the prohibition is seldom applied. In practice, both provisions are mainly applied to exclusionary abuses by firms in a position of dominance. Secondly, the EC competition rules, and particularly Article 82, are not driven by fairness concerns that may be distinguished from the goal of protecting competition. Like US law, EC law respects a dominant firm’s right to forcefully compete on the merits. It does not strive to insulate inefficient competitors from competition. Like US antitrust law, EC law struggles with formulating an adequate test for distinguishing between pro-competitive conduct, or “competition on the merits”, on the one hand, and anticompetitive, exclusionary conduct on the other.203 This is by its very nature a difficult 203 For the difficulty of formulating such a test, see, e.g., Hovenkamp, supra note 35, at 24; Antitrust Modernization Commission, supra note 22, at 81. See also Evans and Padilla, supra note 22, at 73: “. . . the welfare effects of unilateral practices are inherently difficult to assess”.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 161

Principles of Section 2 Sherman Act and Article 82 EC 161 task, given that: the methods of exclusion and competition on the merits will frequently be the same;204 the perfectly legitimate and pro-competitive intent to outperform, and thereby damage or even eliminate competitors may be difficult to distinguish from an intention to exclude by anticompetitive means;205 and assessing a dominant firm’s conduct therefore requires a thorough inquiry into a firm’s conduct and an appraisal thereof based both on intent and likely effect, where both can be uncertain in practice. The current reform initiatives, both in the EU206 and the US,207 stand for another attempt to get the distinction between “competition on the merits” and anticompetitive exclusionary conduct right. Like their brethren in the US, competition policy makers in the EU acknowledge the relevance of economic theory in pursuing this task. However, in translating these economic insights into legal rules, EU policy makers face a somewhat different legal framework: Article 82 protects the “institution” of competition, the competitive process itself, instead of making consumer harm the ultimate reference point. This concept is enshrined in the EC Treaty itself: it follows from Article 3(1)(g) EC, which makes a “system ensuring that competition in the internal market is not distorted” one of the fundamental Treaty goals.208 It is a particular normative underpinning of EC competition law, and not a policy choice that competition lawyers or the Commission would be free to change. This does not imply that EC competition law is insensitive to the concept of efficiency. Rather, it is grounded in the conviction that the undistorted competitive process will generally tend to maximize wealth and consumer welfare, at least in the medium term.209 Judge Posner appears to be an unsuspecting witness for the rationality of the EU’s policy choice. Efficiency, he says, “is the ultimate goal of antitrust, but competition [is] a mediate goal that will often be close enough to the ultimate goal to allow the courts to look no further”.210 A further aspect in which EC competition law markedly differs from the US is its understanding that the process of competition flows from the exercise of 204 Examples are low pricing or refusals to deal. For the more general claim, see Franz Böhm, Wettbewerb und Monopolkampf, Berlin 1933, pp. 9–10; Mestmäcker, supra note 101, at 6: “. . . die rechtliche Behandlung von Marktmacht ist vor allem deshalb so schwierig, weil Inhaber von Macht befähigt sind, die Institute des Privatrechts und die unter den Voraussetzungen freier Konkurrenz legitimen Mittel des Wettbewerbs in den Dienst der Marktbeherrschung zu stellen”. 205 Antitrust Modernization Commission, supra note 22, at 81: “. . . companies routinely attempt to ‘exclude’ competitors from the market simply by producing the best quality product at the lowest price. Accordingly, an observation that a particular firm’s conduct ‘excludes’ its competitor does not answer whether the conduct is harmful to competition or just to the firm’s competitor.” 206 See Discussion Paper, supra note 186. 207 See in particular the transcripts from the hearings on single-firm conduct currently conducted by the FTC, available at http://www.ftc.gov/os/sectiontwohearings/index.shtm. 208 For the relevance of Article 3(1)(g) for the interpretation of Article 82, see Continental Can, supra note 21, para. 23. See also Hoffmann-La Roche, supra note 97, para. 38; Eilmansberger, supra note 21, at 132. 209 Eilmansberger, supra note 21, at 135. 210 Posner, supra note 41, at 29.

(G) Ehlermann 08 Schweitzer

162

6/11/08

11:48

Page 162

Session One: The Sherman Act And Abuse Of Position Under Art 82

individual rights. This is an additional reason why EC law cannot make consumer harm the ultimate test of anticompetitive conduct, as US antitrust tends to do. The EC law approach has been linked to Kantian philosophy.211 Legally, it flows from the EC Treaty’s fundamental purpose: EC competition law protects the opportunities to compete on the merits that result from the realization of the free movement rules. In doing so, it protects the individual rights of those who compete. This does not imply, however, that Article 82 protect competitors instead of competition. Rather, it protects competitors as a part of the competitive process, and only against those harms which are not part of normal “competition on the merits” but result from exclusionary, nonmerit-based acts. The objection that this, and the accompanying focus not on consumer harm but on harm to competition, will lower the threshold of liability, curbing the dominant firm’s incentives to compete and hence chilling competition, can be countered by the observation that an effective protection of competitors against exclusionary acts will increase the incentives of nondominant market players and potential newcomers to invest and compete. This may be particularly valuable in a market environment where the barriers to enter foreign markets frequently remain significant. This leads to another enduring feature of EC competition law, namely its close links with the market integration goal. The nexus between competition principles and market integration in EC law is increasingly criticized.212 However, as with the concept of competition embedded within the Treaty itself, the integration objective is not a feature of the law that the Commission would be free to give up: EC competition law is part of primary Treaty law and is functionally intertwined with the Treaty’s goals. Furthermore, the European focus on market integration cannot be condemned as mere ideology. While the successes of market integration in the EU are immense, the state of the internal market is still a far cry from the unity of the market in the US. National boundaries remain a reality in the EU—a reality that the Sherman Act does not have to struggle with. The conviction, so strongly ingrained in US antitrust law, that markets tend to be self-correcting,213 has never had the same appeal in the EU, where market boundaries frequently still follow the national territories of Member States. This is one of the reasons why Chicago school thinking, based on the assumption that robust, efficiently integrated markets exist and will erode any barriers that might be erected for limited periods of time, has not been perceived as relevant in the 211 Mestmäcker, “Bausteine zu einer Wirtschaftsverfassung—Franz Böhm in Jena”, in ders., Wirtschaft und Verfassung in der Europäischen Union, 2 ed., 2006, pp. 116, at 123–127. 212 For a telling view, see Forrester, supra note 10, at 926, who speaks of market integration as a “civil religion” or a “cult”. 213 See, e.g., Frank Easterbrook, “The Limits of Antitrust”, 63 Texas Law Review 1, 15 (1984). According to Judge Easterbrook, while there is no automatic way to correct wrong decisions of the Supreme Court, and while bad law is thus likely to stick, “[a] monopolistic practice wrongly excused will eventually yield to competition . . . as the monopolist’s higher prices attract rivalry”. In a somewhat weaker version, the same claim is made by Evans and Padilla, supra note 22, at 83–84.

(G) Ehlermann 08 Schweitzer

6/11/08

11:48

Page 163

Principles of Section 2 Sherman Act and Article 82 EC 163 context of EC competition law.214 The different degree of market integration is certainly one of the reasons for the somewhat more proactive attitude towards positions of market power under Article 82 as compared to Section 2 Sherman Act. The actual risks of (successful) exclusion, and consequently the costs of underinclusive tests, may simply be significantly higher in the EU. Another important difference between EC competition law and US antitrust law concerns the different structure of enforcement institutions on which they rely. Institutions and patterns of enforcement, sanctions and procedural rules influence the optimal design of substantive competition law. As seen earlier, renowned US antitrust scholars have stressed that underinclusive rules on single-firm conduct may in fact emerge as a reaction to an environment which provides (overly) strong incentives for private enforcement by making treble damages available and which charges lay juries with the task of applying the relevant tests.215 Indeed, such rules, as well as the high costs of litigation borne by each party irrespective of the outcome of litigation, may influence the cost-benefit analysis that underlies the design of the applicable rules. From that perspective, the US Supreme Court’s persistent concern with rules that avoid “false positives” may well have its reasons. In the EU, on the other hand, the conditions of competition law enforcement are very different. Competition law in Europe thus far has not relied on private enforcement even remotely to the same extent as in the US. Here in the EU, risks of overdeterrence that may result from treble damages or from the alleged unreliability of juries are not a reality. The fact that the costs of litigation are borne by the losing party reduces the extortionary potential of competition law claims. In the absence of a litigation-based rationale for underinclusive rules, there may be good reason for EC competition law to formulate tests that strive to capture instances of abuse of dominance more comprehensively and systematically than is currently the case in US antitrust law. 214 Sir Leon Brittan, European Competition Law: Keeping the Playing-Field Level, Brassey’s, 1992, p. 3. 215 See, e.g., Kovacic, supra note 8, at 63–64: “. . . a fear that mandatory treble damages could provide excessive compensation and create over-deterrence may have induced the courts to design and apply liability standards in a manner that diminishes the private litigant’s prospects for success. The Harvard-inspired forms of judicial ‘equilibration’ to constrain private plaintiffs—the adjustment by the courts of the malleable features of the US antitrust system to offset perceived excesses in characteristics (e.g., mandatory trebling of damages and availability of jury trials) not subject to judicial alteration—can have far-reaching consequences well beyond the resolution of private antitrust cases. This is certainly the case where the method of equilibration is to alter liability rules. The establishment of more permissive substantive liability rules has systemwide effects. The non-intervention presumptions of liability standards that constrain the prosecution of private antitrust cases encumber public authorities alike”. See also Hovenkamp, supra note 35, at 45 et seq., and for a broader critical survey of the private antitrust enforcement regime in the US, see ibid. at 57 et seq. Of particular relevance is Hovenkamp’s summary at p. 76: “Often judges respond to an overly aggressive remedies system by defining substantive violations too narrowly . . . The result often gives us the worst of both worlds, a substantive system that fails to prosecute anticompetitive practices that it is capable of prosecuting, and a remedies system that strikes haphazardly while leaving other, equally serious practices undeterred.”

(G) Ehlermann 08 Schweitzer

164

6/11/08

11:48

Page 164

Session One: The Sherman Act And Abuse Of Position Under Art 82

Stepping back and surveying the differences between Article 82 and Section 2 Sherman Act, many of them appear to be strongly rooted in the normative structure of the rules that apply in different market realities and in different enforcement environments. All of these reasons are valid ones. None of them is incompatible with the effective competition of competitive markets or based on “unsound” economics. German ordoliberalism, so frequently blamed for having infected the Community’s approach to competition with outdated economic theory, has certainly been influential in shaping the law.216 But its influence has not consisted of infusing non-economic fairness concerns or a regulatory philosophy into the interpretation of Article 82. What is indeed close to German ordoliberal thought, rather, is the conception of the competitive process as a process resulting from the exercise of individual economic liberties. While this is contrary to US Chicago school thought, such a concept is by no means irrational or in contradiction with modern economic theory. Many outstanding economists have defended this approach.217 Against this background, it is questionable whether we should uncritically follow calls for convergence between US antitrust law and EC competition law.218 As long as the reasons for divergence are clearly articulated and explained, there appear to be good reasons for them to persist.

216 One of the important influences probably came in the person of Ernst-Joachim Mestmäcker, who was special advisor (sonderberater) to Hans von der Groeben, the first Commissioner for competition, during the formative years and until 1970. 217 See Martin Hellwig, Recht und spontane Ordnung, FS für Ernst-Joachim Mestmäcker zum 80. Geburtstag, Nomos, 2006, pp. 233 et seq.; Amartya Sen, Rationality and Freedom, Harvard University Press, 2003, pp. 3 et seq. 218 For such calls see, inter alia, Temple Lang and O’Donoghue, supra note 9, at 85. In their view, differences between US and EC antitrust law “should be minimized where possible”. See also William J. Kolasky, “What is competition? A comparison of U.S. and European perspectives”, supra note 7, at 53, according to whom “[a] divergence of policies can only breed chaos and confusion, unless there are clearly articulated reasons for the differences”.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 165

VI James S. Venit* Cooperation, Initiative and Regulation— A Cross Cultural Inquiry May 2007

Nietzsche begins his essay on history1 by describing some decidedly ahistorical cows ruminating in a field. Men are not cows. A humble private in the trenches of World War I suddenly begins to consume history books and when asked why points to the shells landing around him and replies: “I am trying to understand how the devil I got here.” The “here” I would like to explore is the paradox of our European antitrust law and why it has gone in directions, particularly as concerns single firm conduct, that themselves appear to be anti-competitive. In what follows I shall seek to develop three ideas: First, the EU and US antitrust agencies do not always mean the same thing when they speak of “competition”, and this is particularly but not exclusively the case when it comes to dealing with single firm conduct in the context of Section 2 of the Sherman Act and Article 82. Second, the difference in approach may not primarily be driven by the difference between a focus on short versus long-term effects, but rather by differences in cultural attitudes about the propriety and social implications of aggressive, individualistic competition and the ability of markets to solve their own problems in the long run. Third, societies that have not experienced robust competition and unrestricted markets tend to have less confidence in both and thus are prone to adopting regulatory approaches that not only reflect this lack of confidence but also tend to reinforce the lack of competition with the negative consequences common to non virtuous circles. In Europe this has led to the formulation of a set of very restrictive rules that appear better designed for regulating the conduct of public monopolies than that of firms who have had to earn their success by competing.

Before beginning, two health warnings. First, generalizations and stereotypes are always dangerous, probably the more so as their apparent accuracy increases. This is particularly the case because trends in enforcement change. The US agencies and courts of the 1960s have also been criticized as too interventionist, and there is no guarantee that current attitudes will remain fixed * Partner Skadden, Arps, Slate, Meagher & Flom LLP and Affiliates. 1 Untimely Meditations: On the Uses and Disadvantages of History for Life (R.J. Hollingdale, trans.), Cambridge University Press, 1983.

(H) Ehlermann 08 Venit

166

6/11/08

11:49

Page 166

Session One: The Sherman Act And Abuse Of Position Under Art 82

in either Europe or the US.2 Second, although what follows may reflect my view that the current European approach to single firm conduct is too interventionist, this article is not an argument for or against one type of social philosophy. Rather it is an attempt to understand, by looking backward and with some sensitivity to cultural biases, how we got to where we currently are.

I. Nineteenth Century Attitudes Toward Cartels and Competition in Germany To understand modern European attitudes toward competition and the role of competition law, it is important to understand the cultural attitude toward cooperation and unbridled competition that evolved in the 19th Century. This story largely centres on Germany and Austria, partly because cartelization does not appear to have been regarded as favourably elsewhere and partly because German historical and cultural attitudes are of particular importance given Germany’s position as the intellectual source for much of what came to be enshrined in the competition provisions of the Treaty of Rome. Review of this history reveals a pervasive cultural attitude toward cartels and single firm economic power whose traces are still present in European antitrust law today. This cultural attitude is reflected in the contrast between cooperation and organization, which were viewed as positive attributes as opposed to individualism and conflict which were not. As will be seen from the discussion that follows, this attitude gave rise to a greater tolerance of cartels in Austria and even more so in Germany during the 19th Century and to hostility to the exercise of single firm economic power which can be traced to modern European law by such manifestations as the decision not to criminalize cartel conduct, reliance on abuse control and the view that single firms exercising economic power have special obligations and should be subjected to special controls. At the beginning of the twentieth century it is estimated that there were about 400 cartels in Germany, which had become known as “The Land of Cartels,”3 with many of these including industrial leaders rather than 2 The fluidity in US attitudes is not inconsistent with the thesis that culturally the US has a greater tolerance than Europe for aggressive individualism. Populism is also an important constituent of the US cultural experience and is not unrelated to the tendency of the US legal system to act as a vehicle for redistributing wealth. Indeed, the very fluidity of US attitudes is supportive of the view that mobility, both social and individual has played a greater role in the US than here. 3 David Gerber, Law and Competition in Twentieth Century Europe, Oxford University Press, 1998, p. 75. Möschel estimates that by 1907, 82% of the coal industry, 50% of crude steel, 90% of the paper industry and 48% of the cement industry were, according to a Reichstag report,

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 167

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 167 defensive alliances of smaller firms.4 The pervasiveness of German cartelization can be traced to a number of economic, political and cultural factors. Among the most important were the sheer speed of German industrialization, the central role played in the German economy by industries well suited for cartel activity such as chemicals, coal and iron, the high degree of industrial concentration, vertical integration, the role of German banks, who preferred ‘stable’ market conditions, and a high proportion of German export activity which led cartelists to defend their existence as necessary to ensure global competitiveness.5 Added to these factors were political elements—the general historical weakness of German liberalism,6 the belief in the necessity of “organizing” the economy, the alliance between business and the State in the task of unification and thereafter a perceived tie between German economic performance and military strength.7 Other forces also contributed to the positive evaluation of cartels. Especially important was the fact that cartelization was largely seen as a positive phenomenon by 19th Century German economists. These economists, all of whom belonged to the “historical” school8 eschewed theoretical approaches and opposed what they referred to as “Manchesterism”, a reductionist version of classical liberal, Anglo-Saxon economics. For these German economists, cartels were both an appropriate and natural response to overproduction and instability and were thus regarded as useful in combating unwanted cyclical fluctuations of the economy.9 There was also an “ethical” or social perspective to these economists’ evaluation of cartels which led to their being viewed as a positive expression of a cooperative social spirit that reduced class conflict.10 Although these economists were fully aware of the negative economic consequences of cartelization, they nevertheless viewed cartels as encouraging positive attributes such as restraint and concern for weaker competitors, and they contrasted this more “ethical” approach with the American trusts, which were viewed as purely exploitative.11 This attitude cartelized. Government rationing gave a further impetus to cartelization during the First World War, and a Cartel Ordinance adopted in 1923 at the height of German hyper-inflation generally encouraged cartels, as did the National Socialists who saw them as useful for planning production. See Wernhard Möschel, “Competition Policy from an Ordo Point of View”, in Alan Peacock and Hans Wilgerodt, eds., German Neo-Liberals and the Social Market Economy, St. Martin’s Press, 1989, pp. 143 et seq. 4 Gerber, cited supra note 3, at 74. 5 Ibid. at 74–75. 6 Ibid. at 72–76. 7 Ibid. at 75 8 Academic life was sufficiently well organized and supported by the Prussian state that economists who did not share the views of the historical school, and were instead interested in a theoretical approach, were unable to secure university appointments in Germany. See ibid. at 83 (citing Friedrich Lutz, Verstehen und Verständigung in der Wirtshaftswisesenshaft, Tubingen, 1967, p. 18). 9 Ibid. at 86. 10 Ibid. at 87. 11 Ibid. at 87–88.

(H) Ehlermann 08 Venit

168

6/11/08

11:49

Page 168

Session One: The Sherman Act And Abuse Of Position Under Art 82

was also manifested in a notable hostility to monopoly, which in economic theory seems difficult to reconcile with a favourable attitude toward cartels. This inconsistency appears to have been the result of the unique ethical and social lens through which cartels were being evaluated, and may have in it the seeds of more modern European theories of economic dependency and the special obligations of dominant firms. The generally favourable view of cartels was also shared by the first German court judgment on the subject, the Saxon Wood Pulp case, decided in 1897.12 In this judgment, the highest German court, relying on the economic literature, concluded in language that anticipates the crisis cartels of the 1970s and 1980s under the Coal and Steel Treaty, that cartels were beneficial in that they preserved firms from bankruptcy and enabled them to maintain adequate prices. As the court put it: “if the firms in a particular branch band together to eliminate or control price reductions among themselves, their cooperation can be seen not only as a justified application of the drive to self-preservation, but also—as a general rule—a service to the public, provided that such prices are continuously so low that economic ruin threatens the firms.”13

The Reichsgericht did, however, note that cartels might be illegal in two circumstances: (i) where they were obviously intended to establish an actual monopoly and to exploit consumers unfairly; or (ii) where these results were actually achieved by the cartel agreements.14 Attitudes in Austria were similar. There too, at the fin de siècle, one finds a similar cultural preference for cooperation and a distrust of unbridled competition. In the case of Austria, these choices were anchored in the collectivist values of Austrian society fostered by cultural and political traditions, not least among them, the Catholic church.15 What is important for the discussion that follows is the connection between a positive cultural disposition toward cartels as an ethical manifestation of a benign collectivist and social spirit and thus as a beneficial form of cooperation and a hostility toward unrestrained competition and a general condemnation of aggressive, individualistic conduct, which, as noted above, was frequently associated with American trusts. As Gerber puts it: “. . . cartels were basically [viewed as] positive institutions. They were seen as superior to both excessive competition and alternate forms of organization such as trusts, because they assured and rewarded important social values such as cooperation rather than baser forces such as individualism and egoism, thus reducing conflicts among competitors and protecting rather than exploiting the weak. This

12 13 14 15

Reichsgericht, [1897] 38 RGZ 155 (Sächsisches Holzstoffkartel). Ibid. at 157, cited in Gerber, supra note 3, at 92. Supra note 12 at 158, cited in Gerber, supra note 3, at 93. Gerber, supra note 3, at 64.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 169

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 169 basic perception of cartels and the competitive process persisted in many parts of Europe until recently, and its echoes continue to reverberate.”16

As we shall see, these reverberations continue to be felt most forcefully today not in the area of cartel conduct, which is now generally viewed as exploitative (although generally not worthy, on the continent, of criminal condemnation outside the area of public bid rigging), but more importantly in the attitude toward aggressive and individualistic competition by single firms.

II. The German Influence on EU Competition Law The special attitude towards cooperation and unbridled competition takes on importance because the road to modern European competition law passes, as should by now be well known, through Germany. And although the rise of the Third Reich, the Second World War and its aftermath were watershed events that strongly influenced the development of German competition law and shook it loose from a large part of its earlier heritage, some elements remained. As we shall see, these reappeared in the competition law which emerged in Germany and which, in turn, largely dominated the development of European competition law in both its formative and subsequent stages. There is, of course, an irony in the current transatlantic discussion about EU and US enforcement standards, since the United States played a significant role in the introduction of a modern competition law to the German legal order following the Second World War. However, German law took its own form and, once it had been adopted, its own direction. The German experience, in turn, had a massive influence on the content and interpretation given to Community law as first formulated in Articles 85 and 86 of the Treaty of Rome and in Regulation 17/62. Although Germany’s adoption of an effective competition law was strongly influenced by the victorious western allies, the forced adoption of decartelization measures imposed by the US and British authorities during the early years of the occupation never fully uprooted or replaced indigenous competition law traditions. As a result, the German antitrust law that was finally adopted in 1957, the same year as the Treaty of Rome, does not strongly resemble US law. Conversely, although there are signs that this influence is now finally weakening, German law and German thinking had a decisive formative influence on the development and early interpretation of EU antitrust law.

16

Ibid. at 111.

(H) Ehlermann 08 Venit

170

6/11/08

11:49

Page 170

Session One: The Sherman Act And Abuse Of Position Under Art 82

The story is told, inter alia, by David Gerber in his 1998 book Law and Competition in Twentieth Century Europe: Protecting Prometheus,17 and is more or less as follows. Once the US abandoned the so-called “Morgenthau plan”, which would have reduced Germany to an agrarian economy, the US and British occupying forces imposed decartelization measures which remained in force until the German competition law was adopted in 1957 after a prolonged, nearly decade-long period of gestation and debate. The law entered into force on 1 January 1958.18 The interim US- and British-imposed anti-cartel legislation was nearly identical. The US version provided that “[e]xcessive concentrations of German economic power, whether within or without Germany and whatever their form or character . . . are prohibited, their activities are declared illegal and they shall be eliminated” except where approved by the military government. Excessive concentration of power included “cartels, combines, syndicates, trusts, associations or any other form of understanding or concerted action between persons which have the purpose or effect of retraining, or of fostering monopolistic control of, domestic or international trade or other economic activity”.19 The reason for the insistence on a strong antitrust law as part of the new post-War order was the perception, shared by many Germans including, as we shall see, the Freiburg school of law and economics (the “Ordoliberal” school), that the pre-War domination of the highly regulated German economy by cartels and powerful single firms had fostered an essentially anti-democratic spirit that had contributed to the rise of National Socialism and eventually the War.20 The fact that German industrial barons had thrown their support behind Hitler at a period when he was still struggling for political legitimacy did 17 See supra note 3. For a shorter version, see also David Gerber, “Constitutionalizing the Economy: German Neoliberalism, Competition Law and the ‘New’ Europe”, 42 American Journal of Comparative Law 25 (1994). 18 These decartelization measures were unpopular to the extent they were perceived as victors’ justice, and some German academics and officials feared that they had a negative effect on German acceptance of forceful indigenous anti-cartel laws. However, inasmuch as industrial resistance to strong anti-cartel legislation delayed its adoption for nearly a decade and resulted in a somewhat watered down version of the original proposal, it is fair to say that there was some fairly entrenched resistance to anti-cartel measures at least on the side of German industry. 19 See Prohibition of Excessive Economic Concentration of Economic Power, Law No. 56 (1947) (Germ.). 20 As noted above, by 1907, 82% of the coal industry, 50% of crude steel, 90% of the paper industry and 48% of the cement industry were, according to a Reichstag report, cartelized. Government rationing gave a further impetus to cartelization during the First World War, and a Cartel Ordinance adopted in 1923 at the height of German hyper-inflation generally encouraged cartels, as did the National Socialists who saw them as useful for planning production. See Möschel, supra note 3 at 143. The general favourable 19th Century attitude toward cartels was accompanied by a hostile attitude toward monopoly which, in economic theory, seems difficult to reconcile absent the unique ethical and social lens through which cartels were evaluated. This hostility to single firm power appears also to have been the forerunner of modern European theories of economic dependency and the special obligations of dominant firms. Gerber, supra note 3, at 86–89.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 171

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 171 nothing to dispel this view. The introduction of a strong competition law was thus seen by the US occupation authority as one of the pillars that would anchor Germany in the western alliance and stimulate both democracy and economic recovery. Many progressive Germans in the academic world and legal profession (particularly the ordoliberals at the University of Freiburg, who had been brooding on these issues at their own risk even prior to and during the War)21 shared this view and also viewed laws that would curb undue economic power as part of the creation of a rule of law that had been missing under previous regimes. This explains the constitutional anchoring of the competition system and the importance of legal certainty for German thinkers and antitrust pioneers in the post-War period. As discussed below, it also accounts for the importance attached to small and medium-sized firms as the backbone of a democratic market economy and as a counterbalance to the monopolistic and perceived anti-democratic tendencies of large firms. American insistence on the adoption of strong antitrust law as part of the new post-War order did not extend to specifying the terms of that law. As a result, the law that was finally adopted in 1957—nearly a decade after the debate on the new law started22—reflected unique German perspectives and traditions and thus bore little conceptual or theoretical resemblance to US antitrust law.23 21 The original leading Ordoliberal thinkers were the economist Walter Eucken and two lawyers, Franz Böhm and Hans Grossman-Doerth. 22 The reason for the delay was the resistance of German industry to a tough anti-cartel law. This resistance led to the inclusion of certain statutory exemptions for export cartels, specialization cartels, and cooperation between small and medium-sized firms that weakened the basic prohibition principle. 23 This despite the fact that, as in the Sherman Act, two of the key provisions of the Gesetz gegen Wettbewerbsbeschränkungen (the “GWB”) dealt with horizontal agreements and single firm conduct, with the third treating vertical agreements. Gerber describes the GWB as “a hybrid that reflected, in addition to ordoliberal ideas, contacts with the US antitrust law and residual influences from prior German experience”. Gerber, supra note 3, at 277. When a number of German officials charged with enforcing the new law visited their US colleagues in the 1950s, one of their primary interests, other than the practicalities and tactics of enforcement, were the price discrimination provisions of the Robinson-Patman Act. See Gerber, supra note 3, at 269–270. According to Gerber, the assumption that the Allied decartelization laws were the “model” for the German competition statute adopted in 1957 “is highly misleading”: “The role of the decartelization laws in the shaping of the German competition law should not . . . be overestimated. . . . The decartelization laws did not ‘become’ the new German competition law in the sense that the laws were identical or even similar, as even a cursory look at the texts reveals. . . . The decartelization laws . . . were not referred to as a model for the new law, and at least in the legislative debates they were seldom used as a reference point for discussion. In addition, the laws were often not strictly enforced . . . [n]or did they become part of German legal culture, retaining for many the ‘odor of an imposed system.’ Finally, they did not introduce German decisionmakers to the idea of a strong, prohibition-based competition law; . . . this was done by the Ordoliberals, whose ideas were already at the center of discussions of competition law.” Ibid. at 269–270. However, it also appears that the US model was frequently cited in defence of the need for a strict anti-cartel law during the long debate between the submission of the first German draft in 1949 and its adoption in 1957. See ibid. at 271. But see also Möschel, who takes the view that “[t]he practical application of the GWB was closely modeled on the antitrust legislation of the United States and it never lost its stigma as a foreign law imposed by the occupation powers”. Möschel, supra note 3, at 145.

(H) Ehlermann 08 Venit

172

6/11/08

11:49

Page 172

Session One: The Sherman Act And Abuse Of Position Under Art 82

Concern about the causal relationship between excessive concentration of economic power and anti-democratic tendencies had already found particularly fertile ground among the intellectuals and academics who comprised the Freiburg school during the War and these thinkers succeeded in implanting many of their philosophical concerns and approaches in the German competition law that was ultimately adopted. Some of the ordoliberals’ fundamental assumptions, particularly concerning single-firm economic power, appear to have had their roots in some of the collectivist socio-ethical attitudes toward competition discussed above, which involved a strong distrust of aggressive, individualistic competition, as well as the distrust of large firms which stemmed, inter alia, from the perceived role of those firms in assisting Hitler’s rise to power. The first component may be explained by the fact that ordoliberalism’s focus was primarily humanistic rather than being rooted in efficiency or economic values.24 The second is the product of one of the twentieth century’s great disasters. Ordoliberalism’s four constituent elements have recently been described as: (i) a competition policy primarily oriented toward the goal of securing individual freedom of action as a value in itself, from which the goal of economic efficiency is merely derived; (ii) a strong role for the State in the preservation of the prerequisites of the competitive system but hesitancy towards overt price regulation; (iii) the shaping of competition policy into a rule of law rather than a mechanism for discretionary decisions; and (iv) the embedding of competition policy into the economic order of a free and open society.25 In particular, ordoliberal values do not rely “on the long-term process of selfhealing of the overall society”. Instead, they protect “the individual economic freedom of action as a value in itself against any impairment of excessive economic power”.26 Many of the same ordoliberal thinkers who played a central role in the adoption and implementation of German antitrust law also very heavily influenced the drafting and interpretation of the competition provisions of the Treaty of Rome.27 This helps explain why the protection of individual freedom of action as a value in itself and the concept of “performance-based competition” have played such central roles in the interpretation of Articles 81 and 82 of the EC Treaty and continue to play a major role in German com24

Gerber, supra note 3, at 36. See Möschel, supra note 3, at 142. 26 Ibid. at 147. Or as was said by Walter Adams: “The freedom of the sheep to coexist with the wolf is meaningless in the absence of a shepherd.” Walter Adams, “Antitrust, Laissez-faire and Economic Power”, in Fritz Neumark et al., eds., Wettbewerb, Konzentration, und wirtschaftliche Macht, Duncker & Humboldt, 1976, at 13. 27 Gerber quotes Otto Schlecht, a leading Germany economic policy official at this time, as claiming that, “[w]ithout this battle [for the GWB] there probably never would have been the prohibition of cartels or the abuse supervision in the EC Treaty”. Gerber, supra note 3, at 344 (citing Otto Schlecht, “Macht und Ohnmacht der Ordnungspolitik—Eine Bilanz nach 40 Jahren Sozialer Marktwirtschaft”, 40 ORDO 303–11 (1989)). 25

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 173

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 173 petition law to this day. The ordoliberals’ contribution also helps explain the unique European approach to single-firm conduct. For, although the ordoliberals certainly did not share the positive view of cartels that prevailed in Germany in the late 19th Century, their attitudes towards single-firm conduct do reflect some of the distrust of aggressive competition that underlay German attitudes toward competition, as well as the specificities of their own historical situation. Recent defenders of the Ordoliberal school have made clear how central the preservation of freedom of action and what they describe as the competitive process is to their approach to competition law.28 As Professor Giorgio Monti, presumably no relation to the former Commissioner, has phrased it, “competition is a process whereby market actors participate in the economy without overwhelming constraints from private and public power. Accordingly, the aim of competition policy is ‘the protection of individual economic freedom of action as a value in itself’ . . .” According to this view, “economic efficiency is the result of the freedom which competition law preserves. . . . [O]rdoliberalism values individual freedom as an end in itself . . .”. A restriction of competition within the meaning of Article 81(1) is “an undue restriction of the economic freedom of the parties or a restriction on other market participants”.29 In the context of Article 82, ordoliberalism gave birth to the doctrines of “performance-based” competition. This concept springs from a distinction drawn by a German scholar, H.C. Nipperdey, between “performance-based competiton” (Leistungswettbewerb) and “impediment competition” (Behinderungswettbewerb). According to Professor Gerber, “the former included conduct that made a firm’s products more attractive to consumers, typically by improving their characteristics or lowering their prices, while the latter referred to conduct designed to impede a rival’s capacity to perform”.30

28 See Giorgio Monti, “Article 81 EC and Public Policy”, 39 Common Market Law Review 1057, 1059 (2002). 29 Ibid. at 1059–1061 (footnotes omitted). Historically, Ordoliberalism can be viewed as a response to excessive concentration of economic power. This reaction manifests itself in its concern for preserving what are perceived as “competitive” market structures and a lack of confidence in the ability of markets to self-regulate. The result is often an over-willingness to intervene in order to save competition from itself or to save competitors from their more successful rivals. 30 Gerber, supra note 3, at 53. Behinderungswettberwerb is best translated as “impediment competition”, rather than exclusionary competition. As discussed below, the poles of the contrast between performance-based competition and impediment competition are not dissimilar to the two standards identified by Professor Einer Elhauge in, respectively, his Yale and Stanford law review articles: conduct that enhances the efficiency of the firm engaging in it and conduct that limits other firms’ efficiency (although this raises the issue of how to treat conduct that both enhances a firm’s efficiency and decreases that of its rivals). Einer Elhauge, “Why Above-Cost Price Cuts to Drive out Entrants are Not Predatory—and the Implications for Defining Costs and Market Power”, 112 Yale Law Journal 681 (2003); Einer Elhauge, “Defining Better Monopolization Standards”, 56 Stanford Law Review 253 (2003).

(H) Ehlermann 08 Venit

174

6/11/08

11:49

Page 174

Session One: The Sherman Act And Abuse Of Position Under Art 82

European law has embraced the concept of performance-based competition31 but it has also embraced other concepts traceable to ordoliberal thought such as the “special obligations of the dominant firm” and the need to protect the competitive structure against competition from strong rivals. As one of the modern ordoliberal proponents puts it: “A system of undistorted competition should be understood as a market system in which superior business performance is the decisive factor for the success of an undertaking (principle of “competition on the merits”).32 That objective is ultimately pursued for the benefit of the consumer, and consumer welfare33 is thus one, if not the foremost, rationale of abuse control.34 Nonetheless, it should not be considered a direct requirement of the types of abuse of interest here (anticompetitive abuses). As was repeatedly confirmed by the Court, Article 82 applies not only to practices directly damaging consumers, but also to conduct which is detrimental to

31 The concept of “means other than normal competition on the basis of the performance of commercial operators,” which is central to the notion of abuse under Article 82 as articulated first in Case 85/76, Hoffman-La Roche v Commission [1979] ECR 461, was developed by Professor Peter Ulmer of Heidelberg University in the mid-1970s. Gerber, supra note 3, at 313. According to Ulmer, the conduct of a firm constituted an abuse when it satisfied two conditions: (i) the conduct was non-performance based competition, i.e., competition that is not based “on the merits”; and (ii) it must restrict what residual competition remains in the market in which the dominant position is held. See also John Kallaugher and Brian Sher, “Rebates Revisited: AntiCompetitive Effects and Exclusionary Abuse under Article 82”, 25 European Competition Law Review 263, 269 (2004). 32 Case 85/76, Hoffman-La Roche v Commission cited in previous footnote, at para. 91; Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359, at para. 69; Case T-65/98, Van den Bergh Foods v Commission [2003] ECR II-4653, at para. 157; Case T-51/89, Tetra Pak Rausing v Commission [1990] ECR II-309, at para. 23; Case 322/81, NV Nederlandsche Banden-Industrie Michelin v Commission (Michelin I) [1983] ECR 3461, at para. 70; Case T-111/96, ITT Promedia v Commission [1998] ECR II-2937, at para. 138; Case 31/80, L’Oreal v PVBA De Nieuwe Amck [1980] ECR 3775, at para. 27; Case T-219/99, British Airways v Commission [2003] ECR II-5917, at para. 241; Case T-203/01, Manufacture Française des Pneumatiques Michelin v Commission (Michelin II) [2003] ECR II-4071, at paras. 54 and 97; Case T-228/97, Irish Sugar v Commission [1999] ECR II-2969, at para. 111; Case T-65/89, BPB Industries and British Gypsum v Commission [1993] ECR II-389, at paras. 94 and 118. 33 In a judgment dated 27 September 2006, which has now been contested before the ECJ (on appeal: Cases C-501/06, C-513/06, C-515/06 & C-519/06), the Court of First Instance identified consumer welfare as the key goal of competition law (without drawing any distinction between Articles 81 and 82) and placed this consideration at the forefront of its analysis of whether there was a restriction of competition by object. Case T-168/01, GlaxoSmithKline Services Unlimited v Commission, [2006] ECR II-2969, on appeal: Cases C-501/06, C-513/06, C-515/06 and C-519/06, not yet decided. The CFI’s judgment suggests that Professor Eilmansberger’s conclusion that consumer welfare is not the goal of competition law no longer reflects Community law as it has now developed. See Thomas Eilmansberger, “How to distinguish good from bad competition under Article 82 EC: In search of clearer and more coherent standards for anti-competitive abuses”, 42 Common Market Law Review 129 (2005). Certainly, Professor Eilmansberger’s interpretation is also out of step with the Commission’s guidelines on the application of Article 81(3), [2004] OJ C101/97, which also places consumer welfare at the centre of the analysis under both Articles 81(1) and 81(3). See discussion below. 34 “Consumer welfare” here refers to the real, not the Chicago, sense of the term. See Alison Jones and Brenda Sufrin, EC Competition Law, Oxford University Press, 2004, 39 et seq.; Commission guidelines on vertical restraints, [2000] OJ C291/1, at para. 7; Commission guidelines on the application of Article 81(3), cited in previous footnote, at para. 13.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 175

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 175 them by its impact on an effective competition structure.35 The protection of the competitive process is thus a goal in itself.36 The underlying idea is that consumers are best served if firms are forced to compete and, at the same time, are protected in their efforts to succeed on the merits.”37

The cited passages accurately reflect the two cornerstones of the ordoliberal approach: the application of Article 81(1) to restrictions on economic freedom and the application of Article 82 to conduct that is perceived as deviating from normal “competition on the merits” and thereby threatening the competitive structure (which Professor Eilmansberger, perhaps with a nod to Professor Fox,38 now extends to include the competitive process). However, the attempt to relate the “superior business performance which is the decisive factor for the success of an undertaking” with the “principle of ‘competition on the merits’” should cause some surprise to the initiated reader. For both “competition on the merits” and its opposite—“means other than normal competition on the basis of the performance of commercial operators”—have acquired very special, indeed, one could say “coded” meanings under Community law and those meanings appear unrelated to conventional notions of “superior business performance”.39 To take one example, in the field of discounts, “competition on the merits” does not include the granting 35 Case 6/72, Europemballage Corp and Continental Can Co Inc v Commission [1973] ECR 215, at para. 26; Hoffman LaRoche, supra note 31 at paras. 89 et seq. and 125; British Airways, supra note 32 at para. 244; Cases 40/73 to 48/73, 50/73, 54/73 to 56/73, 111/73, 113/73 and 114/73, Suiker Unie and Others v Commission [1975] ECR 1663, at para. 518; Michelin I, supra note 32 at para. 71; BPB Industries and British Gypsum, supra note 32 at para. 120; Case T-54/99, max.mobil Telekommunikation v Commission [2002] ECR II-313, at para. 52; Irish Sugar, supra note 32 at para. 232; also Case C-418/01, IMS Health v NDC Health [2004] ECR I-5039, at para. 48. Drexl proposes to explain such consumer emphasis with the concept of dynamic efficiency. See Josef Drexl, “Intellectual Property and Antitrust Law-IMS Health and Trinko-Antitrust Placebo for Consumers Instead of Sound Economics in Refusal-to-Deal Cases”, 35 ICC: International Review of Intellectual Property and Competition Law 788, 802, 804 et seq. (2004). 36 See Luc Gyselen, “Rebates-Competition on the Merits or Exclusionary Practice?”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 287 et seq.; Eleanor Fox, “Abuse of Dominance and Monopolisation: How to Protect Competition Without Protecting Competitors”, in ibid. pp. 69 et seq.; Eleanor Fox, “We Protect Competition, You Protect Competitors”, 26 World Competition 155 (2003). 37 Philip Lowe, “DG Competition’s Review of the Policy on Abuse of Dominance”, in Annual Proceedings of the Fordham Corporate Law Institute 2003 163 (2004); Gyselen, supra note 36 at 287; Duncan Sinclair, “Abuse of Dominance at a Crossroads: Potential Effect, Object and Appreciability Under Article 82 EC”, 25 European Competition Law Review 491, 494 (2004). 38 See both articles by Professor Fox, supra note 36. 39 As discussed more fully below, “competition on the merits”, as the term has been used and defined in Community law, bears very little resemblance to the robust form of aggressive competition that is more readily accepted and encouraged in the United States. To cite just one example that will be developed more fully below, in the context of rebates, “competition on the merits” does not extend to, or take into account, factors such as the value to the seller of its customer’s business performance. Rather, it has been confined to a very narrow supply-side, costbased justification which in practice makes it impossible to defend rebate schemes against allegations of exclusion and discrimination. For examples in the case law, see Michelin II and British Airways, each cited supra note 32.

(H) Ehlermann 08 Venit

176

6/11/08

11:49

Page 176

Session One: The Sherman Act And Abuse Of Position Under Art 82

of rebates that incentivize output or enhance distributors’ performance or reflect the value to a supplier of the superior performance of a party entrusted with the sale of its goods. Rather, “normal competition on the basis of performance” and “competition on the merits” have meant that rebates granted by a dominant firm can only be justified when they correspond to very precise cost savings realized by the firm. 40 This has led one astute observer to remark that European Community case law applicable to above-cost discounting is “rooted in nineteenth-century economic theory which relied on a cost-based theory of value from Ricardo to Marx”.41 A better feel for the classical ordoliberal position in practice—this time dealing with the special obligations of dominant firms and the extent to which it is necessary to assess the actual effects of competition that is not “based on the merits”—can be gained from the following passages taken from the Advocate General’s opinion in the recent British Airways case:42 69. The conduct of a dominant undertaking is not, therefore, to be regarded as abusive within the meaning of Article 82 EC only once it has concrete effects on individual market participants, be they competitors or consumers. Rather, a line of conduct of a dominant undertaking is abusive as soon as it runs counter to the purpose of protecting competition in the internal market from distortions (Article 3(1)(g) EC).43 That is because, as already mentioned, a dominant undertaking bears a particular responsibility to ensure that effective and undistorted competition in the common market is not undermined by its conduct.44 71. What is to be proved is, rather, the mere likelihood of the conduct in question hindering the maintenance or development of competition still existing in the market by means other than competition on the merits, thereby prejudicing the goal of effective and undistorted competition in the common market. With regard, therefore, to rebates and bonuses of a dominant undertaking, it has to be proved that they are capable45 of making it difficult or impossible for that undertaking’s competitors to have access to the market and its business partners to choose between various sources of supply. 87. It is thus sufficient to demonstrate that the rebate or bonus scheme of a dominant undertaking is likely to make it difficult or impossible for its competitors to gain access to the market and its business partners to choose between various 40 As a result, it is not possible to justify discounts by such obvious factors as their ability to incentivize output or reward the performance of the supplier’s distributors. 41 David Spector, “Loyalty Rebates and Related Pricing Practices: When Should Competition Authorities Worry”, forthcoming in World Competition. 42 Opinion of Advocate General Kokott delivered on 23 February 2006, Case C-95/04 P, British Airways v Commission [2007] ECR I-2331. This passage is cited not because it constitutes good law but rather for its presentation of the Ordoliberal approach. 43 Joined Cases 6/73 and 7/73, Istituto Chemioterapico Italiano and Commercial Solvents v Commission [1974] ECR 223, at para. 25. 44 Michelin I, supra note 32, at para. 57. 45 Suiker Unie, supra note 35 at para. 526; Hoffmann-La Roche, supra note 32 at para. 90; Michelin I, supra note 32 at para. 73, second sentence, and para. 85, first sentence. On the criterion of likelihood, see also Case C-7/97, Bronner [1998] ECR I-7791, at para. 38.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 177

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 177 sources of supply or business partners, unless there is an objective economic justification for it. Where there is such a hindrance to remaining competition, it can be assumed that, indirectly, consumers are also disadvantaged. 125. However, proof that quantifiable damage or an actual, quantifiable worsening of the competitive position of individual trading partners of the dominant undertaking actually took place cannot be demanded. That is because, as already stated, Article 82 EC serves primarily to protect competition as an institution.46 Therefore, under subparagraph (c) of the second paragraph also, discrimination amongst trading partners in competition with each other can be regarded as abusive once the conduct of the dominant undertaking is likely in the circumstances of the individual case to cause competition amongst those trading partners to be distorted.

The inability of normative language to provide guidance in monopolization cases has recently been noted by several commentators.47 To their critique of the inadequacy of vague standards, I would add the treacherous use of important words about whose fundamental meaning there is no agreement. Or to put it more bluntly, it should not be surprising that the practical significance of words is destroyed by the evocation of the virtues of protecting competition and the competitive process when those advocating these goals do so on the basis of very different understandings of what they mean by “competition” or the “competitive process”.48 Linguistically, we are in a situation analogous to the one described by Abraham Lincoln in his second inaugural address near the end of the North American Civil War when he noted of the two opposing sides that “Both read the same Bible and pray to the same God; and each invokes His aid against the other . . .”49 In that conflict, it was relatively easy to distinguish the antagonists well in advance by the colour of their uniforms. That may not always be the case when, for one party 46

Footnote omitted. See Einer Elhauge, “Defining Better Monopolization Standards”, supra note 30, at 255 (2003). Elhauge writes: “We’ve all gotten used to a little vagueness in law. Sometimes you just can’t foresee or account for the full complexity of life, and, when that is so, the best the law can do is to define some general guidelines for courts and juries to apply to particular facts. But for decades monopolization doctrine has been governed by standards that are not just vague but vacuous.” See also Barry E. Hawk, “Article 82 and Section 2: Abuse and Monopolizing Conduct”, in Wayne D. Collins, ed., Issues in Competition Law and Policy, ABA Section of Antitrust Law, forthcoming. According to Hawk, “[t]hese definitions offer mostly conclusory labels and provide no firm guidance in hard cases involving competitively ambiguous conduct (like reduced pricing, refusals to deal and product innovations)”. 48 Robert Bork has provided the following alternative definitions of what might be meant by “competition”: (i) the process of rivalry; (ii) the absence of restraint over one person’s or firm’s economic activities by any other person or firm; (iii) that state of the market in which the individual buyer or seller does not influence the price by his purchases or sales; (iv) the existence of fragmented industries and markets; or (v) a shorthand expression or term of art, designating any state of affairs in which consumer welfare cannot be increased by moving to an alternative state of affairs through judicial decree. Robert H. Bork, The Antitrust Paradox—A Policy at War with Itself, Basic Books, 1978, pp. 58–61. 49 In that particular case, as Lincoln goes on to note, the indignant Deity visited a considerable plague on the houses of both sides. 47

(H) Ehlermann 08 Venit

178

6/11/08

11:49

Page 178

Session One: The Sherman Act And Abuse Of Position Under Art 82

to the debate, the “competitive process” means the Schumpeterian striving for and destruction of monopoly power through innovation while for the other it means the protection of the competitive structure and smaller competitors against the competitive threat represented by a larger rival.50

III. The Anticompetitive Paradox Prior to the actual attainment of a monopoly, single firm conduct poses the greatest difficulty for antitrust law because the conduct of the firm resembles and indeed may constitute the embodiment of desirable competition.51 The attitude of antitrust law to this striving for success is contradictory: we want and encourage firms to compete aggressively but we are not always sure whether or when to intervene when they have done too good a job of it. For those who have faith in market forces, there is a reluctance to intervene which is not shared by those who prefer to defend what they perceive as the competitive process.52 I use the language of belief intentionally because behind either commitment lie very different visions of what “competition” is.53 Those who trust the action of market forces tend to rely on the same forces to rectify the situation that may result from an excess of temporary success, at least where entry barriers are not absolute.54 Those who do not have such 50 No economist has so ably anticipated (what was for him) the future as did Joseph Schumpeter. His repatriation would be a welcome event. 51 See Hawk, supra note 47 (“Application of antitrust law to dominant firm conduct implicates the highly difficult task of reconciling competing and conflicting policy considerations, including the various tensions among consumer welfare, fairness, and market integration”). 52 William Baxter was the Assistant Attorney General in the Antitrust Division of the Justice Department who took the decision to close down the DOJ’s Section 2 investigation of IBM’s alleged monopolization of computers. I once had the pleasure of being on a panel with Baxter in the mid-1980s. The former Attorney General expressed satisfaction over the handling of the IBM case before we went onstage, noting “Well, we certainly got that one right, didn’t we?” 53 See Hawk, supra note 47 (“Differing historical contexts such as the greater role of public companies and state-created monopolies in the EU, differing policy considerations such as the EU’s traditional embrace of fairness, and differing underlying economic and juridical assumptions about, among others, market erosion and the capability of authorities and courts to identify and remedy anticompetitive conduct all explain the traditionally broader scope of Article 82 compared with Section 2.”). Hawk identifies the forces that result in these differences as differing historical experiences and differing assumptions about the market and governmental intervention. He also cites Geradin’s well-made observation that “one possible explanation for the greater general willingness in the EC to impose duties to deal on monopolists is that, compared to the United States, more of the monopolies in Europe were created by regulations, government subsidies, or permitted combinations rather than by innovation or other investments”. Damien Geradin, “Limiting the scope of Article 82 EC: What can the EU learn from the U.S. Supreme Court’s judgment in Trinko in the wake of Microsoft, IMS, and Deutsche Telekom?”, 41 Common Market Law Review 1519 (2004). 54 As then-Judge Bork said in Rothery Storage, “[i]f it is clear that Atlas [a furniture moving company] and its agents by eliminating competition among themselves are not attempting to restrict industry output, then their agreement must be designed to make the conduct of their

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 179

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 179 trust favour intervention, even though this may mean imposing limits on the successful firm’s ability to compete and thus may hinder rather than advance “the competitive process”. Thus, for one side, the quest to dominate and destroy one’s rivals is the engine of the competitive process and should not be trammelled.55 For the other it is a destructive force that must be constrained. In either case, cant about protecting competition or the competitive process serves as a convenient but uninformative smoke screen. Little wonder that language buckles under the strain and becomes vacuous. No word can simultaneously designate itself and its polar opposite, and no meaningful guidance can be given with words that have lost their descriptive and normative power because they mean diametrically opposite things to those employing them. How we have come to have such diverse views about the meaning of competition (including the lack of consensus even within the same legal culture) is another question. If one were to distinguish—historically, almost anthropologically—the American from the continental European perspective one might well start with the difference between a society whose formative experience was the frontier56 and one whose first enduring social form after the fall of Rome was feudalism. From these very different starting points there is a line of development57 which, notwithstanding its deviations, individual business more effective. No third possibility suggests itself”. Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d 210 (D.C. Cir. 1986), cert. denied, 479 U.S. 1033 (1987). Similarly, Bork said in The Antitrust Paradox, supra note 48, that “[i]mproper exclusion” is always deliberately predatory and inefficient, which is rare; otherwise, the exclusion is the product of superior efficiency. “There is no ‘intermediate case’ of exclusion . . .” 55 See Bork, The Antitrust Paradox, supra note 48. One very useful way to distinguish between competitive and anticompetitive agreements is to ask whether the firms’ agreement is consistent with the unilateral drive for monopoly. The exercise points to the critical differences between competition and cooperation. It also suggests that the distrust of aggressive competition may have its origins not in distrust of private power but rather in distrust of aggressive business conduct, initiative and innovation. Against this background, it is not unreasonable to question whether intervention aimed at single-firm conduct does not produce a cooperative solution that betrays the essence of competition and may not be the most beneficial one for consumers. Those interested in curbing single-firm conduct in spheres that go beyond output restrictions and consumer exploitation would of course argue that their interventions are designed to protect competition, but that merely raises the question of what one means by “competition”. 56 Although no longer universally accepted, the “Turner thesis” on the role of the frontier is well known to students of American history. See Frederick Jackson Turner, “The Significance of the Frontier in American History” (1893). One could liken the “incessant expansion” of successive frontiers of which Turner spoke to the creative destruction of the entrepreneurs that drives Schumpeter’s model of monopolistic innovation. One might also add to the Turner thesis other individualistic elements—the emigration of those who preferred to exchange the known and the familiar for the new and take their chances, and the absence of an entrenched, inherited aristocracy. 57 We must be careful of over-generalization. Feudalism itself took different forms in different countries in the middle ages and the developmental path of countries such as France, with its early strong centralized administration and 16th Century mercantilist approach which later translated into dirigisme, is very different from that of Germany or England. And other contrasts must be drawn on a nation-by-nation basis. Nevertheless, awareness of the differences between the European and American developmental line is not without value in understanding differences in attitudes and results.

(H) Ehlermann 08 Venit

180

6/11/08

11:49

Page 180

Session One: The Sherman Act And Abuse Of Position Under Art 82

exceptions and the untold specific mutations of its DNA, can be traced to different attitudes toward individual initiative, respect for hierarchy and order, and the use and misuse of public and private power, and ultimately to different views about what constitutes competition and the extent to which it is necessary to intervene in markets. This difference in attitudes yields different results not merely as concerns substantive interpretations but also in enforcement modalities, and indeed in the uses to which the legal system itself should be put. To cite one obvious example, the prevalence of private enforcement (like private litigation in general) in the United States is not the result solely or perhaps even principally of a wise or unwise governmental decision to create a civil disincentive to discourage cartel activity. It is also both a product and a manifestation of the general litigiousness of a society which— as can be inferred from areas as diverse as its divorce and tort laws—has historically exploited and culturally favours, to a far greater extent than continental European legal cultures, the use of the legal system as a means of redistributing wealth, which in turn is a reflection of the mentality and mobility that are characteristic of a socially fluid society whose decisive formative experiences were immigration and the frontier.58 This contrast further suggests that the trust or distrust of markets and/or the competitive process (and indeed the manner in which these terms are defined and understood) may be strongly influenced by the same cultural attitudes and experiences. There is probably another consequence that flows from this divergence in formation and experience. For those societies in which competition is most aggressive (and that therefore least need a dose of invigorating unregulated competition) appear to have a predilection for maintaining that aggressive level,59 whereas societies in which the competitive urge has been more restrained and in which experience has been more slanted toward cooperative solutions or the exercise of public power and which may therefore be more in need of an infusion of something more dynamic have less experience with, and consequentially less tolerance for, such exposure. Some of the flavour of these different responses comes across in Professor Gerber’s description of the western European response to the worldwide economic depression of the 1870s through the 1890s, particularly in countries like Austria and Germany where cartelization became widespread and where, as Gerber puts it, “com58 The interests and profit motives of the private bar are another contributory factor that should not be overlooked, as is clearly evidenced by the fee arrangements in private enforcement class actions. 59 This tolerance no doubt reaches its zenith in Judge Easterbrook’s meditation on intent in A.A. Poultry Farms, Inc. v. Rose Acre Farms, 881 F.2d 1396, 1401-02 (7th Cir. 1989): “Almost all evidence bearing on ‘intent’ tends to show both greed-driven desire and glee at a rival’s predicament. . . . [But] firms need not like their competitors; they need not cheer them on to success; a desire to extinguish one’s rivals is entirely consistent with, often is the motive behind, competition.” But Judge Easterbrook’s approach reflects the awareness that the competitive urge is to triumph over and eliminate one’s rivals and a concern that the alternative—cooperating with rivals or treating them more gently—is likely to produce less advantageous results.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 181

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 181 petition that was not ‘managed’ came to be seen as ‘excessive’ and thus as justification for competitors to join together and to regulate it.”60 As he notes, in reference to Germany: “Social support for the general proposition that cooperation is better than individualism nourished the growth of the cartels.”61

The positive attitude toward the cooperative spirit of cartels has carried with it a contrasting distrust of individualism and of aggressive competition, which is viewed as predatory. This attitude has combined with the ordoliberals’ distrust of large industrial firms and their preference for the Mittlestand (small and medium-sized firms), and together these influences have survived in modern European antitrust law. “The ‘ethical’ advantages of cartels were often contrasted with American trusts. According to the historical school [of 19th Century German economists] cartels encouraged positive ethical attributes such as restraint and honesty in economic conduct and concern for weaker competitors. Trusts on the other hand, were portrayed as naturally rapacious, deceitful and willing to harm not only consumers and competitors but society as a whole.”62

The purpose of the foregoing analysis is not to postulate that there is a uniformity of belief on either side of the Atlantic, nor to deny the changes in antitrust fashion that led someone like Judge Bork to decry the anticompetitive and interventionist trend of the US Supreme Court in the period that preceded the ascendancy of the Chicago School. Nor is it to stake a claim for American superiority. Rather, it is merely a question of, first, acknowledging that we are the products of our histories and cultures and that these condition our predispositions and beliefs; and then trying to track down these origins so that we may better understand them, perhaps thereby to free ourselves somewhat of their constraints.

60 Gerber, supra note 3, at 16-42. Gerber’s chapter on the 19th Century in essence traces the loss of faith in the beneficially self-regulating market espoused by Adam Smith as a result of the economic depression which stretched from the 1870s to the 1890s and which brought with it a distrust of competition and the “market”; and which led to a reliance on cooperation in the form of cartels and government intervention. These developments do not appear to have had their parallel, or at least not as intensely, in the United States during the same period. In the case of fin de siècle Austria, Gerber links the distrust of unbridled competition to the collectivist values of Austrian society, values fostered by cultural and political traditions, not least among them the Catholic church. Ibid. at 64. As concerns Germany, Gerber notes that the combination of the speed and intensity of industrialization and the accompanying social transformations made competition appear particularly menacing, while the depressed conditions that slowed economic growth in the last quarter of the 19th Century made it appear unreliable. The result was the perception that competition worked best when it was regulated. Ibid. at 69. 61 Ibid. at 76. As Gerber notes, the underlying cultural force was the belief that the best way to deal with the dangers of unrestricted competition was to manage the process. Ibid. at 73. 62 Ibid. at 88 (citing Gustav Schmoller, “Das Verhaltnis der Kartelle zium Staat”, 116 Schriften des Vereins für Socialpolitik 237, 267–268 (1906)).

(H) Ehlermann 08 Venit

182

6/11/08

11:49

Page 182

Session One: The Sherman Act And Abuse Of Position Under Art 82

Certainly it would appear that a historical, cultural difference in attitudes has extended into the present and that this helps explain some of the current divergences between the two sides of Atlantic. On the western shore, very aggressive agency enforcement against cartels and very little action against single-firm conduct, whereas on the eastern shore, until recently, a lack of aggressive cartel enforcement but a record of attacking single-firm practices, including in cases that the colleagues on the western shore were not willing to pursue, or at least not with the same degree of engagement and finality. And there are many other signposts one could point to: the hostility under Article 82 to discounting as compared to the rule of reason approach to refusals to deal, the doctrine of the “special obligations” of the dominant firm, the reluctance with which (and the time it has taken for) efficiencies to be accepted as a key part of antitrust analysis,63 and the hostile overreaction to nonhorizontal mergers which has required correction from the courts.64 Lest this critique appear harsh and one-sided, it should be noted that a similar assessment has been delivered by distinguished European commentators (in the same way that some American commentators have criticized as overly non-interventionist the current US approach). At the forefront of these European voices I would put Professor Giuliano Amato, who has provided one of the more thoughtful analyses of the cultural presumptions that underlie the European approach to antitrust in his 1997 book, Antitrust and the bounds of power: the dilemma of liberal democracy in the history of the market. As concerns Article 82, Professor Amato observes that Europeans “were historically accustomed and culturally not opposed to recognizing private power subject to its bridling and balancing through regulation and the direct entry of public power in the economy. . . .” “This set of circumstances,” he notes, “has meant that our antitrust law was from the outset not a negotiation but a limitation of private power, and that from that viewpoint it combined two different natures: the classical “surgical” one, and an entirely European, markedly regulatory one—the acceptance of the dominant position as a position of independence from competitors and in any event something different from and less than total market closure,65 the special responsibility of the firm that holds it, and the treat63 A vice that appears to have been addressed in the Commission’s Notice on Article 81(3), cited supra note 34. 64 See Tetra Pak Rausing, supra note 32; Case T-209/01, Honeywell v Commission [2005] ECR II-5527; Case T-210/01, General Electric Company v Commission [2005] ECR II-5575. 65 It is sometimes said that, under EU law, there is no sanction for the attempt to monopolize and that the EU accepts the existence of dominance and only prohibits its abuse. Given the very low thresholds at which dominance kicks in—40% for the courts, and possibly above 25% for the Commission—and given the mechanical way in which dominance is often attributed solely on the basis of market shares, it can be argued that EU law does in fact block attempts at monopolization, and that it does so at thresholds of market power very notably below those that are used as benchmarks in the US. In other words, the “tolerance” for a dominant position is qualified by the ease with which dominance can be found. See Giuliano Amato, Antitrust and the Bounds of Power: the Dilemma of Liberal Democracy in the History of the Market, Hart Publishing, 1997.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 183

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 183 ment of various types of abuse, are perhaps the clearest examples of this European specificity.”66

Professor Amato’s observations suggest that the European approach has its disquieting contradictions. For the tolerance of public intervention in the economy has two dimensions: on the one hand, it has allowed governmental monopolies a significant role in the economy; on the other hand, it has fostered regulatory intervention in the private sector under the guise of controlling abuse. Apparently, the ordoliberal reticence about State intervention in the economy does not translate into distaste for intervention by an entity created by the State to modify single-firm conduct, provided the entity has the ability to act sufficiently independently of direct State influence. Again in the words of Professor Amato: “As we can see, two different antitrust perspectives emerge . . . and the difference always, and significantly, leads to the same aspect. The first is antitrust law that delays intervention to the last, leaving the market to provide as far as possible by itself for a definition of its own dynamics and its own equilibria: only imminent risk, with no alternatives, of output restriction justifies and permits intervention. The second is antitrust law that seeks to prevent that risk emerging and inserts itself more frequently and earlier into ongoing market dynamics, seeking to influence their structure. The first, in other words, sets the boundary of public power as far ahead as possible, accepting the risk of private power; the second does not accept that risk and instead more runs the risk of preventive intrusions by public power.”67

There may be an incestuous relationship between the tolerance of public power in the economic sphere and the aggressive regulation of private economic operators. But even if the two are independent phenomena, their combination does not necessarily produce a competitive outcome. That is why one of the most hopeful signs at the outset of the discussion about the reform of Article 82 has been the recognition that there may be a vital difference in the appropriate treatment of firms that have been given or have inherited monopoly positions and, by contrast, firms that have earned, through their foresight, skill and industry, a “dominant” market position.68 Commissioner 66 Ibid. at 113. The reference to the “acceptance” of the dominant position calls attention to the fact that under EU law, a dominant position is not prohibited, but is instead regulated. It could be argued that US law is more restrictive in that it prohibits conduct that may give rise to monopoly whereas EU law is willing to live with dominance although not its abuse. However, the reality of any theoretical difference depends on where the thresholds for monopolization and dominance are set. Translated into numerology, a system that prohibits certain conduct only when a firm has 70% or more of a relevant antitrust market is obviously less interventionist than one that imposes constraints as soon as the a firm attains a 40% share. 67 Ibid. at 112. 68 See Mario Monti, “Introductory Statement”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What Is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 3 et seq.; John Vickers, “How Does the Prohibition of Abuse of Dominance Fit with the Rest of Competition Policy?”, in Ehlermann and Atanasiu, ibid. pp. 147 et seq.

(H) Ehlermann 08 Venit

184

6/11/08

11:49

Page 184

Session One: The Sherman Act And Abuse Of Position Under Art 82

Monti alluded to this in his last address to this Workshop in 2005, and John Vickers has commented on it forcefully, noting that: “Appropriate public policy toward firms with actual or potential market power depends on the cause of the market power, not least for reasons of economic incentives. It is one thing to gain market power by innovating or otherwise winning customers over time by serving them well in the product markets; it is quite another thing to get market power by merger or cartel agreement; or to strengthen market power by exclusionary practices; or to hold or inherit a state monopoly. That is an old point, but its implications for the relationship of the prohibition of abuse of dominance to other elements of competition policy perhaps warrant more thought.”69

Sensitivity to this critical distinction suggests an emerging awareness of the need for a different approach to the two types of firms and perhaps suggests, in the case of firms that have earned their positions by innovating or better serving their customers, a greater reluctance to intervene. Such sensitivity is welcome. Concepts such as the “special duty” of the dominant firm and the willingness of the European institutions to find that competition has been eliminated or fatally weakened on the basis of market shares while ignoring market dynamics, or in contexts where this seems improbable, reinforce the impression that constraining competitive conduct has been the guiding principle underlying the application of Article 82 and that it may have produced anticompetitive, or at any rate less competitive results than would a less interventionist approach.70 Certainly, the interventionist regulatory inclination identified by Professor Amato can, as Professor Vickers has recognized, give rise to a system where incentives are skewed: dominant firms lose the motivation to compete and rivals are encouraged not to enhance their own competitiveness but rather to restrict the ability of the market leader to compete by seeking regulatory protection, which can take the form of a price umbrella or other ill-advised remedies, in a way that mirrors in reverse one of the standards for determining abuse. As Professor Amato has put it: “Again, our abuses of dominant positions express a lasting, unchanging severer antitrust orientation than the American ones, and no economic analysis has been able to modify or mitigate the “special responsibility” of firms which, surrounded by weaker competitors, are strongly restricted from conduct that in the USA would today be classed as “aggressive” competition. This means that by a long path that is all our own, and quite far away from the one that at the turn of the century set up the notion of efficiency on the other side of the Atlantic as openness of market

69

See Vickers, cited previous footnote, at 155. Judge Bork’s polemic against overzealous enforcement of the Sherman Act was motivated in part by concerns about the negative impact on American competitiveness, and concerns about loss of competitiveness informed the Chicago School’s agenda. 70

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 185

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 185 processes, we have arrived at adopting one that has a similar aim regarding abuses and continue to be loyal to it, even assuming decisive regulatory nuances in doing so”.71

Professor Amato is not alone in his concern. In the report submitted by the EAGCP in the context of the Article 82 reform,72 the message was resoundingly anti-interventionist: “We should not fall into the trap of active intervention and fine-tuning; whenever possible, competition is to be preferred to detailed regulation as the best mechanism to avoid inefficiencies and foster productivity and growth.”73 “The standard for assessing whether a given practice is detrimental to ‘competition’ or whether it is a legitimate tool of ‘competition’ should be derived from the effects of the practice on consumers. If we think of ‘competition’ as a regime in which the different suppliers contend to sell their products to participants on the other side of the market, then the benefits reaped by the other side of the market will themselves provide a measure of how well ‘competition’ works.”74 “Competitors themselves should not be protected from competition by the authority’s intervention. In each case, the competition authority must assess these matters without prejudice for any particular structure.”75 “In focusing on consumer welfare, one must not fall into the trap of seeing competition policy as a tool of active policy intervention designed to correct the inefficiencies associated with monopolies and oligopolies so as to maximize some measure of welfare. Competition policy is based on the principle that competition itself is the best mechanism for avoiding inefficiencies, so the competition authority should not try to let its own intervention replace the role of competition in the market place.”76

*

*

*

In recent years, considerable progress has been made in the reform of Article 81. While decentralization has not resolved the issue of what constitutes a restriction of competition, nor the theoretical problems caused by the bifurcation of Article 81, it does resolve the procedural and structural problem and was probably the most the Commission could do without a Treaty amendment. And it is certainly encouraging that, with the advent of decentralization, the national courts have shown a propensity for abandoning some of the Commission’s most prized but questionable dogmas, and have begun charting their own course. The pharmaceutical sector is one notable example, and 71

Supra note 64 at 114. Economic Advisory Group on Competition Policy (EAGCP), “An Economic Approach to Article 82” (July 2005), available at www.shh.fi/~stenback/eagcp_82_2005_july_21.pdf. 73 Ibid. at 3. 74 Ibid. at 8. 75 Ibid. at 9. 76 Amato, supra note 64 at 10–11. 72

(H) Ehlermann 08 Venit

186

6/11/08

11:49

Page 186

Session One: The Sherman Act And Abuse Of Position Under Art 82

there have been a slew of recent national court judgments outlining an approach that would not have been expected by the Commission.77 Similarly, in its Notice on the application of Article 81(3), the Commission has for the first time put consumer welfare, rather than individual liberty of action, at the heart of Article 81(1). The Commission takes the position, in the context of a discussion of restrictions by object, that: “Restrictions by object such as price fixing and market sharing reduce output and raise prices, leading to a misallocation of resources, because goods and services demanded by customers are not produced. They also lead to a reduction in consumer welfare, because consumers have to pay higher prices for the goods and services in question.”78

However, institutional anomie has resulted in a situation in which the Commission finds itself in the position of pleading cases before the Community courts on the basis of principles that appear inconsistent with its most recent pronouncements in policy documents such as the aforementioned Notice. Without naming names, this author has been embroiled in a pending case where the Commission on appeal to the Court of Justice has disputed both the principle that restrictions by object are based on a presumption and that the high likelihood of consumer harm is indispensable to the finding of an infringement by object. Similarly, the Commission’s Janus-like Article 82 Discussion Paper, which manifests a desire to go forward but also backward at the same time, is 77 In France, the Court d’Appel de Paris has upheld a decision of the Conseil de la Concurrence involving a refusal to supply parallel traders. The Conseil had concluded that “on ne voit pas quelle justification permettrait à un opérateur économique d’imposer à un producteur qui ne dispose pas de la liberté de fixer ses prix pour les produits destinés à être utilisés sur un territoire, d’appliquer d’une manière générale les mêmes conditions de vente pour des produits destinés exclusivement à d’autres territoires où les conditions de marché sont différentes.” The Court concurred, noting that “il n’est pas abusif pour un laboratoire de défendre ses intérêts commerciaux en refusant de livrer ses produits à un prix administré à un opérateur qui ne vend aucun produit sur le marché national pour lequel la réglementation du prix a été élaborée et qui ne recherche ce produit qu’à la condition que le prix fixé par les pouvoirs publics en vue d’un usage sur le territoire national lui permette de le revendre sur un marché étranger avec profit.” Judgment of the Cour d’Appel de Paris (1ère chamber—Section H) of 23 January 2007, Société Pharma-Lab, S.A.R.L. v. Conseil de la Concurrence. In Spain, in a judgment confirming the dismissal by the Tribunal de Defensa de la Competencia (TDC) of a complaint against pharmaceutical companies that refused to supply a parallel trader, Spain’s second highest court (Audiencia Nacional) declared that “parallel exports mainly benefit wholesalers, who obtain disproportionate, unexpected and exceptionally high profits (‘wind-fall profits’). In other terms, parallel exports do not constitute any direct benefit for consumers that pay the same price for the pharmaceutical product, whether it originates or not from a parallel import.” Judgment of the Audiencia Nacional of 26 January 2005, Laboratorios Farmacéuticos, unofficial translation of the Spanish text: “[L]as exportaciones paralelas benefician principalmente a los mayoristas, que obtienen ganancias desproporcionadas, inesperadas y excepcionalmente elevadas (“wind-fall profits”). Es decir, las exportaciones paralelas no representan ningún beneficio directo para los consumidores que pagan el mismo precio por el producto farmacéutico, proceda o no de una importación paralela.” 78 Supra note 34 at para. 21.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 187

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 187 another example. Here we have the endorsement of an “effects based” analysis followed by equivocations about the need to show recoupment in predation cases79 (on the grounds that intent is sufficient), the adoption of an economic approach to loyalty rebates followed by the proposal of average total cost test for assessing their exclusionary nature (again on the grounds that the intent behind loyalty is always exclusionary), and exceptions to the “as efficient” competitor test that could swallow it up.80 Such institutional resistance to change involves two elements. One is the apparently perceived necessity to take into account the existing jurisprudence and to maintain the “old line” in litigated cases that were started prior to the reform, the Commission’s pleadings in the British Airways case being an excellent example.81 The other is either a difference in the speed of adaptation or, more simply, the resistance of officials within the bureaucracy to change. The ability of either form of inertia to derail genuine reform should not be underestimated.

79 This seems counterproductive since, if recoupment were required, dominant firms that predate would have an incentive not to raise prices after eliminating or disciplining a rival because doing so would furnish proof of illegality. Thus, eliminating the recoupment requirement arguably encourages predation, since the ability to recapture profits is not a constituent element of the infraction. Conversely, if recoupment were a requirement and the firm did not raise price, its price cutting, if sustainable, would become more permanent and consumers would profit. Or if the firm knew that it would not eventually be able to raise price, there would be less incentive to engage in predation in the first place. 80 The ambivalence with which the Commission faces the task of reform was summed up by one of the drafters of the Discussion Paper, who noted at a conference that the Commission faced the following alternative: if it followed the economists’ recommendations it would bring very few cases although they would be good ones; if it did not follow those recommendations, it would continue to bring a lot of cases but most of them would be bad. Perhaps, he said, the best approach was something in the middle. However, since that remark was made there have been signs, most notably in a presentation by the Commission’s Chief Economist at a conference in Brussels in December 2006, that the Commission is now moving less ambiguously in a single direction toward meaningful reform. 81 Case C-95/04 P, British Airways v Commission [2007] ECR I-2331. The judgment of the ECJ is disappointing but it should not be surprising. The question of whether rivals could match BA’s discounts and recover their costs was not meaningfully analyzed. Instead, the anticompetitive effect was found in the formal nature of the discount, which made it necessary for rivals to grant “significantly higher rates of discount or bonus” in order to match the dominant firm’s discount. See ibid. at para. 75. According to the Court, in such a situation there is no need “to examine whether BA’s conduct had caused prejudice to consumers”. All that need be examined is “. . . whether the bonus schemes at issue had a restrictive effect on competition . . .” Ibid. at para. 107. In other words, the offensive effect arises from the nature or the form of the discount. This approach contrasts sharply with the one taken in the US case involving the same parties and the same rebate programme. There, the Second Circuit ruled that “[r]ewarding customer loyalty promotes competition on the merits.” Virgin Atlantic Airways v. British Airways, 257 F.3d 256, 265 (2d Cir. 2001). On the other hand, in what would appear to constitute an important first step toward a more flexible approach, the ECJ took the view that an examination of the efficiency justifications of the rebate system was warranted, stating in paragraph 86 of the judgment that “[i]t has to be determined whether the exclusionary effect arising from such a system, which is disadvantageous for competition, may be counterbalanced, or outweighed, by advantages in terms of efficiency which also benefit the consumer”.

(H) Ehlermann 08 Venit

188

6/11/08

11:49

Page 188

Session One: The Sherman Act And Abuse Of Position Under Art 82

The modernization discussions in both the US and the EU have generated an enormous amount of print and testimony over the appropriate standards in Section 2/Article 82 cases.82 Subject to one comment on efficiencies, that debate lies beyond the scope of this paper. What is interesting about this discussion, however, is the extent to which it appears to circle around the two poles of Leistungswettbewerb and Behinderungswettberwerb first identified by Nipperdey, provided they are stripped of the special meaning they have acquired under Community law over the past forty years.83 Within this framework, there appears to be broad and general agreement on at least some of the easier questions. Thus, one can identify Leistungswettbewerb with the proposition which has more or less universal assent to the effect that, all other things being equal and assuming that the conduct can be clearly identified, one should not condemn conduct that increases a firm’s output and/or efficiency and resulting ability to deliver consumer benefits.84 If words mean

82 For a comprehensive survey, see Hawk, supra note 47. The modernization discussion in the US has generated a substantial amount of commentary concerning the appropriate tests to be used to establish illegal exclusionary practices. Consumer welfare tests, profit sacrifice tests, but for tests, balancing tests, “as efficient competitor” tests and efficiency tests have been proposed and discussed at length. See generally, Elhauge, supra note 47; “General Approaches to Defining Abusive/Monopolistic Practices—Roundtable”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007; Herbert Hovenkamp, “Exclusion and the Sherman Act”, 72 University of Chicago Law Review 155 (2005); Herbert Hovenkamp, “Signposts of Anticompetitive Exclusion: Restraints on Innovation and Economies of Scale”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007; Marina Lao, “Defining Exclusionary Conduct Under Section 2: The Case for Non-Universal Standards”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007; A. Douglas Melamed, “Exclusionary Conduct Under the Antitrust Laws: Balancing, Sacrifice and Refusals to Deal”, 20 Berkeley Technology Law Journal 1247 (2005); A. Douglas Melamed, “Exclusive Dealing Agreements and Other Exclusionary Conduct—Are There Unifying Principles”, 73 Antitrust Law Journal 375 (2006); Mark R. Patterson, “The Sacrifice of Profits in Non-Price Predation”, 18 Antitrust 37 (2003); Emil Paulis, “The Burden of Proof in Article 82 Cases”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007; Mark S. Popofsky, “Defining Exclusionary Conduct: Section 2, the Rule of Reason, and the Unifying Principle Underlying Antitrust Rules”, 73 Antitrust Law Journal 435 (2006); Stephen Salop, “The Controversy over the Proper Antitrust Standard for Anticompetitive Exclusionary Conduct”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007; Steven C. Salop, “Exclusionary Conduct, Effect on Consumers, and the Flawed Profit-Sacrifice Standard”, 73 Antitrust Law Journal 311 (2006); “Symposium—Identifying Exclusionary Conduct Under Section 2”, 73 Antitrust Law Journal 1 (2006); John Vickers, “Abuse of Market Power”, 115 Economic Journal F244-F261 (2005); Gregory J. Werden, “Identifying Exclusionary Conduct Under Section 2: The No-Economic Sense Test”, 73 Antitrust Law Journal 413 (2006); Gregory Werden, “Identifying Single-Firm Exclusionary Conduct: From Vague Concepts to Administrable Rules”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007. 83 Bork draws the same distinction in The Antitrust Paradox, taking the view that only predation should be proscribed. Supra note 48. 84 One should also add another, more neutral category (but one that is vital for the preservation of competitive markets), which is that one should not condemn conduct that is not dissimilar from that engaged in by rivals, thereby putting a cap on the “special obligations of the dominant firm” standard.

(H) Ehlermann 08 Venit

6/11/08

11:49

Page 189

Cooperation, Initiative and Regulation—A Cross Cultural Inquiry 189 what they say, commentators as diverse as Professor Elhauge,85 for example, and the ordoliberals share this view. They would also seem to share the converse view that conduct that merely hinders the ability of firms to compete on the merits (in the ordoliberal formulation) or raises rivals costs or impedes their efficiency86 without enhancing the output or efficiency of the firm engaging in it, again assuming all other things are equal and that the conduct can be clearly identified and its effects established, merits sanction. The difficulty here, as Professor Fox recognized two years ago, is in distinguishing between the two.87 Greater difficulties also arise in dealing with the more troublesome category of conduct that both increases one firm’s efficiency but also decreases the efficiency (or raises the costs or impedes the ability to compete) of rival firms. On this issue, views appear to diverge not merely between commentators but apparently over time on the part of the same commentator.88 Here I would only note that it would seem reasonable not to condemn conduct that is efficiency enhancing, for the same reasons that one has efficiency justifications under Article 81(3) and considers efficiencies in mergers. If this view is accepted, efficiency justifications should not be structured in a way that in practice makes it virtually impossible for the dominant firm to invoke them successfully. There is some urgency to the need for reform. Competition law is not applied or enforced in a vacuum. Enforcement decisions have a real impact on business conduct, and the incentives and disincentives created directly affect the way in which firms comport themselves and compete. In addition to concerns about over and under enforcement, Type I and Type II errors, difficulties in assessment, fact gathering and judgment, this factor should not be ignored in evaluating the pros and cons of different enforcement approaches. Certainly, untimely early intervention itself has the potential for substantial anticompetitive harm by giving firms the wrong incentives—discouraging 85 See supra note 47 at 256 (where a defendant has improved its own efficiency in order to make a better product, it should be free to sell that product at any above-cost price it wants, even though that may shrink rivals’ market shares to a size that leaves them less efficient). 86 See ibid.: “Exclusionary conduct should be illegal if it would further monopoly power by impairing the efficiency of rivals even if the defendant did not successfully enhance its own efficiency.” 87 See Fox, Abuse of Dominance and Monopolisation, supra note 36. 88 See Einer Elhauge, “Why Above-Cost Price Cuts to Drive out Entrants are Not Predatory—and the Implications for Defining Costs and Market Power”, cited supra note 30, at 23–24 (2003). See also Michael Levine, “Airline Competition in Deregulated Markets: Theory, Firm Strategy and Public Policy”, 4 Yale Law Journal on Regulation 393, 395, 400–401, 403–405 (1987). Elhauge’s analysis is of particular interest in that he argues that airline conduct that has been deemed predatory by regulatory authorities can be explained by hub and spoke economics, which make it more efficient for one airline to provide both the non-stop and connecting flights in the hub-spoke system. According to Elhauge, given the economics of the industry, “reactive above-cost price cuts by hub incumbents, exit by the single route entrant and restoration of higher prices can be explained by fully desirable, competitive behavior”. Elhauge, cited above, at 27. But see also Elhauge, supra note 47 at 256, where Professor Elhauge articulates a test under which conduct would be condemned if it impaired rivals’ efficiency, whether or not it enhanced the monopolist’s efficiency.

(H) Ehlermann 08 Venit

190

6/11/08

11:49

Page 190

Session One: The Sherman Act And Abuse Of Position Under Art 82

innovation or price competition and encouraging rivals to complain and seek regulatory assistance rather than to innovate or discount on their own. And of course it may also impose on the antitrust authority quasi-regulatory tasks for which it is ill-suited, as it wades into ruling on such regulatory matters as licensing fees and discount structures. Last, sharp differences in the attitudes toward the same conduct in different jurisdictions are unworkable and unacceptable from the perspective of undertakings concerned in markets that are global. The purpose of these reflections has not been to provide answers to vexing questions but rather to take stock of where we are and how we got there. Development has its awkward phases. The Treaty of Rome was adopted 50 years ago in a world very different from the one that now preoccupies us, and its competition rules reflected a Community that was emerging not only from a ruinous war and bitter national rivalries but from a tradition in which State enterprise, State ownership and State intervention had played a significant economic role and, in the case of the Member State which had the greatest single influence in the shaping of its competition law, a tradition with a marked distrust of aggressive, unbridled competition. These traditions have found their way into European law, particularly as it impacts on and seeks to constrain single firm conduct in ways whose benefits may be questioned. The passage of time, the strengthening of the private sector, the increased interaction between competition authorities both within the Community and between it and third countries have gradually led to the recognition that a serious reform of some of these attitudes and their legal emanations is required. That reform is underway. If some recent signs can be relied on, there is reason to believe that it may well alter significantly the European peculiarity with respect to single firm conduct in pro-competitive ways that can only be welcomed.

Afterword I began the limited research involved in this paper expecting that I would find, in late 19th Century attitudes, a certain bias in favour of cooperation and a distrust of aggressive, unrestrained competition and that these values would be linked to broader historical, social and cultural factors. There is always a danger in finding what one seeks and that danger is increased when one relies largely on secondary as opposed to primary sources as I have had to do. I am nevertheless and obviously indebted to Professor Gerber for the research he has undertaken and for his attentiveness to the influence of German ideas, including those that achieved their expression in the last 25 years of the 19th Century, on the development of European competition law.

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 223

I David L. Meyer*

The Development and Communication of Competition Law Standards Applicable to Single-Firm Exclusionary Conduct: A Perspective Based on U.S. Experience June 2007

I appreciate the opportunity to share some of the views and experiences of the US Department of Justice on the tremendously important issue of competition law standards for single-firm conduct that harms competition (what I will often refer to here as “single-firm exclusionary conduct”). This is an area of competition law that has seen tremendous recent debate in Europe, the United States, and elsewhere in the world. As we all know, the European Commission has been engaged in a lengthy process aimed at producing guidelines for the application of Article 82. Meanwhile, across the Atlantic, the Department of Justice and the Federal Trade Commission have been conducting extensive hearings to consider the many complex issues posed by legal standards under Section 2 of the Sherman Act. And, of course, as time moves on, these standards continue to evolve as courts apply the underlying statutory framework in the context of particular enforcement actions—both private and governmental. I will focus on four lessons that the US experience offers in this area. First, despite some significant challenges, competition law should address singlefirm conduct that harms competition. Second, in the single-firm setting, where the costs of incorrect, hard-to-apply, or ambiguous standards are likely to be high, competition law should tread carefully, and there is a particularly acute need for standards that provide useful guidance to businesses and their counselors. Third, the US experience shows that enforcement agencies, working within the framework of a flexible set of core statutory provisions, can use various forms of guidance both to shape the underlying legal standards and to affect the behavior of private parties. And, fourth, US experience also suggests several important attributes of effective single-firm conduct standards— standards aimed at protecting the competitive process without sacrificing beneficial competition in the process.

* Deputy Assistant Attorney General, U.S. Department of Justice

(J) Ehlermann 08 Meyer

224

6/11/08

11:49

Page 224

Session One: The Sherman Act And Abuse Of Position Under Art 82

I. Competition law should address potentially anticompetitive single-firm conduct. First let me emphasize that we in the United States believe strongly that legal standards governing single-firm conduct are an important component of a comprehensive and effective competition law regime. Such standards have been part of US antitrust law for more than 100 years. Section 2 of the Sherman Act was a one-sentence statutory provision enacted in 1890 to prohibit “monopolization” and “attempted monopolization”. As interpreted by the courts over the past century, these offenses have two basic elements: (a) the firm must have a very high degree of durable market power, referred to as “monopoly power”, or, in an attempt case, it must have the “dangerous probability” of achieving that power; and (b) the firm must have maintained or achieved, or must be threatening to achieve, such power in a manner that is illegitimate, in the sense that it is inconsistent with innovation, skill, hard work, luck, or hard-fought competition on the merits. The first of these requirements—monopoly power—has never been defined with precision. It connotes the power unilaterally to control price for some substantial period of time, and it is typically associated with market shares above 70 percent combined with significant barriers to entry. The second of these requirements (i.e., point (b) above) has presented one of the most vexing issues in all of US antitrust jurisprudence. In early cases the requirement was often referred to as the “willful” acquisition or maintenance of monopoly power, as distinct from having monopoly “thrust upon” the firm. Modern cases recognize that all firms strive for dominance, and courts thus require that the striving be shown to be illegitimate. These cases tend to refer to the plaintiff’s need to prove that the defendant engaged in “exclusionary or predatory” conduct that harmed competition without contributing to competition on the merits. Although these well-established elements leave many questions unanswered, I note two important implications of this legal test. First, Section 2 does not condemn mere possession of monopoly power, resulting for example from a firm’s success in developing a product that catches the consumer’s fancy. A corollary is that the law also does not condemn the exploitation of monopoly power acquired legitimately: thus, a firm that has developed a wildly successful product—and perhaps even obtained a valid patent covering the innovation—can charge whatever it wishes and may choose not to share the profits with distributors or assist others in developing a competing product.

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 225

Standards for Single-Firm Conduct: a US Perspective

225

Why is it important for competition law to be able to reach at least some forms of single-firm exclusionary conduct? A modern foundation of competition law is the economic principle that the exercise of durable market power can distort the allocation of resources and harm consumers. Although competition law is not an appropriate mechanism for regulating the pricing, quality and output decisions of firms that possess such power, it does appropriately reach the means by which firms achieve or maintain that power. There is an international consensus that competition law must address agreements that have, likely will have, or tend almost always to have unreasonably anticompetitive effects. Such agreements would include naked price-fixing agreements, some mergers and acquisitions, and under some circumstances other horizontal or vertical arrangements. However, not all mechanisms that unreasonably enhance or entrench market power involve coordination among separate firms. When a firm with monopoly power acts alone to impede competition—via predatory pricing, for example—competition law should provide a framework for evaluating the competitive consequences of that firm’s behavior. The important role that competition law plays in policing potentially anticompetitive single-firm conduct is emphasized by several government enforcement actions: the DOJ’s cases against Microsoft, American Airlines, and Dentsply are a few examples. More recently, the US antitrust agencies have devoted considerable attention to Section 2 in a series of hearings aimed at helping us to advance our thinking with respect to unilateral conduct and to better inform our judgment about when it is appropriate to bring enforcement actions under Section 2. But the US agencies’ views of the significance of single-firm conduct rules should not be judged solely by the number of cases they bring. In the United States, Section 2 is most often enforced by private plaintiffs. The plaintiffs in these cases are often competitors who claim they have been hindered by the alleged monopolist’s behavior. The development of legal standards governing single-firm exclusionary conduct in the United States has evolved in significant part through the adjudication by US courts of these private claims. Important recent examples include the LePage’s case,1 a Third Circuit decision that addressed bundled rebates, albeit in a manner that failed to provide a constructive legal standard; the Trinko case,2 a Supreme Court case that addressed a monopolist’s alleged refusal to assist a rival; and the Weyerhaeuser case,3 where the Supreme Court recently held that an objective price-cost test applies in the context of claims of predatory over-bidding by an alleged monopsonist. The Department, joined by the FTC, participated in these three cases (and many others) as an amicus curiae, providing the 1

LePage’s v. 3M, 324 F.3d 141 (3d Cir. 2002). Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004). 3 Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., 549 U.S. 312 (2007), 127 S.Ct. 1069. 2

(J) Ehlermann 08 Meyer

226

6/11/08

11:49

Page 226

Session One: The Sherman Act And Abuse Of Position Under Art 82

Supreme Court with its thoughts on the proper legal standard and the appropriateness of the cases as a vehicle for further developing Section 2 law. Because private enforcement is so robust, it plays an important role in shaping the law and in influencing the behavior of businesses that must comply with the law. As a result, the US antitrust agencies’ efforts include not just investigating and bringing single-firm exclusionary conduct cases in appropriate circumstances but also providing guidance to businesses and their counselors and influencing the development of the law in cases brought by private plaintiffs. This is one of the objectives of our single-firm conduct hearings.

II. In the single-firm setting, where the costs of incorrect, hard-to-apply or ambiguous standards are likely to be high, there is a particularly acute need for standards that provide useful guidance to businesses and their counselors. When we in the United States think about proper competition law standards for single-firm exclusionary conduct, we apply the same analytical framework that we apply to other types of conduct that potentially harm competition. Essentially, we ask how the applicable statutory framework can be applied in a manner which is faithful to the text and to legislative intent, and which entails the greatest benefit and lowest costs to the policies of competition underlying the law. The foundational principle—the bare minimum condition—is that the legal standard must not penalize conduct unless it threatens harm to consumers and the competitive process. Choosing a particular market outcome or protecting a particular competitor—even a national champion—cannot be the goal. In the single-firm setting, there are several other considerations that bear upon the proper legal standards. It is one thing to say that the test should be whether “competition” or “consumer welfare” is harmed, but it is quite another to convert this goal into an effective enforcement regime. Unlike naked horizontal agreements to restrain trade, for example, much single-firm conduct is vigorously pro-competitive and should be encouraged. Competitive striving—whether through the introduction of new or improved products, enhancements to quality, innovative distribution schemes, or aggressive price competition—brings significant benefits to consumers. Moreover, the “opportunity to charge monopoly prices—at least for a short period—is what attracts ‘business acumen’ in the first place; it induces risk taking that produces innovation and economic growth”.4 As a result, as US 4

Trinko, supra note 2, at 407 (2004).

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 227

Standards for Single-Firm Conduct: a US Perspective

227

courts observed in the middle of the last century, competition law should not turn on those who, through dint of hard work or hard competition, manage to succeed and thereby secure dominant positions in their fields.5 Compounding the problem, it is notoriously hard to distinguish between the means of legitimate competition and those of “illicit exclusion”. Part of the problem is that the conduct itself is hard to distinguish—the same act can be either pro-competitive or anticompetitive, depending on the circumstances. It is thus impossible to enact rules that hinge legality solely on the type of conduct the dominant firm undertakes. Another part of the problem is that the evil we are concerned about (harm to competition) ordinarily arises only as an indirect result of rivals being disadvantaged and leaving the market—but this is as much a hallmark of vigorous competition as of monopolization. There is thus a significant risk that single-firm conduct rules that are too general—such as an admonition that dominant firms “shall not harm competition”—will be impossible to understand and apply with precision. Even if we could safely conclude that enforcers or adjudicators will never apply such rules incorrectly to falsely condemn pro-competitive conduct (although in the United States such “false positives” do occur), such a rule would still cause harm by creating uncertainty on the part of firms engaging in the rough and tumble of daily competition. Without knowing where the line of legality is drawn, they will pull back so as to avoid inadvertently crossing it. In the process, they will inevitably pull their competitive punches and refrain from engaging in the kinds of vigorous competition that would otherwise have benefited consumers. The “chilling” problem would be reduced if all single-firm conduct could be treated like a major merger or similar corporate transaction—a significant one-time event in the life of the firm that is generally susceptible to resourceintensive ex ante review by competition authorities. But single-firm exclusionary conduct takes myriad forms. Businesses make decisions about their competitive strategies on an ongoing basis. They would neither tolerate nor expect competition authorities to pre-screen those decisions and provide, as the agencies usually do for mergers, the sort of enforcement signal that (in cases where the agency does not intend to challenge the behavior) gives the parties comfort that they are free to proceed with little enforcement risk. Businesses must instead make their own calculations of legal risk and channel their behavior accordingly, based on what they and their counselors can discern from the available resources—the statutes, case law, enforcement patterns, and other communications—about where the line of legality is drawn. The imperative in the single-firm conduct setting of structuring legal rules that provide sufficient practical guidance to avoid “chilling the very conduct the antitrust laws are designed to protect” is reflected in many of the US 5 See United States v. Aluminum Co. of America, 148 F.2d 416, 430 (2d Cir. 1945) (“The successful competitor, having been urged to compete, must not be turned upon when he wins.”).

(J) Ehlermann 08 Meyer

228

6/11/08

11:49

Page 228

Session One: The Sherman Act And Abuse Of Position Under Art 82

courts’ more recent Section 2 judgments. Weyerhaeuser adapted the pricecost test of predatory pricing law to claims of monopsony over-bidding.6 And Trinko emphasized concerns about chilling pro-competitive conduct in underscoring the narrow circumstances in which a monopolist might have an obligation to deal with its rival.7 The same considerations motivate the enforcement agencies in the United States to do what they can to help the law develop in a manner that yields sensible legal standards while simultaneously providing guidance to businesses about how those standards should be applied.8

III. The U.S. experience shows that enforcement agencies, working within the framework of a flexible set of core statutory provisions, can use various forms of guidance both to shape the underlying legal standards and to affect the behavior of private parties. In thinking about how an agency can most constructively provide guidance addressing single-firm exclusionary conduct rules without undermining its legitimate enforcement prerogatives, it is useful to consider the experience that US competition agencies have had with guidelines and similar policy statements in many other areas of competition law. Over the past fifteen years, the US Department of Justice and the Federal Trade Commission have jointly issued seven formal sets of guidelines.9 Over the preceding quarter century, the Justice Department alone had already issued another eight.10 We have learned a great deal from that experience. I will first discuss the lessons learned in general terms and then, in the next section, I will address the application of those lessons to guidance for single-firm exclusionary conduct. First, we have learned that guidelines and other formal policy statements can serve many purposes. Their principal goal is to declare the agency’s 6

Weyerhaeuser, Slip at 12. Trinko, Slip at 10–11, 14. 8 See Thomas Barnett, “The Gales of Creative Destruction: The Need for Clear and Objective Standards for Enforcing Section 2 of the Sherman Act”, Opening Remarks for the Antitrust Division and Federal Trade Commission Hearings Regarding Section 2 of the Sherman Act (Washington, D.C., 20 June 2006). 9 The two agencies jointly issued guidelines or similar policy statements on horizontal mergers in 1992 (revised in 1997), on competition in health care industries in 1994 (revised in 1996), on licensing of intellectual property in 1995, on international operations in 1995, and on collaborations among competitors in 2000. These are all available from the agencies’ websites. 10 The Department issued guidelines or similar policy statements on mergers in 1968 and 1982 (revised in 1984), on international operations in 1972, 1977, and 1988, on research joint ventures in 1980, and on vertical restraints in 1985. 7

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 229

Standards for Single-Firm Conduct: a US Perspective

229

enforcement intentions and to assist the business community in an area in which policy is evolving or simply unclear. Guidelines can reassure the business community by declaring that certain conduct will not be subject to an enforcement action, and they can forewarn the business community by declaring that certain conduct will be subject to an enforcement action. Sometimes the business community has perceived significant risks of antitrust enforcement that were not really there, and we responded with policy statements providing reassurance, especially when we thought that the conduct at issue was apt to be efficient and was being chilled by perceived risks. The objective of providing reassurance was particularly important in the merger field in the 1980s, when merger policy was in transition and moving away from the era of Von’s Grocery.11 In that case, the government challenged a merger yielding a firm with 7 percent of the market, and the dissenting Justice observed that the only apparent consistency in US merger policy was that the government always wins.12 Reassurance was likewise a key objective of guidelines in the 1980s and 1990s that educated businesses about changing attitudes toward efficiency-enhancing intellectual property licensing restrictions. The 1994 health care guidelines arguably served both objectives: putting participants in this vital industry on clearer notice that the antitrust laws applied to them, while at the same time providing reassurances regarding the broad scope of conduct permitted under those laws. Although providing guidance to the business community has been a main purpose of competition guidelines in the United States, such policy statements serve other purposes as well. Perhaps most obviously, our guidelines assist practitioners in counseling clients and they help frame the discussion when practitioners appear before the agencies on behalf of their clients. Wellcrafted guidelines establish a framework for analysis and a vocabulary for discussion, which tends to make our investigations more efficient as parties conduct analyses of, and focus our attention on, the facts most likely to be dispositive. This has been particularly valuable in merger enforcement, where the time pressures associated with our review are often acute. Our guidelines are also designed to play a role in the evolution of competition law. In many settings, courts have looked to agency policy statements for guidance and have been willing to follow that guidance where the agencies’ policies are well supported by case law, scholarship, or the agency’s own wellarticulated reasoning. One major success in this regard is the “hypothetical monopolist” paradigm for market definition, which was introduced in the Justice Department’s 1982 Merger Guidelines. This simple analytical tool seemed to many at the time to be both a sharp break with case law precedent and a wholly impractical theoretical construct. But over time views changed markedly. For many years now, not just enforcers but counselors and courts 11 12

United States v. Von’s Grocery, 384 U.S. 270 (1966). Id. at 301 (Justice Stewart dissenting).

(J) Ehlermann 08 Meyer

230

6/11/08

11:49

Page 230

Session One: The Sherman Act And Abuse Of Position Under Art 82

in the US—and not just in the US but around the world—have been defining relevant antitrust markets by asking whether a hypothetical monopolist could profitably increase price by a small but significant amount for a nontransitory period. This success provides a good illustration of how guidelines issued by enforcers can change the law and influence behavior. Guidelines and similar policy statements can also be useful in communicating with agency staffs. In a large agency, the best way to get everyone on the same page is to do so literally: by putting policies and procedures in writing. This serves two important functions. On the one hand, the process of writing down policies usefully frames unanswered questions, exposes ambiguities, and generally focuses the attention of an agency’s policy makers. On the other hand, having written policies serves to guarantee that everyone in the agency gets exactly the same message. This is especially important when there is significant turnover within the organization. Second, we have learned from experience that our guidelines provide advice that is at its most concrete and useful when they set forth clear statements of conduct that will not be challenged except in extraordinary circumstances—what are often referred to as antitrust “safety zones” or “safe harbors”. Our horizontal merger guidelines spell out clear concentration measures and indicate that transactions in unconcentrated markets, or that have only modest effects on concentration, are highly unlikely to raise serious competitive concerns. Likewise, our 1993 statements of Antitrust Enforcement Policy in the Health Care Area spell out six explicit safety zones for various categories of conduct—such as group purchasing arrangements where the participants account for less than 35% of purchases and 20% of sales in the downstream market—that will not be challenged because they are likely to generate pro-competitive benefits and highly unlikely to cause anticompetitive harm. We are cognizant, however, that the articulation of such safety zones must be accompanied by a very clear message that conduct outside those zones will not automatically be condemned—lest our guidance induce parties to act too conservatively by limiting their conduct to that covered by the safe harbors. Third, we have found that parties act not only on guidance we convey explicitly in the text of our guidelines—i.e., the analytical framework, statements of presumptions and safe harbors—but also on the enforcement attitudes revealed in the tone and tenor of those guidelines. The audience of our guidelines includes both experienced competition law practitioners, who will carefully parse the precise language used to articulate our enforcement policy, as well as many who know little of competition law and on whom subtle nuances may be wasted. Those with less intimate knowledge of the inner workings of competition law often will form impressions based on what they perceive to be the enforcement attitudes reflected in the guidelines. Guidelines and other policy statements always send a message—whether overtly or subliminally—of reassurance, hostility, or something in between. When we craft

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 231

Standards for Single-Firm Conduct: a US Perspective

231

our guidelines we must work hard to ensure that the intended message gets through. For example, in our Antitrust Guidelines for Collaborations Among Competitors, we were careful to send the right messages about different categories of conduct: emphasizing a hostile view of naked agreements and a neutral analytical stance towards other competitor collaborations that have pro-competitive potential. Clearly, it does not serve the ends of competition policy to send a message of hostility with respect to conduct that is most often pro-competitive, or to send a message of reassurance with respect to conduct that is most often anticompetitive. Fourth, we have learned that guidelines can be a two-edged sword. To be useful to the business community, guidelines must limit an agency’s discretion by setting out general limiting principles and by establishing specific limits on the potential application of the relevant provisions of law. But these same limits can end up binding the agency in those cases when it concludes an exception is warranted. We have found that the courts are responsive to arguments by targets of an agency’s enforcement actions that enforcement of the antitrust laws against their conduct would be inconsistent with the agency’s officially stated policy. This presents a substantial challenge in drafting guidelines, especially because it is impossible to anticipate and analyze all possible circumstances. Striking the right balance is always difficult, and in some complex areas it may prove infeasible to assist the business community substantially without unduly limiting an agency’s discretion. In addition, zealous advocates for the parties sometimes assert inconsistencies between the agency’s enforcement action and its stated policies even when such inconsistencies do not actually exist. Experience has taught that exquisite care must be taken in the drafting of guidelines to limit opportunities for creative advocates to assert false inconsistencies. Language that appears to its drafters to be unambiguous may be understood in a different way by others, so guidelines require extensive vetting that affords ample occasions to discover and correct failures to communicate. Finally, as suggested by some of the preceding observations, issuing formal guidelines or policy statements is not always the right thing to do. To be sure, a basic principle of good government is that agencies should communicate their practices and policies to the public. Businesses directly affected by agency actions should be informed about what the agencies do and why. But if the legal framework is too complex to express in simple terms, or if clear safe harbors are too difficult to articulate, providing formal guidance that reserves sufficient enforcement flexibility for the agency might entail creating even more uncertainty than existed without the guidelines—or an atmosphere in which perceived risks are elevated rather than reduced. In such settings the agency’s desire to be transparent might better be served by communicating in less formal ways. Sometimes, actions can speak louder than words; agency enforcement decisions, especially if accompanied by detailed explanations of the underlying facts and reasoning, can inform the public more effectively

(J) Ehlermann 08 Meyer

232

6/11/08

11:49

Page 232

Session One: The Sherman Act And Abuse Of Position Under Art 82

than general policy statements. The US agencies have also communicated their views on enforcement policy in other ways, for example by posting court filings on the agency’s web site13 and by promoting public speeches by agency officials.14

IV. Some Thoughts on the Role of Guidelines Addressing Competition Law as Applied to Single-Firm Exclusionary Conduct What do these lessons tell us about how an agency can productively advance the goal of certainty with respect to standards governing single-firm conduct by dominant firms? This is a subject that the US agencies are evaluating carefully based on what we learned from our year-long series of hearings on single-firm conduct standards, and we look forward to communicating those findings to the public in some fashion in the not-so-distant future.15 Given the present state of the law on single-firm exclusionary conduct in both the United States and Europe—and in particular given the lack of clear standards applicable to the broad array of conduct that might form the basis of a claim that a dominant position was achieved or maintained through improper means—our experience suggests that guidelines on single-firm conduct would be most useful if they substantially increased clarity in this important and difficult area of competition policy. Guidelines that merely articulate alternative theories or that simply list factors considered in the agency’s analysis might be of some use in steering parties and courts away from inappropriate modes of analysis, but they would not provide much useful concrete guidance to businesses. Similarly, although guidelines might usefully articulate modern thinking on the ways in which single-firm conduct can be exclusionary, and though they may caution businesses and their counselors to pay attention to competition law risks when they structure their affairs, such guidelines would not significantly assist the business community in formulating their strategies. Nor would they significantly assist practitioners in counseling their clients or significantly assist the courts in wrestling with difficult cases. Useful guidelines on single-firm exclusionary conduct should do more, by helping those firms understand when they need not be sensitive to competition 13 Especially useful are the briefs in these cases that were filed in the courts of appeals. These briefs are available at: http://www.usdoj.gov/atr/public/appellate/appellate.htm. 14 Recent speeches by DOJ officials are available at http://www.usdoj.gov/atr/public/speeches/ speeches.htm. 15 Extensive materials relating to the hearings are available at: http://www.usdoj.gov/atr/ public/hearings/single_firm/sfchearing.htm.

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 233

Standards for Single-Firm Conduct: a US Perspective

233

law rules, and by assisting them in avoiding risks without unduly sacrificing desirable pro-competitive benefits. In an area of otherwise open-ended or hard-to-understand rules, guidelines can do this only by both setting out a general set of principles that will guide analysis and establishing specific limits on the application of those principles to the facts. Equally important, the general tenor of any guidelines should communicate a message that is not susceptible to misinterpretation as being unduly hostile to most single-firm conduct, but that instead adequately reflects the fact that in most circumstances the kinds of conduct addressed by the guidelines would be pro-competitive rather than harmful to consumers (even if such conduct might in extreme cases constitute an abuse of dominance). Guidelines should not leave firms that possess only modest market share with the impression that they might well be found dominant and that much of their day-to-day unilateral conduct could violate the law. Guidelines on abuse of dominance could most effectively assist the business community, practitioners, courts, and even the administering agencies by setting out safe harbors. Such safe harbors would generally be warranted when conduct is sufficiently unlikely to be harmful that reserving the ability to challenge it does not justify the administrative costs of further analysis under the applicable legal framework, or the potential consumer harm associated with erroneously condemning the conduct, or the chilling effect associated with uncertainty. Two categories of safe harbors would appear most promising. The first and most important safe harbor would address the threshold question of “dominance”. Such a safety zone would provide tremendous “bang for the buck” because it would tell all businesses that are not realistically “dominant” that they needn’t worry about the law’s prohibitions at all. Of course, selecting an appropriate level for a market share safe harbor is a delicate matter. Adopting a safe-harbor market share threshold that is very high might create substantial legal certainty but at an unacceptably high cost in terms of consumer harm from exclusionary conduct by firms that would fall within the safe harbor yet have the unilateral ability to injure the competitive process. But it does not follow that a safe harbor should employ a very low market share threshold. A market share safe harbor enhances legal certainty only to the extent that it affords real and reliable safety to firms that might otherwise have perceived a non-trivial risk of being found dominant. Establishing a safe harbor that applies only to firms with very low market shares might well increase business uncertainty by suggesting a greater likelihood that competitors just outside the safe harbor will be found dominant. Creating a rebuttable presumption that a firm with less than some given market share lacks dominance might provide some useful guidance, but our experience tells us that that benefit would be limited. A mere presumption does not provide sufficiently useful assurances to the business community and it does not reduce the burdens placed on the enforcement institutions

(J) Ehlermann 08 Meyer

234

6/11/08

11:49

Page 234

Session One: The Sherman Act And Abuse Of Position Under Art 82

(agencies and courts) that must evaluate claims of abuse by firms that have little likelihood of being dominant. So where should the safe harbor threshold for dominance be? Speaking from US experience, where the legal standard is “monopolization”, our case law would strongly suggest that there is little possibility of a firm having monopoly power unless it holds more than 70 percent of the market. Even when a firm’s market share is well above this level, there is often healthy debate whether its power is sufficiently real and durable. To the extent that competition law is concerned with the ability of a single firm acting unilaterally to cause significant harm to the competitive process, it is hard to imagine a firm having that ability if its market share is under 50–60 percent. Another important challenge is to provide safe harbors for particular forms of conduct engaged in by firms that are potentially dominant. Four years ago, former Competition Commissioner Mario Monti declared: “To define the notion of ‘exclusionary abuse,’ one should therefore make a distinction between competition on the merits, or ‘normal’ competition, and other types of behaviour.”16 Useful guidelines would flesh out this important concept both as a matter of general principle and in terms of specific safe harbors for limited categories of conduct. We recognize that this is not such an easy task, and that difficulty is one of the many reasons it is hard to produce useful guidelines on abuse of dominance. DG Competition’s Discussion Paper17 articulated circumstances in which particular practices should be presumed to infringe Article 82. There is no doubt that presumptions of this sort can play a useful role in abuse of dominance guidelines. Where particular conduct by dominant firms is very likely to harm the competitive process, and where it very clearly does not constitute potentially pro-competitive competition on the merits, it is valuable to send a clear signal that the conduct is likely to be problematic. An entirely openended inquiry into competitive effects would maximize uncertainty in the business community as well as the burden placed on the agencies and courts. However, it is important to ensure that presumptions of infringement are not too easily invoked, even when they are rebuttable. Otherwise, firms may find it too risky to adopt strategies that include conduct that would presumptively (though rebuttably) infringe Article 82 (or Section 2), even when such strategies would benefit competition and consumers. Firms would correctly perceive that overcoming a presumption would necessarily be an uncertain prospect, and that it could entail significant expense. For purposes of this paper it would not be productive to analyze whether a safe harbor would be appropriate for each of the particular categories of 16 Comments of Mario Monti, “Introductory Remarks”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds, European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 5 et seq. 17 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf.

(J) Ehlermann 08 Meyer

6/11/08

11:49

Page 235

Standards for Single-Firm Conduct: a US Perspective

235

conduct that might implicate Article 82 (or Section 2), especially since the work of the US agencies on this subject is far from complete. But I will address one type of conduct—predatory pricing. With respect to this practice, the competition enforcement community has a large body of scholarship and substantial case experience to draw on, so it is relatively easy to say what abuse of dominance guidelines could usefully do. First, guidelines should unequivocally state that a dominant firm engages in lawful competition on the merits when its aggressive price cutting does not go below its cost. Second, guidelines should articulate as clearly as possible how the agency would go about comparing prices to costs. Finally, guidelines should address the issue of recoupment. As probably everyone here is aware, the predatory pricing case law in the United States, but not in Europe, imposes on plaintiffs the burden to demonstrate a reasonable prospect that the defendant could recoup the losses it suffered while charging prices below its costs. Our experience has been that the recoupment requirement usefully focuses a Section 2 predatory pricing case on the critical issues. For example, in some cases it may be clear that recoupment is infeasible because the dominant firm’s scheme of offering low prices has already lasted too long for it to make economic sense as an investment in obtaining future monopoly power. The impossibility of recoupment in such a case would be powerful evidence that the conduct merely reflects persistent and intense competition on the merits, and it would eliminate the need to address more difficult issues such as a comparison of the firm’s prices with its costs.18

V. Conclusion Developing sound legal standards for single-firm exclusionary conduct is hard work. The variety of means by which a dominant firm might conceivably harm the competitive process is quite broad. But those same means very often—and likely most often—are used by firms without market power (as well as firms with such power) to carry on their business in the manner they deem most efficient, most productive, and most profitable. In the vast majority of cases, therefore, those means reflect rather than override the competitive process and benefit rather than harm consumers. As a result, uncertainty

18 The dominance requirement of Article 82 does not necessarily subsume the recoupment requirement. Recoupment requires not merely that a competitor be dominant at the outset; its predatory pricing must sufficiently enhance or preserve its dominance to make the short-term losses pay off as an investment in future market power. Hence, recoupment poses a question distinctly different from the question of dominance, and guidelines on predatory pricing should reflect this insight.

(J) Ehlermann 08 Meyer

236

6/11/08

11:49

Page 236

Session One: The Sherman Act And Abuse Of Position Under Art 82

regarding the line of illegality is likely to sacrifice exactly what competition law is designed to protect. Legal standards must be cognizant of this basic conundrum, and they must strive to provide greater certainty to businesses. In the United States the enforcement agencies are in the midst of studying how the US regime can be improved without giving up the ability to challenge single-firm conduct in those situations where it truly threatens to harm competition. Our broad experience with enforcement guidelines suggests one path that holds promise, both in the United States and here in Europe: guidance that assures most businesses that their day-to-day striving to succeed will not run afoul of competition law rules, while preserving the agencies’ ability to act where necessary to challenge bona fide threats to the competitive process.

(K) Ehlermann 08 Vickers

6/11/08

11:49

Page 237

II John Vickers* Speaking Note, Panel II (“A reformed approach to Article 82 and the US practice: an overall appreciation”) June 2007

Introduction Claus originally asked me to provide an overall “appreciation” of the Commission’s new draft guidelines on Article 82. It turns out, as we all know, that the draft guidelines are so new as not yet to have come into existence, but exist they surely will—indeed must—or else a great opportunity will have been missed by the Commission. It is in fact much easier to discuss nonexistent guidelines than actual ones since, as with per se approaches to antitrust, one is less trammelled by facts. And I hope, but did not risk checking with Claus, that one can respectably get away with submitting a much shorter written contribution in advance of the Workshop than would otherwise have been appropriate.1 This minuscule paper has four parts. The first offers a brief appreciation (in both senses) of the Commission’s draft Guidelines on the assessment of nonhorizontal mergers, and their relevance to the Article 82 guidelines to come. Next are some comments on the threshold of dominance, where the staff discussion paper of December 2005 may inadvertently have created an unfortunately over-interventionist impression. Third, I offer some remarks about possible guiding principles such as consumer harm, the sacrifice (or “but for”) test and the “as-efficient competitor” standard. Finally, in view of the reference in the remit of this panel to US practice, which is largely for David Meyer to discuss, I will say a word on one transatlantic issue: comparative institutional dynamics.

* Drummond Professor of Political Economy, Oxford University. 1 I discuss abuse and dominance, respectively, at greater length in “Abuse of market power”, 115 Economic Journal 244 (2005) and “Market power in competition cases”, 2 European Competition Journal 3 (2006).

(K) Ehlermann 08 Vickers

238

6/11/08

11:49

Page 238

Session One: The Sherman Act And Abuse Of Position Under Art 82

The draft non-horizontal merger Guidelines The Commission’s draft non-horizontal merger Guidelines, published in February 2007, are relevant to the future guidelines on Article 82 because the two need to be consistent. Indeed, the economics of how non-horizontal mergers can lessen competition (or “significantly impede effective competition”, if you will) are closely related to the economics of exclusionary abuses. There are several features of the draft Guidelines that I hope will be echoed in the guidelines to come on Article 82. First, the draft non-horizontal merger Guidelines are explicitly consumeroriented throughout.2 Second, competitor protection is pithily rejected: “the fact that a merger affects competitors is not in and of itself a problem” (paragraph 16). Accordingly, “foreclosure” is not necessarily a problem. The concern is about anticompetitive foreclosure—i.e., foreclosure leading to consumer detriment (see paragraph 18). Third, the draft Guidelines are clear that non-horizontal mergers provide substantial scope for efficiencies (paragraph 13). The same is also true of a range of vertical practices other than mergers. Fourth, the draft Guidelines set out the principal theories of harm to competition. Thus, for example, section IV.A explains how vertical mergers may lessen competition to the detriment of consumers by means of (i) input foreclosure and (ii) customer foreclosure. Fifth, the draft Guidelines identify some of the crucial questions of fact that theories of harm to competition must face. A good example is the statement in paragraph 111 that “[i]t is only when a sufficiently large fraction of market output is affected by foreclosure resulting from the merger that the merger may significantly impede effective competition”.3 In sum, the approach to merger appraisal reflected in the draft guidelines is to state theories of harm to competition and consumers, then to see whether the market facts fit those theories, and then to make an overall assessment, keeping in mind possible efficiencies, of the likely overall impact on competition and consumers. There is a lot to be said for approaching Article 82 cases in very much the same way.

2

With the term “consumer” encompassing both intermediate and final consumers. This is said in a section on conglomerate mergers but the point is surely general, and indeed paragraph 73 on vertical mergers has more or less the same form of words. 3

(K) Ehlermann 08 Vickers

6/11/08

11:49

Page 239

Speaking Note

239

Approach to dominance Whatever one’s concept of abuse, it can happen only in the presence of substantial market power. Absent collusion, a firm can have substantial market power only if it has a large share of a market that is well-defined in an economic sense. The first of these statements does not itself determine whether dominance analysis has value because the propositions (a) “If no dominance, then no abuse”, and (b) “If abuse, then dominance”, are logically equivalent. In practice, however, (a) wins, and dominance analysis should continue to come first in casework. That is not least because, again as with mergers, it allows the efficient screening out of relatively unproblematic cases. In the area of Article 82 this seems especially important to do. If and when abuse analysis is reached, dominance findings should remain centre stage, for an “abuse” with a tenuous nexus to market power is probably not abuse. Within dominance analysis, market definition and market share analysis have their place so long as it is remembered that substitutability is a matter of degree,4 and that a large market share is a necessary but not a sufficient condition for substantial market power. So a market share that is not large is a good indicator of an absence of dominance, and hence of an absence of abuse. It was therefore a bit surprising that the Commission Staff Discussion Paper of December 2005 flirted with market shares as low as 25%. Having stressed helpfully that market power “may derive from several factors which, taken separately, are not necessarily determinative”, the Discussion Paper spoke of (single-firm) dominance as being: (i) “very likely” with a share persistently above 50% provided that rivals have much smaller shares, (ii) “more likely” with a 40–50% share than below 40%, but (iii) “not likely” below 25%. A firm with 25% of an over-broad “market” could be dominant, but that is a reason to define markets well, not a reason why guidelines need to discuss the possibility of (single-firm) dominance where a firm’s market share is only 25%. So on market shares, which could anyway do with de-emphasis, why not simply say that dominance is more likely above 50% than below, and that it is not likely below 40%, provided that the market is well-defined? Anyhow, I suspect that most (good) cases will be in territory well north of 50%, and that there will also be evidence of dominance much more compelling than market share figures alone can be.

4

For an elaboration of this point, see the 2006 paper cited supra note 1.

(K) Ehlermann 08 Vickers

240

6/11/08

11:49

Page 240

Session One: The Sherman Act And Abuse Of Position Under Art 82

Approach to abuse The discussion above of the draft non-horizontal merger Guidelines identified five features that I hope will be echoed in the contemplated Article 82 guidelines: —the consumer leitmotiv; —explicit recognition that the task is not to condemn foreclosure but to distinguish anticompetitive from pro-competitive foreclosure; —allowing clear scope for efficiency defences (objective justification in Article 82 terms) and, where appropriate, balancing; —clarity about theories of harm to competition and consumers; and —a corresponding indication of key factual issues in cases. In the context of exclusionary conduct under Article 82, the second of these points is the classic and inescapable issue of what is, and what is not, “competition on the merits”. At least three types of general standards have been debated: (a) the sacrifice/“but for” test, (b) the “as-efficient competitor” test, and (c) the consumer harm test.5 Each comes in various specifications, for example according to whether the test is seen as necessary and/or sufficient for a finding of exclusionary abuse. Traces of all three standards can be found in the (relatively well-developed) case law on predatory pricing. The cost benchmarks can be related to the “as-efficient competitor” test. Below-avoidable-cost pricing entails short-run sacrifice. Subsequent consumer harm follows if there is recoupment, which under Brooke Group must be shown in US law, but which has not generally been recognized as a condition of abuse in EC law. Provided that recoupment is interpreted broadly—e.g., so as to include other-market recoupment via reputation effects—it might be good practice for the Commission to address it explicitly in cases. Even where the likelihood of recoupment could reasonably be inferred from dominance, there would be little harm in saying that, and it would underline the ultimate consumer orientation of the Commission’s competition policy. The sacrifice/“but for” test has a certain intuitive appeal, but if it is to have a central role in guidelines, at least three issues need to be addressed. The first is circularity. It would obviously be unhelpful simply to say that conduct is anticompetitive if-and-or-only-if it entails profit sacrifice but for its anticompetitive effect. That would leave the fundamental issue—the meaning of “anti-competitive”—hanging in mid-air. How to bring it down to earth? Second is the point that not all abuse entails profit sacrifice even in the short run. Third is the question of the counterfactual: sacrifice relative to what? Not full-blown short-run profit-maximization, presumably. But then what? 5

These are discussed at greater length in the 2005 paper cited supra note 1.

(K) Ehlermann 08 Vickers

6/11/08

11:49

Page 241

Speaking Note

241

The question of the counterfactual must also be addressed by standards cast in terms of consumer harm. Failure to maximize consumer well-being cannot be the litmus test of abuse. But it would certainly seem wrong for conduct to be considered abusive if consumers would probably do no worse with that conduct than with what the firm would reasonably do if the conduct were prohibited. The “as-efficient competitor” standard avoids some of these difficulties, but not others. In essence, this standard says that conduct by a dominant firm is abusive if, without objective justification, it would tend to foreclose firms no less efficient than the dominant firm (at operating the business in question). One main line of price squeeze case law, for example, seems to accord with this principle. In short, this standard says that the foreclosure that is abusive is that of as-efficient competitors. Such foreclosure is more than likely to be anti-consumer. A difficulty, however, is that under theories of “divideand-rule” exclusion, a dominant firm might, for example, operate a series of exclusive contracts precisely to deny rivals economies of scale and thus to remain more efficient than those rivals in operating each incremental unit of business. This needs to be allowed for. The “as-efficient competitor” standard is not a panacea. The fact that the question of what is competition on the merits is not easy is all the more reason squarely to address it in the guidelines. Likewise in cases, it seems far preferable to state and test theories of harm to competition and consumers against the facts than to condemn conduct for its form.

Institutional dynamics At the outset I stressed the importance of the Commission taking the initiative by way of preparing guidelines towards a reformed approach to Article 82. Otherwise, there is a serious risk of inertia in the context of EC institutional dynamics, or lack thereof. The interpretation of US antitrust law in the light of new understanding has undergone considerable change in certain periods, notably (but not only) the treatment of non-price vertical restraints in the 1970s. (That flexibility, depending on whether and how it is exercised, has pros and cons.) But the US institutional setting is perhaps more open to change. The US agencies for the most part bring cases in court rather than taking administrative decisions (though the FTC may pursue administrative proceedings). Most US cases, including a number that have shaped the law, stem from private actions. And US judges openly disagree with each other. By contrast, apart from the fact that the European Court of Justice does not always share the Opinion of the Advocate General, the Community Courts present a unanimous front. So too, incidentally, does the European

(K) Ehlermann 08 Vickers

242

6/11/08

11:49

Page 242

Session One: The Sherman Act And Abuse Of Position Under Art 82

Central Bank, unlike the US Federal Open Market Committee and (more so) the Monetary Policy Committee of the Bank of England. This may be natural and even desirable, given that the EC consists of national Member States, but a source of institutional dynamics may be lost, or at least obscured, making it all the more important for the Commission to take the lead. The Courts’ readiness to accept a reformed approach to Article 82 is of course a great unknown. Few signs are to be found in this year’s judgments from the CFI in Wanadoo and from the ECJ in British Airways. On the other hand, in September 2006, the CFI, in the Spanish GlaxoSmithKline case, rejected formalistic arguments in an Article 81 context, as had Advocate General Jacobs in the Greek GSK case concerning Article 82. And the CFI merger judgments in 2002 gave considerable impetus to a reformed approach to the ECMR following the Commission’s launch of its policy review in that area. Whatever may emerge from Microsoft, the Commission needs to press the Article 82 review to a conclusion, in the shape of guidelines and of course casework. Say what you do, and do what you say.

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 243

III Jochen Burrichter* A Reformed Approach to Article 82: The Impact on Private Enforcement June 2007

I. Private Enforcement of Article 82 1. Consequences of the Courage and Manfredi Judgments Although the Courage1 and Manfredi 2 judgments of the European Court of Justice dealt with the enforcement of Article 81, there can be no doubt that the principles set forth in these judgments also apply to the private enforcement of Article 82. In the words of the Court in Courage, the “full effectiveness” of Article 82 and the practical effect of the prohibition laid down in Article 82 would be put at risk if it were not open to any individual to claim damages for losses caused by an abuse of a dominant position. Consequently, national law has to provide all means necessary for the efficient private enforcement also of Article 82.

2. Relevant Areas of Private Enforcement of Article 82 a) Invalidity of Abusive Agreements Although Article 82 does not provide for an equivalent of Article 81(2),3 its effectiveness requires that national law provide for the nullity of abuses of dominance in the form of agreements or any other acts which may have effects in civil law, e.g., unilateral declarations by which a party terminates an agreement. Most national laws provide for the nullity of any acts which violate legal prohibitions such as that contained in Article 82. * Partner, Hengeler Müller. 1 C-453/99, Courage v Crehan [2001] ECR I-6297 2 Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi and Others v Lloyd Adriatico and Others [2006] ECR I-6619. 3 Article 81(2) states that “[a]ny agreements or decisions prohibited pursuant to this article shall be automatically void”.

(L) Ehlermann 08 Burrichter

244

6/11/08

11:50

Page 244

Session One: The Sherman Act And Abuse Of Position Under Art 82

b) Injunctive Relief Even more than in cases of violations of Article 81, the effectiveness of Article 82 requires an efficient system of injunctive relief against abuses of dominance at an early stage. The experience of German competition law in cases of abuse of a dominant position, and especially of a position of “superior” market power below the threshold of dominance, is an example of very efficient control of abuses at an early stage by means of injunctive relief granted by the courts. Injunctive relief should not require that the abuse of a dominant position has already occurred if it is imminent. Nor should it be necessary to show that damages have already been caused, or to establish fault on the part of the dominant firm. For an efficient application of Article 82 it is also helpful to have the possibility of obtaining preliminary injunctions, provided that, if the court in the ordinary procedure subsequently decides not to confirm such an injunction, the plaintiff compensates the defendant for damages suffered as a consequence of the preliminary ruling.

c) Claims for Damages Private enforcement of claims for damages in cases of infringements of Article 82 is as important as in the case of infringements of Article 81. However, plaintiffs are faced with even greater difficulties as regards access to evidence and satisfying the burdens of proof with respect to: —the existence of dominance; —the abuse of dominance; —the existence of damages and the amount of claims; —the causal link between the abuse of dominance and the damages; and —fault. For an efficient private enforcement of Article 82, it is important to establish a reasonable system of rebuttable presumptions or rules that shift the burden of proof.

d) Stand-alone and Follow-on Actions It follows from the clandestine character of cartels that most private enforcement against infringements of Article 81 has been and very likely will continue to consist of follow-on actions after a cartel has been detected by competition authorities, in particular as a consequence of successful leniency programmes. By contrast, infringements of Article 82 are not normally clandestine acts. The abusive practices of dominant firms are committed in the open marketplace, and their economic consequences are immediately relevant for competitors, purchasers, suppliers and final customers. It is therefore

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 245

A Reformed Approach to Article 82: Impact on Private Enforcement 245 very likely that an efficient system of private enforcement of Article 82 will lead to more stand-alone actions compared to the field of Article 81. Such stand-alone actions, in turn, can be expected to relieve the competition authorities of their workload to a greater extent than actions in connection with Article 81. The benefits of efficient private enforcement of Article 82 should encourage efforts to enhance such enforcement, notably by means of workable guidelines. They also underline the need to introduce rebuttable presumptions or workable rules for shifting the burden of proof, provided such procedural devices are economically reasonable.

3. Standing The Courage and Manfredi judgments provide useful guidance for purposes of defining who is entitled to initiate a lawsuit in cases of violation of Article 82. The following rules can be established:

a) Actual Purchasers/Suppliers Abusive prices or tying and bundling practices are directly aimed at actual purchasers and/or suppliers. As the Court stated in Courage,4 and as was later confirmed in the Manfredi judgment,5 any individual suffering losses from the violation of the competition provisions is entitled to claim damages provided that there is a causal link between the abusive practice and the loss suffered. Consequently, actual purchasers and suppliers have sufficient standing to bring actions against abusive practices by dominant firms.

b) Potential Purchasers and Suppliers In cases of abusive refusals to supply, there is no question that the party requesting the supply has sufficient standing to claim damages or to initiate an action for injunctive relief. It is especially in these cases where injunctive relief may become an important means for the private enforcement of Article 82.

c) Actual or Potential Competitors In cases of exclusionary abuses, predatory pricing, single branding, rebates, or tying and bundling, the standing of actual or potential competitors has to 4 5

Supra note 1, paras. 24–26. Supra note 2, para. 61.

(L) Ehlermann 08 Burrichter

246

6/11/08

11:50

Page 246

Session One: The Sherman Act And Abuse Of Position Under Art 82

be decided. Again, under the Courage and Manfredi judgments, there is no doubt that actual or potential competitors have sufficient standing to bring private actions, namely as regards applications for injunctive relief. In these cases, the question is whether a plaintiff has sufficient standing in a case where the abuse of the dominant position relates to a market other than that in which the plaintiff is active. For example, the German courts have ruled that standing is not limited to participants on the dominated market but may also extend to participants on “third markets” if a causal link between the allegedly abusive behaviour and the negative impact on competition in the third market can be established.

d) Purchasers/Suppliers of Actual Competitors Whereas the above cases seem relatively easy to assess, the question is whether purchasers and suppliers of actual competitors which are only indirectly affected by the abusive practices of the dominant firm have sufficient standing. It follows from the Courage and Manfredi judgments that the standing of such purchasers cannot be ruled out. In these cases, the “passing on” defence problems arise if the direct purchaser or competitor also claims damages.

4. Causation In order to establish that the abusive practices have caused losses, the plaintiff has to provide sufficient evidence that the defendant’s behaviour was the relevant cause for such losses. In rapidly changing market situations this may put a difficult burden of proof on the plaintiff. This is especially so if the plaintiff’s business has collapsed and the plaintiff alleges that this collapse was caused by the defendant’s abusive behaviour. A defence to be expected in such cases will point to the plaintiff’s own inefficient business decisions. In cases where the plaintiff alleges to have been deterred by the dominant firm from entering into a certain market, claiming the loss of anticipated profits, the burden of proof includes a demonstration of the plaintiff’s preparedness to enter the market in question, given that there may be many other reasons why an operator chooses not to enter a market.

5. Calculation of Damages The methods of quantifying damages may differ across cases concerning the various forms of abusive practices:

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 247

A Reformed Approach to Article 82: Impact on Private Enforcement 247

a) Abusive Prices In the case of abusive prices, direct purchasers may quantify their damage by calculating the difference between the abusive price and the hypothetical “but for” price. Here the same difficulties that are associated with Article 81 cases are bound to arise. The “before and after” method compares prices using a temporal parameter, while the “yardstick” approach compares the prices in the market affected by the abusive practices to prices in similar but unaffected markets. All methods show deficiencies if market conditions are unstable or if markets are too different to isolate the effects of the dominant undertaking’s conduct. A cost-based method would calculate the dominant undertaking’s per-unit cost of production, adding an appropriate profit margin. It goes without saying that the establishment of an “appropriate profit margin” may be highly controversial.

b) Tying and Bundling In cases of abusive tying and bundling, the damage may be calculated by reference to the price for the unwanted products or services, or by reference to that price plus additional losses suffered due to reduced sales when the buyer is induced to buy fewer units of the wanted goods. If the buyer is a competitor, the damages could be calculated by referring to the losses of profits incurred due to the restricted demand for its product.

c) Predatory Pricing In cases of predatory pricing, competitors may lose market share and sales, and may eventually be driven out of the market. Damages sustained by such competitors would have to be calculated on the basis of lost profits.

d) Single Branding and Rebates Where the abuse takes the form of single branding or an unlawful rebate scheme, the dominant undertaking’s competitors lose market share and profits because the purchasers in the downstream market will divert their purchases away from the competitors and towards the dominant undertaking. Damages will therefore be calculated as the difference between the profit made during the period of the rebate scheme and the likely profit which could have been made in its absence.

(L) Ehlermann 08 Burrichter

248

6/11/08

11:50

Page 248

Session One: The Sherman Act And Abuse Of Position Under Art 82

e) Refusal to supply In the case of a refusal to supply, the undertaking which has been denied access to products or services will presumably request injunctive relief and must therefore establish that the criteria under which the defendant must grant access are satisfied. This can be substantiated by referring to the access conditions that have applied to third parties. Where damages are sought, the basis for the calculation would be the loss of profits during the period in which access has been abusively denied.

f) Margin Squeeze In abusive “margin squeeze” cases, a vertically integrated dominant firm supplies third parties active on the downstream market and uses its dominance in the upstream market to set prices in a way that renders downstream rivals’ activities unprofitable. The plaintiff must establish its damages by calculating the difference between its actual profit margin and its hypothetical margin.

6. Joint and Several Liability Joint and several liability could only arise in cases of collective abuses of joint dominance, which may be rare in practice.

II. Impact of Draft Article 82 Guidelines (should they emerge) on Private Enforcement In assessing the impact of the (future) draft guidelines on private enforcement, the simple question is whether guidelines would make private enforcement easier or not. Having regard especially to the important role of private enforcement in fighting against abusive practices at an early stage through injunctive relief, it is doubtful whether guidelines would make life easier for plaintiffs.

1. Proving Dominance In previous court practice, the main evidence adduced to establish dominance has been the defendant’s market share in the relevant market. Although this

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 249

A Reformed Approach to Article 82: Impact on Private Enforcement 249 poses difficult questions, at least a competitor plaintiff normally has access to the relevant facts in order to submit sufficient evidence of the market shares of the allegedly dominant firm. As the courts have established certain guidelines as to which market shares indicate a dominant position, any weakening of the dominance test by giving less relevance to market shares and referring to competitive constraints faced by the allegedly dominant firm makes it more difficult for a plaintiff to prove market dominance. The more “intangible” the factors that have to be covered by a plaintiff, the more difficult it will be for him to prove dominance. Only actual or potential competitors, who tend to be familiar with market conditions, may be able to meet such a challenge. The courts could in practice rule that, where the plaintiff submits facts establishing a market share indicative of dominance, the burden of proof is shifted to the defendant. The defendant would then have to show that, in spite of its high market share, there are competitive constraints that exclude such a dominant position. With such an allocation of the burden of proof, the consequences of rendering the concept of dominance more complex could be alleviated.

2. Proving Abuse a) General Approach Thus far in its jurisprudence, the European Court of Justice has basically applied a form-based approach, although it has not embraced per se rules for all forms of possible abusive conduct. It goes without saying that this approach is advantageous for a plaintiff. If a certain type of conduct is generally regarded as harmful and therefore abusive, the burden of proof for establishing a justification for the behaviour rests with the defendant. In its contemplated future guidelines, the Commission intends to move away from a strict form-based approach to an effects-based approach. The envisaged approach concentrates on whether the conduct by the dominant firm results primarily in benefits or in harm to consumers (“consumer welfare balancing test”). The conduct will only be regarded as abusive if the disadvantages outweigh the advantages of the alleged abuse, which may take the form of lower prices, improved quality or improved total welfare from cost savings. It goes without saying that such an effects-based approach makes life more difficult not only for the competition authorities (which, however, have ample means of investigating market conditions) but especially for private plaintiffs, who in most cases will not have access to the facts necessary for the overall assessment and who, in most European jurisdictions, have no procedural means such as discovery to collect the necessary information.

(L) Ehlermann 08 Burrichter

250

6/11/08

11:50

Page 250

Session One: The Sherman Act And Abuse Of Position Under Art 82

b) Welfare Balancing Test and Burden of Proof The disadvantages of the effects-based approach could be avoided by appropriate rules concerning the burden of proof. If the alleged abusive behaviour objectively increases the dominant firm’s market power in the first instance, the burden of proving that the disadvantages for consumers are outweighed by efficiencies clearly should shift to the defendant. This rule would be in line with the rationale of Article 2 of Regulation 1/2003, under which the undertaking claiming the benefit of an exemption under Article 81(3) bears the burden of proving that the conditions of that exemption are fulfilled. It makes sense to apply this principle by analogy for purposes of establishing whether a prima facie case of abuse may be rebutted.

c) Establishing the Counterfactual and its Effects As already mentioned, the new approach for the assessment of abuses requires the plaintiff not only to submit evidence of the behaviour allegedly distorting competition but also to analyze its anticompetitive effects. In comparison to the counterfactual, references to “realistic alternatives in the light of standard industry practice” are of no great help. There may already be doubts as to what constitutes “standard industry practice”. The closer the standard (and also abusive?) industry practice is to the actual conduct of the allegedly dominant undertaking, the less likely it is that the conduct will be found to increase the dominant firm’s market power. The difficulties associated with defining the counterfactual will clearly increase the factual burden on private plaintiffs in Article 82 cases. Those difficulties will increase if the relevant thresholds for the likelihood of the conduct to produce consumer harm are increased. In cases of injunctive relief, only an ex ante likelihood can be a practical test in most cases. The thresholds for the assumption of likelihood of consumer harm should not be raised higher than those already established by the Community Courts and applied, for example, in the British Airways case.6 The same applies to the time-frame of the analysis. To require a demonstration of negative effects well into the future would make a plaintiff’s task still more difficult. In practice, only foreseeable effects can be taken into consideration. In proceedings before a court, long-run effects are very difficult to evaluate, and consequently they may be discounted or disregarded.

6 Case T-219/99, British Airways v Commission [2003] ECR II-5918, upheld on appeal: Case C95/04 P, [2007] ECR I-2331.

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 251

A Reformed Approach to Article 82: Impact on Private Enforcement 251

d) Efficiency Justification Applying the general principles of Article 2 of Regulation 1/2003, there should be no doubt that the defendant must prove a sufficient likelihood of pro-competitive effects for the consumer.

e) “As Efficient Competitor” Screen For price-based abuses, the Commission introduces the “as efficient competitor” screen. Only conduct which would exclude an “equally efficient competitor” is regarded as abusive. An “as efficient competitor” is a hypothetical competitor having the same costs as the dominant firm. It goes without saying that, for a plaintiff in a private action, this test is difficult to meet. At the outset the plaintiff needs to have reliable information on the pricing conduct and the costs of the dominant undertaking. For that purpose, according to the Commission, both the revenues and costs of the dominant undertaking are relevant. In practice, a plaintiff may have access to information on the dominant undertaking’s revenues, but rarely on its cost structure. It must be emphasized that the rules of civil procedure in most Member States of the Community are very restrictive as regards granting access to information which constitutes business secrets of another undertaking. The reference of the Commission to the cost structure of “apparently efficient competitors”7 will not make the plaintiff’s task any easier, as he will not have access to information concerning either the cost structure of the dominant firm or the cost structure of other competitors. Nevertheless, a competitor could try to use its own cost structure, although this would beg the question of whether it is in fact an “efficient competitor”. The basic question which the Commission identifies is whether, under the “as efficient competitor” screen, all competitive advantages of the dominant firm have to be compared to its competitors, including benefits from network effects and information asymmetries, better access to financing or better economies of scale. The consequence of taking into account such competitive advantages would basically mean that only a “clone” of the dominant firm would qualify as an “as efficient competitor”. This would have effects on both the factual and the legal side of price-related abuses. On the factual side, proving the infringement would be more difficult since not only cost-related issues—which are already difficult enough to establish—come into play, but also “intangible” issues that are difficult to measure objectively. On the legal side, including all competitive advantages would raise the bar the plaintiff must pass in order to meet the “as efficient competitor” screen. 7 See DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf, paras. 67, 103 and 164.

(L) Ehlermann 08 Burrichter

252

6/11/08

11:50

Page 252

Session One: The Sherman Act And Abuse Of Position Under Art 82

A solution to part of the problem could consist of shifting the burden of proof to the defendant, at least with regard to the relevance of the other kinds of competitive advantages described above which it enjoys.

3. Impact on Standing As noted above, according to the Courage and Manfredi judgments, national law must entitle any individual to claim damages for losses suffered as a result of violations of EC competition law, including Article 82. The focus of any future guidelines on consumer harm cannot affect the interpretation of EC law by the Court of Justice. Nevertheless, guidelines may have de facto consequences, and this also applies with respect to standing. The more remote a potential plaintiff is from the market, the more difficult it will be, e.g., for an indirect purchaser or the purchaser of goods from a competitor, to substantiate a complaint based on Article 82.

4. Impact on Fault National law may require fault on the part of the defendant in connection with a claim for compensation of damages, whereas fault is normally not a condition for injunctive relief ordering the dominant firm to cease and desist from its abusive behaviour. It goes without saying that, the more “good faith” defences against alleged infringements of Article 82 are open to a defendant in an Article 82 damages case, the better his chances are of “excusing” his behaviour by referring to “reasonable” advice received from, e.g., economists indicating that the behaviour would not violate Article 82, thereby excluding fault.

5. Causation and Damages The difficulties in defining the counterfactual, which have been set out above, will also have consequences for the establishment of causation and of the amount of damages. Defendants could bring forward the argument that damages should be quantified by reference to the losses that the plaintiff would have suffered if the defendant had chosen to undertake the cause of action which was the least costly or most beneficial for the defendant, assuming that both would have been lawful. The court would have to establish a hypothetical course of action in order to decide on the causation issue and on the amount of the damages claimed. The reference to a counterfactual “realistic

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 253

A Reformed Approach to Article 82: Impact on Private Enforcement 253 pricing scheme” could also reduce the amount of damages to be claimed by a plaintiff in comparison to a counterfactual scenario of “no rebates at all”. The “as efficient competitor” screen will have an especially serious impact on the amount of damages which the plaintiff will be able to claim. Since damages have to be assessed by using the “but for” scenario, the present situation takes into account the scenario the plaintiff would have been in but for the infringement. If the plaintiff already competed on the market before the allegedly abusive behaviour, the court can compare the revenue gains by the claimant pre-infringement and during the infringement period. This information is usually easily available to the plaintiff. This approach will significantly change with the application of the “as efficient competitor” screen. If the plaintiff successfully establishes dominance, abuse and causation, it is foreseeable that the defendant will also refer to the “as efficient competitor” screen with regard to the calculation of damages. The benchmark of the damages suffered by an equally efficient competitor will certainly limit the amount of damages to be claimed by a plaintiff.

III. Conclusion Under the contemplated new approach to the application of Article 82, which is based on effects and on harm to consumers, the arsenal of defences of the dominant undertaking as a defendant will be enlarged. It will also be more difficult for a plaintiff, before initiating a lawsuit, to ascertain whether the claim will ultimately be successful, since the effects-based approach requires the provision of information and a degree of economic analysis based on hypothetical assumptions and scenarios. It is clear that, compared to a formbased approach, an economic approach increases the difficulties for plaintiffs and may discourage private enforcement. To a certain extent, these effects can be reduced by developing reasonable rules for shifting the burden of proof based on sound economic findings.

(L) Ehlermann 08 Burrichter

6/11/08

11:50

Page 254

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 287

I Simon Bishop* Loyalty Rebates and “Merger Standards”: A roadmap for the practical assessment of Article 82 investigations September 2007

1. Introduction In December 2005, the European Commission published the DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (“Discussion Paper”).1 That paper set out a proposal for reform of the competitive assessment for these types of commercial practices, i.e., exclusionary practices of dominant firms. The introduction and general tone of that paper, reflecting the comments made by the Commissioner for Competition at Fordham 2005, suggest a move towards an effects-based approach whereby consumer welfare and economic efficiency are the guiding principles in the application of EC competition law. In particular, the Discussion Paper states that: With regard to exclusionary abuses the objective of Article 82 is the protection of competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources. [. . .] In applying Article 82, the Commission will adopt an approach which is based on the likely effects on the market.

The adoption of an effects-based regime would, if realized, be a positive development, and one which would reflect developments in other areas of European competition law.2 Past experience and economic analysis have * RBB Economics. I have benefited from many discussions on the issues discussed in this paper with colleagues at RBB Economics, in particular Adrian Majumdar and Derek Ridyard. However, the views expressed herein are my own and do not necessarily reflect those of RBB Economics. In addition, Mel Marquis provided many insights and comments throughout the drafting of this paper, for which I am very grateful. 1 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (December 2005), http://ec.europa.eu/comm/competition/antitrust/art82/ discpaper2005.pdf. 2 See, e.g., Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, [2004] OJ C31/03; Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings, http://ec.europa.eu/comm/competition/mergers/legislation/notices_ on_substance.html; Commission Reg. 2790/99/EC of 29 December 1999 on the application of

(N) Ehlermann 08 Bishop

288

6/11/08

11:50

Page 288

Session Two: A Reformed Approach to Exclusionary Conduct

shown that the current form-based approach adopted in Article 82 investigations fails to properly protect consumer interests, as it can often discourage pro-competitive practices. Adopting an effects-based approach is particularly important in the assessment of exclusionary abuses for the following reasons. First, most commercial practices that have potential exclusionary effects are also likely to have efficiency-enhancing properties, even when undertaken by dominant firms. A form-based approach therefore runs a serious risk of preventing the attainment of those efficiency gains and suppressing desirable and legitimate commercial freedoms, while not allowing sufficient flexibility to prevent innovative exclusionary strategies. The adverse consequences for competition and therefore consumers from such over-intervention should be obvious. Second, exclusionary practices necessarily harm competition through harm to competitors. This makes it particularly difficult for form-based rules to distinguish between practices that reduce consumer welfare from those that simply harm rival firms through the normal competitive process. To put the point another way, it should be recognized that the loss of market share (perhaps even resulting in market exit), as a reflection of normal competition, is a critical feature of the efficient operation of competitive markets. This is true even in the presence of a dominant firm. In this respect, the Discussion Paper apparently sets out with clear policy objectives. It states:3 [I]t is competition, and not competitors as such, that is to be protected. Furthermore, the purpose of Article 82 is not to protect competitors from dominant firms’ genuine competition based on factors such as higher quality, novel products, opportune innovation or otherwise better performance.

However, as soon as one moves away from the current form-based approach, this raises the obvious question of how to discriminate between pro-competitive and anticompetitive behaviour on the part of a dominant firm. There has been considerable debate as to the appropriate monopolization standard that should be adopted to distinguish anticompetitive conduct from pro-competitive conduct.4 For example, the “no economic sense” test and the Article 81(3) of the Treaty to categories of vertical agreements and concerted practices, [1999] OJ L336/21. 3 Paragraph 54 of the Discussion Paper, cited supra note 1. 4 Herbert Hovenkamp, “Signposts of Anticompetitive Exclusion: Restraints on Innovation and Economies of Scale”, in Barry Hawk, ed., 2006 Fordham Corporate Law Institute, Juris Publishing, 2007, chapter 18; Gregory Werden, “The ‘No Economic Sense’ Test for Exclusionary Conduct”, 31 Journal of Corporation Law 293 (2006); Einer Elhauge, “Defining Better Monopolization Standards”, 56 Stanford Law Review 253 (2003); Douglas Melamed, “Exclusive Dealing Agreements and Other Exclusionary Conduct—Are There Unifying Principles?, 73 Antitrust Law Journal 375 (2006); John Vickers, “Abuse of Market Power”, 115 The Economic Journal 244 (2005).

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 289

Loyalty Rebates and “Merger Standards”

289

“exclusion of an equally efficient rival” test have both been advocated for discriminating between efficient pro-competitive behaviour on the part of a dominant firm and behaviour that is anticompetitive. The “no economic sense” test would not condemn the conduct of dominant firms unless it can be established that that conduct would make no economic sense but for its tendency to eliminate or lessen competition. The “equally efficient rival” test is based on the idea that conduct can only be deemed to represent anticompetitive exclusion if it were to deter entry or force from the market equally efficient competitors. Exclusion under any other circumstances would, according to the test, simply reflect efficient market outcomes. Although such tests provide useful insights into the type of conduct that competition law ought to prohibit, neither of these two tests, nor indeed any other test that has been advanced, is without defects. From the author’s perspective, the appropriate standard against which we should be assessing conduct (albeit often implicitly) is the consumer welfare test: are consumers made worse or better off by a given business practice, taking into account the dynamic nature of competition? This is not to say that competition authorities and practitioners actually seek to directly measure consumer welfare; such a task would likely not be possible in the majority of cases. Rather, it represents a guiding principle. The debate on appropriate monopolization standards and the competing merits of the various tests advanced is therefore best seen as providing a proxy for implementing the consumer welfare standard by seeking to identify “rules of thumb” to help discriminate between pro-competitive and anticompetitive business conduct rather than providing a roadmap for practical assessment. Indeed, the debate surrounding the appropriate monopolization standards is to a large extent academic in the sense that it rarely ventures into the difficult terrain of providing practical steps for actually implementing these standards. This paper is not concerned, therefore, with the merits and drawbacks of each of the various monopolization standards that have been advanced. Rather, it argues that we should focus on the economic principles of the business conduct under investigation and assess that conduct with reference to actual market evidence. In short, this paper argues that the experience of assessing the likely competitive effects of mergers in practice provides the appropriate roadmap for introducing an effects-based assessment into Article 82. By “merger standards”, I refer to two important characteristics of EU merger control; the use of available market evidence; and the presumptions that are adopted in the competitive assessment. As is routinely the case in the area of merger control, much more attention needs to be given to the actual facts of the industry under consideration. In the merger context, empirical market evidence is routinely used to assess current competitive conditions and hence to assess the likely impact of the merger under investigation. Indeed, this has been a cornerstone of the success of EU merger control.

(N) Ehlermann 08 Bishop

290

6/11/08

11:50

Page 290

Session Two: A Reformed Approach to Exclusionary Conduct

The need to introduce “merger standards” into Article 82 extends beyond just consideration of actual market evidence. It also extends to the presumptions that are adopted in the competitive assessment. The competition concern raised by conduct that is deemed to be exclusionary is that it results in market share-shifting away from small rivals of the dominant firm and in consequence reduces the ability of rival firms to compete to such an extent that they are marginalized or driven from the market altogether. In such circumstances, exclusionary conduct by dominant firms might result in increased prices through this foreclosure effect.5 But it is important to recognize that short-run effects that harm competitors also bring direct benefits to consumers. It is therefore important to note that any competition concerns arise indirectly in the long run via the marginalization of competitors rather than arising directly and in the short run. The analysis of potential anticompetitive exclusionary effects must therefore carefully specify the conditions that give rise to the purported outcome, and must go beyond a mere theoretical assessment by accounting for observed industry characteristics and behaviour. This competition concern is analytically identical to standard competition concerns raised by non-horizontal mergers, where the primary concern is also the potential for marginalizing competitors.6 However, although marginalization represents a potential competition concern, it is well recognized that non-horizontal mergers in most instances are pro-competitive. In other words, the (economic) presumption is that non-horizontal mergers do not result in anticompetitive foreclosure even if rival firms are worse off post-merger. Share-shifting (i.e., a reduction of a rival’s market share) should therefore represent only a necessary but not a sufficient condition for concluding that the conduct of a dominant firm should be deemed an exclusionary abuse under Article 82. In short, to establish “anticompetitive foreclosure” (i.e. harm to competition) it is insufficient merely to establish harm to competitors.7 What is important is whether, as a result of the conduct under investigation, rivals are still unable to exert an effective competitive constraint on the dominant firm’s conduct in terms of the prices and quality of products or services it supplies. As an illustration of how such an effects-based approach can be applied in practice, this paper focuses on the assessment of loyalty rebates. The term “loyalty rebate” is capable of encompassing a wide class of discount schemes. 5 Alternatively, entry or expansion may be deterred, allowing the firm to preserve its market power. 6 The same can also be said of the standard competition concerns that are raised by vertical restraints. 7 As the Court of First Instance has noted, for example in its Tetra Laval judgment, such theories require clear and convincing evidence. See Cases T-5/02 and T-80/02, Tetra Laval BV v Commission [2002] ECR II-4381, upheld: Cases C-12/03 P and C-13/03 P, Commission v Tetra Laval [2005] ECR I-987.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 291

Loyalty Rebates and “Merger Standards”

291

Although not precisely defined, a loyalty rebate scheme represents a form of non-linear pricing in which the unit price of a good declines when a buyer’s purchases meet a buyer-specific minimum threshold requirement. It is this characteristic that differentiates a loyalty rebate from other discount schemes or forms of price discrimination.8 This paper focuses on single-product rebate schemes but the principles, with appropriate modification, also apply to the assessment of multi-product rebate schemes.9 Loyalty rebates probably constitute one of the most discussed and hotly debated areas of European competition policy. Traditionally, loyalty rebates have been considered to represent one of the most pernicious anticompetitive practices in which a dominant firm can engage. Whereas the presumption in this area of merger control is that share-shifting rarely gives rise to competition concerns (via the marginalization of competitors),10 the presumption in the area of loyalty rebates under Article 82 is that any share-shifting necessarily gives rise to anticompetitive outcomes. Indeed, there is even a presumption that a loyalty rebate scheme offered by a dominant firm necessarily involves the anticompetitive marginalization of competitors where those rivals are growing.11 Consequently, in its decisional practice, the Commission, supported by the European Courts, adopts in all but name a per se prohibition on a dominant firm employing a loyalty rebate scheme.12 However, that hostile policy stance has been challenged by a number of commentators, including some Commission officials (as evidenced by a softening of the per se stance presented in the Discussion Paper).13 These commentators recognize that loyalty rebates are often employed for pro-competitive reasons even when offered by dominant firms. Hence, although it is acknowledged that under certain circumstances loyalty rebate schemes can give rise to anticompetitive outcomes, these commentators argue that the per se approach is unjustified and gives rise to significant false positives and has the adverse consequence of chilling normal competition.14 8 A loyalty rebate need not necessarily be conditional on reducing purchases from rival suppliers. Whether this is the case depends on the precise form of the rebate scheme employed. 9 See Bruce Kobayashi, “The economics of loyalty discounts and antitrust law in the United States”, 1 Competition Policy International 115 (Autumn 2005). 10 This presumption is articulated in the Commission’s Non-Horizontal Merger Guidelines, cited supra note 2. 11 See Commission Decision 2000/75 of 14 July 1999, Virgin/British Airways [2000] OJ L30/1, upheld: Case T-219/99, British Airways v Commission [2003] ECR II-5917; upheld on further appeal: Case C-95/04 P, [2007] ECR I-2331. 12 That hostility is also reflected in the stance adopted by many national competition authorities in Europe. 13 For example, the Discussion Paper (supra note 1) states, at paragraph 138: “A supplier may use single branding obligations and rebate systems for efficiency enhancing reasons and for anticompetitive reasons and they may have efficiency enhancing effects and anticompetitive effects.” 14 It is important to note that an effects-based system will permit firms to engage in conduct that is pro-competitive, but it will not enable firms to engage in more anticompetitive behaviour. It is a failure of understanding on the part of the critics of an effects-based approach that there

(N) Ehlermann 08 Bishop

292

6/11/08

11:50

Page 292

Session Two: A Reformed Approach to Exclusionary Conduct

This paper seeks to address two related issues. First, it attempts to identify the source of the fundamental difference of opinion between these two camps: those that broadly support the current form-based approach, which effectively prevents dominant firms from employing loyalty rebates (the “traditionalists”); and those that support a less dogmatic and less formalistic approach towards loyalty rebates, even when employed by dominant firms, on the basis that economic principles clearly demonstrate that such conduct often represents normal and efficient competitive behaviour. The key area of disagreement between the traditionalists and the progressives is the alleged loyalty-inducing effects of loyalty rebate schemes and its consequences for competition. Since any such scheme, whether entered into by a dominant firm or by a non-dominant firm, provides retailers with incentives to promote the sales of that firm they can, in some sense, always be said to be loyaltyinducing. For the traditionalists such loyalty-inducing effects provide a sufficient basis for finding harm to competition. But as this paper explains, pointing to the loyalty-inducing effects of a loyalty rebate scheme is simply stating the obvious; by itself, it says nothing about the actual exclusionary effects of such schemes even when practised by dominant firms. This leads to the second and critical issue. As noted above, if we are to move away from the current per se approach, how do we determine under what circumstances a loyalty rebate scheme will give rise to anticompetitive outcomes? In other words, what does a more economic effects-based approach actually entail in practice? Accepting that loyalty rebate schemes offered by even dominant firms can represent normal pro-competitive behaviour and accepting also that under some circumstances they can lead to a marginalization of competitors which results in adverse consequences for consumers, what factors need to be taken into consideration when seeking to determine when a loyalty rebate scheme is anticompetitive and when it merely represents normal competition? The difficulty arises from the need to distinguish between conduct that merely makes like hard for competitors and conduct that makes life hard for competitors and as a consequence leads to adverse outcomes for consumers. This paper argues that much more attention needs to be given to actual market evidence. In particular, consideration of observed industry evidence will require one to demonstrate that the incentives provided by loyalty rebate schemes (the so-called “suction effect”) actually gives rise to anticompetitive effects, rather than simply assuming that that is the case.15

is a significant difference between the frequency of intervention and the appropriateness of intervention. These critics assume that a reduction in the frequency of intervention necessarily represents a weakening of competition law. This stance fails to take into account the costs of over-intervention and ignores basic principles of competition economics. 15 As explained in Section 5 of this paper, the alleged anticompetitive effects of loyalty rebate schemes often generate testable hypotheses.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 293

Loyalty Rebates and “Merger Standards”

293

The remainder of this paper is organized as follows. Section 2 outlines the current approach to assessing the competitive effects of loyalty rebates under Article 82. This approach can be summarized very simply; if a firm is held to be dominant then it can’t offer any rebate scheme that involves a “loyalty inducing” effect. The current policy stance simply assumes that any loyalty rebate scheme operated by a dominant firm necessarily results in the foreclosure of competition to the detriment of consumers; this effectively results in a per se prohibition on dominant firms offering loyalty rebates. Section 2 also seeks to understand the primary competition concern that underpins the current hostile approach to loyalty rebates. It focuses on the alleged loyalty-inducing properties of such schemes and how that is held to translate into competitive harm. In brief, the traditional view assumes that loyalty rebates have a share-shifting effect that necessarily translates into adverse consequences for consumers. Section 3 explains why the current policy towards loyalty rebates is flawed, and in particular why the effective per se prohibition on dominant firms offering loyalty rebates is likely to give rise to significant false positives (i.e., preventing behaviour that is not anticompetitive) with consequent adverse effects for consumers. The current policy stance is flawed for the following reasons. First, it places far too much emphasis on dominance and, in effect, renders the definition of the relevant market the critical, indeed, defining, step in the competitive assessment. If a firm is held to be dominant, any loyalty rebate scheme (regardless of its actual impact on smaller rival firms) is held to result in anticompetitive foreclosure with adverse consequences for consumers. As such, the current interpretation of dominance under Article 82 precludes any effects-based analysis. The issue of dominance and the role that it plays in the assessment of Article 82 might therefore be characterized as the elephant in the room in the context of the debate on the reform of Article 82. Indeed, whilst those matters are in urgent need of attention, the current policy debate largely ignores it. Until they are explicitly addressed, the entire policy debate to date, interesting as it is, will continue to be largely academic. Second, the current approach completely ignores the various procompetitive rationales that a firm might have for offering loyalty rebates. Loyalty rebate schemes can provide a powerful and natural instrument of competition.16 Loyalty rebate schemes can eliminate double marginalization, provide incentives for customers to reduce the divergence in incentives that exist between suppliers of goods and services and those that distribute those products, and permit more efficient recovery of fixed costs. In effect, loyalty rebate schemes can play efficiency-promoting roles similar to those that have been associated with vertical restraints. 16 See, e.g., Kobayashi, cited supra note 9; Simon Bishop, “Delivering benefits to consumers or per se illegal?: Assessing the competitive effects of loyalty rebates”, in Konkurrensverket, The Pros and Cons of Price Discrimination, Konkurrensverket (Swedish Competition Authority), 2007, Chapter 3.

(N) Ehlermann 08 Bishop

294

6/11/08

11:50

Page 294

Session Two: A Reformed Approach to Exclusionary Conduct

Third, the fundamental premise underlying the hostile approach to loyalty rebates is itself flawed. Essentially, the current approach is overly concerned with a customer’s incentives at or around the “last unit” of purchase. Given the inherent discontinuity in any loyalty rebate scheme, it is inevitable that at some point a customer will have incentives to try to promote the products of one firm over another. But, as explained, it is inappropriate to focus on such “static” incentives when the customer nears the target threshold. Rather, the focus should be on “dynamic” incentives, i.e. on the customer’s effective options over time and the impact that the level at which the target threshold is set has for a customer’s incentives. Having set out what is wrong with the current policy toward loyalty rebate schemes operated by dominant firms, Section 5 attempts to redress the inadequacy of the current approach by outlining a suggested framework for how loyalty rebates should be assessed in practice. A central element of that competitive framework is the need to take into account observed market evidence. In nearly all investigations involving loyalty rebates, the matter concerns actual competition between a large firm and one or more smaller rivals as opposed to whether a loyalty rebate scheme deters new entry. This means that it is possible to assess directly whether rivals have indeed been placed at such a competitive disadvantage as to be marginalized because it is possible to observe what is actually happening in the market. We are able routinely to conduct such an empirical assessment when considering mergers; there is no reason why the same approach shouldn’t be applied to the competitive assessment of loyalty rebates.17 The lack of emphasis on actual market evidence represented a significant flaw in the Discussion Paper. As noted above, there is also a need to bring the presumption that loyalty rebate schemes necessarily lead to the marginalization of competitors, and hence a reduction in the effectiveness of competition, into line with the presumption that vertical mergers are generally pro-competitive. A genuine effects-based analysis would consider this directly rather than relying on stark ad hoc rules. Section 6 provides a summary and outlines three practical steps that would help improve policy by bringing the competitive assessment of loyalty rebate schemes into line with the methods and policy stances adopted in merger control. These are: stop focusing on the last unit (i.e., adopt a more dynamic view of the competitive impact of loyalty rebates) and recognize that a rival’s loss of market share does not inevitably lead to foreclosure; start to recognize the elephant in the room (i.e., reconsider the role that dominance plays in the competitive assessment)—dominance represents a useful screen in the competitive assessment but no more than that; and start being “positive” (i.e., start publishing positive decisions rather than just negative ones); the publication of both positive and negative decisions in the area of merger control 17

82.

Similar considerations apply to the assessment of other exclusionary abuses under Article

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 295

Loyalty Rebates and “Merger Standards”

295

has contributed significantly to the development and improvement of EC competition policy in this area.

2. The current policy towards loyalty rebates The Article 82 case law can fairly be characterized as being extremely hostile towards any dominant firm offering a loyalty rebate scheme. The current approach adopts a two-step process: the assessment is first concerned with whether the firm under investigation enjoys a dominant position; and second, if dominance is deemed to have been established, it is considered whether the conduct of the dominant firm constitutes an abuse of that dominant position. In theory, there is much to commend this two-step approach, as the requirement to establish dominance in principle provides a useful screen: if the firm is not dominant then there can be no anticompetitive effects and therefore no further analysis is required. The second step is purely formalistic, as the assessment focuses on the form of the conduct: the rebate scheme can only be justified according to the case law if the discounts granted are cost-based. For example, in Michelin I,18 the Commission found that Michelin had abused a dominant position in replacement tyres for trucks and buses in the Netherlands through the provision of off-invoice discounts and end-of-year rebates based on performance targets. The Commission stated that: “[W]ith the exception of short term measures, no discount should be granted unless linked to a genuine cost reduction in the manufacturer’s costs. The compensation paid to Michelin dealers must be commensurate with the tasks they perform and the services they actually provide, which reduce the manufacturer’s burden. In addition the system of discounts and bonuses agreed must be clearly confirmed to each dealer when the sales contract is presented and concluded.”

The end result is that the current European antitrust policy of loyalty rebate schemes under Article 82 can be stated very simply; if you are dominant, then any loyalty rebate scheme will be deemed to represent an abuse of that dominant position. This hostility translates into an effective per se prohibition against the offer of loyalty rebates by dominant firms.19 The remainder of this section attempts to understand the basis for this hostile stance. As noted above, the term “loyalty rebate” is capable of encompassing a wide class of discount schemes. Although not precisely defined, a loyalty rebate 18

Case 322/81, NV Nederlandsche Banden-Industrie Michelin v Commission [1983] ECR 3461. There are some who argue that Article 82 does not embody a per se approach. Such arguments, in my view, are simply not tenable when one considers the line of reasoning adopted in the decisions and judgments of the Commission and the European Courts, respectively. 19

(N) Ehlermann 08 Bishop

296

6/11/08

11:50

Page 296

Session Two: A Reformed Approach to Exclusionary Conduct

scheme represents a form of non-linear pricing in which the unit price of a good declines when a buyer’s purchases meet a buyer-specific minimum threshold requirement. It is this characteristic that differentiates a loyalty rebate from other discount schemes or forms of price discrimination.20 Loyalty rebate schemes can therefore take many forms, including the following categories:21 • An exclusivity discount: The buyer obtains a discount only by purchasing all its needs from the supplier.22 • An individualized quantity discount: Each buyer is offered a discount conditional on purchasing a given quantity within a particular reference period. In this case the targets may differ for buyers of different sizes. • A growth discount:23 The buyer is given a discount if its purchases in the current period exceed its purchases in the relevant past period by a given amount. Whether meeting such growth targets results in the buyer increasing or retaining the market share of the supplier depends on whether the targets are set at a higher level than the growth in the overall market. • A bundled discount: The target relates to purchases across a range of products. In this case, it might be necessary for the buyer to purchase a certain amount of another product supplied by the firm in question in order to qualify for the rebate where the target amounts to a large portion of needs of that product. The term “loyalty rebate” therefore covers a wide range of discount schemes. What each of these different schemes has in common is that it provides incentives for customers to purchase more products or services from the firm offering the loyalty scheme. All loyalty rebate schemes thus have the “effect of inducing loyalty”. Put simply, all loyalty rebate schemes create incentives for customers to purchase more from the firm employing the loyalty rebate scheme, and they are therefore likely to result in share-shifting. The loyalty-inducing effects of a loyalty rebate scheme can be seen in the following hypothetical representation. Firm X has a variable unit cost of production of €1 and annual fixed costs of €1m (which for the purposes of this discussion we can assume to be sunk).24 This firm is assumed to set a list price of €2 per unit sold but also offers a 5% discount if the customer reaches a certain target threshold, so that its effective price per unit purchased then falls 20 As already explained, a loyalty rebate need not necessarily be conditional on reducing purchases from rival suppliers. Whether this is the case depends on the precise form of the rebate scheme employed. (Some typical examples are listed below.) 21 This paper focuses on single product rebate schemes but, as noted above, the principles, with appropriate modification, also apply to the assessment of multi-product rebate schemes. See Kobayashi, cited supra note 9. 22 Effectively, this is a market share discount where the threshold is set at 100 per cent. 23 This is the type of rebate scheme that was at the centre of the BA/Virgin decision, cited supra note 11. 24 It is worth noting that the analysis presented here does not assume dominance. This by itself should make it clear that the competitive assessment needs to go beyond a consideration of the incentive properties of a loyalty rebate scheme even when operated by a dominant firm.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 297

Loyalty Rebates and “Merger Standards”

297

to €1.90 per unit. (This example assumes that the threshold is based on a share of total purchases but the analysis would be similar if the target threshold were based on year-on-year growth or on achieving an absolute level of sales). As the customer increases its volume of sales from the firm offering the loyalty rebate, the incentives to purchase even more from that firm also increase. Indeed, at some point, the effective price for purchasing additional volumes becomes negative. This is the so-called “suction effect”. This can be illustrated graphically in Figure 1.

Figure 1: Properties of loyalty rebate schemes

Figure 1 shows three diagonal lines sloping up from the origin (point A).25 • The steepest line (running from A through E) shows the customer’s expenditure when buying at the list price of €2 per unit. Under the loyalty rebate scheme, this is the price the customer pays until it can show that it has reached the target threshold set by the dominant supplier. At that point, the customer qualifies for a rebate equivalent to the discount of 5% on all purchases, which is shown as the vertical distance from the top of the list price line to point C. • The middle line (from A to C) shows the amount that the customer actually pays if it agrees and adheres to the terms of the loyalty rebate, i.e. the effective unit price falls to €1.90. • The bottom line indicates the total variable costs incurred by the supplier in supplying the required volume of goods or services. The slope of the line indicates the marginal cost. 25

Note: the lines in the above figure are not drawn to scale.

(N) Ehlermann 08 Bishop

298

6/11/08

11:50

Page 298

Session Two: A Reformed Approach to Exclusionary Conduct

Figure 1 shows that once the customer has purchased volumes equivalent to or greater than B* then the effective unit price for purchasing the rest of its requirements from Firm X becomes negative; the value of the prospective year-end rebate is greater than the price the customer needs to pay for the extra units. This implies that, viewed from the perspective of a customer that has already purchased B* from Firm X, the incentives to make additional purchases from Firm X are so great that rival firms would find it extremely hard to contest the customer’s demand above B*; in effect, rival firms would need to offer negative prices to contest such sales. This is the so-called “suction effect” or “loyalty-inducing effect” of such rebate schemes.26 In summary, loyalty rebates can provide powerful incentives for a firm’s downstream customers to purchase additional volumes from that firm. In principle, such incentives can lead to customers shifting purchases from smaller rivals to such an extent that these rival firms are marginalized. But as the hypothetical example demonstrates, all loyalty rebates have these properties, regardless of whether the firm offering the rebate scheme is held to be dominant. This means that it is an insufficient basis for concluding that a loyalty rebate scheme gives rise to anticompetitive outcomes by highlighting the “suction effect” of such schemes. Indeed, it is important to be clear about what is meant by the marginalization of competitors. The loss of volumes will only marginalize a competitor if it adversely affects the pricing decisions of that competitor, for example by leading to an increase in short-run marginal costs. In other words, there can be no presumption that the share-shifting to which a loyalty rebate scheme may give rise actually results in the marginalization of competitors. Indeed, in its draft Non-Horizontal Merger Guidelines, the Commission states (in accordance with established economic literature) that non-horizontal mergers predominantly give rise to benign or pro-competitive effects. This raises an important policy issue; namely, the discrepancy between the assessment of foreclosure under the Merger Regulation as set out in the draft Guidelines and its assessment under Article 82. Whereas the presumption set out in the draft Guidelines recognizes that harm to competitors does not necessarily, or even often, translate into harm to competition, the competitive assessment under Article 82 often takes quite a different stance: anything which harms competitors is necessarily assumed to harm competition. For example, any loyalty rebate scheme employed by a firm held to be dominant is assumed to harm competition.27 But loyalty rebate schemes can provide the kind of efficiency-promoting roles that have been identified with 26 For a formal depiction of the suction effect, see Frank Maier-Rigaud, “Article 82 Rebates: Four Common Fallacies”, 2 European Competition Law Journal 85 (2006). 27 See, e.g., the Opinion of Advocate General Kokott in British Airways, cited supra note 11, para. 68.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 299

Loyalty Rebates and “Merger Standards”

299

vertical restraints and vertical mergers; loyalty rebate schemes can be employed to eliminate double marginalization, provide incentives for customers and reduce the divergence between the incentives of suppliers and those of the distributors that sell their products. Having markedly different approaches to assessing the issue of foreclosure is unjustified from the perspective of the underlying economics. It will be interesting to see whether the declared intention to move towards a more effects-based approach in applying Article 82 will lead to the elimination of this discrepancy.

3. The current assessment of loyalty rebates is flawed As noted in Section 2, the current policy towards loyalty rebates is extremely hostile; if a firm is held to be dominant then it is effectively per se illegal for it to offer loyalty rebates to its customers. This section explains why the current policy towards loyalty rebates is flawed, and in particular it explains why it is likely to give rise to significant false positives (i.e., why it is likely to deter behaviour that is not anticompetitive). In summary, the current policy stance is flawed for the following reasons. • It places far too much emphasis on dominance. • The current approach completely ignores the pro-competitive rationales for which firms offer loyalty rebates. • The fundamental premise underlying the hostile approach to loyalty rebates is itself flawed. The remainder of this section expands on each of these points.

3.1 Too much focus on dominance As noted in Section 2, the establishment of dominance not only lies at the heart of the current competitive assessment of loyalty rebates under Article 82 but actually represents the only important consideration in the whole investigation. There are two fundamental problems with the current approach. First, the assertion that harm to competitors necessarily translates into harm to competition is at odds with the economic literature and also at odds with the case law in other, but related, areas of policy, particularly in the assessment of non-horizontal mergers. Moreover, the actual competitive effects of a particular loyalty rebate scheme, to the extent that they are considered at all, are assumed rather than assessed. For example, in British Airways, the Advocate General stated that any adverse impact on competitors of a firm held to be dominant can be

(N) Ehlermann 08 Bishop

300

6/11/08

11:50

Page 300

Session Two: A Reformed Approach to Exclusionary Conduct

presumed to harm “competition as such (as an institution)”.28 This approach has been and continues to be supported by the Courts.29 Second, establishing dominance is itself highly problematic. Dominance in most cases is primarily determined by a firm’s market share, and consequently the definition of the relevant market plays a critical role in the competitive assessment under Article 82. Although there is now broad consensus (in theory if not in practice) that the hypothetical monopolist test provides the appropriate framework for defining relevant markets,30 this does not imply that defining relevant markets is always a straightforward task. Indeed, experience shows that this is rarely the case and that a detailed case-by-case assessment is usually required. The difficulties of defining the relevant market in Article 82 investigations are compounded by the existence of what is commonly termed the “cellophane fallacy”. The cellophane fallacy has two implications; the first is well-known and the second is less well-known and frequently ignored. • An unthinking application of the hypothetical monopolist test (a test originally formulated in the context of merger control) will tend to result in relevant markets being defined too broadly. In consequence, the use of market shares in such cases will tend to understate the market strength of the firm under investigation. • More importantly, in many cases the available evidence will simply be unable to discriminate between two plausible and therefore competing definitions of the relevant market; the available evidence can be consistent with both a narrow definition of the relevant market and with a wide definition of it. The second implication effectively means that in many cases we simply can’t know with any degree of certainty whether a firm is or is not dominant based solely on an assessment of market shares.31 Neither the parties nor the competition authorities definitely prove their case.32 28 See, the Opinion of Advocate General Kokott in British Airways, cited supra note 11, paras. 68 and 86. 29 See, e.g., Case 85/76, Hoffmann-La Roche & Co. AG v Commission [1979] ECR 461; Case T-203/01, Manufacture française des pneumatiques Michelin v Commission (Michelin II) [2003] ECR II-4071. 30 Unfortunately, a surprising number of people continue to misunderstand the SSNIP test and erroneously consider it to represent merely one way of defining relevant markets. The incoherent discussion in the CAT judgment in Aberdeen Journals provides a particularly apt example of this misunderstanding of basic principles. Those who hold this view are however unable to provide an alternative definition that accords with anything but pure ad hoc reasoning. Judgment of 23 June 2003 of the Competition Appeals Tribunal, Aberdeen Journals v Director General of Fair Trading, Case No. 1009/1/1/02, [2003] CAT 11. 31 For a discussion of the additional difficulties of defining relevant markets in Article 82 investigations, see Simon Bishop and Simon Baker, “The Role of Market Definition in Dominance and Monopoly Enquiries”, Office of Fair Trading Economic Discussion Paper 2 (November 2001). 32 Due to the existence of the cellophane fallacy, market evidence may in some cases be construed in such a way that it suggests a market definition that is too wide; one should therefore

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 301

Loyalty Rebates and “Merger Standards”

301

In light of the difficulties that market definition raises in the context of Article 82 inquiries, the definition of the relevant market and any conclusions on dominance should play primarily a screening role to eliminate from consideration the activities of those firms that are not dominant under any plausible market definition. Article 82 inquiries should focus predominantly on whether the conduct under investigation constitutes an abuse. In principle, the assessment of dominance could be expanded to give more regard than has hitherto been the case to other factors.33 However, as experience in merger control illustrates, taking these other factors into account is best done as part of an in-depth investigation. In short, we shouldn’t put all our eggs into the market definition/dominance basket. Moreover, placing market definition at the centre of any investigation risks market definitions being reverse-engineered: “I don’t like what you are doing, so you must be dominant.” For these reasons, there is a compelling case to retain the establishment of dominance as a safe harbour: only if the firm might be held to be dominant is there a need for further investigation of the effects of its business conduct. In other words, assessing whether conduct represents an abuse would be based on the competitive effects of the particular business conduct under investigation and not on whether the firm is dominant. In short, the issue of dominance and the role that it plays in the assessment of Article 82 might be characterized as the elephant in the room.34 Dominance and what role it should play in the competitive assessment under Article 82 is an issue which urgently needs to be addressed. Unfortunately, despite the urgent need to re-think these matters, the current policy debate largely ignores them. And ironically so, since the implications of a finding of a dominant position under the current approach represent a significant and perhaps insurmountable obstacle to the implementation of a genuine effects-based approach in the enforcement of Article 82. A genuine effects-based approach recognizes that the same business conduct, even when practised by dominant firms, can in some circumstances be anticompetitive and in others pro-competitive.35 For example, although in certain circumstances the business conduct of offering additional payments for reaching certain targets (i.e., loyalty rebates) may foreclose competition, with adverse consequences for consumers, in many if exercise caution in interpreting such evidence. Unfortunately, as the Aberdeen Journals case illustrates so well (see supra note 30), the cellophane fallacy is often invoked to unreasonably dismiss all evidence. 33 For example, the industry might be characterized by bidding competition, in which case even very high market shares might not confer significant market power. 34 For a more detailed discussion of this issue, see Simon Bishop, “The Elephant in the Room: Re-Thinking the Role of Dominance in Article 82 Investigations”, European Competition Journal, forthcoming. 35 This is also recognized in the law—even dominant firms are able to engage in what is termed “normal competition”. Unfortunately there has been no attempt by either the Commission or the European Courts to provide an operational definition of that term.

(N) Ehlermann 08 Bishop

302

6/11/08

11:50

Page 302

Session Two: A Reformed Approach to Exclusionary Conduct

not most other circumstances the same business conduct yields significant benefits to consumers. The fact that these business practices are routinely adopted by non-dominant firms demonstrates clearly that they often represent normal competitive practices. This is not to say that such business conduct when practised by a dominant firm cannot lead to anticompetitive outcomes. It is only to state that the fact that non-dominant firms engage in such conduct demonstrates that such conduct can be pursued for benign reasons. This implies that we need to assess the business conduct of a firm in its particular context, even when that firm has been found to be dominant.

3.2 Pro-competitive benefits of loyalty rebates The underlying pro-competitive business motivation for employing loyalty rebates is to sell more products or services at prices which increase profits. Seeking additional sales so as to increase profitability is a standard motivation for all firms. As noted above, the proposition that there are pro-competitive motivations for employing loyalty rebate schemes is clearly demonstrated by the fact that such pricing practices are also employed by firms that are not dominant. As is well accepted in antitrust economics, any business practice employed by a firm with no significant market power cannot give rise to anticompetitive outcomes.36 The various pro-competitive rationales can be categorized as follows: • providing incentives for customers to supply complementary services; • inducing customers to lower prices to end consumers (i.e. reducing double marginalization); and • efficient fixed cost recovery (Ramsey pricing37). The first two reasons are akin in many respects to vertical restraints. The third rationale is to enable the firm to increase output. One way a firm might seek to achieve higher sales volumes is to lower the price of its product. However, an across-the-board price reduction might not be attractive due to its effect on the profitability of the firm. This is particularly true where production involves significant fixed costs and low variable costs. Loyalty rebate schemes provide one method that permits firms to make greater sales to customers without reducing the prices on all units sold.

36

There are strong logical grounds for equating significant market power with dominance. Frank Ramsey, “A Contribution to the Theory of Taxation”, 37 Economic Journal 47 (1927). 37

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 303

Loyalty Rebates and “Merger Standards”

303

3.2.1 Providing incentives for customers to supply complementary services A key pro-competitive rationale for a firm to employ loyalty rebates is to align the incentives of customers with that of the supplying firm.38 Customers can add substantial value to a supplier’s products or services by providing additional complementary services. These complementary services include promoting the product in store, providing detailed product information to customers (sometimes including a demonstration of how to use the products or services), keeping an appropriate stock of product so that at any time final consumers are able to purchase the product they require, or simply putting more effort into selling the products or services of the supplier. These services create benefits for the supplier while their costs mainly accrue to the customer. A customer chooses the level of these complementary services by considering its own marginal benefit received from and its own marginal cost from providing such services. Since the customer has no inherent incentive to take into account the supplier’s interest in deriving such benefits, in general the customer will provide a lower level of services than the supplier would wish. A direct solution to this problem would involve the supplier writing a contract with the customer specifying the level of services to be provided and the reimbursement from the supplier to the customer. However, there are several concerns with this method. First, specifying service levels in a contract is difficult because such services cannot be measured objectively. Second, it is costly for the supplier to engage in direct monitoring of the provision of these services. Third, even if monitoring is achievable, enforcement in case of contract breach has additional costs. Finally, and importantly, the customer has an informational advantage over the supplier for determining the appropriate combination of these services to expand demand, as local demand conditions for the product might be highly variable. Any method that aligns customers’ incentives with those of the supplier would allow the supplier to achieve efficient levels of service provision at different locations in a decentralized manner. A loyalty rebate scheme provides an efficient method to achieve this goal. Greater effort by the customer may be necessary to increase the supplier’s share of its customer’s total purchases of that product. Provided the customer achieves the target threshold set by the supplier, the supplier shares with the customer the benefits of the sale expansion achieved. In short, the discounts made available in the rebate scheme seek to align the incentives of the customer with those of the supplier. 38 The need to align the interests of suppliers and customers is sometimes referred to as the principal-agent problem. This issue was first addressed by the Nobel Laureate, James Mirrlees. Among his other relevant works, see Mirrlees, “An Exploration in the Theory of Optimum Income Taxation”, 38 Review of Economic Studies 175 (1971).

(N) Ehlermann 08 Bishop

304

6/11/08

11:50

Page 304

Session Two: A Reformed Approach to Exclusionary Conduct

3.2.2 Inducing customers to lower prices to end consumers (i.e., reducing double marginalization) In many settings, the customers of suppliers employing loyalty rebates (in particular retailers) add a margin to the wholesale price of products or services when they set their own prices. This margin includes the profit margin of the supplier’s customer as well as the customer’s costs. Since the supplier’s wholesale price already includes a profit margin, the price to the end consumer ends up suffering from double marginalization. As a result, the price to the end consumer is too high compared to what the supplier would choose if it had its own integrated network, to the detriment of both the supplier and end consumers. Double marginalization could be eliminated if the supplier charged a wholesale price equal to the marginal cost of producing its products or services. In that case, each customer of the supplier would make the same retail pricing decision that the supplier would make if the customer were vertically integrated with the supplier. But of course, in that case, the supplier would be left with either lower profits or (most likely) significant losses. A loyalty rebate scheme allows the supplier to disentangle the average wholesale price from the marginal wholesale price to a specific customer. Any customer of the supplier knows that reaching the target will imply a lower price. The customers will therefore have incentives to moderate their mark-up over the wholesale price, and in consequence output will be expanded. The customer thereby increases the likelihood of reaching the target threshold, upon which the supplier shares part of the benefits it realizes with the customer.

3.2.3 Efficient fixed cost recovery (price discrimination) As first noted by Ramsey, charging different prices to different customers (i.e., price discrimination) can represent an efficient way to recover fixed costs.39 Loyalty rebate schemes provide one method of price discrimination that permits the efficient recovery of fixed costs. Many industries are characterized by high fixed costs and (relatively) low marginal costs of providing goods and/or services. In consequence, firms in fixed cost recovery industries charge prices in excess, often well in excess, of short-run marginal costs. The likelihood that such pricing will entail some loss of static efficiency (i.e., certain consumers will be dissuaded from consuming the product even though they value it higher than the marginal cost of supply) must be traded off against the risk that the product would not exist at all if investors were not offered the prospect of fixed cost recovery at the time when the necessary 39

See Ramsey, cited supra note 37.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 305

Loyalty Rebates and “Merger Standards”

305

investments were made. Where businesses face this problem of fixed cost recovery, and where they also face a number of markets or market segments in which they sell their product, it becomes likely that they will charge different price-cost margins on different transactions or in different segments of the markets they serve, i.e. firms will engage in price discrimination. For this reason, industries characterized by price differences that cannot be fully explained by differences in the costs of supply are an extremely pervasive phenomenon in real world markets.

3.3 It’s about more than the “last unit”! As discussed in Section 2, the current hostile policy stance towards loyalty rebate schemes stems primarily from focusing solely on the incentive properties of a rebate scheme at or around the threshold of eligibility for the rebate. This section explains why that focus on the “static” incentives of the scheme is inappropriate. Focusing on the last unit means that the competitive assessment fails to address other important issues, and in so doing fails to reflect the actual competitive dynamics observed in many industries. This section addresses two critical issues which lie at the core of an effects-based competitive assessment but which are absent from the Commission’s current analysis of the competitive effects of loyalty rebates. • First, I explain why loyalty rebate schemes do not necessarily have exclusionary effects even when offered by dominant firms. • Second, I provide a more complete analysis of the incentives at or around the threshold. Whereas the traditionalists focus on the situation where a customer is close to reaching a target, in practice the situation can be more complex. Finally, the circumstances under which loyalty rebates might indeed permit the exclusion of equally efficient rival firms are considered.

3.3.1 Loyalty rebate schemes are not necessarily exclusionary Section 2 presented an analysis of a loyalty rebate scheme offered by Firm X. Figure 1 is reproduced below for ease of reference. Assuming that a customer has already purchased volumes greater than B, the ability of rival firms to compete for volumes between that amount and the customer’s total annual requirements will be extremely limited; the effective price for making additional purchases above B (but below C) is negative. But rather than focusing on the incentives of the rebate scheme at a particular point in time (e.g., once the customer has purchased volumes greater

(N) Ehlermann 08 Bishop

306

6/11/08

11:50

Page 306

Session Two: A Reformed Approach to Exclusionary Conduct

Figure 1: Properties of loyalty rebate schemes

than B), it is interesting and indeed more appropriate to analyze the way in which this loyalty rebate scheme affects the customer’s incentives throughout the period to which the loyalty rebate applies. As we shall see, a customer’s incentives depend critically on the point at which the customer is making its purchase decisions throughout the qualifying period. Once the customer has reached point B, the marginal price of buying the rest of its requirements from the dominant firm is zero, so rival suppliers would need to work very hard to contest this slice of business. Beyond point B (at point E, for example), the effective marginal price to the customer of buying the rest of its requirements from the dominant firm is actually negative, since the value of the prospective year-end rebate is greater than the price the customer needs to pay for the extra units. But if one considers the choices available to consumers at some lower level of purchases, such as D on the dominant firm’s list price curve, the incentives are less powerful. Although it is true that at this point it becomes expensive for the customer to contemplate shifting demand to a rival because to do so would result in the forfeit of the opportunity to earn the substantial loyalty rebate at the end of the year, the price for additional purchases is still positive. Figure 1 illustrates this incentive effect by reference to the slope of the line drawn between (in this case) point D and the discounted end-period point C. Having bought the majority of its requirements from the dominant firm at list price, the full value of the prospective year-end loyalty discount should be spread across the remaining purchases, giving a lower marginal price. Point D in Figure 1 has been constructed so that this marginal price (the slope of the line from D to C) is the same as the dominant firm’s variable costs.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 307

Loyalty Rebates and “Merger Standards”

307

Going back further in time to the point where the customer has made no purchases from the supplier, the incentive effects of the loyalty rebate scheme might be even weaker, and indeed the rebate may provide no incentives to divert sales at all. At point A (representing the time prior to agreement of a supply contact or the beginning of the year to which the threshold applies), the price that is offered by the loyalty rebate scheme is essentially the discounted price of €1.90 if the customer stays “loyal” to the dominant firm and meets the target threshold. If the customer can easily meet this target by buying enough of its requirements either from the dominant firm or from a rival supplier, this is a very simple choice. The rival will win the business if it can offer a price for the year that beats the dominant firm’s offer of €1.90, and the fact that the rebate is expressed as a loyalty deal does not imply that it is necessarily exclusionary. This discussion makes it clear that the likely competitive effects arising from the incentives of a loyalty rebate scheme (even a retrospective rebate scheme) are dependent on the customer’s available effective options. First, if customers are in a position to evaluate the offer from the origin (point A), and if they have the clear option to trade off the dominant firm’s loyalty rebate offer against similar offers made by rival suppliers, then the shape of the discount profile tells us nothing about its effect on competition. As long as the customer can credibly threaten to buy enough of its requirements from the dominant firm’s rivals, it will also be able to assess the value-for-money offered by the dominant firm’s loyalty rebate offer against the alternative of switching demand to the dominant firm’s rivals. In terms of Figure 1, the customer will simply compare the dominant firm’s discount offer (shown by point C in Figure 1) with the alternative offer made by rival suppliers. Having done so, the particularly powerful incentives that come into play at or close to the “last unit” may not arise; customers respond to the incentives provided by the scheme as a whole rather than those that arise as their level of purchases approach the threshold.

3.3.2 Incentive properties of loyalty rebates are complex As noted above, both the Commission and the European Courts presume that loyalty rebate schemes inevitably involve customers switching part of their requirements to the dominant firm offering the loyalty rebate scheme.40 But underlying this strong presumption are three assumptions as to the behaviour of customers. • First, customers are able to disregard the interests of their own clients (i.e., the final consumers). 40 Neither the Commission nor the Courts have explained why the same “loyalty inducing” effects of such rebate schemes do not apply equally to non-dominant firms.

(N) Ehlermann 08 Bishop

308

6/11/08

11:50

Page 308

Session Two: A Reformed Approach to Exclusionary Conduct

• Second, the use of target thresholds gives each customer an overwhelming incentive to favour the sales of products of the dominant firm over sales of products supplied by other firms. • Third, what is true for an individual customer applies to the market as a whole. With respect to the first assumption, it should be noted that in most industries, intermediate customers possess some ability to influence the purchase decisions of their own customers. This might take the form of pre-sales advice, in-store placement or retail discounts (i.e., price inducements). It is for precisely this reason that suppliers seek to provide appropriate incentives to their customers to promote their products rather than those of a competitor. This simply represents a standard form of competition between suppliers. For example, in the BA/Virgin case,41 the Commission effectively assumed that the only parameter of competition between airlines was on the incentive schemes provided to travel agents; the price of air travel, the quality of service and other forms of competition (e.g., advertising) were simply not considered to be relevant. Moreover, in many industries, a supplier’s customers will themselves operate in a competitive market. A supplier’s customers compete with one another to deliver the best possible service to their own clients. Competition may take place with respect to quality of product or service, the ability to meet consumer requirements and of course the price of those products or services. An intermediate customer that is unable to fulfil final consumers’ requirements will be unlikely to retain that business over time. With respect to the second assumption (that of an “overwhelming incentive”), even if it were supposed that a supplier’s customers were able to exercise considerable control over the purchases of their own clients, it is not clear that the structure of loyalty rebate schemes necessarily results in the dominant firm being favoured over competitors. If a customer can choose among suppliers, with no adverse effects on its business, its preference will accord with its expectations as to which products will contribute to the highest revenue possible over the year. But this does not imply that the customer will always necessarily choose to buy from the dominant firm. This can be seen in the following simple example. In this example, it is assumed that, in this case, customers can, with no adverse effects on their business, sell products or services that yield the highest revenue to the retailer. Consider two suppliers, a dominant supplier A and a competing but smaller supplier B. Both suppliers operate loyalty rebate schemes which are based on customers reaching a target threshold. Once that target threshold has been reached, a retrospective discount is granted on all purchases. Table 1 shows the target thresholds for the two suppliers, above which additional commissions become payable. In this simplified example, the 41

Cited supra note 11.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 309

Loyalty Rebates and “Merger Standards”

309

loyalty rebate scheme of each supplier contains only one target threshold for each retailer.42 The additional revenues (discounts) received are assumed to be the same for both suppliers at €20 per unit if the threshold is met. The threshold for Supplier A is set at 220 which, reflecting its larger size of operations, is four times the threshold set by Supplier B. Table 1 Incentives for customers under different sales scenarios

Target Threshold Revenue per unit (€) Current Sales Scenario #1 Scenario #2 Scenario #3

Supplier A

Supplier B

220 20

55 20

210 230 230

50 50 60

Table 1 shows the current level of sales of a particular customer in three different scenarios. In scenario # 1, the customer has not yet reached the target threshold for either supplier. In scenario # 2, the customer has already exceeded its target for Supplier A but not for Supplier B, and in scenario # 3, the customer has exceeded the target thresholds for both suppliers. Now suppose the customer has the ability to choose between the two suppliers for the purpose of buying 10 additional units. The customer’s choice of supplier will clearly depend on which scenario we are considering. Assuming that the list price is the same for the products of both suppliers, the customer will seek to direct sales towards the supplier that generates the greatest increase in discount arising from the structure of the loyalty rebate scheme. Table 2 shows the amount received by the customer agent from the performance element at the current level of sales shown in Table 1 and the increment if the customer sells an additional 10 units of Supplier A’s products or, rather, 10 additional units of Supplier B’s products. Table 2 Choices of customers under different sales scenarios Supplier A Current Increment Scenario #1 Scenario #2 Scenario #3

4,600 4,600

4,400 200 200

Supplier B Current Increment 1,200

1,200 1,200 200

42 Note that the consideration of multiple thresholds does not significantly alter the conclusions to be drawn from this example.

(N) Ehlermann 08 Bishop

310

6/11/08

11:50

Page 310

Session Two: A Reformed Approach to Exclusionary Conduct

In scenario # 1 in table 2, selling an additional 10 units of Supplier A’s products generates additional revenue of €4,400 (i.e., 220 times €20) compared to only €1,200 additional revenue (i.e., 60 times €20) if he sells 10 more units of Supplier B’s product. Clearly, the customer, if he has a choice, will choose to sell 10 more units of Supplier A’s products. But in scenario # 2, this is no longer the case. The customer has already reached the threshold target for Supplier A and therefore the additional revenue generated by the performance reward scheme will be €200 (i.e., 10 times €20). This is less than the incremental revenue earned from selling an additional 10 units of product supplied by Supplier B. In this case, the additional sales take the customer over the target threshold of Supplier B, yielding additional revenues of €1,200. The customer, if he has a choice, will therefore in this scenario choose to sell additional units supplied by Supplier B. In scenario # 3, where it has met both the threshold targets, the customer will be indifferent as to which supplier’s products he sells. This simple example highlights another important element of the analysis which is typically totally disregarded in the competitive assessment of loyalty rebates, namely, the level at which the target is set. If suppliers set too high a target then its customers’ purchasing decisions will simply not be affected; why should customers bother to direct sales to the dominant entity if they have no prospect of reaching the target? Conversely, if the supplier sets the target threshold too low then the customer may be able to reach the target without engaging in any directional selling or other efforts on behalf of the dominant supplier. In this case too the target threshold has no affect on customer incentives. It is by no means obvious that the customer will always seek to favour the larger supplier even when the strong assumption is made that the customer can choose, without detriment, which products to sell. Hence, with respect to the second assumption, despite the Commission’s presumption that loyalty rebate schemes necessarily bind customers to the dominant firm, a much more careful analysis than is suggested by the approach adopted by both the Commission and the Courts is required. In assessing the competitive effects of loyalty rebate schemes, the assessment must go beyond merely stating that the scheme gives customers incentives to purchase greater volumes from the dominant firm; that should be selfevident. As the above simple examples demonstrate, this is true even for retrospective rebate schemes that set certain targets for customers and grant those customers that meet such targets a discount that applies to all units and not just the additional ones purchased above the target. With respect to the third assumption, it is also incorrect to assume that what applies to an individual customer necessarily applies to the market as a whole. The concept of foreclosure relates to the market as a whole; what matters is whether there is sufficient room for a rival to operate efficiently. This can occur if either: the loyalty rebate covers only part of each customer’s requirements; or rival firms can obtain sufficient volumes from those cus-

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 311

Loyalty Rebates and “Merger Standards”

311

tomers that choose not to respond to the incentives provided by the loyalty rebate of the firm held to be dominant. Indeed, in many markets in which firms offer loyalty rebates, some customers will choose to purchase primarily from the market leader and others will choose to purchase primarily from rival firms. Moreover, the choices that individual customers make as to which products they purchase in any given year may vary.43

3.3.3 When might loyalty rebates become exclusionary? The analysis corresponding to Figure 1 stated that, at the beginning of the rebate scheme, equally efficient rivals can compete effectively against the dominant firm. But a different analysis might apply if the appeal of the dominant firm’s brand is such that the customer has a strong pre-disposition to buy a substantial proportion of its requirements from the dominant firm, and therefore the best outcome that a rival supplier can realistically expect to achieve is to secure, say, one third of the customer’s requirements. Such sales might be called the dominant firm’s assured base. In the presence of an assured base, a loyalty rebate scheme can have a significant effect on a customer’s choices. Starting from a point part-way along the customer expenditure profile, the marginal price for additional purchases from the firm can be much lower. If that marginal price falls below the supplier’s marginal cost of supply, the fear of exclusionary effects becomes greater because the rebate scheme involves pricing those contestable units below the supplier’s avoidable costs. This does not in itself establish that the rebate scheme is abusive, since a finding of some prices below variable costs does not in itself establish an exclusionary economic effect. It does, however, begin to build a picture of exclusionary effects. To see the impact of such an assured base of sales for the competitive effects of a loyalty rebate scheme offered by a firm held to be dominant, consider Figure 2. Figure 2 shows a loyalty rebate scheme in which the firm charges €20 per unit for all purchases up to 50 units and then offers a retrospective rebate of €10 per unit for all sales above 50. Marginal costs are assumed to be constant at €5 per unit. Overall, the loyalty rebate scheme involves above-cost pricing—at any volume of sales, total revenue exceeds total cost of production. For example, at 100 units, revenues equal €1,000 and costs equal €500. However, it is clear that if the firm has an assured base of sales of 50 units (for whatever reason) then the overall loyalty rebate scheme can be seen to be exclusionary in the sense that the additional revenues earned on the extra 50 units of sales are below the incremental costs of making those sales 43 This was a feature of the UK travel agency industry in terms of which airline each travel agent chose to promote in any given year.

(N) Ehlermann 08 Bishop

312

6/11/08

11:50

Page 312

Session Two: A Reformed Approach to Exclusionary Conduct

Figure 2: The impact of an assured base of sales on the competitive effects of rebate schemes

(incremental revenues equal zero, whilst incremental costs are €250). In effect, the loyalty rebate provides a mechanism for the firm to target the effective discount on those volumes that are “open” to competition. Although an assured base of sales may be very difficult to identify in practice and may differ across customer types, this stylized framework nevertheless provides a useful starting point. First, it demonstrates that a loyalty rebate scheme can, in principle, be used to target a lower price on a range of sales open to competition in the same way that a dominant firm can selectively reduce prices in one market where it faces entry but leave prices high in another market where it does not face the prospect of competition.44 Second, it provides a framework for assessing the likelihood of foreclosure by “allocating” the discount inherent in the loyalty rebate scheme to the range open to competition. This approach allows safe harbours to be devised on the assumption that: (a) an assured base of sales exists; (b) sales could be made at the list price for that assured base; and (c) the appropriate measure of cost does not vary significantly according to the identification of the assured base. These assumptions permit the whole of a discount to be attributed to the range of sales assumed to be open to competition. The larger the assumed assured base, the easier it will be to engage in exclusion. If we are confident that the discount 44 There is also a link to the tying and bundling literature. The discount on the range open to competition can be thought of as being conditional on the pre-purchase of the assured base.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 313

Loyalty Rebates and “Merger Standards”

313

has been over-allocated to that range and the implied price for sales in this range exceeds the appropriate measure of cost, this will typically indicate that the discount scheme does not give rise to foreclosure concerns. In summary, reference can be made to the same economic principles that have been used to analyze predatory pricing by considering whether the discounted price lies above the appropriate measure of cost. If so, the discount most likely represents a form of price competition, and an equally efficient rival could match the discounted price. This will be the case if both of the following two tests are met. • Does total expenditure under the retrospective rebate scheme cover total avoidable cost? • Where total expenditure under the retrospective rebate scheme does cover total avoidable cost, is there evidence of targeting below-cost prices in a range open to competition? In the above example, only the first of these tests was fulfilled.

3.3.4 Summary Loyalty rebate schemes do not necessarily give rise to “suction effects”. A more dynamic and complete competitive assessment reveals that the incentive properties of any given scheme are more complex. However, this section did outline the circumstances under which a loyalty rebate scheme might indeed be exclusionary.

4. Assessing competitive effects in practice: a suggested approach This section seeks to set out a practical framework for determining whether a loyalty rebate gives rise to competition concerns. Once we move away from the current per se prohibition and towards an effects-based system, we are immediately faced with the question: how do we discriminate between pro-competitive and anticompetitive behaviour of dominant firms? After defining the relevant market (hopefully taking into account all the inherent issues in that step) and after identifying whether the company has significant market power, an effects-based assessment of the conduct of dominant firms should embody the following consumer-welfare based tests/considerations.45 45 As noted in the main text, we continue to consider defining the relevant market and establishing dominance to provide a useful step in the analysis. Defining the relevant market provides a useful preliminary step to the assessment of the competitive effects of the conduct in question. Nota

(N) Ehlermann 08 Bishop

314

6/11/08

11:50

Page 314

Session Two: A Reformed Approach to Exclusionary Conduct

An effects-based assessment needs to provide a coherent explanation of anticompetitive harm. Such an explanation is particularly important in the context of exclusionary abuses (i.e., the type of conduct covered by the Discussion Paper), where the need to distinguish between harm to competitors and harm to competition is particularly acute. As noted above, each category of potential exclusionary abuse has, first and foremost, important pro-competitive and efficiency-enhancing benefits. This important fact is all too often forgotten and the Discussion Paper does nothing to reverse this deficiency in the proposed competitive analysis. To guard against this tendency, a central element of any investigation should explain how the conduct of the dominant firm leads to the foreclosure of competitors and why the consequent change in market structure ultimately enables the dominant firm to increase price above the competitive levels. By requiring competition authorities to spell out the theory of harm in each case, it becomes clear that establishing harm to competitors is a necessary (but not sufficient) element of any theory of anticompetitive harm. This is particularly the case where nondominant firms in the same industry are offering similar rebate schemes; such empirical observations would indicate that loyalty rebate schemes have a pro-competitive rationale. This is not to say that the rebate scheme offered by a dominant firm could not have exclusionary effects but merely to state that the hurdle for concluding competitive harm should be raised accordingly as compared to the situation where only the dominant firm offers a rebate scheme.46 At a minimum, a proper effects-based system should have regard to actual market evidence. Almost by definition, a proper effects-based assessment will consider whether market evidence is consistent with competitors being harmed. But in most instances, loyalty rebate schemes give rise to testable hypotheses. Whether rivals are harmed can be assessed empirically by considering the behaviour of the dominant firm’s customers over time. For example, when assessing the competitive effects of, say, a loyalty rebate scheme, if it can be seen that individual customers switch from supplier to supplier from year to year this would indicate that customers are not “loyal” to suppliers. It is also critically important to recognize that foreclosure is an aggregate concept: foreclosure takes place, if at all, at the market level and not at the individual customer level. Moreover, considering what has happened to competing suppliers’ market shares over time is indicative of whether the business conduct of the dominant firm under investigation has indeed had a market-foreclosing effect by marginbene, this is in direct contrast to the approach put forward in the paper prepared by the EAGCP. The EAGCP paper appears to argue that the establishment of dominance is not required in an effects-based analysis. See EAGCP, “An economic approach to Article 82” (July 2005), available at http://ec.europa.eu/comm/competition/publications/studies/eagcp_july_21_05.pdf. 46 But even here, it cannot be assumed that the rebate scheme has exclusionary effects, for the reasons described in Section 4.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 315

Loyalty Rebates and “Merger Standards”

315

alizing competitors. Where the evidence shows that competitors are able to increase their market share, this raises a strong economic presumption that the competitive conduct under investigation is not foreclosing competition. This presumption appears to be at odds with European Court judgments where the central issue is the protection of the structure of the market so that any adverse impact on competitors is deemed to harm “competition as such (as an institution)”.47 This latter definition of foreclosure is explicitly aimed at protecting not consumers but rather competitors, on the assumption that long run harm to consumers is to be expected—an assumption that is itself predicated upon a finding of dominance. Such views would appear to be in conflict with the views expressed by the Courts in the context of merger control.48 However, the converse is not true: a decline in a rival firm’s market share is consistent with but does not prove exclusionary conduct on the part of the dominant firm. It is important to be clear about what is meant by the marginalization of competitors. Simply noting that a dominant firm’s rivals are adversely affected in the sense that they will find it harder to make sales is insufficient to conclude that these firms are marginalized. Indeed, harming competitors is an integral part of the competitive process: the harder one firm competes, the harder it is for its rivals to win sales. Firms can be said to be marginalized only if the ability to compete for future sales is permanently reduced even if prices were to be increased above competitive levels. Sustainable price increases can occur only if either competitors are forced to withdraw existing products permanently from the market or the marginal cost of supply is permanently adversely affected.49,50 To conclude that the conduct of the dominant firm is exclusionary, one therefore needs to consider whether the dominant firm is increasing its market share by competing on the merits or through anticompetitive conduct. In order to distinguish between these two possibilities, regard is usually given to whether the competitors are equally efficient competitors. If rival firms can be shown to be less efficient than the dominant firm, it is unlikely that the loss of market share is the result of anticompetitive conduct. In order to determine whether competitors are equally efficient competitors, reference is made to cost benchmarks: if the dominant firm can be shown to be pricing above the appropriate cost benchmark, then exclusionary conduct can be ruled out.

47

Opinion of Advocate General Kokott in British Airways, cited supra note 11, paras. 68 and

86. 48 See, e.g., Cases T-5/02 and T-80/02, Tetra Laval BV v Commission [2002] ECR II-4381, upheld: Cases C-12/03 P and C-13/03 P, Commission v Tetra Laval [2005] ECR I-987. 49 Nota bene, although the loss of sales may lead to an increase in a firm’s average cost, this will only result in that firm’s marginalization if its unit costs are adversely affected or if it ceases to invest in the development of new products. 50 Nota bene, this is consistent with the approach adopted in Commission Decision of 21 January 2004, GE/Amersham, Case No COMP/M.3304.

(N) Ehlermann 08 Bishop

316

6/11/08

11:50

Page 316

Session Two: A Reformed Approach to Exclusionary Conduct

This raises the question of which cost benchmark should be used. An obvious cost benchmark is average avoidable cost (AAC), since pricing below AAC involves incurring losses that could otherwise be avoided. Such pricing has the potential to exclude “as efficient” competitors. In contrast, selling output at a price above AAC is normally consistent with short-term profit maximizing behaviour, since at these prices the dominant firm covers the variable cost of producing the additional output and makes a positive contribution to fixed costs. Accordingly, pricing behaviour whereby a firm sells part or even all of its output at prices below ATC but above AAC is widely used in competitive markets. Although such prices can in principle raise competition concerns in the long term,51 the presumption should be that such pricing conduct represents normal competition.52

5. Conclusions Based on the foregoing, I would offer the following three policy recommendations. • First, start focusing on the dynamic incentive properties of loyalty rebate schemes rather than focusing on the last unit. • Second, start recognizing the elephant in the room, i.e., re-examine the role that the concept of dominance plays in the competitive assessment. This is critical for an attempt to introduce an effects-based approach to assessing Article 82 investigations. Currently, once dominance has been found, it is presumed that any harm to competitors necessarily translates into harm to competition. This is not only contrary to an effects-based assessment; it also ignores the difficulties with defining the relevant market which are raised by the cellophane fallacy. • Third, start being “positive”, i.e., start publishing clearance decisions. This is the only way to really change towards an effects-based system. In effect, these policy recommendations boil down to bringing Article 82 into line with merger control. There are two aspects in which the standards of merger control should be adopted in relation to Article 82. First, it needs to be recognized that the competitive assessment of loyalty rebates is in a number of respects analogous to the competitive assessment of vertical restraints and vertical mergers. But whereas the share-shifting that arises under vertical 51 For example, it might be argued that an “as efficient” rival is unable to fully recover fixed costs and is forced to exit the market as a result of financial constraints, 52 This issue is analogous to the assessment of conglomerate mergers, where potential longrun harm has to be contrasted with immediate observable short-term consumer benefits. See, e.g., the judgment of the CFI in Tetra Laval, cited supra note 48

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 317

Loyalty Rebates and “Merger Standards”

317

mergers, for example, is not presumed to give rise to competition concerns, the opposite is the case when assessing loyalty rebates. Second, the competitive assessment needs to use market evidence to examine testable hypotheses that arise from examining the competitive effects of loyalty rebates. For example, if a year-on-year growth target is alleged to lead to rival firms being forced from the marketplace since customers inevitably focus more and more on the dominant firm’s products and services, let’s examine recent market evidence to see whether this is indeed the case.

(N) Ehlermann 08 Bishop

6/11/08

11:50

Page 318

(O) Ehlermann 08 Allendesalazar

6/11/08

11:51

Page 319

II Rafael Allendesalazar* Can We Finally Say Farewell to the “Special Responsibility” of Dominant Companies? June 2007

Of all the stimulating questions included by Professor Ehlermann in his background paper, I was particularly attracted by one of the topics he raised, i.e., the concept of a “special responsibility” imposed on dominant companies not to allow their conduct to impair genuine undistorted competition in the market. The specific question Professor Ehlermann asks in this respect is: What is the guidance provided by the concept of “special responsibility” of the dominant company, in order to identify anti-competitive behaviour? Reading the background paper, I was initially surprised by the fact that Professor Ehlermann had included this question under the heading of “dominance”, whereas I would normally have thought of “special responsibility” as a notion related rather to the second tier in the application of Article 82 EC, i.e., the concept of abuse. But when going through the precedents to prepare this paper, I was even more surprised to see that the first time the Court of Justice referred to this special responsibility of dominant firms, in its famous Michelin I judgment of 1983,1 it had done so while analyzing whether Michelin was in a dominant position, and rejecting the company’s argument, supported by the French Government, that it was being penalized for the quality of its products and services. Notwithstanding this first judgment, in subsequent cases the special responsibility of dominant companies has been mentioned when analyzing the abusive nature of the commercial practices at question, rather than in the context of dominance. Should we draw any practical consequence from this different use of the concept of special responsibility? I’m not sure. But it could be argued that the application of the concept of “special responsibility” would be less controversial if it were restricted to indicating when a company should be aware that, given its market power (or to use legal terminology, due to its dominant position), its commercial practices could have an exclusionary effect, rather

* Partner, Martínez Lage & Asociados and Howrey. 1 Case 322/81, NV Nederlandsche Banden-Industrie Michelin v Commission [1983] ECR 3461, para. 57.

(O) Ehlermann 08 Allendesalazar

320

6/11/08

11:51

Page 320

Session Two: A Reformed Approach to Exclusionary Conduct

than applying this concept as a normative example of abuse. In other words, a company would be subject to this special responsibility whenever it has an unconstrained degree of market power enabling it to adopt unilateral commercial practices capable of producing anticompetitive effects. This interpretation would have the advantage of bridging the concepts of dominance and abuse by linking the degree of dominance with the capability of engaging in conduct that restricts competition, in what the Commission’s Discussion Paper on Article 82 describes as sliding scales:2 the degree of special responsibility of the company would depend on the higher or lower degree of its market power and on the higher or lower capability of the specific conduct to produce anticompetitive effects. There is a downside to this interpretation. Once the focus is set on the effects of a firm’s conduct, it is (at least theoretically) possible that unilateral conduct by a non-dominant company will restrict competition.3 In sliding scale terms, this would occur when the conduct is highly likely to produce exclusionary effects, even though it is pursued by companies with a low degree of market power. This damaging conduct would not be prohibited by Article 81, due to its unilateral nature, nor by Article 82, which of course only applies to dominant companies. Yet this criticism seems more theoretical than practical: it is difficult to imagine a company successfully adopting and maintaining unilateral abusive conduct if it does not enjoy a high degree of market power, i.e., if it is not dominant. We come, then, to the question raised by Professor Ehlermann: does the concept of “special responsibility” provide any real and useful guidance to the dominant company in order to identify which kinds of behaviour can be considered anticompetitive and which are merely “normal competition” (whatever this expression may mean)? The answer is clearly no. The concept of “special responsibility” neither helps to recognize when conduct will be considered abusive nor explains why it will be prohibited. For example, it does not allow a company to know, when 2 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf, para. 59. 3 The group of experts that commented on Article 82 referred to this possibility:

“Our proposed effect-based approach also allows us to capture in a balanced and meaningful way the notion of special responsibility of a dominant firm. The reference to such responsibility is often intended to prohibit some practices when exerted by a dominant firm, while considering them lawful if practiced by smaller competitors. Once we focus on the exclusionary effects of market practices, the notion of special responsibility naturally emerges from the analysis, in that certain practices are to be prohibited when they determine exclusionary effects, while they are lawful as long as no competitive harm is involved. Since in this analysis we do not need to assess the existence of dominance separately, the special responsibility implicitly applies to any conduct and firm that (is able to) interfere and distort the competitive process of entry into the market.” Report by the EAGCP, “An economic approach to Article 82”, p. 15. The report is available at http://ec.europa.eu/comm/competition/publications/studies/eagcp_july_21_05.pdf.

(O) Ehlermann 08 Allendesalazar

6/11/08

11:51

Page 321

Farewell to the “Special Responsibility” of Dominant Companies? 321 establishing its discount systems, whether its reference periods will be considered “relatively long” within the meaning of Michelin I,4 or to predict when offering bundled rebates for its different brands may lead to considering such rebates as loyalty-enhancing and therefore abusive.5 The essence of this special responsibility is to prohibit some practices when exerted by a dominant firm while considering them lawful if practiced by smaller competitors. Coming back to the first appearance of this “special responsibility” in Michelin I, the Court there stated that: “A finding that an undertaking has a dominant position is not in itself a recrimination but simply means that, irrespective of the reasons for which it has such a dominant position,6 the undertaking concerned has a special responsibility not to allow its conduct to impair genuine undistorted competition on the Common Market.” 7.

The Commission, supported by the Community Courts, has repeatedly broadened the scope of the special responsibility with almost every new Article 82 decision, depriving this concept of any role it may have had in providing guidance as to when and why a specific unilateral practice should be considered anticompetitive. The concept of special responsibility has thus become a mere litany chanted by the Commission and the Courts to conclude that certain kinds of conduct that are ubiquitous in the marketplace will nonetheless be prohibited when practiced by dominant companies and will give rise to immense fines. In other words, under the traditional form-based approach to Article 82, the Commission and the Courts have relied on the doctrine of special responsibility to sanction perfectly sound commercial practices that are otherwise considered pro-competitive, without having to prove the anticompetitive effects of such practices.

4 In the Michelin I case (cited supra note 2), one year was considered too long whereas this same duration was accepted in the British Gypsum case. See Notice pursuant to Article 19 (3) of Council Regulation No 17 concerning an application for negative clearance or exemption pursuant to Article 85 (3) of the EEC Treaty—Case No IV / 32.929—British Gypsum—‘Super Stockist Scheme’, [1992] OJ C321/9. In British Airways, six months was considered an acceptable reference period (see Commission Decision 2000/75 of 14 July 1999, Virgin/British Airways [2000] OJ L30/1, upheld: Case T-219/99, British Airways v Commission [2003] ECR II-5917; upheld on further appeal: Case C-95/04 P, [2007] ECR I-2331), whereas in Coca-Cola Export, the permitted reference period was three months. See Commission Press Release IP/90/7 of 9 January 1990. 5 The Coca-Cola Company, for instance, has been obliged to apply separate stocking commitment and rebates schemes for regular Cola, Cola Light and orange carbonated soft drinks. See Commission Decision of 22 June 2005, Coca-Cola, Case COMP/39.116. 6 The fact that this doctrine is applicable irrespective of the reason for which the company has reached its dominant position clearly indicates that it applies not only to former legal monopolies but also to companies that have acquired a dominant position by competing in the market. There are strong reasons, however, suggesting that residual monopolies left over from the age of State-granted exclusive rights should be subject to tougher limitations. 7 Michelin, supra note 2, para. 57.

(O) Ehlermann 08 Allendesalazar

322

6/11/08

11:51

Page 322

Session Two: A Reformed Approach to Exclusionary Conduct

For instance, in the British Gypsum8 and the Atlantic Container Line9 judgments, the ECJ stated that the special responsibility prevents dominant companies from using commercial practices that are perfectly standard in the market, “notwithstanding the fact that they are adopted by most, if not all, of their competitors”. But it can be argued that, by imposing this limitation, the Court compels the dominant companies to behave precisely as a dominant company, i.e., to act to an appreciable extent independently of its competitors. The same can be said of the duty imposed on the dominant company to refrain from applying commercial practices even where they have been specifically requested by its clients. By having to refuse its client’s request, paradoxically the company is again forced to exert its market power vis-à-vis its customers. Following the Ice-cream cabinet case,10 we also know that dominant companies may even have to refrain from business activities that “contribute to an improvement in production or distribution of goods”. But wouldn’t such conduct normally be regarded as efficient, for instance in the context of Article 81(3)? Isn’t this obligation somehow at odds with the efficiency defence explicitly recognized in the Discussion Paper, and somehow also in the British Airways11 judgment? Furthermore, the special responsibility imposed on dominant companies has also required them to ignore the basic economic and commercial logic of discounts, by making it abusive to offer specific discounts only to marginal clients instead of extending them across the board, as in Irish Sugar,12 or by stating the general principle that rebates can be abusive if they are not related to any specific cost savings or efficiency gains (British Airways). As a last example of this expanding interpretation of the notion of the special responsibility doctrine, the recent Wanadoo13 judgment reminds us that this doctrine may even preclude a dominant company from aligning its promotional prices with those of its non-dominant competitors, that is, it may preclude them from using the so-called “meeting competition defence”, which had at least theoretically been recognized in previous case law. It is therefore safe to conclude that the concept of special responsibility does not provide any guidance either for determining when a company can be subject to Article 82 or for identifying or explaining why sound commercial

8

Case C-310/93 P, BPB Industries and British Gypsum v Commission [1995] ECR I-865. Joined cases T-191/98, T-212/98, T-213/98 and T-214/98, Atlantic Container Line v Commission [2003] ECR II-3275. 10 Case T-65/98, Van den Bergh Foods v Commission, [2003] ECR II-4653. 11 Case C-95/04 P, British Airways v Commission, [2007] ECR I-2331. 12 Case T-228/97, Irish Sugar v Commission [1999] ECR II-2969. 13 Case T-340/03, France Télécom v Commission [2007] ECR II-107, on appeal: Case C-202/07 P, not yet decided. 9

(O) Ehlermann 08 Allendesalazar

6/11/08

11:51

Page 323

Farewell to the “Special Responsibility” of Dominant Companies? 323 practices that are totally common in the marketplace become abusive when applied by companies burdened with such a special responsibility. The consequence of such an unrestrained application of the concept of special responsibility is that, ultimately, proof of dominance is almost sufficient to establish an abuse. Even members of the Chief Economist’s office14 have recognized this.15 This problem is reinforced by the fact that neither the Commission’s decisional practice nor the case law of the Community Courts provides an alternative consistent definition of abuse, particularly with respect to exclusionary abuses. On the contrary, the application of Article 82 has been driven by vague slogans such as “special responsibility”, “competition on the merits”, “normal competition”, and “level playing field”, the exact meanings of which have never been defined. The diagnosis offered by Einer Elhauge in the context of US monopolization law (“monopolization doctrine currently uses vacuous standards and conclusory labels that provide no meaningful guidance about which conduct will be condemned as exclusionary”16) is perfectly applicable to the traditional practice of the Commission and the Community Courts in Article 82 cases. A couple of years ago, the OECD’s paper “Competition on the Merits”17 pointed out that the days in which competition authorities can rely on such slogans without drawing bitter criticism are dwindling. And Professor Joskow has explained that “antitrust policy needs to evolve in a way that firms receive clear signs from these enforcement institutions, so that they are able to determine where to draw the line between behaviour and market structures that are likely to be legal and those that are illegal”.18 The Commission’s Discussion Paper on Article 82, or at least the general principles set out in its initial sections, can be regarded as a first step towards this desirable goal. The fact that the Paper patently ignores the concept of special responsibility is in itself noticeable, even though it still has recourse to some of the other slogans mentioned above such as “normal competition” and “competition of the merits”. Furthermore, the most recent judgments

14 Miguel de la Mano, “The dominance concept: new wine in old bottles”, FTC/DOJ Hearings on single-firm conduct, Washington DC, 7 March 2007. 15 It is worth noting that, while under the traditional formalistic approach to Article 82 the special responsibility implies that dominance is pivotal to the application of this provision, to the point that little or no attention is given to the anticompetitive effects of the conduct, under a purely economic approach the same special responsibility would imply that any anticompetitive conduct should be prohibited, regardless of whether the company involved is dominant or not. See the EACGP’s report, supra note 4. 16 Einer Elhauge, “Defining better monopolization Standards”, 56 Stanford Law Review 253 (2003). 17 Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=875360. 18 Paul Joskow, “Transaction Cost Economics, Antitrust Rules and Remedies”, 18 Journal of Law, Economics and Organization 95 (2002).

(O) Ehlermann 08 Allendesalazar

324

6/11/08

11:51

Page 324

Session Two: A Reformed Approach to Exclusionary Conduct

seem to show that the CFI and the ECJ are also trying to avoid relying on the concept of special responsibility. The disappointing and poorly argued Wannadoo judgment19 does mention it, but only by quoting the contested decision of the Commission and not by including the concept in its own reasoning. The Discussion Paper clearly opts for the “equally efficient firm” test to distinguish exclusionary conduct from healthy competition. This is certainly a first step in the right direction, although it is unfortunate that the Paper does not explain why it has opted for this specific test and under which circumstances it believes that it is superior to other tests such as the “no economic sense” test, the profit sacrifice test, the consumer welfare test or Professor Elhauge’s efficiency test. However, this initial optimism waters down when reading the second part of the Discussion Paper, where the analysis of several specific exclusionary practices is kept within the stringent orthodoxy of the existing case law. This conflict between the new economic approach to Article 82 and the way it has traditionally been applied is clearly reflected in the Opinion by Advocate General Kokott in the Court’s British Airways case, where she stated that any reorientation by the Commission in the application of Article 82 will have to remain within the framework prescribed for it by the provision itself as interpreted by the ECJ.20 This proves that, if the application of Article 82 is to be reoriented towards a new economic approach, a goal I believe is now unanimously accepted, it will have to be done by applying a great deal of amnesia as regards most of its precedents and as regards the case law of the Community Courts. But it is impossible to make an omelette without breaking eggs. Furthermore, if the Commission wants to follow this path, it should plunge into it as soon as possible, making its position explicit and clear, using watertight arguments that can overcome the traditional approach embedded in its own former decisional practice and in the judgments of the ECJ and the CFI. If the Commission does not act decisively, then, given the increasingly decentralized application of EC competition law, national courts and national authorities will probably feel obliged by Article 16 of Regulation 1/2003 to refer to the traditional formalistic approach to Article 82, instead of applying the new, and for some time, judicially untested economic approach.

19 20

Supra note 15. See paragraph 28 of the Opinion of Advocate General, in British Airways, supra note 13: “In this context it is immaterial how the Commission intends to define its competition policy with regard to Article 82 EC for the future. Any reorientation in the application of Article 82 EC can be of relevance only for future decisions of the Commission, not for the legal assessment of a decision already taken. Moreover, even if its administrative practice were to change, the Commission would still have to act within the framework prescribed for it by Article 82 EC as interpreted by the Court of Justice.” (footnote omitted)

(O) Ehlermann 08 Allendesalazar

6/11/08

11:51

Page 325

Farewell to the “Special Responsibility” of Dominant Companies? 325 To this extent it is important for the Commission to change not only the way it has been applying Article 82 in terms of substance, but also the way it handles its administrative practice. As previously mentioned, enforcement institutions should endeavour to provide firms with clear signs of which kinds of conduct are likely to be legal and which kinds are likely to be condemned, a policy objective that is even more vital when an authority has publicly undertaken a new approach to certain practices. Presently the Commission only publishes decisions declaring the existence of abusive practices. No publicity (or only very limited publicity) is ever given to decisions in which it rejects complaints. Therefore, the precedents provide all sorts of examples of conduct that has been considered abusive. It is about time to start identifying clearly which ubiquitous practices should not be treated as anticompetitive and to create safe harbours in which all companies, whether dominant or not, will be able to compete. The best practical way the Commission can do so is by publishing decisions repelling complaints based on Article 82.21 Given the Courts’ record in sustaining all the Commission’s decisions applying Article 82, it could be hoped that, if the new approach is clearly and substantially reasoned, the Courts will set aside many of their own judgments and endorse a reoriented approach to Article 82; an approach whereby the concept of special responsibility will finally disappear, or at least lose the major role it has played in the application of Article 82 in the last decade.

21 The Commission should also speed up the publication of its sanctioning decisions, which recently have dragged on for many months. Thus, it has not yet published any decision adopted under the “new approach” to Article 82. An appalling example is the Tomra decision (Commission Decision of 29 March 2006, Prokent/Tomra, Case COMP/38.113) adopted almost a year and a half ago and as of September 2007 still unreported. If, as mentioned by a DG COMP official, this decision “can be considered an important step towards the envisaged reform of the application of Article 82 EC Treaty” (Frank Maier-Rigaud and Dovile Vaigauskaite, “Prokent/Tomra, a textbook case? Abuse of dominance under perfect information”, Competition Policy Newsletter No. 2 (2006)), such an important decision should be immediately published to allow companies, practitioners and other enforcement authorities to grasp the extent and depth of the reformed application of Article 82. The present excessive delays in publishing these decisions cannot be explained either by reasons of confidentiality (many other competition authorities are perfectly able to deliver public versions of their decisions immediately) or by difficulties of translation (hundreds of other Community acts are published every day in the OJ in 23 languages). [Editor’s note: the lengthy decision is now available at http://ec.europa.eu/comm/ competition/antitrust/cases/index/by_nr_76.html#i38_113.]

(O) Ehlermann 08 Allendesalazar

6/11/08

11:51

Page 326

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 327

III Robert O’Donoghue* Verbalizing a General Test for Exclusionary Conduct under Article 82 EC August 2007

This paper considers the vexed issue of how to verbalize a generally-applicable definition of exclusionary conduct under laws relating to unilateral behaviour (or “monopolization”, in US antitrust law parlance), such as Article 82 EC. The issue is of obvious practical importance: business firms that are, or may be, dominant need reasonably clear guidance on fundamental matters such as the restrictions on their unilateral pricing conduct and to what extent they are obliged to deal with third parties with whom they do not wish to deal. Equally, in many sectors such as recently-liberalized utility sectors, new entrants face serious market entry obstacles as a result of abusive vertical foreclosure, e.g., discrimination, denial of access, quality degradation, etc. There is thus a clear need both for enforcement and for rules that companies themselves can reasonably apply ex ante. The issue is also highly topical. The EU Commission’s review of Article 82 is continuing, with a draft set of guidelines expected later in 2007. The US antitrust agencies also concluded, on 8 May 2007, their year-long hearings on single-firm conduct.1 Unfortunately, however, the basic test for exclusionary conduct has not been clarified to any great extent. The application of Article 82 to exclusionary conduct has been described as “vague” and “conclusory”.2 Similar criticisms have been voiced in the United States,3 where leading antitrust thinkers have described the treatment of unilateral abusive conduct as “the biggest substantive issue facing antitrust”.4 Given the uncertain and vague nature of current definitions of exclusionary behaviour, lawyers and economists have made a number of recent * Barrister, Brick Court Chambers, London and Brussels (robert.odonoghue@brickcourt. co.uk). Sections of this paper draw from Robert O’Donoghue and A. Jorge Padilla, The Law And Economics Of Article 82 EC, Hart Publishing, 2006. 1 Comprising 18 days of hearings, 28 different panels, and 130 panellists drawn from a range of stakeholders (e.g., lawyers, economists, business people, academics, and business historians). 2 Robert O’Donoghue and A. Jorge Padilla, The Law And Economics Of Article 82 EC, Hart Publishing, 2006, p. 177. 3 See Einer Elhauge, “Defining Better Monopolization Standards”, 56 Stanford Law Review 253 (2003). 4 Richard A. Posner, “Vertical Restraints and Antitrust Policy”, 72 University of Chicago Law Review 229 (2005).

(P) Ehlermann 08 O’Donoghue

328

6/11/08

11:51

Page 328

Session Two: A Reformed Approach to Exclusionary Conduct

proposals that seek to offer a unified definition of exclusionary conduct.5 These include the “profit sacrifice” test and its close cousin the “no economic sense” test, the “equally efficient competitor” test, the “consumer welfare” test, and the “limiting production to the prejudice of consumers” test. These tests obviously have much in common, but their principal proponents have also sought to argue that alternative tests are flawed and mutually exclusive. This paper outlines some general considerations that should guide the framing of useful, operational rules for unilateral conduct, and in particular the relevance of error costs, the need for administrable ex ante rules, and the problems of imperfect information (Section I). It then considers the main elements of each of the principal tests proposed for the assessment of exclusionary conduct, followed by the principal criticisms aimed at them (Section II). The role of efficiencies is also then considered, as this is a central element in many of the proposed tests (Section III). Finally, a tentative conclusion is set out on the usefulness of each test as an operational rule (Section IV).

I. Designing Optimal Rules Legal rules, including competition law rules, should be designed in a way that makes their enforcement efficient and practical.6 This is premised on the notion that perfect information allowing competition authorities and courts to weigh the pro-competitive and anticompetitive effects of a practice in every case will almost never be available. A related point is that, for unilateral conduct, it is not reasonable to expect firms to subject everyday business decisions to detailed balancing analysis to scrutinize their compatibility with competition law. For example, firms operating on an EU-wide basis cannot reasonably be expected to assess the consumer welfare impact of offering different prices and terms and conditions in each Member State, in order to avoid unlawful discrimination claims under Article 82(c). In most instances, the sheer multiplicity of different countries, customers, prices, and factors that lead to pricing changes make this wholly impractical. Antitrust commentators therefore propose that legal rules should be guided by several

5 For an overview of the main tests, see John Vickers, “Abuse of Market Power,” Speech to the 31st conference of the European Association of Research in Industrial Economics, Berlin, 3 September 2004. See also Organisation for Economic Co-operation and Development, “Competition on the Merits,” Background Note (May 2005), available at http://papers.ssrn. com/sol3/papers.cfm?abstract_id=875360. 6 See, e.g., Richard Posner, “An Economic Approach To Legal Procedure And Judicial Administration” 2 Journal Of Legal Studies 399 (1973). For a recent application more specific to unilateral conduct, see David Evans, “How Economists Can Help Courts Design Competition Rules: An EU and US Perspective”, 28(1) World Competition 93 (2005).

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 329

A General Test for Exclusionary Conduct under Article 82 EC 329 considerations that make their enforcement optimal. Two particular considerations bear mention: (1) error costs; and (2) form versus effects rules.

A. Error Costs In competition law, the optimal theoretical legal rule would always identify anticompetitive conduct while always allowing pro-competitive conduct to go unpunished. Unfortunately in practice, virtually all legal rules suffer from imperfections as a result of lack of information, enforcement, and other sources of potential error (e.g., lack of expertise). These difficulties mean that competition rules must grapple with the dangers of allowing anticompetitive practices to go unpunished (so-called “false negatives” in statistical parlance) and/or treating pro-competitive practices as anticompetitive (so-called “false positives”). Which of the two errors is likely to be more costly depends on whether a particular practice is, on balance, more likely to lead to harm or good. On a very general level, there are probably good reasons why the risk of overdeterrence is greater than under-deterrence in the context of Article 82. Competition laws are generally more hostile to collusive arrangements between firms (be they mergers or other agreements) than they are to unilateral conduct. This is mainly on the assumption that competitive harm is generally more likely to occur from two or more firms agreeing to limit their output than unilateral action by one firm. Put differently, it is one thing for a firm to acquire market power through superior products or skill, but quite another for two or more firms to restrict competition between them in favour of cooperative arrangements that confer market power. This distinction is not necessarily hard and fast—many dominant firms may acquire a monopoly position by means other than skill and foresight (e.g., where the government grants them special or exclusive rights)—but it is probably correct, as a general matter, to treat market power that results from an agreement between two or more firms differently from the way unilateral action by a firm with market power is treated. Regarding specific types of abusive practices, Article 82 already incorporates certain rules that reflect concerns for over-deterrence and underdeterrence. A good example concerns unconditional price cuts by a dominant firm. Price competition is of course to be encouraged. At some point, however, price competition may cause harm to consumers, where for example a dominant firm charges low prices to cause rivals’ exit, and later recoups its investment through increased prices in future. While the precise measurements differ, economists have long argued that firms should be presumed to be acting lawfully when prices are above production costs, usually marginal cost or some analogous measure. This insight is captured by the first rule on

(P) Ehlermann 08 O’Donoghue

330

6/11/08

11:51

Page 330

Session Two: A Reformed Approach to Exclusionary Conduct

predatory pricing in the AKZO case, that prices below average variable cost (a proxy for marginal cost) are presumed to be exclusionary.7 (However, even that presumption has been relaxed in recent years given the legitimate reasons why a dominant firm might price below cost for a period, e.g., in order to introduce new products, or to build a network.) Some economists have also devised theoretical models showing that prices above cost can sometimes harm consumer welfare.8 The basic idea is that less efficient rivals can bring about reductions in price that are sufficient to compensate for their relative inefficiency, as well as the notion that many of them will become as or more efficient over time. But this insight has not led to general restrictions on above-cost price cuts under Article 82, precisely because of the very high risk of wrongly condemning aggressive, but legitimate, price competition. Instead, such price cuts have been condemned in only exceptional cases, usually where the firm in question is a virtual monopolist and/or the pricing strategy is part of a series of abusive acts with the same aim.9 Many commentators would argue that even this exception goes too far and risks false positives. But there is nonetheless a strong consensus that above-cost unconditional price cuts should be presumed lawful in all but extreme cases.

B. Form Versus Effect An evaluation of the risks of false negatives and positives, and prior beliefs about the degree of benefit and harm of particular practices, has led to the application of different types of tests for antitrust rules. At one extreme are practices subject to per se legality or illegality rules, i.e., the practice is deemed lawful or unlawful without the need for a detailed inquiry into its actual or likely effects on competition. A per se rule may be absolute in the sense that no exceptions are permitted or modified in the sense that a rebuttable presumption of legality or illegality applies. Per se rules are only appropriate where: (1) experience and logic suggest that the benefit/harm resulting from a practice is so clear and unambiguous that there is no point in wasting court or regulatory resources in investigating its effects; and (2) the risk of false positives or false negatives is small. The only universally accepted per se antitrust rule is hard core horizontal price-fixing. At the other extreme lie “rule of reason” inquiries, where the benefits and harm caused by a practice are carefully evaluated. This inquiry may be struc7

See Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359, paras. 70–71. See Mark Armstrong and John Vickers, “Price Discrimination, Competition, and Regulation”, 41(4) Journal of Industrial Economics 335 (1993). 9 See, e.g., Joined Cases C-395/96 P and C-396/96 P, Compagnie Maritime Belge Transports SA, Compagnie maritime belge SA and Dafra-Lines A/S v Commission [2000] ECR I-1365. 8

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 331

A General Test for Exclusionary Conduct under Article 82 EC 331 tured, in the sense that conduct is evaluated through a series of screens to distinguish lawful and unlawful conduct, or it may be unstructured in that harm and benefits are simply assessed and compared. Economists overwhelmingly agree that a rule of reason (or effects) based approach is appropriate when dealing with unilateral conduct, and they have criticized past policy under Article 82 EC for its excessive reliance on form over effects. A recent report by the Economic Advisory Group on Competition Policy on Article 82 EC— which was commissioned by the Chief Economist of DG Competition—proposes an effects-based approach for the following reasons:10 “A more consistent approach would start out from the effects of anticompetitive conduct . . . and consider the competitive harm that is inflicted on consumers. Adopting such an effects-based approach would ensure that these various practices are treated consistently when they are adopted for the same purpose. In contrast, a form-based approach creates the risk that they will be treated inconsistently, with some practices possibly enjoying a relatively more lenient attitude (e.g., because of different standards). Arbitraging among these different treatments may facilitate exclusion, or induce the dominant firm to adopt alternative exclusionary methods, which may well inflict a higher cost on consumers.”

Another way of looking at these types of rules is to consider whether unilateral practices should be assessed on the basis of their form or actual or likely effect. Historically, a number of practices under Article 82 could have been regarded as being subject to per se illegality rules. Exclusive dealing, for example, was subject to a strong presumption of illegality in earlier cases such as Suiker Unie and Hoffmann-La Roche.11 This presumption has been relaxed in recent cases, most notably in Van den Bergh,12 as a result of which exclusive dealing under Article 82 is now more aptly characterized as being based on a rule of reason, much in the same way as a rule of reason applies under Article 81. Similarly, regarding predatory pricing, the AKZO case suggested that pricing below average variable cost is subject to a per se rule. This finding has also been relaxed in recent cases, in line with economic thinking indicating that pricing below average variable cost may have a non-exclusionary explanation. For example, in Wanadoo, the Commission did not conclude that the dominant firm’s prices were unlawful from the mere fact that they were below average variable cost for a significant period; rather, the Commission also looked at the strategic rationale for those prices and their effects on competition. The

10 See Report by the Economic Advisory Group on Competition Policy, “An Economic Approach to Article 82,” July 2005, available at: http://ec.europa.eu/comm/competition/publications/studies/eagcp_july_21_05.pdf., at p. 6. 11 See Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114/73, Coöperatieve Vereniging “Suiker Unie” UA and others v Commission [1975] ECR 1663; Case 85/76, Hoffmann-La Roche & Co AG v Commission [1979] ECR 461. 12 See Case T-65/98, Van den Bergh Foods Ltd v Commission [2003] ECR II-4653.

(P) Ehlermann 08 O’Donoghue

332

6/11/08

11:51

Page 332

Session Two: A Reformed Approach to Exclusionary Conduct

current rules might therefore best be described as constituting a regime of modified per se illegality. Finally, individualized retroactive loyalty rebates that apply over a relatively long reference period were also effectively subject, in the past, to a per se illegality rule (absent objective justification).13 Again, however, this rule has been substantially relaxed and has shifted towards a rule of reason-type inquiry, with some structural screens to eliminate unproblematic cases.14 In sum, there are now virtually no practices that could be described as per se unlawful under Article 82. In between the per se and pure effects approaches lies an intermediate approach where simple error-cost analysis, based on economic evidence, is used to structure administrable legal rules (a “structured rule of reason”). Such rules can help bridge the gap between a formal (and most likely prescriptive) approach and a pure effects-based approach (which is unlikely to yield much by way of business certainty ex ante). Take above-cost unconditional price cuts. On the basis of economic theory, the case may be made that such price cuts can harm consumer welfare in certain circumstances. But no clear legal rule has been devised to say when harm to consumer welfare occurs. Absent a clear rule, restricting unconditional above-cost price cuts is likely to greatly chill price competition. The optimal solution might therefore be to do nothing, even if in so doing certain anticompetitive practices thereby escape censure. However, under Article 82, the situation is one of modified per se legality: unconditional above-cost pricing is presumed legal absent exceptional circumstances (e.g., action by a virtual monopolist, industries with very high fixed-variable costs ratios).15 This debate has also found its way into the discussions of the Article 82 reforms. But the choice sometimes posited between an approach based on legal form and one based on economic effects is to some extent false and ignores a number of practical considerations. Relying only on legal form almost certainly leads to incorrect conclusions by ignoring the mixed eco13 See Denis Waelbroeck, “Michelin II: A Per Se Rule Against Rebates by Dominant Companies?” 1 Journal of Competition Law and Economics 149 (2005). 14 See Commission Decision of 29 March 2006, Prokent/Tomra, Case COMP/38.113, available at http://ec.europa.eu/comm/competition/antitrust/cases/index/by_nr_76.html#i38_113. It has been claimed by the Commission that a more lenient approach was applied in this case. See Frank Maier-Rigaud and Dovile Vaigauskaite, “Prokent/Tomra, a textbook case?”, Competition Policy Newsletter No. 2 (2006), pp. 19–24. The Court of Justice in British Airways seems to have applied something akin to a per se approach to retroactive rebates, but the case seems largely reflective of the Court’s unwillingness to re-assess what it classified as findings of fact by the Court of First Instance in the same case. The Court’s judgment does not in any event necessarily preclude an effects-based approach, since it held that the discounts must have an “exclusionary effect”. See Case C-95/04 P, British Airways v Commission [2007] ECR I-2331, para. 68. 15 See Commission Decision 93/82 of 23 December 1992, Cewal, Cowac and Ukwal [1993] OJ L34/20, upheld: Joined Cases T-24/93, T-25/93, T-26/93, and T-28/93, Compagnie Maritime Belge Transports SA and Others v Commission [1996] ECR II-1201; upheld on further appeal: Joined Cases C-395/96 P and C-396/96 P, Compagnie Maritime Belge Transports SA and Others v Commission [2000] ECR I-1365.

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 333

A General Test for Exclusionary Conduct under Article 82 EC 333 nomic effects of many unilateral practices. More importantly, if a form-based approach were adopted, it would be of limited practical importance, since per se legality guidance would almost certainly be offered only for a handful of practices that companies probably already know are legal. (As noted above, almost no practice is per se illegal now under Article 82, so telling firms what they certainly cannot do is likely to be of limited use in practice.) In other words, by its nature, general guidance for dominant firm conduct will not be permissive. Equally, however, proponents of an economic effects analysis also need to recognize that the law would be much less clear than it is already if each case depended on an assessment of the economic benefits and harm produced by a given type of conduct, most of which can only be assessed ex post (if at all). Economists sometimes underestimate the importance of legal certainty to businesses.

II. The Principal Suggested Tests for Exclusionary Conduct The search for clearer standards in abuse of dominance cases has led to a proliferation of different general tests for exclusionary conduct. Many of these tests are closely related—certainly more so than some of their proponents would seem to suggest. The main tests are as follows: • Profit sacrifice/no economic sense. A first test is based on the notion of profit sacrifice, meaning that exclusionary conduct requires a firm to deliberately forego a more profitable course of action.16 A closely-related cousin is the “no economic sense” test, which would treat conduct as exclusionary if it would make no economic sense but for its tendency to exclude rivals.17 • Equally-efficient competitor test. A second test is the equally efficient competitor test.18 This holds that the only conduct that is exclusionary is that which would exclude an equally or more efficient rival. Conduct that excludes less efficient rivals is deemed to be competition “on the merits”, on the ground that the competitive process would result in the elimination of such undertakings in any event. 16 The profit sacrifice test was originally proposed by industrial economists in the early 1980s. See Janusz Ordover and Robert Willig, “An Economic Definition of Predation: Pricing and Product Innovation”, 91 Yale Law Journal 8 (1981). The test was intended to provide an objective, transparent, and economically based framework for assessing exclusionary unilateral behaviour. The economists defined exclusionary behaviour as a “response to a rival that sacrifices part of the profit that could be earned under competitive circumstances were a rival to remain viable, in order to induce exit and gain consequent additional monopoly profits”. Ibid., pp. 9–10. 17 See Gregory Werden, “The ‘No Economic Sense’ Test for Exclusionary Conduct,” paper submitted to the British Institute of International and Comparative Law, 5th Annual Antitrust dialogue, London, 9–10 May 2005. 18 Richard Posner, Antitrust Law, 2nd ed., University of Chicago Press, 2001, pp. 194–95.

(P) Ehlermann 08 O’Donoghue

334

6/11/08

11:51

Page 334

Session Two: A Reformed Approach to Exclusionary Conduct

• Consumer welfare test. A third suggested test is a test based on consumer welfare. Under this test, only conduct that harms consumer welfare, or harms consumer welfare more than it enhances efficiency, is considered exclusionary.19 This test is expressly rooted in the objective of maximizing consumer welfare. The main elements of the foregoing tests, and the principal criticisms, are outlined below. But the differences between these tests should not be overstated. In essence, each seeks: on the one hand, to identify situations in which conduct is inefficient and hence anti-consumer; and, conversely, to allow efficient conduct that yields consumer benefits over time. There is also a fourth test: • Limiting production. This final test is expressly grounded in the wording of Article 82(b), which states that “limiting production, markets or technical development to the prejudice of consumers” is illegal. In its most basic form, this test suggests that foreclosure or handicapping of competitors, by which competition is reduced still further, is illegal if harm to consumers would actually or likely result.20

A. Profit Sacrifice/No Economic Sense Elements of the profit sacrifice test. The profit sacrifice test assumes that a firm would not rationally engage in exclusionary conduct unless it considers that any short-term sacrifice of profits would be more than compensated for by the expected gains as a result of the exclusion or discouraging of rival firms if the conduct is successful. The most obvious example concerns predatory pricing. The theory is that a firm would not knowingly sell below cost unless it had a reasonable expectation that short-term losses will be offset by the additional profits that follow in the medium- to long-term from the exclusion of rivals. The issue of recoupment—which arguably plays a role under Article 82—in effect seeks to measure whether profit sacrifice would be rational by assessing whether the longer term gains of below-cost pricing are likely to outweigh its

19 See Andrew Gavil, “Exclusionary Distribution Strategies by Dominant Firms: Striking a Better Balance”, 72 Antitrust Law Journal 3 (2004); Steven Salop, “Section 2 Paradigms and the Flawed Profit-Sacrifice Standard”, available at www.crai.com.au/agenda/salop%20paper.pdf; Maurits Dolmans, “Efficiency Defences Under Article 82 EC Seeking Profits Or Proportionality? The EC 2004 Microsoft Case in the Context of Trinko,” 24th Annual Antitrust And Trade Regulation Seminar, NERA, Santa Fe, New Mexico 8 July 2004 (on file with author). 20 See John Temple Lang and Robert O’Donoghue, “The Concept Of An Exclusionary Abuse”, Global Competition Law Centre Research Papers on Article 82 EC, College of Europe, July 2005, p. 38, available at http://www.coleurop.be/content/gclc/documents/GCLC% 20Research%20Papers%20on%20Article%2082%20EC.pdf.

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 335

A General Test for Exclusionary Conduct under Article 82 EC 335 short-term costs. (This is not, however, a strictly mathematical question, but involves a more general inquiry as to whether the predator could increase or maintain market power in the future.) Recent antitrust case law in the US has endorsed a profit sacrifice test to some extent, but judicial acceptance of the test has been mixed. In American Airlines,21 the Department of Justice (as plaintiff) argued that the appropriate inquiry in a predatory pricing case was whether incrementally-added capacity was money losing, even if the service provided by the incumbent airline remained profitable overall on the relevant city pair. The 10th Circuit held that, even under the standard advanced by the DOJ itself, it had not been demonstrated that the additions of capacity at issue were unprofitable.22 In Trinko,23 the DOJ (as amicus curiae) advocated essentially the same sacrifice test for assessing unilateral refusals to deal. Although the Supreme Court’s majority opinion did not expressly refer to the sacrifice test, it justified past cases in which a duty to deal was imposed on the basis that the defendant had foregone a more profitable course of conduct in refusing to deal. For example, in its discussion of Aspen Skiing, the Court attached importance to the fact that the defendant had refused to deal even when the requesting party offered a price equal to the retail price charged by the defendant downstream. It pointed to the defendant’s willingness to forego shortterm benefits through “[t]he unilateral termination of a voluntary (and thus presumably profitable) course of dealing”, and its “unwillingness to renew the ticket even if compensated at retail price”, as facts that suggested its “distinctly anticompetitive bent”. As a result, the DOJ has indicated that it plans to assert the sacrifice standard with renewed confidence following Trinko.24 Criticisms of the profit sacrifice test. The profit sacrifice test has been criticized in important respects. The first set of criticisms is fundamental in nature. Certain commentators have argued that the test is flawed in two critical respects.25 First, they argue that a number of types of conduct do not involve profit sacrifice, but have been recognized as exclusionary. For example, filing a false or overbroad patent application may be cheaper than filing a correct and properly-defined one. The same point can be made about other forms of non-price predation (e.g., falsely disparaging a rival), reprisal

21 United States v AMR Corp, 140 F. Supp. 2d 1141 (D. Kan. 2001), affirmed, 335 F.3d 1109 (10th Cir. 2003). 22 See R. Hewitt Pate, “The Common Law Approach and Improving Standards for Analyzing Single Firm Conduct”, Fordham International Antitrust Conference, New York, 23 October 2003. 23 Verizon Communications Inc v Law Offices of Curtis V. Trinko LLP, 540 U.S. 398 (2004). 24 See J. Bruce McDonald, Deputy Assistant Attorney General, Antitrust Division, “The Struggle For Standards”, presented before the American Bar Association Section of Antitrust Law, Spring Meeting, Washington, D.C., 1 April 2004 (quoting Trinko, cited in previous footnotes (emphasis in original)). 25 See Elhauge, supra note 3.

(P) Ehlermann 08 O’Donoghue

336

6/11/08

11:51

Page 336

Session Two: A Reformed Approach to Exclusionary Conduct

abuses, and anticompetitive forms of raising rivals’ costs. A profit sacrifice test would therefore seem to wrongly exclude such abuses.26 A second criticism is that a profit sacrifice test could capture a number of forms of highly desirable market activity. For example, in the area of intellectual property or major investments in tangible property, the initial investments would typically be unprofitable but for the prospect of later monopoly returns reaped by (lawfully) excluding competitors. A literal application of the sacrifice test might treat such investments as predatory despite the fact that, in general, they clearly benefit consumer welfare by offering a better market option. Of course, it might be argued that good sense would prevail in such circumstances and that the conduct in question would be seen as creating dynamic benefits. But a rule that contains exceptions based on the notion that “we will know them when we see them” is precarious. Another set of criticisms concerns the ease of application and predictability of the profit sacrifice test in practice. A number of difficulties are said to arise.27 First, there is the problem of determining the sacrifice: a sacrifice relative to what? It is not clear, for example, whether the profit sacrifice requires a firm to opt for the most profitable course of action to avoid a finding of exclusionary conduct or whether it should be required to have passed on a more profitable alternative. (Presumably, it is the latter.) It is also not clear what degree of sacrifice would be sufficient to establish exclusionary conduct or whether the rule is a strict one, i.e., is any profit sacrifice automatically abusive? Second, many abuse of dominance cases do not involve extending a monopoly and increasing profits; instead, they concern actions designed to maintain a monopoly. For example, a reprisal abuse may be carried out simply to make clear to rivals and customers that aggressive competition (or actions by customers to support rival firms, or complaints to competition authorities, etc.) will meet with an immediate response by the dominant firm. In such cases there may be no additional profits resulting from the dominant firm’s conduct, but it may serve to insulate an existing dominant position from future erosion. This also exposes a related problem: that the dominant firm’s current prices may already be above the competitive level. In this case, the absence of higher prices as a result of the exclusionary conduct could falsely show an absence of exclusionary effect on the grounds that no sacrifice occurred (a variant of the “cellophane fallacy”). Finally, although one of the main benefits of the profit sacrifice standard is said to be its objectivity,28 it is argued that the test would in practice be highly subjective and speculative. In its most basic form, the profit sacrifice test asks 26 See Brief for the United States and Federal Trade Commission as Amici Curiae Supporting Petitioner, Verizon Communications, Inc v Law Office of Curtis V. Trinko LLP, No. 02-682 (docketed US Sup. Ct. Dec. 13, 2002). 27 See Salop, supra note 19. 28 See, e.g., Mark Patterson, “The Sacrifice of Profits in Non-Price Predation”, 18 Antitrust 37 (2003).

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 337

A General Test for Exclusionary Conduct under Article 82 EC 337 a court to assess the dominant firm’s likely conduct in the hypothetical absence of an ability to raise prices. This is hypothetical, speculative, and uncertain. Different outcomes could be imagined on the basis of the same set of facts. The no economic sense test as a variant of profit sacrifice. Criticism of the profit sacrifice test has led the DOJ to argue for a variant of it: the “no economic sense” test. The DOJ has argued in recent antitrust cases that it is relevant to ask whether the conduct would make economic sense for the defendant but for its tendency to eliminate or lessen competition. According to the Department, conduct is not exclusionary or predatory if it would not make economic sense for the defendant but for its tendency to eliminate or lessen competition.29 The DOJ contends that this type of “sacrifice” is a more accurate measure for exclusionary abuses than profit sacrifice, as it entails a choice between a business strategy that would make no business sense but for the probability that the conduct would create or maintain monopoly power.30 The no economic sense test certainly addresses some of the criticisms of the profit sacrifice test. In particular, it does not characterize as unlawful every departure from short-run profit maximization. This would permit investments that confer long-term benefit by allowing a firm to retain exclusive control over its inventions, something that could in theory be regarded as suspect under the profit sacrifice standard. But it is clear in some instances, particularly those involving network effects, that the elimination of competition, and the survival of a single competitor, is beneficial. Investments in network effects of this kind would only make economic sense if rivals are eliminated. It is not clear how such cases would be treated under the no economic sense test, but they could in theory result in a finding of exclusionary conduct. (Again, however, one would hope that common sense would prevail in this instance.) A further problem is that the no economic sense test involves an assessment of the range of options that were open to the company at the time it embarked on a particular course of conduct. In most cases, the best evidence of a company’s options will be its business plans. Assessing whether a course of conduct made sense only if competitors were eliminated on the basis of such plans is extremely difficult in practice. For example, assume that a firm enters a new market in which initial capital costs are very high, and assume that it needs to acquire scale and scope economies and learning experience to reduce costs and achieve profitability, i.e., the firm needs to acquire volume. There is no effective way in this scenario of distinguishing between volume growth that is justified in itself and volume growth that is predicated, in whole or in part, on the elimination of a rival. It may be that, in this instance, the firm would pass the no economic sense test on the basis that there was a reasonably 29 30

Brief for the United States and Federal Trade Commission, supra note 26. Ibid.

(P) Ehlermann 08 O’Donoghue

338

6/11/08

11:51

Page 338

Session Two: A Reformed Approach to Exclusionary Conduct

anticipated non-exclusionary reason for its strategy (even if that turned out to be wrong). But this is not obvious and, even if it were, this approach runs the risk of being under-inclusive by allowing exclusionary conduct in the vital early stages of a new market to go unchecked. Of course, these types of cases are difficult under any test, but the point is that, in this regard, the no economic sense test does not appear to have any unique advantages over the profit sacrifice test, or any of the others discussed here.

B. Equally Efficient Competitor Test Elements. Exclusionary conduct has also been defined as conduct that would exclude an equally efficient rival firm. This definition was originally proposed by Judge Posner in his seminal book, Antitrust Law. The latest edition of the book offers the following definition of exclusionary conduct:31 “[T]he plaintiff must first prove that the defendant has monopoly power and second that the challenged practice is likely in the circumstances to exclude from the defendant’s market an equally or more efficient competitor. The defendant can rebut by proving that although it is a monopolist and the challenged practice is exclusionary, the practice is, on balance, efficient . . . [P]ractices that will only exclude less efficient firms, such as a monopolist’s dropping his price nearer to (but not below) its costs, are not actionable, because we want to encourage efficiency. Only when monopoly power is used to discourage equally or more efficient firms and thus perpetuate a monopoly not supported by superior efficiency should the law step in. Even then, it should be alert to the possibility that the exclusionary effect of the monopolist’s practice is offset by efficiency gains.”

The equally efficient competitor test certainly has some basis under Article 82. For example, the AKZO predatory pricing rules are grounded in the economic insight that a profit-maximizing dominant firm should be allowed to price down to the level of its average variable costs. This applies even if the dominant firm’s costs are lower than those of rivals. Similarly, the test usually applied in margin squeeze cases—whether the dominant firm’s own downstream arm could trade profitably if it had to pay the same input prices as third parties—relies on an equally efficient competitor test. Indeed, in Bronner, Advocate General Jacobs made clear that “the primary purpose of Article [82] is to prevent distortion of competition—and in particular to safeguard the interests of consumers—rather than to protect the position of particular competitors”.32 Consumers are generally best served by the most efficient firms, i.e., those with the lowest costs. 31

Posner, supra note 18, pp. 194–96. See Opinion of Advocate General Jacobs in Case C-7/97, Oscar Bronner GmbH & Co KG v Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co KG, Mediaprint Zeitungsvertriebsgesellschaft mbH & Co KG and Mediaprint Anzeigengesellschaft mbH & Co KG [1998] ECR I-7791, para. 58. 32

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 339

A General Test for Exclusionary Conduct under Article 82 EC 339 Criticisms of the equally efficient competitor test. A number of criticisms can be made of the equally efficient competitor test. First, examining whether a firm with similar levels of efficiency compared to the dominant firm will generally be of limited relevance where firms compete on the basis of differentiated products. In this circumstance, the products’ perceived qualities will typically be a more important parameter of competition than relative efficiency levels. Second, less efficient competitors can, in theory, enhance consumer welfare when the increased competition they bring in the market benefits consumers more than the cost of their relative inefficiency.33 For this reason, the duties imposed on dominant firms under Article 82 are not limited to equally efficient competitors, but, exceptionally, may include duties towards less efficient firms. Under the CEWAL line of case law, unconditional price cuts that remain above the dominant firm’s average total costs may be abusive in certain circumstances. Rules seeking to place restrictions on unconditional above-cost price cuts can be criticized, mainly on the grounds that they are in practice likely to chill desirable competition, but it is undeniable that a number of existing rules under Article 82 assume that less efficient firms can confer a net benefit on consumer welfare. This suggests that the “equally efficient rival” test could not be unreservedly accepted under Article 82. A third problem concerns the definition of equal efficiency where the dominant firm has a first mover or some other cost advantage over new entrants, or where rivals have not yet reached their minimum efficient scale. For example, in the area of conditional above-cost pricing schemes (e.g., loyalty rebates), much of the current uncertainty in the law stems from how to treat economies of scale for purposes of defining an “equally efficient firm”. The objection in such cases is usually that the dominant firm’s large volume of past sales gives it a scale or scope advantage over rivals and that, by extending this advantage to marginal units and customers, the dominant firm can sometimes offer prices at the margin that a rival only competing for the marginal units cannot match. While there is some merit in the view that only the most efficient firm should serve a customer, the Commission has taken the opposite approach in several cases. Fourth, for certain types of abuses, the concept of equal efficiency is of limited use when assessing the legality of conduct. For example, in the case of false declarations by a dominant firm to regulatory approval agencies, or concealment of essential patents within the context of standard setting organizations, rivals’ relative efficiency will be of little relevance if the action in question materially limits their access to the market. Of course, a less efficient firm is, all things being equal, likely to be more adversely affected by conduct of this kind by a dominant firm than an equally efficient one, but this issue 33 See, e.g., Armstrong and Vickers, supra note 8. See also Aaron Edlin, “Stopping AboveCost Predatory Pricing”, 111 Yale Law Journal 941 (2002).

(P) Ehlermann 08 O’Donoghue

340

6/11/08

11:51

Page 340

Session Two: A Reformed Approach to Exclusionary Conduct

goes more to the effects of the practice rather than the definition of operational rules as to when certain conduct is abusive or not. Finally, the equally efficient competitor test may require complex balancing exercises in some cases where the conduct would harm an equally efficient firm, but generates efficiencies sufficient to offset this harm. Of course, efficiencies, if asserted, would also need to be assessed under any other test used for exclusionary abuses, but the point is that the outcome of the equally efficient competitor test may not always be predictable by a firm at the time when it embarks on a particular course of action.34

C. Consumer Welfare Test Elements. The third test seeks to shift the focus away from the economic motivation for the alleged exclusionary conduct, and from the relative efficiency of competitors, towards an assessment of whether the dominant firm’s practices had, or are likely to have, a material adverse effect on consumer welfare.35 Under this test, exclusionary conduct would violate Article 82 if “it reduces competition without creating a sufficient improvement in performance to fully offset these potential adverse effects on prices and thereby prevent consumer harm”.36 In other words, only conduct that produces net anticompetitive effects would be regarded as exclusionary. An analysis of whether the conduct causes net harm to consumer welfare would take account of all available information relevant to the likely effects of conduct on consumers. The most relevant evidence consists of output and 34 Another efficiency-based test, proposed by Professor Elhauge, focuses on whether the alleged exclusionary conduct increases the firm’s dominance because it enhances its own efficiency or only because it limits rivals’ production. As he explains: “The proper monopolization standard should instead focus on whether the alleged exclusionary conduct succeeds in furthering monopoly power (1) only if the monopolist has improved its own efficiency or (2) by impairing rival efficiency whether or not it enhances monopolist efficiency . . . which would permit the former conduct and prohibit the latter.” Elhauge, supra note 3, at 253. This test certainly has certain advantages over some of the other tests. In particular, it avoids complex ex post balancing acts where conduct is both exclusionary and efficient. It also treats efficiency-enhancing conduct by a dominant firm as presumptively lawful, whereas in certain situations under the profit sacrifice test it might not be (e.g., long-term investments in intellectual property or intangible assets). Vague phrases such as “competition on the merits” and “normal competition” are also avoided under this test. But use of this test as an operational rule remains untested in practice. It would still need to be decided whether conduct is efficiency-enhancing or efficiency-reducing, which is often complex. Moreover, the articulation of the test by Elhauge focuses mainly on refusals to deal where the tradeoff between short-term static efficiency and long-term dynamic efficiency is probably clearer (i.e., with dynamic efficiency generally being considered more important). It is not clear how it would apply in other situations where the tradeoffs to be made are less capable of measurement. In any event, Elhauge’s test has been endorsed by the OECD: see “Competition on the Merits”, supra note 5, para. 65. 35 See the references to Gavil, Salop and Dolmans, supra note 19. 36 Salop, supra note 19.

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 341

A General Test for Exclusionary Conduct under Article 82 EC 341 prices, but quality and innovation may also play a role. There is some divergence among proponents of the consumer welfare test regarding whether efficiencies should be assessed on the basis of whether they simply outweigh any inefficiencies,37 or whether they should be subject to a more detailed proportionality inquiry.38 The latter has two elements. First, it would have to be shown that any harm caused by the dominant firm’s conduct is necessary to achieve the overall efficiencies. Second, it would have to be assessed whether the harm caused to competition is disproportionate when compared to any benefits that it brings. The consumer welfare test certainly has some pedigree in the case law in Europe and elsewhere. In the various Microsoft proceedings, the US Court of Appeals and the Commission each essentially applied a consumer harm standard to the various practices alleged.39 The Court of Appeals elaborated a multi-stage analysis of the various alleged exclusionary practices. First, the plaintiff has to show that consumers would be harmed. Second, if such harm is shown, the defendant may offer a pro-competitive justification for its conduct. Third, the pro-competitive justification can be rebutted by the plaintiff, or its positive impact on consumers can be shown to be outweighed by its negative effects on consumers.40 On the facts, the Court performed little actual balancing, since it was clear, on the evidence, that most of Microsoft’s conduct was overwhelmingly anticompetitive or pro-competitive. For example, regarding the claim that Microsoft had deceived Java developers about the Windows-specific nature of the tools, the Court noted that no efficiency justification had been advanced by Microsoft. A more explicit balancing exercise was undertaken by the Commission in the European Microsoft case. The Commission found that, in relation to the tying of Windows Media Player (WMP) with the Windows operating system (OS), Microsoft had “not submitted adequate evidence to the effect that tying WMP is objectively justified by pro-competitive effects which would outweigh the distortion of competition caused by it . . . [W]hat Microsoft presents as the benefits of tying could be achieved in the absence of Microsoft tying WMP with Windows.”41 In particular, the Commission found that: (1) ease of use could be achieved without tying (OEMs could do the bundling at no cost to Microsoft);42 (2) distribution efficiencies were minor and did not

37

See Gavil, supra note 19. See Phillip E. Areeda and Herbert Hovenkamp, Antitrust Law, 2nd ed., Aspen Publishers, 2002, para. 651a. 39 See United States v Microsoft, 253 F.3d 34, 59 (D.C. Cir. 2001); Commission’s Decision of 24 March 2004, Microsoft, Case COMP/37.792 (summary published in [2007] OJ L32/23), upheld: Case T-201/04, Microsoft v Commission, judgment of the Court of First Instance of 17 September 2007, not yet reported as of this writing. 40 Microsoft, 253 F.3d at 59. 41 Case COMP/37.792, Microsoft, supra note 39, para. 970. 42 Ibid. 38

(P) Ehlermann 08 O’Donoghue

342

6/11/08

11:51

Page 342

Session Two: A Reformed Approach to Exclusionary Conduct

outweigh the distortion of competition;43 (3) there was no evidence of technically superior performance due to code integration of WMP with the OS;44 and (4) platform efficiency (i.e., the desire to keep applications focused on Microsoft interfaces) was not a recognized efficiency.45 More generally, the consumer welfare test, if accepted, would unify the principles concerning mergers and other agreements with those under Article 82. Cooperative agreements between firms that restrict competition are subject to an express balancing act under Article 81(3), including an assessment of whether the anticompetitive effects are necessary and proportionate to achieve the alleged efficiencies. The Commission’s Guidelines on Article 81 provide that, “for an agreement to be restrictive by effect it must affect actual or potential competition to such an extent that on the relevant market negative effects on prices, output, innovation or the variety or quality of goods and services can be expected within a reasonable degree of probability”.46 Similarly, efficiencies under EC merger control are subject to the condition that they: (1) benefit consumers; (2) result from the merger; and (3) are verifiable. A sliding scale is also applied, i.e., mergers with the greatest scope for causing consumer harm also require the most compelling evidence of counterbalancing efficiencies.47 Criticism of the consumer welfare test. The consumer welfare test presents a number of difficulties. While balancing pro-competitive and anticompetitive effects may be appropriate when firms choose to conclude an agreement—in particular when, as under the merger control rules, that agreement is subject to mandatory prior approval—judging unilateral conduct in the same way is precarious and might lead to haphazard outcomes. A firm embarking on a course of unilateral conduct ex ante may be unsure as to where the balance between pro-competitive and anticompetitive aspects lies and when such effects will materialize. Much would depend on the effect of a practice on the dominant firm’s rivals, which the dominant firm cannot generally be expected to know. Moreover, what a firm expects ex ante may of course turn out to be different from what occurs ex post. These problems are most likely to be acute in markets in which technology evolves rapidly and where new entry is a strong feature, since actual market outcomes may differ materially from many firms’ expectations. Proponents of the consumer harm test have responded to these criticisms with several clarifications. First, they argue that any balancing would not turn courts and competition authorities into central planners, i.e., comparing the harm to consumer welfare with the benefits to producer welfare. Second, they 43

Ibid., paras. 956 et seq. Ibid., para. 958. 45 Ibid., para. 962. 46 See Guidelines on the application of Article 81(3) of the Treaty, [2004] OJ C101/97, para. 24. 47 See Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, [2004] OJ C31/5, paras. 80–86. 44

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 343

A General Test for Exclusionary Conduct under Article 82 EC 343 say that courts and competition authorities would not be required to apply sophisticated quantitative techniques to measure the probability and weight of certain effects. Instead, they would apply a “preponderance of evidence” approach to determine whether the benefits and harm are each proven by the evidence and, if so, to compare which is greater. Third, proponents argue that firms would be judged not ex post, but by reference to the types of effects that were reasonably foreseeable ex ante, even if they turned out to be wrong. Finally, issues of uncertainty are said to be exaggerated. Most cases, they say, can be resolved at an early stage without the need for complex balancing exercises, either because there is no material harm to consumers or because it is clear that the conduct in question is overwhelmingly harmful or beneficial. But even with these clarifications, it is clear that the consumer harm test would likely present significant complexities in many cases, and it certainly would not be easy for a dominant firm to apply ex ante.

D. Limiting Production to the Prejudice of Consumers Elements. Article 82(b) states that, where a firm is dominant, “limiting production, markets or technical development to the prejudice of consumers” is illegal. Unlawful foreclosure or handicapping of competitors, by which competition is reduced still further, “limit production” in this sense. Limiting production is the most frequent and important category of abuse in practice, since it broadly covers any type of exclusionary conduct that limits rivals’ possibilities and causes harm to consumers. And just as firms can compete in a myriad of ways, so too they can seek to exclude rivals through a multiplicity of strategies. The vast majority of infringement decisions under Article 82 have concerned exclusionary abuses and, therefore, Article 82(b). Article 82(b) captures the key feature of exclusionary conduct, namely, that it makes competitors’ products or services less attractive or less available, rather than simply making the dominant company’s product better or more available. Offering better or cheaper products must generally be legal, however great the difficulties it causes to rivals. In contrast, creating difficulties for competitors in other ways is not. In basic terms, the application of these principles to the most common pricing and non-pricing abuses is relatively straightforward. The bare wording of Article 82(b) does not of course solve every problem or question, but it at least provides some satisfactory underlying principles, as well as consistency. Take the example of dominant patent holders in the area of pharmaceuticals impeding generic rivals’ market access through changing product registration documents. Under generic substitution legislation in the EU, less expensive generic drugs can (and, in some cases, must) automatically be substituted for branded equivalents. This rule, however, does not necessarily

(P) Ehlermann 08 O’Donoghue

344

6/11/08

11:51

Page 344

Session Two: A Reformed Approach to Exclusionary Conduct

apply to generics of different dosages, different formulations (e.g., substitution of instant for continuous release), or different delivery systems (e.g., capsules for tablets). Observers have noted that by withdrawing a particular version of a branded drug, for which there are approved generic substitutes, in favour of another version, for which there are not, the extent of generic substitution can be decreased (or eliminated), thereby extending the life cycle of the branded monopoly drug. The withdrawal of a branded drug which a generic applicant has referred to and relied on in its own application for approval would potentially extend the life cycle further (i.e., in the absence of data on which to rely, the generic manufacturer would need to conduct its own clinical trials albeit with a decreased incentive to do so). That kind of fact pattern would be analyzed as follows under the “limiting production” test: • The first step in analyzing the conduct is to assess its impact on generic competition, i.e., whether in fact there is a prima facie case of “limiting production”. If the impact of the conduct is to seriously restrict the scope of generic competition, then the focus turns to the evaluation of the pro-competitive justifications for the conduct. A number of pro-competitive justifications can be imagined (e.g., safety/quality concerns). • Where the issue concerns the replacement of an existing product with a new dosage or method of administration, the second step is to consider whether the new replacement product is advantageous relative to the replaced product, i.e., whether there is a prima facie case of “prejudice to consumers”. If there is no such advantage, it might be presumed that the only reason for the introduction of the new version and the discontinuation of the old version is to limit generic competition. Conduct of this kind is likely to be regarded as anticompetitive. • Assuming that the new formulation has advantages relative to the original product, the focus then shifts to whether there are legitimate reasons why the branded company would not offer both the new and the old product, thereby allowing the new product to compete on the merits with generic versions of the original product. Among these might be diseconomies of scope in producing the two products, including manufacturing diseconomies, additional inventory costs, etc. Arguments may be raised that the limited withdrawal of the old product (i.e., withdrawal from only certain markets) suggests that anticompetitive motives rather than diseconomies are driving the decision. • Fourth, in assessing whether such reasons are sufficient to justify eliminating the original product, a useful starting point is to ask whether the same decision would have been taken in the absence of generic competition. If the answer is yes, then it can be presumed that the costs of supporting two products are not outweighed by the consumer benefits measured as the revenue from the continued sale of the original product, thereby providing a

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 345

A General Test for Exclusionary Conduct under Article 82 EC 345 legitimate business justification for its discontinuance. If the answer is no, then the question becomes whether the volume of the original product sold in the face of generic competition and competition from the new formulation is so low as to make it uneconomical to continue to offer it. That is, whether standing alone it can be demonstrated that the original product’s revenue does not justify its continued production and sale. (It is certainly a legitimate business justification to discontinue a product if the product is losing money.) If, on the other hand, but for the effect on generic competition, it would have been profitable for the branded company to continue offering the original product alongside the new formulation, discontinuance of that product will likely be found on balance to be anticompetitive. Evidence that both products are sold profitably in other countries might tend to contradict an assertion that it is unprofitable to continue to sell both products. • Finally, in addition to objective economic evidence, courts and the antitrust enforcement agencies will also consider evidence of the subjective intent of the branded company in pulling the version of the product subject to generic competition. Thus, if there are internal memoranda, correspondence, e-mails etc. which suggest that the motivation for the decision to pull the original product is to reduce generic competition, these documents will be considered good evidence that the conduct is on balance anticompetitive notwithstanding any proffered legitimate business justifications. Provided the evidence is reasonably clear and unambiguous, it should inform the assessment of the motivation for the withdrawal or amendment of the registration documentation. Criticisms. The “limiting production” test can be criticized as lacking the necessary specificity for it to be useful as an operational rule. The test still begs the question of when “limiting” is legal and when it is not. The added requirement that “limiting” is only prohibited if it causes prejudice to consumers is a vital clarification, but it still rather begs the question. However, the other proposed tests are equally beset by the same problems and, arguably, even more so. A second problem is that there are various exceptions to the rule that limiting rivals’ production is legal unless consumers are also harmed. Normally, an above cost price is legal since any “limitation” is the result of greater efficiency, i.e., lower costs. But as noted earlier, the Community Courts have held in CEWAL that above-cost price cuts may be abusive in certain circumstances. Again, however, these criticisms are probably more accurately directed at the application of the case law rather than the underlying rationale of the limiting production test. Indeed, the logic of the limiting production test is that preventing a dominant firm from discounting down to its marginal cost of production (or some proxy) would itself cause “prejudice to consumers”.

(P) Ehlermann 08 O’Donoghue

346

6/11/08

11:51

Page 346

Session Two: A Reformed Approach to Exclusionary Conduct

D. Efficiency Defences Role in Article 82 cases. A central issue arising from the application of the various proposed tests for exclusionary conduct concerns the role of efficiencies in the assessment of conduct with a mixture of exclusionary and efficiency effects. Failure to pass the “no economic sense” test implicitly accepts that the conduct lacks a business justification other than exclusion. Equally, the consumer welfare test requires consideration of whether the harm caused by a practice is offset by efficiencies. Proponents of the equally-efficient competitor test also accept that conduct which harms equally-efficient competitors may have offsetting efficiency benefits. Finally, the “limiting production” test has an express limb concerning harm to consumers. In sum, the role of efficiencies is directly or indirectly relevant to the principal exclusionary abuse tests that have been proposed. Although Article 82 does not contain an exemption clause similar to Article 81(3), it is well established that “objective justification” can immunize conduct that would otherwise be an abuse under Article 82.48 Objective justification has a number of different facets under Article 82: (1) situations in which the dominant firm’s conduct is objectively necessary because of factors external to the dominant firm’s conduct (e.g., safety concerns); (2) situations in which the dominant firm takes defensive measures to protect its commercial interests (e.g., meeting competition); and (3) situations in which the dominant firm’s conduct is justified by efficiencies. Category (3) is focused on here since it is the most complex and important in practice. At least in theory, it is accepted that conduct that would otherwise be regarded as exclusionary may be justified by the fact that it creates offsetting efficiency benefits. In the area of predatory pricing, price cuts may be introduced to increase demand for the dominant firm’s products. For example, short-term promotional offers are intended to allow consumers to become familiar with a product in the hope that, when they do, they will, recognizing the product’s quality, pay a higher price once the promotional phase ends. Loss-leading has a similar rationale in that price cuts on certain products are offered in order to increase demand for other (complementary) products. 48 See, e.g., Case 40/70, Sirena Srl v Eda Srl and others [1971] ECR 69, para. 17; Case 24/67, Parke, Davis and Co v Probel, Reese, Beintema-Interpharm and Centrafarm [1968] ECR 55; Case 78/70, Deutsche Grammophon Gesellschaft mbH v Metro-SB-Großmärkte GmbH & Co KG [1971] ECR 487; Case 395/87, Ministère public v Jean-Louis Tournier [1989] ECR 2521; Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207; Case 77/77, Benzine en Petroleum Handelsmaatschappij BV and others v Commission [1978] ECR 1513, paras. 33–34; Case 311/84, Centre belge d’études de marché/Télémarketing (CBEM) v SA Compagnie luxembourgeoise de télédiffusion (CLT) and Information publicité Benelux (IPB) [1985] ECR 3261; Magill TV Guide/ITP, BBC and RTE, [1989] OJ L78/43; Eurofix-Bauco v Hilti, [1988] OJ L65/19; Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755; Case T-228/97, Irish Sugar plc v Commission [1999] ECR II-2969; Case C-163/99, Portugal v Commission [2001] ECR I-2613.

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 347

A General Test for Exclusionary Conduct under Article 82 EC 347 Price cuts may also be intended to expand the market or increase efficiency in other ways, such as by acquiring learning experience to reduce costs over time or by creating network effects. A slightly different defence is that, in situations of excess capacity, the maximum market price may not exceed any firm’s relevant costs, in which case the least inefficient option is to sell below cost for a period until the market corrects itself.49 Objective justification is also a central issue in the case of exclusive dealing, loyalty rebates, and other vertical restraints. Such restraints may be motivated by the need to recover fixed costs more efficiently, the need to provide optimal incentives to retailers, and the need to ensure that customer-specific investments by the dominant firm are adequately protected.50 Many of the same justifications underpin price discrimination by a dominant firm, which should generally be presumed to be efficient where it expands output more than in situations in which uniform prices apply.51 The same comment can be made in respect of tying/bundling, which is almost always motivated by some efficiency considerations (though it may create strong foreclosure too). In essence, all of these defences seek to put forward explanations of why the conduct in question is efficient or justified by some legitimate consideration other than the dominant firm’s interest in excluding competitors. Conditions for an efficiency defence as per the Discussion Paper. Proof of efficiencies requires a number of different steps, now outlined in detail in DG Competition’s Discussion Paper on Article 82.52 According to this document, efficiencies may lead to a finding that Article 82 does not apply if four cumulative conditions are satisfied. These conditions essentially mirror those applicable under Article 81(3). First, the dominant firm must show that the conduct was undertaken to improve the production or distribution of products or to promote technical or economic progress (e.g., by improving the quality of its product or by obtaining specific cost reductions), or to generate another efficiency. Such claims must be substantiated, though the evidence may be less demanding in the case of qualitative efficiencies such as improvements in distribution. Second, the dominant firm must show that the conduct is indispensable to achieve the alleged efficiencies. It is for the dominant company to demonstrate that there are no other economically practicable and less anticompetitive alternatives to achieve the claimed efficiencies, taking into account the market conditions and business realities facing the dominant company. The dominant company is not required to consider hypothetical or theoretical alternatives. The Commission will only contest the claim where it is reasonably clear that there are realistic and attainable alternatives, when viewed in the overall 49

See O’Donoghue and Padilla, supra note 2, Chapter 5, section 5.6. Ibid., Chapter 7. 51 Ibid., Chapter 11. 52 See DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf, paras. 84 et seq. 50

(P) Ehlermann 08 O’Donoghue

348

6/11/08

11:51

Page 348

Session Two: A Reformed Approach to Exclusionary Conduct

context of the dominant firm’s conduct and the market realities faced by the dominant firm at the time it made the relevant decision. The dominant company must explain and demonstrate why seemingly realistic and less restrictive alternatives would be significantly less efficient. Third, the dominant company needs to show that efficiencies brought about by the conduct concerned outweigh the likely negative effects on competition and in particular the likely harm to consumers that the conduct might otherwise have. This will be the case when the Commission, on the basis of sufficient evidence, is in a position to conclude that the efficiencies generated by the conduct are likely to enhance the ability and incentive of the dominant company to act pro-competitively for the benefit of consumers. The fourth condition makes clear that it is not sufficient for the efficiencies to outweigh the negative effects on competition: on balance, consumers must also benefit from the conduct concerned. This reflects the consideration that Article 82 protects competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources. This requires that the pass-on of benefits must at least compensate consumers for any actual or likely negative impact caused to them by the conduct concerned. If consumers in an affected relevant market are worse off following the prima facie abusive conduct, the conduct cannot be justified on efficiency grounds. In making this assessment, the Commission states that it must be taken into account that the value of a gain for consumers in the future is not worth the same as a present gain for consumers. In general, the later the efficiencies are expected to materialize in the future, the less weight the Commission or national authorities can assign to them. This implies that, to be considered as a countervailing factor, the efficiencies must arise in the short-term. The incentive on the part of the dominant company to pass efficiency gains on to consumers is often related to the existence of competitive pressure from the remaining firms in the market and from potential entry. This incentive may often be already small as a result of the dominant position. The greater the actual or likely negative effects on competition, the more the Commission or national authorities have to be sure that the claimed efficiencies are substantial, likely to be realized, and to be passed on, to a sufficient degree, to the consumer. It is therefore highly unlikely that prima facie abusive conduct of a dominant company with a market position approaching that of a monopoly, or with a similar level of market power, can be justified on the ground that efficiency gains would be sufficient to outweigh its actual or likely anticompetitive effects and would benefit consumers. Similarly, in a market where demand is very inelastic it is highly unlikely that abusive conduct of a dominant company strengthening its dominant position can be justified on the ground that efficiency gains would be sufficient to counteract the actual or likely anticompetitive effects and would benefit consumers. The final condition is that competition in respect of a substantial part of the products concerned must not be eliminated. The Discussion Paper states that

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 349

A General Test for Exclusionary Conduct under Article 82 EC 349 when competition is eliminated the competitive process is brought to an end and short-term efficiency gains are outweighed by longer-term losses stemming, inter alia, from expenditures incurred by the dominant company to maintain its position (rent seeking), misallocation of resources, reduced innovation, and higher prices. This recognizes the fact that rivalry between undertakings is an essential driver of economic efficiency, including dynamic efficiencies in the shape of innovation. Ultimately, according to the Commission, the protection of rivalry and the competitive process is given priority over possible pro-competitive efficiency gains. Thus, it is highly unlikely that abusive conduct of a dominant company with a market position approaching that of a monopoly could be justified on the ground that efficiency gains would be sufficient to counteract its actual or likely anticompetitive effects. This accords with the sliding scale applied to efficiencies under EC merger control rules. By way of guidance, the Discussion Paper suggests that a 75% market share would substantially eliminate competition in circumstances where there is no effective competition from other actual competitors in the market (e.g., because they have higher costs or because they face capacity constraints). Evaluating the approach to efficiency defences under Article 82 EC. The explicit recognition of a potential efficiency defence under Article 82 is clearly welcome, since a good deal of unilateral conduct will have a mixture of positive and negative effects on consumers. But significant difficulties remain. A first point is that, despite the general recognition of objective justification in the case law, there are very few cases under Article 82 in which objective justification has actually been accepted by courts and competition authorities. This may reflect the fact that the basis for the defence put forward in several cases was not strong enough, but more likely suggests that there is something of a disconnect between theory and practice on objective justification. Efficiency defences have typically been rejected with cursory analysis by the Community institutions and without any indication of the analytical framework they have in mind—including most recently by the Court of First Instance in Wanadoo.53 This deficiency should be addressed, since a defence that is recognized in theory, but not in practice, is the same as no defence. 53 See Case T-340/03, France Télécom v Commission [2007] ECR II-107; on appeal: Case C-202/07, not yet decided. The case concerned predatory pricing. The CFI held that, in the case of pricing below average variable cost (AVC), the “only interest which the undertaking may have in applying such prices is that of eliminating competitors”. (para. 197). This formulation would appear to exclude the possibility that prices below AVC may be non-exclusionary/efficient in certain circumstances. Recent economics literature has raised the possibility that pricing below AVC may have valid business justification in certain circumstances. Reasons suggested include network effects and learning-by-doing efficiencies. In such cases, recovery of the investment in below-cost sales stems from efficiency-enhancing factors (e.g., higher product quality or lower cost) rather than from increased profits through eliminating or disciplining a rival. See Patrick Bolton, Joseph F. Brodley, and Michael H. Riordan, “Predatory Pricing: Strategic Theory and Legal Policy” 88 Georgetown Law Journal 2239 (2000).

(P) Ehlermann 08 O’Donoghue

350

6/11/08

11:51

Page 350

Session Two: A Reformed Approach to Exclusionary Conduct

A second point is that, although the theory of balancing pro-competitive and anticompetitive effects sounds straightforward, it is often anything but. In theory, the exercise is easy: the amount of the benefits (increased consumer welfare) is compared with welfare loss caused by the exclusion of rivals (e.g., reduced consumer surplus or deadweight loss). Whichever is larger determines the outcome. But in practice it may not be easy, or indeed possible, for a dominant firm to make such detailed assessments at the time it decides on its commercial strategy, in particular if this involves detailed knowledge of the effects of a particular practice on rivals and more generally on consumer welfare. One commentator summarizes these practical concerns as follows:54 “The problem, however, is that neither economic actors nor law enforcement entities are omniscient. Given real world limitations, market-wide balancing tests that seek to assess the benefits and competitive harms of exclusionary conduct are intractable for courts and antitrust agencies, and even more so for firms trying to decide in real time what conduct is permitted and what is prohibited. Prospective defendants cannot be expected to know in real time, ex ante, whether their efficiency-generating conduct will cause disproportionate harm to their rivals or consumers because, in order to know that, the defendants would have to know more than they can be expected to know about consumer demand, their rivals’ costs and prospects for innovation and for mitigation of harm, future entry conditions, and the like. From the perspective of the defendants, therefore, a balancing test would likely either be ignored, impose excessive transaction costs (a kind of tax on entrepreneurship), or result in excessive caution. There is little reason to expect that a balancing test would create optimal ex ante incentives for marketplace behaviour.”

Third, it is questionable whether it is correct in law to require the dominant firm to show that the efficiencies outweigh the anticompetitive effects. Article 2 of Regulation 1/2003 provides that “the burden of proving an infringement of . . . Article 82 . . . shall rest on the party or the authority alleging the infringement”. In Syfait, Advocate General Jacobs made clear that proof of objective justification means that “certain types of conduct on the part of a dominant undertaking do not fall within the category of abuse at all”.55 It is true that Article 2 of Regulation 1/2003 places the burden of proving the benefit of the conditions of Article 81(3) on the defendant, but it clearly does not apply the same principle to Article 82 and objective justification. In these circumstances, it should be for a plaintiff or competition authority to show that the anticompetitive effects outweigh the efficiencies, since otherwise no abuse has been proven. The dominant firm should simply bear the initial burden of producing colourable evidence to substantiate the efficiencies claimed 54 See A. Douglas Melamed, “Exclusionary Conduct Under the Antitrust Laws: Balancing, Sacrifice, and Refusals to Deal” 20 Berkeley Technology Law Journal 1249 (2005). 55 See Opinion of Advocate General Jacobs in Case C-53/03, Synetairismos Farmakopoion Aitolias & Akarnanias (Syfait) and Others v GlaxoSmithKline plc and GlaxoSmithKline AEVE [2005] ECR I-4609, para. 53.

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 351

A General Test for Exclusionary Conduct under Article 82 EC 351 (mainly because only it has access to such evidence).56 The legal burden should then shift to the plaintiff or competition authority. Fourth, a particular problem arises because the last condition of the efficiency defence under Article 82—i.e., the condition that the conduct must not “substantially eliminate competition”—is in practice likely to preclude the availability of the defence. Under Article 82, an efficiency defence would be raised in circumstances where, first, a firm is already dominant and, second, its conduct has been found to have an actual or likely exclusionary effect on competition. Although the Community institutions have made clear that dominance does not necessarily mean that competition is “substantially eliminated”,57 a finding of dominance and material foreclosure effect under Article 82 may, for practical purposes, mean that competition is substantially eliminated. But even then, the conduct in question may still enhance consumer welfare overall. In other words, there appears to be a logical contradiction between the substantive test for abuse and the availability of an efficiency defence: because of the addition of the “substantial elimination of competition” condition, the former seems to preclude the latter. It is not clear why the Discussion Paper proposes to transpose the conditions for exemption under Article 81(3) to the efficiency defence under Article 82. The most likely reason is consistency, but the Commission has very rarely exempted under Article 81 arrangements that created dominance (except, perhaps, in the case of revolutionary new technologies and markets). Fifth, many economists question whether balancing anticompetitive effects is meaningful where those effects are a function of a restriction that is indispensable to achieve the stated efficiency.58 For example, if exclusive dealing is indispensable to achieve an efficiency (e.g., to justify a customer-specific investment), the source of the anticompetitive concern is the same as the source of the efficiency. And yet, under the Discussion Paper’s fourth condition for an efficiency defence—no elimination of competition—an efficient

56 See Joined Cases C-204/00 P et al., Aalborg Portland A/S and Others v Commission [2004] ECR I-123, paras. 78–79 (“[I]t should be for the party or the authority alleging an infringement of the competition rules to prove the existence thereof and it should be for the undertaking or the association of undertakings invoking the benefit of a defence against a finding of an infringement to demonstrate that the conditions for applying such defence are satisfied, so that the authority will then have to resort to other evidence. Although according to those principles the legal burden of proof is borne either by the Commission or by the undertaking or association concerned the factual evidence on which a party relies may be of such a kind as to require the other party to provide an explanation or justification, failing which it is permissible to conclude that the burden of proof has been discharged.”). 57 See Commission Notice—Guidelines on the application of Article 81(3) of the Treaty, [2004] OJ C101/97, para. 106; Joined Cases T-191/98 and T-212/98 to T-214/98, Atlantic Container Line AB and Others v Commission [2003] ECR II-3275, para. 939 (“As the concept of eliminating competition is narrower than that of the existence or acquisition of a dominant position, an undertaking holding such a position is capable of benefiting from an exemption.”). 58 See RBB Economics, “Selective Price Cuts and Fidelity Rebates”, Economic Discussion paper prepared for the Office of Fair Trading (July 2005), paras. 4.146 et seq.

(P) Ehlermann 08 O’Donoghue

352

6/11/08

11:51

Page 352

Session Two: A Reformed Approach to Exclusionary Conduct

clause would probably be condemned in this instance. A related problem is the suggestion in the Discussion Paper that an efficiency defence will generally be unavailable at market shares in excess of 75%. It is not clear why this statement was added: either conduct is efficient or it is not. Choosing a 75% upper limit is therefore arbitrary and pointless. Thus, in formulating an efficiency defence under Article 82, the Commission would appear to have followed the wording of Article 81(3) for no good reason other than general notions of consistency (despite the wording of the two Articles being clearly (and in my view, deliberately, different). A sounder position is simply to apply a sliding scale (as occurs under merger control laws) whereby the greater the market power harm, the greater the need for offsetting consumer benefits. A final important point is that the scope and availability of an efficiency defence cannot be looked at in isolation from the substantive rules that apply for specific practices. For example, the non-discrimination clause in Article 82(c) has sometimes been applied in a mechanical fashion, with little regard for the many legitimate reasons why firms charge different prices or offer different terms. The effect of this rule may be to require dominant firms to put forward efficiency justifications for everyday business decisions. But most differences in prices or terms result from the relative skills and bargaining power of customers and do not have (or need) a formal efficiency justification beyond this. It makes no sense to require a dominant firm to provide an elaborate explanation for something as innocuous as other firms’ negotiation skills or bargaining power. The same could be said of conditional discount schemes. The broad (and arguably vague) rules historically applied by the Commission in respect of such practices—whether the discounts have a “fidelity-building” effect—require dominant firms to offer efficiency justifications for many practices that have a simple, obvious, and pro-competitive logic. Applying a detailed efficiency justification test in every instance to such practices is unnecessary and wrong.

IV. Assessment of the Various Tests for Exclusionary Conduct The various attempts at verbalizing a general test for exclusionary conduct deserve enormous credit. Most importantly, they all recognize a number of elementary, but fundamental, points. In the first place, they show the need for a unified definition of exclusionary conduct. A range of disparate tests that depend on how the conduct happened to be classified would be haphazard, intellectually incoherent, and would lead to differing outcomes. Second, these general tests all recognize the most fundamental point: that it is analytically wrong to equate harm to competitors with unlawful exclusion, since

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 353

A General Test for Exclusionary Conduct under Article 82 EC 353 legitimate competition has the same effect, but is positively encouraged. Thus, each test recognizes that unilateral conduct must have some inherent element of unlawful exclusion before it can even begin to be considered abusive. While unlawful exclusion and competition often look similar in appearance, the issue of disentangling them cannot be ducked. In particular, it is not simply a question of comparing whatever harm to rivals the conduct happens to cause against any offsetting benefits, since, as noted, this could impugn competition itself. Another point that emerges from the foregoing is that the principal tests are much closer in terms of their basic articulation of the problem and solution than their proponents appear to consider. (Proponents’ willingness to identify fundamental differences between the tests is in part a function of the usual dialectic that occurs in these situations: a proposed test will always be made to look more compelling by identifying “fatal” flaws in competing ideas.) In essence, each test seeks to identify conduct that harms rivals for reasons other than the dominant firm’s superior efficiency and also makes provision for the possibility that consumers can benefit from efficiencies that offset any exclusion. That said, some of the tests clearly fare better than others in terms of their usefulness. Although the equally-efficient rival test has a clear pedigree in certain Article 82 case law (particularly in predatory pricing and margin squeeze cases), a number of difficulties remain. First, the Community institutions have not consistently followed this test, even in pricing abuse cases. As noted, abovecost price cuts may be illegal under Article 82 in exceptional circumstances. The Article 82 Discussion Paper also clearly contradicts this general test in several places. First, it suggests that conduct that harms competitors who are not yet as efficient, but who might become so, could be regarded as abusive.59 Second, its treatment of above-cost conditional rebate practices clearly applies a test that is not based on an equally-efficient competitor, but seeks to modify that test by applying it only to a fraction of market output known as the “contestable” share.60 Third, the as-efficient competitor test is clearly less useful— and some would argue useless—where firms compete by offering differentiated products. Fourth, it is not clear how the equally-efficient competitor test should be applied where the dominant firm has activities in multiple related markets (creating economies of scope/common costs) whereas rivals are standalone producers of one product. In these circumstances, rivals may be as efficient in one market but much less efficient overall. As noted above, the Discussion Paper seeks to compensate for problems of this kind by assessing the as-efficient competitor test over a smaller fraction of market output. However, unfortunately this sounds like regulation. Finally, the equallyefficient competitor test is of limited use in the case of non-price abuses, except 59 60

Discussion Paper, supra note 52, para. 67. Ibid., paras. 153 et seq.

(P) Ehlermann 08 O’Donoghue

354

6/11/08

11:51

Page 354

Session Two: A Reformed Approach to Exclusionary Conduct

perhaps to assess possible anti competitive effects. There is some basis in economic theory for saying that less-efficient competitors can improve consumer welfare in certain circumstances. But applying an equally-efficient competitor test with so many possible exceptions may legitimately lead one to wonder whether the exception in fact would become the rule. The consumer harm test undoubtedly asks the correct theoretical question for assessing unilateral conduct: does it cause net harm to consumers? It also has some pedigree in the decisional practice and case law under Article 82— most recently Microsoft—and is consistent with the overall assessment of anticompetitive effects under Article 81 and the substantive analysis under EC merger control. But whether all unilateral conduct should be subject to such an overarching inquiry is questionable. What unilateral conduct a firm can engage in without violating the law should be subject to clear rules in all but exceptional cases, without the need to balance exclusionary effects against pro-competitive aspects. It is true of course that not all firms are dominant, so the point should not be exaggerated; however, it is routine, in my practical experience, for firms with market shares as low as 30% to assume as a matter of risk management that they could be regarded as dominant. Although proponents of the consumer harm test have made its operational features as useful as possible (and made clear that the assessment should focus on the situation ex ante, not ex post), complex and precarious balancing acts are still likely to be necessary in marginal cases, where the cost of error is also likely to be high. It also bears emphasis that the vast majority of enforcement is not done in courts or before competition agencies, but in a private sphere where business firms and their advisors scrutinize everyday practices for compatibility with competition laws. My experience in private practice for the last decade or so gives me very little confidence that a consumer welfare test can be easily applied by firms in real time. Moreover, if issues of proportionality come into play, economics contributes very little by way of predictability and the outcomes will represent matters of policy rather than precision. Of course it might be argued that much the same exercise is sometimes conducted under Article 81 and EC merger control. But cases involving agreements are different in the sense that the firms can always choose not to conclude an agreement, or to structure a different one, or to amend some aspect of their agreement to comply with objections under competition law (e.g., by offering a merger remedy). The firms are also much more likely to have detailed knowledge of the effect of an agreement on their output and to be able to quantify the synergies created by cooperation. Indeed, this is the primary motivation for merger activity and forms of horizontal cooperation. The same cannot generally be said of most unilateral conduct. The quantitative techniques used in merger control to assess the price effects of a merger are much more difficult to apply to the unilateral conduct of a single firm and, again, it is reasonable to ask whether firms making everyday business decisions should be subject to such burdens. Finally, the counterfactual compar-

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 355

A General Test for Exclusionary Conduct under Article 82 EC 355 ison of the state of competition in the absence of the agreement is also generally clearer in the case of agreements because the agreement will coordinate the activities of firms that previously had independent activities in respect of the matters subject to the agreement. Both the profit sacrifice and the no economic sense tests are useful in that they seek to move the debate on abusive conduct away from a subjective assessment of what is competition on the merits towards a more objective measure of whether conduct is profit maximizing or economically rational but for its ability to exclude. But the criticisms of both tests are compelling, and several fundamental conceptual questions remain unanswered in respect of these tests. First, it is not clear how the tests would apply where the sacrifice and recoupment occur simultaneously. For example, a margin squeeze does not imply a direct loss on each unit sold as in the case of predation, since a loss of sales by raising the upstream price may be compensated for by downstream volume gains. It is of course true that there is an opportunity cost on each unit not sold to the downstream competitor (by raising the input price, the dominant firm may reduce the size of the downstream market and therefore sell less upstream output), but it is not clear how this would fare under profit sacrifice/no economic sense. Second, in monopoly maintenance cases, there may be no overall profit sacrifice/increase, leading to a false negative finding that no exclusionary conduct has taken place.61 Third, there are good reasons to believe that the profit sacrifice/no economic sense tests would also be difficult to implement in practice because they require a comparison against a market in which the dominant firm had no power over price (or some other but-for benchmark). The experience of the US courts applying a similar test in the American Airlines case shows the complexities involved (although some of these are a function of the complex nature of hub and spoke versus route pair competition in the airline sector). Indeed, some commentators have suggested that it is not even clear whether the sacrifice test is a single unified substantive standard for assessing all exclusionary conduct or simply a more objective measure of the defendant’s intent, or the likely effects of a practice. As Sir John Vickers notes, “while the sacrifice test might be useful in assessing wilfulness or intent, it does not naturally yield a substantive standard of what behaviour is exclusionary. There is no escape from the fundamental question of what is [exclusionary]”.62 In other words, the sacrifice test may, at best, constitute a useful characterization of certain types of abuses—in particular, pricing abuses—it is not, in itself, capable of identifying exclusionary conduct and clearly distinguishing it from legitimate conduct. At the very least, however, it is reasonable to assume that the profit sacrifice/no economic sense tests

61 62

See Salop, supra note 19. See Vickers, supra note 5.

(P) Ehlermann 08 O’Donoghue

356

6/11/08

11:51

Page 356

Session Two: A Reformed Approach to Exclusionary Conduct

remain tied to a number of open questions that would likely further complicate Article 82 in the short to medium term. This leads to the conclusion that the “limiting production” test under Article 82(b) is probably the best test overall. First, it is expressly based, as it should be, on the words of the EC Treaty, and does not require the Commission or the Community Courts to invent new tests not stated anywhere in the Treaty. Second, it captures the two fundamental insights of any sensible definition of exclusionary conduct. The first is that it ensures that only conduct that results in output limitation (or “limiting production”) is considered exclusionary. All exclusionary conduct results in the limitation of either the dominant firm’s production, or, more likely, that of competitors (either because rivals are forced to exit the market or remain in the market but face marginalization due to increases in their costs caused by the dominant firm’s strategic actions). Although the Community Courts do not always refer to specific clauses of Article 82 in their judgments, multiple judgments have confirmed that Article 82(b) captures both types of limitation, i.e., it prohibits a dominant enterprise from limiting the production, marketing or development of its competitors, as well as its own.63 The Commission has also applied Article 82(b) in its seminal decision in Microsoft,64 where Microsoft’s conduct was characterized as limiting innovation to the prejudice of consumers. The second key insight captured in Article 82(b) for defining exclusionary conduct is that the only output limitation of interest under Article 82 is that which causes prejudice to consumers. Article 82(b) makes it clear that a dominant company may limit its rivals’ 63 See, e.g., Joined Cases 40 to 48, 50, 54 to 56, 111, 113 and 114–73, Coöperatieve Vereniging “Suiker Unie” UA and others v Commission [1975] ECR 1663, paras. 399, 482–83, and in particular paras. 523–27; Case 41/83, Italy v Commission (British Telecommunications) [1985] ECR 873; Case 311/84, Centre belge d’études de marché—Télémarketing (CBEM) v SA Compagnie luxembourgeoise de télédiffusion (CLT) and Information publicité Benelux (IPB) [1985] ECR 3261, para. 26; Case 53/87, Consorzio italiano della componentistica di ricambio per autoveicoli and Maxicar v Régie nationale des usines Renault [1988] ECR 6039; Case 238/87, AB Volvo v Erik Veng (UK) Ltd [1988] ECR 6211; Joined Cases C-241/91 P, Radio Telefís Éireann (RTE) and Independent Television Publications Ltd (ITP) v Commission (Magill) [1995] ECR I-743, para. 54; Case C-41/90, Klaus Höfner and Fritz Elsner v Macrotron GmbH [1991] ECR I-1979, para. 30; Case C-55/96, Job Centre coop arl [1997] ECR I-7119, paras. 31–36; and Case C-258/98, Giovanni Carra and Others [2000] ECR I-4217. For commentary, see John Temple Lang, “Abuse of Dominant Positions in European Community Law, Present and Future: Some Aspects”, in Barry Hawk, ed., Fifth Fordham Corporate Law Institute, Law & Business, 1979, pp. 52, 60; John Temple Lang, “Anticompetitive Non-Pricing Abuses Under European and National Antitrust Law”, in Barry Hawk, ed., 2003 Fordham Corporate Law Institute, Juris Publication Inc., 2004, pp. 235–340; Christopher Bellamy and Graham Child, European Community Law of Competition, 2nd ed., Sweet & Maxwell, 1978, pp. 754–55; Lenaert Ritter, David Braun and Francis Rawlinson, EC Competition Law—A Practitioner’s Guide, 2nd ed., Kluwer, 2000, pp. 362–63; Michel Waelbroeck and Aldo Frignani, European Competition Law, Transnational Publishers Inc., 1999, p. 551; Pierre Mercier, Olivier Mach, Hubert Gilliéron and Simon Affolter, Grands Principes du Droit de la Concurrence: Droit Communautaire: Droit Suisse, Dossiers de Droit Europeen No. 7, Helbing Lichtenhahn Verlag/Editions Bruylant, 1999, pp. 260–65. 64 See Microsoft, supra note 39, Section 5.3.1.3.1 (“Microsoft’s refusal to supply limits technical development to the prejudice of consumers”) and paras. 693 et seq.

(P) Ehlermann 08 O’Donoghue

6/11/08

11:51

Page 357

A General Test for Exclusionary Conduct under Article 82 EC 357 possibilities if no prejudice to consumers results, such as by offering better products or lower prices. Of course, this simple formulation does not answer every question or cater for every nuance but it at least allows a consistent frame of reference. Third, the emphasis on “prejudice to consumers” in Article 82(b) would expressly allow for efficiency defences for conduct that limits rivals’ production. The status of such defences is not clear under the profit sacrifice test or equally efficient competitor test. For example, an exclusive dealing obligation by a dominant firm will usually limit rivals’ production or access to the market, but it may have a valid defence if the dominant firm is making a substantial customer-specific investment and needs some assurance that the customer will buy from it to justify the investment. Pricing below cost may also have a consumer benefit in limited circumstances. Promotional pricing is legitimate where a new product requires consumer familiarity before customers can appreciate its enhanced qualities. Customer familiarity with the product in question during the promotional pricing phase may render them loyal and therefore willing to pay a higher price in the future because of the product’s added qualities. In such circumstances, the low price is intended to allow customers to try the product and see if they like it. Higher future prices do not depend on competitors’ exclusion, but on the product’s enhanced characteristics over existing products. Article 82(b) also has a clearer normative content for defining exclusionary conduct than any of the broad definitions used by the Community institutions. “Normal competition,” as per Hoffmann-La Roche, is a vague phrase, not least because the Commission has rejected the notion that a common practice within an industry would necessarily constitute “normal competition” if carried out by a dominant firm.65 “Competition on the merits” and “genuine undistorted competition” are also vague, since not all competition on the basis of price, quality, and functionality is allowed under Article 82. And, as noted above, the term “special responsibility” is simply an overall label for conduct that is abusive if carried out by a dominant firm. Article 82(b) is also more precise and certain than saying that each practice should be judged according to its positive and negative effects on consumer welfare. Economists sometimes underestimate the importance of legal certainty. This general principle of Community law requires that firms should, to the extent possible, be able to judge whether their conduct is legal or not when they decide to embark on a particular course of conduct.66 65 Ibid., footnote 877 (citing Case 322/81, NV Nederlandsche Banden Industrie Michelin v Commission [1983] ECR 3461, para. 57; Case T-111/96, ITT Promedia NV v Commission [1998] ECR II-2937, para. 139). 66 See Takis Tridimas, The General Principles of EC Law, Oxford University Press, 1999, pp. 165–66; John Temple Lang, “Legal Certainty and Legitimate Expectations as General Principles of Law”, in Ulf Bernitz and Joakim Nergelius, eds., General Principles of European Community Law, Kluwer Law International, 2000, pp. 163 et seq. For non-competition cases, see Case C-313/99, Gerard Mulligan and others v Minister of Agriculture and Food, Ireland and

(P) Ehlermann 08 O’Donoghue

358

6/11/08

11:51

Page 358

Session Two: A Reformed Approach to Exclusionary Conduct

Fourth, to the extent necessary or useful, Article 82(b) is sufficiently flexible to incorporate elements of the other proposed tests. Although these tests have no clear legal basis under EU competition law, they may be useful in order to verify the “limiting production” test. In particular, Article 82(b) does not prevent the use of the profit sacrifice, equally efficient competitor, or consumer welfare tests where they are valid and useful. Moreover, in circumstances where each test is said to exclude the application of the other tests, and all tests remain essentially untested under Article 82, it is important that Article 82(b) should be relied upon as the basic test for defining exclusionary conduct. This applies not least because economists can and do change their views or evolve them over time. Finally, to the extent relevant, the limiting production test is also similar to the test proposed by the leading treatise on US antitrust law for defining exclusionary conduct.67

Attorney General [2002] ECR I-5719; Case C-63/93, Fintan Duff, Liam Finlay, Thomas Julian, James Lyons, Catherine Moloney, Michael McCarthy, Patrick McCarthy, James O'Regan, Patrick O'Donovan v Minister for Agriculture and Food and Attorney General [1996] ECR I-569, paras. 19–20. 67 See Areeda and Hovenkamp, supra note 38, para. 651a (defining exclusionary conduct as acts that “(1) are reasonably capable of creating, enlarging, or prolonging monopoly power by impairing the opportunities of rivals; and (2) that either (2a) do not benefit consumers at all, or (2b) are unnecessary for the particular consumer benefits that the acts produce, or (2c) produce harms disproportionate to the resulting benefits”) (emphasis added).

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 389

I Andrea Coscelli and Geoff Edwards* Efficiency and Anticompetitive Effects of Tying July 2007

I. Introduction This paper proposes a framework for the analysis of tying practices undertaken by dominant companies under Article 82 with a view to identifying the circumstances in which tying practices are most likely to be harmful. A number of recent contributions have explained why it is essential to focus on the rationale behind the practices adopted by firms with market power and to carefully analyze the effect these practices have on actual and potential competitors and the competitive process more generally. In particular, it is important to recognize that tying practices are not per se harmful to consumers and may have net positive effects. We therefore propose that a rule of reason approach should be adopted when investigating tying cases under Article 82. We also note that our proposed approach is consistent with the rule of reason approach proposed by Frédéric Jenny. This approach is reflected in the Discussion Paper on the application of Article 82 to exclusionary abuses published by DG Competition in December 2005.1 In the Issues Paper distributed in advance of this Workshop, it is suggested that a “consumer welfare balancing test” for abuse cases under Article 82 could be defined as follows: “Conduct cannot be abusive unless it is likely to harm consumers. . . . It is now commonly agreed that few practices (including retroactive rebates and tying) can be considered as generally harmful. The evaluation of these practices has to move away from the evaluation of forms towards the evaluation of effects (to identify those circumstances in which indeed consumers are likely to be harmed). The likely harm to consumers could be demonstrated by showing that the dominant firm will

* Dr Andrea Coscelli is a Vice-President and Dr Geoff Edwards is a Principal in the European Competition Practice of CRA International, based in London. The views expressed in this article are personal to the authors and are not necessarily the views of CRA International. We would like to thank Valter Sorana for very helpful comments on this paper. 1 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses, 2005, available at: http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005. pdf.

(R) Ehlermann 08 Coscelli and Edwards

390

6/11/08

11:52

Page 390

Session Two: A Reformed Approach to Exclusionary Conduct

have higher market power than in the counterfactual scenario, and that the conduct will not lead to sufficient countervailing efficiencies.” (emphasis in the original)

We agree fully with the substantive test being proposed.

1.1 Definitions Bundling and tying are associated concepts. Bundling refers to the practice of selling two products together. Bundling can take several forms. “Mixed bundling” occurs where the two products2 are sold separately but they are also available together at a discount (for example, two pay TV packages may be sold separately but also at a discount to a consumer subscribing to both at the same time). By contrast, “pure bundling” occurs where the two separate products are only sold together and are not available separately. Tying is another special form of bundling which occurs when one product (the “tying” product) is sold only on condition that the second product (the “tied” product) is also purchased, although the second product may also be sold individually. A pure bundle can then be seen as a two-way tie in which each product is sold only on condition that the other product is also purchased. Tying (conditioning the sale of one product on the sale of another) may be effected through contractual terms, through technological integration or through pricing (economic tying). To illustrate how the different approaches could generate exactly the same economic effect, consider the case in which each potential consumer does not want more than one unit of each product and suppose there were a law against explicit contractual ties. In this case, sellers could achieve the same effect by setting the price for one of the products equal to the price for both of the products taken together. The implicit price for the second product is thus zero,3 and the consumer may as well take the two products together, even if the consumer attaches no value to the second product. In fact, an exclusionary tie effectuated through pricing will occur whenever the bundle discount is so large that it would not be profitable for competitors selling only the tied product to match that discount. In other words, it will occur where the implicit price for the tied product when it is sold in the bundle is below the incremental cost of producing the tied product.4 The practices of mixed bundling and refusal to supply are also closely associated with tying. As discussed above, selling a mixed bundle but setting the implicit price of the second product in the bundle below its incremental cost 2 Paragraphs 185 to 187 in DG Comp’s Discussion Paper (cited in previous footnote) discuss how one can determine, in practice, whether products are distinct. 3 In fact, any price for the second product that is below the long run incremental cost of producing the product means that, in practice, a tie occurs. 4 See Barry Nalebuff, “Exclusionary Bundling”, 50(3) The Antitrust Bulletin 321 (2005), for a discussion of tying through bundled pricing (which he calls “exclusionary bundling”).

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 391

Efficiency and Anticompetitive Effects of Tying 391 means that the first product is effectively “tying” the second product, and refusal to supply the “tying” product independently has exactly the same effect as tying the two products together. This means that the antitrust treatment of these practices needs to be perfectly consistent and generate exactly the same outcome when the practices generate the same effects. In this paper we use tying as a shorthand, but the arguments would equally apply to bundling practices (or to refusal to supply) generating the same effects.

1.2 Sample selection When discussing the relative importance of efficiency and “anticompetitive” motives driving tying policies, there is an inherent sample selection problem to address. The reason is that, while all the Article 82 decisions adopted by the European Commission or National Competition Authorities (collectively, “the Authorities”) against dominant companies are in the public domain, there are many complaints brought by rival firms against dominant companies where bundling and tying concerns represent the main theory of harm that are not taken up or are later abandoned by the Authorities without a formal case closure decision that is publicly available. In these cases, it is likely that the decision by the relevant Authority not to pursue the particular complaint was significantly influenced by the presence of an efficiency rationale for the specific conduct being analyzed. This means that a focus on the role played by the “efficiency defence” in the handful of cases where an infringement of Article 82 is found by the Authorities would significantly underestimate the role actually played by efficiency motives in convincing the Authorities whether or not to pursue particular cases. The reason is that these are precisely the handful of cases where it is difficult to find efficiency justifications for the particular conduct carried out by the dominant company, or where there is the need to very carefully balance efficiencies against anticompetitive effects. Another interesting aspect of the enforcement activity in this area is that it is almost always linked to complaints from rivals who claim to be weakened or excluded by the tying practices undertaken by the dominant company. It is therefore the business decision of actual or potential entrants to fight a legal battle that acts as a pre-condition to a possible antitrust investigation. Unless one believes that complaints are more likely to arise in those markets where potentially the most significant detriment to consumer welfare linked to the tying practices exists, there is a very real risk that the enforcement activity will not focus on the markets where the most significant potential detriment arises.5 5 The UK Office of Fair Trading (OFT) has recognized this bias by adopting its “Competition Prioritisation Framework”—a set of criteria used to assess which cases should be taken up for investigation, and which cases already under investigation should be continued, in order to allow the OFT to focus its finite resources on “high impact” cases. See OFT, Competition Prioritisation

(R) Ehlermann 08 Coscelli and Edwards

392

6/11/08

11:52

Page 392

Session Two: A Reformed Approach to Exclusionary Conduct

In practice, there is also a potential asymmetry in enforcement activity between: (i) instances where entry takes place and particular tying activities are subsequently introduced which have the effect of substantially undermining the entrant’s expansion or which cause the entrant to exit; and (ii) tying practices put in place before entry occurs which have the effect of “strategically” increasing barriers to entry. In some cases, the latter practices might generate greater consumer detriment, but they may be less likely to be the subject of an antitrust investigation. For instance, some technological ties might in practice allow a dominant company to monopolize a neighbouring market, and a detailed investigation might show that there was no efficiency rationale for the tie. Nonetheless, the reaction of potential rivals might simply be to focus their R&D efforts in other markets where the potential payoff is more significant, and in the absence of complaints the market might remain monopolized for years without being subject to any enforcement activity.

1.3 Roadmap of the paper There is general agreement among economic commentators that tying may be motivated by cost savings, quality improvements or other efficiencies, by the desire to price discriminate among heterogeneous customers (with ambiguous effects, a priori, on consumer welfare) or by “strategic” (potentially anticompetitive) rationales. It is also possible that more than one of these rationales applies in a particular case. We discuss efficiency rationales and price discrimination in Section 2, the main “strategic” rationales in Section 3 and a possible rule of reason approach to individual (and complex) cases in Section 4.

2. Efficiency rationales and price discrimination There is an inevitable degree of circularity in the assessment of the rationale behind certain practices, especially in the absence of contemporaneous internal (strategy) documents discussing the rationale behind certain choices. The reason is that, often, certain practices are defined as being driven by an Framework (October 2006), available at http://www.oft.gov.uk/shared_oft/press_release_ attachments/compcriteria.pdf. The initial step for the OFT is to consider whether it is best placed to tackle the problem, or whether the problem is better addressed in another forum. This includes consideration of the opportunities for private enforcement. Subsequent steps for consideration before opening an investigation are: 1) an estimate of direct consumer benefit from an OFT investigation; 2) the likely deterrent effect of an investigation; 3) aggravating or mitigating factors; 4) the need to set policy or precedent or the opportunity to take advantage of a high profile sector; 5) the demands of an investigation on OFT resources; and 6) the likelihood of success.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 393

Efficiency and Anticompetitive Effects of Tying 393 efficiency rationale because the companies implementing them are active in fiercely competitive markets, so it would be entirely irrational to expect “strategic” effects to arise. This is a correct inference to make, but what is not clear is the extent to which it is then appropriate to argue that similar practices by firms with significant market power 6 are equally motivated by efficiency reasons, as these firms may simultaneously stand to gain from strategic moves. In this section we briefly review efficiency rationales for tying that have been identified in the literature. We note that, while the Commission’s Discussion Paper on the application of Article 82 acknowledges that there may be efficiency explanations for tying behaviour, the Discussion Paper’s list of possible efficiencies is somewhat narrower than what is found in the literature, and the list does not capture all of the arguments the commentators have raised.7

2.1 Transaction cost savings Tying can reduce transaction costs in a number of ways. For instance, an important reason why tying is a pervasive feature of competitive markets is that savings in customer search costs can be realized by tying component products together into bundles that customers find most convenient. To give a “well-worn” example, people generally do not want to buy shoes for their left foot separately from shoes for their right. If left and right shoes were not sold together as a pair, consumers would have to spend time and effort to search for matching pairs. These search costs would raise the implicit price of the pair of shoes and would by the same token reduce demand. Shoe manufacturers and distributors save consumers these search costs and promote efficient matching by offering only pure bundles of left and right shoes (or to use the terminology of tying, they tie the sale of left shoes to the sale of right shoes and vice versa). Shoes offer a somewhat trivial example, but it should be readily apparent that search cost savings exist for almost any consumer product. For example, automobiles are typically purchased as a bundle of a very large number and variety of components—engine, transmission, chassis, seats, wheels, tyres, CD player and more. Most consumers value the tying together of these components, in part because it avoids for them the time and effort of searching for each component separately. Tying can also reduce “oversearching” or “sorting” costs that arise when customers are allowed to “cherry pick” from a pool of similar, but imperfectly homogenous items. For example, when mushrooms are sold loose in 6 7

We use “significant market power” and “dominance” interchangeably in the paper. See supra note 1, paras. 178 and 204–206.

(R) Ehlermann 08 Coscelli and Edwards

394

6/11/08

11:52

Page 394

Session Two: A Reformed Approach to Exclusionary Conduct

supermarkets at a given price per kilogram, the first customer will choose those that are least blemished, since they will be undervalued at the average price. Later customers will arrive to find that the average quality of the remaining mushrooms has fallen so far that they are overvalued at the average price. These customers may not make a purchase that day and the market may not clear. It may be costly for the seller to price every item individually, as this would require close inspection to determine each item’s value and individual labelling of the items. One solution commonly observed is that a group of similar items of varying quality is tied together, so that customers are forced to take both overvalued and undervalued items in an average priced bundle. Tying in this fashion can reduce search and sorting costs incurred by customers and may also increase total consumption. This explanation for tying was first proposed by Kenney and Klein (1983)8 to explain the movie studio practice of block booking groups of movie features. This involved movie distributors requiring exhibitors to contract to take a number of movie features for a given period at an average price, rather than letting them pick and choose particular films. In the United States, block booking practices were found to be illegal and were prohibited by the US Supreme Court on the basis that they extended monopoly power from desirable features to undesirable features.9 These judgments were later criticized by Stigler, who noted that the monopoly extension analysis makes no sense: the blocks of films were average priced and so the distributor should be able to collect as much revenue by setting higher prices for the more desirable films and lower prices for the less desirable ones.10 Contracts between distributors and exhibitors were entered into before the films’ production and before the true value of each film could be ascertained, and features were therefore said to be “blind sold”. Kenney and Klein argued that the practice of block booking was used to prevent exhibitors from exploiting the informational advantage that arises once the film quality is known by accepting only desirable films and rejecting undesirable ones.11 Kenney and Klein also illustrated their theory with a case study of the diamond marketing arrangements of De Beers. As the precise value of an individual uncut diamond is uncertain and difficult to measure, and as customers typically demand a large number of uncut diamonds, the De Beers company developed a system of bundling together a number of “imperfectly homogenous” diamonds (i.e.,

8 Roy W. Kenney and Benjamin Klein, “The Economics of Block Booking”, 26(3) Journal of Law and Economics 497 (1983). 9 See United States v. Paramount Pictures, Inc., 334 U.S. 131, 156 (1948); United States v. Loew’s, Inc., 371 U.S. 38 (1962). 10 George J. Stigler, “United States v. Loew’s, Inc.: A Note on Block Booking”, Supreme Court Review 152 (1963). 11 Supra note 8. Their argument relies on there being prohibitive costs of entering into an ex post contractual mechanism whereby the price for each film is determined only after its value is known.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 395

Efficiency and Anticompetitive Effects of Tying 395 diamonds of inherently different, but roughly similar value) at an average price in order to reduce search and sorting costs. Kenney and Klein note that block packaging can operate only when the seller enjoys high brand name capital; the value of De Beers’ reputation ensures the correct incentives for De Beers not to cheat customers with bundles of low average value. They argue that it is this very efficiency of De Beers’ selling practices that explains its dominance of the world’s gem-quality uncut diamond market. A third transaction cost explanation for tying is that too many options can be overwhelming for customers and may lead to no purchase decision at all.12 Tying reduces the option set. If the effect of this is to reduce indecision and increase consumption, then welfare will be enhanced.

2.2 Production and distribution cost savings Tying can also save on production and distribution costs through economies of scale and scope. Returning to the automobile example, it would be possible for car manufacturers to sell each component of a car separately and allow customers to assemble the car themselves. Only a very small minority of customers would find own-assembly desirable. Most would be faced with considerably greater costs of assembly than those faced by the manufacturer, and most would therefore prefer to buy a fully assembled car at a price that reflects the manufacturer’s lower assembly costs. This explains why bundling occurs for automobiles, but it does not fully explain why the option of mixing and matching components is generally limited—i.e., it does not explain why components are generally “tied” together. The explanation for this lies in production and distribution cost savings from standardization. A certain level of standardization of product sets can save costs both in assembly and in the administration of customer orders; allowing customers to pick and choose each and every component would impose significant administrative costs on the car manufacturer.13 Moreover, there would be considerably greater costs in distribution if distributors were required to stock a complete range of car components.

12 See Barry Nalebuff, “Bundling, Tying and Portfolio Effects: Part 1—Conceptual Issues”, DTI Economics Paper No.1, February 2003 at p. 32, referring to recent work in behavioural decision theory. 13 See ibid. at page 31. Some high-end car manufacturers might allow for greater customization while charging higher prices to reflect the additional costs. Mass market manufacturers offer more standardized products at correspondingly lower prices.

(R) Ehlermann 08 Coscelli and Edwards

396

6/11/08

11:52

Page 396

Session Two: A Reformed Approach to Exclusionary Conduct

2.3 Compatibility cost savings Where product interfaces are complex and not standardized, it may be costly for a manufacturer of two or more component products to design interfaces between the components so that each component will be compatible with components manufactured by competitors. The process of standardization can itself be costly in the consultation, negotiation and documentation of interfaces. In addition, standardization may require modifications in the design of the components that compromise their efficient functioning. In this case, tying provides a technological solution to minimize these compatibility costs.14

2.4 Protection of intellectual property Tying may also serve to protect intellectual property rights in the tying product.15 Achieving compatibility between the tying product and competitors in the tied product might require disclosure to those competitors of proprietary information in the design of the tying product. For example, suppose that in order to develop applications to run on an operating system it is necessary to have access to proprietary source code for the operating system. The operating system manufacturer might be reluctant to share the source code with competitors in application software for fear that those competitors, or others, would use the source code to develop rival operating system software, undermining the returns to the proprietary information. In this case, tying the operating system together with applications provides customers with the functionality that they desire while preserving the operating system manufacturer’s intellectual property. While the short-term consequences may be higher prices and reduced choice for consumers, the premise of intellectual property laws is that protection of intellectual property is in the long-term interests of consumers because it provides the necessary incentives for future valuable innovations.

2.5 Quality assurance Tying can also preserve product quality and can protect a firm’s reputation and brand name. Where two products are complements, and where the 14 For another discussion of compatibility cost savings as a reason for tying, see Jean Tirole, “The Analysis of Tying Cases: A Primer,” 1(1) Competition Policy International 1, 15 (2005). 15 See ibid.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 397

Efficiency and Anticompetitive Effects of Tying 397 quality of one product may influence the actual or perceived quality of the second product, tying is a way for manufacturers of the second product to maintain its quality and quality perception.16 For example, a manufacturer might tie customer service, after-sales support or distribution to the main product if there is a risk that low cost and low quality third-party supply of the associated services may reflect poorly on the main product, damaging its reputation and reducing its demand. For instance, pay TV broadcasters usually exercise a tight control over the set-top boxes distributed in the marketplace by the manufacturers, as any technical problems with the reception of the programmes that might be due to the set-top boxes will likely generate complaints from the subscribers to the pay TV broadcaster and not to the manufacturer. In addition, where customers or courts lack the information necessary to determine liability for malfunctions of a system of two complementary products, a manufacturer may prefer to tie the two products together rather than risk exposure to liability for the faulty designs of third parties.17

2.6 Legitimate competitive response In some situations tying products together may be an efficient response in a competitive market situation. For example, in markets where products are durable but require frequent replenishment or ongoing maintenance, competing firms may develop strategies of tying the durable product with the replenishment or maintenance service and offering low (possibly below-cost) prices for the durable product while recovering losses sustained on the durable product via charges for the replenishment or maintenance of the product. Common examples are printers and toner cartridges, mobile phone handsets and minutes of talk time, and razors and razorblades. The same argument has also been applied to tying of aftermarket sales.18 If the market for the durable product is competitive, competition should drive the overall system prices for consumers towards the overall system costs.19 Such strategies may be legitimate responses to consumer preferences to reduce their up-front commitments to a product or to “time-shift” their overall payments towards the future. In the absence of such strategies, total consumption would be lower, as many consumers would not purchase the durable product. However, if the firm is dominant in the “primary” market, the Authorities have sometimes taken a suspicious view of low prices charged in the “primary” market to capture “follow on” sales. For instance, in the 16

For a general discussion, see Nalebuff, supra note 12, at pages 32–33. See Tirole, supra note 14. 18 Carl Shapiro, “Aftermarkets and Consumer Welfare: Making Sense of Kodak”, 63 Antitrust Law Journal 483 (1995). 19 See ibid. 17

(R) Ehlermann 08 Coscelli and Edwards

398

6/11/08

11:52

Page 398

Session Two: A Reformed Approach to Exclusionary Conduct

Napp case in the UK,20 the Authorities ruled that it was unlawful for the dominant company to force its smaller rivals to engage in “system competition”. The “rule of reason” approach to these cases is discussed in Section 4 below.

2.7 Eliminating double marginalization A final important efficiency explanation for tying is the elimination of double monopoly mark-ups in a vertical chain of supply to the customer or in a system of complementary products. Where there is a monopolist at one level of the supply chain, or of one product in a system of complementary products, the absence of perfect competition in the other stages of production, or in the complementary products, will result in both lower profits to the monopolist and higher prices for consumers than if the monopolist were the sole supplier and able to set a single monopoly price for the system.21 In this situation, consumers stand to benefit from lower prices if the monopolist ties the vertical stages or complementary products together. While this rationale holds only where the products are complements, and where competition in each of the complementary products is imperfect, it is nonetheless an important efficiency rationale for tying behaviour when these conditions hold.

2.8 Price discrimination Even if none of the above efficiency explanations for tying fits the facts of a particular case, tying is not necessarily an exclusionary practice. Tying may instead be employed to facilitate price discrimination, possibly without any exclusionary intent or effect. Price discrimination is a practice designed to increase seller revenues in the presence of heterogeneous consumers. Although it does not generally have an exclusionary motive, price discrimination in oligopolistic markets has, a priori, ambiguous effects on consumer welfare and therefore requires careful analysis in the circumstances of the particular case to determine whether consumers will gain from the behaviour. 20 See the Decision of the DG of Fair Trading, Case No CA98/2/2001, Napp Pharmaceutical Holdings Ltd and Subsidiaries (Napp), 30 March 2001, paras. 148–159. 21 A monopolist selling complementary products will charge less overall than the sum of the prices of independent monopolists selling each product separately. This is because, in the case of two complementary products, a lower price for one product stimulates demand for the other. Only when this positive externality is internalized within a single firm is pricing set efficiently. The original insight comes from Augustin Cournot, Reserches sur les Principes Mathématiques de la Théorie des Richesses, Hachette, 1838. The result is quite general—it does not depend on specific functional forms of demand or cost functions, and it does not require the products to be perfect complements. See Nalebuff, supra note 12, at p. 38.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 399

Efficiency and Anticompetitive Effects of Tying 399 The literature has identified two forms of price discrimination facilitated by tying practices. The first relates to the extraction of greater value from customers with heterogeneous preferences across two products that are not necessarily functional complements. The second relates to the use of complementary products to meter the usage of customers and to segment the market according to customer valuations for the system.

2.8.1 Homogenizing preferences Where customers have differing preferences for two products, tying the products together can encourage greater consumption than selling the products separately and can enable sellers to extract greater consumer surplus.22 The aim of the tie in this instance is to homogenize preferences. For example,23 suppose there are two customers, Fred and Wilma, and two pay TV channels, Bedrock Bowls and Larry Lava. Suppose the marginal cost to Bedrock Sky Broadcasting (BSB) of supplying each channel is zero. Fred is willing to pay 10 clams for BB and 4 for LL. Wilma is willing to pay only 4 clams for BB but will pay 11 for LL. If the channels could only be sold separately, then BSB’s best price strategy would be to sell BB for 10 and LL for 11 clams. If the two channels can be sold together at a single price of 14, BSB can increase its revenue from 21 to 28 clams. The extra value extracted by BSB comes from the additional channel consumption that the tied offer generates. Consumer surplus will actually increase in this case by 1 clam, as Wilma values the two channels at 15. However, it is also possible to construct examples where this form of price discrimination decreases consumer welfare.24

2.8.2 Metering usage Where some customers use a durable product more intensively than other customers, they are likely to have a different willingness to pay for the product. Setting a single price for the product will fail to extract consumer surplus 22 The original insight is from George J. Stigler, supra note 10. Stigler argued that block booking of feature movies was a rational response of movie studios to the imperfectly observable variation in preferences for movies in different geographic markets. By selling movies in blocks, the studios could extract greater revenues than by selling each movie separately. 23 This example is adapted from David S. Evans and Michael Salinger, “Why do Firms Bundle and Tie? Evidence from Competitive Markets and Implications for Tying Law”, 22(1) Yale Journal on Regulation 37, 49–50 (2005). 24 For example, suppose that Fred and Wilma’s values for LL and BB respectively are 6 rather than 4. In this case, if the two channels had to be sold separately, BSB’s profit maximizing prices would be 6 clams for each channel. Both Fred and Wilma would purchase both channels and BSB’s revenue would be 24 clams. If the channels could be sold together, BSB would set a single price of 16 and earn 32 clams. But in this case Fred’s consumer surplus would fall from 4 to 0 and Wilma’s from 5 to 1; total consumer welfare, measured by consumer surplus, would fall by 8 clams.

(R) Ehlermann 08 Coscelli and Edwards

400

6/11/08

11:52

Page 400

Session Two: A Reformed Approach to Exclusionary Conduct

from customers with higher valuations for the product. At the same time, customers with lower valuations will not purchase the product. Ideally, a firm in this situation would like to identify high- and low-valuation customers at the time of sale and charge them varying up-front sums to reflect their full valuations. When this is not possible, an alternative is to find some way to meter usage of the product and to charge prices that vary accordingly. One way to do this is to tie the durable product to a complementary product that must be used in a given proportion to usage of the durable product, and then to charge a reduced price for the durable product while charging higher prices for the complementary product, thereby extracting greater consumer surplus from those customers with greater intensity of use. This is referred to as “metering” or “requirements tying”.25 A familiar illustration is of a printer manufacturer tying toner cartridges to its printers through patented designs so as to extract greater toner revenues from high-usage customers than from those with low usage. The theory has also been applied to the tying of after sale services such as maintenance.26 In some cases it may be debatable whether a tie is necessary for the purpose of metering, and whether an alternative and more direct metering method and charging structure for the main product could be used instead.27 Where a less costly method of metering exists, greater suspicion of a price discrimination motivation for the tying behaviour may be justified. However, it is likely that there will be instances where tying does indeed offer a more efficient and practical metering method than direct metering.28 Again, the consumer welfare effects of requirements tying are a priori ambiguous; some customers may be better off under metering while others may be worse off.29

25 This explanation for tying was developed by Meyer L. Burstein, “The Economics of Tie-In Sales,” 42 Review of Economics and Statistics 68 (1960). For a formal exposition, see Lester G. Telser, “A Theory of Monopoly of Complementary Goods,” 52 Journal of Business 211 (1979). 26 See Zhiqi Chen and Thomas W. Ross, “Refusals to Deal, Price Discrimination and Independent Service Organizations,” 2(4) Journal of Economics and Management Strategy 593 (1983). 27 See Nalebuff, supra note 12, at pp. 74–77. Nalebuff argues that, if direct metering is possible, this should be encouraged over a tying arrangement which could have anticompetitive collateral effects in the tied product market and even in the tying product market (if, in order to enter the latter market, a firm needs access to independent firms in the tied market). 28 See ibid. Nalebuff acknowledges that direct metering may be difficult due to measurement issues and added costs in adding counters to products. 29 It is possible to construct examples where no customer is worse off. See Simon Bishop and Mike Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, Sweet and Maxwell, 2002, pp. 212–213. Bishop and Walker develop a specific example in which, without metering, a low-value customer will not purchase the durable product and a high-value customer will have her entire consumer surplus extracted, whereas with metering, the low-value customer will purchase the product but have his entire consumer surplus extracted, while the high-value customer will derive a positive consumer surplus.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 401

Efficiency and Anticompetitive Effects of Tying 401

2.8.3 Ambiguity of consumer welfare effects Since the consumer welfare effects of price discrimination are a priori ambiguous, we agree with Tirole (2005),30 who maintains that, where tying conduct is supported by a rationale of price discrimination, it is appropriate to apply a rule of reason analysis to determine what the net effects on consumers would be on the facts of the particular case. We also agree with Nalebuff (2003), according to whom price discrimination schemes can introduce a number of inefficiencies caused by customer’s efforts to try to work around tying schemes.31 These factors (among others) should be considered in the rule of reason analysis.

3. Anticompetitive rationales—exclusionary tying As we have seen above, tying can have efficiency explanations or can be motivated by the additional revenues to be earned from price discrimination (with ambiguous effects on consumer welfare). Tying can alternatively (or at the same time) have exclusionary effects with respect to the supply of the tied product or of the tying product (or even of products that may not yet exist), and this “strategic” rationale may be the main motive behind the decision to tie the two products together. We begin this section by explaining the circumstances under which we can be reasonably confident that tying does not have an anticompetitive motivation. This derives from the Chicago school critique of the classical antitrust hostility towards tying arrangements, a critique developed in particular in the 1960s and 1970s. We then move on to review post-Chicago learning with respect to the circumstances in which tying may have anticompetitive purposes and effects. As mentioned above, these effects may be felt in the tied market, in the tying market, or in other markets altogether.

3.1 When tying is not likely to be anticompetitive: Chicago school insights The Chicago school critique provides some insights into the conditions under which tying could be said to be less likely to have an anticompetitive motivation. Chicago school economists argued that if a firm is a monopolist in 30 31

Supra note 14. Nalebuff, supra note 12, at pp. 77–79.

(R) Ehlermann 08 Coscelli and Edwards

402

6/11/08

11:52

Page 402

Session Two: A Reformed Approach to Exclusionary Conduct

relation to one product, and if customers also purchase another product that is supplied competitively, there is nothing for the monopolist to gain from requiring the customer to purchase both products together.32 In this particular situation there is only ever one monopoly rent that can be extracted. Similarly, if two products are complements and there is no independent demand for the tied product, such as the components of an automobile (chassis, engine, steering wheel, etc.) or a desktop operating system and an internet browser, then there is again only one available monopoly rent in the system and the monopolist of the tying product has incentives for the tied product to be supplied as efficiently as possible so that it can extract all available rents in the system through the price for the monopoly product.33 Chicago school scholars argued that if, in either of these situations, a monopolist supplier of the “tying” product decides to tie the two products together, it is unlikely that it does so in order simply to leverage market power into the tied product market. However, post-Chicago learning has questioned the generality of the Chicago school approach. In Section 3.2 we explore what happens when the Chicago school conditions are relaxed, which may be more realistic in many markets. Post-Chicago scholars have also identified the limited frame of reference of the Chicago theory. Implicit within the Chicago model is a static concept of strategic behaviour; the model demonstrates the irrationality of seeking additional profits in the tied product market when the tying product market is monopolized, but it ignores the possibility of future rent streams that may derive from control of the tied product. In Section 3.3 we review the recent development of dynamic models which illustrate that concerns for competition may arise even when the Chicago school conditions hold. First, there may be an incentive for a monopolist in the tying product to leverage its market power into the market for the tied product in order to exclude competition in the market for the tying product or in new and emerging markets that are connected to the tied product. This was essentially the US Department of Justice’s allegation against Microsoft with regard to the tie of the company’s operating system with its internet browser. Second, where products are durable and the tied product is characterized by product upgrades, a mono32 Suppose a customer independently values each of two products, A and B, at €10 (with no value for a second unit of either product), and suppose that the marginal cost of producing each product is €5. Suppose further that a firm has a monopoly over the production of A, but B is supplied competitively. If the monopolist supplies A alone, it can set a price of €10 and earn a profit of €5 from the customer. If the monopolist instead requires that the customer take B together with A, then the most that the customer will pay for this bundle will be €15 and the monopolist will gain nothing. This is because, at any higher bundle price, the customer stands to earn greater consumer surplus by purchasing B on its own for €5 from a competitive supplier. See Aaron Director and Edward H. Levi, “Law and the Future: Trade Regulation,” 51 Northwestern University Law Review 281 (1956); Richard A. Posner, Antitrust Law: An Economic Perspective, University of Chicago Press, 1976; Robert H. Bork, The Antitrust Paradox, Basic Books, 1978. 33 See Posner, cited in previous footnote.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 403

Efficiency and Anticompetitive Effects of Tying 403 polist may have an incentive to leverage its market power into the tied product market in order to capture future profits in the upgrades. Microsoft’s operating system and application programmes may provide an illustration of this. If a plausible theory of harm can be developed along one of these lines, there may be legitimate concerns regarding anticompetitive effects even when the Chicago school conditions hold.

3.2 Relaxing the Chicago school conditions Whinston’s seminal paper34 was the first to explore what happens when the Chicago school conditions are relaxed. He explored the consequences of relaxing, first, the condition of perfect competition in the tied product market, and second, an implicit assumption underlying the Chicago school’s complementary products theory, namely that the monopolist’s product in the tying market is essential for all uses of the tied product. In each case, he showed formally that tying can be used as a strategic tool to leverage monopoly power from the tying market into the tied market; in other words, it can be profitable for a monopolist of the tying product to exclude competitors in the tied product. The main intuition in Whinston’s models is that, in certain circumstances, tying can deny competitors in the tied product market the scale necessary to cover fixed costs, and can thereby deter entry or drive existing competitors out of the tied product market. In each case, the monopolist incurs costs as long as rivals remain in the tied product market (relative to selling the products separately) and the tying conduct is only profitable if exclusion is effected and prices can subsequently rise. It is therefore the exclusionary effect of the tying behaviour that makes tying profitable for the monopolist. Note that, as with predation, actual exclusion of rivals may not be necessary for the monopolist to achieve anticompetitive effects; the mere disciplining of rivals’ conduct in the tied product market may be sufficient.35

3.2.1 Tying with independent demands and imperfect competition in the market for the tied product The first model that Whinston explores assumes that there are two products that are demanded independently and that competition for the tied product is imperfect. In particular, suppose there are fixed costs in production of the tied product so that there are scale economies in the production process; the 34 Michael Whinston, ”Tying, Foreclosure and Exclusion”, 80 (4) American Economic Review 837 (1990). 35 See, e.g., ibid., p. 844, footnote 10.

(R) Ehlermann 08 Coscelli and Edwards

404

6/11/08

11:52

Page 404

Session Two: A Reformed Approach to Exclusionary Conduct

structure of the tied product market is therefore oligopolistic. Since the two products are demanded independently, there are, in principle, two markets and therefore two “monopoly” rents available.36 If the monopolist commits to tying the two products together, then in order to make sales of its monopolized product it must also make sales of the tied product. This will lead the monopolist to compete more aggressively and to lower its implicit price of the tied product to attract sales away from rivals in the market for the tied product.37 If the impact on rivals’ sales is sufficiently large, the rivals’ profits may fall below the level that justifies entry or continued operation in the presence of fixed costs, and exclusion may result. Tying in this case may or may not be a profitable strategy for the monopolist. On the one hand, the monopolist stands to gain from converting the tied market from an oligopoly into a monopoly (or into an oligopoly with fewer competitors). On the other hand, the monopolist will only offer the two products together in a tie, and must therefore sacrifice some sales of the monopolized product to customers with low valuations for the tied product. Note, importantly, that in this model the monopolist will only stand to earn greater profits if rivals are excluded from the tied product market. If rivals remain in the tied product market, then the monopolist would do better to sell the products separately rather than tying them together.38 To identify when the exclusionary strategy just described is a likely explanation of tying behaviour, it is important to consider whether it would be both credible and profitable for the monopolist. The strategy will only exclude rivals if the monopolist can make a credible (irreversible) commitment to the tie, such as in the form of a technological tie in the product design or in the production process. Otherwise, the rival will realize that if it enters or remains in the market the monopolist will do better to undo the tie. The strategy will be profitable only if customers have a reasonably high valuation for the tying product and if the monopolist’s version of the tied product is of a reasonably high quality. If a significant proportion of customers do not place great value on the tying product, and/or regard the rival’s tied product as significantly superior in quality, then the tie will be ineffective in excluding the rival because the monopolist will be unlikely to capture a sufficient proportion of the rival’s sales (customers will simply ignore the monopolist’s tied

36 The rent in the market for the tied product might not actually be a “monopoly” rent if rivalry in the tied product market is disciplined but not excluded and the market thus remains an oligopoly. 37 To see the monopolist’s incentives to compete more aggressively when the products are tied, note that, when the products are sold separately and the monopolist fails to sell a unit of the tied product, the monopolist loses only the profit on that unit of the tied product, whereas when the products are tied together, failing to sell a unit of the tied product causes the monopolist to lose both the profit on the tied product and the monopoly profit on the tying product. 38 This is because, by tying the two products together, the monopolist sacrifices sales of the monopolized product and also receives a lower implicit price for the tied product.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 405

Efficiency and Anticompetitive Effects of Tying 405 offer). More generally, these are some of the key factors one would assess under a rule of reason analysis. Finally, although there is likely to be a loss of consumer welfare in this scenario of exclusionary tying (since, following exclusion, prices for the tied product will rise and product variety will be diminished), there would also be efficiency gains in the form of avoided fixed costs. Therefore, the overall welfare impact is unclear prior to an evaluation in the context of a particular market.

3.2.2 Tying with complementary demands when the monopolist’s product is not essential Whinston also examined an alternative model in which the two products are complements for one group of customers, but where there is another group of customers with demand only for the tied product or who attach only a small value to the monopolist’s product.39 The monopolist’s product is therefore not essential for all customers of the tied product. Here again, there is the possibility of extracting a second “monopoly” rent from those customers with independent demand for the tied product. By tying the products together, the monopolist ensures that the first group of customers (who view the two products as complements) will acquire its version of the tied product, thereby reducing the size of the tied product market available to rivals. If a large enough proportion of customers view the products as complements, there may be insufficient residual demand for rivals in the tied product market to cover their fixed costs of operation, leading to exclusion. Again, note that the monopolist will only stand to earn greater profits if rivals are excluded from the tied product market. Prior to exclusion, the monopolist will incur costs relative to selling the products independently. Following exclusion, the monopolist can earn greater profits by setting higher prices for the tied product when it is sold separately to customers with demand only for that product while the threat of the monopolist to tie the products again will deter entry and allow the monopolist to continue extracting the rent. In this way, the monopolist may leverage its market power in relation to the tying product so as to obtain and exploit market power in the market for the tied product. 39 To give a familiar example, suppose that there is only one hotel on a resort island and that dinner in the hotel’s associated restaurant is included in the room price. Hotel guests view the hotel and restaurant on the island as complements. However, suppose there is another group of customers for restaurants on the island, such as locals and visitors for the day from neighbouring islands. These customers have demand for restaurants independent of demand for hotel accommodation on the island. Replacement parts markets can also be characterized as instances in which the monopolist’s tying product is not essential for use of the tied product. For example, one component of a typical stereo system is an LED display. When the stereo system is first purchased it is purchased as a bundle of components including the display. After this initial purchase, however, if the display fails a customer may seek a replacement part. For this customer, demand for the replacement part is independent of demand for the other components of the stereo system.

(R) Ehlermann 08 Coscelli and Edwards

406

6/11/08

11:52

Page 406

Session Two: A Reformed Approach to Exclusionary Conduct

Once again, tying will only exclude rivals if a credible commitment can be made to the tie. Whether a tie in this situation would be profitable for the monopolist depends on the sizes and valuations of the two groups of customers. A sufficiently large group of customers with independent demand for the tied product will allow an efficient rival to enter and remain in the tied product market. Unless there is exclusion, there would be nothing for the monopolist to gain from tying the products. If there is exclusion, then consumer welfare will be harmed due to higher prices for the tied product, although again, aggregate welfare may rise or fall. Note also that the Whinston model assumes homogeneous products in the tied market, while there will almost invariably be some degree of differentiation in real-world markets, and this will have a significant effect on the profitability of a potential exclusionary strategy.

3.3 Dynamic models Even when the Chicago school conditions hold in a static sense, dynamic models of tying can be developed to reveal the possibility of strategic motivations related to separate streams of future rents that may be earned if the monopolist can gain control of the tied product market.

3.3.1 Tying to exclude competition in the tying market or in an emerging market Carlton and Waldman (2002)40 show that tying of complementary products can be used to preserve a monopoly position in the tying market or to allow a monopolist to transfer its monopoly to a newly emerging market. The intuition for tying to preserve market power in the market for the tying product is straightforward. Suppose that a monopolist in a primary product market also operates in a complementary product market, and suppose that a potential entrant has the ability to supply a superior complementary product. It follows from the Chicago school logic that the monopolist would generally have no incentive to tie its products together, as it can benefit from the effect that the superior complementary product offered by its rival will have on demand for the monopoly product. However, if the entry of an independent supplier of the complementary product enhances the prospect of rival entry into the monopolist’s primary product market in a subsequent period, then the monopolist may have an incentive to tie its products together to exclude entry into the complementary product market, thus blocking the path 40 Dennis W. Carlton and Michael Waldman, “The Strategic Use of Tying to Preserve and Create Market Power in Evolving Industries,” 33(2) RAND Journal of Economics 194 (2002).

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 407

Efficiency and Anticompetitive Effects of Tying 407 into the primary market and preserving its monopoly.41 Carlton and Waldman show that the tying conduct may exclude entry into the complementary product market in one of two ways; it may deny the entrant the necessary scale to cover the fixed costs of entry, or alternatively, even if there are no fixed costs of entry into the complementary product market, if there are network externalities on the demand side of the complementary product market then the tie may have a similar exclusionary effect. A similarity between this model and the earlier Whinston models is that the tying conduct has to be irreversible to be effective. If the tie is reversible, then entry will not be deterred. A significant difference between this model and Whinston’s models is that, here, there is only one monopoly rent available in a system of complementary products. It is the preservation of that single monopoly rent in the future that motivates the exclusionary conduct. The consumer welfare concern is that, in the long run, consumers will be harmed by having to pay higher prices for the system and by experiencing less variety. This stylized model captures a number of the key elements of the case brought by the US Department of Justice against Microsoft. In that case, it was alleged that Microsoft had tied its internet browser to its operating system in order to exclude a rival browser (Netscape Navigator) from the complementary product market and thus to preserve Microsoft’s monopoly over the primary market, i.e., the market for operating systems. The analysis with regard to extending a monopoly position into a newly emerging market is similar. Suppose the newly emerging market is for a product that is associated with the same complementary product as the primary market. Tying of the primary and complementary products may again reduce the profitability of an entrant in the complementary product market below what is necessary to cover its fixed costs of entry, and this may deny a rival in the newly emerging market an independent source of the necessary complementary product. In this way, a monopolist in the primary market can transfer its monopoly to a newly emerging market. Alternatively, if the complementary product market is characterized by demand side network externalities, tying of the primary and complementary products may exclude rivals in the complementary product and again enable the monopolist to transfer its monopoly from the primary market to the newly emerging market. Choi and Stefanadis (2001)42 model a somewhat different setting in which there is a single potential entrant for each of multiple complementary 41 For example, if the products are complements and if the monopolist refuses to supply its complementary product for use in combination with rival versions of the primary product, then entry into the primary product market may depend on the availability of an independent supplier of the complementary product. In Carlton and Waldman’s model, the returns to the rival in the primary market depend on capturing more of the surplus associated with the superior complementary product, so if the superior complementary product is not available, the rationale for entry into the primary market falls away. 42 Jay Pil Choi and Christodoulos Stefanadis, “Tying, Investment and the Dynamic Leverage Theory,” 32 RAND Journal of Economics 52 (2001).

(R) Ehlermann 08 Coscelli and Edwards

408

6/11/08

11:52

Page 408

Session Two: A Reformed Approach to Exclusionary Conduct

products that are currently monopolized by one firm. The authors show that, if the monopolist ties the products together it can preserve its monopoly position by reducing the probability of entry, as entry would then require the coordination of successful innovation in all of the complementary products.43 The Office of Fair Trading (“OFT”) in the UK used a similar framework in its characterization of Genzyme’s abuse.44 The OFT argued that Genzyme’s policy of selling the drug Cerezyme (in respect of which Genzyme was dominant) together with the associated home care services at the same price to the National Health Service for use by patients suffering from Gaucher disease amounted to a tie, and that it effectively excluded rivals from the potentially competitive market for the provision of home care services to Gaucher patients. According to the OFT, this also had the effect of increasing the barriers to entry in the monopolized market for the tying product (the supply of treatments to Gaucher patients, where Cerezyme was the leading treatment), as Genzyme’s “close relationship” with Gaucher patients through the delivery of home care services would have made it more difficult for a rival supplier to convince patients to switch to its product. The Genzyme case highlights a key issue common to many other cases: is there a relevant antitrust market for services associated with the supply of a main product where those services may be insourced or outsourced? That is, whenever dominant companies decide not to outsource particular activities (or to bring back into the “house” activities previously outsourced) in competitive markets, do they “monopolize” the “tied” market related to the supply of these services? For instance, suppose that a dominant pay TV company initially issues a tender to acquire installation services from third parties to offer at subsidized prices to its customers. When the original contract with a third party expires, the company decides it would be more efficient to bring this activity back in-house. Is this an illegal tie, given that the company has now “monopolized” the market for the supply of installation services to its pay TV customers? We believe that there is a very significant risk that antitrust concerns regarding these types of in-sourcing/outsourcing activities by dominant companies would prevent dominant companies from efficiently reorganizing the perimeter of their business, as a number of other factors affecting in-sourcing/outsourcing decisions change over time.

43 This is a formalized extension of Oliver E. Williamson, “Assessing Vertical Market Restrictions: Antitrust Ramifications of the Transaction Cost Approach,” 127 University of Pennsylvania Law Review 953 (1979), where it was noted that tying forces an entrant to enter both markets, and that this may be more costly and hence less likely. 44 See Judgment of 11 March 2004 of the Competition Appeal Tribunal, Genzyme Ltd v. Office of Fair Trading, Case No. 1016/1/1/03, [2004] CAT 4.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 409

Efficiency and Anticompetitive Effects of Tying 409

3.3.2 Tying with durable products and upgrades for the tied product In a subsequent paper, Carlton and Waldman (2005)45 considered a situation where the products are durable and product upgrades are a characteristic of the tied product. They showed that, even if the products are complements and the monopolist’s product is essential for consumers of the tied product, there might still be an incentive for the monopolist to tie the products together. In particular, there might be an incentive to exclude rivals in the market for the tied product in order to capture future rents from the sale of upgrades of the tied product. This observation essentially relaxes another implicit assumption in the Chicago school analysis, namely that all sales are made in one period and there are no profits to be realized in a later period. If upgrades are sold in a second period, then in order for the monopolist to ensure that it captures the rents from selling those upgrades it must ensure that it is the firm that sells the upgrades. If a rival could develop a superior tied product, the only way for the monopolist to ensure that it sells upgrades in the second period is to exclude the rival in the first period. Carlton and Waldman also showed, as might be expected, that the incentives for the monopolist to tie the two products together are stronger when switching costs are added to the model.

3.4 No-cost exclusionary tying Tirole has argued that exclusionary tying cases are best characterized as predation rather than tying per se.46 However, as Nalebuff has observed, an important distinction between tying cases and standard single-product predation cases is that, in some cases, exclusionary tying may be effected with no profit sacrifice by the monopolist.47 Nalebuff focuses on the case of economic ties (tying through mixed bundle pricing) and describes two ways in which a monopolist may price its products with the effect of excluding equally efficient rivals in the market for the tied product at no cost to itself. First, the monopolist may under-price the competitive (tied) product. Following Nalebuff’s notation, let the monopoly price of product A (the tying product) be m, and let the competitive price of product B (the tied product) be c. Nalebuff assumes a special case where the two products are consumed in equal proportions. In this case, consumers will be indifferent between buying a bundle of A and B at a price of m+c and buying A and B 45 Dennis W. Carlton and Michael Waldman, “Tying, Upgrades and Switching Costs in Durable-Goods Markets,” NBER Working Paper 11407 (2005). 46 See Tirole (2005), supra note 14. 47 See Nalebuff, supra note 4. See also Barry Nalebuff, “Tied and True Exclusion: Comment on Jean Tirole’s ‘The Analysis of Tying Cases: A Primer’”, 1(1) Competition Policy International 41 (2005).

(R) Ehlermann 08 Coscelli and Edwards

410

6/11/08

11:52

Page 410

Session Two: A Reformed Approach to Exclusionary Conduct

separately at prices of m+e and c-e respectively. Setting prices in this way will not affect the monopolist’s profitability or consumer welfare in the short term; the monopolist will continue to earn the full monopoly rent on each sale of product A whether it sets unbundled prices of m and c or m+e and c-e. However, rivals in product B cannot compete at a price of c-e and will be excluded from that market. The monopolist would therefore stand to gain, and consumers would lose in the long term, if by excluding competition in the B market the monopolist can preserve its monopoly over product A. The important difference with predation cases is that the monopolist in a tying case can use the price of the monopolized product to instantaneously offset any losses from under-pricing the competitive product, so that no short-term profit sacrifice is implied. Second, the monopolist may over-price the monopoly (tying) product. Lowering the price of product B below the competitive level may not even be necessary to effectuate no-cost exclusion. Here, Nalebuff assumes products consumed in fixed but not necessarily equal proportions. However, for simplicity of exposition we will continue to consider the case of equal proportions. If the stand-alone price for A is set at m+e and the A+B bundle price is m+c, then for any positive value of e the customer is better off buying B from the monopolist than from an equally efficient rival B producer. Although the stand-alone price of A is above the profit-maximizing monopoly price, there will be no sacrifice of profits in equilibrium, as no customers will purchase A on a stand-alone basis. Rivals will be excluded from the B market because, while they would charge c, the implicit price faced by customers would be c+e. Again, in contrast to predation cases, customers’ dependence on the monopolized product A can be exploited so that exclusion may be effected in the market for product B with no short-term profit sacrifice.

3.5 Mixture of efficiency and possible “anticompetitive” motives While it is useful to separately list the “efficiency” and “strategic” considerations in tying cases, as we do in the sections above, the reality is that most antitrust cases analyzed in detail by the Authorities tend to have both efficiency and “strategic” motives behind specific tying practices. An interesting example is discussed in a recent EC merger decision, Amer/Salomon.48 In this decision, the Commission discusses a fairly recent trend according to which suppliers of winter sports equipment tie the sale of alpine skis and alpine bindings. Historically, skis and bindings were sold on a separate, stand-alone basis and the market leaders in the sale of skis differed from the market leaders in the sale of bindings. 48

Commission Decision of 12 October 2005, Amer/Salomon, Case No COMP/M.3765.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 411

Efficiency and Anticompetitive Effects of Tying 411 According to the Amer Group (owner of the Atomic brand) and Salomon, there were technical and economic motivations for the development, since “ski manufacturers have identified the ski-binding interface as an important factor in the overall performance of a ski. To make this interface more efficient, each manufacturer has developed its own interface. The positions of screws are now specific to each ski manufacturer so that only one brand of bindings can match.”49 When the tying practices began, the two markets (alpine skis and alpine bindings) were highly competitive oligopolies with a reasonably large number of suppliers selling many differentiated products in each. It would therefore seem that the move must have been entirely driven by efficiency considerations. Nonetheless, there were positive margins in each market and the effect of the tying practices was twofold: (a) the market leaders in alpine skis managed to significantly increase their share in the market for alpine bindings (i.e., the market for the “tied” product) and (b) the main manufacturers of alpine bindings were weakened, and some of them in the end were purchased by manufacturers of alpine skis. While the size of the “efficiency” benefit associated with the technical tie to the end consumers was a matter of debate, it seems clear that the leading manufacturers of the “tying” products (alpine skis) gained significantly from the decision to tie the “tied” product (alpine bindings). This example shows that, even in situations where the degree of market power is limited, it is very difficult to disentangle pure efficiency reasons from more “strategic” reasons that might be labelled “anticompetitive” during an antitrust investigation. For instance, in this case one might be worried that the market structure, after all the leading manufacturers began to tie their sales of alpine skis and bindings, is less competitive and/or less conducive to innovation than the previous (less concentrated) market structure, and that this has caused some consumer detriment.50

4. Analyzing tying under a rule of reason approach From the preceding sections we have seen that there can be efficiency or price discrimination explanations for tying, and that it is also theoretically possible for tying to exclude or discipline competitors, with adverse effects on consumers. We note that Nalebuff prefers a per se rule to a rule of reason approach in cases where a monopolist creates a tie that equally efficient rivals 49

Ibid., para. 16. It is clearly difficult to compare two oligopolistic market structures with many players in each in a situation where innovation and branding are key competitive variables. 50

(R) Ehlermann 08 Coscelli and Edwards

412

6/11/08

11:52

Page 412

Session Two: A Reformed Approach to Exclusionary Conduct

cannot match and a significant share of the tied market is foreclosed (and the monopolist could reasonably have understood that such foreclosure would be the consequence of its pricing).51 However, we would favour a (structured) rule of reason approach in order to open up the analysis to the possibility of counterbalancing efficiency explanations for the tie that benefit consumers.52 A rule of reason approach would also be sufficiently flexible to properly analyze tying cases with essentially empirical assessments focusing on the market facts discussed in Section 4.1 below. There is no set of necessary and sufficient conditions that can be derived from the economic analysis to allow a decision maker to easily adjudicate on a tying claim, so we believe a detailed empirical analysis under a (structured) rule of reason remains the most appropriate approach.53 We suggest that there should logically be two steps to an analytical assessment of tying behaviour under a (structured) rule of reason approach.54 First, as a preliminary step, it must be established whether the tying conduct has excluded or disciplined efficient competitors, or whether there is a dangerous probability that it will do so. Only if there is a likely adverse effect on efficient competitors is there a need to investigate further. Second, it must be assessed whether the adverse effect on competitors is harmful to, or simply part of, the process of competition, i.e., it must be determined whether harm to consumers is, or is likely to be, the ultimate effect.55 An issue in mixed bundling cases is whether there is a potential foreclosure concern when a dominant firm selling a bundle of products prices each pro51

See supra notes 4 and 47. We speak of a “(structured) rule of reason” in this paper because we consider that the rule of reason approach we envisage should have a logical structure. On the other hand, we leave it to legal commentators to decide how “structured” the approach must be to help the courts (i.e., in relation to the extent to which burdens of proof and legal presumptions need to be rigidly determined in the administrative or civil procedure). In this sense, we would agree with the observations made by Lowe and by Neven at this Workshop with respect to the use of safe harbours and rebuttable presumptions as opposed to a more formal “structured rule of reason” as described by Ahlborn and Padilla (in this Volume). 53 Interestingly, the imposition of a per se rule on tying in the US has simply shifted part of the substantive assessment that would take place under rule of reason to the discussion of whether there are two distinct products and whether the two products are tied (i.e., whether there is standalone demand for the tied product.) 54 A similar two-step approach has been suggested by Willig. See Robert Willig, “(Allegedly) Monopolizing Tying via Product Innovation”, oral remarks at the FTC/DOJ’s Sherman Act Section 2 Joint Hearing on Understanding Single-Firm Behavior: Tying Session, 1 November 2006. 55 Willig’s two-step approach (see previous footnote) begins with an assessment of whether there is “harm to competition” and then asks whether the practice is “part of competition” (essentially an allowance for business justifications and efficiency effects). Unfortunately, “harm to competition” can be a particularly vague term if not carefully defined. For example, in this instance, does it mean harm to competitors, harm to the dynamic process of competition, harm to social welfare, harm to consumer welfare, or something else altogether? In the context of Willig’s two-step approach, it seems to us that what Willig is referring to by this term is actually “harm to efficient competitors”, which in the absence of any efficiency effects would be likely to harm consumers through a reduction in output and increase in prices. 52

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 413

Efficiency and Anticompetitive Effects of Tying 413 duct above long-run average incremental cost (LRAIC) but causes efficient single-product rivals to be excluded because they cannot recover their common set-up costs. The issue here is that the dominant firm exploits economies of scope (and possibly economies of scale) that are not immediately available to single-product rivals. Our view is that a rule of reason test should be applied in this context, as in some instances it might be more appropriate to adopt a “reasonably efficient competitor” test than an “equally efficient competitor” test in order to protect efficient single-product (or smaller) rivals until they are able to expand and achieve the economies of scale and/or scope needed to compete on an equal footing with the dominant company. This means that a “safe harbour” for prices above LRAIC would not be appropriate.

4.1 Theory of exclusion: will the tie have exclusionary effects? The first step, we suggest, in a (structured) rule of reason approach is to ask whether the tying conduct in question would be likely to exclude or discipline rivals so as to cause consumer detriment in the absence of any efficiency effects. In other words, this initial step involves the development of a theory of exclusion and empirical support for this theory wherever possible. This is a key step in the assessment of when tying is likely to be harmful. We first note that an essential factor in the assessment of a tie is whether the firm in question has market power in the market for the tying product (and the burden of proof for this assessment would obviously lie with the Authorities). Unless the firm’s primary product is viewed by customers as having few alternatives and the firm thereby has significant market power to use as leverage, there can be no question of an anticompetitive tie. The following discussion therefore assumes that there is significant market power in relation to the tying product. A large range of factors determine the impact that a tie is likely to have on rivals in the market for the tied product, many of which are amenable to assessment with empirical evidence. We will begin with supply-side conditions in the tied product market, and then we discuss demand-side conditions and counter-strategies that may be available to rivals in the tied product market.56 Finally, we consider whether the tie is likely to lead to exclusion in the market for the tying product or in emerging markets.

56 The analysis of competition in the market for the tied product deals with the question of whether in fact the tied product is a distinct product and involves concepts such as the demand for “pure heterogeneous bundles” and for “mixed heterogeneous bundles”.

(R) Ehlermann 08 Coscelli and Edwards

414

6/11/08

11:52

Page 414

Session Two: A Reformed Approach to Exclusionary Conduct

4.1.1 Supply-side conditions in the tied product market On the supply side, higher marginal (unit) costs and higher fixed costs in producing the tied product each increase the likelihood that a tie will have exclusionary effects. • Marginal costs. When two products are tied together, customers that require the tying product pay a zero implicit (effective) price for the tied product. Rival producers of the tied product will be able to compete for these customers only if they can differentiate their version of the tied product sufficiently so as to convince these customers to add the rival’s product to the version that they have already purchased from the monopolist through the tie. In other words, the rival’s product must offer additional value to these customers that exceeds the marginal cost of production and therefore the price that the rival must charge for the product. Higher marginal costs thus require greater differentiation in the rival’s tied product in order for the rival to make sales to these customers, and increase the likelihood of exclusion in the tied product market.57 • Fixed costs. Fixed costs imply economies of scale in production. In order for a rival to enter and remain in the tied product market, it must make sufficient sales and sufficient margins on those sales to recover its fixed costs. Tying by a monopolist in the primary market reduces the residual market available to rivals in the tied product. The higher the fixed costs, the less likely the rival will be able to achieve a minimum efficient scale in the presence of a tie. Since R&D costs are fixed costs, exclusion is more likely in industries where R&D plays a particularly important role. Learning curve effects may be seen as analogous to fixed costs, requiring rivals to achieve economies of scale in production of the tied product and contributing to a greater likelihood that a tie will lead to exclusion.

4.1.2 Demand-side conditions in the tied product market Recall that customers which purchase the tying product face a zero implicit (effective) price for the tied product. For a given tied product market size, the higher the proportion of tied product demand subject to the tie, the smaller will be the residual market available to rivals that produce only the tied product, and the more likely it is that those rivals will be excluded. A number of factors are relevant to an assessment of the size of the residual market that remains available to rivals that offer only the tied product. • Proportion of customers with demand for both products. If there is no scope for product differentiation in the tied product, and if all customers consume 57

This factor is discussed in Tirole (2005), supra note 14, at pp. 9–10.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 415

Efficiency and Anticompetitive Effects of Tying 415











the two products in fixed proportions (i.e., if there is no independent demand for the tied product) then a tie implies exclusion, as no customers will remain available to rivals in the tied product. The smaller the proportion of customers of the tied product that also require the monopolist’s primary product (and so the larger the proportion of customers with independent demand for the tied product) the greater the residual market available to rival producers of the tied product. Identity of customers subject to the tie. Some customers may be more readily accessible to entrants seeking to operate on the tied product market. For example, in the telecommunications industry, entrants can typically serve business customers and residential customers in dense urban areas at considerably lower cost than rural customers. These customers are therefore particularly important in determining whether a tie leads to exclusion of rivals in the tied product market. Large customers. Large customers in the tied product market that have independent demand for the tied product may be particularly important in determining the size of the residual market for the tied product. If their demand is sufficiently large, these customers may have the ability to guarantee rivals sufficient sales to remain as competitors in the market, or they may be able to sponsor new entry. Scope for product differentiation. The more that rivals in the tied product market can differentiate their products from the tied product of the monopolist, the greater their scope to make sales and to earn positive margins above marginal costs and thereby recover fixed costs. Note that greater heterogeneity in customer preferences will provide greater scope for product differentiation. Also, the greater the value to customers of the tied product relative to the tying product, the smaller the product differentiation in the tied product that will be needed to overcome the effects of the tie. Network effects. Network effects arise where a customer’s value for a product increases with the number of other users of the product. Where there are network effects in the tied product market, a tie may serve to “tip” the tied product market to the monopolist and deprive rivals of the opportunity to develop network externalities of their own. One part of the allegation in the Microsoft case brought by the US DOJ was that network effects led to a tipping of the market toward Microsoft’s Internet Explorer and away from Netscape at a critical moment in the development of the internet browser market. Two-sided markets. Two-sided markets are a variant of markets with network effects. In two-sided markets, such as (non-direct) advertising, card systems or platform software, one side of the market will be more attracted to the product the larger the size of the other side of the market; it is therefore necessary to get both sides of the market on board to be successful. For example, advertisers will pay more for advertising in a magazine the larger the number of readers; retail merchants will be more willing to pay for a

(R) Ehlermann 08 Coscelli and Edwards

416

6/11/08

11:52

Page 416

Session Two: A Reformed Approach to Exclusionary Conduct

card system the larger the number of retail customers that use the particular type of card; and application software developers will be more willing to write software for a particular platform the larger the platform base in the market. Tirole (2005) explains that there may be scope in two-sided markets for rival producers of the tied product to survive a tie on one side by a monopolist of the primary product, even if rivals cannot differentiate their products in the eyes of that side of the market, as there may be scope for differentiation in supply to the other side.58

4.1.3 Counter-strategies in the tied product market Rivals in the market for the tied product might be able to develop effective strategies to counter tying by a monopolist in the primary product market. For example, a rival in the tied product market may form its own tie with an alternative producer in the primary product market. In some circumstances it might even be possible for a rival in the tied product market to sponsor entry or to enter the primary market itself as a rival to the monopolist. Alternatively, the rival might look to tie its product with a third product that the monopolist in the primary market does not control. For example, a tie by a monopolist in print and television advertising might be countered by a rival in television advertising that ties its product with radio or internet advertising.59 Also relevant here, under many models of tying (though not all), is whether the monopolist’s tie is credibly irreversible.60 If it is not, rivals in the tied product market will be aware that if they remain in the market the monopolist will be likely to earn lower profits under the tying arrangement and will not rationally maintain the tie. Exclusion in these situations is therefore more likely to be effected if the monopolist makes a pre-commitment to the tie, such as through a technological tie in the product design. By contrast, contractual ties and economic ties (i.e., mixed bundling with a low implicit price for the tied product) can more easily be undone.

4.1.4 Is there a plausible dynamic theory of exclusion? Consider the possibility that a tie may be designed to exclude competition in the tying market or in a newly emerging market. The key question here is 58

See Tirole (2005), supra note 14, at pp. 9–10. Slade shows that Canadian newspapers with monopoly power in local areas tend to tie the provision of advertising services to the purchase of advertising space much more frequently than the newspapers published in multi-newspaper cities. Margaret Slade, “The Leverage Theory of Tying Revisited: Evidence from Newspaper Advertising”, 65 Southern Economic Journal 204 (1998). 60 As we have seen, in some models tying may be used to maximize short term profits, in which case there is no need to make an irreversible commitment to the tie. 59

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 417

Efficiency and Anticompetitive Effects of Tying 417 whether the tie increases the costs and reduces the prospects for entry into the monopolist’s primary market (the tying product market). In other words, “but for” the tie, would there be more competition in the market for the tying product? An allegation that the monopolist is seeking to exclude competition in the tied product market in order to retain its monopoly in the tying product market should be accompanied by a credible theory of how the tie reduces the prospect of greater competition in relation to the tying product. It seems to us that this theory of exclusion should consist of at least two elements. First, it needs to be established that existing barriers to entry in the tying product market are not insurmountable, and that in the absence of the tie it is possible, or perhaps probable, that rivals would enter that market and constrain the market power of the current monopolist. Second, the link between the tied product market and the tying product market must be explained, such that greater competition in the tied product market will facilitate entry into the tying product market. For example, in the Microsoft case, the US DOJ argued that, although Microsoft already enjoyed a near monopoly position in operating systems, it perceived a real threat from rival operating systems running Java-based applications. One line of argument was that, if Netscape were allowed to remain a significant player in the browser market, it would facilitate the development of Java-based applications that would run on any operating system, and over time this would attract more and more customers away from Microsoft’s operating system.

4.2 Theory of harm: will the tie cause long-run consumer harm? Once it has been established that exclusion has occurred or is likely to occur, it is necessary to assess whether that exclusion will cause consumers to suffer harm in the long run. Consumer harm does not follow exclusion automatically. As we have seen, tying that has exclusionary effects might generate efficiencies enjoyed directly by consumers or passed on to them indirectly, or it might facilitate price discrimination, which, as pointed out several times, has a priori ambiguous effects for consumer welfare. Therefore, to establish anticompetitive tying conduct it is necessary for the relevant Authority to demonstrate both a theory of exclusion and a theory of harm, and to support each theory with empirical evidence wherever possible. We suggest that an assessment of whether a tie will cause long-run harm to consumers should be conducted in two stages. • The first stage is to consider whether the circumstances surrounding the tie are consistent with any of the efficiency or price discrimination motivations presented in Section 2 of this paper (or any other non-strategic motivation that might be proposed). If no such explanations are presented by the

(R) Ehlermann 08 Coscelli and Edwards

418

6/11/08

11:52

Page 418

Session Two: A Reformed Approach to Exclusionary Conduct

dominant company, then it would seem reasonable for the relevant Authority to reach a presumption that the tying conduct is motivated by a strategy of exclusion and leverage of market power. • If non-strategic motivations are plausible, then since tying conduct could simultaneously serve both strategic and non-strategic purposes, with both negative and positive effects on consumers, the final stage is to perform a balancing assessment of these likely effects.

4.2.1 Threshold conditions: is there likely to be harm from exclusion? Threshold conditions for an investigation of tying behaviour follow the Chicago school insights reviewed earlier. Where there is strong competition in the tied product market, or where the two products are consumed in fixed proportions and the monopoly product is essential, so that there is only one monopoly rent across the two products, and where dynamic theories of exclusion (such as those relating to future streams of rents in the tying product, in newly emerging markets, or in upgrades of the tied product) do not appear plausible, the relevant Authority could presume that tying is unlikely to be motivated by a desire to exclude competitors and subsequently to lower output and raise prices in the tied product market. In these circumstances, even if rivals in the tied product market are excluded, efficiency or price discrimination motives for tying are more likely. For example, in the Genzyme case,61 the market for home care services (which includes home care for Gaucher patients and home care for patients with other illnesses) was highly competitive. Genzyme would not have been able to extract any additional rents by integrating the supply of Cerezyme with home care services for Gaucher patients, except to the extent that those rents derived from greater efficiency and quality assurance in the delivery of Cerezyme. This would be a case where it would seem reasonable to conclude that, even if exclusion of competitors in home care services might occur, consumers would not be harmed by the tie.62 We note that exclusion is more likely when there are high R&D costs, as the tie reduces the ability of rivals in the tied product market to recover these fixed costs. The tie may also discourage R&D aimed toward new products that rely on the tied product as a complementary product. Tying is therefore likely to be particularly problematic in industries characterized by innovation competition, as tying can dampen the incentives of rival firms to innovate. There might also be cases where the exclusionary effects of tying on actual and potential rivals are limited, as rivals will simply adapt their commercial offerings to the structure adopted by the main player in the market. The 61 62

Supra note 44. We discuss the OFT’s theory of harm in Section 3.3.1 above.

(R) Ehlermann 08 Coscelli and Edwards

6/11/08

11:52

Page 419

Efficiency and Anticompetitive Effects of Tying 419 question in these cases is whether there has been a significant detriment in consumer welfare as a result of moving from one competitive equilibrium to another.

4.2.2 Are there plausible efficiency rationales? If a review of the threshold conditions indicates that tying might be an exclusionary (predatory) anticompetitive strategy, then it is necessary to consider whether the conduct is motivated by efficiency rationales. This should involve careful consideration of each of the efficiency rationales discussed earlier in Section 2, and any further rationales that the impugned firm might suggest. That is, we agree with Philip Lowe’s comment that the dominant firm needs to explain to the relevant Authority its “theory of benefit”. Again, an emphasis should be placed on empirical verification and quantification of efficiencies. If no efficiency rationale appears plausible (including any efficiency rationale advanced by the dominant firm), then it would be reasonable for the relevant Authority to conclude that the tie has an anticompetitive motive and would have harmful effects on long-run consumer welfare.63

4.2.3 Balancing assessment If efficiency or price discrimination rationales are identified as plausible, it is then appropriate to balance the conduct’s likely benefits for consumers against the likely harm. One important issue to consider here is whether an alternative method was open to the monopolist to achieve the claimed efficiencies with less (or no) anticompetitive effects, in line with the EU principle of “objective necessity” as applied by the European Commission in the Microsoft case.64 If it can be established that the tying arrangement is indispensable to achieving the claimed efficiencies, the next issue is the extent to which those efficiencies are passed through to final consumers.

4.3 Policy considerations We believe that detailed guidelines on the application of Article 82, and in particular on how the Commission will approach future tying cases, would be extremely useful and timely. In that regard, we agree with the comments made 63 We note that it is not so simple to reach a conclusion on the effects of the tie on overall social welfare. Where there are fixed costs in production and downward sloping average cost curves, exclusion may save society the duplication of fixed costs. 64 See Christian Ahlborn, David Bailey and Helen Crossley, “An Antitrust Analysis of Tying: Position Paper”, GCLC Research Papers on Article 82 EC 166, 196 (2005).

(R) Ehlermann 08 Coscelli and Edwards

420

6/11/08

11:52

Page 420

Session Two: A Reformed Approach to Exclusionary Conduct

by Philip Lowe and by Damien Neven, which seem to indicate that such guidelines may indeed be expected in the future. We believe these guidelines should follow the very helpful format adopted in the recent draft guidelines on the assessment of non-horizontal mergers.65 The reason is that we believe it would be useful at this stage to provide the (national) Authorities, as well as lawyers and economists working on these matters in the EU, with detailed guidelines describing the issues the Commission would focus on when analyzing specific cases. Some legal commentators have argued that guidelines are not necessary and that the Authorities should simply bring cases and let the Community Courts change the case law on tying (if necessary) through their judgments. Even leaving aside the current time-lag between the Authorities’ decisions and the Courts’ judgments in non-merger cases, we disagree with this approach, as we believe that tying cases are very fact-specific and that it is difficult to provide correct guidance through a few judgments on complicated tying cases.

65 Available at http://ec.europa.eu/comm/competition/mergers/legislation/draft_nonhorizontal_ mergers.pdf.

(S) Ehlermann 08 Rey

6/11/08

11:52

Page 421

II Patrick Rey* On the Right Test for Exclusive Dealing June 2007

I. Background: Modernizing the enforcement of Article 821 The EAGCP report argued in favour of an economics-based approach to Article 82 to match the reforms made in connection with Article 81 and merger control. In particular, we supported an effects-based rather than a form-based approach to competition policy. Such an effects-based approach should focus on the presence of anticompetitive effects that harm consumers, and it should be based on an examination of the economic reality of the markets in question; it should rely on sound economics, and it should be grounded on facts. One of the main merits of an effects-based approach is that it avoids confusing the protection of competition with the protection of competitors. Competition is a means to an end, which is the satisfaction of consumer needs. It is a process that forces firms to be responsive to consumers’ needs with respect to price, quality, variety, and so forth. Over time, it also acts as a selection mechanism, with more efficient firms replacing less efficient ones. The application of competition rules should therefore contribute to fostering this process and, in particular, it should place consumers at the heart of the analysis when agreements and conduct are assessed. The Court of First Instance recalled this principle in September 2006 in a case involving a restriction on parallel trade,2 an area in which the overriding principle of market integration had until then led the Commission (and the Community Courts) to follow a quasi-per se illegality approach. The CFI asserted that, although “parallel trade must be given a certain protection,” this is the case only insofar as “it gives final consumers the advantages of effective competition in terms of supply or price”.3 The court then added:

* University of Toulouse 1 This introduction builds partly on the EAGCP report “An economic approach to Article 82”, available at http://ec.europa.eu/comm/competition/publications/studies/eagcp_july_21_ 05.pdf. All errors or omissions are mine. 2 Case T-168/01, GlaxoSmithKline Services v Commission [2006] ECR II-2969; on appeal: Cases C-501/06, C-513/06, C-515/06 and C-519/06, not yet decided. 3 Ibid., para. 121.

(S) Ehlermann 08 Rey

422

6/11/08

11:52

Page 422

Session Two: A Reformed Approach to Exclusionary Conduct

“Consequently, while it is accepted that an agreement intended to limit parallel trade must in principle be considered to have as its object the restriction of competition, that applies [only] in so far as the agreement may be presumed to deprive final consumers of those advantages.”4

These observations, and the specific situation of the markets concerned, led the court to reject the qualification of a restriction by object in the application of Article 81(1), and to condemn the Commission for failing to review in detail the economic evidence and reasoning put forward by the parties when applying Article 81(3). Thus, the court rejected the traditional doctrinal stance that any measures restricting parallel trade must automatically be condemned as inimical to the single market. That is, even for parallel trade restrictions—one of the last two types of restrictions (together with minimum resale price maintenance) which are still blacklisted by the Commission after its review of the application of Article 81 to vertical agreements—the court is urging the Commission to take proper account of the actual or likely effects of those provisions on the markets. The application of Article 82 to exclusive dealing and other potentially exclusionary practices should reflect the same principle: it should foster the competitive process as a means to enhance consumer welfare, rather than as a means to protect particular competitors, suppliers or intermediaries. This is all the more important—and unfortunately all the more difficult—given that, like most of the practices usually falling under Article 82, exclusive dealing can also, in many contexts, contribute to enhancing efficiency and foster innovation, and can thus serve rather than impede the competitive process. Thus, for example, a predator may try to get rid of a rival through a price war; and yet, one of the most often mentioned benefits of competition is precisely to exert downward pressure on prices. Thus, one cannot decide whether a “low” price is, by itself, a good thing or a bad thing. Similarly, tying is present everywhere: who is willing to build her or his own watch, car, etc., from assorted spare parts? Still, in specific circumstances, the same practice may serve anticompetitive purposes. Exclusive dealing is no exception, since, as further discussed below, it can be adopted for a series of efficiency-based reasons but can also be used in some situations to raise entry barriers and distort competition. Given this basic ambiguity, a purely form-based approach is likely to either generate many false positives, which would impede the competitive process and thus harm consumers, or to miss those cases in which a practice is driven by anticompetitive purposes. In contrast, an effects-based approach ensures that competition rules will be enforced in an effective manner, without unduly thwarting pro-competitive strategies. By the same token, it generates consistency, since any specific practice is assessed in terms of its outcome, so that 4

Ibid.

(S) Ehlermann 08 Rey

6/11/08

11:52

Page 423

On the Right Test for Exclusive Dealing 423 practices leading to the same result will be subject to a comparable treatment. This makes it more difficult for companies to circumvent competition policy constraints by playing around with different commercial practices. In terms of procedure, these considerations led us in the EAGCP to emphasize the need for a verifiable and consistent account of significant competitive harm, based on sound analysis and grounded on facts. This requirement is an essential step for distinguishing when a practice is indeed used for anticompetitive purposes rather than for pro-competitive, efficiency-enhancing considerations. In particular, since many practices can have pro-competitive as well as anticompetitive effects, merely alluding to the possibility of an anticompetitive story is not sufficient. The required ingredients of the story must therefore be properly spelled out and shown to be present. While this first step may be perceived as somewhat constraining the competition authority’s leeway, it is however necessary to ensure the consistency of the treatment of the various practices that can serve the same anticompetitive effects. It also contributes to enhancing predictability and, consequently, it contributes to providing legal certainty and making competition policy enforcement more effective. The intrinsic ambiguity emphasized above also counsels in favour of a balanced approach, where potential pro- and anticompetitive effects are treated equally (although possibly assessed sequentially, starting with anticompetitive effects). This would contrast with an “efficiency defence” approach, since that approach would tend in practice to put more weight on anticompetitive effects. Furthermore, in contrast to a form-based approach, an effects-based approach needs to put less weight on a separate verification of dominance, except as a screening device. The structural indicators traditionally used in the first steps of the procedure are merely proxies for dominance. They may provide an appropriate measure of market power in some instances, but not in others, in which case competition authorities’ intervention is likely to be inappropriate. In contrast, if an effects-based approach provides evidence of an abuse which is only possible if the firm has a position of dominance, that can be regarded as more direct and reliable evidence of dominance. This is not to say that traditional considerations about the presence or absence of dominance become moot; rather, they become part of the procedure for establishing competitive harm resulting from the practice under investigation. Likewise, an effects-based approach naturally captures the notion of “special responsibility” in the sense that certain practices would be prohibited when carried out by a dominant firm but would be lawful if practiced by smaller competitors. Focusing on the exclusionary effects of market practices gives a robust foundation to this notion, since certain practices will then be prohibited when they generate exclusionary effects, whereas they will be permitted as long as no competitive harm is involved. Below I discuss the implications of relying on an effects-based approach in exclusive dealing cases.

(S) Ehlermann 08 Rey

424

6/11/08

11:52

Page 424

Session Two: A Reformed Approach to Exclusionary Conduct

II. An effects-based approach to exclusive dealing As discussed above, when a complaint arises or when a competition authority suspects an abuse of dominance, the first question should be: what is the nature of the competitive harm involved in that case? Insisting on a consistent anticompetitive scenario brings several benefits. First, while creative imagination can generate an infinite number of practices, there are not that many types of competitive harm and, for each one, the established toolbox of relevant, consistent arguments is relatively limited. Second, this requirement allows a clear identification of the key facts that need to be checked. Third, this approach guarantees a consistent treatment of alternative practices that could serve the same anticompetitive purpose. Thus, overall, identifying the nature of the competitive harm at stake can facilitate and speed up the investigation process while promoting high levels of predictability.

1. The Chicago critique: capability versus incentives Since many practices can serve pro-competitive, efficiency-enhancing purposes as well as anticompetitive purposes, alluding to the mere possibility of an exclusionary effect is not sufficient. One needs to spell out a scenario in which the dominant firm would indeed have an incentive to engage in the practice for anticompetitive reasons. While this may seem a rather weak requirement (after all, why would a firm pass up an opportunity to enjoy extra market power?), this is not as obvious as it may sound. Consider, for example, the case of a manufacturer supplying a good to a given customer, and denote by C the cost of producing the good and by S the surplus it gives the customer. Suppose further that a potential entrant may enter with a superior technology that yields a cost advantage, A. In the absence of entry, the incumbent manufacturer could exploit its market power and appropriate the full value of the good it produces (by charging a price equal to the customer’s surplus, S) and thus earn a profit equal to S – C. If instead entry occurs, competition drives the price down to the incumbent’s cost (which represents the best price it can offer), C. The customer then obtains a net surplus equal to S – C, while the entrant drives the less efficient incumbent out of the market and earns a profit reflecting its cost advantage, A. To prevent entry, the incumbent manufacturer could try to lock in the user by means of an exclusive dealing contract. However, anticipating that it will no longer benefit from competition if it signs an exclusive contract, the customer will not accept an exclusive contract at a price higher than C. It follows that such an exclusive dealing contract cannot be profitable for the manu-

(S) Ehlermann 08 Rey

6/11/08

11:52

Page 425

On the Right Test for Exclusive Dealing 425 facturer. That is, there is no exclusive dealing contract that would be both profitable for the manufacturer and acceptable to the customer. This illustration is the essence of the so-called “Chicago critique” to many traditional exclusionary concerns; it generalizes to a variety of situations in which competing manufacturers deal with several customers or intermediaries such as retailers. Imposing, for example, a “single-branding” obligation on a retailer, thus preventing that retailer from carrying rival brands, certainly appears at first glance to benefit the manufacturer. By the same token, this benefit for the manufacturer would appear to come at the expense of the retailer (at least in the absence of any negative externalities from one brand to another, as well as any diseconomies of scope that may arise from carrying multiple brands, and so forth) and at the expense of consumers. However, the retailer will be unwilling to accept the single-branding obligation without compensation. The general thrust of the argument is that there is a single source of profit in the vertical structure as a whole, and moreover it “takes two to tango”: therefore, a restraint that would exclude rivals but deprive contracting partners from the additional profits that they could derive from dealing with the rivals in question is unlikely ever to be adopted—unless, the critique goes, it generates some efficiency in the relationship, in which case antitrust authorities should not step in. In particular, a manufacturer that is tempted to impose exclusive dealing on its retailers, for example, must take into account the compensation it will have to grant to these retailers for their lost trade opportunities; put another way, it will have to take into account that it could extract better terms from these retailers if this loss of profitability were not imposed on them. This critique has triggered economists to reconsider the foreclosure argument and to put it on firmer ground.5 But it demonstrates the abovementioned need to not only describe the possible effects of exclusive dealing on rivals, but moreover to spell out a consistent scenario explaining the anticompetitive rationale for the parties involved. That it to say, capability is not sufficient—all the more so since many practices can have both adverse effects and positive ones—due attention must be paid as well to the parties’ incentives to engage in this or that practice. This can be contrasted with some of the wording of the Discussion Paper that was circulated by DG Comp.6 For example, paragraphs 58–59 refer to 5 For example, Aghion and Bolton have shown that, in contexts such as that illustrated by the above simple example, the incumbent manufacturer could use semi-exclusive contracts (e.g., in the form of penalties for breach) to extract some of the entrant’s efficiency gains, which can then be shared with the customer; furthermore, in the presence of uncertainty about the magnitude of these efficiency gains, the practice may well result in the exclusion of moderately more efficient rivals. See Philippe Aghion and Patrick Bolton, “Contracts as a Barrier to Entry”, 77 American Economic Review 388 (1987). 6 See DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses, document released for public consultation in December 2005, available at the Commission’s website: http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf.

(S) Ehlermann 08 Rey

426

6/11/08

11:52

Page 426

Session Two: A Reformed Approach to Exclusionary Conduct

dominance, capability, incidence, network effects, scale/scope economies, and so on, and then the passage concludes that these factors will lead to the presumption of “likely effects”. This basically switches the burden of proof to the defendant, who must then establish an “efficiency defence”, without any prior account made by the enforcer of any specific anticompetitive “story”. This can be seen, for example, in paragraph 149 on single-branding obligations, which states: “Where the dominant company applies [such an] obligation to a . . . substantial part of the market, the Commission is likely to conclude that the obligation has a market distorting foreclosure effect and thus constitutes an abuse of the dominant position.”

To be sure, one could interpret this sentence as suggesting that a properly reasoned analysis will indeed be likely when ascertaining the existence of a consistent exclusionary scenario; however, the same paragraph 149 then goes on to state: “In its assessment the Commission will not only look at the capability of the obligation, the degree of dominance and level of the tied market share, but will also take into account evidence why for particular reasons no market distorting foreclosure effect may result.”

This tends to turn the analysis “upside down”: once “capability” is established, rather than insisting on a consistent scenario of an actual anticompetitive effect, or studying possible alternative efficiency motivations, the paragraph merely asks whether there are “particular reasons” why the presumed market distorting effects would not be there.

2. Anticompetitive scenarios: vertical foreclosure The kind of competitive harm usually involved in exclusive dealing cases corresponds to “vertical foreclosure”, i.e. the exclusion of a competitor from a market that is vertically related to the market on which the incumbent firm is dominant. Usual features characterizing vertical foreclosure concerns are the following: (i) the firm in question controls a “bottleneck” facility, that is, an input that is necessary to operate in upstream or downstream markets; and (ii) the firm forecloses these vertically related markets by denying or otherwise limiting access to the bottleneck input. For example, the owner of infrastructure such as rail tracks or port facilities may deny access to this infrastructure to rail transportation service operators, thereby preventing them from providing their services in an effective way. In other cases, a key retailer may control access to consumers in a particular area; entering into an exclusive purchasing deal with a given manufacturer then de facto denies rivals access to those consumers.

(S) Ehlermann 08 Rey

6/11/08

11:52

Page 427

On the Right Test for Exclusive Dealing 427 Exclusive dealing is only one among many other instruments in the “forecloser’s toolbox”. The bottleneck owner can, for example, integrate vertically into the target market and refuse to deal with potential competitors, as in the case of airlines’ computerized reservation systems. Alternatively, the forecloser may make the bottleneck input incompatible with competitors’ products or technologies, or it may engage in tie-ins and refuse to unbundled the tied products. In the presence of economies of scope or scale requiring cooperation among firms in the same market, a dominant group of firms may put their competitors at a disadvantage by refusing to cooperate. In the absence of integration, and short of exclusive deals with given partners, the bottleneck owner can favour some competitors over others; this discrimination can be explicit, e.g., through individualized rebates, or implicit, e.g., through loyalty programmes or growth-based rebates that are formally available to all but tailored to the needs of specific users. Similarly, substantial quantity discounts may allow the survival of only a few customers; for instance, a large enough fixed fee can transform a potentially competitive downstream industry into a natural monopoly industry. The traditional foreclosure concern is that the owner of the bottleneck input may leverage its market power in related markets. There again, however, the Chicago School critique applies: there is a single final market and thus only one profit to be reaped, which the dominant firm can secure by exerting its market power over the bottleneck. Therefore, the dominant firm has no incentive as such to distort competition in the other markets. On the contrary, imperfect competition in these markets may actually create distortions and reduce the profitability of the bottleneck, e.g., by reducing the variety or quality of the goods and services produced. While the Chicago critique is correct, anticompetitive effects may still arise in specific circumstances.7 For example, the bottleneck owner may deter competition in a vertically related market to protect its home market. This may happen in situations where entry in the home market is facilitated by competition in the other market.8 Alternatively, a bottleneck owner may face a commitment problem, which makes it difficult for the firm to exercise its monopoly power without engaging in exclusionary practices: once it has sold access to a first downstream firm, it has an incentive to provide access to other firms as well, even though downstream competition will then reduce the first competitor’s profits; this opportunistic behaviour will however be anticipated, and this reduces the bottleneck owner’s profit. More generally, competition in downstream markets may “percolate” upstream and prevent the upstream bottleneck firm from making a profit. To address this commitment 7 For an overview of the modern economic literature on vertical and horizontal foreclosure, see Patrick Rey and Jean Tirole, “A Primer on Foreclosure”, in Mark Armstrong and Robert Porter, eds., Handbook of Industrial Organization, 3rd edition, North Holland, 2007. 8 See Dennis Carlton and Michael Waldman, “The Strategic Use of Tying to Preserve and Create Market Power in Evolving Industries”, 33 Rand Journal of Economics 194 (2002).

(S) Ehlermann 08 Rey

428

6/11/08

11:52

Page 428

Session Two: A Reformed Approach to Exclusionary Conduct

problem, the bottleneck owner may then wish to restrict or eliminate competition in the downstream markets through the types of practices mentioned above,9 e.g., by entering into an exclusive dealing agreement with a particular firm—in order to restore its market power in the upstream market, rather than to leverage it in the downstream market. However, it should be kept in mind that here again, exclusive dealing constitutes only one of the possible options open to the owner of the bottleneck. In this context, it should be pointed out that banning discrimination would help the bottleneck owner to resist demands for selective price cuts, and that this would contribute to the maintenance of high prices. Vertical integration also constitutes an alternative solution to the upstream firm’s commitment problem. Interestingly, this line of reasoning suggests that foreclosure may be more of a concern when the bottleneck lies upstream rather than downstream, where “downstream” and “upstream” are defined with reference to the proximity of end users—the downstream segment being the closer to final consumers. Indeed, the opportunism described above applies when an upstream dominant firm deals with competing downstream firms. It does not arise when the dominant firm is in the downstream segment, at the direct interface with final consumers, since the firm then enjoys both monopoly power vis-à-vis consumers and monopsony power vis-à-vis its suppliers. The firm therefore does not need to distort competition in the upstream market in order to exert its market power (on both sides), and would instead presumably benefit from intense competition among its suppliers; any access restriction is then likely to be driven by efficiency reasons. By contrast, if the dominant firm is upstream and it relies on downstream intermediaries to reach final consumers, then, as just discussed, competition for final consumers among these downstream intermediaries may dissipate profits and prevent the upstream bottleneck from fully exercising its market power; in that case, the upstream firm may indeed have an incentive to distort or limit competition in the downstream market. This “scenario” may therefore support foreclosure concerns in the case of access to an essential facility, but not when, say, a key retailer or travel agency enters into special agreements with particular product suppliers or service providers. Another possible scenario relies on buyers’ coordination problems,10 in situations where new entrants need a minimum share of the market in order to operate a viable business (e.g., because of large fixed costs). Suppose, for example, that there are multiple buyers and that an incumbent supplier offers each buyer a (small) rebate in exchange for exclusive dealing. A buyer will accept this offer if the rebate is larger than his expected loss due to the reduced 9 See Oliver Hart and Jean Tirole, “Vertical Integration and Market Foreclosure”, Brookings Papers on Economic Activity (Microeconomics) 205 (1990). 10 See, e.g., Aghion and Bolton, supra note 5; Eric B. Rasmussen, J. Mark Ramseyer and John S. Wiley, “Naked Exclusion”, in 81 American Economic Review 1137 (1991); Ilya Segal and Michael Whinston, “Naked Exclusion: Comment”, 90(1) American Economic Review 296 (2000).

(S) Ehlermann 08 Rey

6/11/08

11:52

Page 429

On the Right Test for Exclusive Dealing 429 likelihood of entry, but typically he will not take into account the negative externality on the other buyers’ resulting restriction of choices. If buyers cannot coordinate their behaviour, entry might then be deterred even though it would benefit buyers as a whole. Exclusive dealing contracts can thus be anticompetitive when buyer coordination is needed. While this line of reasoning was initially developed in the contexts where the “buyers” were not competing against each other, it is tempting to apply it as well to situations where the “buyers” are themselves competing in a downstream market, as in the case of retailers that may carry one or several brands. Interestingly, however, Fumagalli and Motta have stressed that, in such a case, there may be less of a need for buyer “coordination” when one retailer, for example, is able on its own to cover a substantial part of the market when being offered competitive wholesale conditions.11 In that case, the incumbent would have to lock in most if not all of the retailers, and then the original Chicago critique applies again. Spelling out a precise anticompetitive scenario helps to identify the required key ingredients. For example, the last scenario just described requires a number of facts: • there must be many small buyers; • there must be no communication or coordination between them on procurement of inputs. This would not be the case where, for instance, supermarkets delegate the power of listing products to a centralized agency and even negotiate the transaction on aggregate sales with this agency; • “buyers” must not be competing against each other, or else they must be unable to increase much their respective market shares even when being offered a competitive advantage over rival retailers; and • on the supply side, the magnitude of the efficient size for an entrant should also be evaluated. If it appears that a competitor can enter the market even if it sells only a small number of units, for instance because the required fixed costs are small, then exclusive dealing contracts are unlikely to prevent entry. On the other hand, exclusive dealing contracts are more likely to be anticompetitive where there are large fixed costs. As can be seen from this example, spelling out the specific anticompetitive scenario not only permits a consistency check, it also helps to identify those particular instances where there is a serious risk of consumer harm. In the various anticompetitive scenarios described above, the intervention of competition authorities may foster competition in the related markets and thus in the industry as a whole. This intervention can benefit consumers, e.g., through lower prices in the short run or through higher rates of innovation in 11 See Chiara Fumagalli and Massimo Motta, “Buyers’ coordination, exclusive dealing and entry, when buyers compete”, London CEPR Discussion Paper (2002) and 96 American Economic Review 785 (2006).

(S) Ehlermann 08 Rey

430

6/11/08

11:52

Page 430

Session Two: A Reformed Approach to Exclusionary Conduct

the related markets in the long run. However, such intervention also affects the bottleneck owner’s rate of return. In the long run it may thus have an adverse impact on that firm’s incentives to invest or innovate, and it may, for example, impede the development of key infrastructure. Competition authorities may therefore wish to refrain from prosecuting foreclosure conduct when it compensates the bottleneck owner for its investment or innovative activity. This is similar to the logic underlying the patent system: prospective licensees are not willing to pay much for the use of a new technology if they know that the licensor will “flood the market” with similar licences; therefore, mandating access through additional licences would reduce the innovator’s profitability and consequently its incentives to invest in R&D. The source of the bottleneck owner’s market power may thus help to determine when or whether a competition authority should intervene. In particular, intervention aimed at preventing foreclosure, and consequently at reducing the bottleneck owner’s profit, seems more warranted when the firm’s market power derives from increasing returns to scale or scope (as may be the case with respect to a bridge, a stadium, or a news agency), or when it derives from a historical accident, than when it results from an innovative strategy.

3. Pro-competitive motivations Competition authorities should also verify whether there are efficiency defences justifying the exclusive dealing arrangement. In particular, even when exclusive dealing generates anticompetitive effects, it may still give rise to countervailing efficiencies. Competition authorities should intervene only if this possibility can be ruled out. Various lines of efficiency defences can be put forward for exclusive dealing arrangements. For example: —Free-riding on the investment of the dominant firm. This argument is a variant of the argument described above of forbearance as a reward to investment, and it is basically linked to the need of the dominant firm to recoup marketing or similar expenses that benefit its partners. —Protecting reputation. A related argument stems from the fear of being associated with inferior products or services which might hurt the firm’s reputation. Misbehaviour by a rival may spoil the reputation of the dominant firm as well as that of its other partners. —Relationship-specific investments. A manufacturer may, for example, refrain from investing in information and training for its distributors if this investment can then be used to benefit its rivals. The manufacturer may then insist on an exclusive arrangement in order to protect such an investment. Similarly, a manufacturer may adopt an exclusivity clause in order

(S) Ehlermann 08 Rey

6/11/08

11:52

Page 431

On the Right Test for Exclusive Dealing 431 to promote “retailer loyalty”, i.e. to encourage the retailer to tailor its promotional efforts to the manufacturer’s product. —Excessive entry. In the absence of foreclosure, excessive entry might occur, generating an inefficient duplication of fixed costs, due to so-called “business-stealing” effects. When contemplating whether to enter the relevant market, a firm does not take into account the fact that part of its prospective customers will simply switch away from existing products; the revenue generated by the entrant’s product may thus exceed its social value.12 However, the validity of this argument may be difficult to assess in practice, since the characterization of the socially optimal number of firms is generally a complex matter. —More intense head-to-head competition. In addition to the foregoing efficiency considerations, exclusive dealing may well lead rivals to compete more intensely against each other. In particular, competition “for the market” may lead a manufacturer to grant better wholesale terms to retailers, which are then at least partly passed on to consumers. Thus, one should not necessarily conclude, from the mere observation of actual market shares, that competition is not working in the interest of end users. In addition, competition “in the market” between rival manufacturers, each engaged in exclusive deals with different retail networks, may well be more effective in disciplining firms and contributing to consumer welfare than when rival brands are carried by the same retailer, which can act as a “common agent” for supposedly competing manufacturers. There again, the diversity of arguments and of cases in which they may or may not be relevant calls for an assessment based on the economic reality of the market in which exclusive dealing takes place.

III. Concluding remarks To sum up, there is a basic ambiguity regarding most of the practices usually involved in Article 82 cases. This ambiguity calls for an effects-based approach that balances pro- and anticompetitive effects. Such an approach requires as a first step a consistent story of anticompetitive behaviour resulting in consumer harm, based on sound reasoning and supported by the facts of the case. Spelling out a coherent story, which should moreover distinguish between mere capability and incentives to engage in anticompetitive conduct, ensures consistent enforcement. 12 For detailed analyses of this issue, see Steven Salop, “Monopolistic Competition with Outside Goods”, 10 Bell Journal of Economics 141 (1979); N. Gregory Mankiw and Michael D. Whinston, “Free Entry and Social Inefficiency”, 17(1) Rand Journal of Economics 48 (1986).

(S) Ehlermann 08 Rey

432

6/11/08

11:52

Page 432

Session Two: A Reformed Approach to Exclusionary Conduct

This approach can be contrasted with the analysis proposed in DG Competition’s Discussion Paper. For example, when dealing with all-units conditional rebates, the Discussion Paper quickly focuses on the so-called “required share”, defined as the “share of customers’ requirements on average the entrant at least should capture so that the effective price is at least as high as the average total cost of the dominant company”.13 This prompts several queries. For example, this presumes that the most effective way for entering the market consists of seeking a comparable share of the various consumers’ orders. Yet in practice, one would expect a new entrant to “target” specific customers (e.g., those that are more inclined to be interested in the entrant’s alternative offer), rather than trying to supply all of them uniformly; in that case, the targeted customers would be more likely to meet the “required share”, while the other ones, for which the threshold might not be met, would be less relevant anyway. In addition, the focus on “average total cost”, rather than the alternative cost measures suggested for other practices, raises some concern about the overall consistency of the proposed approach across practices that may well be used interchangeably for similar anticompetitive purposes. But more importantly, this proposed approach bypasses any account of why the firm adopted such rebate schemes in the first place. The discussion focuses on “capability”, illustrated by methodological boxes on how to measure required shares, without discussing whether or how the firm could and would successfully offer these rebates for anticompetitive purposes. The Discussion Paper then proposes that the Commission should assess the “commercially viable share” that an efficient competitor or entrant can be expected to supply.14 This again calls into question the presumed “entry scenario”, which appears to rely on more or less uniform supply of the various customers. In addition, no mention is made of a need to explain the reasons why an efficient competitor could not achieve the same share as the incumbent. This reinforces the risk of confusion between protecting competition, for the benefit of consumers, and protecting competitors. Not only should these reasons be accounted for, one would moreover need to check their consistency with a coherent story of anticompetitive harm. 13 See DG Discussion Paper, supra note 6, at para. 155: “As a first step the Commission will endeavour to calculate how big a share of customers’ requirements on average the entrant at least should capture so that the effective price is at least as high as the average total cost of the dominant company (“the required share”). [. . .] In case the shares of the customers’ requirements purchased from actual rivals are smaller than the required share, the rebate scheme is likely to have a foreclosure effect where there is in addition no indication that these rivals are less efficient.” 14 See ibid., para. 156: “[T]he Commission will endeavour to assess the commercially viable share an efficient competitor or entrant can be expected to supply and to compare this with the required share. . . . Where the required share exceeds the commercially viable share the rebate system is likely to have a foreclosure effect which reduces competition as the effective price that results from the rebate system over this commercially viable share will be below the average total cost of the dominant company.”

(T) Ehlermann 08 Melamed

6/11/08

11:52

Page 433

III A. Douglas Melamed 1 Thoughts about Exclusive Dealing July 2007

Exclusive dealing, which in Europe is sometimes called “single branding”, refers to arrangements pursuant to which a firm (“the Firm”) induces an input supplier to deal with it but not with its competitors. In its simplest and perhaps most common form, a manufacturer enters into an agreement with a retailer pursuant to which the retailer purchases and resells the manufacturer’s products and agrees not to deal in competing products.2 The term “exclusive dealing” can also be used to refer to less precise and formal arrangements, such as those that provide for partial exclusivity (e.g., where at least 75% of the retailer’s sales must be of the Firm’s products) or induce exclusivity by non-contractual threats (e.g., where the Firm announces that it will not deal with retailers that do business with its competitors).

Injury to Competition and the Easy Cases Exclusive dealing can injure competition if it excludes competitors of the Firm or raises their costs and thereby enables the Firm to gain or maintain market power that it would otherwise not have. Exclusive dealing can exclude or raise the costs of competitors by denying them access to needed inputs, forcing them to obtain more costly inputs, or causing their sales to diminish and thus driving them below efficient scale. Proof that the exclusive dealing has injured or is likely to injure competition in this way should be a precondition to condemnation of the exclusive dealing under the antitrust or competition laws. Absent such effects, exclusive dealing is benign from an antitrust or competition law perspective; and

1 Partner, Wilmer Cutler Pickering Hale and Dorr LLP. Einer Elhauge, Rick WarrenBoulton and Greg Werden provided valuable comments on an earlier draft of this paper. 2 In this example, the retailer is the input supplier, and the input is retail distribution services. While it might seem odd to refer to a downstream firm as an input supplier, the uniform use of the term “input supplier” simplifies exposition by enabling analysis of the issue without regard to whether the constrained entity is upstream or downstream from the firm that induced the exclusivity.

(T) Ehlermann 08 Melamed

434

6/11/08

11:52

Page 434

Session Two: A Reformed Approach to Exclusionary Conduct

because (as discussed below) exclusive dealing can be efficient and can benefit consumers, condemning it without proof of injury to competition will almost surely result in false positives that would reduce efficiency, competition and welfare. On the other hand, exclusive dealing can be readily condemned under the antitrust laws if (i) it causes actual or likely injury to competition, and (ii) the exclusive dealing provides no plausible efficiency benefit under the circumstances. If these conditions are proven, exclusive dealing can be regarded as a form of “naked exclusion”—conduct that excludes rivals, creates or maintains market power and cannot be justified.3

Benefits from Exclusive Dealing Exclusive dealing can have a variety of benefits for efficiency and competition. In fact, because exclusive dealing is commonly used in circumstances in which there appears to be no realistic prospect that it will create market power, it is presumably widely thought to provide efficiency benefits. Some might suggest that efficiency should be the default explanation for exclusive dealing. Exclusive dealing can provide numerous types of benefits, not all of which can be catalogued here. Broadly speaking, the most common benefits appear to fall into two related categories. The first has to do with aligning the incentives of the Firm and the input supplier. For example, if a retailer sells the Firm’s products and not those of its competitors, it will have an incentive to maximize the sales of those products, and the Firm will thus have increased incentives to deal with the retailer. The second has to do with protecting and thus inducing efficient, relationship-specific investments. For example, a distributor (the Firm, in this example) might need assurance that the manufacturer will not dilute the distributor’s sales by dealing with multiple distributors before it will invest in a new and more efficient distribution facility. Or the Firm might need assurance that a retailer will not divert customers

3 As a logical matter, one could reverse the presumption implicit in these two paragraphs and argue that the law should condemn all exclusive dealing that cannot be shown to have efficiency benefits under the circumstances, without regard to proof of injury to competition, in order to guard against the risk of false negatives. US law wisely rejects this approach (i) because efficiencies from exclusive dealing are so common that an injury-to-competition screen is needed to guard against false positives, and (ii) because of a more generalized belief that government should refrain from intervening in commercial arrangements absent proof of a good reason to intervene (a belief which in turn derives from the risk of errors of any such intervention and of capture of the government by factions, the direct costs of frequent government intervention, and indirect demoralization costs of frequent government intervention).

(T) Ehlermann 08 Melamed

6/11/08

11:52

Page 435

Thoughts about Exclusive Dealing

435

to competitors’ products before it will support a promotional campaign designed to attract more customers to the retailer. Although exclusive dealing can provide benefits, it might not be necessary for those benefits. For example, under some circumstances, the Firm might be able to align retailer incentives for point of sale services by paying directly for those services, rather than by requiring the retailer not to deal with competitors. Or the Firm might reasonably need exclusive relationships with a number of distributors but might not need such relationships with every distributor. For purposes of this paper, exclusive dealing will be regarded as providing benefits only if there is no other way that comparable benefits could be provided without a material increase in cost and with materially less risk of injury to competition. A Firm that is likely to obtain market power as a result of exclusive dealing is, all other things being equal, more likely to use exclusive dealing to achieve benefits that could be achieved in other and perhaps less restrictive ways. Thus, a careful inquiry as to whether exclusive dealing is really needed to obtain the benefits is probably most appropriate where the incremental market power created or likely to be created by the exclusive dealing is greatest.

Balancing Harms and Benefits The difficult question for antitrust law is how to treat exclusive dealing that both injures competition and provides benefits. As explained above, if it does not injure competition or if it does injure competition but provides no benefits, the antitrust issue is simple. But what if exclusive dealing both causes harm and provides benefits? (1) One approach would condemn exclusive dealing if the injury to competition were substantial, regardless of the benefits. There is language in DG Competition’s Discussion Paper on Article 824 which appears to support a condemn-if-harmful approach.5 Such an approach could be defended on normative grounds (i.e., taking the view that preserving rivalry is preferable regardless of its effects on economic welfare) or on the empirical ground that

4 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf. 5 See ibid., para. 54 (“The essential objective of Article 82 . . . is the protection of competition . . . . [and] to ensure that these competitors are able to expand in or enter the market and compete . . . .”); para. 84 (efficiency defense applicable only if “competition in respect of a substantial part of the products concerned is not eliminated”); para. 91 (“Ultimately the protection of rivalry and the competitive process is given priority over possible pro-competitive efficiency gains.”)

(T) Ehlermann 08 Melamed

436

6/11/08

11:52

Page 436

Session Two: A Reformed Approach to Exclusionary Conduct

such a rule is most likely to enhance welfare.6 The normative rationale is, at least from a US perspective, not an appropriate or desirable objective of the antitrust laws, which are intended to prevent conduct that impairs economic welfare. The empirical rationale, which is in effect a categorical judgment that the harm outweighs the benefits in general, seems to be little more than a guess of, at best, dubious accuracy. To be sure, a short-term efficiency gain can be outweighed by long-term welfare losses from resulting market power in individual cases. But, especially in times of rapid technological change and economic transformation, market power will often prove to be transitory and will often be eroded by forces that cannot be anticipated. A law that prohibited efficient exclusive dealing whenever it seems likely to injure competition would thus deter a wide range of efficient conduct that might lead, at most, to only shortlived market power. Perversely, such a law would be especially likely to deter the most efficient conduct (because that conduct is most likely to disadvantage rivals and lead to market power) and would encourage rivals to seek refuge from efficient competition by seeking legal remedies rather than by competing in the marketplace. Because the empirical premise for the condemn-if-harmful approach seems unlikely to match reality, the approach is unsound. Moreover, even if it were true that exclusive dealing which threatens competition is in general likely to cause harm that outweighs the benefits, a general rule condemning all such exclusive dealing would still be problematic. Such a rule would be desirable only if it could be shown to be superior to a rule that sought to distinguish those cases in which the harm exceeds the benefits from those in which the benefits exceed the harm and to condemn only the former. (2) An alternative approach would be to have different rules for different kinds of exclusive dealing, based on empirical judgments about the likely benefits and harms of the particular kind of agreement. For example, exclusive dealing agreements could be prohibited if more than a certain percentage of input suppliers are constrained, and they could be permitted if their duration were less than a certain length of time. While such more refined categories could in principle reduce the likelihood of empirical error, they are still not likely to correlate closely with the welfare implications of the arrangements.7 Moreover, different rules for different categories would have other disadvantages as well. In particular, they would: lead to decisions based on

6 The latter is the rationale stated in the Discussion Paper. See ibid., para. 91 (“When competition is eliminated . . . short-term efficiency gains are outweighed by longer-term losses . . . .”). 7 For example, a safe harbor for short-term agreements would fail to prohibit a monopolist that induced exclusive dealing without any agreement at all simply by threatening not to deal with input suppliers that were not exclusive. See, e.g., Lorain Journal Co. v. United States, 342 U.S. 143 (1951).

(T) Ehlermann 08 Melamed

6/11/08

11:52

Page 437

Thoughts about Exclusive Dealing

437

formalisms; induce parties to make inefficient changes to their conduct in order to increase the likelihood that the conduct would be deemed to fall into a more lenient category; invite disputes about how to categorize conduct; and increase transaction costs, both ex ante (when businesses are trying to decide how to act) and ex post (when categorization disputes arise in litigation). At the very least, there needs to be a default rule to give guidance to businesses, agencies and courts about how to assess the lawfulness of a new form of exclusive arrangement that does not seem to fall into any of the more precise categories. (3) Yet another approach would be literally to balance the harms and the benefits from the exclusive dealing at issue in the individual case. Several commentators in the US have advocated such an approach.8 While such an approach almost tautologically makes sense as a theoretical matter, it makes little sense as a practical matter. Imagine that the Firm implements an exclusive dealing program which predictably will reduce its costs or increase its sales by a certain amount and which is likely to harm competitors of the Firm both by increasing sales of the Firm’s product and by increasing the competitors’ costs. In order to balance the harms against the benefits, an antitrust agency or court would have to quantify both. Quantifying the benefits (to the Firm and its customers, including new customers generated by the exclusive dealing) with acceptable accuracy might be feasible if only a little accuracy were required for the determination to be acceptable. If consumer welfare were the measure, quantifying the benefits would require estimating consumer surplus in the world without the exclusive dealing and the extent to which that surplus would be increased by the exclusive dealing. Increased consumer surplus would reflect (i) any outward shift in the demand for the Firm’s products because of enhanced quality of or point of sale services for those products, and (ii) any outward shift in the supply curve, attributable to the exclusive dealing, together with (iii) any changes in those factors over time. Quantifying the harms would be more difficult. If total welfare were the measure, the agency or court would have to weigh impacts rippling throughout the economy—surely an impossible task. If consumer welfare were the measure, the agency or court would have to measure the welfare loss to consumers who would have done business with the Firm’s competitors, but for the exclusive dealing, and who are harmed by the exit or increased costs of those competitors. The calculation would have to take into account, among other things, the benefits to some of those customers who shift to purchasing the Firm’s products, the possibility that perhaps unanticipated new entry or innovation spurred by the Firm’s success will offset the harm to competition,

8 See, e.g., Steven Salop, “Exclusionary Conduct, Effect on Consumers, and the Flawed Profit-Sacrifice Standard”, 73 Antitrust Law Journal 311 (2006).

(T) Ehlermann 08 Melamed

438

6/11/08

11:52

Page 438

Session Two: A Reformed Approach to Exclusionary Conduct

and the likely duration of the harm (which presumably must be discounted to present value). Balancing thus seems like an impossible task. In fact, although US courts often pay lip service to balancing the harms and benefits resulting from exclusionary conduct, they rarely if ever actually try to balance them.9 Balancing is at best an invitation to make arbitrary decisions based on guesses, imperfect information, or bias.10 Even if balancing could be done by agencies or courts with acceptable objectivity and accuracy, a legal rule based on balancing would still be unwise. The impact of antitrust rules on the millions of decisions that businesses make every day and that never become the subject of investigation or litigation is far greater than, and far more important than, the impact of antitrust rules on cases that are the subject of government investigation or litigation. While it is important that antitrust cases be fairly and correctly decided, it is even more important for antitrust rules to give useful and sound guidance to the business community. Antitrust law ought therefore to strive for rules that minimize the sum of: (i) the costs businesses must incur to determine what the law requires; (ii) the costs of beneficial conduct that the rules deter; and (iii) the costs of the harmful conduct that the rules do not deter. Balancing tests plainly do not minimize such costs. If a business in real time has to inquire whether a contemplated efficiency-generating exclusive dealing arrangement will create more harm than benefit before implementing the arrangement, it will: (i) have to spend substantial resources to investigate that question; (ii) likely need to postpone implementing the arrangement; (iii) avoid the arrangement regardless of its efficiencies because of an unwillingness to bear the legal risk or an inaccurate estimate of costs and benefits; and/or (iv) ignore the law and hope that the arrangement does not become the subject of an investigation or lawsuit. None of these is a desirable outcome. 9 In the Microsoft case, for example, the court appeared to adopt a balancing test: “[I]f the monopolist’s procompetitive justification stands unrebutted, then the plaintiff must demonstrate that the anticompetitive harm of the conduct outweighs the procompetitive benefit.” United States v. Microsoft Corp., 253 F.3d 34, 59 (D.C. Cir. 2001) (per curiam). But the court did not engage in any real analysis of that step, and it is not clear what the court meant by that language. The court found that almost all of the defendant’s allegedly anticompetitive conduct had no justification at all and concluded that the conduct was illegal without any balancing. It simply asserted, without explanation, that the one or two aspects of the challenged conduct that it found had some pro-competitive effect were for that reason lawful. Thus, notwithstanding its use of the rhetoric of balancing, the court might in substance have meant simply that conduct will be found to be anticompetitive only if it has no pro-competitive benefits at all. 10 Proponents of balancing sometimes assert that balancing is simply the familiar “rule of reason” test that has long been used to assess joint ventures and other horizontal restraints. In fact, however, the rule of reason test in such cases is more like the “no economic sense” test discussed below than the balancing test. The rule of reason in horizontal cases asks whether the arrangement is likely to increase or reduce the output of the parties to the agreement. By contrast, balancing in cases of exclusionary conduct requires a balancing of the impact of the arrangement on the parties to the arrangement against the impact on third parties—competitors and their customers. That is a far more daunting task.

(T) Ehlermann 08 Melamed

6/11/08

11:52

Page 439

Thoughts about Exclusive Dealing

439

The Beauty of Looking Inward Decisions in US monopolization cases frequently state that a defendant is entitled to the fruits of its success or its dominance if that dominance was obtained by “superior product [or] business acumen”11 or by “skill, foresight and industry”.12 The implication is the mirror image of the condemn-ifharmful rule discussed above. The implied rule would permit the use of exclusive dealing arrangements, regardless of their effect on competition, if under the circumstances they constitute a form of competition on the merits—which presumably refers to competition on the basis of efficiency. A rule that focuses on the conduct of the Firm has two principal virtues. First, it is much easier to apply, especially by businesses in real time, than a balancing test. Instead of trying to analyze the benefits of the exclusive dealing arrangement, conditions in the market, the ability of rivals to respond, the likelihood of innovation and entry, and the likely time dimensions of all the above, the Firm—or an antitrust agency or court—needs to analyze only the efficiency of the exclusive dealing arrangement for the Firm itself. Second, a rule that focuses on the conduct of the Firm reflects an important normative principle rooted in the liberal tradition in the US. It is that businesses should be free to conduct themselves as they see fit unless they engage in damnable conduct. And whether conduct is damnable depends on the nature of the conduct, not on its competitive consequences. Competition on the merits refers to competition on the basis of efficiency. But it need not and should not be a flabby concept that permits any conduct that has any efficiency properties. Such a rule would permit the Firm to use arrangements which have some efficiency benefits but which, on balance, would not make sense for the Firm absent their tendency to exclude rivals from the market and create or maintain market power for the Firm. So, competition on the merits should mean conduct that makes business sense for the Firm—because it is profitable for the Firm, taking into account opportunity costs—regardless whether it creates or maintains market power for the Firm. This is sometimes called the “No Economic Sense” (“NES”) test. As a general matter, the NES test ensures that equally efficient competitors will not be excluded and condemns only conduct that reduces welfare in a static sense. If applied sensibly, it presents little risk of false positives.

11 12

United States v. Grinnell Corp., 384 U.S. 563, 570–71 (1966). United States v. Aluminum Co. of America, 148 F.2d 416, 430 (2d Cir. 1945).

(T) Ehlermann 08 Melamed

440

6/11/08

11:52

Page 440

Session Two: A Reformed Approach to Exclusionary Conduct

The “No Economic Sense” Test Applied to Exclusive Dealing The principle that competition on the merits is lawful means that exclusive dealing is permissible, regardless of whether it injures competition, if it would have been profitable for the Firm even if it had not injured competition, after taking into account available alternatives such as non-exclusive or less exclusive arrangements with the input suppliers or exclusive arrangements with fewer input suppliers. Its application thus requires a comparison of the benefits derived by the Firm from the exclusive dealing—increased sales, reduced costs, more predictable input supplies, or whatever—against the cost to the Firm of that arrangement. The basic cost to the Firm of an exclusive dealing arrangement is a result of the following simple fact: exclusive dealing is costly to the input supplier because it has to forego dealing with the Firm’s competitors. Those costs borne by the input supplier could lead to a downward shift in its demand for dealing with the Firm, or it could induce the input supplier to insist on explicit compensation for the exclusive dealing. Either way, those costs should be internalized to the Firm, which will presumably take them into account in determining whether to induce its input suppliers to deal exclusively with it. For example, if the Firm were to give a retailer a choice of dealing exclusively or non-exclusively, the retailer will presumably require a lower price for dealing exclusively.13 That price differential is the cost to the Firm. The Firm presumably compares that cost with the benefits from exclusive dealing in deciding whether to choose an exclusive dealing arrangement. In applying the NES test, that comparison is undertaken without regard to any benefits to the Firm from the possibility that the exclusive dealing arrangement will create or maintain market power for the Firm. In other words, the NES test begins with the recognition that an exclusive dealing arrangement imposes an opportunity cost on input suppliers and asks whether that cost is fully defrayed by the efficiencies generated by the arrangement.14 In the case of exclusive dealing and other types of exclusionary vertical agreements, the costs incurred by the Firm should be a rough proxy for the sum of the costs incurred by the input suppliers and competitors of the Firm as a result of the exclusive dealing. If the exclusive dealing harms or threatens to harm the Firm’s competitors, they will presumably be willing to pay to some or all of the input suppliers compensation up to the amount of the harm to induce them to reject the exclusive dealing. The opportunity cost to the 13

This of course is the premise of loyalty discount and bundling cases. Another way of putting the question is to ask whether the Firm is able to induce the input supplier to agree to the exclusive dealing only by sharing with it the fruits of the market power created or preserved by the exclusive dealing. 14

(T) Ehlermann 08 Melamed

6/11/08

11:52

Page 441

Thoughts about Exclusive Dealing

441

input suppliers of exclusive dealing will reflect these potential defensive payments from the Firm’s competitors and will thus reflect the harm to those competitors from the exclusive dealing. And that opportunity cost should be included in the costs of exclusive dealing that are ultimately borne by the Firm. If the benefits, calculated without regard to the prospect of excluding rivals and gaining or preserving market power, exceed those costs, the exclusive dealing would be efficient, and it would be permissible under the NES test. If the costs exceed those benefits, and if the exclusive dealing excludes or raises the costs of rivals and thus creates or is likely to create market power for the Firm, then the exclusive dealing will be unlawful. In that event, the exclusive dealing would be inefficient; it would not make sense for the Firm but for the likelihood that it will injure competition and create or maintain market power for the Firm. Care must be taken not to confuse the legitimate benefits of an arrangement with its anticompetitive consequences. For example, suppose the Firm implements an exclusive dealing arrangement that entails large up-front capital costs (perhaps for a distribution facility) that can be defrayed only if the Firm’s sales are large enough in effect to dominate the market. It might be tempting to condemn the arrangement on the ground that the investment made no sense but for the market-dominating success. But the NES test calls for a more refined analysis. It asks whether the defendant’s product could generate enough sales to make the investment profitable at competitive prices that do not reflect the demise of rivals. If so, the profitability of the investment would not depend on any price increase enabled by an increase in the Firm’s market power; it would instead reflect the efficiency of the arrangement, and the arrangement would pass muster under the NES test.15

The Myth of Costless Exclusive Dealing A recent article argues that the NES test makes no sense for exclusive dealing.16 Although the article is mistaken in many respects and is based in part on a misunderstanding of the NES test,17 it does adumbrate an interesting 15 See A. Douglas Melamed, “Exclusive Dealing Agreements and Other Forms of Exclusionary Conduct—Are There Unifying Principles?”, 73 Antitrust Law Journal 375, 395–96 (2006). 16 Jonathan M. Jacobson and Scott A. Sher, “The No Economic Sense Test Makes No Sense for Exclusive Dealing”, 73 Antitrust Law Journal 779 (2006). 17 Among other things, the article asserts that the NES test would condemn exclusive dealing that enables the Firm to take sales from its competitors. But the NES test condemns exclusive dealing not because it enables the Firm to take sales from rivals (which is of course often a principal goal of lawful competition), but when its profitability for the Firm depends on creating or maintaining market power for the Firm.

(T) Ehlermann 08 Melamed

442

6/11/08

11:52

Page 442

Session Two: A Reformed Approach to Exclusionary Conduct

argument to the effect that exclusive dealing is often costless to a dominant firm. The premise of the argument is that the dominant Firm can give input suppliers an all-or-nothing choice between dealing exclusively with the Firm or not dealing with it at all and that the input suppliers will choose exclusive dealing even if the price charged by the firm is not reduced (or, in a scenario where the Firm is buying inputs, even if the price paid by the Firm is not increased). The argument is not fleshed out in the article, but the logic is presumably as follows. The input suppliers obtain surplus from their dealings with the Firm. The all-or-nothing offer enables the Firm to capture some of the surplus (in the form of benefits from exclusive dealing) that the input suppliers would otherwise realize from trading non-exclusively with the Firm. The input suppliers will nevertheless accept the exclusive deal because the surplus still available to them exceeds that available from not dealing with the Firm. Thus, the argument goes, the NES test will fail to condemn exclusive dealing even where the overall costs exceed the benefits because the costs will not be borne by the Firm and will thus be neglected when the NES test is applied. This argument is factually overstated and reflects a misunderstanding of the NES test. Exclusive dealing is costless to the Firm only if all of the following conditions are met: (i) the Firm is already dominant in the market in which it does business with the input supplier and is thus uniquely able to generate surplus for the input supplier; (ii) the Firm makes an all-or-nothing offer to only the inframarginal input suppliers who have no realistic option of rejecting it; and (iii) the Firm has no other way of discriminating among the input suppliers. Condition (i) should be self-explanatory. Condition (ii) probably requires explanation. There is presumably a downward sloping demand curve of input suppliers demanding an opportunity to deal with the Firm.18 An all-ornothing offer would likely induce the marginal input suppliers—located near where the demand and supply curves intersect—to reject the offer because the offer reduces the net value of dealing with the Firm.19 Also, the imposition of an exclusive dealing requirement should itself increase the number of marginal input suppliers and/or the likelihood that marginal input suppliers will reject the all-or-nothing offer because competitors of the Firm will value those suppliers more and will be willing to pay more to induce those suppliers to deal with them to the extent that exclusive dealing denies the competitors access to other (inframarginal) input suppliers. Decisions by marginal 18 If not—for example, if there is only one input supplier—the Firm and the input supplier would be a bilateral monopoly. In that event, the cost to the Firm of securing the exclusive dealing of the input supplier would in all likelihood reflect the opportunity cost to the input supplier of not being able to deal with third parties. 19 Alternatively, individual input suppliers might have downward sloping demand curves and, if permitted by the Firm, might for example accept the all-or-nothing offer for some but not all of their facilities.

(T) Ehlermann 08 Melamed

6/11/08

11:52

Page 443

Thoughts about Exclusive Dealing

443

input suppliers not to deal with the Firm would impose costs on the Firm. Moreover, to the extent that marginal input suppliers reject the all-or-nothing offer, they will be entirely available to competitors of the Firm, and the exclusive dealing arrangements will be less likely to injure competition. Condition (iii) is critical. The no-cost argument assumes that the input suppliers are in effect realizing consumer surplus from dealing with the Firm and that they will accept the all-or-nothing offer to avoid losing the entire surplus. But dominant firms rarely if ever deal with important input suppliers by selling (or buying) off posted price lists, and they are presumably able to discriminate among input suppliers on the basis of price and other factors. (Indeed, as noted, the premise of the costless exclusive dealing argument is that the Firm can discriminate among the input suppliers by offering an all-or-nothing offer of exclusive dealing to some but not all of them.) If the Firm could engage in such price (or other) discrimination, it would be able to capture the surplus from the input suppliers without exclusive dealing.20 In that event, the Firm would have no additional leverage with which costlessly to force the input supplier to agree to such an arrangement. And if the Firm could discriminate on price but chose to induce exclusive dealing instead, then the foregone gains from price discrimination would be the cost to the Firm of the exclusive dealing. In short, exclusive dealing is likely to be costless to the Firm in only very rare situations. As a general matter, exclusive dealing imposes costs on the input supplier; it will not change the bargaining relationship between the Firm and the input supplier; and it will thus impose costs on the Firm. Moreover, the criticism of the NES test would be misplaced even if exclusive dealing could be made costless to the Firm by means of all-or-nothing offers. The criticism is based on an understanding of the NES test that is too wooden and mechanical. The NES test is not just or even primarily an arithmetic formula. It begins with the question of whether the defendant’s conduct made sense without regard to the exclusion of rivals and the creation of market power. If the defendant does not have a factually based, legitimate explanation for the conduct that excludes rivals and creates market power, the conduct will fail the NES test without any need to calculate benefits and costs. Calculations come into the picture only when the defendant can point to legitimate benefits but there is reason to suspect that those benefits do not justify the costs. In the case of exclusive dealing obtained through the imposition by the Firm of an all-or-nothing offer, the first question would be whether there is a good explanation, not just for exclusive dealing, but for the all-or-nothing offer. If an input supplier rejected the offer and the Firm refused to deal with it, the question would be whether there is a good reason for the refusal to deal. 20 The price discrimination could take the form of differential fixed fees if the Firm wanted to capture surplus without affecting the input suppliers’ variable costs.

(T) Ehlermann 08 Melamed

444

6/11/08

11:52

Page 444

Session Two: A Reformed Approach to Exclusionary Conduct

If the input supplier rejected the all-or-nothing offer and countered with an offer to pay more to the Firm in order to deal with it non-exclusively, and if the Firm then rejected the counteroffer, the question would be whether there is a legitimate reason for the Firm to pass up the increased payments; at the very least, those foregone payments would be an opportunity cost incurred by the Firm. In other words, proper application of the NES test focuses on opportunity costs and inquires whether there is any aspect of the Firm’s conduct that both excludes rivals and cannot be explained as furthering legitimate purposes. The point is not that the NES test will never permit exclusive dealing that might reduce welfare overall. There might, for example, be instances in which the Firm will be able to justify an all-or-nothing policy. But the costs of such cases are likely to be far less than the transaction and error costs of other rules that might be used to determine the lawfulness of exclusive dealing.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 445

IV Calvin S. Goldman, Navin Joneja and Elisabeth E.C. Yuh* A Canadian Perspective on Tied Selling and Exclusive Dealing June 2007

I. Introduction Canadian competition law differs somewhat from the European and American frameworks in that its statutory regime expressly sets out the circumstances in which tied selling and exclusive dealing practices are deemed impermissible. Nevertheless, these provisions have fuelled vigorous discussion among practitioners, policymakers, economists and enforcement agencies about the appropriate applicable standards. There remains no consensus about their interpretation or application. This contribution provides a general overview of Canadian provisions related to tied selling and exclusive dealing, and of relevant case law. It also explores some of the efficiency considerations for tied selling, an issue recently discussed in greater detail at a symposium hosted by the Competition Bureau. In addition, it examines the current state of the law in Canada regarding the test for determining the existence of impermissible exclusive dealing, a topic re-invigorated by the recent Federal Court of Appeal decision in Canada Pipe.1

II. Canada’s Institutional Framework A. The Competition Act Canadian competition law is embodied in the Competition Act,2 adopted in 1986.3 The Competition Act comprises both criminal and non-criminal (also * Blake, Cassels & Graydon LLP, Toronto. 1 Commissioner of Competition v. Canada Pipe Ltd., 2006 FCA 233, [2007] 2 F.C.R. 3. 2 R.S.C. 1985, c. C-34. 3 There is a surviving body of common law causes of action relevant in the competition law context, such as intentional interference with economic interest and restraint of trade. Exclusive dealing can occur in the contractual context where a contract contains a restrictive covenant of

(U) Ehlermann 08 Goldman

446

6/11/08

11:53

Page 446

Session Two: A Reformed Approach to Exclusionary Conduct

termed “civil”) provisions. Damages can only be awarded for violations of the criminal provisions.4 Where conduct contravenes the civil provisions, the principal remedy is injunctive relief. Competition-related laws have a long history in Canada, dating as far back as 1889. The purely criminal law approach set out in the 1910 Combines Investigation Act5 did not work well, as the enforcement authorities could not prevail even in the most serious cases. This led to calls for reform of the system. The approach that prevailed in Canada prior to 1976 can be called the “bifurcated judicial model”. It involved the Director of Investigation and Research6 undertaking investigative functions and, where appropriate, recommending formal enforcement action to the federal Attorney General under the criminal provisions of the Combines Investigation Act. Reviewable practices were subject to civil review initiated by the Director of Investigation and Research and heard before a specialized adjudicative body called the Restrictive Trade Practices Commission. Canada’s Competition Act in its present form was largely created in 1986. These 1986 amendments made a major break from the previous competitionrelated legislation, which focused on combines primarily through a criminal sanctions approach. The 1986 legislation, while retaining the existing criminal offences, substantially overhauled the merger and monopolization provisions by designating them as civilly reviewable matters and by making effective enforcement possible through a more flexible process before the Competition Tribunal. The Supreme Court of Canada approved of this new approach, concluding that the federal Parliament could enact civil competition laws under its constitutional power to regulate trade and commerce.7 The 1986 amendments also introduced an explicit purpose clause which, as indicated by the Federal Court of Appeal in Canada Pipe,8 should guide the trade. Provincial courts have traditionally examined the reasonableness of such clauses in light of the temporal and geographical limitations as well as the scope of the clause. See Nordenfelt v. Maxim Nordenfelt Guns & Ammunition Co., [1894] A.C. 535 (U.K.H.L.). Claims brought under the Competition Act and claims under the common law are not mutually exclusive. In B-Filer Inc. v. The Bank of Nova Scotia, 2005 Comp. Trib. 31, Madam Justice Simpson held that matters under Part VIII of the Competition Act fell within the exclusive jurisdiction of the Competition Tribunal. Consequently, a superior court of a province did not have the jurisdiction to entertain any application brought on the basis of breaches of the Competition Act. Likewise, the Competition Tribunal could not review any application based on the grounds of breach of contract or common law causes of action. However, a decision by a superior court of a province did not foreclose any proceedings before the Tribunal, and vice versa. 4 Section 36 of the Competition Act allows a private party who has suffered loss as a result of conduct that contravenes a criminal competition offence (or as a result of contempt of an order issued by the Competition Tribunal or by another court under the Competition Act) to recover damages. 5 S.C. 1910, c. 9. 6 A position since renamed “Commissioner of Competition”. 7 General Motors of Canada Ltd. v. City National Leasing Ltd., [1989] 1 S.C.R. 641. 8 Canada Pipe, supra note 1, at para. 48.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 447

A Canadian Perspective on Tied Selling and Exclusive Dealing 447 interpretation of individual statutory provisions, including those relating to tied selling and exclusive dealing: “1.1 The purpose of this Act is to maintain and encourage competition in Canada in order to promote the efficiency and adaptability of the Canadian economy, in order to expand opportunities for Canadian participation in world markets while at the same time recognizing the role of foreign competition in Canada, in order to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy and in order to provide consumers with competitive prices and product choices.” 9

The provisions governing tied selling and exclusive dealing, both of which are civilly reviewable matters, are now contained in Part VIII of the Competition Act, entitled “Matters Reviewable by the Tribunal”, under the sub-heading “Restrictive Trade Practices”. This Part also contains provisions on the abuse of dominance, refusal to deal, consignment selling, and delivered pricing laws, as well as the substantive merger review provisions under the Competition Act. Tied selling and exclusive dealing (along with a provision governing certain market restrictions) are specifically dealt with under section 77 of the Act. Historically, the courts and commentators viewed such vertical contractual constraints in conjunction with other exclusionary conduct connected with monopolistic behaviour. However, more recent views suggest that these practices—in the absence of market dominance—may actually enhance efficiency and, as a consequence, competition authorities should consider adopting a restrained approach towards their enforcement.10 This policy debate shapes the interpretation of section 77, and helps explain why tied selling and exclusive dealing were included as civil reviewable matters rather than as criminal offences.11 As discussed below, they stand independently in the Canadian legislation, and without an explicit requirement to show market dominance. It need only be shown that the conduct is engaged in by a major supplier and that it is having an effect of a substantial lessening of competition. In theory, this would appear to give the provisions greater potential reach, but in practice the Canadian Competition Bureau does not pursue cases under these provisions very often. The focus has primarily been on abuse of dominance; tied selling and exclusive dealing are usually pursued in parallel with such dominance allegations (as discussed in Section V. below).

9

Competition Act, supra note 2. William S. Comanor, “Vertical Market Restrictions, and the New Antitrust Policy”, 98 Harvard Law Review 983 (1985). 11 Calvin S. Goldman and John D. Bodrug, eds., Competition Law of Canada, Juris Publishing, 2004, at § 5.06[2]. 10

(U) Ehlermann 08 Goldman

448

6/11/08

11:53

Page 448

Session Two: A Reformed Approach to Exclusionary Conduct

B. The Competition Tribunal Decisions regarding conduct that allegedly violates the Competition Act are made by the Competition Tribunal, a statutorily created specialized court. The Competition Tribunal’s exclusive jurisdiction covers civil provisions including Part VIII of the Competition Act, which includes tied selling and exclusive dealing as well as mergers, abuse of dominance and other vertical matters.12 The Competition Tribunal was established in 1986 and replaced the Restrictive Trade Practices Commission. Parliament’s intention was that the Tribunal would act as a specialized adjudicative body uniquely suited to deal with non-criminal competition matters.13 Members of the Tribunal are comprised of both judges from the Federal Court and lay people. Judicial members preside over hearings, with panels composed of three to five members with at least one lay member. All appeals from the Tribunal are to the Federal Court of Appeal.14 The Competition Tribunal did not inherit its predecessor’s extensive investigative powers, which included the ability to authorize searches.15 The Competition Tribunal exercises those powers required for “the attendance, swearing and examination of witnesses, production and inspection of documents” like a superior court of record.16 It may also make findings of fact, issue remedial orders and approve the terms of negotiated consent agreements.17

C. Private Action Rights for Tied Selling and Exclusive Dealing In 2002, the Competition Act was amended to allow for a private right of action for claims of tied selling and exclusive dealing through the enactment of section 103.1. Accordingly, at present a claim for tied selling may be

12 Applications under Part VII.I are also subject to the exclusive jurisdiction of the Tribunal, as are references brought under s. 124.2(2) of the Competition Act. 13 Criminal offences under the Competition Act are prosecuted before superior courts. The Attorney General of Canada retains the exclusive jurisdiction to prosecute. The Commissioner’s enforcement mandate is limited to referring such a matter to the Attorney General. 14 Competition Tribunal Act, R.S.C. 1985, c. 19 at s. 13. 15 Combines Investigation Act, R.S.C. 1970, c. C-23 at s. 10(3). This provision was later deemed to be constitutionally invalid by the Supreme Court of Canada in Hunter v. Southam, [1984] 2 S.C.R. 145. 16 Competition Tribunal Act, supra note 14 at s. 8(2). Production orders and search warrants sought by the Commissioner must also be obtained from a judge of a superior or county court. See Competition Act, supra note 2 at s. 11(1). 17 Competition Act, supra note 2 at s. 105.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 449

A Canadian Perspective on Tied Selling and Exclusive Dealing 449 brought before the Tribunal either by the Commissioner or by an aggrieved private party.18 This represents a significant step for Canadian competition law. Prior to these amendments, private rights of action, as embodied in section 36 of the Competition Act, were available only for conduct contrary to the Act’s criminal provisions, for example horizontal agreements between competitors. For conduct that contravenes other provisions of the Act, the only recourse left available to an aggrieved party was to press the Commissioner to undertake action or to rally with five other complainants and file an application to commence an inquiry with the Commissioner.19 The introduction of a Canadian private right of action has not unleashed a floodgate of claims. This may be in part because section 103.1 carefully circumscribes the ability of a private party to commence an action for tied selling or exclusive dealing. A private party cannot pursue an application to the Tribunal concurrently with the Commissioner for the same matter under the tied selling or exclusive dealing provisions.20 In addition, the Commissioner may not bring an application for an order under the tied selling, exclusive dealing or abuse of dominance provisions based on substantially similar facts. Moreover, leave to make an application must first be obtained. The Tribunal has the discretion to grant leave where “it has reason to believe that the applicant is directly and substantially affected in the applicant’s business by any practice referred to in one of those sections that could be subject to an order under that section”.21 The applicant seeking leave needs to provide sufficient credible evidence of what is alleged to give rise to a bona fide belief by the Tribunal. This is a lower standard of proof than proof on a balance of probabilities, which is the standard applicable to the later decision on the merits.22 In considering an application for leave, the Tribunal cannot draw any inference from the fact the Commissioner has or has not taken any

18 A private right of action was also made available for claims of refusal to deal under section 75 of the Competition Act but not for the more general abuse of dominance provision. 19 Competition Act, supra note 2 at s. 9(1). 20 However, in this case, a private litigant may seek leave from the Competition Tribunal to intervene. The decision to grant leave falls within the Tribunal’s discretion. However, where leave is granted, the private litigant must satisfy the four elements of the test set out in Canada (Commissioner of Competition) v. United Grain Growers Ltd, 2002 Comp. Trib. 35 at para. 17. Accordingly: (1) the matter alleged to affect that person seeking leave to intervene must be legitimately within the scope of the Tribunal’s consideration or must be a matter sufficiently relevant to the Tribunal’s mandate; (2) the person seeking leave to intervene must be directly affected; (3) all representations made by a person seeking leave to intervene must be relevant to an issue specifically raised by the Commissioner; and (4) the person seeking leave to intervene must bring to the Tribunal a unique or distinct perspective that will assist the Tribunal in deciding the issues before it. 21 Ibid. at s. 103.1(7) (emphasis added). 22 Construx Engineering Corporation v. General Motors of Canada, 2005 Comp. Trib. 21 at para. 7.

(U) Ehlermann 08 Goldman

450

6/11/08

11:53

Page 450

Session Two: A Reformed Approach to Exclusionary Conduct

action.23 If leave is granted, the same fact will also have no bearing in considering the substance of an application.24 Initiating a private action allows the private party to have greater control over the process. However, as of this writing—five years after the private rights of action were established—there have been very few cases brought in Canada under the tied selling and exclusive dealing provisions. Part of the reason for the dearth of case law may be a result of the inherent challenges of marshalling evidence in advance demonstrating that the impugned behaviour had anticompetitive effects within the meaning of these provisions. Jochen Burrichter notes that this is a challenge faced by competition authorities in Europe as well. This challenge is magnified for private litigants who, unlike the competition authorities, do not benefit from the exercise of investigative powers.25 Additionally, a private litigant may not have access to the material gathered by the Commissioner pursuant to the exercise of her formal powers.26 Furthermore, an unsuccessful private litigant may be subject to paying substantial costs,27 without the possibility of seeking damages.28 The only remedy available to a private claimant is injunctive relief. As a result, there is little jurisprudence to date that would provide more clarity and certainty regarding these practices. This is somewhat unfortunate since a private party may be in the best position to detect a reviewable practice, particularly if their stake in the matter is substantial. More recently, the OECD has recommended that Canada ought to further expand the right of private parties to pursue competition law enforcement. This was after the OECD examined the 2002 amendments to the Act that expanded the right of private access and found them to be a step in the right direction, but inadequate. The OECD report noted that “permitting private parties to obtain conduct orders and damages for violations of the Act’s civil provisions (other than those relating to mergers) would be the most effective means of supplementing governmental enforcement and deterring anticompetitive conduct”.29 The Competition Bureau’s need to allocate its limited resources to other enforcement priorities is likely another reason for the lack of case law on this subject. In addition, unlike the abuse of dominance provisions, there are no 23

Competition Act, supra note 2 at s. 103.1(11). Ibid. at s. 77(7). 25 See Jochen Burrichter, A Reformed Approach to Article 82: The Impact on Private Enforcement, in this Volume. 26 Notably, pursuant to sections 11, 15 and 16 of the Competition Act, information obtained as a result of the exercise of the Commissioner’s formal powers may be privileged, in which case it could not be disclosed to a private litigant. 27 Competition Tribunal Act, supra note 14 at s. 8.1(1). 28 Competition Act, supra note 2 at s. 77(3.1). 29 Organization for Economic Cooperation and Development, “Canada—Report on Competition Law and Institutions (2004)”, 18 January 2005, DAF/COMP(2005)4 at 32. 24

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 451

A Canadian Perspective on Tied Selling and Exclusive Dealing 451 agency guidelines in Canada with respect to the tied selling and exclusive dealing provisions. While the 2002 amendments to the Act expanding the right of private access to situations involving a violation of section 77 are useful, more can be done in Canada to provide clarity and guidance.

II. Elements of the Tied Selling and Exclusive Dealing Provisions Under the Competition Act, tied selling and exclusive dealing are defined as distinct reviewable practices. Section 77 sets out in detail the requisite elements of tied selling and exclusive dealing. In contrast, the American definitions of tied selling and exclusive dealing arise from judicial development under the Sherman Act and the Clayton Act.30 Notwithstanding this distinguishing feature, the concepts used to analyze the law of tied selling and exclusive dealing have evolved to become fairly similar, and indeed the Competition Tribunal has relied upon American case law in certain circumstances. Given their similar legislative evolution, the tied selling and exclusive dealing practices share common elements. Treatment by the Competition Tribunal of one provision largely has been interpreted as applicable to the other. We examine below the respective definitions of tied selling and exclusive dealing and their common elements.

A. Definition of Tied Selling Under section 77(1), “tied selling” is defined as: “(a) any practice whereby a supplier of a product, as a condition of supplying the product (the “tying” product) to a customer, requires that customer to (i) acquire any other product from the supplier or the supplier’s nominee, or (ii) refrain from using or distributing, in conjunction with the tying product, another product that is not of a brand or manufacture designated by the supplier or the nominee, and

30 Sherman Act, 15 U.S.C. §§ 1–7 at §§ 1 and 2 and Clayton Act, 15 U.S.C.A. §§ 12-27 at § 3. Although rare, tied selling actions can also be brought under section 5 of the Federal Trade Commission Act. As the Sherman Act expresses a general prohibition against any unreasonable restraint in trade, the US Congress enacted the Clayton Act with the express objective of capturing anticompetitive tying arrangements. See Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2 (1984) at footnote 21.

(U) Ehlermann 08 Goldman

452

6/11/08

11:53

Page 452

Session Two: A Reformed Approach to Exclusionary Conduct

(b) any practice whereby a supplier of a product induces a customer to meet a condition set out in subparagraph (a)(i) or (ii) by offering to supply the tying product to the customer on more favourable terms or conditions if the customer agrees to meet the condition set out in either of those subparagraphs.”

To substantiate a claim of tied selling, it is necessary to establish: (1) the existence of “two separate products”; (2) the existence of a practice of tying; (3) that the tying is engaged by a major supplier or that it is widespread in the market; and (4) anticompetitive effects leading to a substantial lessening of competition. The first two elements, which are specific to tied selling, are discussed below.

1. Separate Products By definition, tied selling requires two distinct products, one tied to the other.31 Though superficially simple, the question of whether there are two separate products occupied a majority of the Tribunal’s tied selling analysis in the Tele-Direct case.32 In this case, the Director of the Competition Bureau alleged that the respondents, Tele-Direct (Publications) Inc. and Tele-Direct (Services) Inc., engaged in tied selling by, as a condition of supplying advertising space in telephone directories, inducing customers to acquire another product from the respondents, namely advertising/marketing services. The Tribunal adopted the test set out by the US Supreme Court in Jefferson Parish Hospital District No. 2 v. Hyde to determine whether there were two separate products. According to the Supreme Court, “no tying arrangement can exist unless there is a sufficient demand for the purchase of [the tied product] separate from [the tying product] to identify a distinct product market in which it is efficient to offer [the tied product] separate from [the tying product]”.33

31 Canada (Director of Investigation and Research) v. Tele-Direct (Publications) Inc. (1997), 73 C.P.R. (3d) 1 (Comp. Trib.), at 118–165. 32 Ibid. The Tele-Direct case concerned an application brought by the Director under both the abuse of dominance and tied selling provisions, seeking an order from the Competition Tribunal against Tele-Direct (Publications) Inc. and Tele-Direct (Services) Inc., affiliated companies. The Director alleged that respondents illegally bundled advertising space (i.e. the printing, publishing, and distribution of a directory) with their advertising services (i.e. the provision of advice and support with respect to design, placement, and content of ads) in Yellow Pages directories. The Tribunal undertook the two-stage inquiry described above to determine whether services and advertising space formed distinct products over any segment of the market. In particular, the Tribunal considered evidence from advertiser witnesses, economic and practical efficiencies arguments, and profitability studies that were accepted as having a weak connection to cost studies. In the end, the Tribunal found that there were in fact two different products over a segment of the market that included more than just the commissionable accounts. The other conditions having been met, the Tribunal made a finding of impermissible tied selling which substantially lessened competition in the advertising services market. 33 466 U.S. 2 (1984) at 21–22, endorsed in Tele-Direct, supra note 31 at 119.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 453

A Canadian Perspective on Tied Selling and Exclusive Dealing 453 This two-stage inquiry asks: (i) whether there is sufficient demand from buyers to acquire the allegedly tied products separately from each other; and if so, (ii) whether separating the products would result in their efficient supply. In adopting this two-pronged approach based on demand and efficiency, the Tribunal rejected the use of other tests which could protect unjustifiable tied selling. For instance, the functional dependency test, which would regard as a single product (or allow the tying of) any two goods which were heavy complements or useless without each other, has been explicitly rejected.34 Whether there is separate demand for products is determined by evidence from actual purchasers about their preferences, along with explanations as to why they hold those preferences.35 The actual behaviour of purchasers (for example whether they actually do purchase the products separately where possible) is also relevant.36

2. Tying Condition The definition of tied selling also requires the existence of a tying condition. The tying condition need not entirely preclude the possibility of acquiring the tied product elsewhere. It is sufficient that it establishes a form of coercion. In addition, there is no need for the condition to be embodied in an express clause inducing or requiring a tied sale. The coercion may arise as a result of a combination of benefits and/or economic penalties which have the effect of limiting the effective choice of supplier.37

34 Tele-Direct, supra note 31, at 120–121. However, a certain lesser degree of complementarity between the products will not preclude a finding of two separate products. 35 Ibid. at 136–137. 36 Ibid. at 139. 37 Canada (Director of Investigation and Research) v. NutraSweet Co. (1990), 32 C.P.R. (3d) 1 (Comp. Trib.) [NutraSweet] at 54; Tele-Direct, supra note 31 at 122–124 and 172–173. In TeleDirect, the respondents pointed to the absence of an express contractual condition which bound clients to purchase advertising space and services as one inseparable product. Clients were not prohibited from purchasing their advertising services from a supplier other than Tele-Direct. Rather than endorsing a literal definition, the Tribunal preferred to take a purposive approach. It found that a condition within the meaning of section 77 may arise absent an express provision where there is any type of inducement, particularly economic, to force a customer to acquire two otherwise distinct products from the same supplier. Customers were effectively coerced into buying both advertising space and services as a package from Tele-Direct rather than acquiring only advertising space and obtaining advertising services from another supplier. Accordingly, the Tribunal made a finding that the definition of “tied selling” had been satisfied. The Tribunal’s discussion in relation to the nature of the required condition for tied selling would equally apply to the other reviewable conduct enunciated at section 77, such as the exclusive dealing arrangements.

(U) Ehlermann 08 Goldman

454

6/11/08

11:53

Page 454

Session Two: A Reformed Approach to Exclusionary Conduct

B. Definition of Exclusive Dealing According to the legislative history, the exclusive dealing provision was introduced in 1976 to address situations where the practice of exclusive dealing “deprives the market of products which are in demand and which would produce needed price competition in the market”.38 Under section 77(1), “exclusive dealing” means: “(a) any practice whereby a supplier of a product, as a condition of supplying the product to a customer, requires that customer to (i) deal only or primarily in products supplied by or designated by the supplier or the supplier’s nominee, or (ii) refrain from dealing in a specified class or kind of product except as supplied by the supplier or the nominee, and (b) any practice whereby a supplier of a product induces a customer to meet a condition set out in subparagraph (a)(i) or (ii) by offering to supply the product to the customer on more favourable terms or conditions if the customer agrees to meet the condition set out in either of those subparagraphs.”

Anticompetitive exclusive dealing is not limited to absolute exclusions. This provision also addresses situations where a supplier permits a customer to deal only in one or two insignificant articles obtained from sources other than the supplier,39 or where a supplier provides something less than 100 percent of the customer’s requirements.40 Accordingly, permitting a customer to deal in some other products which do not constitute a significant threat to a supplier’s competitive position will not in itself immunize the supplier from the exclusive dealing provision.41 In order to sustain a claim of impermissible exclusive dealing, it is necessary to demonstrate some element of coercion which obliges customers to deal only or primarily in the products of the supplier. The coercive element can be satisfied not only where (actual or implied) threats have been made, but also where there is any type of inducement or advantage.42

38 Standing House of Commons Committee on Finance, Trade and Economic Affairs, Minutes of Proceedings and Evidence, (3 December 1974), at 50. 39 Ibid. at 50. 40 See Mark Q. Connelly, “Exclusive Dealing and Tied Selling Under the Amended Combines Investigation Act”, 14 Osgoode Hall Law Journal 521 (1976), at 534. 41 Goldman and Bodrug, supra note 11, at §5.06[2]. 42 Competition Act, supra note 2 at ss. 77(1)(a) and (b).

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 455

A Canadian Perspective on Tied Selling and Exclusive Dealing 455

C. Other Elements of the Tied Selling and Exclusive Dealing Provisions Before the Tribunal can issue a remedial order in the case of either tied selling or exclusive dealing, the following additional elements must be satisfied: (i) the practice must be engaged in by a major supplier or must be widespread in a market; and (ii) there must be an exclusionary effect as a result of which competition is likely to be lessened substantially. These elements are discussed below.

1. “Major Supplier or Widespread in the Market” The statute does not define what threshold is required for a supplier to be considered “major” or for when a practice is deemed to be “widespread”. However, indicators of whether a supplier is a major one include its market share, its financial strength and its record as an innovator.43 Absent any evidence of countervailing power, a major supplier will also be equated with a participant that possesses market power in the market for the tying product.44 A participant may be considered a “major supplier” within the meaning of section 77 if its actions either result in a substantial lessening of competition or have an appreciable or significant impact on the market where it sells.45 The Competition Tribunal has established that a practice will be found to be widespread when virtually all customers are affected by the impugned practice.46 This “major supplier” requirement is a key distinguishing element setting the exclusive dealing and tied selling legislative provisions apart from the abuse of dominance provision. The latter requires a demonstration that the party “substantially or completely control[s] (. . .) a class or species of business”,47 which has been equated with market power.48 Market power is defined as the “ability to set prices above competitive levels for a considerable period”.49 While market share will be an important factor in assessing whether a party has market power, other factors, such as barriers to entry, will also be taken into account. 43

Canada (Director of Investigation and Research) v. Bombardier Ltd. (1980), 53 C.P.R. (2d) 47. Tele-Direct, supra note 31 at 34. 45 Bombardier, supra note 43 at 55; NutraSweet, supra note 37 at 55. 46 NutraSweet, supra note 37 at 55. 47 Competition Act, supra note 2 at s. 79(1)(a). 48 Tele-Direct, supra note 31 at 107. 49 Canada (Director of Investigation and Research) v. Laidlaw Waste Systems Ltd., (1992), 40 C.P.R. (3d) 289 (Comp. Trib.); Canada (Director of Investigation and Research) v. D&B Companies of Canada Ltd. (1996), 64 C.P.R. (3d) 216 (Comp. Trib.); Competition Bureau, Enforcement Guidelines on the Abuse of Dominance Provisions. 44

(U) Ehlermann 08 Goldman

456

6/11/08

11:53

Page 456

Session Two: A Reformed Approach to Exclusionary Conduct

2. Exclusionary Effects This element requires the demonstration that the practice at issue (has or) is likely to: (a) impede entry into or expansion of a firm in a market; (b) impede introduction of a product into or expansion of sales of a product in a market; or (c) have any other exclusionary effect in a market, as a result of which competition is, or is likely to be, lessened substantially.50 Whether the tied selling or exclusive dealing practice is likely to impede entry or expansion and result in a substantial lessening of competition is considered simultaneously. The Competition Tribunal has held that: “Whether exclusive dealing [or tied selling] by a supplier impedes expansion or entry of competitors in the market is most easily and meaningfully considered as part of the determination of whether there is or is likely to be substantial lessening of competition as a result of that practice.”51

The statutory test to determine the existence of a substantial lessening of competition is similar to the test contained in the abuse of dominance provisions of the Act; hence, the examination will involve principally the same considerations.52 According to the Tribunal, the question will be whether the practice enhances or preserves the supplier’s market power, and the degree to which such conduct creates barriers to entry or expansion in the market. The element of exclusionary effect is met if a practice of exclusive dealing or tied selling adds to or preserves market power.53 Fundamental to this assessment, as the Tribunal has indicated, is the degree to which the anticompetitive acts have created entry or expansion barriers. For example, it would be relevant if a new entrant with a smaller market share has been precluded from increasing its market share as a result of the allegedly anticompetitive acts.54 For tied selling, the anticompetitive effects are to be measured in terms of the market for the tied product, not the tying product. The Tribunal has found that, since tying involves leveraging from the tying product market to the tied product market, it is only sensible to assess the effects of the practice, that is, the substantial lessening of competition, in the target or tied product market.55

50

Competition Act, supra note 2 at s. 77(2). NutraSweet, supra note 37 at 56. But see the decision by the Federal Court of Appeal in Canada Pipe (see section V below). 52 Ibid. 53 Ibid. at 47. 54 Canada (Director of Investigation & Research) v. BBM Bureau of Measurement (1980), 60 C.P.R. (2d) 26; NutraSweet, supra note 37, at 47–48. 55 Tele-Direct, supra note 31 at 175. We note, however, that economists have postulated alternative theories of anticompetitive harm based on tied selling. See, example e.g., Michael D. Whinston, “Tying, Foreclosure and Exclusion”, 80 American Economic Review 837, 855–56 (1990). 51

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 457

A Canadian Perspective on Tied Selling and Exclusive Dealing 457 In Tele-Direct, the Tribunal used as a comparator the value of business in the market of the tied product which, “but for” the tying practice, would be available to competitors.56 In that case, since the “total of the [business] found to be tied add[ed] up to well in excess of 50 percent of the current . . . market”, the “amount of revenue affected by the tie [was] undoubtedly sufficient to conclude that there is a substantial lessening of competition”.57 The Tribunal also found a direct correlation between the degree of market power and the effect of an anticompetitive act on the market. In other words, the greater the market power, the more likely the impugned anticompetitive action is likely to have a significant anticompetitive effect and thus the more likely it is to be scrutinized by the Tribunal. “Where a firm with a high degree of market power is found to have engaged in anticompetitive conduct, smaller impacts on competition resulting from that conduct will meet the test of being “substantial” than where the market situation was less uncompetitive to begin with. In these circumstances, particularly Tele-Direct’s overwhelming market power, even a small impact on the volume of consultants’ business, of which there is some evidence, by the anti-competitive acts must be considered substantial.”58

D. Remedies and Exceptions 1. Remedies The typical remedy for an anticompetitive practice of tied selling or exclusive dealing is an order by the Tribunal directing the supplier to unbundle the tied products or prohibiting the supplier(s) from continuing to engage in the exclusive dealing. However, the Competition Tribunal is neither limited nor bound by the remedies outlined in the Commissioner’s application and has the discretion to fashion any remedy that it deems appropriate to “overcome the effects [of the practice] in the market or to restore or stimulate competition in the market”.59 In specific circumstances, interim injunctive relief is also available. On ex parte application of the Commissioner,60 the Tribunal may make an interim order where: (1) injury to competition that cannot adequately be remedied by the Tribunal is likely to occur; (2) a person is likely to be eliminated as a 56 This test was affirmed in the abuse of dominance context in Canada Pipe, supra note 1 at para. 26, leave to appeal to Supreme Court of Canada refused 31637 (11 May 2007). The Federal Court of Appeal stated that “[e]ach statutory element must give rise to a distinct legal test, for otherwise the interpretation risks rendering a portion of the statute meaningless or redundant”. 57 Tele-Direct, supra note 31, at 174. 58 Ibid. at 126. 59 Competition Act, supra note 2 at s. 77(3); NutraSweet, supra note 37 at 58. 60 Competition Act, supra note 2 at s. 103.3(1).

(U) Ehlermann 08 Goldman

458

6/11/08

11:53

Page 458

Session Two: A Reformed Approach to Exclusionary Conduct

competitor; or (3) a person is likely to suffer a significant loss of market share, revenue or other harm that cannot adequately be remedied by the Tribunal.61 Interim orders are initially effective for ten days.62 They are subject to both extension and revocation.63

2. Exceptions The statute outlines specific circumstances where tied selling and exclusive dealing are permissible, even though the requisite elements of the anticompetitive acts as described above may have been met.64 To begin with, the language of the provision specifies that these acts are illegal only where the supplier requires an exclusivity relationship or imposes tied products. It would appear that where the exclusivity or bundling request originates from the customer, the request would not run afoul of the Competition Act. a. Temporary Exclusive Dealing A party may legally engage in temporary exclusive dealing arrangements for the purpose of facilitating the entry or introduction into a market of a new supplier or product where the following conditions are met: (1) the exclusive dealing is engaged in for only a limited period of time; (2) the period of time is a reasonable one; and (3) the behaviour is engaged to facilitate the entry of a new supplier or new product in the market. The Competition Tribunal has not yet had the occasion to consider the meaning of “reasonable period of time” under this section. It appears likely, however, that the relevant time is in each case a question of fact to be considered in relation to the time needed to enter the market, taking into account prevailing industry conditions. It is presumed that a firm which has become a “major supplier” has likely exceeded that time period.65 b. Technological and IP Rights Tied selling arrangements that are “reasonable having regard to the technological relationship between or among the products to which it applies” are 61

Competition Act, supra note 2 at s. 103.3(2). Ibid. at s. 103.3(4). 63 Ibid. at ss. 103.3(5.1) and (5.3). 64 Ibid. at s. 77(4). There are two other exceptions that should be noted. Pursuant to s. 77(4)(c), tied selling is permissible in certain circumstances for the purposes of securing a loan. Tied selling and exclusive dealing arrangements are permissible among affiliates where the parties are affiliated within the definition provided for by s. 77(5). 65 On these points, see Goldman and Bodrug, supra note 11 at § 5.06[3]; Michael Trebilcock et al., The Law and Economics of Canadian Competition Policy, University of Toronto Press, 2002, at 449–50. 62

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 459

A Canadian Perspective on Tied Selling and Exclusive Dealing 459 not prohibited.66 While the “technological” exemption for tied selling arrangements has been subject to judicial interpretation, the term “reasonable” has not been clarified. In BBM, the respondent, BBM, tried to argue that there was a reasonable technological relationship between radio and television audiences because the preparation and provision of data had common “administration costs” and “process costs”. The Restrictive Trade Practices Commission held that the technological defence could only arise where there is a reasonable requirement to sell the two products together for technological reasons. This might be the case where there might be injury or damage resulting from not using the tying product in conjunction with the tied product. Any technological benefits from producing the products together were irrelevant.67

IV. Efficiency Considerations with respect to Tied Selling There is no efficiency exception or defence for tied selling expressly set out in section 77. However, from a purely business (and economic) perspective, efficiency may be a valid justification for tying two possibly distinct products. As a result, efficiency could be a factor considered at the stage of determining whether there is in fact a single tied product or two unrelated and distinct products. As tying is in fact prevalent in competitive markets, it is a fair assumption that it provides efficiencies in many circumstances. Some of these efficiencies include the elimination of some transaction costs, reduction in searching and sorting, and “variable proportion” where, for instance, it eliminates inefficient input substitution when one product is characterized by market power. There are various other legitimate reasons why firms engage in tying. For example, the aim might be to reduce production, distribution or marketing costs, or to facilitate the entry and market penetration of a new product. After several decades of economic investigation into the competitive effects of tying, there is still a debate as to whether there should be any presumption on the part of competition authorities that tying is anticompetitive.68 As discussed at a symposium held by the Competition Bureau in March of 2007, the Chicago School’s theory on tying postulates that tying has nothing to do with trying to achieve a monopoly in the tied product because the monopolist’s profits remain the same for complementary products whether or not products are tied—monopoly profits can only be extracted once. However, this position 66

Competition Act, supra note 2, at s. 77(4)(b). BBM, supra note 54, at 33–34. 68 See Christian Ahlborn, David S. Evans and A. Jorge Padilla, “The antitrust economics of tying: a farewell to per se illegality,” 49 Antitrust Bulletin 287, 329 (2004). 67

(U) Ehlermann 08 Goldman

460

6/11/08

11:53

Page 460

Session Two: A Reformed Approach to Exclusionary Conduct

does have its critics. For example, one theory argues that a firm enjoying monopoly power in the tying product might have an anticompetitive incentive to tie when the market for the tied good is imperfectly competitive if tying keeps potential rivals out of the market for the tied product or, alternatively, helps the monopolist to preserve its market power in the tying product. This “foreclosure” theory suggests that, through tied selling, a monopolist deprives its competitors in the market for the tied product of adequate scale, thereby lowering their profits below the level that would justify remaining active in (or, alternatively, entering) that market.69 Another possible issue is that, in certain circumstances, a commitment to tying through a physical tie involving incompatible products may be a means to leverage monopoly power. This may be the case when products that are not exclusively complementary are tied, or where some use of the tied product is discrete from use of the tying product. It could prevent other producers of the tied good from entering the market or, if they already participate in the market it could force them to exit. Although this situation could possibly reduce the monopolist’s short-term profits in the tying product, any such losses would be offset by increased profits in the tied product as prices rise due to a lack of competition in the market for the tied product. A further example, at least in theory, that is cited by critics involves analyzing tying in a two-stage market. This is the situation where the monopoly component is essential only for a period of time, but not thereafter. By physically tying its products from the outset, the monopolist commits to depriving its rival of sales in the earlier period, which may force the rival out of the market for the longer term. This, in turn, would reduce the monopolist’s profits in the initial period, but it may increase its profits in the second period by more than it lost in the first period. In any event, although certain economic theories have provided examples of situations where tying may be anticompetitive, they do not disturb the evolving consensus that tying is a constant feature of economic life in dynamic, competitive markets and that the primary motivations for this form of strategic behaviour are usually the realization of substantial efficiencies and increased consumer welfare. In line with this economic discussion, the Competition Bureau is currently grappling with the question of under what circumstances tied selling in the intellectual property context should be prohibited by the Competition Act, particularly where a patented product is tied to an unpatented product. There is, of course, no easy or clear-cut answer. Efficiency considerations appear to have played a large role in the views expressed at the Competition Bureau’s recent symposium. In this regard, we note that while there is no efficiency exception to tied selling pursuant to section 77, the abuse of dominance 69 Whinston, supra note 55. See also Einer Elhauge and Damien Geradin, Global Competition Law and Economics, Hart Publishing, 2007, at 501.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 461

A Canadian Perspective on Tied Selling and Exclusive Dealing 461 provision does have an explicit exception for the exercise of IP rights as well as practices resulting from “superior competitive performance”,70 and the overarching purpose clause in the Competition Act does make reference to efficiency-based considerations.71

V. The Current Test in Canada for Exclusive Dealing The two leading cases in Canada concerning the exclusive dealing provisions of the Competition Act are NutraSweet72 and Canada Pipe.73 One telling feature of these cases, and of the exclusive dealing provision generally, is that claims for exclusive dealing are discussed in both cases alongside claims for abuse of dominance outlined in section 79. There has yet to be a litigated case in Canada where the exclusive dealing provision has been considered on its own, although there is now a private right of action for exclusive dealing (as discussed above), and this right does not extend to abuse of dominance. As a result of this parallel treatment the Competition Tribunal has, to a certain degree, conflated the exclusive dealing practice with abuse of dominance.74 For example, the Tribunal has confirmed that the types of anticompetitive behaviour listed in section 78 are not exhaustive, and they arguably would encompass arrangements such as exclusive dealing.75 This approach thus recalls that of the European Court of Justice, which long ago held that Article 82 EC “merely gives examples, not an exhaustive enumeration of the sort of abuses of a dominant position prohibited by the Treaty”.76 In fact, in economic terms, the exclusive dealing provision closely resembles sections 78 and 79 of the Competition Act, which deal with abuse of dominance. For instance, section 78 lists (for the purposes of the abuse of dominance provision), as an example of an “anti-competitive act”, the “acquisition by a supplier of a customer who would otherwise be available to a competitor of the supplier . . . for the purpose of impeding or preventing the competitor’s 70

Competition Act, supra note 2 at s. 79(4). Ibid. at s. 1.1. 72 NutraSweet, supra note 37. 73 Canada Pipe, supra note 1; Canada (Commissioner of Competition) v. Canada Pipe, 2005 Comp. Trib. 3. 74 Trebilcock, supra note 65, at 448. Damien Neven suggested, in his oral remarks at this Workshop, that the methods developed in merger control to assess a transaction’s anticompetitive effects are transferable to the determination of whether behaviour is truly exclusionary in the antitrust context. Damien Neven, “A reformed approach to exclusionary conduct”, presented at the 12th Annual Competition Law and Policy Workshop, Robert Schuman Centre, EUI, Florence, 8 June 2007. Neven’s presentation is reproduced in this Volume at pages 274 et seq. 75 Trebilcock, supra note 65, at 505. 76 Case 6/72, Europemballage Corporation and Continental Can Company Inc. v Commission of the European Communities [1973] ECR 215, para. 26 71

(U) Ehlermann 08 Goldman

462

6/11/08

11:53

Page 462

Session Two: A Reformed Approach to Exclusionary Conduct

entry into, or eliminating the competitor from, a market” and “requiring or inducing a supplier to sell only or primarily to certain customers, or to refrain from selling to a competitor, with the object of preventing a competitor’s entry into, or expansion in a market”. This notion of locking up customers or suppliers, and preventing access by competitors, has been the hallmark feature of the exclusive dealing cases brought before the Competition Tribunal.77 However, as noted below, the Tribunal in Canada Pipe treated the same practice differently under the abuse of dominance and exclusive dealing provisions, finding that the practice in question did not constitute an “anticompetitive act” under the abuse provisions but that it did meet the definitional requirement of exclusive dealing under section 77.

A. Canada (Director of Investigation and Research) v. NutraSweet Co. In NutraSweet, the NutraSweet Company (“NSC”) produced and sold aspartame worldwide to various customers who manufactured foods and beverages using aspartame. In Canada, NSC supplied aspartame for over 90% of the market’s needs. As the expiry date of its patent in Canada approached, NSC had systematically begun to insert exclusivity clauses into all of its supply contracts. In addition to these exclusivity clauses, NSC’s customers were further incentivized to use NSC as a result of substantial financial inducements, including logo allowances and cooperative marketing allowances. These “fidelity” rebates encouraged the promotion of logos in products made exclusively with aspartame. If a product was made with a blend of different sweetener, the logo could not be used and the customer was not entitled to these rebates. The Director brought an application under sections 77 and 79, claiming these practices were contrary to the exclusive dealing and abuse of dominance provisions. The Director had attempted to infer from the inclusion of the exclusivity clauses that customers were effectively hostages to NSC because a refusal to accept this term or to comply with it could have resulted in a refusal to supply. The Competition Tribunal stated that the mere existence of an exclusivity clause was not in itself conclusive of the existence of a condition of exclusivity. Whether customers are really subjected to the influence of a supplier had to be determined from the context surrounding the inclusion of that clause. However, the Tribunal held that the fidelity rebates in the supply contracts constituted clear financial inducements to customers both to deal only in the respondent’s brand of aspartame and to refrain from using another producer’s aspartame. Thus, as a result of the combination of financial 77 NutraSweet, supra note 37; Laidlaw Waste Systems, supra note 49; Canada Pipe (Comp. Trib.), supra note 73.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 463

A Canadian Perspective on Tied Selling and Exclusive Dealing 463 inducements and exclusivity clauses, the definitional requirement of “exclusive dealing” was met: “Therefore we conclude that the financial incentives and the exclusivity clause amount to exclusive dealing within the meaning of para. 77(1)(b): the customers clearly agreed to deal only or primarily in the products of NSC and in return received various rebates whose existence depends on exclusive use of NutraSweet brand aspartame.”78

With respect to the key question of whether there had been a substantial lessening of competition, the Tribunal referred to its reasoning regarding the same issue as discussed under the abuse of dominance claim. In that discussion, the Tribunal phrased the test as follows: “In essence, the question to be decided is whether the anticompetitive acts engaged in by NSC preserve or add to NSC’s market power. The issue with respect to the contract terms associated with exclusivity . . . is the degree to which these anticompetitive acts add to the entry barriers into the Canadian market and, additionally therefore, into the industry.” 79

Based on the evidence, which included the fact that the exclusive use clauses appeared in virtually all of NSC’s contracts and thus covered 90 percent of the Canadian market, the Tribunal was convinced that “the exclusivity in NSC’s contracts, which includes both the clauses reflecting an agreement to deal only or primarily in NutraSweet brand aspartame and the financial inducements to do so, impedes ‘toe-hold’ entry into the market and inhibits the expansion of other firms in the market”.80 Accordingly, the requisite exclusionary effect element was satisfied. The Tribunal was not persuaded by NSC’s assertion that the supply contracts were renewable on a yearly basis. Nor did it agree that exclusivity promoted efficient supply distribution and lower inventory costs when managed by a single firm, although the Tribunal did leave the door open where an industry has such special characteristics that may justify this claimed source of cost-savings. Notably, the Tribunal decided not to label this argument as one based on “efficiencies”: “This line of reasoning, it should be noted, is not an ‘efficiency defence’. It leads, rather, to the conclusion that customers are, on balance, better off as a result of exclusivity and that they pass these cost savings on to consumers. Under exclusivity customers are able to negotiate a band of minimum and maximum purchases. Without it, the customers would presumably have to commit to a specific volume and would have to hold inventories to satisfy higher-than-anticipated demand, or would have to make higher-than-required purchases in the event that requirements were less than anticipated. An executive of a major customer stated that the broad band negotiated by the customer in its exclusive contract does not mean that this is an important consideration; the customer is, in any event, quite capable of 78 79 80

NutraSweet, supra note 37 at 54. Ibid. at 47 (emphasis supplied). Ibid. at 48.

(U) Ehlermann 08 Goldman

464

6/11/08

11:53

Page 464

Session Two: A Reformed Approach to Exclusionary Conduct

accurately forecasting demand for its product. Whatever the customer’s abilities, the Tribunal does not see much merit in the respondent’s line of reasoning. It can always be claimed that the risk and cost of holding plant and inventory are reduced if there is a single supplier rather than several. Unless it can be shown that an industry has special characteristics that make this claimed source of cost savings important, there is no reason to give it any weight.”81

Perhaps the Tribunal felt that it could not label this argument as an “efficiency” defence because of the absence of any explicit reference to an efficiency justification under section 77. However, it is clear that the Tribunal felt the need to explain its economic rationale on this particular point. This may suggest that the Tribunal was in fact mindful of the possible efficiencybased justifications of exclusive dealing, even if such considerations are not explicitly set out in section 77. The Tribunal also dismissed arguments that exclusivity was necessary to protect against free riding, albeit in a rather abbreviated fashion, simply noting that “NSC is [not] entitled to any more protection against competition than it was able to obtain through patent grants that provided it with a considerable head start on potential competitors”.82

B. Commissioner of Competition v. Canada Pipe Ltd. The exclusive dealing provisions were considered again, and for the first time by the Canadian Federal Court of Appeal (“FCA”), in the more recent case of Canada Pipe.83 At issue in Canada Pipe was the stocking distributor program (“SDP”) offered by Bibby Ste-Croix—a division of Canada Pipe, which provided quarterly and annual rebates to distributors who bought certain products exclusively from that division. The Commissioner of Competition contended that the SDP constituted a practice of exclusive dealing contrary to section 77, as well as an abuse of dominance contrary to section 79. With respect to exclusive dealing, the Competition Tribunal had found that Bibby Ste-Croix was a “major supplier” and that the practice at issue met the definition of exclusive dealing under section 77; however, it did not find that the SDP was an anticompetitive act for purposes of the abuse of dominance provisions. In any event, there was no finding of a substantial lessening or prevention of competition because of the apparent ease of entry, low switching costs and lack of exclusionary effect.84 The Commissioner appealed to the FCA, which engaged in an extensive analysis applying the tests mandated by the language of the Act. The FCA, in a decision issued in June 2006, went out of its way to note that section 77 and sec81 82 83 84

NutraSweet, supra note 37 at 51–52. Ibid. at 52. Canada Pipe, supra note 1. Canada Pipe (Comp. Trib.), supra note 73, at paras. 263–70.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 465

A Canadian Perspective on Tied Selling and Exclusive Dealing 465 tion 79, although based on similar logic and fundamental principles, contained different statutory language regarding the test for exclusionary effects. For example, although not germane to this particular case, the FCA noted that, on the one hand, the wording of section 79 appears to embrace all past, present and future effects of the impugned anticompetitive act, while on the other hand, section 77 seems limited to present and future (but not past) effects.85 Moreover, the FCA observed a further distinction between the two provisions. In the opinion of the FCA, the term “impede” found at section 77 called for a “broader perspective” relative to the scope indicated by the term “prevent” used in section 79. As a result, the scope of section 77 may be more encompassing than initially interpreted by the Tribunal. After taking stock of the differences between the language of sections 77 and 79, the FCA explained that the Tribunal had erred with respect to section 77 because it had misconstrued the appropriate test to apply to the determination of a substantial lessening of competition for the same reasons that it had erred with respect to the abuse of dominance claims. Accordingly the FCA referred to and relied upon the discussion with respect to the abuse of dominance in explaining the appropriate standard for a substantial lessening or prevention of competition: “In order to achieve the inquiry dictated by the statutory language of paragraph 79(1)(c), the Tribunal must compare the level of competitiveness in the presence of the impugned practice with that which would exist in the absence of the practice, and then determine whether the preventing or lessening of competition, if any, is ‘substantial’. This comparison must be done with reference to actual effects in the past and present, as well as likely future effects. Only through such a comparative approach can the Tribunal determine, as the statutory provision requires, whether the impugned practice ‘has had, is having or is likely to have the effect of preventing or lessening competition substantially.’ ”86

Following its publication, there has been some debate over whether the FCA’s ruling actually altered the exclusionary effects test first established in NutraSweet. The novel aspect of the Canada Pipe decision is the establishment of the “but for” test to determine the substantial lessening of competition.87 The Federal Court of Appeal clarified that the language of the provision required an examination of whether the relevant market would be substantially more competitive “but for” the alleged anticompetitive practice. As a result, the essence of the test is to compare the state of the market in the presence of the impugned practice with the same market without it. Leave to appeal the FCA’s decision was recently denied by the Supreme Court of Canada.88 Accordingly, the matter is remanded to the Tribunal for 85 86 87 88

Canada Pipe, supra note 1 at para. 94. Ibid. at para. 37. Canada Pipe, supra note 1 at paras. 36–38. Supra note 56.

(U) Ehlermann 08 Goldman

466

6/11/08

11:53

Page 466

Session Two: A Reformed Approach to Exclusionary Conduct

further consideration in line with the FCA’s reasoning. The extent to which efficiency or other business justifications are considered by the Tribunal in its assessment of exclusive dealing remains to be seen. On the one hand, there is no explicit statutory provision that allows for such considerations in section 77 (as there is for “superior competitive performance” under the abuse of dominance provision in section 79), and the FCA has made it clear that section 77 is a distinct statutory provision that must be read on its own. On the other hand, however, the purpose clause of the Competition Act, which the FCA has said should guide the interpretation of statutory provisions throughout, seems to allow for such considerations. Decisions like NutraSweet and Canada Pipe contribute to the ongoing debate surrounding the appropriate treatment of loyalty-based discounts or other similar rebate programs by dominant players. These cases demonstrate the difficulty in assessing whether a practice is truly anticompetitive and whether it has the types of effects on competition that merit prohibition.

VI. Conclusion In Canada, the provisions governing tied selling and exclusive dealing have been part of the relevant competition law statute for over 30 years. Yet there remain relatively few decided cases concerning these provisions and no cases have been brought by the Commissioner outside of the abuse of dominance context. There are no agency guidelines or other forms of official guidance available to help interpret these provisions. The Competition Bureau continues to wrestle with the task of finding the appropriate dividing line between anticompetitive tied selling or exclusive dealing practices and conduct that has a legitimate business rationale and basis. In Canada Pipe, the Federal Court of Appeal did provide some guidance by clarifying the test for a substantial lessening of competition, albeit primarily in the context of abuse of dominance. However, the Court also stated that the exclusive dealing provision (and, presumably, the companion tied selling provision) must be interpreted in its own light, keeping in mind all of the aims enunciated in the purpose clause of the Competition Act. Moreover, tied selling and exclusive dealing claims may now be brought before the Competition Tribunal by private parties, a further indication that the rules applying to these practices are stand-alone provisions. While there has been little case law to date, it will be interesting to see how the legal and economic debate concerning these provisions continues to evolve in Canada and across other jurisdictions.

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 467

A Canadian Perspective on Tied Selling and Exclusive Dealing 467

APPENDIX: Relevant Provisions of the Competition Act Purpose of Act 1.1 The purpose of this Act is to maintain and encourage competition in Canada in order to promote the efficiency and adaptability of the Canadian economy, in order to expand opportunities for Canadian participation in world markets while at the same time recognizing the role of foreign competition in Canada, in order to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy and in order to provide consumers with competitive prices and product choices.

Exclusive Dealing, Market Restriction and Tied Selling Provisions Definitions 77. (1) For the purposes of this section, “exclusive dealing” “exclusive dealing” means (a) any practice whereby a supplier of a product, as a condition of supplying the product to a customer, requires that customer to (i)

deal only or primarily in products supplied by or designated by the supplier or the supplier’s nominee, or (ii) refrain from dealing in a specified class or kind of product except as supplied by the supplier or the nominee, and (b) any practice whereby a supplier of a product induces a customer to meet a condition set out in subparagraph (a)(i) or (ii) by offering to supply the product to the customer on more favourable terms or conditions if the customer agrees to meet the condition set out in either of those subparagraphs; “market restriction” means any practice whereby a supplier of a product, as a condition of supplying the product to a customer, requires that customer to supply any product only in a defined market, or exacts a penalty of any kind from the customer if he supplies any product outside a defined market; “tied selling” means (a) any practice whereby a supplier of a product, as a condition of supplying the product (the “tying” product) to a customer, requires that customer to

(U) Ehlermann 08 Goldman

468

6/11/08

11:53

Page 468

Session Two: A Reformed Approach to Exclusionary Conduct (i)

acquire any other product from the supplier or the supplier’s nominee, or (ii) refrain from using or distributing, in conjunction with the tying product, another product that is not of a brand or manufacture designated by the supplier or the nominee, and (b) any practice whereby a supplier of a product induces a customer to meet a condition set out in subparagraph (a)(i) or (ii) by offering to supply the tying product to the customer on more favourable terms or conditions if the customer agrees to meet the condition set out in either of those subparagraphs. Exclusive dealing and tied selling (2) Where, on application by the Commissioner or a person granted leave under section 103.1, the Tribunal finds that exclusive dealing or tied selling, because it is engaged in by a major supplier of a product in a market or because it is widespread in a market, is likely to (a) impede entry into or expansion of a firm in a market, (b) impede introduction of a product into or expansion of sales of a product in a market, or (c) have any other exclusionary effect in a market, with the result that competition is or is likely to be lessened substantially, the Tribunal may make an order directed to all or any of the suppliers against whom an order is sought prohibiting them from continuing to engage in that exclusive dealing or tied selling and containing any other requirement that, in its opinion, is necessary to overcome the effects thereof in the market or to restore or stimulate competition in the market. Market restriction (3) Where, on application by the Commissioner or a person granted leave under section 103.1, the Tribunal finds that market restriction, because it is engaged in by a major supplier of a product or because it is widespread in relation to a product, is likely to substantially lessen competition in relation to the product, the Tribunal may make an order directed to all or any of the suppliers against whom an order is sought prohibiting them from continuing to engage in market restriction and containing any other requirement that, in its opinion, is necessary to restore or stimulate competition in relation to the product. Damage awards (3.1) For greater certainty, the Tribunal may not make an award of damages under this section to a person granted leave under subsection 103.1(7). Where no order to be made and limitation on application of order

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 469

A Canadian Perspective on Tied Selling and Exclusive Dealing 469 (4) The Tribunal shall not make an order under this section where, in its opinion, (a) exclusive dealing or market restriction is or will be engaged in only for a reasonable period of time to facilitate entry of a new supplier of a product into a market or of a new product into a market, (b) tied selling that is engaged in is reasonable having regard to the technological relationship between or among the products to which it applies, or (c) tied selling that is engaged in by a person in the business of lending money is for the purpose of better securing loans made by that person and is reasonably necessary for that purpose, and no order made under this section applies in respect of exclusive dealing, market restriction or tied selling between or among companies, partnerships and sole proprietorships that are affiliated. Where company, partnership or sole proprietorship affiliated (5) For the purposes of subsection (4), (a) one company is affiliated with another company if one of them is the subsidiary of the other or both are the subsidiaries of the same company or each of them is controlled by the same person; (b) if two companies are affiliated with the same company at the same time, they are deemed to be affiliated with each other; (c) a partnership or sole proprietorship is affiliated with another partnership, sole proprietorship or a company if both are controlled by the same person; and (d) a company, partnership or sole proprietorship is affiliated with another company, partnership or sole proprietorship in respect of any agreement between them whereby one party grants to the other party the right to use a trade-mark or trade-name to identify the business of the grantee, if (i)

the business is related to the sale or distribution, pursuant to a marketing plan or system prescribed substantially by the grantor, of a multiplicity of products obtained from competing sources of supply and a multiplicity of suppliers, and (ii) no one product dominates the business. When persons deemed to be affiliated (6) For the purposes of subsection (4) in its application to market restriction, where there is an agreement whereby one person (the “first” person) supplies or causes to be supplied to another person (the “second” person) an ingredient or ingredients that the second person processes by the addition of labour and material into an article of food or drink that he then sells in

(U) Ehlermann 08 Goldman

470

6/11/08

11:53

Page 470

Session Two: A Reformed Approach to Exclusionary Conduct association with a trade-mark that the first person owns or in respect of which the first person is a registered user, the first person and the second person are deemed, in respect of the agreement, to be affiliated.

Inferences (7) In considering an application by a person granted leave under section 103.1, the Tribunal may not draw any inference from the fact that the Commissioner has or has not taken any action in respect of the matter raised by the application.

Abuse of Dominant Position Definition of “anti-competitive act” 78. (1) For the purposes of section 79, “anti-competitive act”, without restricting the generality of the term, includes any of the following acts: (a) squeezing, by a vertically integrated supplier, of the margin available to an unintegrated customer who competes with the supplier, for the purpose of impeding or preventing the customer’s entry into, or expansion in, a market; (b) acquisition by a supplier of a customer who would otherwise be available to a competitor of the supplier, or acquisition by a customer of a supplier who would otherwise be available to a competitor of the customer, for the purpose of impeding or preventing the competitor’s entry into, or eliminating the competitor from, a market; (c) freight equalization on the plant of a competitor for the purpose of impeding or preventing the competitor’s entry into, or eliminating the competitor from, a market; (d) use of fighting brands introduced selectively on a temporary basis to discipline or eliminate a competitor; (e) pre-emption of scarce facilities or resources required by a competitor for the operation of a business, with the object of withholding the facilities or resources from a market; ( f ) buying up of products to prevent the erosion of existing price levels; (g) adoption of product specifications that are incompatible with products produced by any other person and are designed to prevent his entry into, or to eliminate him from, a market; (h) requiring or inducing a supplier to sell only or primarily to certain customers, or to refrain from selling to a competitor, with the object of preventing a competitor’s entry into, or expansion in, a market; (i) selling articles at a price lower than the acquisition cost for the purpose of disciplining or eliminating a competitor; ( j) acts or conduct of a person operating a domestic service, as defined in subsection 55(1) of the Canada Transportation Act, that are specified under paragraph (2)(a); and

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 471

A Canadian Perspective on Tied Selling and Exclusive Dealing 471 (k) the denial by a person operating a domestic service, as defined in subsection 55(1) of the Canada Transportation Act, of access on reasonable commercial terms to facilities or services that are essential to the operation in a market of an air service, as defined in that subsection, or refusal by such a person to supply such facilities or services on such terms. Regulations (2) The Governor in Council may, on the recommendation of the Minister and the Minister of Transport, make regulations (a) specifying acts or conduct for the purpose of paragraph (1)( j); and (b) specifying facilities or services that are essential to the operation of an air service for the purpose of paragraph (1)(k). Prohibition where abuse of dominant position 79. (1) Where, on application by the Commissioner, the Tribunal finds that (a) one or more persons substantially or completely control, throughout Canada or any area thereof, a class or species of business, (b) that person or those persons have engaged in or are engaging in a practice of anti-competitive acts, and (c) the practice has had, is having or is likely to have the effect of preventing or lessening competition substantially in a market, the Tribunal may make an order prohibiting all or any of those persons from engaging in that practice. Additional or alternative order (2) Where, on an application under subsection (1), the Tribunal finds that a practice of anti-competitive acts has had or is having the effect of preventing or lessening competition substantially in a market and that an order under subsection (1) is not likely to restore competition in that market, the Tribunal may, in addition to or in lieu of making an order under subsection (1), make an order directing any or all the persons against whom an order is sought to take such actions, including the divestiture of assets or shares, as are reasonable and as are necessary to overcome the effects of the practice in that market. Limitation (3) In making an order under subsection (2), the Tribunal shall make the order in such terms as will in its opinion interfere with the rights of any person to whom the order is directed or any other person affected by it only to the extent necessary to achieve the purpose of the order.

(U) Ehlermann 08 Goldman

472

6/11/08

11:53

Page 472

Session Two: A Reformed Approach to Exclusionary Conduct

Administrative monetary penalty (3.1) Where the Tribunal makes an order under subsection (1) or (2) against an entity who operates a domestic service, as defined in subsection 55(1) of the Canada Transportation Act, it may also order the entity to pay, in such manner as the Tribunal may specify, an administrative monetary penalty in an amount not greater than $15 million. Aggravating or mitigating factors (3.2) In determining the amount of an administrative monetary penalty, the Tribunal shall take into account the following: (a) the frequency and duration of the practice; (b) the vulnerability of the class of persons adversely affected by the practice; (c) injury to competition in the relevant market; (d) the history of compliance with this Act by the entity; and (e) any other relevant factor. Purpose of order (3.3) The purpose of an order under subsection (3.1) is to promote practices that are in conformity with this section, not to punish. Superior competitive performance (4) In determining, for the purposes of subsection (1), whether a practice has had, is having or is likely to have the effect of preventing or lessening competition substantially in a market, the Tribunal shall consider whether the practice is a result of superior competitive performance. Exception (5) For the purpose of this section, an act engaged in pursuant only to the exercise of any right or enjoyment of any interest derived under the Copyright Act, Industrial Design Act, Integrated Circuit Topography Act, Patent Act, Trade-marks Act or any other Act of Parliament pertaining to intellectual or industrial property is not an anti-competitive act. Limitation period (6) No application may be made under this section in respect of a practice of anti-competitive acts more than three years after the practice has ceased. Where proceedings commenced under section 45 or 92 (7) No application may be made under this section against a person (a) against whom proceedings have been commenced under section 45, or (b) against whom an order is sought under section 92

(U) Ehlermann 08 Goldman

6/11/08

11:53

Page 473

A Canadian Perspective on Tied Selling and Exclusive Dealing 473 on the basis of the same or substantially the same facts as would be alleged in the proceedings under section 45 or 92, as the case may be.

Private Action Provision Leave to make application under section 75 or 77 103.1 (1) Any person may apply to the Tribunal for leave to make an application under section 75 or 77. The application for leave must be accompanied by an affidavit setting out the facts in support of the person’s application under section 75 or 77. Notice (2) The applicant must serve a copy of the application for leave on the Commissioner and any person against whom the order under section 75 or 77 is sought. Certification by Commissioner (3) The Commissioner shall, within 48 hours after receiving a copy of an application for leave, certify to the Tribunal whether or not the matter in respect of which leave is sought (a) is the subject of an inquiry by the Commissioner; or (b) was the subject of an inquiry that has been discontinued because of a settlement between the Commissioner and the person against whom the order under section 75 or 77 is sought. Application discontinued (4) The Tribunal shall not consider an application for leave respecting a matter described in paragraph (3)(a) or (b) or a matter that is the subject of an application already submitted to the Tribunal by the Commissioner under section 75 or 77. Notice by Tribunal (5) The Tribunal shall as soon as practicable after receiving the Commissioner’s certification under subsection (3) notify the applicant and any person against whom the order is sought as to whether it can hear the application for leave. Representations (6) A person served with an application for leave may, within 15 days after receiving notice under subsection (5), make representations in writing to the Tribunal and shall serve a copy of the representations on any other person referred to in subsection (2).

(U) Ehlermann 08 Goldman

474

6/11/08

11:53

Page 474

Session Two: A Reformed Approach to Exclusionary Conduct

Granting leave to make application under section 75 or 77 (7)

The Tribunal may grant leave to make an application under section 75 or 77 if it has reason to believe that the applicant is directly and substantially affected in the applicants’ business by any practice referred to in one of those sections that could be subject to an order under that section.

Time and conditions for making application (8)

The Tribunal may set the time within which and the conditions subject to which an application under section 75 or 77 must be made. The application must be made no more than one year after the practice that is the subject of the application has ceased.

Decision (9)

The Tribunal must give written reasons for its decision to grant or refuse leave and send copies to the applicant, the Commissioner and any other person referred to in subsection (2).

Limitation (10) The Commissioner may not make an application for an order under section 75, 77 or 79 on the basis of the same or substantially the same facts as are alleged in a matter for which the Tribunal has granted leave under subsection (7), if the person granted leave has already applied to the Tribunal under section 75 or 77. Inferences (11) In considering an application for leave, the Tribunal may not draw any inference from the fact that the Commissioner has or has not taken any action in respect of the matter raised by it. Inquiry by Commissioner (12) If the Commissioner has certified under subsection (3) that a matter in respect of which leave was sought by a person is under inquiry and the Commissioner subsequently discontinues the inquiry other than by way of settlement, the Commissioner shall, as soon as practicable, notify that person that the inquiry is discontinued.

(W) Ehlermann 08 Paulis

6/11/08

11:53

Page 515

I Emil Paulis* Article 82 EC and Exploitative Conduct June 2007

1. Introduction The history of the views concerning the correct balance between enforcement against exploitative abuses and against exclusionary abuses is an interesting one. In the early days of European competition policy, some commentators were of the opinion that, from a legal perspective, Article 82 (then Article 86) was exclusively concerned with exploitative abuses.1 This debate effectively ended with the judgment of the European Court of Justice (ECJ) in Continental Can, which clarified that Article 82 also covers exclusionary abuses. The Court stated that Article 82 “is not only aimed at practices which may cause damage to consumers directly, but also at those which are detrimental to them through their impact on an effective competition structure”.2 Many contributors to the recent lively debate on Article 82 take a view almost completely opposite to that of the early commentators mentioned above, arguing that enforcement of Article 82 should concentrate on exclusionary abuses; competition authorities should thus either abstain from or be very restrained in enforcement actions against exploitative abuses.3 In this paper I will present my views on the proper role of enforcement against exploitative conduct under Article 82 EC. This includes both some thoughts on the general characteristics of markets in which it makes sense to

* Director for Policy and Strategic Support, DG Competition, European Commission. The views expressed herein are those of the author and do not necessarily reflect those of the Commission. 1 See, e.g., René Joliet, Monopolization and Abuse of Dominant Position, Martinus Nijhoff, 1970. 2 Case 6/72, Europemballage Corporation and Continental Can Company, Inc. v Commission [1973] ECR 215, para. 26. 3 See, e.g., David Evans and Jorge Padilla, “Excessive Prices: Using Economics to Define Administrable Legal Rules”, 1 Journal of Competition Law and Economics 97 (2005); Bruce Lyons, “The Paradox of the Exclusion of Exploitative Abuse”, Centre for Competition Policy Newsletter, Spring 2007; Massimo Motta and Alexandre de Streel, “Exploitative and Exclusionary Excessive Prices in EU Law”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What Is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 91 et seq.; Robert O’Donoghue and Jorge Padilla, The Law and Economics of Article 82 EC, Hart Publishing, 2006.

(W) Ehlermann 08 Paulis

516

6/11/08

11:53

Page 516

Session Three: Exploitative Conduct And The Remedies

intervene, as well as a short discussion of methodologies for identifying exploitative conduct in such markets. I distinguish here between “exploitative” and “discriminatory” conduct. The paper does not deal with discriminatory conduct, which raises particular issues not necessarily related to exploitation, such as discrimination as part of exclusionary conduct and discrimination based on nationality. The ECJ in Continental Can made clear that the potentially abusive practices mentioned in the text of Article 82 do not form “an exhaustive enumeration of the sort of abuses of a dominant position prohibited by the Treaty”.4 Nevertheless, most discussions of exploitative abuses take as their starting point Article 82(a), which states that an abuse may consist in “directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions”. Exploitative conduct may thus both concern pricing and “other” trading conditions. Most recent contributions focus on the pricing aspect— and then only on selling prices. I will also mostly discuss this aspect, which usually is called “excessive” pricing.

2. Why take action against exploitative conduct? At first sight it may seem strange to question why competition authorities should take enforcement actions against exploitative conduct. After all, many competition authorities seem to think that competition policy is ultimately about protecting consumer welfare. Commissioner Kroes, for example, stated in her 2005 Fordham speech that “Article 82 enforcement should focus on . . . behavior that has actual or likely effects on the market, which harms consumers”.5 However, she went on to say that “it is sound for our enforcement policy to give priority to so-called exclusionary abuses, since exclusion is often at the basis of later exploitation of customers”. This is what one commentator has called the “paradox of the exclusion of exploitative abuse”.6 Various arguments have been brought forward to explain this “paradox”. They may be divided into two types. One type of argument focuses on the practical difficulties of competition authorities to intervene against excessive prices. The other type of argument focuses more on what one (only slightly polemically) could call the “positive effects” of excessive prices in a market economy. My discussion below will show that I have more sympathy for the first type of argument than for the second. 4

Supra note 2, para. 26. Neelie Kroes, “Tackling Exclusionary Practices to Avoid Exploitation of Market Power: Some Preliminary Thoughts on the Policy Review of Article 82”, 29 Fordham International Law Journal 593 (2005). 6 Lyons, supra note 3. 5

(W) Ehlermann 08 Paulis

6/11/08

11:53

Page 517

Article 82 and Exploitative Conduct 517 There are two basic reasons why enforcement actions against excessive prices are particularly difficult—and especially so for a “generalist” competition authority. First, determining whether a specific price is “excessive” involves complicated comparisons of prices with costs of production and investment. This may involve difficult decisions about the profitability of a dominant firm. Determining whether a price is excessive may also involve difficult comparisons with whatever useful “benchmark” prices can be identified. Some of the problems involved in these comparisons—for example, the issue of cost allocation in multi-product firms—are also present for other price-based abuses. However, when these problems are “solved” for the other abuses, the price/cost question becomes relatively “simple” in that the issue is whether the price is higher or lower than some well-defined cost measure. To determine whether excessive pricing has taken place, there is another layer of complication since it has to be decided whether a price—which may be higher than all relevant cost measures—is in fact too high. According to many commentators, such a decision will necessarily be somewhat arbitrary, unless one takes the rather draconian position that any price over some well-defined cost benchmark is excessive. Second, intervening against excessive pricing may entail the risk of a competition authority finding itself in the situation of a semi-permanent quasiregulator. The authority may have to come back time and again to the pricing of the dominant firm when cost or other conditions change in the industry, something that a “generalist” competition authority is much less equipped for than proper regulators with their deep knowledge of and continuous involvement in their industries. An authority may be able to establish a simple price comparison rule that can avoid such a situation. An example of such a rule could be that the dominant firm cannot charge more (or only X % more) in market A than it does in market B where the freely determined price in market B is for some reason more acceptable than the freely determined price in market A. There may still be recurring problems where such a rule can be imposed; the dominant firm may come back after a few years, claiming that conditions have changed and that the rule needs to be revised. But at least the problems seem of a lesser magnitude than a rule establishing a link between price and costs, as costs are normally less easy to observe than other prices. I consider these practical difficulties so convincing and the risk of competition authorities arriving at the wrong result so great that enforcement actions against exploitative conduct in my view should only be taken as a last resort. In many markets, prices may temporarily be high but once market forces have had the time to play out the prices will come back to more normal levels. In such cases it would be unwise to run the risk of taking a wrong decision, and furthermore to spend enforcement resources, on solving a problem that would solve itself over time anyway. Note that price erosion can occur even where entry barriers are so high that there may be a dominant firm operating on the market. Of course, it may be that a dominant firm tries to prevent

(W) Ehlermann 08 Paulis

518

6/11/08

11:53

Page 518

Session Three: Exploitative Conduct And The Remedies

prices from dropping by raising entry barriers artificially. In such a situation it is wiser for the competition authority to tackle these entry barriers directly, since they will likely amount to an exclusionary abuse. However, if the market is characterized by such “natural” entry barriers that it is unlikely that market forces over time will bring prices down, enforcement actions aimed directly against excessive prices may be appropriate. I will be a bit more specific on this shortly. First, I would like to be clear about what I am not arguing. The other argument—besides the practical difficulties that I have just discussed—was the “positive effects” argument. This has perhaps most famously been expressed by Justice Scalia in his opinion in Trinko,7 where he argued that “[t]he mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system. The opportunity to charge monopoly prices—at least for a short period—is what attracts ‘business acumen’ in the first place; it induces risk taking that produces innovation and economic growth”.8 I realize that Scalia did soften his argument slightly by adding: “at least for a short period”. This makes it rather difficult to argue vehemently against his position, since it is not entirely clear what Scalia meant by “a short period”. Let us therefore pretend, for the sake of argument, that this apparent qualification does not appear in the judgment. The argument would then be that monopoly profits are good, perhaps even necessary, because that is what attracts the type of risk taking and investment that drives innovation and economic growth. I do not want to argue against human nature, and I therefore fully accept that much risk taking and investment is indeed done in the hope of achieving significant financial returns.9 What I do object to is the idea that the only—or the “optimal”—way of giving dominant firms the correct incentives is to allow them to charge monopoly profits without any possibility of intervention on the part of competition authorities. Taking that argument to the extreme, it would seem difficult to justify interventions against exclusionary abuses. Restraining the possibility of dominant firms to exclude others probably reduces their profits, which—according to this logic—would have a negative impact on risk taking and innovation. It seems to me that the position behind having provisions such as Article 82 can only be that “not all monopoly profits are good”. It is then up to competition authorities—of course under the control of courts—to figure out which monopoly profits are “good” and which are not. And here I fully accept that—as a matter of administrability—it may be very difficult to make that call when deciding whether to intervene against excessive prices. So much so that some legal systems do not provide for that possibility at all. 7

Verizon Communications, Inc v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004). Ibid. at 407. 9 I suspect, however, that many innovators are just as motivated by the sheer joy and stimulus of the innovative process itself. 8

(W) Ehlermann 08 Paulis

6/11/08

11:53

Page 519

Article 82 and Exploitative Conduct 519 This raises the obvious question of why the possibility is envisaged in Article 82 EC but not in the Sherman Act or in other US antitrust laws. One explanation may be found in the fact that Article 86 does not prohibit the acquisition of dominance through unilateral abusive behaviour, and that this justifies greater protection against direct exploitation of consumers by dominant firms. After all, it can be considered that the reward of a dominant position need not come from unfair exploitation of consumers, as the dominant firm has enough incentives resulting from the revenues gained from its higher volume of sales at a price which will in most cases be at a supracompetitive level anyhow.

3. What markets are candidates for intervention? Several commentators have recently given their opinion on what kinds of markets could be candidates for intervention by competition authorities against excessive prices. Let me quickly give a flavour of some of these contributions before offering my own thoughts. Motta and de Streel10 argue that intervention should be limited to industries where there are: (a) high and non-transitory barriers to entry; and (b) monopoly or near-monopoly situations due to current or past exclusive or special rights. They add two “institutional issues”. First, there should be no effective way for the competition authority to eliminate the entry barriers. Second, there should be no sector-specific regulator. Evans and Padilla11 suggest, as candidates for intervention, circumstances where: (a) the firm enjoys a (near) monopoly position in the market; (b) the prices charged by the firm widely exceed its average total costs; and (c) there is a risk that those prices may prevent the emergence of new goods and services in adjacent markets. O’Donoghue and Padilla conclude the chapter on excessive prices in their book on Article 82 by stating that:12 “[t]here is a growing consensus on the need to identify administrable limiting principles to ensure that Article 82(a) is enforced only when strictly necessary, i.e. minimising the likelihood of false convictions. The emerging consensus is that intervention should be restricted to industries: (1) protected by high barriers to entry; (2) where one firm enjoys considerable market power; and (3) where investment and innovation play a relatively minor role.”13 10

Supra note 3. Supra note 3. 12 Supra note 3, at p. 638 (footnotes omitted). 13 Later on the same page they reformulate these conditions slightly to: “(1) the market is protected by high barriers to entry; (2) consumers have no credible alternatives to the products of the dominant firm; and (3) firms compete in a mature environment, where investment and innovation play little or no role”. 11

(W) Ehlermann 08 Paulis

520

6/11/08

11:53

Page 520

Session Three: Exploitative Conduct And The Remedies

Let me continue by commenting briefly on these proposals—and in the process I will formulate my own views. The one element common to the proposals is the presence of high entry barriers. I agree that competition authorities should not intervene in markets where it is likely that normal competitive forces over time will eliminate the possibilities of a dominant company to charge high prices. Intervention should therefore indeed be limited to markets characterized by very high and long-lasting barriers to entry— although I would add barriers to expansion. Motta and de Streel would furthermore require that there be a monopoly or a near-monopoly situation due to current or past exclusive or special rights. This condition has two parts: first, it requires “a monopoly or a nearmonopoly”; second, this market position must be due to “current or past exclusive or special rights”. A “monopoly or near-monopoly” is presumably something more than a dominant position. If there are very high and longlasting barriers to entry and expansion, I believe that we can be confident that the position of the dominant firm will not be challenged for the foreseeable future. I must admit that I do not then see the need to introduce a category of “monopoly”—or “near monopoly”, for that matter—into our analysis. However, it almost sounds as if Motta and de Streel think that the combination of a dominant position and “high and non-transitory barriers to entry” more or less automatically will lead to a situation of “monopoly or near monopoly”, so perhaps in reality our views on this point are not that far apart. However, I am not convinced by the proposal to restrict intervention to situations where current or past exclusive or special rights are the cause of the dominant position (or, according to Motta and de Streel, “monopoly or near monopoly”). I don’t see why competition authorities need be restricted in such a way. My angle would rather be to focus once again on the question of “high and long-lasting barriers to entry and expansion”. A legal monopoly would certainly qualify as a situation where such barriers exist. But high and longlasting barriers to entry and expansion would normally also be found in situations of natural monopoly. These are, of course, also the sectors where there often are or have been exclusive or special rights, so there will be a certain overlap anyway. But still, I prefer to focus directly on the barriers to entry and expansion, and not to be restricted as to which barriers I can look at. As to the “institutional issue” of whether we should intervene when there is a sector-specific regulator, I do not think that the Commission can say that it will never intervene whenever there is such a (national) regulator. Of course, I hope that, in practice, we would not have to do so. However, there are examples of the Commission intervening even though a national regulator either had decided not to intervene or had positively endorsed the behaviour of a dominant company.14 These were not excessive pricing cases, but the princi14 See Commission Decision 2003/707 of 21 May 2003, Deutsche Telekom AG [2003] OJ L263/9, upheld on appeal: Case T-271/03, Deutsche Telekom AG v Commission, judgment of the

(W) Ehlermann 08 Paulis

6/11/08

11:53

Page 521

Article 82 and Exploitative Conduct 521 ple is the same. The Commission should maintain the option to intervene when a national regulator is not acting or when it takes decisions that are not in conformity with Community law. The other institutional issue mentioned by Motta and de Streel is that there should not be an effective way for the competition authority to eliminate the entry barriers. They are not referring here to the obvious example of strategic entry barriers, which can be attacked by the competition authority as exclusionary abuses. What they have in mind are “current legal barriers”, where they argue that it “may be more cost-effective for the authority to lobby the government to lift the barriers and liberalise effectively the sector than to open several exploitative abuse cases”. I entirely agree with the general proposition that competition authorities should always intervene in the most effective way. At the Community level the considerable efforts spent on liberalizing certain sectors can be seen as an example of what Motta and de Streel propose. However, it may be that, generally, such “lobbying efforts” by competition authorities are more effective at the national level than at the Community level. Consumers may pay excessive prices for quite some time before Community legislation removes the legal barriers. It may therefore be necessary to work on two fronts at the same time, that is, by both “lobbying” and intervening directly against excessive prices. Evans and Padilla propose a condition that I have not discussed so far, namely that intervention should only take place if the excessive prices are such that they may prevent the emergence of new goods and services in adjacent markets. They point out that this condition “corresponds” to the ECJ’s “new product test” in Magill 15 and IMS Health.16 However, they would apply this test in situations beyond the exclusionary refusal to supply cases such as Magill and IMS Health to situations where it is consumers that are not served because prices are too high. They mention General Motors and British Leyland as cases consistent with this approach. While I find the idea interesting, I do not think the Commission can restrict itself to intervening only in such cases. Finally, O’Donoghue and Padilla introduce yet another idea that competition authorities should only intervene where investment and innovation play a “minor role.” I think that in practice this will indeed often be the case. In many markets with considerable investment and innovation, barriers to entry and expansion may be high, but not necessarily long-lasting. Furthermore, as I will discuss below, investment costs should be taken into account when Court of First Instance of 10 April 2008, not yet reported. See also Commission Decision of 4 July 2007, Wanadoo España/Telefónica, Case COMP/38.784; on appeal: Case T-336/07, Telefónica SA and Telefónica de España, not yet decided. 15 Joined Cases C-241/91 and C-242/91 P, Radio Telefis Eireann (RTE) and Independent Television Publications Ltd (ITP) v Commission [1995] ECR 743. 16 Case C-418/01, IMS Health GmbH & Co OHG v NDC Health GmbH & Co KG [2004] ECR I-5039.

(W) Ehlermann 08 Paulis

522

6/11/08

11:53

Page 522

Session Three: Exploitative Conduct And The Remedies

determining whether prices are excessive. Prices may need to be considerably above production cost in order to finance innovation costs. However, one should not forget that there may be, for example, network effects in industries with considerable investment and innovation. I am not convinced that competition authorities should be prevented from intervening in such industries— even if they are very innovative. Where does this discussion leave me? I’m afraid that, for the moment, I can only see one reasonable criterion that we can use to identify markets that could be candidates for interventions against excessive prices, and that is that there should be very high and long-lasting barriers to entry or expansion. The other conditions that have been brought forward in the debate are interesting, and I have some sympathy for all of them. But I do not feel comfortable with restricting the area of possible intervention to only markets that fulfil these conditions on top of the requirement of entry or expansion barriers that I just mentioned.

4. How do we identify excessive prices? The discussion above focused on which markets could be candidates for possible intervention against excessive pricing by a dominant firm. However, even after identifying such a market, there remains the difficult task of determining whether the prices charged by the dominant firm in this market are in fact excessive. A useful starting point is the test indicated by the European Court of Justice in United Brands:17 “The questions therefore to be determined are whether the differences between the costs actually incurred and the price actually charged is excessive, and, if the answer to this question is in the affirmative, whether a price has been imposed which is either unfair in itself or when compared to competing products.”

A high profit margin may result both from high prices and from low costs. The second limb of the test addresses this issue. The test also implies that high prices, for instance compared to prices in other markets, by themselves are not abusive if they do not lead to an excessive difference between price and costs. High prices could, for instance, be explained by differences in cost conditions. Before going a bit more into detail with how to implement this test, it is useful to ask a fundamental question. How is a dominant firm actually supposed to behave if it wants to avoid being accused of charging excessive prices? Is it 17 Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207, para. 252.

(W) Ehlermann 08 Paulis

6/11/08

11:53

Page 523

Article 82 and Exploitative Conduct 523 supposed to behave “as if” it were in a competitive “equilibrium”? In other words, is it supposed to behave as if it had no market power? Or is it only supposed to refrain from exploiting its market power “completely”, so that prices “somewhat” over the competitive equilibrium prices are accepted but prices that are “excessively” so remain prohibited. The advantage of the “as if” comparison is that, at least in theory, there is a clear comparator. If prices are above the competitive equilibrium prices, they are excessive. This avoids the messy and somewhat arbitrary decision of when a high price is too high. However, I don’t think the “as if” argument works. Market power is found on a continuum. Firms with some market power—but not enough to be called dominant—face no risk of intervention from competition authorities if they charge prices above the competitive level. The more market power they have, the higher the prices they can charge. It would seem a bit odd if somebody who attained a lot of market power—and therefore had a dominant position—suddenly had to lower its prices drastically to the competitive level. It seems to me that the model we must have in mind is instead that once a firm arrives at a certain level of market power it is implicitly told that it now cannot exploit its market power any further. The use of market power is, so to speak, “capped” at a certain level. This model implies that only very large deviations from competitive conditions may be indicative of abusive pricing. Of course, this invites the natural question of whether I can say how large is “very large”. I’m afraid not, at least not if I am supposed to give a number, like “prices are excessive if they are X % over costs”. I don’t think I am willing at this stage to provide a “safe harbour” either. Cost and profit margin calculations are so complicated—and so far we have little experience in doing them—that I do not feel able to do so. Of course, you and others may be able to change my mind, but for the moment I prefer to remain at a more conceptual level. I know that this position leaves me open to criticism—or to joking comparisons with the remark of US Supreme Court Justice Potter Stewart that he could not define which pornographic materials rose to the level of obscene (and hence constitutionally unprotected) speech, but that he knew it when he saw it.18 However, this is how it has to be for the moment. I will not spend time mentioning the various price and cost comparisons that may be relevant. I think most of you are well aware of them, and I don’t think I have much new to add. However, I would like to come back briefly to the point concerning investments and innovation. When calculating the profit margin, it is important to give proper consideration to the investment risks involved in the industry concerned. Profit margins are typically calculated on the basis of the prices and costs of products that are sold on the market. And 18 See Jacobellis v. Ohio, 378 U.S. 184, 197 (1964) (Stewart, J., concurring) (recently recalled by Michael Salinger, “Moneyball and Price Gouging", address to Antitrust Committee of Boston Bar Association, Boston, 27 February 2006, at p. 6).

(W) Ehlermann 08 Paulis

524

6/11/08

11:53

Page 524

Session Three: Exploitative Conduct And The Remedies

the products of a dominant company will often be among the most successful of those products that are actually brought to market. However, in many industries there are substantial risks involved in developing such products. There may be several unsuccessful products for each product that is successfully brought to market. Intervening too easily against the pricing of the few successful companies in such industries could risk chilling welfare-enhancing investment efforts.

(X) Ehlermann 08 Roller

6/11/08

11:53

Page 525

II Lars-Hendrik Röller* Exploitative Abuses August 2007

1. Introduction This paper investigates the role of exploitative abuse under Article 82. Unlike exclusionary practices, there is relatively little economic commentary on the proper role of exploitative abuses in competition policy,1 even though the legal perspective in Europe has consistently emphasized the importance of enforcing antitrust action in the arena of exploitative practices.2 In this paper I will address the following two questions: (i) what are the effects of antitrust action against exploitative conduct?; and (ii) when should antitrust action be taken against exploitative abuses? Before addressing these questions I’d like to make some general observations about Article 82. Let me begin by stating that any anticompetitive behaviour involves both exclusionary and exploitative abuses. Anticompetitive conduct must ultimately lead to exploitation. In fact, the sole purpose of firms engaged in exclusionary practices is to increase their market power, which in turn will allow firms to increase their rents. The typical anticompetitive story is thus one in which a firm reduces competition through anticompetitive means (exclusion) in order to reap the benefits of higher market power (exploitation). In this sense, exclusionary and exploitative abuses are part of the same economic logic for antitrust, and it is no surprise that both types of behaviours are at the core of Article 82. Against the backdrop of this basic economic logic regarding how markets work, it is important to understand that while it is correct to say that “one excludes in order to exploit”, it is not correct to say that all exploitation is rooted in exclusionary abuse. In fact, the essence of pro-competitive behaviour is to increase market power, which will in turn increase rents by means * ESMT and Humboldt University. 1 See, e.g., David S. Evans and A. Jorge Padilla, “Excessive Prices: Using Economics to Define Administrable Legal Rules”, 1(1) Journal of Competition Law and Economics 97 (2005); Massimo Motta and Alexandre de Streel, “Excessive Pricing and Price Squeeze under EU Law”, in ClausDieter Ehlermann and Isabela Atanasiu, eds., Competition Law Annual 2003: What is an abuse of a dominant position?, Hart Publishing, 2006, pp. 91 et seq. 2 For a historical account of the balance between exclusionary and exploitative abuses under Article 82, see the paper by Emil Paulis, “Article 82 and exploitative conduct”, this Volume.

(X) Ehlermann 08 Roller

526

6/11/08

11:53

Page 526

Session Three: Exploitative Conduct And The Remedies

of exploiting this very market power. In fact, if there were no possibility to ever exploit one’s market power, there would be no incentive to compete.3 Thus, pro-competitive behaviour must also involve exploitation (“positive effects”). Another possibility is that some government has—or has in the past— granted an exclusive right, as in the case of network industries (post, telecommunications, rail, etc.). As a result, firms may hold strong market positions and may price accordingly, if left unregulated. In this situation, exploitative abuses would not be due to exclusionary conduct. The above considerations suggest that both pro- and anticompetitive behaviour involve exploitation. Since the goal of antitrust is to discriminate between pro- and anticompetitive conduct, it follows that the mere existence of exploitative behaviour is insufficient to identify anticompetitive practices. By contrast, the existence of exclusionary practices does provide a sound basis for identifying anticompetitive behaviour. As a result, one might conclude that it is indeed solely exclusionary conduct that antitrust should focus on. In this spirit, I now turn to the first question, in the context of which I consider what the economic basis for antitrust action under exploitative conduct might be (Section 2). Then I return to the above point on the proper role of exploitative abuse in Europe (Section 3).

2. What are the effects of antitrust action against exploitative conduct? Perhaps the most widely held view in favour of antitrust action on the grounds of excessive pricing is that it leads to lower prices and thereby increases consumer surplus.4 It is argued that such price reductions benefit consumers directly and in the short-run. Given that consumers are at the heart of competition policy, interventions in cases of excessive pricing are therefore on this view an integral element of antitrust enforcement. Despite a certain appeal of the above argumentation, there are very credible arguments as to why excessive pricing interventions in fact will not benefit consumers. Whether prices are higher or lower under antitrust enforcement is fundamentally associated with the analysis of the relevant counterfactual. In other words, one has to address what the prices would have been (both in the short and long run, i.e., is the price reduction temporary or 3 This includes situations where firms are under the threat of bankruptcy. In this case, engaging in pro-competitive behaviour allows firms to increase their rents, even though they are close to zero (as opposed to negative). 4 This section will couch the discussion on exploitative abuses in terms of excessive pricing, which constitutes the vast majority of all antitrust cases dealing with such abuses.

(X) Ehlermann 08 Roller

6/11/08

11:53

Page 527

Exploitative Abuses

527

not?) in the absence of antitrust actions (or the threat of it). In other words, one has to take the likelihood of entry into account and how antitrust intervention affects this. Whether entry takes place or not depends on the existence and magnitude of entry barriers. There are a number of different types of entry barriers, such as strategic entry barriers (e.g., excess capacity and first mover advantages), legal entry barriers (e.g., access regulation, IPRs), absolute cost advantages, asymmetric information, etc. If such entry barriers are not significant, high post-entry margins attract entry. In other words, the market does self-correct. Since antitrust enforcement against excessive pricing lowers post-entry margins, it will tend to discourage entry and negatively affect the market’s propensity to self correct, which may ironically lead to higher prices. For this reason, it makes sense to restrict antitrust interventions in the arena of excessive pricing to situations where entry barriers are so high that entry (absent antitrust action) is unlikely.5 A further fundamental objection against the use of excessive pricing interventions is that they negatively affect investments and/or innovation. By lowering prices ex post, incentives to invest or to innovate ex ante are reduced or even eliminated. As was mentioned above, the essence of pro-competitive behaviour is to increase market power, which will in turn increase the incentive to invest or innovate. In industries where innovation or large up-front investments play a major role, consumers ultimately do not benefit from excessive pricing interventions. A final reason why intervention in excessive pricing cases may not benefit consumers is that such cases entail a very complex assessment. It is difficult to say what constitutes an “excessive” price without referring to practically all aspects of a market and its competitive environment. Invariably, price-cost tests raise a number of difficult methodological issues, such as how to allocate costs, how to define efficient costs, how to find the proper benchmark, and how to determine the nature of competition (including demand). An example of these difficulties of empirically assessing price-cost margins can be illustrated by a recent analysis of past data from the European airline industry.6 In a structural approach, all the above aspects of competition are 5 It is also important to take into account the extent to which prices in related markets are affected by the intervention. For example, some commentators have claimed, in the recent debate on regulation of charges for international roaming, that the impact of regulating such charges will lead to higher prices in domestic markets. I am not aware of much analytical work in this area of antitrust economics, let alone systematic empirical evidence. Absent formal analysis, a reasonable conjecture is to assume that firms will raise prices in related markets whenever antitrust action lowers prices in complementary markets, while the opposite happens when products are substitutes. 6 Damien J. Neven, Lars-Hendrk Röller, and Zhentang Zhang, “Endogenous Costs and Price-Cost Margins: an Application to the European Airline Industry”, 54 Journal of Industrial Economics 351 (2006). The data analyzed in this paper cover the period from 1976 to 1994. This period is particularly well suited to analyze the impact of unions. Notice also that low-cost airlines did not play a major role at that time.

(X) Ehlermann 08 Roller

528

6/11/08

11:53

Page 528

Session Three: Exploitative Conduct And The Remedies

taken into account (demand, cost, and the nature of competition in the industry) in order to properly estimate price-cost margins. In addition, the authors allow for “endogenous costs”, i.e., costs linked to product market competition and associated with labour unions in the following way. If price-cost margins are high, unions bargain for higher wages; this, in turn, raises costs, thereby lowering those price-cost margins. If, on the other hand, price-cost margins are low, union demands are moderate, which will exert upward pressure in price-cost margins. In equilibrium, all these factors are taken into account and balanced. The main result of the study is as follows. Observed prices in the European airline industry are virtually identical to monopoly prices, even though observed margins are consistent with competition. The reason is that costs had been inflated to the point that European consumers were faced with de facto monopoly prices. In other words, the empirical evidence suggests that prices have been “excessive”, while price-cost margins are “normal”. In sum, antitrust action under the banner of exploitative abuse needs to be applied with great caution. Not only is it difficult to identify price-cost margins properly, but it is also far from obvious what an “excessive price” would be and how antitrust action would then benefit consumers.

3. When should antitrust action be taken against exploitative abuses? Let me now return to the statement made in the introduction, namely that, since pro- and anticompetitive behaviour both involve exploitation, it is indeed solely exclusionary conduct that antitrust should focus on. If anything, the discussion summarized in the previous section would certainly support such a policy. However, I do believe that antitrust action against exploitative abuse does have a role to play, although only under certain circumstances.

(i) “Gap cases” In principle, an abuse case must identify anticompetitive conduct that results in increased market power, relative to the counterfactual. As was mentioned above, it is the road to dominance that is important in order to identify profrom anticompetitive conduct. If dominance (or for that matter, any kind of market power) is obtained through competition on the merits, then this is good for consumers; otherwise, it is not. By contrast, Article 82 only applies to firms that are already dominant. In other words, anticompetitive conduct

(X) Ehlermann 08 Roller

6/11/08

11:53

Page 529

Exploitative Abuses

529

that leads to a dominant position cannot be caught in Europe under Article 82 as an exclusionary abuse. This is an enforcement “gap”, since it is precisely the way in which dominance is acquired that matters in terms of economic effects.7 I would suggest that antitrust enforcement against exploitative abuses can be used to close this important gap. That is, exploitative abuse cases should be based on acquiring a dominant position through anticompetitive exclusionary conduct. In this way, exploitative abuse cases lead us back to investigating exclusionary conduct, which is in fact the proper way to identify anticompetitive practices. By focusing on the road to dominance via anticompetitive behaviour, exploitative abuse cases are firmly grounded on the way markets work; they do not make it necessary to decide what is “excessive” from an ex post point of view. Note that this approach is very much in line with the observation that there are many exploitative abuse cases in sectors with formerly state-owned monopolies. Perhaps this observation has something to do with the fact that these firms did not obtain their dominant positions based on merit alone, but rather by virtue of a public policy decision (which itself was likely based on a theory of natural monopoly). To the extent that the road to dominance matters, these scenarios should in fact be considered gap cases, even though here dominance was not achieved by anticompetitive exclusionary conduct on the part of the undertaking. Overall, there appear to be three main advantages in defining exploitative abuse as acquiring dominance as a result of an exclusionary abuse. First, it is in line with sound economics. Second, it avoids the standard debate on what is “excessive” (which I believe is impossible to define operationally). And third, it closes the gap described above under Article 82.

(ii) “Mistakes” A second reason for antitrust action against exploitative abuses may be “mistakes”. In other words, for some reason an antitrust authority may not have effectively prosecuted an exclusionary abuse. In this situation, one may argue that, since the exclusionary abuse was not caught, or perhaps because there was uncertainty as to the effect of the exclusionary conduct (a type II error), there is need for a second “shot” at enforcement, this time on the basis of an exploitative abuse, (a “two shot” policy). A related argument is that the mistakes are due to uncertainty. Given that exclusionary abuses are based on “likely effects”, there is of course also 7 There may be other gaps under Article 82, such as anticompetitive conduct below the level of dominance, or anticompetitive behaviour by an oligopoly.

(X) Ehlermann 08 Roller

530

6/11/08

11:53

Page 530

Session Three: Exploitative Conduct And The Remedies

a likelihood that the agency will get it wrong. As a result, one may argue that two instruments are better than one. In other words, exploitative abuses can be stopped whenever the exclusionary abuse has had an actual effect, even where it was previously deemed unlikely. On the other hand, one may legitimately wonder whether the better policy response is not to eliminate the mistakes (or reduce the uncertainty), rather than to maintain a two shot policy. It may be, for example, that once an exclusionary abuse has taken place the damage has been done and remedies imposed at a later stage may come too late. Moreover, a two shot policy may give rise to over-enforcement, reducing type II errors but leading to more type I errors.8

(iii) When to take action Summarizing the above argumentation, one may conclude that exploitative action under Article 82 is limited to very special circumstances. Building on the above discussion, the following circumstances for antitrust action in cases of excessive pricing can be identified.9 (a) (b) (c) (d) (e)

significant entry barriers market is unlikely to self-correct no (structural) remedy is available no regulator or regulatory failure “gap cases” or “mistake cases”

[=> otherwise advocacy] [=> otherwise ex ante regulation] [=> otherwise exclusionary abuse]

The above conditions are cumulative, i.e., they are necessary conditions and must all be met. The first two conditions (a) and (b) have already been discussed and are self-evident. If there are no entry barriers, or if the market would self-correct, there is no compelling need to intervene. The market is best left to itself and consumers will benefit. Conditions (c) and (d) have not been addressed above, but have been mentioned by other authors. They are based on the fact that remedies in exploitative abuse cases (especially excessive pricing) essentially amount to price regulation and that there are better placed instruments and institutions to deal with this. Specifically, (c) specifies that if there are structural remedies available—such as removing relevant entry barriers, opening markets, liberalizing, etc.—then the proper policy would be advocacy in favour of these

8 Note that mistakes may go either way, i.e., there are both type I and II errors under exclusionary abuse cases. A two shot policy based on uncertainty about the likely effects would control the actual effects ex post. In principle, there will be cases where both type I and type II errors may be found. Nevertheless, one would presume that a two shot policy would only be applied when type II errors have occurred (i.e., where an exclusionary abuse was not caught). 9 Again, I concentrate on excessive pricing, as this is the most prominent area.

(X) Ehlermann 08 Roller

6/11/08

11:53

Page 531

Exploitative Abuses

531

structural remedies, as opposed to ex post intervention in exploitative abuse cases. To the extent that exploitative abuse cases can be used to achieve structural remedies, or to the extent that exploitative abuse cases are complementary to advocacy action, exploitative abuse cases under Article 82 would be helpful—but only because they support a structural remedy. Similarly, condition (d) specifies that regulatory-type antitrust action (as in excessive pricing cases) is only warranted if there is no regulatory agency, or if the regulator does not operate effectively. Where there is a regulator, intervention by an antitrust authority is thus limited, as specialized regulatory institutions are likely to have superior regulatory know-how. However, antitrust intervention remains possible, not least because industry-specific regulators are more likely to be subject to regulatory capture (and hence may fail to protect consumer welfare) than an antitrust authority, whose competence extends horizontally across most or all economic sectors. Finally, condition (e) refers to “gap cases”. As argued above, this category refers to cases where dominance is a result of exclusionary conduct, and not the other way around. If this condition is not met, then any relevant anticompetitive conduct can be addressed by taking action to remedy the exclusionary abuse. In other words, if the defendant did not achieve dominance by means of exclusionary behaviour, then antitrust intervention is unlikely to be able to discriminate between pro- and anticompetitive conduct. It will run into methodological problems, and it will be difficult to implement in line with sound economics. Recall that there is also the possibility of achieving dominance through government action, and these are also gap cases. Condition (e) also mentions “mistake cases”, primarily for completeness. However, the above discussion has expressed some doubt as to the appropriateness of arguing that a second shot is needed.

4. Concluding remarks I have argued that excessive pricing cases should be limited to certain special circumstances. Focus should be on “gap cases”, where anticompetitive exploitation exists if and only if exclusionary abuse—or government actions—have led to a dominant position. Specifically, it is the road to dominance that becomes important in those cases. Overall, there appear to be three main advantages in defining exploitative abuse, in particular excessive pricing, as acquiring dominance through exclusionary abuse or governmental action. First, it is in line with sound economics (i.e., it is more capable of discriminating between pro- and anticompetitive conduct). Second, it avoids the standard debate on what is “excessive” (which

(X) Ehlermann 08 Roller

532

6/11/08

11:53

Page 532

Session Three: Exploitative Conduct And The Remedies

in my view, as I mentioned earlier, is impossible to define operationally10). Third, it closes a gap under Article 82. Finally, let me submit that the above approach, which directs exploitative abuse cases towards analyzing exclusionary behaviour, is in line with an effects-based approach under Article 82.

10 How do we identify excessive pricing? What is the standard? Recall that any standard based on costs is methodologically doubtful. Note that the approach proposed in this paper does provide a benchmark, which is rather natural and consistent with the rest of Article 82. The benchmark for excessive prices is simply the price that would prevail “but for the exclusionary conduct”.

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 533

III Amelia Fletcher and Alina Jardine* Towards an Appropriate Policy for Excessive Pricing June 2007

Introduction In the US, competition law does not cover excessive pricing. In the EU, the Commission has typically exercised its prosecutorial discretion not to take excessive pricing cases, and has often stated that it is not its role to become a price regulator. Nevertheless, there have been strong hints that the contemplated EC draft guidelines on Article 82 of the EC Treaty will provide guidance on exploitative abuse—which most typically equates to “excessive pricing”1—alongside the expected guidance on exclusionary abuse. This clearly raises the issue of what the appropriate policy for excessive pricing should be under Article 82. This brief paper examines this question. The paper sets out the key arguments for and against a relatively laissezfaire approach towards high pricing. A number of conclusions are drawn from this discussion as to an appropriate policy approach to excessive pricing. The paper then argues that, while it may sometimes be appropriate to intervene against excessive pricing, the authorities should wherever possible endeavour to address the causes of the abuse—that is, the market circumstances that allow the excessive pricing to occur—rather than using price regulation to address the symptoms. Such interventions arguably have more in common with consumer policy than with traditional competition policy, but the objective is the same: to improve the functioning of competition, for the benefit of consumers. It should be noted that this paper focuses on exploitative excessive pricing, which is not in itself harmful to the structure of competition in the relevant

* Amelia Fletcher is Chief Economist and Alina Jardine is an Economic Advisor at the UK’s Office of Fair Trading (OFT). The views expressed here are their own and do not necessarily represent those of the OFT. 1 Article 82(a) gives as a specific illustration of abuse: “directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions”. While there has been at least one case based on the unfair trading conditions part of this clause (Commission Decision 2000/12 of 20 July 1999, 1998 Football World Cup, [2000] OJ L5/55), cases have more usually revolved around “unfairly high” or “excessive” pricing.

(Y) Ehlermann 08 Fletcher

534

6/11/08

11:54

Page 534

Session Three: Exploitative Conduct And The Remedies

market. It does not cover situations in which excessive pricing may be exclusionary, such as margin squeeze or constructive refusal to supply. However, it should be noted that in practice many excessive pricing cases are essentially about exclusion rather than exploitation, and moreover the line between the two can sometimes be unclear.2

Arguments for a laissez-faire approach to high prices In many jurisdictions, exploitative pricing is either not proscribed by competition law or rarely prosecuted. This reflects the fact that there are a number of arguments against intervening against high prices through competition law.

High prices may be an important market signal First, if a dominant firm is earning excessive profits in a given market, this will typically act as an important market signal to attract new entrants into the market. In the absence of substantial barriers to entry, therefore, any intervention that reduces the profits of an incumbent might not only be unnecessary but could actually prolong the monopoly situation by blocking efficient market signals to potential entrants. For this reason, it would be a sensible policy approach not to intervene against high prices if one expects them to stimulate successful new entry within a reasonable period.

High profits in one market may be given away elsewhere Secondly, for multi-product firms, or firms in multi-sided markets, it will often be the case that a significant proportion of the “excessive profits” that are made in one market are, in practice, given away in another market. Consider, for example, an electric toothbrush manufacturer, whose toothbrushes are designed so that they can only be used with the manufacturer’s proprietary brush heads. The manufacturer is effectively a monopolist in the supply of its own brush heads, and accordingly will typically price them fairly high. However, so long as the manufacturer faces effective competition in the 2 For example, the recent Albion Water case in the UK, which was argued in terms of excessive pricing, was essentially about the exclusion of a potential intermediate supplier (Albion Water), rather than a straight exploitation case. See Competition Appeals Tribunal, judgment of 18 December 2006, Albion Water Limited v Water Services Regulation Authority, Case No. 1046/2/4/04, [2006] CAT 36.

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 535

Towards an Appropriate Policy for Excessive Pricing 535 primary electric toothbrush market, we should expect it to compete much of this “excess profit” away. This will be true even if consumers do not engage in any form of “life cycle” assessment of product pricing. This does not necessarily mean that it will always be inappropriate to intervene against high prices in markets where a substantial proportion of the profits from those high prices are effectively passed back to consumers. In some cases, such a pricing structure may nevertheless be inefficient, for example if the consumers who receive the passed-back profits are not the consumers that pay the high prices, or if the pricing structure substantially distorts consumers’ choices (which can in turn affect the nature and degree of competition). However, we do mean to suggest that a sensible policy approach would be to refrain from intervening against high prices for one element of a firm’s product portfolio unless careful consideration is given to: the extent of competition the firm faces with respect to the other elements of its portfolio; the extent to which any “excess profits” are effectively returned to the consumer; and the extent to which the pricing structure substantially distorts consumers’ choices.

Assessment of excessive pricing is difficult Thirdly, excessive pricing can be extremely difficult to assess. This is true ex post when examining a particular case, but the issues are still more extreme when trying to set clear rules that allow ex ante competition law compliance. The key problem here is that it is not clear what the appropriate benchmark should be. One obvious option is the “competitive price”. But how does one define the competitive price in a market that is not competitive? Should dominant firms really be required to price at levels which would obtain under vigorous (Bertrand) price competition, when such prices would not be observed in non-cooperative oligopolies not subject to Article 82? In United Brands,3 the Court of Justice used the alternative concept of “economic value”: “charging a price which is excessive because it has no reasonable relation to the economic value of the product supplied is [. . .] an abuse.”

But there are serious problems with this test too. In economic terms, if a person is willing to pay a given price for a product, then the economic value of that product to that person must be at least as high as that price. It is worth noting that the UK Court of Appeal recently overturned a High Court finding of excessive pricing, on the basis that the “economic value” of a product

3 Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207.

(Y) Ehlermann 08 Fletcher

536

6/11/08

11:54

Page 536

Session Three: Exploitative Conduct And The Remedies

should take account of its value to the buyer.4 Since this latter concept equates economically to what a buyer would be willing to pay for the product, this line of reasoning taken to its logical conclusion would seem to suggest that—based on the “economic value” test—excessive pricing cases cannot be brought if the buyer ever actually buys the product. One can, of course, look at historic margins, margins in other regions or countries for comparison or margins earned in similar industries. “But”, say the advocates of a laissez-faire approach to excessive pricing, “none of these is perfect and sometimes none is possible.” Moreover, even if one does observe differences in margins, should this necessarily imply abuse, or should some degree of differential margin be acceptable?

Price regulation is highly distortive Fourthly, and perhaps most importantly, it is argued that fines for excessive pricing effectively equate to price regulation. Indeed, sometimes such regulation is explicit. The concern here is that price regulation can distort competition, investment and R&D, to the detriment of consumer welfare. In particular, ongoing price regulation of a dominant firm can: (i)

Inhibit entry/expansion by competitors: Potential competitors are less likely to enter or expand in the market if they find it hard to compete against the low prices of the regulated firm. This is especially likely to be a problem where a dominant firm has important incumbency advantages, since new entrants would typically seek to win market share in such circumstances by undercutting the incumbent’s high prices. (ii) Distort incentives for efficiency: Suppose a firm has higher profits than its competitors primarily because it is highly efficient, for example if it has adopted state-of-the-art production processes. If such a firm runs the risk of being penalized for having excessive profits, and having its productivity gains expropriated by being forced to lower its prices, this could potentially discourage firms from improving efficiency (and productivity) in the first place. (iii) Distort investment incentives, including incentives to innovate: More generally, a dominant firm’s market position will often derive partly from its past investments. And in some cases, the firm will require the high prices associated with its dominant position to recoup these past investments. From an ex ante perspective, future recoupment also serves as the incentive for those investments. This incentives issue is likely to be especially 4 England and Wales Court of Appeal (Civil Division), judgment of 2 February 2007, Attheraces (UK) Limited v. The British Horseracing Board Limited, Case No. A3/2006/0126, [2007] EWCA Civ 38.

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 537

Towards an Appropriate Policy for Excessive Pricing 537 important for R&D investments, which can be highly risky, costly, and slow to generate revenues. Intellectual property rights are specifically designed to provide innovating firms with a degree of market power, and so to stimulate upfront R&D investment through the “prize” of higher than normal future profits. Any reduction in future profits—or a greater risk of these profits being regulated—could clearly jeopardize such incentives. (iv) Distort pricing incentives: Depending on the specific design of the price regulation, there is a risk that regulation can facilitate anticompetitive pricing. For example, suppose a dominant firm is regulated on its average price over two markets, and it faces potential competition in one of those markets. It will have an incentive to cut price in the contestable market to make entry unprofitable, and this predatory behaviour will be costless because it can be funded by raising price in the non-contestable market and preserving the average price at the regulated level. In terms of these distortions, it seems likely that the “deterrent” effect of excessive pricing rules—whereby dominant firms in the economy endeavour to keep prices below their “best guess” as to what constitutes excessive pricing—has the potential to be substantially more problematic than the ex post regulation of those dominant firms whose pricing has explicitly been found to be exploitative. In the absence of excessive pricing rules, firms set prices to maximize profits. If they are concerned that their profit-maximizing prices might be seen as exploitative under competition law, though, this could lead them to alter their pricing behaviour in all sorts of unpredictable and distortive ways. By contrast, where competition authorities engage in ex post regulation of infringing firms, any distortions can be taken into account on a case-by-case basis and can, to some extent, be avoided by careful intervention design. The distortions associated with the “deterrent” effect of excessive pricing rules provide a good policy argument for minimizing this deterrent effect, in particular by steering clear of imposing fines for excessive pricing and of allowing private damages actions in respect of such behaviour, since each of these strengthens firms’ incentives to abide by competition law. By limiting available sanctions to the imposition of ex post penalties, such as future price regulation, firms are likely to be less concerned about breaching excessive pricing rules, and as such the associated distortions across the economy should be greatly reduced.5 5 Given the very limited numbers of excessive pricing cases we observe (even in Europe), one might anyway expect that there is not a substantial ex ante deterrent effect on pricing in practice. However a recent (and as yet unpublished) study by the OFT on the deterrence effect of the UK competition regime found a not insignificant amount of deterrence of pricing behaviour due to concerns it might be found to unlawfully exploit dominance (although it is also notable that the deterrence levels for excessive pricing were substantially lower than for any other form of abuse).

(Y) Ehlermann 08 Fletcher

538

6/11/08

11:54

Page 538

Session Three: Exploitative Conduct And The Remedies

Another concern highlighted above was the risk that price regulation might inhibit entry or expansion by competitors, and so prolong the dominant firm’s market position. This is potentially a serious issue. However, it is worth noting that it would be less likely to arise in practice if the policy approach were adopted of only intervening in markets where one does not expect the high prices to stimulate successful new entry within a reasonable period. Under this policy, price regulation should not occur where competitors are realistically willing and able to enter or expand through undercutting the dominant firm’s prices, and so become a real constraint on the dominant firm.

Summary of policy implications In summary, one might reasonably conclude from the above arguments that a sensible policy approach towards excessive pricing would have the following characteristics: • There should be no intervention against high prices if one expects them to stimulate successful new entry within a reasonable period. • In examining high prices for one element of a firm’s product portfolio, it is important also to consider carefully the pricing of other elements of its portfolio, the competition the firm faces in those other markets, and the impact on consumers’ choices. • In order to reduce deterrence, firms should not face fines for excessive pricing, and should not face the risk of private damages actions in respect of such behaviour. None of the above are currently explicitly (or even implicitly) incorporated within EC competition policy. Their adoption would therefore go a long way towards meeting the concerns set out above. Of the three, the third would probably be the most controversial. However, vigorous advocates for a laissez-faire approach to excessive pricing would argue that the concerns set out above are so serious that it is wrong for competition law to proscribe high pricing at all. They would typically accept that there may need to be explicit regulation for certain areas of natural monopoly—such as the utilities—but they would argue that such regulation should be done carefully by specialist regulators. The rest of the economy should be left alone, not least because the risks of careless and ill-informed intervention outweigh any potential benefits.

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 539

Towards an Appropriate Policy for Excessive Pricing 539

Arguments for a degree of intervention against excessive pricing Against the laissez-faire view, a number of arguments can be made as to why competition law should cover excessive pricing.

Good fit with the objectives of competition policy Firstly, there is the simple philosophical point that there is a clear potential for consumer harm from excessive pricing. Indeed, it is arguable the primary rationale for competition policy is—in the end—to limit the potential for such exploitative behaviour, and in so doing to benefit consumers. As such, there is a good fit between a law against excessive pricing and the overarching objectives of competition policy. This point is given further weight by a recent paper by Akman (2007)6 which reviews the travaux préparatoires (preparatory documents) associated with the initial drafting of Article 82 (then Article 86) in order to find out the “legislative intent” of the provision. Akman concludes that: “. . . the provision was intended to apply to only ‘exploitative’ abuses and not ‘exclusionary’ abuses. Their main worry being ‘increasing the size of the pie’. . . .”7

Price regulation of this sort occurs elsewhere in competition and consumer law Secondly, it is pointed out that certain forms of high pricing are already proscribed under other parts of competition and consumer law. For example: (i)

The law on exclusionary abuse of dominance can require dominant suppliers to ensure that their pricing is fair and reasonable: High prices charged by an upstream supplier to a downstream firm can constitute a constructive refusal to supply or margin squeeze if they restrict or distort the ability of the latter to compete on the relevant downstream market.

6 Pinar Akman, “Searching for the Long-Lost Soul of Article 82EC”, CCP Working Paper 6-5 (March 2007), available at http://www.ccp.uea.ac.uk/publicfiles/workingpapers/CCP07-5.pdf. 7 Ibid. Akman also provides evidence that there was dispute during the early years of Article 82’s implementation as to whether it applied to only exploitative abuses or included exclusionary abuses as well. For example, Joliet (later a judge of the ECJ) was of the view that it merely covered exploitative abuses, the test of legality being not the interference with other firms’ freedom to compete and the use of “exclusionary” practices to achieve and hold power, but rather whether there is monopolistic exploitation of the market. See René Joliet, Monopolization and Abuse of Dominant Position, Martinus Nijhoff, 1970. Joliet apparently reached this conclusion by studying the examples listed in Article 82.

(Y) Ehlermann 08 Fletcher

540

6/11/08

11:54

Page 540

Session Three: Exploitative Conduct And The Remedies

(ii) Fair and reasonable pricing may in some cases be necessary to justify the application of Article 81(3): This may be the case in particular with respect to the exemption of horizontal agreements. There may be two possible rationales for such requirements: (a) To ensure that the horizontal agreement does not eliminate competition. An example concerns standard-setting organizations, in respect of which the Commission’s Guidelines on Horizontal Cooperation Agreements specify that: “To avoid elimination of competition in the relevant market(s), access to the standard must be possible for third parties on fair, reasonable and non-discriminatory terms.”8 (b) To ensure that the restriction of competition is no greater than is required to achieve the benefits. An example here is the UK LINK case (2000),9 in which the OFT concluded that an agreement between UK banks, relating to the cross-charge levied for the use of each others’ ATM machines, would qualify for an exemption so long as the charge was set no higher than required for cost recovery. (iii) The regulation of unfair contract terms requires that hidden charges be “fair”: Where consumers face additional charges that are not covered by the “core terms” of a standardized contract with a supplier, these may contravene consumer law if they are not “fair”. “Fair” in this situation has regularly been interpreted as “cost-reflective”. A recent example in the UK was the OFT’s threatened action against the default fees charged by credit card issuers for late payment of invoices, which were considered to be substantially above cost.10 In addition, in the UK, the Enterprise Act 2002 empowers the authorities to carry out market investigations where there are “features” of a market that prevent, restrict or distort competition. Such investigations regularly examine markets in which prices are found to be excessive, and one potential outcome is to impose price regulation remedies. However, in practice the UK Competition Commission11 has typically preferred remedies designed to make the markets involved work more effectively, which should in turn lead to lower prices, as opposed to regulating prices directly. This alternative approach to addressing excessive pricing is discussed further below.

8

[2001] OJ C3/2, para. 174. OFT Decision of 11 May 2000, LINK Interchange Network Ltd, Case No. CP/0642/00. 10 See http://www.oft.gov.uk/news/press/2006/credit-cards. In the event, the credit card companies reduced their charges in order to avoid action, so the case did not go to court. 11 Market investigations are carried out by the UK Competition Commission, following a reference from the OFT or another designated body. 9

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 541

Towards an Appropriate Policy for Excessive Pricing 541

Assessment difficulties are overstated Thirdly, while it is true that assessment of excessive pricing can sometimes be difficult, it would be wrong to overstate these difficulties. On the one hand, they do mean that it can be difficult to set out guidance that draws a clear line between excessive pricing and lawful pricing. On the other, there will nevertheless be cases where the excessive pricing is sufficiently extreme that it is relatively easy to demonstrate, based on a variety of different measures. An example is the OFT’s Napp case (2001),12 which related to the excessive pricing of a sustained release morphine product. This case was arguably not a typical excessive pricing case since there were two elements to the case: exclusionary low pricing to the hospital sector and excessive pricing to the “community sector” (i.e., the pharmacy sector). While the OFT chose to run the case as two separate abuses, this particular instance of excessive pricing could instead have been framed as ongoing recoupment from Napp’s predatory strategy, rather than as an abuse in its own right. Nevertheless, the case demonstrates well how one can sometimes observe clear water between “excessive” pricing and “competitive” pricing. In assessing Napp’s pricing, the OFT benchmarked Napp in several different ways: • Price-cost margins were compared both across Napp’s activities and with those of Napp’s competitors. • Prices were compared not only across activities and with those of competitors, but also across countries and over time. Napp’s prices and margins were found to be high—and by some margin— relative to all of these different comparators. On this basis, and without facing any serious difficulties in assessment, the OFT found that Napp had abused its dominant position.

Potential distortions from price regulation are overstated Fourthly, the policy approach set out at paragraph 23 above would go a long way towards addressing the potential distortions highlighted as arising from price regulation. In particular, the deterrence effect would be greatly reduced, as would the risk of price regulation inhibiting a realistic prospect of entry or expansion by competitors. Of continuing concern, however, are the potential distortive effects of ex post price regulation on upfront investment. Especially important here is investment in R&D, since R&D can play a crucial role in improving 12 OFT Decision of 30 March 2001, Napp Pharmaceutical Holdings Ltd and Subsidiaries (Napp), Case No. CA98/2/2001, upheld on appeal: Competition Appeals Tribunal, judgment of 15 January 2002, Napp Pharmaceuticals Holdings Limited and Subsidiaries and Director General of Fair Trading, Case No. 1001/1/1/01.

(Y) Ehlermann 08 Fletcher

542

6/11/08

11:54

Page 542

Session Three: Exploitative Conduct And The Remedies

consumer welfare over the long term, and intellectual property rights are specifically designed to generate a degree of apparently “excessive” profits in order to incentivize firms to engage in costly and risky R&D in the first place. To some extent, these various concerns can be ameliorated if competition authorities bear in mind the risks of intervention. They should intervene only after careful consideration of the potential distortive impact of any proposed regulation, and only in a manner that ensures appropriate returns to sunk investments. However, in the case of R&D, the importance of ensuring incentives for dynamic competition may mitigate in favour of a stronger policy approach: specifically, there should be no intervention against excessive prices for an innovative product within that product’s patent life. It is noteworthy that in the Napp case, the OFT argued that the patent period provides an opportunity for recoupment of ex ante investment. While not made explicit, the tenor of this discussion suggests that the OFT would not have brought the case had the drug still been within patent.13

Price regulation is not the only remedy! Finally (and one of the key messages of this paper), it is important to recognize that price regulation is not the only possible remedy to a finding of excessive pricing. In the following section, we discuss an alternative approach to addressing exploitative abuse, which aims to address the causes of high prices, rather than the symptoms, and as such is less likely to result in market distortions. The approach is grounded in the “demand side” of the market, and thus arguably has more in common with consumer policy than with traditional competition policy, or even sectoral price regulation. As such, for some competition authorities it may tread relatively new ground and involve new thinking. However, the benefits are potentially large, not least in terms of limiting the extent of ongoing price regulation.

An alternative approach to remedying exploitative abuse As a general rule, exploitative abuse over a prolonged period occurs only where there are barriers to entry or expansion, preventing competitors from undercutting the dominant firm and eroding its market position. 13 Note that this argument assumes that the patent regime is well designed in terms of achieving the right balance between positive effects for dynamic efficiency and negative effects for competition, or at least that it is not for competition policy to try and correct any design failures of the patent regime. This assumption, however, is clearly controversial.

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 543

Towards an Appropriate Policy for Excessive Pricing 543 Sometimes these barriers relate purely to the supply side of the market. For example, potential competitors may lack access to crucial IP or they may face insurmountable regulatory barriers to entry.14 Often, however, the most serious barriers to entry and expansion will be to some extent related to the demand side of the market: that is, related to the characteristics and behaviour of buyers. These include: (i)

High switching costs: whereby buyers find it difficult or costly, or generally lack the incentive, to switch between suppliers. (ii) Lack of shopping around by customers: whereby buyers find it difficult or costly, or generally lack the incentive, to shop around between suppliers. (iii) Lack of comparable information across suppliers: such that buyers are unable to make rational decisions across suppliers. (iv) Asymmetric information between firms and customers: whereby buyers cannot observe all product characteristics (most usually, product quality) and therefore are overly reliant on supplier reputation or past experience with a given supplier. In such situations, the authorities may be able to put in place remedies which alleviate the relevant demand side problems and in doing so activate buyers to drive up the degree of competition in the market. This should in turn lead to lower prices and eliminate any need for price regulation. A nonexclusive list of possible interventions is provided in Box 1 below. As mentioned above, in the UK, the Enterprise Act 2002 empowers the Competition Commission to carry out market investigations. These are specifically designed to take a holistic approach to looking at why markets are not working well, examining causes of problems as well as symptoms, and reviewing demand side issues of the type above alongside supply side issues. The Competition Commission also has a wide range of sanctions available, the main constraint being that its interventions must be proportionate to the problems it identifies in the market. However, competition authorities in other Member States do not have such wide-ranging powers. The ability to bring excessive pricing cases under Article 82 would therefore seem to provide a useful and important avenue for other authorities to bring the sort of interventions listed in Box 1.15 Such interventions—where effective—have the potential to generate far greater benefits for consumers than price regulation, since enhanced competition will typically be more effective at driving up quality, service and innovation and driving down costs. At the same time, they do not have the downsides of price regulation described above. 14 It should be noted that where competition problems arise due to regulatory barriers to entry, competition authorities have an important competition advocacy role to play in trying to reduce such barriers. 15 An alternative would be for other jurisdictions to adopt a market investigation law similar to that in the UK, but this would require far-reaching legislative change.

(Y) Ehlermann 08 Fletcher

544

6/11/08

11:54

Page 544

Session Three: Exploitative Conduct And The Remedies

Conclusions Having set out the arguments for and against a laissez-faire approach towards excessive pricing, this paper proposes that a sensible policy approach would have the following characteristics: (i)

(ii)

(iii)

(iv) (v) (vi)

There should be no intervention against high prices if one expects such prices to stimulate successful new entry within a reasonable period. In examining high prices for one element of a firm’s product portfolio, it is important also to consider carefully the pricing of other elements of its portfolio, the competition the firm faces in those other markets, and the impact on consumers’ choices. In order to reduce deterrence, firms should not face fines for excessive pricing, and should not face the risk of private damages actions in respect of such behaviour. There should be no intervention under Article 82 against the high prices of an innovative product within its patent period. More generally, competition authorities should consider carefully the effect of any ex post intervention on ex ante investment incentives. Competition authorities should seek alternative remedies to price regulation, which are designed to address demand side problems and thereby to activate competition in the market.

While these various changes to policy would not overcome all of the concerns raised by advocates of a laissez-faire approach to excessive pricing, they would go a long way towards doing so. In the view of the authors, adoption of such an approach would tip the balance towards maintaining a policy of intervening against excessive prices under Article 82. One important factor in reaching this conclusion is that many jurisdictions do not have other levers to press in order to seek alternative demand side remedies designed to make markets work more effectively, of the sort set out below in Box 1. Box 1: Possible demand side interventions to enhance entry/expansion 1. Interventions to facilitate switching • Cancellation rights (improved, clearer, easier to apply) • Number (or similar) portability—where customers are attached to an element of their present product

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 545

Towards an Appropriate Policy for Excessive Pricing 545 • Customer information portability—where new suppliers face information asymmetries/disadvantages relative to current suppliers • Improved ease/speed of switching 2. Interventions to facilitate shopping around • • • •

Cooling off periods (improved, clearer, easier to apply) Written quotations, lasting for a fixed period Search routines and price comparison websites Consumer education and advocacy (about the benefits that shopping around could bring)

3. Interventions to facilitate comparisons between offerings • • • •

Clear upfront prices (for all elements of the package) Standardized prices (e.g., APRs for credit products) Standardized products (e.g., in insurance—base line cover) Price comparison tables (e.g., as provided for loan products on the website of the UK Financial Services Authority)

4. Interventions to overcome asymmetric information problems • • • • •

Warranties schemes Complaints/redress schemes Minimum quality standards Licensing schemes Legislation that encourages the use of independent advisors (as in financial services) • Codes of practice

(Y) Ehlermann 08 Fletcher

6/11/08

11:54

Page 546

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 547

IV

Ian S. Forrester, QC* Sector-Specific Price Regulation or Antitrust Regulation— A Plague on Both Your Houses? August 2007

Mercutio, a friend of Romeo, is stabbed by Tybalt as part of the endless feuding between the warring houses of Montague and Capulet in the Verona of Shakespeare’s play, and before dying cries out, “A plague on both your houses”.1 My initial reaction to the choice posed in relation to remedies was similar to that of Mercutio. I had great reservations about sector-specific price regulation and its impact on competition, and great reservations about the fitness of competition law enforcers to be price-setters. In the process of reflecting in the past weeks on the specific problems presented by both routes, and in light of some personal experience, my views have shifted somewhat. Enthusiasts of competition law generally believe it can do good and can remedy identified abuses. Are there areas where using currently available competition law remedies will simply do more harm than good, areas where it should not go? Al Gore, a now fondly-remembered political leader, suggested that government should modestly set as a goal doing no harm. The unintended benefits for the collectivity of private enterprise, and strong scepticism about the economic prudence of government, had long ago already been voiced by Adam Smith. He wrote: “Every individual necessarily labours to render the annual revenue of the society as great as he can. He generally, indeed, neither intends to promote the public interest, nor knows how much he is promoting it. By preferring the support of domestic to that of foreign industry, he intends only his own security; and by directing that industry in such a manner as its produce may be of the greatest value, he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention. Nor is it always the worse for the society that it was no part of it. By pursuing his own interest he frequently promotes that of the society more effectively than when he really intends to promote it.”2 * Queen’s Counsel at the Scots Bar; Visiting Professor, University of Glasgow; White & Case, Brussels. Warm thanks are expressed to Jacob Borum, Sandra Keegan, Tom Winsor and others unnamed for their contribution to this paper. The opinions expressed are wholly personal. 1 Romeo and Juliet, Act III, Scene 1. Shakespeare wrote “A plague o’ both your houses”, rather than the today more easily understood “on”. 2 Adam Smith, The Wealth of Nations, Book IV, chapter 2.

(Z) Ehlermann 08 Forrester

548

6/11/08

11:54

Page 548

Session Three: Exploitative Conduct And The Remedies

The celebrated Scot concluded that: “kings and ministers . . . are themselves always, and without any exception, the greatest spendthrifts in the society.”

We cannot assume the state or the civil service or regulatory bodies to be any better than anyone else at identifying what will be the consequences of regulatory initiatives. The conclusions of the well-informed and well-intentioned officials can be completely wrong. Putting matters differently, I have no confidence whatever that, when public bodies regulate, they will do so successfully. This is not to be discourteous; not to deny the inevitability of regulation; and not to preach for an unregulated Stone Age. It is a reflection of experience: agencies can make mistakes; agencies advance their own interests as well as the public interest; agencies rarely dissolve themselves; agencies are slow; agencies too rarely look back to consider whether their intervention was, in the light of experience, useful; agencies may be captured by the industry for whose oversight they are responsible, or they will at least (legitimately) hesitate to assume the industry’s status quo is wrong.3 I note an echo of this in an opinion of Justice Brennan,4 who rejected the argument of a city that its power-generating activities, being in the public interest, were entitled to a kind of immunity from the antitrust laws. “[The City’s] argument that their goal is not private but public service is only partly correct. Every business enterprise, public or private, operates its business in furtherance of its own goals . . . the economic choices made by public corporations in the conduct of their business affairs, designed as they are to assure maximum benefits for the community constituency, are not inherently more likely to comport with the broader interests of national economic well-being than are those of private corporations acting in furtherance of the organization and its shareholders.”

The question of whether ex ante regulation is preferable to the unforeseeability of ex post competition law, with specific reference to the telecommunications sector, was debated in the 1998 Workshop in this series.5 On that occasion I voiced the proposition that, at least in the context of the regulation of the telecommunications markets, it was very doubtful that the talents, care and regulatory comprehensiveness deployed by national regulators were usefully being deployed. I feared that these elements were oriented towards regulating the market (honestly, conscientiously, carefully) rather than creating 3 There is a whole area of enquiry called Public Choice Theory on this theme. For those who want the joy of reading an elegantly literate paper by a master theorist, I commend Richard Posner’s paper on “Theories of Economic Regulation”, NBER Working Paper no. W0041, Center for Economic Analysis of Human Behavior and Social Institutions, National Bureau of Economic Research Inc., New York (May 1974). Also published in 5(2) Bell Journal of Economics and Management Science 335 (1974). 4 City of Lafayette v. Louisiana Power & Light Co., 435 U.S. 389, 403 (1978). 5 See Claus-Dieter Ehlermann and Louisa Gosling, eds., European Competition Law Annual 1998: Regulating Communications Markets, Hart Publishing, 2000.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 549

Sector-Specific Price Regulation or Antitrust Regulation

549

a competitive market. That there have been constraints imposed on such an unfortunate outcome is a credit to the design and functioning of the regulatory framework put into place in 2002.6 So, I begin with a stern disclaimer, noting the imperfections and inescapable unreliability of public entities when facing regulatory choices. By regulating too closely they can prevent or discourage the emergence of something useful, profitable or enjoyable. It should not be assumed that entrusting matters to a public sector-specific body will achieve the desired good result. At a minimum, this should be a matter of regular ex post verification. The situations where price control may be a useful option could include an evidently permanent monopoly which is not capable of being duplicated. The permanent way underlying a railway network with track and signaling devices is one example: price controls may be the best means to decide how much train services operators should pay for running a certain service three times per weekday on a certain route. Use of runways at an airport might be another, geographically much narrower example. Restaurants or car parks or bank-tellers at an airport may need protection against a greedy lessor seeking to capitalize on limited airport space. I am sure there are other situations in which price-setting may be appropriate.

6

Directive 2002/21/EC of the European Parliament and of the Council of 7 March 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive); Directive 2002/19/EC of the European Parliament and of the Council of 7 March 2002 on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive); Directive 2002/20/EC of the European Parliament and of the Council of 7 March 2002 on the authorisation of electronic communications networks and services (Authorisation Directive); Directive 2002/22/EC of the European Parliament and of the Council of 7 March 2002 on universal service and users’ rights relating to electronic communications networks and services (Universal Service Directive); Directive 2002/77/EC of 16 September 2002 on competition in the markets for electronic communications networks and services (Text with EEA relevance); Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector.

These Directives, which constitute the present telecoms regulatory framework, are available at http://ec.europa.eu/information_society/policy/ecomm/info_centre/documentation/legislation/ index_en.htm. See also Commission Communication on the Review of the EU Regulatory Framework for electronic communications networks and services (COM(2006)334), 29 June 2006; Implementation of the Regulatory Framework in the Member States 2005—11th Report, 20 February 2006; and Implementation of the Regulatory Framework in the Member States— 10th Report, 2 December 2004. All pertinent documents are available at http://ec.europa. eu/information_society/policy/ecomm/info_centre/documentation/communic_reports/index_en. htm.

(Z) Ehlermann 08 Forrester

550

6/11/08

11:54

Page 550

Session Three: Exploitative Conduct And The Remedies

Does Article 82 offer a better chance of getting things right? The presence of Article 82 in European law (and other competition laws around the world modeled on that of Europe) distinguishes our law from US law. It is directed against two kinds of abusive conduct, exploitative and exclusionary. Modern antitrust theory holds that we should focus on the latter and generally leave it to the marketplace to remedy the former. It holds that competition enforcers should not be minded to help individual competitors, and should not see themselves as horse race handicappers whose notion of perfect competition is a robustly contended race whose winner is uncertain. Competition authorities, on this theory, should stay away from “price gouging” cases. They should be very reluctant to take on pure excessive pricing cases. As long ago as 1977, Mario Siragusa7 was suggesting that Article 82 should be directed mainly to exclusionary conduct and might be limited, with respect to exploitative abuses, to prohibiting a narrow range of conduct. He suggested limiting cases of abuse not involving distortions of competition to those enumerated in Article 82(b). So the notion is by no means new. However, we have to accept that the judgments of the European Courts on dominance do not commonly read as if they had been drafted by Chicago School lawyers or economists. The classic definition of dominance is: “A position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately of consumers.”8

And such a firm is burdened by a duty: “A finding that an undertaking has a dominant position . . . simply means that, irrespective of the reasons for which it has such a dominant position, the undertaking concerned has a special responsibility not to allow its conduct to impair genuine undistorted competition in the common market.”9

So even if we feel that Article 82 ought principally to be applied to exclusionary abuses and ought not to be used as a price control instrument, the European Court of Justice has made it clear that the Article applies to exploitative abuses. 7 Mario Siragusa, “Application of Article 86: tying arrangements, refusals to deal, discrimination and other cases of abuse”, in Regulating the behaviour of monopolies and dominant undertakings in Community law, 1977 Bruges Week, Cahiers de Bruges, N.S. 36, De Tempel, Bruges (1977). 8 Case 27/76, United Brands v Commission [1978] ECR 207, para. 65. 9 Case 322/81, NV Nederlandsche Baden-Industrie Michelin v Commission [1983] ECR 3461, para. 57 (emphasis supplied).

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 551

Sector-Specific Price Regulation or Antitrust Regulation

551

There is another reason not to eschew the pursuit of exploitative abuses. The enforcement of competition law in such countries as Kenya, Albania and Burkina Faso (all members of the International Competition Network) will necessarily involve the occasional checking of attempts to price-gouge in such areas as milk, taxis, bread and other basic staples. Indeed, the Greek Minister of Economy showed that he was not immune to the populist charms of competition law when he called for an enquiry into the supposedly suspicious fact that a number of banks increased their lending rates one after the other. This may be a kind of abuse of the competition law process, but I suppose we are better off with an enquiry being opened, conducted and closed than with ministerially imposed bank rates. So even though we may be rather embarrassed by old cases like United Brands,10 we cannot escape the fact that, from time to time, complaints about alleged exploitative pricing abuses have to be investigated. The Court’s statement that dominant companies are under a “special responsibility” suggests that, by virtue of their position, they have to pursue more pro-competitive behaviour (presumably in order to make life a little easier for competitors) than non-dominant players are required to do. When teaching students, I have commonly used the simile of the large jungle animal being obliged to tread warily lest it eliminate smaller creatures. Eleanor Fox considers such an image to be outdated: I fear she may be over-optimistic. This brings us to the debate on whether the competition rules or sectorspecific rules are better adapted to achieving effective remedies. To state what is obvious but fundamental, competition law enforcement should condemn and remedy and possibly punish breaches of the law, either existing ones or newly-recognized ones. It is not intended in the first place to change how the marketplace functions, although that will normally be a consequence of its invocation. As a general principle, parties to infringing conduct are free to choose how they will terminate their infringements. Sector-specific regulation is intentionally interventionist, prescribing precisely what operators must do.

The strengths and weaknesses of national regulatory authorities I begin by considering certain features of how competition principles have been applied in two particular marketplaces, telecommunications and railways. In the field of e-communications, technological developments and Commission legislative pressure have eliminated many aspects of the monopolies formerly held by incumbents, but the level of much new telecoms 10

Supra note 8.

(Z) Ehlermann 08 Forrester

552

6/11/08

11:54

Page 552

Session Three: Exploitative Conduct And The Remedies

regulation remains maximalist rather than minimalist (and in my humble opinion not sufficiently de-regulatory).11 More specifically and more alarmingly, the forthcoming Roaming Regulation,12 which has been the subject of final agreement between the European Parliament and the Council, cannot be reconciled with the supposedly de-regulatory philosophy of the rest of EU telecoms legislation. I suggest below that populism has on this occasion corrupted sound lawmaking. I further suggest that sector-specific regulatory bodies should not use the pricing tool to protect the public interest without first considering how other tools could enhance competitiveness. The regulator could compel the monopolist (or holder of significant market power) to practise a policy of transparency of terms, conditions and pricing; or a policy of non-discrimination in respect of commercial terms and conditions offered; or to apply separation of cost accounting rules which accurately reveal the weighting of the various cost elements collectively exploited, thereby facilitating a transparent evaluation of specific cost elements and pricing strategies. Unfortunately, a pricing obligation is seen as more effective and precise, while telecoms experience suggests that once established, obligations are not readily dismantled. To quote Richard Posner (no less!), “[b]ehind each scheme of regulation could be discerned a market imperfection the existence of which supplied a complete justification for some regulation assumed to operate effectively and without cost”.13 Yet “regulation is not positively correlated with the presence of external economies or diseconomies or with monopolistic market structure”.14 I submit that we should not make the assumption that society functions better due to sector-specific measures applied by national officials on a standing basis. The permanence of such measures is one of their greatest drawbacks. Regulatory authorities are usually staffed by technicians, engineers and others more familiar with the technology than they are familiar with the consequences of the technology in the market.15 11 There have been admirable statements from, for example, then-Commissioner Liikanen: “I hope regulators in applying these rules will remember that this regulatory framework is about rolling back regulation and promoting competition. The framework must work in a way which supports and does not stifle emerging markets and services. It is now up to national regulators to abide by this Common Position and to provide a sound justification whenever they deviate from it.” Press Release IP/04/528 of 23 April 2004, available at: http://europa.eu/rapid/ pressReleasesAction.do?reference=IP/04/528&format=HTML&aged=1&language=EN&guiLa nguage=en. I am not sure how much Liikanen has been heeded. 12 Regulation 717/2007 of the European Parliament and of the Council of 27 June 2007 on roaming on public mobile telephone networks within the Community and amending Directive 2002/21/EC, [2007] OJ L171/32. 13 Posner, cited in footnote 3 above: 5(2) Bell Journal of Economics and Management Science 335, 336 (1974). 14 Ibid. 15 For an illuminating appreciation of what regulators examine in coming to their decisions, see: http://ec.europa.eu/information_society/policy/ecomm/article_7/index_en.htm. See also: http://circa.europa.eu/Public/irc/infso/ecctf/library.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 553

Sector-Specific Price Regulation or Antitrust Regulation

553

The regulators of electronic communications for a country are always much more numerous and better-funded than the regulators of competition for the entire economy of the same country. This may mean that we have invested too cautiously in competition regulation, or that we are spending too much on e-communications regulation. The e-communications regulators may be over-regulating in the drive towards competition. Maybe it would be possible to draw conclusions about the utility of this regime by examining the market share, or diminution of market share, of the incumbent, or by observing the number of competing suppliers for these services. For example, on 7 February 2006, the Commission adopted the “Communication on Market Reviews under the EU Regulatory Framework— Consolidating the internal market for electronic communications”,16 which reflects the growing competitivity of many of the markets that were formerly the object of monopoly. However, I suspect that the return on investment of official diligence in the cause of de-regulating telecoms markets might seem disappointing. Regulators commonly have a preoccupation with costing which does not encourage them to take a non-cost approach to particular problems. In past years, they devoted much skill to eliminating cross-subsidies from the costs 16 Commission Communication on Market Reviews under the EU Regulatory Framework: Consolidating the internal market for electronic communications, COM(2006) 28 final, 6 February 2006 (issued pursuant to Article 7 of the Framework Directive, cited supra note 6):

“The regulatory framework for electronic communications set up in 2002 involved a major overhaul in regulatory approach, linking sector-specific regulation and competition law in a novel way. The previous, more mechanistic approach to regulation was replaced by an economic approach where regulation is based on competition law principles. It requires national regulators to conduct “market reviews” in order to determine whether a particular market should be regulated. Where a given market is susceptible to ex ante regulation and a regulator finds one or more undertakings to have significant market power or ‘SMP’ (equivalent to ‘dominance’ under competition law) on that market, it must impose appropriate regulation. Conversely, regulation must not be imposed, or existing regulation must be withdrawn, where no undertaking is found to have SMP. The market review process is subject to scrutiny by the Commission under the Community consultation mechanism established under Article 7 of the Framework Directive.” (Paragraph 1.—Introduction) (emphasis in original) “Article 7 of the Framework Directive requires national regulatory authorities (NRAs) to notify the Commission and other NRAs of their findings as to market definition and SMP analysis and the regulatory obligations they intend to impose (or remove) (their ‘proposed measures’). The Commission then has one month in which to assess the notification of the proposed measures (the ‘phase one’ procedure). The vast majority of cases are handled within this one month period by a letter to the NRA concerned, which may contain Commission comments as to how the measures in question could be further improved. In cases where the Commission considers that, in terms of market definition or SMP analysis, the proposed measures would create a barrier to the single market or if it has serious doubts as to the measures’ compatibility with Community law (and in particular the common policy objectives that all NRAs should pursue), it can conduct a more detailed investigation lasting a further two months (‘phase two’). Following this in-depth investigation, should its concerns be confirmed, the Commission may require the NRA to withdraw the draft measures (‘veto’ decision) and possibly resubmit the market analysis in question at a later stage.” (Paragraph 2.—Article 7 Procedures: overview)

(Z) Ehlermann 08 Forrester

554

6/11/08

11:54

Page 554

Session Three: Exploitative Conduct And The Remedies

claimed by the incumbent operator for different kinds of services in the bundle of universal service. It was not so obvious that they were well adapted to fine-tuning their area of operations. I diffidently suggest that regulators have a reluctance to find that effective competition will be sustainable in the absence of regulation, and that they prefer to arrange matters so as to place the incumbent at some disadvantage. There is also a problem of lack of expertise (and autonomy) in national telecoms regulators. Unfortunately, very few are as able, conscientious and independent as Ofcom in the UK or its counterparts in the Republic of Ireland or in France. The Commission has commented on this in its annual reports on the implementation of the regulatory framework (see, e.g., the 10th, 11th and 12th Implementation Reports, starting with the 12th17). The Commission has expressed concern about the “independence and impartiality” of the Polish authority, and about State influence in Slovakia. The quality of the analysis of reports and the rigour of the conclusions of a well-resourced expert agency and those of a small, weak agency will vary greatly: “NRAs Independence In general the NRAs have consolidated their authority and independence. Doubts have been expressed, however, in the case of Slovakia regarding separation of regulatory functions and control of State ownership of market players. A fresh concern regarding the independence and impartiality of the new NRA has arisen in Poland, moreover, in the light of the abolition of its predecessor and the scope of the government’s powers of dismissal. The extent of political influence over day-today regulatory decisions in some Member States is an issue calling for further examination. New entry and cross-border investment will only reach their full potential where the independence and impartiality of the regulator can be relied on by the market. Market Reviews The process of notification and consultation of the Commission and other NRAs under Article 7 of the Framework Directive is a key tool for ensuring that the benefits of consistent regulatory policy feed through to all European users.”18

The competition and sectoral rules governing the electronic communications, information technology and broadcasting fields are a perpetual source of change and controversy as technology evolves. Rules developed in the 17 http://ec.europa.eu/information_society/policy/ecomm/implementation_enforcement/ annualreports/12threport/index_en.htm. 18 Commission Communication on European electronic communications regulation and markets 2006 (12th report), COM(2007) 155, 29 March 2007. The Communication is available from the Commission’s website at: http://ec.europa.eu/information_society/policy/ecomm/ implementation_enforcement/annualreports/12threport/index_en.htm.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 555

Sector-Specific Price Regulation or Antitrust Regulation

555

context of scarcity (limitations on the number of available broadcast channels, the number of local telephone providers, and other natural monopolies) ought to be re-examined in the context of technological abundance to see whether the objective underlying the rule requires its continued application. This argument has failed to find much favour with the Member States, the European Parliament or Commission: hence the new TV Without Frontiers Directive. Many academic and commercial commentators debate the topic, most notably in internet space.19 In the realm of broadcasting we may distinguish between two kinds of intervention, namely, economic and cultural. Economic intervention is being gradually rolled back by Community law. The cultural specificity of public service broadcasting and audiovisual regulation at EU level (I speak as someone who from childhood has been informed, entertained and formed in different ways by the BBC, an extraordinary public service broadcaster) is a topic with which some Member States and regulators have difficulty in grappling. But that is for another day. Regulatory agencies try to apply their abilities to settling precisely-defined controversies, such as the terms on which an incumbent must provide subscriber information to the publisher of a rival telephone directory. They are familiar with the market and very ready to decide individual problems with confidence. I believe they are approaching their roles without a consistent view of the pro-competitive environment which must be the overall goal. When we look at the decisions taken by the 27 national regulators in telecoms, it would be my submission that they have a tendency to resort to pricing remedies as the first modality rather than as the last, reversing the apparent intent behind the Community legislation. Annex IV to the Commission’s Communication on Article 7 gives an overview of the many remedies imposed around Europe.20 I submit that there is a preponderance of remedies relying on the full suite of possibilities including price controls, rather than less intrusive remedies. My concern is that, by failing to discern clearly the competitive trends in telecoms markets while faithfully trying to apply at local level to local controversies very detailed cost-driven solutions to fundamental or minor issues of competition, the National Regulatory Authorities, in the belief they are doing the right thing, are imposing their own rigidities and inflexibilities on a

19 See, for example, the long list of public and private bodies that submitted comments in response to the Consultation on the European Commission’s ‘Communication on the Review of the EU Regulatory Framework for electronic communications networks and services’ (adopted in June 2006): http://ec.europa.eu/information_society/policy/ecomm/info_centre/documentation/public_consult/review_2/index_en.htm. See also the external expert studies in that respect for the European Commission: http://ec.europa.eu/information_society/policy/ecomm/info_ centre/documentation/studies_ext_consult/index_en.htm#2006. 20 Commission Communication on Market Reviews under the EU Regulatory Framework, supra note 16.

(Z) Ehlermann 08 Forrester

556

6/11/08

11:54

Page 556

Session Three: Exploitative Conduct And The Remedies

national basis. If we look at some of the published telecoms decisions,21 using the market data that is published with the decisions, we may conclude that regulators have a bias towards regulation (over-regulation?) via price remedies rather than imposing the lightest possible obligation. I further take the liberty of suggesting that the Commission itself is not immune to the temptation of controlling rather than liberating. Where national regulators consult the Commission, there may be a tendency within the Commission to recommend action or inquiry rather than recommending or demanding inaction. Thus, DG Competition may say “do not bless this practice”, when DG Information Society says “no problem”. To be fair, there are also cases where DG Competition says “do not regulate”,22 and officials correctly emphasize that the trend is clearly in the direction of discouraging regulation. But as with all tides, there are ebbs and flows. There appear to have been at least a few cases where there were differing views, as the Commission’s comments on these cases seem to indicate that more than one opinion could be expressed.23 Circumstances could arise where, for a number of reasons, the Commission’s services could be driven by different objectives. For example, DG Competition may be handling a competition enforcement action on a particular market that has been notified to the Commission for approval under Article 7 of the Framework Directive.24 DG Competition officials will perfectly naturally wish to ensure that the outcome of the ex ante case is not at odds with their own position in the ex post case, regardless of the merits of the ex ante case. Another example could be where an intervention might be difficult to justify on the basis of competition law but where such an intervention might nonetheless be desirable in terms of enforcement goals. This might be the case for a joint dominance problem,

21

See http://circa.europa.eu/Public/irc/infso/ecctf/library. See http://ec.europa.eu/information_society/newsroom/cf/itemdetail.cfm?item_id=2190, news report of 29 September 2005, where the Commission announced that: “The 52 alternative telephone network operators providing “call termination” services (meaning services operators buy from each other to enable their customers to make calls to different networks) within Germany have significant market power on their own individual networks, as does Deutsche Telekom AG, confirms the European Commission in a decision under Article 7 of the EU framework on electronic communications today. The German regulator, Bundesnetzagentur (BNetzA), recently carried out a fresh market analysis, after the Commission found in May 2005 that the earlier BNetzA (then RegTP) analysis lacked evidence of the absence of significant market power of the alternative network operators. In the new notification, BNetzA took into account the arguments made by the Commission in the veto decision, in particular with respect to the effects of regulation on the incumbent. As a result, BNetzA will regulate the termination market for fixed calls.” 23 The Dutch wholesale broadband access letter was one such example (available at: http://circa.europa.eu/Public/irc/infso/ecctf/library?l=/nederland/registeredsnotifications/nl2005 0281/fileserve_enrpdf/_EN_1.0_&a=d), whilst the Spanish mobile comments letter (currently under appeal) is another (http://circa.europa.eu/Public/irc/infso/ecctf/library?l=/espaa/ registeredsnotifications/es20050330/final_enpdf_15/_EN_1.0_&a=d). 24 See supra note 6. 22

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 557

Sector-Specific Price Regulation or Antitrust Regulation

557

where the burden of proof for ex ante intervention is thought to be not so high as under ex post interventions.25 The relevance to my proposition of this phenomenon is that competition enforcement is not a neutral, abstract discipline. It is of course shaped by policy agendas. I will now describe a recent regrettable example where an established philosophy or policy was disregarded for political ends. As of the beginning of June 2007, the European Commission, for the first time in fifty years, set the retail price of a commonly-used service.

The Regrettable Example of the Roaming Story: Populism and Pricing Remedies In 1999 it was agreed that telecommunications price controls should gradually fade away, leaving market forces to take over.26 But in the summer of 2007, we have heard that, in a new political “breakthrough”, the Commission and European Parliament have tackled the problem of “excessive” roaming charges. Who could be opposed to such a happy result? I submit that the intelligent consumer should have doubts, and the student of Community law should share those doubts. Mobile telephone operators charge customers for originating calls but not for receiving calls in their own country in Europe. (Landline operators normally apply no charge for receiving calls.) “Roaming charges” apply when the customer goes to another Member State and makes or receives calls there. The customer is charged not only for each call so made but also for all calls received. The charges imposed by mobile operators for providing these roaming services are considerably higher than the charges for “normal” domestic calls. For those wishing to check charges from Florence to home, I recommend “Europe’s Information Society Thematic Portal: Sample post-paid voice tariffs for travellers from Italy”.27 The roaming charges have been a source of considerable profit for the mobile telephone companies. Every mobile operator follows a similar policy. There has been a lengthy competition enquiry into possible price coordination on roaming charges. Indeed, there were raids on most of the big mobile telephone operators in 2001 in search of cartel behaviour, and then there were enquiries as to domestic dominance and abuse. However, despite a lot of enquiry, it appears that 25 Once a firm or firms have been identified as dominant, there is no need to justify an intervention under the ex ante rules, where issues of abuse or consumer harm should not arise. 26 See the Commission’s report on the 1999 Review of the then-Telecoms Package, available at: http://europa.eu.int/ISPO/infosoc/telecompolicy/review99/com2000-239en.htm (in particular, section 3, “Access and Interconnection”). 27 See http://ec.europa.eu/information_society/activities/roaming/index_en.htm.

(Z) Ehlermann 08 Forrester

558

6/11/08

11:54

Page 558

Session Three: Exploitative Conduct And The Remedies

the Commission has found no evidence. So why should mobile companies not be free individually and unilaterally to impose a higher tariff for certain services—as a reflection of commercially intelligent, self-regarding pricing strategies that take account of consumer demand and available alternative services? Read what the European Commission says about its breakthrough: “What has the Commission done so far? In 1999, the Commission launched a wide antitrust sector inquiry into the level of retail and wholesale roaming prices in Europe because of concerns that they were excessive. This inquiry led to the opening of separate antitrust proceedings under EC Treaty rules on abuse of monopoly power (Article 82) against mobile operators in Germany and in the UK for excessive wholesale international roaming tariffs. These proceedings are still under way. ... An update of the website (http://europa.eu.int:information society/roaming) in March 2006 showed that not much has happened since October 2005. In some cases prices even went up. ... The European Parliament voiced concerns and asked the Commission to develop new initiatives in order to reduce the high costs of cross-border telephony. It was against a background of complete lack of industry-driven progress over many years, that the Commission decided that action was required. ... Operators have offered reductions in the form of various packages, but as these are too little, too late, the Commission tabled on 12 July 2006 an EU regulation which will cut the cost of using your mobile abroad within the EU. The proposed EU regulation would not fix an ideal price for roaming charges, but would ensure that mobile roaming charges are not unjustifiably higher than the charges at home.”28

If this sounds nice, let us recollect some law. Ex ante burdens in the telecommunications sector can be placed only on companies that have been found to have significant market power.29 In order to have significant market power in this context, a company must be deemed to enjoy dominance within the meaning of Article 82 read in conjunction with the Framework Directive.30 In addition, three clear conditions have to be met: first, the market must be one that tends over time not to produce competitive outcomes;

28 Excerpts from a European Commission Information Society and Media Fact Sheet 59 (“Switch on to a Eurotariff ”) (July 2006). 29 Commission guidelines on market analysis and the assessment of significant market power under the Community regulatory framework for electronic communications networks and services (“SMP Guidelines”), [2002] OJ C165/6, at para 5. See also Articles 14–16 of the Framework Directive 2002/21/EC, supra note 6. 30 SMP Guidelines, cited in previous footnote.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 559

Sector-Specific Price Regulation or Antitrust Regulation

559

furthermore, it must be a market that exhibits high barriers to entry; and third, if the market is found to be uncompetitive, an intervention on the basis of competition law must be deemed to be inadequate.31 When we look at the outcomes of national regulators’ examinations of the market for roaming charges, we find that not one market player was found to have a dominant position. So ex ante obligations could not be imposed on mobile operators compatibly with the philosophy of the electronic communications regulatory framework. The Roaming Regulation itself 32 produces some charming words incapable of being fitted comfortably into a competition law analysis. With my own comments interspersed, I quote: “. . . Although some operators have recently introduced tariff schemes that offer customers more favourable conditions and lower prices, there is still evidence that the relationship between costs and prices is not such as would prevail in fully competitive markets.”33 “The creation of a European social, educational and cultural area based on the mobility of individuals should facilitate communications between people in order to build a real ‘Europe for Citizens’.”34 [This is central planning law, not competition law!] “This Regulation is not an isolated measure, but complements and supports [in ways not explained or demonstrated], insofar as Community-wide roaming is concerned, the rules provided for by the 2002 regulatory framework for electronic communications. That framework has not provided national regulatory authorities with sufficient tools to take effective and decisive action with regard to the pricing of roaming services within the Community, and thus fails to ensure the smooth functioning of the internal market for roaming services. This Regulation is an appropriate means of correcting this situation.”35 [I respectfully disagree: the premise is that the market for roaming services is not competitive because there is excessive pricing; if there is excessive pricing but no one company is dominant, and we still see no competitive outcome, then perhaps there is an Article 81 problem. But DG Competition can find no such problem. This suggests that there is a lack of jurisdiction to use either Article 81 or Article 82.]

31 Commission Recommendation on Relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC, [2003] OJ L114/45, at para. 9. See also Articles 14–16 of the Framework Directive 2002/21/EC, supra note 6. 32 Regulation (EC) No 717/2007 of the European Parliament and of the Council of June 27 2007 on roaming on public mobile networks within the Community and amending Directive 2002/21/EC [on a common regulatory framework for electronic communications networks and services], [2007] OJ L171/32. Available at http://eur-lex.europa.eu/LexUriServ/site/en/oj/2007/ l_171/l_17120070629en00320040.pdf. 33 Regulation 717/2007, supra note 12, paragraph 1. 34 Ibid., paragraph 2. 35 Ibid., paragraph 4.

(Z) Ehlermann 08 Forrester

560

6/11/08

11:54

Page 560

Session Three: Exploitative Conduct And The Remedies

“The Eurotariff should be set at a level which guarantees a sufficient margin to operators and encourages competitive roaming offers at lower rates. Providers should actively offer a Eurotariff to all their roaming customers, free of charge, and in a clear and transparent manner.”36 [Price-fixing for crowd-pleasing, I submit. Since when does the Commission have the right to set prices in an apparently competitive market? If the market is uncompetitive due to dominance or due to concertation, let action be taken.]

We are not told why the Commission considered that regulation was appropriate, proportionate and fulfilled the standards of better regulation (that is, only regulate when absolutely necessary). We are told that roaming prices were “excessive” even though all NRAs that examined their markets found no dominance.37 I regret to have to suggest that this regulatory intervention was a case of opportunistic populism in the guise of consumer protection. I have no professional interest in this particular battle, other than as a disinterested observer who believes in a more robust role for market forces in the regulation of market outcomes. I expect my family’s mobile telephone bills will slightly increase as prices domestically will go up. Roaming charges presumably cross-subsidize domestic calls. So the (maybe more prosperous) customers who travel abroad paid a bit more and the (maybe less prosperous) domestic customer paid a bit less.38 The result of this triumph of the political process on roaming will presumably be to raise the overall costs for mobile communications, notably for poorer EU citizens who do not travel to other Member States as often as Members of the European Parliament. I suspect that DG Competition tried to voice reservations about this new policy, and that its voice was disregarded. As noted above, political policy can override wise competition policy. As Chancellor Merkel observed: “From this summer onwards, mobile telephone customers will already have the benefit of permanently reduced roaming charges. [. . .] To me, this is a good example of a Europe which achieves results, a Union whose action confers practical benefits on its citizens.

36

Regulation 717/2007, supra note 12, paragraph 18. On 19 July 2007, it was announced that DG Competition had decided to close proceedings against Vodafone, O2 and T-Mobile, which had been accused of charging excessive prices for wholesale roaming charges. See http://europa.eu/rapid/pressReleasesAction.do?reference=IP/ 07/1113&format=HTML&aged=0&language=EN&guiLanguage=en. One may regard this closure as confirming the weakness of the Article 82 cases. 38 Consider another example of “differential pricing”. Restaurants commonly charge less for lunch, when everyone must eat, than for dinner when a smaller number choose to eat. The food is often the same. Those who eat in the evening may be thought better able to pay than those who eat out at noon. Restaurants individually lack significant market power but they very likely observe each other’s pricing patterns with satisfaction. Should we be shocked and try to correct matters by adopting a unique legislative initiative, or even by applying the competition rules? I suggest not. Among other considerations of consumer welfare, if the price of dinner is forced down, the price of lunch will very likely rise. 37

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 561

Sector-Specific Price Regulation or Antitrust Regulation

561

Sometimes technical legislation is hard for people to understand [. . .] but anyone who has one of these in their pockets [she took out a mobile phone] can grasp the benefits of this regulation.”39

I submit that we may see in this episode a bad precedent for the remedying of supposed competition problems. The ex ante mechanism established to permit regulation in the event of a market failure indicated that no intervention was warranted; but thereafter another mechanism was invoked. The process was politicized, with copious use of the media to dramatize the debate. The Regulation was adopted despite legislative provisions to the contrary. It was intended to change the result which market forces were delivering. And the consumer advantages of the supposed remedy (cheaper prices for those who most need cheaper prices) are by no means evident. The relevance for this gathering is that regulators may intend to do the right thing but do not always do so in the face of political pressure. And that is true at the stage of remedies as well as earlier in the proceedings.

An alternative regulated sector: railways In order to have a basis for comparison from a completely different industry, let us consider the railways in the UK. Some sixty years ago, the various UK railway companies were nationalized into a vast industrial empire run by the British Railways Board, which employed, managed or coordinated transport police, luxury hotels, ferries, steelmaking and engineering, as well as owning track, rolling stock and locomotives. The railways were spared the rigour of privatization during the Thatcher years (although the size of the empire was trimmed), and it was only in 1992 (after Mrs Thatcher had passed from office) that the railways themselves were to be transferred to private hands. Under the 1993 Railways Act, what used to be “British Rail” was broken up into a large number of separate, newly-created operations. The track itself as well as the signals and the stations went to Railtrack, with the maintenance of the track and signals going to British Rail Infrastructure Services. The Office of the Rail Regulator has the role of supervising competition-related aspects

39 http://www.europarl.europa.eu/news/expert/infopress_page/008-8367-178-06-26-90120070626IPR08366-27-06-2007-2007-true/default_en.htm. For further references to the saga, see the Commission’s press releases: “International Mobile Roaming: how will the new “Eurotariffs” reduce the cost of using a mobile phone in the European Union?”, available at http://europa.eu/rapid/pressReleasesAction.do?reference=MEMO/07/251&format=HTML&ag ed=0&language=EN&guiLanguage=en, and “EU Roaming Regulation enters into force across all 27 Member States on 30 June”, http://europa.eu/rapid/pressReleasesAction. do?reference= IP/07/870&format=HTML&aged=0&language=EN&guiLanguage=en.

(Z) Ehlermann 08 Forrester

562

6/11/08

11:54

Page 562

Session Three: Exploitative Conduct And The Remedies

of the newly constructed rail industry. The Rail Regulator was an economic regulator (questions of safety were not then part of the Regulator’s remit), with powers to supervise the contractual terms under which access could be granted to the track, for example. The UK regime is endowed (in railway matters as well as in others) with large resources and great technical expertise. In other Member States, rail regulatory authorities may not be independent or may be merely advisors of a minister of transport; making reliable and wellfounded decisions would be very challenging for a less well-funded regulator. The complex history of the privatization and regulation of the UK freight and passenger railway industry is not for this paper. Relevant for our purpose is that the Office of the Rail Regulator had both ex ante price control powers and ex post competition powers, which allowed it to react to complaints. I asked Tom Winsor, who was then Rail Regulator, to compare his Office’s experiences in using these two kinds of power. He felt that his ex ante powers were highly preferable to his competition law enforcement powers. This conclusion deserves some explanation. The Rail Regulator was under a duty when licensing under the Railway Act 1993 to consider whether to proceed under the Competition Act 1998. Although the competition powers looked strong, and although the detailed statutory instruments had been prepared by masters in the art of drafting from the Rail Regulator’s point of view, they led inexorably into an arena dominated by clever lawyering which seemed never to reach a conclusion on the merits. For example, in 2001, Enron (then in its heyday) complained that the freight operator EWS had abused its dominant position in charging for the rail transport of petrochemicals. That investigation consumed an inordinate amount of time and legal fees, and was finally concluded in November 2006. So the only completed competition case in the UK railway sector pursuant to the new legislative regime consumed more than five years and involved platoons of lawyers and economists, arguments over access to the file and the whole procedural apparatus of courteoussounding combat which characterizes British public sector enquiries into weighty matters. By contrast, the regulatory regime established in the rail sector involved a large element of price control which worked successfully. This was done via access contracts, unlike the telecoms or water industries. (The Regulator could also establish the terms of access to the network for dependent users. For example, particular stations were leased to the train operator which made the most use of the station; it was then necessary to regulate what and how that facility owner charged the tenant.) There was a very wide margin of appreciation for the regulator in establishing and enforcing licence terms. Thanks to Tom Winsor, I include below a chart (in the public domain) showing how access charges for railway services were established. The table demonstrates that the criteria were understandable and the outcome of the calculations was a known and transparently-arrived-at figure. Fare control was not left to the Rail Regulator, except in the case of open

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 563

Sector-Specific Price Regulation or Antitrust Regulation

563

Periodic Review Process The current access charges for franchised passenger services were set before privatization for the period to March 2001 (Control period 1). The periodic review will reset these charges for a further five years (Control period 2). It will also establish the asset base upon which Railtrack will continue to earn a reasonable return (Control periods 3 and beyond).

Regulatory Asset Base (RAB) The asset value on which Railtrack is expected to earn a reasonable return. This consists of the value at privatization plus the cost of enhancements

X Cost of Capital (COC) The rate of return which Railtrack is expected to earn on the RAB

Maintenance & renewal activity The amount of work which Railtrack needs to do between 2001 and 2006 to maintain and renew the network in appropriate condition

X Efficiency improvements The annual rate at which Railtrack can be expected to improve the efficiency of its maintenance and renewal activity

Required level of Profit The amount Railtrack needs to finance its investments (calculated by multiplying the RAB by the COC)

Revenue requirement The amount of revenue Railtrack needs from passanger access charges, freight, government grants or property sales/rental between 2001 and 2006

Maintenance & renewal spend The amount Railtrack needs to spend between 2001 and 2006 in order to maintain and renew the network in appropriate condition

access operators40 who were able to gain access to the network without elaborate government approval (although in practice no fares were prescribed in this context). The Department of Transport regulators do have power to set 40 See Directive 91/440/EEC of 29 July 1991 on the development of the Community’s railways, [1991] OJ L237/25, requiring all EU Member States to separate the management of railway operation and infrastructure from the provision of railway transport services, the idea being that the track operator would charge the train operator a transparent fee to run its trains over the network, and anyone else could also run trains under the same conditions (i.e., open access).

(Z) Ehlermann 08 Forrester

564

6/11/08

11:54

Page 564

Session Three: Exploitative Conduct And The Remedies

standard class peak hour tickets. (There was a complaint about Virgin Trains, as to London—Manchester fares, but it was concluded that Virgin was not dominant.) Car parks and their charges were the subject of other complaints: local monopolies are not infrequently the occasion of controversy. Thus, the Regulator could consider the condition and capacity of the network, the likely intensity of use of the network and the works necessary to improve performance, and it then could set charges for access to the rail network. The train operators were not opposed to a pricing mechanism. They needed to know what they would have to pay and needed to have rapid solutions to problems, including very specific questions about unreasonable charges. The Railways Act 1993 provided that every contract between Railtrack and the train operators for the use of the network must be approved by the Rail Regulator. Only in quite trivial cases was there an exemption. These rules were made more rigorous by a so-called Work Code with detailed rules about liability, quality of service and the like. The Regulator prescribed detailed pricing rules.41 The quality of the stewardship by Railtrack of its responsibilities was closely supervised and train operators were empowered to seek prompt, adequate and effective remedies. So the train operators, some of whom might have been nervous about making a competition law complaint, favoured the transparency, reliability and predictability of the regulatory regime over the competition regime. As already noted, the competition route was slow, painful, lawyer-dominated, and had all the formal apparatus of modern litigious combat. The regulatory regime was thorough, predictable, quite swift and reliable. So it is not the case that a pricing regime will necessarily be capricious, rigid or unfair. I find this admittedly anecdotal example to be instructive. I therefore come to the question of whether price regulation can be better than the use of the competition rules.

Regulation, antitrust and consumer protection legislation The purpose of sector-specific regulation for present purposes is to shape the marketplace in such a way that competition can develop and become sustainable. Regulation is an ex ante process, aiming at dictating conduct and commercial behaviour, as a substitute for real competition until the market 41 One notable difference between the rail and telecoms sectors relevant to the appropriateness of price regulation is the “replicability” of the regulated asset. Price controls have a particular ability to affect the “make or use” / “build or buy” decisions of competing firms where replication of the asset is possible. Should a new telecoms firm rent services over the infrastructure of the incumbent operator, or should it build its own infrastructure? This will depend crucially on traffic volumes and the price charged for renting as against the cost of building its own infrastructure. Where the underlying asset is plainly not replicable, as in the case of a rail network, errors in setting the price would have less serious consequences.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 565

Sector-Specific Price Regulation or Antitrust Regulation

565

can work properly. Competition law aims at forbidding and punishing anticompetitive conduct. It applies ex post (except of course for merger control). Article 82 should condemn specific acts deemed abusive that have occurred in the past. Regulation imposes a particular course of conduct on an undertaking (dominant or non-dominant) on the theory that this will ensure the better development of competition. The reality of experience is that regulation becomes permanent, and that regulators may inadvertently handicap the market forces that they are trying to release. And as competition law is used in more novel and complex fields, it becomes difficult to say whether an intervention is meant to punish an identified abuse, or whether it aims instead to adjust the manner in which a market functions. So the ex ante and ex post regimes can overlap. Consumer protection legislation should imply that customers are sufficiently informed to take decisions about competing offers based on knowledge and value for money, and thereby to obtain the benefits promised by unrestricted competition. Free competition maximizes consumer welfare. Consumer protection legislation aims, among other things, at safeguarding interests that competition alone would not ensure, and at removing the asymmetry of information and of bargaining power between producers and consumers. Consumer legislation tends to be contract-focused. It focuses on the rights of the consumer in connection with the purchase of a product or service. I submit that Community rules on consumer protection appear to be quite fragmented. It is not easy to see overarching big principles. “The concern for uniformity and effectiveness of Directive 93/13 in procedural matters as opposed to the lack of interest as to the substantive definition of fairness may suggest that the ECJ does not see a need for uniformity in the substantive concepts of the Directive; this is equivalent to stating that . . . the ECJ does not see Directive 93/13 as a measure of market integration, but simply as one of consumer protection: in order to ensure consumer protection, effectiveness and uniformity of remedies are far more important than the uniformity of substantive concepts.”42

All three types of regulatory measures are meant to maximize consumer welfare. However, whereas consumer protection legislation focuses on the individual consumer and contractual rights, competition law and sector-specific regulation generally focus on enhancing consumer welfare higher up the chain by ensuring that the market is characterized by healthy competition.43 Conventional Commission logic is that, as long as there is competition, benefits such as lower prices and better products will flow down to the consumer automatically. Although this is often the case, it is not always so. The Commission has been accused of focusing on competitors (i.e., maintaining a 42 Paolisa Nebbia, “Unfair Contract Terms in European Law: a Study in Comparative and EC Law”, in Modern Studies in European Law, Volume 15, Hart Publishing, 2007, pp. 171 et seq. 43 Sector-specific legislation is not without its own specific consumer protection measures. See, for example, Chapters II and IV of Directive 2002/22/EC (the Universal Service Directive), cited supra note 6.

(Z) Ehlermann 08 Forrester

566

6/11/08

11:54

Page 566

Session Three: Exploitative Conduct And The Remedies

certain number of competitors in the market) without convincingly showing that its intervention actually benefits the consumers. Recently, the Commission has indicated that consumer welfare will play a more prominent role in Article 82 enforcement policy.44 The complainant has a very important role in driving forward new Article 82 cases. It must often be hard for Commission officials to be sure whether they are helping complaining competitors (known, bleeding, lamenting, wellresourced) or victim customers/consumers (unknown, probably unaware of the controversy, not easy to poll, largely silent). This leads to the question of the adequacy of pricing remedies under Article 82.

Pricing remedies in previous Article 82 cases The Commission has established price-related remedies in a number of refusal-to-supply cases. In Zoja/CSC—ICI (“Commercial Solvents”),45 the Commission imposed two different obligations on Commercial Solvents (“CSC”) and Istituto Chemioterapico Italiano (“ICI”) when ordering them to end their infringement of Article 82 EC: • One obligation was precise and could be implemented immediately. This was the obligation to supply products to satisfy Zoja’s most urgent needs without further delay. Article 2 ordered CSC/ICI to “immediately furnish 60,000 kg of nitropropane or 30,000 kg of aminobutanol at a price not exceeding the maximum price they charged for these two products”. • The second obligation, less precise, was the obligation to supply Zoja on a long-term basis. Article 2 of the Decision ordered CSC/ICI to “submit for the approval of the Commission, within a period of two months from the notification of the present decision, proposals relevant to the subsequent supply of Zoja”. This same procedure was applied in Hugin/Liptons,46 concerning a refusal to supply spare parts for Hugin cash registers to Liptons. The Commission noted in its Decision that “[t]he price for such spare parts should be an appropriate market price between that which is currently charged by Hugin AB to Hugin UK and that which is currently charged by Hugin UK to end users in the United Kingdom and which allows to Hugin UK an adequate margin of profit and to Liptons a reasonable trade discount”. Article 3 of the Decision 44 See Press Release IP/05/162 of 19 December 2005: “The Commission wants to concentrate its resources on those anti-competitive practices that are most likely to cause harm to consumers. [. . .] The proposals made in the Discussion Paper on Article 82 would in a similar way imply a strong focus on those abuses of dominant positions most likely to harm consumers.” 45 [1972] OJ L299/51. 46 [1978] OJ L22/23.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 567

Sector-Specific Price Regulation or Antitrust Regulation

567

then required Hugin to submit proposals relating to the resumption of supplies of spare parts for Hugin cash registers to Liptons. In Magill TV Guide/ITP, BBC and RTE (“Magill”),47 the Commission ordered ITP, BBC and RTE to supply to each other and to third parties, on request and on a non-discriminatory basis, their individual advance weekly programme listings, and to permit reproduction of those listings by such parties. “ITP, BBC and RTE shall bring the infringements as mentioned in Article 1 to an end forthwith by supplying each other and third parties on request and on a nondiscriminatory basis with their individual advance weekly programme listings and by permitting reproduction of those listings by such parties. This requirement does not extend to information in addition to the listings themselves, as defined in this Decision. If they choose to supply and permit reproduction of the listings by means of licences, any royalties requested by ITP, BBC and RTE should be reasonable. Moreover, ITP, BBC and RTE may include in any licences granted to third parties such terms as are considered necessary to ensure comprehensive high quality coverage of all their programmes, including those of minority and/or regional appeal, and those of cultural, historical and educational significance. The parties are therefore required, within two months from the date of notification of this Decision, to submit proposals for approval by the Commission of the terms upon which they consider third parties should be permitted to publish the advance weekly programme listings which are the subject of this Decision.”48 (emphasis added)

In NDC Health/IMS Health: Interim Measures (“IMS Health”),49 the Commission ordered IMS Health to grant a licence to undertakings wishing to use the 1860 brick structure (a map of Germany drawn up for the convenience of the pharmaceutical industry). In Recital 215 of its Decision, the Commission noted that it was important to ensure that any fee charged by IMS Health was “reasonable”, and stated: “The order therefore provides for any royalties to be paid for these licences to be determined by mutual agreement between IMS and the party requesting the licence or failing that, by a Decision of the Commission on the basis of a determination by one or more independent experts.” The Commission instructed IMS Health to seek an agreement on the “reasonable” royalties to apply, failing which the royalties would be set by a further Commission Decision. The Commercial Solvents, Hugin/Liptons, Magill and IMS Health Decisions therefore show that, where a Commission Decision imposes an imprecise obligation (such as an obligation to supply something on “reasonable” terms), the Decision sets up a procedure to translate the imprecise obligation imposed by the Decision into a more specific obligation. Article 5 of the 2004 Microsoft Decision50 follows the same approach. 47

[1989] OJ L78/43. Article 2, Commission Decision 89/205 of 21 December 1988, Magill TV Guide/ITP, BBC and RTE, [1989] OJ L78/43. 49 [2002] OJ L59/18. 50 Available at http://ec.europa.eu/comm/competition/antitrust/cases/decisions/37792/en.pdf. On appeal: Case T-201/04, not yet decided as of this writing. 48

(Z) Ehlermann 08 Forrester

568

6/11/08

11:54

Page 568

Session Three: Exploitative Conduct And The Remedies

In Magill and IMS Health, the question of reasonable royalties in effect became unimportant, as in each case the Commission Decision was suspended by interim measures. My impression is that pricing problems were not crucial in Commercial Solvents or Hugin/Liptons (the latter of which was overturned on appeal51).

Excessive pricing cases Although even the earliest drafts of what became Articles 85 and 86 (now Articles 81 and 82) of the Treaty contemplated condemnation of the exploitation by dominant players of abusive price demands,52 examples of the interpretation of the concept and the prosecution of excessive pricing cases have been rare. General Motors Continental 53 and British Leyland 54 involved demanding high prices for the issuance of a certificate whose denial had the effect of delaying the registration of a car imported from another Member State, a special case. In United Brands, we heard that “charging a price which is excessive because it has no reasonable relation to the economic value of the product supplied . . . is an abuse”.55 This method of comparing costs and prices has been used in other cases, such as Deutsche Post,56 perhaps more confidently than in United Brands, an early case where the Article 82 analysis was perhaps tainted by the zeal to achieve market integration. Two very interesting cases arrived as references from national courts. First there was Bodson v Pompes Funèbres,57 concerning the grant by a municipality of exclusivity over certain funeral services. In that case, in order to determine the reasonableness of the price the Court envisaged a comparison by the national court between prices charged where concessions had been granted and prices charged in other municipalities where there was open competition. Shortly thereafter the Court decided Lucazeau v SACEM,58 where a significant difference in price between Member States was said to be an indication of an abuse which the enterprise should be in a position to justify on the basis of objective differences between the Member States concerned. 51 Case 22/78, Hugin Kassaregister AB and Hugin Cash Registers Ltd v Commission [1979] ECR 1869. 52 See Ian Forrester, “The Modernisation of EC Antitrust Policy: Compatibility, Efficiency, Legal Security”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, 2001, pp. 115 et seq. 53 Case 26/75, General Motors Continental NV v Commission [1975] ECR 1367. 54 Case 226/84, British Leyland plc v Commission [1986] ECR 3263. 55 Case 27/76, United Brands Co & United Brands Continental BV v Commission [1978] ECR 207, para. 250. 56 Commission Decision 2001/892 of 25 July 2001, Deutsche Post AG [2001] OJ L331/40. 57 Case 30/87, Bodson v Pompes Funèbres [1988] ECR 2479. 58 Joined Cases 110, 241 and 242/88, Lucazeau v SACEM [1989] ECR 2811.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 569

Sector-Specific Price Regulation or Antitrust Regulation

569

It is beyond the scope of this paper to give any comprehensive account of the competition cases on excessive pricing. I do submit, however, that the authorities are few and not easy to assemble into a single coherent body of law. It is very difficult to predict how a modern excessive pricing case could be brought in the abstract, without very helpful and pertinent comparative factors. That is not a criticism but rather a recognition of the very great difficulty a competition agency must experience in confronting conflicting opinions on whether a price is fair or unfair. Hence my uneasiness about the likely consequences of using a competition authority to intuit the appropriate level of a price.

Microsoft: a painful example of disagreement over pricing as a remedy I have personal knowledge of a celebrated controversy concerning a pricing remedy in an Article 82 case, Microsoft.59 The judgment of the Grand Chamber of the Court of First Instance in Case T-201/04 will be announced in the near future, and it will no doubt be long and interesting. The merits of the case will be argued for some years to come, in this venue and others. I have no wish to venture into those merits, other than to note some of the pricing and remedy difficulties encountered, as a source of guidance on pricing remedies. On 24 March 2004, the Commission issued a decision condemning Microsoft for having unlawfully exploited its near monopoly in the market for PC operating systems. The Commission imposed the highest fine ever imposed on a single company: 497 million euros.60 According to the Commission, Microsoft had abused its dominant position on the market for PC operating systems in two ways. First, by refusing to license certain “interoperability” technology to its competitors so they could develop “work group server” operating systems which would compete directly with Microsoft’s server operating systems (commonly called the “interoperability” or “refusal to supply” infringement). The infringement lay in not acceding in September 1998 to a request for technical divulgation from a competitor. Second, by bundling together two products which supposedly ought to have been offered separately, namely the PC operating system and an enhanced media player, Windows Media Player (commonly called the “product integration” or “tying” infringement). This infringement lay in bringing out an upgraded version of Windows that incorporated video streaming without offering a version of Windows that did not include that enhancement. 59 60

Case T-201/04, Microsoft v Commission, not yet decided as of this writing. See supra note 50.

(Z) Ehlermann 08 Forrester

570

6/11/08

11:54

Page 570

Session Three: Exploitative Conduct And The Remedies

As a remedy for the “tying” infringement, the company was ordered to prepare a special version of the Windows operating system without the functionality referred to as Windows Media Player, and then to launch it in commerce. This task was accomplished without a pricing controversy: the Commission had confirmed that Microsoft could offer the fully functioning version at the same price as the version without Windows Media Player. As to the interoperability abuse, Microsoft was obliged to draw up a “complete and accurate specification” of certain technology. This description of the characteristics of how Microsoft’s server operating systems provided certain services to other computers in a network would then be licensed to other companies who would write their own software in accordance with the specification. The goal of the licence was to endow competitors with the capacity to build software which could serve as a “drop-in replacement”, a perfectly operational alternative to Microsoft’s server operating system. The company entrusted the drawing up of the specification to several hundred engineers, many of whom had retired after creating the software in the early 1990s. There have been disagreements, irrelevant for present purposes, about the scope of the specification to be prepared and about what format should be adopted for the specification. As to the terms of disclosure, Article 5 of the Decision states: “(a) Microsoft Corporation shall, within 120 days of the date of notification of this Decision, make the Interoperability Information available to any undertaking having an interest in developing and distributing work group server operating system products and shall, on reasonable and non-discriminatory terms, allow the use of the Interoperability Information by such undertakings for the purpose of developing and distributing work group server operating system products;” (emphasis supplied)

Regarding the level of royalty, the Decision states: “The requirement for the terms imposed by Microsoft to be reasonable and nondiscriminatory applies in particular: [. . .] (ii) to any remuneration that Microsoft might charge for supply; such a remuneration should not reflect the ‘strategic value’ stemming from Microsoft’s market power . . .”61

These broad notions were then amplified by so-called “pricing principles” to which Microsoft acceded upon the Commission’s request in June 2005. The pricing principles were intended for use by the Trustee in resolving any dispute that might arise with a prospective licensee about the appropriate level of royalties for the specification.62 In the event of dispute, the Trustee could consider four factors:

61

See supra note 50, para. 1008. The WSPP Pricing Principles are available for download from the Microsoft website. See: http://www.microsoft.com/downloads/details.aspx?FamilyId=5A9CACB3-E823-4522-967CA6EE3CD3995D&displaylang=en. 62

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 571

Sector-Specific Price Regulation or Antitrust Regulation • • • •

571

whether the protocols represent Microsoft’s own creation; whether the creations by Microsoft constitute innovation; a market valuation of comparable technologies; and any other factors deemed appropriate by the Trustee.

Even though there have not been any disputes with licensees about royalty rates, the principles have given rise to a number of points of debate. For example, the Commission holds that it need consider market comparators only if the technology has been found to be innovative in the sense that it is superior to all other technologies on the market. In addition, it has been argued that no royalty should be paid for technology which solves problems particular to Microsoft software as opposed to problems of general applicability. Further, it has been argued that the price should be set at a level which permits viable competition between the licensee and Microsoft: on this theory, since the licensees could need as long as two years to implement the specification by building their own software, the price of the licence should— it is argued—be set low enough to make that effort worthwhile. The pricing principles were intended to guide the Trustee in the event of a specific dispute about what would be a reasonable royalty. No such dispute has arisen. Some ten licensees have agreed on the royalties they will pay without recourse to dispute resolution procedures. However, the principles are being used as an absolute over-arching standard against which to judge the royalty rates put up for negotiation by the company. There was sharp debate over the level of innovation reflected in the licensed technology. The company said that it had 73 patents either already granted to it or pending. The Trustee and the Commission’s experts, TAEUS, concluded that the technology embodied little innovation, and that only a few small pieces of the specification merited any compensation at all.63 PricewaterhouseCoopers supplied evidence to the effect that the royalty rates proposed by Microsoft were 30% below market norms for similar technology. Microsoft recalled that the proposed royalty rates were merely a starting point for negotiation, and that it was “open for business” in dealing with prospective licensees; and that in fact it had concluded licences with several companies. The Commission nonetheless has taken the view that the proposed maximum royalty rate of 5% was not “reasonable” since, according to the Commission, there was no “significant innovation” in the licensed material. Thus, the asserted infringing conduct consisted of proposing as a basis for negotiation a royalty rate of 5% of the licensee’s selling price for whatever product the licensee made based upon the licensed technology.64 63 See Press Release IP/07/269 of 1 March 2007: “Competition: Commission warns Microsoft of further penalties over unreasonable pricing as interoperability information lacks significant innovation”. 64 In August 2006, Microsoft submitted a proposal on licensing terms for the protocol specifications after several discussions, stating that Microsoft was “willing to entertain any reasonable price offer from any potential licensee, and that we are willing to be flexible to meet any unique

(Z) Ehlermann 08 Forrester

572

6/11/08

11:54

Page 572

Session Three: Exploitative Conduct And The Remedies

Indeed, the Commission’s position (adopted after the Decision of March 2004) is that the technology should be made available free of charge, or for a nominal fee, on a worldwide basis. (In a separate controversy, the Commission is demanding that the disclosure be effected on an Open Source basis, that is, without imposing any confidentiality constraints upon the licensee and thereby ensuring that Microsoft’s technology is placed in the public domain.65) The Commission’s view is that there is virtually no innovation in the 51 protocols (but that they are essential to remain viably in the market). According to the Trustee, “all of the described features were considered either to have been Microsoft implementations of prior developments by others, or to have been anticipated by prior developments and to be immediately obvious minor extensions to that prior work”. The company disagrees with these factual conclusions, pointing to the amount of time and effort deployed in creating the technology whose characteristics must be divulged in the licensed specification. Nevertheless (and now we reach the point relevant to this Workshop), the company said it was willing to charge lower royalties so as to avoid being penalized, and accordingly it asked what lower figures would be acceptable and not unreasonable or excessive. The Commission was unwilling to prescribe a figure, on the grounds that it is not a price regulator. The procedural twists of the Commission’s proceedings pursuant to Article 24 of Regulation 1/2003 do not for present purposes need to be described in detail. Suffice it to record that on 1 March 2007 the Commission issued a Statement of Objections, threatening a daily penalty, backdated to 16 December 2005, in an amount which was initially 500,000 euros per day, then 2,000,000 euros per day during 40 days ending 30 July 2006, then 3,000,000 euros per day from 31 July 2006 to date.66 As of this week (beginning 3 June 2007), the threatened penalty would total more than 1,100,000,000 euros. The company is thus threatened with the largest penalty in world competition law history for proposing as a basis for commercial negotiations a royalty rate which is higher than the public authority deems appropriate.

business needs of potential licensees”. See Microsoft Statement on European Commission Action on Protocol Pricing, available at http://www.microsoft.com/presspass/press/2007/mar07/ 03-01PricingProtocolPR.mspx. The relevant licensing arrangement may simplistically be outlined as follows. The four options are: (i) a licence for all intellectual property rights in the WSPP protocols and the protocol specifications (“All IP”); (ii) a licence limited to Microsoft’s patents on the WSPP protocols (“Patent Only”); (iii) a licence limited to Microsoft’s trade secrets disclosed in the protocol specifications (“Trade Secret Only”); and (iv) a licence limited to the IDL files (“IDL Only”). Microsoft further divided the protocols into Gold, Silver and Bronze price categories based on the degree of innovation. (A fourth category includes protocols, not necessarily innovative, for which there will be no royalty.) 65 This is the subject of Case T-313/05, Microsoft v Commission, removed from the register on 27 November 2007. 66 Press Release IP/07/269 of 1 March 2007.

(Z) Ehlermann 08 Forrester

6/11/08

11:54

Page 573

Sector-Specific Price Regulation or Antitrust Regulation

573

The company says its technology is the fruit of hundreds of person-years of effort and that it is covered by a number of patents from various jurisdictions and contains valuable know-how,67 so it is valuable.68 The Commission says that the technology is not innovative in the sense of being superior to other technologies, and that—to the extent innovation is present—the technology is oriented to solving problems unique to Microsoft technology. There are hundreds of pages of expert evidence on the licensed technology (which is itself thousands of pages long), on the pricing principles to apply, on the patents, on commercial norms, and on the parties’ negotiations. Despite all this material, the parties are far apart but locked in a painful embrace. The episode seems to be very regrettable. What can we infer from it for the purposes of this gathering? The dispute seems to confirm the limits of the ability of competition authorities to reach convincing decisions on very specific, very detailed questions of price-setting. It is impossible for any normal person to know whether a royalty of 9%, 5% or 0% is appropriate for the delivery of a mass of details about how to build a “drop-in replacement” for a large and hugely complex product. The authority understandably will not wish to be a price regulator, yet the company understandably wants to be told what figure above zero will not be regarded as excessive by reference to the level of creativity revealed in the compulsory divulgation. We must suppose that in a future case a dominant player, faced with a duty to license certain material by way of a remedy, may confront a huge discrepancy of opinion as to the value, utility and technical significance of that material. The company believes that its technology is valuable. The Commission thinks it ought to be delivered without charge or for a nominal fee. Complainants are vocal. From the target company’s perspective, it has the worst of both worlds. It is essentially subject to price regulation, but is not informed of what is the regulated price. It is exposed to hefty fines if it gets the figure wrong. Since it will often be impossible to guess what can be accepted as a “reasonable rate”—a necessarily rather subjective determination—the dominant company may be forced to set its royalty rate below a competitive market level in order to avoid the risk of a gigantic fine. The Commission may feel unable to name an acceptable royalty precisely because it refuses to see itself as a price regulator. Imposing or threatening to impose a fine of over 1,000,000,000 euros for proposing as a basis for negotiation a royalty rate which is deemed excessive 67 Roger Milgrim, “Commission Proposed Capital Punishment—By Definition—for Trade Secrets, A Uniquely Valuable IP Right”, 88 Journal of the Patent and Trademark Office Society, No. 11 (November 2006), at 919 et seq. 68 See Microsoft Statement on European Commission Action on Protocol Pricing, 1 March 2007, available at http://www.microsoft.com/presspass/press/2007/mar07/03-01PricingProtocolPR. mspx; and Brad Smith, News Conference Transcript Regarding European Commission Action on Protocol Pricing, 1 March 2007, available at http://www.microsoft.com/presspass/exec/bradsmith/ 03-01-07ECProtocol.mspx.

(Z) Ehlermann 08 Forrester

574

6/11/08

11:54

Page 574

Session Three: Exploitative Conduct And The Remedies

can hardly be interpreted as a sign that pricing remedies are apt for handling by competition authorities in a hotly-contested Article 82 case.69

Conclusions Article 82 is available in cases of both exploitative and exclusionary abuses. That duality of function may seem strange in light of modern antitrust theory in sophisticated jurisdictions, where for the moment it is unfashionable to pursue excessive pricing cases. Fashions can change. Article 82 is capable of being used improperly, or at least imprudently, with the goal of altering how a marketplace is working. Article 82 should be used for the repression of abuses, for the prohibition and punishment of serious economic crimes. I submit that there may be cases where the “crime”, objectively viewed, does not look like a foreseeable category of prohibited activity. The underlying motive of the public authority may be an understandable reluctance to let market forces remedy the problem. I have mentioned occasions where the competition concerns of the European Commission seem in effect to be pursuing sector-specific adjustments. I have personal experience of the very unsatisfactory outcome when a pricing remedy is used by a competition authority in a complex case where there is a mass of evidence about the price parameters which should be used. I began this paper with the expectation that sector-specific regulators were too focussed on narrow details and not on broad questions of competition, too permanent and prone to survive the proper end of their mission; and that competition enforcers were ill-equipped to achieve success in handling price-related remedies if these became controversial. My experience of the controversy last mentioned gives me the opinion that there is less risk of arbitrariness and passion if the matter is entrusted to a sector-specific regulator. Ex ante price regulation will be based on market practice and known market parameters. I therefore find myself endorsing the proposition that such a mechanism may be a less bad alternative than an Article 82 remedy. 69 On 27 February, 2008, the Commission imposed a fine of EUR 899 million on Microsoft for having proposed as a basis for negotiation an unreasonably high royalty level for the licence of trade secrets without the right to any patents. The technology thus offered for licence was deemed insufficiently innovative. (In an effort to identify a figure which would be regarded as not unacceptable by the Commission, a number of percentages including 0.5%, 1.0%, 0.65% and 0.4% were put forward, but the Commission declined to specify any royalty figure on the grounds that it was not a price regulator.) The controversies over the lawful amount of the royalties for categories of licence have now been concluded. The fine relates to a period ending in 2007, and is the object of Case T-167/08. I note that on 17 September 2007 the Grand Chamber of the CFI rendered judgment on the merits of the original Commission Decision of 24 March 2004, finding heavily in favour of the Commission.

(Za) Ehlermann 08 Blumenthal

6/11/08

11:54

Page 575

V William Blumenthal*

Discussant Comments on Exploitative Abuses under Article 82 EC June 2007

I am delighted to be able to join you for this morning’s session on “Exploitative Conduct and the Remedies—The Interface between Regulation, Antitrust, and Consumer Protection—Is Price Regulation the Answer?” I have been asked to serve as discussant for papers by Amelia Fletcher1 and Ian Forrester,2 who address possible policy responses to excessive pricing by dominant firms. In the course of my comments, I’ll also touch upon issues raised by the paper on related issues by Emil Paulis.3 In considering policy towards “excessive pricing” by dominant firms, I begin with this recognition: not all market failures can be effectively redressed through competition policy, and competition policy is not necessarily the right tool to wield against every form of market failure. In elaborating on this point and, more generally, on the appropriate relationship between competition policy and other regulatory tools, I often begin4 with the scholarly work of Stephen Breyer, now a Justice on our

* General Counsel, Federal Trade Commission, Washington, D.C. These views are those of the speaker and do not necessarily represent the position of the Federal Trade Commission or of any individual Commissioner. 1 Amelia Fletcher and Alina Jardine, “Towards an Appropriate Policy for Excessive Pricing”, in this Volume. 2 Ian S. Forrester, “A Plague on Both Your Houses”, in this Volume. 3 Emil Paulis, “Article 82 EC and Exploitative Conduct”, in this Volume. 4 See, e.g., William Blumenthal, “The Relationship between Competition Agencies and Other Units of Government¸ before International Seminar: Review of Anti-Monopoly Law” (19 May 2006), available at: http://www.ftc.gov/speeches/blumenthal/20060519Mofcom-ADBFinal.pdf. That speech, delivered at a multi-day seminar in Hangzhou to discuss China’s proposed AntiMonopoly Law, was directed at issues facing China as it seeks to adopt a competition law regime—things such as the allocation of responsibilities between competition authorities and sectoral regulators, the problems of state aid and regional preferences, and the tendency “to limit the agency’s jurisdiction by excluding certain industries or certain segments of the economy, often on grounds that those industries or segments are ill-suited for competition . . . because they are ‘natural monopolies’.” Ibid. at 16. That is, the issue in Hangzhou was that governments often adopt a public utility-style regulatory tool when a market-based competition tool would be adequate and preferable. Today’s panel is addressing the reciprocal problem. That is, the issue here in Florence is the interest of competition lawyers in adopting an antitrust-based tool when it may not be adequate and when some alternative regulatory mechanism might be preferable.

(Za) Ehlermann 08 Blumenthal

576

6/11/08

11:54

Page 576

Session Three: Exploitative Conduct And The Remedies

Supreme Court, on the issue of regulatory “matches” and “mismatches”.5 During the late 1970s and early 1980s, while a law professor at an earlier stage of his career, Justice Breyer developed a list of marketplace problems that might justify intervention and a separate list of possible regulatory tools, and he observed that certain tools were best suited to certain problems.6 Where regulation was unsuccessful, as was often the case in the United States during that era, the reason could often be traced to selection of the wrong tool for the particular problem. In Justice Breyer’s words, “regulatory failure sometimes means a failure to correctly match the tool to the problem at hand”.7 His list of marketplace problems includes natural monopoly, rent control, spillovers, information inadequacies, and moral hazard. The regulatory tools include cost-of-service ratemaking, nationalization, taxes, marketable rights, information disclosure, standard-setting, and antitrust. As practiced today by reasonably sophisticated governments, virtually all of those regulatory tools include competitive effects as an element of analysis, but they are not “competition policy” in the sense of antitrust. At the Federal Trade Commission, for example, we have a Bureau of Competition and a separate Bureau of Consumer Protection. Many of the interventions by the Bureau of Consumer Protection involve market failures arising from information inadequacies. In conducting our work relating to those interventions, we routinely involve economists in our Bureau of Economics, and we routinely assess the competitive effects of possible agency actions; but most of the remedies involve some form of information disclosure, as distinct from the forms of antitrust remedy ordinarily sought in matters brought by our Bureau of Competition. There is a tendency on the part of competition lawyers to view competition enforcement (in the sense of antitrust) as pure, and to view sectoral economic intervention as tainted. I accept that, at least directionally, but the distinction is not so clearly drawn as we competition lawyers commonly think. From the perspective of the economy as a whole, competition enforcement will generally be superior; it qualifies as the default regulatory tool. From the perspective of particular circumstances of market failure, however, the answer will be less clear. Competition enforcement may sometimes be inadequate. Some industries are inherently monopolies—they have room for only one player, due to either their cost structure or perhaps network effects. In those industries some form of utility-style regulation may be a better tool. The Fletcher and Forrester papers do an excellent job of collecting and summarizing the fundamental difficulties in using competition law to inter5 See, e.g., Stephen Breyer, Regulation and its Reform, Harvard University Press, 1982. The book elaborates on and develops views initially presented in a law review article that is often easier to locate today: Stephen Breyer, “Analyzing Regulatory Failure: Mismatches, Less Restrictive Alternatives, and Reform”, 92 Harvard Law Review 547 (1979). 6 See Breyer, Regulation and its Reform, cited in previous footnote, at 192. 7 Ibid. at 191.

(Za) Ehlermann 08 Blumenthal

6/11/08

11:54

Page 577

Discussant Comments on Exploitative Abuses under Article 82 EC

577

vene against high prices and other forms of exploitation of monopoly power. The Paulis paper, on which I am not a designated commentator, does as well. Taken together, the three papers provide as complete a list as I have seen in any single source. I do not try to reconstruct the list comprehensively here, but the key points seem to be these: • Considered in terms of the economic system as a whole, the opportunity to charge high prices and earn monopoly profits, at least for a short period, is desirable in that it attracts investment and business talent and yields innovation and growth. • Considered in terms of the particular market, high prices are a signal indicating that the market may currently be characterized by undersupply, and suppressing that signal will deprive the economy of warranted entry and capacity expansion. • Assessing whether a price is truly elevated is difficult for several reasons, which defy the articulation of clear legal rules and deprive the business community of needed guidance: —Identifying the benchmark against which price should be measured presents complex policy questions, —Insofar as the benchmark involves cost, measuring cost poses operational difficulties, and —For multi-product firms and for firms in multi-product markets, even measuring the appropriate “price” can pose operational difficulties. • The legal regime required to address exploitative prices is equivalent to price regulation and is highly distortive. • Intervention against exploitative prices challenges the institutional capabilities of a competition enforcement agency, a point to which I will return below. Because of these many difficulties, this morning’s papers uniformly realize that intervention against exploitation should be an exceptional use of Article 82, and probably an extraordinarily exceptional use. From an outsider’s perspective, it appears that the discussion within Europe is asking what are the narrow, focused circumstances—what are the extensive preconditions that must be satisfied—before intervention would be warranted. I would respectfully ask you to consider whether contemplating even limited intervention against exploitative abuses might be going too far. In expressing this view, I do not mean to rely on our Section 2, which, unlike your Article 82, declines to recognize exploitative abuse. Nor is the view based on differences in enforcement practices between our jurisdictions. While European and US enforcement officials have sometimes emphasized different factors and considerations in our respective speeches and policy statements, we all recognize the public benefit in permitting vigorous marketplace behavior, and actual instances of intervention have been infrequent on both sides of the Atlantic. I have been heartened at this Workshop to hear

(Za) Ehlermann 08 Blumenthal

578

6/11/08

11:54

Page 578

Session Three: Exploitative Conduct And The Remedies

numerous speakers note the role of private enforcement in the US; credit our government interventions, some of which are lower-profile and not prominently reported in the media;8 and generally acknowledge (contrary to what is sometimes said in Europe) that the US agencies have not abdicated enforcement in the field of unilateral conduct. No, in cautioning against even limited intervention by competition agencies against high prices, I am focusing here principally on considerations of institutional design, with an eye towards the match/mismatch analysis developed by Justice Breyer. Simply put, we need to question whether competition agencies have the competence to engage in classical price-and-profits utility-style regulation. As generalist agencies, we lack the right people. We lack the skill sets. We lack adequate industry expertise. Experience has shown that when we’ve tried to step into the role of utility-style regulators, we’ve bungled the task. There obviously are some commonalities between competition enforcement and classical regulation.9 And competition authorities necessarily must address price considerations in certain other contexts, wholly aside from exploitation. Elsewhere in the dominance field, price/cost relationships obviously matter for assessment of allegations of predation. For dominance cases arising from licensing practices, a benchmark price may be needed to accomplish a one-shot remedy.10 In the merger context, the divestiture price will 8 See, e.g., Biovail Corp., No. C-4060 (F.T.C. 23 April 2002) (complaint and consent order), available at: http://www.ftc.gov/os/caselist/c4060.shtm; Union Oil Co. of Cal., No. 9305 (F.T.C. Mar. 4, 2003) (complaint), available at: http://www.ftc.gov/os/caselist/d9305.shtm, and the related case in Chevron Corp., File No. 051-0125 (Aug. 2, 2005) (complaint and consent order), available at: http://www.ftc.gov/os/caselist/0510125/0510125.shtm; Bristol-Myers Squibb Co., No. C-4076 (F.T.C. 18 April 2003) (complaint and consent order), available at http://www.ftc. gov/os/caselist/c4076.shtm; Valassis Communications, Inc., No. C-4160 (F.T.C. 19 April 2006) (complaint and consent order), available at: http://www.ftc.gov/os/caselist/0510008/0510008. shtm; Rambus, Inc., No. 9302 (F.T.C. 2 August 2006) (opinion on liability), available at: http://www.ftc.gov/os/adjpro/d9302/index.shtm; Brief for the U.S. and FTC as Amici Curiae Supporting Plaintiffs-Appellants, In re DDAVP Direct Purchaser Antitrust Litigation, No. 065525-cv (2d Cir. filed 25 May 2007); and this administration’s continued activity in United States v. Microsoft Corp., Civ. Action No. 98-1232 (CKK) (D.D.C.); United States v. Dentsply Int’l, Inc., 399 F.3d 181 (3d Cir. 2005); and United States v. AMR Corp., 335 F.3d 1109 (10th Cir. 2003). The rate of unilateral interventions in the United States is roughly comparable to the rate in the late 1970s. See William E. Kovacic, “The Importance of History to the Design of Competition Policy Strategy: The Federal Trade Commission and Intellectual Property Law”, 30 Seattle University Law Review 319 (2007), available at: http://www.ftc.gov/speeches/kovacic/ 2007intersection.pdf. Kovacic collects statistics and concludes: (a) “Measured simply by the number of cases that allege the Sherman Act § 2 offenses of monopolization or attempted monopolization, the FTC’s enforcement actions over the past five years constitute the agency’s most ambitious program in roughly thirty years,” ibid. at 324–25 (footnotes omitted), and (b) “When the number of cases and the observable outcomes are both taken into account, the FTC’s program of monopolization and attempted monopolization cases since 2001 arguably has no parallel in the agency’s history,” ibid. at 326 (footnote omitted). 9 The two fields are combined, for example, in certain leading texts. See, e.g., W. Kip Viscusi, Joseph Harrington and John Vernon, Economics of Regulation and Antitrust, 4th ed., MIT Press, 2005. 10 See, e.g., Rambus, Inc., No. 9302 (F.T.C. Feb. 5, 2007) (opinion on remedy), available at http://www.ftc.gov/os/adjpro/d9302/index.shtm.

(Za) Ehlermann 08 Blumenthal

6/11/08

11:54

Page 579

Discussant Comments on Exploitative Abuses under Article 82 EC

579

sometimes be relevant to assessment of the adequacy of the proffered divestiture. So, too, will be the transfer prices for inputs or outputs to be sold between a divested business and a previously-integrated retained business. None of these circumstances, however, has the recurring and pervasive character of the regulation that will be needed to redress exploitative abuse in the form of excessive price. The problem is described with thought and candor in the Paulis paper: “[I]ntervening against excessive pricing may entail the risk of a competition authority finding itself in the situation of a semi-permanent quasi-regulator. The authority may have to come back time and again to the pricing of the dominant firm when cost or other considerations change in the industry, something a ‘generalist’ competition authority is much less equipped for than proper regulators with their deep knowledge of and continuous involvement in their industries.”11

Paulis finds “these practical difficulties so convincing and the risk of competition authorities arriving at the wrong result so great that enforcement actions against exploitative conduct . . . should be taken only as a last resort”.12 Some have suggested that competition authorities could develop the needed competencies and might be superior to classical regulators in intervening against excessive pricing in traditionally unregulated sectors. On this view, competition authorities would be more limited in their interventions. They would be more likely to look to market mechanisms before adopting more intrusive steps, they would be more sensitive to the distortions they were causing, and they would be more willing to recede from intervention after markets had corrected adequately. They also would be less susceptible to capture. All of these points have merit. But as a matter of institutional design, they come at an unacceptable cost. Developing the competencies within the competition agency and adopting a price-control mindset among the staff invariably will affect the character of the agency. It predictably will become more regulatory and less market-focused in its mainstream processes. If it is to succeed in the limited interventions that borrow from classical regulation, it is likely to sacrifice competence or at least judgment in the wider set of other interventions. And that is a cost I am not willing to pay. Let me conclude, then, with these observations: The availability of an effective remedy has been a recurring issue in the dominance field for many decades.13 In some instances, an effective, focused 11

Paulis, supra note 3, at 3. Ibid. 13 See, e.g., United States v. United Shoe Machinery Corp., 110 F. Supp. 295 (D. Mass. 1953), affirmed per curiam, 347 U.S. 521 (1954); United States v. Aluminum Co. of America, 91 F. Supp. 333 (S.D.N.Y. 1950). Cf. 4 Phillip E. Areeda, Antitrust Law: An Analysis of Antitrust Principles and Their Application ¶ 1402, at 9-10 (1986) (“unilateral behavior is not only omnipresent, but also often difficult to evaluate or remedy by any means short of governmental management of the enterprise”). 12

(Za) Ehlermann 08 Blumenthal

580

6/11/08

11:54

Page 580

Session Three: Exploitative Conduct And The Remedies

remedy will be feasible. It typically will address a particular restraint that can be excised through a prohibitory injunction. As the range of business conduct that must be addressed broadens, however, or if the injunction moves beyond prohibitions into affirmative obligations, the likelihood that a remedy will be successful becomes more remote. A remedy that entails ongoing regulation of prices and profits by courts or competition authorities is almost certain to fail, for the reasons of competence and resources noted above. Governments have a number of regulatory tools at their disposal for responding to perceived market defects. If a particular monopoly presents a problem that is so severe and intractable that enforcement officials believe the only effective remedy would entail ongoing monitoring and supervision of price, we should be asking whether a sectoral regulator with the appropriate competencies is available. And if none is, we should be asking whether the market failure is really of such a character that one should be constituted.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 617

I Marc van der Woude* Unfair and Excessive Prices in the Energy Sector June 2007

Introduction The concept of competition refers to a process whereby firms dispute the favour of their customers, ideally by proposing better products at the lowest possible price. This consumer welfare-creating process is to a large extent Darwinian in nature: those who cannot compete must die. The law of supply and demand is ruthless and so is competition law. Principles of fairness and justice are extraneous to competition law: the lion eats the deer. Still, EC competition law is not amoral. This is illustrated by the very first example given by Article 82 EC, according to which “directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions” can constitute an abuse. The concept of fairness as a constituent element of EC competition law may disturb those who analyze the competitive process through a scientific lens. Fairness does not relate to economic effects and cannot be measured or quantified. It is a concept that appeals to ethics and norms. The dichotomy between competition as an amoral, welfare-creating process that can be the subject of economic research, and the normative concept of fairness does not necessarily lead to a contradiction. Fairness becomes relevant where the competitive process has ceased to play its welfare-creating role, i.e., where monopolies prevail over perfect competition. When confronted by a monopoly, abstention is the customer’s only choice, and for some goods, such as food, clothing and housing, abstention is not considered a realistic choice. In scenarios where the monopolist faces an inelastic demand curve, fairness is probably the customer’s only safety buoy. The energy sector is one of the areas in which these kinds of situations occur. In our modern societies, customers expect that the light will turn on, when they flick the switch, and that their houses will be heated when they activate their boilers. As candles and jumpers do not offer realistic alternatives, demand is close to inelastic, at least in the short term. Moreover, the energy sector is characterized by the existence of monopolies not only as regards distribution and transport facilities which cannot easily be duplicated but also with respect to production and supply businesses. As shown by the * Erasmus University and Stibbe.

(Zc) Ehlermann 08 van der Woude

618

6/11/08

11:55

Page 618

Session Three: Exploitative Conduct And The Remedies

Commission’s report on the energy sector inquiry, these businesses remain heavily concentrated and are quasi-monopolistic in some Member States. In fact, the sector inquiry originated on the basis of complaints by large industrial users about rising prices. This contribution will focus on the possibilities to assess the compatibility of these high prices with EC competition law under Article 82(a) EC: when do energy prices become unfair within the meaning of this provision? This question will be addressed in five sections. The first section further elaborates upon the dichotomy referred to above. It reflects some personal views about the meaning of Article 82(a). The second section contains a brief overview of the interpretation given to Article 82(a) by courts and competition agencies. The third section tries to identify pricing issues in the energy sector which could possibly qualify for an assessment under Article 82(a). The fourth section deals with some recent price-related interventions by national competition authorities in the energy sector. The last section summarizes the findings of the paper.

I. Conceptual observations The origins of Article 82(a) EC The concept of dominance has been operative in Community law since 1952. Article 66 of the European Coal and Steel Treaty (signed in Paris on 18 April 1951 and effective for 50 years as from 24 July 1952) authorized the Commission to address recommendations to undertakings that held a dominant position shielding them against effective competition, if those undertakings used (not abused) that position for purposes contrary to the objectives of the Treaty. This provision was designed to control those undertakings which had already acquired a dominant position prior to the entry into force of the ECSC merger control rules or which had acquired such dominance through internal growth. According to Joliet, this power was “directed at the misuse of power (for instance through unilateral fixing of unduly high prices or monopolistic reduction of output)”.1 Article 82 has made the power to intervene against price practices more explicit. However, the founding documents of the EC Treaty apparently do not provide much guidance as to the reasons and economic ideas which led the authors to include this provision in the Treaty. David Gerber notes that the competition law provisions were constitutional in nature. Legal practice was supposed to give these provisions their normative content.2 1

René Joliet, Monopolization and abuse of dominant position, Martinus Nijhoff, 1970, p. 218. David Gerber, Law and competition in twentieth century Europe, Oxford University Press, 2001 (first published 1998), p. 345. 2

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 619

Unfair and Excessive Prices in the Energy Sector 619

Dominance and monopoly pricing The ECJ has systematically defined the concept of dominance by referring to the independence of the dominant firm vis-à-vis its competitors, customers and suppliers. This independence implies that the dominant firm is a price setter and not a price taker. The fact that it can set its prices as a function of its own cost considerations is indeed one of the features of a monopoly; the industry demand curve is its own demand curve. This means that monopoly pricing is inherent to the concept of dominance. It also means that unfair prices within the meaning Article 82(a) cannot logically correspond to monopoly prices in the sense just described. If this provision were to aim at monopoly pricing, it would implicitly condemn the dominant position which Article 82 assumes to exist. Joliet notes in this respect that a condemnation of monopoly pricing would imply that all pricing by dominant firms would necessarily be abusive.3 In other words, an unfair price should logically exceed a monopoly price. To introduce a neologism, an unfair price is a “supra-monopoly” price.

Dominance and unfairness As already mentioned above, fairness is a concept that cannot be measured in economic terms. It is a philosophical, ethical and normative notion which transcends economic analysis and which has a role to play in competition law where the market mechanism has ceased to function. As any normative principle, fairness implies a balance of interests. In this case, it is a balance between, on the one hand, the interest of the customer to have access to a certain good or service, and, on the other hand, the commercial interests of the dominant firm. It seems to me that several elements are relevant in striking this balance. Firstly, the degree of dominance should be such that the customer cannot realistically turn to alternative sources of supply. Nor should they be in a position to hope for market entry. The dominance should be monopolistic or quasi-monopolistic in nature and effectively protected by high barriers to entry. Or, as Article 66(7) ECSC puts it, the dominant position must shield the firm from effective competition. Secondly, the goods in question must be indispensable, i.e., they must be goods that customers cannot afford not to have. But what is indispensable? Obviously, goods satisfying basic objective human needs, such as food or medical care, are more likely to be indispensable for end consumers than luxury goods. It is interesting to note in this respect that a substantial part of the (national) case law on excessive pricing deals with pharmaceuticals. However, 3

Joliet, supra note 1, pp. 242–243.

(Zc) Ehlermann 08 van der Woude

620

6/11/08

11:55

Page 620

Session Three: Exploitative Conduct And The Remedies

a good or service can also be indispensable if it is a necessary input for a downstream activity or if it is an “essential facility”. The third element, related to the second, is inelasticity of demand. If customers cannot and will not turn to alternative sources of supply, the supplier will not lose sales volume when he increases his prices, even beyond monopoly price levels. In fact, it would be irrational for him not to increase his price, as doing so will enable him to substantially increase his turnover and profits. Fourthly, what is the conduct of the dominant firm over time? Monopolies can be temporary in nature. They can occur as a result of strikes, accidents or plant outages affecting the ability of competitors to supply the market. Customers may consider it abusive to take advantage of such “windfall” monopolies, but they also have the means to discipline the temporarily dominant firm once the market has returned to a competitive situation. If the market can sanction real or perceived injustice, there is no need to rely on public fairness provisions. Time is also relevant to assess supra-monopoly pricing by permanently dominant firms. If such pricing occurs only temporarily, the justification for outside intervention under Article 82(a) is less apparent than in a situation of systematic overcharging. Fifthly, one can never exclude the possibility of objective reasons which may explain supra-monopoly pricing. For example, exceptional investment plans could possibly offer such justifications.

Unfairness and other abuses; the subsidiary nature of Article 82(a) EC When assessing unreasonably high prices under Article 82(a), it is important to distinguish between two scenarios: one where such pricing conduct takes place in isolation and the other where that conduct forms part of a wider scheme or practice. The first scenario is unlikely to occur very often. Relying on the criteria developed above, it assumes that a series of exceptional conditions are fulfilled: systematic supra-monopoly pricing by a lasting monopolist supplying indispensable goods and facing an inelastic demand curve. As illustrated by the case law cited below, unreasonably high prices are more often part of a wider strategy of the dominant firm. They may be a means to raise funds in the monopolized market in view of financing predatory and/or discriminatory pricing policies in competitive markets. Overpricing may also occur to bar parallel trade if the monopoly concerns a service which is indispensable for that trade to occur. Or it may be a constituent element of an economic tying practice, encouraging customers to buy the cheaper bundle rather than its highly priced components. I submit that Article 82(a) should not apply in a scenario of combined restrictive practices. As already mentioned above, the concept of fairness is a

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 621

Unfair and Excessive Prices in the Energy Sector 621 notion of last resort. It should prevail only where the competitive process has (nearly) ceased to function. However, where a still functioning competitive process is threatened by a series of restrictive practices, antitrust enforcement should focus on these practices. The priority should lie with ensuring a system of undistorted competition within the meaning of Article 3(1)(g) EC. Article 82(a) is therefore of a subsidiary nature. It applies only as a means of last resort.

Proof of unfairness As discussed below, it is important to distinguish two elements when assessing unfair prices: first, the definition of unfair pricing; and second, the evidence demonstrating that the elements of that definition are met. Proof of unfair prices may be obtained by comparing the prices of the dominant firm to costs, or to prices charged by the dominant firm in other periods or markets, or to prices charged by comparable undertakings in the same market or in different markets. It should be noted, however, that such differences are elements of the evidence showing that the conditions of the definition are met, but that they do not coincide with that definition. For example, the difference between prices and costs is not unfair; it is just an element that may point to the existence of unfair prices.

Unfairness and enforcement There is also a practical reason to support the view described above. The application of Article 82(a) is a delicate and complicated task. As a normative concept, fairness cannot be quantified. Something is unfair or it is not unfair, but it will be hard or impossible to determine a criterion as to when a price starts or ceases to be unfair. The cursor can be placed on many different points along a wide spectrum. In addition, the circumstances which led the administrative or judicial body to consider that the prices were unfair may change over time. These findings raise questions as to the enforceability of Article 82(a). A cease and desist order may have a repressive effect, but it will offer limited guidance for the future. Its future effect will at best be equivalent to a maximum price regulation for the dominant firm.4 Even so, Regulation 1/20035 4 See François Lévêque, “Antitrust Enforcement in the Electricity and Gas Industries: Problems and Solutions for the EU”, 19(5) The Electricity Journal 27 (2006) (“It must not become an instrument for fixing administered prices. It would be ironic for the pursuit of liberalization of the electricity and gas sectors to lead to return to regulated prices in power generation.”). 5 Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, [2003] OJ L1/1.

(Zc) Ehlermann 08 van der Woude

622

6/11/08

11:55

Page 622

Session Three: Exploitative Conduct And The Remedies

offers both the Commission and national competition authorities more enforcement options. The dominant firm can avoid repressive action by proposing commitments. These commitments may not only contain a code of conduct as to future pricing, they can also contribute to a structural solution of the supra-monopoly pricing problem. Finally, if that problem persists, Article 7 of Regulation 1/2003 offers the Commission the means to impose structural remedies, which may include divestitures.

Conclusions The Treaty and the founding documents do not offer guidance as to the interpretation of Article 82(a). Since the Treaty assumes the existence of dominance and, hence, of monopolies, Article 82(a) should logically not prohibit monopoly prices. Conceptually, unfair prices are supra-monopolistic prices. The conditions under which supra-monopolistic pricing can be considered unfair are exceptional: the claimant or enforcement authority should have to show that the dominant firm enjoys a lasting (quasi-)monopolistic position in respect of indispensable goods, and that it faces an inelastic demand curve and systematically charges prices exceeding the monopoly price. Given the exceptional nature of such a scenario, and given the need to ensure undistorted competition, enforcement action under Article 82 should primarily focus on practices that harm a competitive process that still exists. Article 82(a) should be triggered only after that process has ceased to play its welfare creating role. As regards enforcement action, structural solutions contributing to the revival of that process should be preferred. Let us now compare this proposed test with the case law of the ECJ and the decisional practice of the Commission and national authorities.

II. Interpretation by the institutions a) The case law of the Court of Justice The early case law The notion of unfair prices entered Community competition law in the Court’s case law of the 1960s and early 1970s, which dealt with the level of royalties for IP rights.6 In Sirena v Eda, the ECJ held that: “As regards the 6 In Case 24/67, Parke Davis and Co. v Probel, Reese, Beintema-Interpharm and Centrafarm, [1968] ECR 55, the ECJ paved the way for this reasoning by determining that “although the sale

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 623

Unfair and Excessive Prices in the Energy Sector 623 abuse of a dominant position, although the price level of the product may not of itself necessarily suffice to disclose such an abuse, it may however, if unjustified by any objective criteria, and if it is particularly high, be a determining factor.” 7 In General Motors, the Commission used the expression “excessive” prices for the first time, suggesting that an excessive price would be unfair. Although the ECJ annulled the decision, it upheld the proposition that under Article 82 an abuse “might lie, inter alia, in the imposition of a price which is excessive in relation to the economic value of the service provided”.8

United Brands: an orientation on costs The ECJ reiterated and specified this principle in United Brands by ruling that “charging a price which is excessive because it has no reasonable relation to the economic value of the product supplied would be [an abuse under Article 82]”.9 Ever since, United Brands has been the main judicial precedent for the application of Article 82 to unfair pricing. However, in that case the ECJ did not specifically set out how the economic value of a product should be determined. The judgment describes a two-stage test for determining whether a price is reasonably related to the economic value of the product supplied: “[t]his excess could, inter alia, be determined objectively [. . .] by making a comparison between the selling price of the product in question and its cost of production, which would disclose the amount of the profit margin. The questions therefore to be determined are whether the difference between the costs actually incurred and the price actually charged is excessive, and, if the answer to this question is in the affirmative, whether a price has been imposed which is either unfair in itself or when compared to competing products.”10

This two-limb test corresponds to the test proposed in section I above. The first part of the Court’s test could be seen as a means to identify the existence of supra-monopoly pricing, whereas the second part of the test concerns the questions whether this excessive price is unfair. Comparisons with prices of competing products may offer guidance to determine unfairness. The case law is not clear as to whether a cost/price analysis is always required to identify the existence of unfair prices. One of the grounds which price of the protected product may be regarded as a factor to be taken into account in determining the possible existence of an abuse, a higher price for the patented product as compared with the unpatented product does not necessarily constitute an abuse”. 7 Case 40/70, Sirena S.r.l. v Eda S.r.l. and others [1971] ECR 69. 8 Case 26/75, General Motors Continental NV v Commission of the European Communities [1975] ECR 1367. 9 Case 27/76, United Brands v Commission [1978] ECR 207, para. 250. 10 Ibid., paras. 251–252.

(Zc) Ehlermann 08 van der Woude

624

6/11/08

11:55

Page 624

Session Three: Exploitative Conduct And The Remedies

led the ECJ to annul the Commission’s decision in United Brands concerns the fact that the Commission had failed to analyze UBC’s cost structure. This suggests that mere price comparisons do not suffice to establish the existence of unfair prices.

SACEM: comparisons In the SACEM cases, however, the ECJ focused on comparisons rather than on cost/price analysis. In Lucazeau v Sacem, the ECJ held that “When an undertaking holding a dominant position imposes scales of fees which are appreciably higher than those charged in other Member States and where a comparison of the fee levels has been made on a consistent basis, that difference must be regarded as indicative of an abuse of a dominant position. In such a case it is for the undertaking in question to justify the difference by reference to objective dissimilarities between the situation in the Member State concerned and the situation prevailing in all the other Member States.”11

Nevertheless, it should be noted that the SACEM cases dealt with the level of royalties for music in discothèques, and that a cost/price analysis for artistic creations is hard to carry out. Moreover, the ECJ was relatively cynical as regards the cost and profit levels put forward by SACEM to justify the price differences. “It is apparent from the documents before the Court that one of the most marked differences between the copyright-management societies in the various Member States lies in the level of operating expenses. Where—as appears to be the case here, according to the record of the proceedings before the national court—the staff of a management society is much larger than that of its counterparts in other Member States and, moreover, the proportion of receipts taken up by collection, administration and distribution expenses rather than by payments to copyright holders is considerably higher, the possibility cannot be ruled out that it is precisely the lack of competition on the market in question that accounts for the heavy burden of administration and hence the high level of royalties.”12

I interpret this paragraph as an implicit criticism of the relevance of a cost/price analysis for establishing the existence of unfair prices. Within the limits set by the shareholder, a monopolist is indeed free to determine its costs and profits. The absence of profits does not offer conclusive evidence as regards the absence of unfair prices. In Bodson v Pompes Funèbres, the ECJ also preferred a comparative approach to a cost/price analysis of the firms’ own cost structure.13 The comparison focused on prices charged by other concession holders: 11 Joined Cases 110/88, 241/88 and 242/88, François Lucazeau and others v SACEM and others [1989] ECR 2811, para. 25. 12 Ibid., para. 29. 13 Case 30/87, Corinne Bodson v Pompes funèbres des régions libérées [1988] ECR 2479, para. 31.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 625

Unfair and Excessive Prices in the Energy Sector 625 “The French Government and PFRL have denied that the prices charged by the subsidiaries of Pompes Funèbres Générales are unfair. The documents before the Court do not contain any information enabling that problem to be resolved. Since over 30 000 communes in France have not granted to an undertaking the concession to provide ‘external services’ for funerals, but have left that service unregulated or operate it themselves, it must be possible to make a comparison between the prices charged by the group of undertakings which hold concessions and prices charged elsewhere. Such a comparison could provide a basis for assessing whether or not the prices charged by the concession holders are fair.”

b) The Commission’s decisional practice Apart from its decisions in General Motors14 and British Leyland,15 which essentially dealt with parallel trade rather than unfair pricing issues, the Commission has never condemned unfair pricing practices. It is not unreasonable to argue that its latest decisions discourage more than encourage enforcement under Article 82(a). The Euromax v IMAX decision16 can be seen as an implicit rejection of price comparisons as a means to identify unfair pricing. In that case, the Commission rejected a complaint brought against the allegedly excessive rental fees charged for the IMAX system. It held that price comparisons with competitors were inadequate in this case because “[t]he differences in type of contracts and performances compared are . . . too great to allow a sufficiently qualified and appropriate comparison”. Moreover, Euromax had not provided a consistent comparison with other products, that is to say, it had failed to point to “a comparison that involves the same products with similar quality and functionality and subject to the same terms and conditions. This is the standard, which is required by the Court of Justice.”17 In short, the complainant failed to identify the adequate benchmark. In the parallel Scandlines case,18 the Commission is equally sceptical as regards cost/price analysis to establish the existence of unfair pricing. The case concerned port charges in the port of Helsingborg, which Scandlines considered to be unfair. In a relatively long decision, the Commission rejected Scandlines’ complaint. First, the Commission explained the difficulties in finding out which costs of Helsingborg’s integrated port activities could be 14

Supra note 8. Case 226/84, British Leyland Plc v Commission [1986] ECR 3263. 16 Commission Decision of 25 March 2004, Euromax v IMAX, Case COMP/37.761, available at http://ec.europa.eu/comm/competition/antitrust/cases/index/by_nr_75.html#i37_761. 17 Ibid. 18 Commission Decision of 23 July 2004, Scandlines Sverige AB v Port of Helsingborg, Case COMP/36.568, available at http://ec.europa.eu/comm/competition/antitrust/cases/index/by_nr_ 73.html. 15

(Zc) Ehlermann 08 van der Woude

626

6/11/08

11:55

Page 626

Session Three: Exploitative Conduct And The Remedies

allocated to the accommodation required for the ferry line to Denmark. Second, once it had identified the relevant costs, the Commission specified that “[t]he fact that the port charges would be non-cost based or the pricing non transparent do not constitute as such as abuses under Article 82 of the EC Treaty.” According to the Commission, more is required: “It is important to note that the decisive test in United Brands focuses on the price charged, and its relation to the economic value of the product. While a comparison of prices and costs, which reveals the profit margin, of a particular company may serve as a first step in the analysis (if at all possible to calculate), this in itself cannot be conclusive as regards the existence of an abuse under Article 82 EC Treaty.”19 “. . . even if it were to be assumed that the profit margin . . . is high (or even ‘excessive’), this would not be sufficient to conclude that the price charged bears no reasonable relation to the economic value of the services provided. The Commission would have to proceed to the second question as set out by the Court in United Brands, in order to determine whether the prices charged to [. . .] are unfair, either in themselves or when compared to other ports.”20

Third, the Commission stressed in Scandlines that demand-side considerations should also be considered in the assessment of the economic value of the product or service. Customers may be prepared to pay a premium for the unique service offered by the Helsingborg port. Fourth, when assessing the economic value of the service, due consideration should also be given to the need to recover large initial investments, the intangible value of the product itself and any opportunity costs. Based on all the above considerations, the Commission concluded that: “[i]n the present case, the economic value of the product/service cannot simply be determined by adding to the approximate costs incurred in the provision of this product/service as assessed by the Commission, a profit margin which would be a pre-determined percentage of the production costs. The economic value must be determined with regards to the particular circumstances of the case and take into account also non-cost related factors such as the demand for the product/service [(i.e. the valuation by the customers and consumers of the product/service)] [or, e.g., costs of capital].”21 “As a consequence, finding a positive difference between the price and the approximate production costs exceeding what Scandlines claims as being a reasonable margin, would not necessarily lead to the conclusion that the price is unfair, provided that this price has a reasonable relation to the economic value of the product/service supplied.”22

It is not unreasonable to submit that the Euromax and Scandlines decisions do not reflect a great degree of enthusiasm to deal with excessive or unfair 19 20 21 22

Ibid., para. 102. Ibid., para. 158. Ibid., para. 232. Ibid., para. 233.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 627

Unfair and Excessive Prices in the Energy Sector 627 pricing issues and that future action in this area is probably not high on the Commission’s priority list.

c) National decisions: the Napp case In contrast with the Commission, national authorities have been less reluctant to interfere against unfair prices.23 A substantial part of these interventions concerns pharmaceutical products. Both in Germany and in The Netherlands, the regulators sought to control vitamin and valium prices on the basis of competition law in the late 1970s. Finding the right standard for assessing unfairness was also the key issue in these cases. German courts rely on market comparisons as a first method to spot excessive prices, but they require proper adjustments to take account of possible differences between the markets in question. The ruling of the British Competition Appeals Tribunal (CAT) in the Napp case24 probably offers one of the most elaborate decisions on unfair prices at the national level, although the pricing conduct here formed part of a wider abuse. The case concerned Napp, a pharmaceutical company with market shares above 90% in a market shielded by high barriers to entry. Relying on United Brands, the UK’s Office of Fair Trading (OFT) found that Napp had charged prices exceeding those that would have prevailed on a competitive market and that there was no actual or potential competitive pressure to bring these prices down to competitive levels within a reasonable period. These high prices in the “community segment” of the market (i.e., where the product was sold through pharmacies) served, inter alia, to finance predatory rebates in the hospital segment.25 In order to establish the difference between Napp’s prices and the prices which could have prevailed in a competitive environment, the OFT relied on various comparisons: (i) between Napp’s prices and own costs; (ii) between Napp’s prices and the costs of its most profitable competitor; (iii) between Napp’s prices and those of its competitors; and (iv) between the prices charged by Napp in the relevant market and the prices it charged in other markets. It resulted from these comparisons that:

23 The Dutch authority has been particularly active in the field of abusive pricing issues. See Erik H. Pijnacker Hordijk and Yvo De Vries, “Onbillijk hoge prijzen als vorm van misbruik van een economische machtspositie onder het Europese en het Nederlandse mededingingsrecht”, 2002 SEW 430. 24 Competition Appeals Tribunal, judgment of 15 January 2002, Napp Pharmaceuticals Holdings Limited and Subsidiaries and Director General of Fair Trading, Case No. 1001/1/1/01. 25 This means that the Napp case is not a pure unfair pricing case. Compare Commission Decision 2001/892 of 25 July 2001, Deutsche Post AG (“Deutsche Post II”) [2001] OJ L331/40.

(Zc) Ehlermann 08 van der Woude

628

6/11/08

11:55

Page 628

Session Three: Exploitative Conduct And The Remedies

• “Napp’s prices in the community segment are typically around . . . [between 30 to 50] per cent higher than its competitors”; • “Napp’s price in the community segment has remained the same since the launch of MST in 1980, . . ., notwithstanding the expiry of its formulation patent in 1992”; • “Napp’s list price (less wholesale discount) in the community segment of the market is on average over 1400 per cent higher than its price in the hospital segment of the market for . . . tablets, where Napp faces competition”; • “Napp’s highest level of discount, the list price in the community segment is on some tablets over 2000 per cent higher than Napp’s hospital prices”; • “Napp’s prices in the community segment are over 500 per cent higher than its prices for export on a contract manufacture basis”; • “Napp’s gross profit margin on sales to the community segment is . . . [in excess of 80] per cent, compared with a margin of around . . . [between 30 and 50] per cent on Napp’s other products sold to the NHS”; and that • “Napp’s gross profit margin of . . . [in excess of 80] per cent on sales to the community segment compares with a gross profit margin of . . . [less than 70] per cent for Napp’s next most profitable competitor”. In its appeal against the OFT’s decision, Napp argued that the relevant time frame to assess its pricing policy should correspond to the lifetime of the product and that the profits made on MST were needed to finance R&D costs for and losses on other drugs in its product portfolio. The CAT did not share these views and rejected the appeal. It considered that launching and promotion costs had been recouped a long time ago, that prices should be assessed in each relevant market, and that there were no objective reasons to price excessively for a period of twenty years, and for more than ten years after the expiry of Napp’s product patent. In taking this decision, the CAT stated that it was not bound to indicate precisely the price level that would be regarded as excessive.

d) Conclusions It follows from this overview that unfair pricing cases are exceptional and that most precedents are negative in the sense that they do not find sustainable grounds on which to ascertain unfair prices. Enforcement activity is more important at national than at Community level. At both levels, United Brands is the main precedent. The two-prong test developed by the ECJ in United Brands corresponds to the idea of “supra-monopoly” pricing developed in section I above; monopoly profits alone do not suffice to establish unfair pricing. Conversely, SACEM shows that the absence of monopoly profits does not necessarily constitute proof of the absence of unfair pricing. Comparisons in time, between

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 629

Unfair and Excessive Prices in the Energy Sector 629 geographic and product markets and between competitors can offer other evidence. Whatever the test used by the authority, it will be difficult to apply. “The fact that the exercise may be difficult is not, however, a reason for not attempting it”, as stated by the CAT in Napp.26 The Napp ruling offers a convincing example of a combination of converging parameters that reveals the existence of unfair pricing in the normative sense as proposed in section I above. The case meets nearly all the proposed criteria. It is now time to examine how these principles are to be applied, more specifically, in the energy sector.

III. Issues in the energy sector a) The sector inquiry Significant rises in gas and electricity wholesale prices and persistent complaints about barriers to entry and limited consumer choice led the Commission to open an inquiry into the functioning of the European gas and electricity sectors in June 2005. On 10 January 2007, the Commission presented its final Report.27 The conclusions were not really surprising: despite the liberalisation process in these sectors, there is no integrated European energy market and indeed there are few well-functioning national markets. Although high prices prompted the energy inquiry, the conclusions of the Report pay relatively minor attention to pricing as a stand-alone issue. The section on gas mostly deals with pricing patterns and the indexation of gas prices to oil prices. The electricity chapter also discusses interaction of electricity prices and primary fuel prices. Both chapters allude to the pricedistorting effects of public price regulation. The section of the report on price formation only refers to possibly unfair pricing conduct in respect of so-called “windfall” profits”, i.e. where electricity producers include the market prices of freely obtained CO2 certificates in their electricity sales prices. The Report refers in this respect to the pending case before the German FCO, which is briefly discussed below.28 26

Napp, supra note 24, at para. 392. The four-part Report is available at: http://ec.europa.eu/comm/competition/sectors/energy/ inquiry/index.html. For a summary of the Report, see Communication from the Commission— Inquiry pursuant to Article 17 of Regulation (EC) No 1/2003 into the European gas and electricity sectors (Final Report), available at http://eur-lex.europa.eu/LexUriServ/site/en/com/2006/ com2006_0851en01.pdf. 28 Press release of the Bundeskartellamt concerning a possible abuse of a collective dominant position by incumbents E.ON and RWE (20 December 2006). The English-language press release is available at http://www.bundeskartellamt.de/wEnglisch/News/Archiv/ArchivNews2006/2006_ 12_20.php. 27

(Zc) Ehlermann 08 van der Woude

630

6/11/08

11:55

Page 630

Session Three: Exploitative Conduct And The Remedies

It is not difficult to grasp why excessive pricing is not a major theme in the Report. The sector inquiry focuses on the causes underlying the relatively high price levels in the gas and electricity markets: market concentration, vertical foreclosure, market integration and insufficient transparency. This structural approach explains why high prices are dealt with in the section of the Report dealing with market concentration. Competition rules can be applied to remedy some of these underlying problems, for example by challenging long-term exclusivity arrangements or a discriminatory use of infrastructure. As regards Article 82, in my view the Report identifies three areas for future enforcement action: (i) non-transparent use of network infrastructure leading to congestion problems; (ii) long-term contracts concluded by incumbents with gas and electricity producers, foreclosing access to essential inputs; and (iii) long-term contracts concluded by incumbents with major customers, foreclosing access to downstream markets. These three options for enforcement efforts primarily concern exclusionary practices and not exploitative conduct within the meaning of Article 82(a). This makes sense. Combating exclusionary practices contributes to making markets work, and if markets work unfair prices should logically not arise. As proposed under Section I above, Article 82(a) is only of a subsidiary nature. It enters into play when the market mechanism has ceased to play its welfarecreating role. The findings of the sector inquiry are worrying, but they do not justify the conclusion that markets in these sectors do not work at all. Although future enforcement action will primarily address exclusionary practices, there are two issues which may possibly qualify for an examination under Article 82(a) and which have already been the subject matter of enforcement action at national level (see Section IV below).

b) Networks The first issue concerns the use of transport and distribution networks. These facilities are often regarded as natural monopolies in the sense that their duplication is not justified in economic terms. They are also regarded as essential facilities because access is required to enter downstream supply markets. These features and their ownership by vertically integrated companies have led the Community and national legislators to regulate the access conditions of these networks.29 This sector-specific regulation has a dual aim. Its first objective is to ensure equal access by third parties. In other words, tariffs and other access conditions should not be discriminatory. Its second objective concerns the subject 29 See Marc van der Woude, “Competing Gas Pipeline Infrastructure”, in International Energy Law and Taxation Review, Sweet and Maxwell, 2005, pp. 12 et seq.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 631

Unfair and Excessive Prices in the Energy Sector 631 matter of this contribution: tariffs should also be “proportionate”.30 Even completely unbundled network operators can still have an incentive to charge disproportionate tariffs for access to their network. Tariff regulation has become an inherent element of the network operating business. It can take various shapes, but nearly all of its manifestations are cost-oriented. The network operator is allowed to obtain a reasonable return (often in a 5% to 7% range) on his regulatory asset base. In addition, some Member States have chosen models that try to simulate competition by cutting admissible cost by an efficiency reduction factor (also called the “X-factor”) or by benchmark methods comparing different network operators. The Dutch system even went as far as imposing turnover regulation instead of tariff regulation, until the Dutch economic appeals court annulled the methodology on the simple ground that it no longer dealt with tariffs.31 All these cost-oriented tariff regulations seek to constrain price-setting by the network operator and, hence, to avoid monopoly pricing; the operator’s cost curve is supposed to intersect the demand curve at a lower point than the point that would have prevailed in the absence of tariff regulation. This type of tariff regulation is considerably stricter than the fairness test imposed by Article 82(a). As mentioned above in section I and confirmed by the ECJ in United Brands, unfair prices are prices that exceed monopoly prices. The rationale of Article 82(a) does not relate to efficiency or cost cutting, but to fairness. This finding makes it difficult to rely on Article 82(a) as an alternative to ex ante tariff regulation.

c) Wholesale markets Marginal cost pricing in the electricity sector Price formation for gas and electricity differs. Prices for gas correspond to commodity prices.32 Simplifying, one could say that competition rules, including Article 82(a), can be applied to gas in the same way as in any other commodity industry. By contrast, electricity price formation is complex.33 Market prices are set as a function of the marginal costs of the last plant to be dispatched. Paragraph 370 of the sector inquiry Report notes the following: “the SRMC 30 See Article 23(4) of Directive 2003/54/EC of the European Parliament and of the Council of 26 June 2003 concerning common rules for the internal market in electricity and repealing Directive 96/92/EC, [2003] OJ L176/37. 31 Trade and Industry Appeals Tribunal (CBb), judgment of 30 November 2006, GTS v. NMa. 32 The sector inquiry Report (supra note 27) also contains a study of the price hike on the UK gas markets. The study did not reveal the existence of abusive pricing practices. 33 The Report deals with various markets: generation, forward markets and spot markets. This section will focus on generation.

(Zc) Ehlermann 08 van der Woude

632

6/11/08

11:55

Page 632

Session Three: Exploitative Conduct And The Remedies

[short-run marginal cost] of the price setting unit determines not only the revenues of the owner of the marginal plant, but also of all other operators with e.g. nuclear, lignite or run-of-river units”. This marginal cost pricing method generates considerable revenues for these other operators, even if they did not set the price. In the same paragraph, the Report notes that “whilst their marginal costs are often significantly lower, it is generally argued that they need a higher price than the marginal costs to recover the higher fixed costs associated with base load generation” (footnote omitted). The functioning of wholesale markets is further complicated by their high degree of concentration. In some countries, the previous monopolist still enjoys market shares in the 90% to 100% range. Lack of interconnection capacity and illiquidity of wholesale markets are considerable entry barriers that contribute to the protection of those positions. According to the Report, the combination of high concentration and lack of liquidity enables dominant generators to manipulate wholesale prices either by withdrawing capacity or “by imposing high prices when they know that their production is indispensable to meet demand”.34 In the latter scenario, “it is possible to raise prices (‘excessive pricing’) even with a relatively small portfolio because [of] the structure of the generation assets and indispensability of certain assets to meet demand at parts of the merit curve, or in certain locations in the network. The higher the concentration in the relevant parts of the merit curve concerned the greater is the scope for influencing prices . . .”.35

The difficulty of applying Article 82(a) Seen from this angle, the wholesale markets in these countries seem to be obvious candidates for scrutiny under Article 82(a). There are several factors, however, that complicate the analysis. These factors concern the system described above, involving marginal cost pricing according to which the latest dispatched plant sets the market price. This system implies that large market shares are not necessarily indicative of market power. The sector inquiry revealed “the possibility that companies with a limited share in generation capacity might have market power at certain moments”.36 Market power is determined, inter alia, by the diversity of a generator’s production portfolio. An operator owning most of the marginal plants in the merit order of production plants is more likely to be the market’s price-setting firm than a larger operator that owns base load plants located more at the “left hand side” of the merit order. It would be a bizarre claim to accuse the price-setting firm of excessive pricing, if it only prices in terms 34 35 36

Report, supra note 27, para. 403. Ibid., para. 405 (footnote omitted). See also para. 428. Ibid., para. 408

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 633

Unfair and Excessive Prices in the Energy Sector 633 of its marginal costs. The price-setting firm is unlikely to make “excessive” profits. High profits are more likely to be found with the owners of the base load plants, which can sell their electricity at a high market price significantly exceeding the marginal costs of their production facilities. But these owners are not necessarily the price-setting firms and, even where they are, the additional revenues may be needed, as mentioned above, to cover their fixed costs and their need to invest in new capacity. The need to recover fixed costs and to ensure investment possibilities raises the fundamental question as regards the time frame under which pricing practices should be assessed. A rigorous or ruthless application of Article 82(a) aimed at combating short- or medium-term pricing, may deprive the accused firms of their ability to invest and hence may exacerbate the lack of liquidity. Investments should be taken into consideration as regards both the firm’s individual investment requirements and the industry investment cycle.

Price-setting frequency and volumes sold at clearing price Even so, the sector inquiry Report offers some interesting materials to assess market power in the generation markets and, possibly, to overcome the difficulties identified above. Paragraphs 429 to 436 combine two elements to establish whether a firm is dominant in the market. The first element deals with the price setting frequency: how often does an operator set the clearing price? “Hypothetically, if only one operator ‘sets the price’ most of the time, it means that there are very few, if any, alternative offers around the clearing price most of the time. The operator builds up knowledge about the inelasticity of demand on a specific part of the supply curve where he operates by comparing his bids with the exchange clearing price. If demand is very inelastic, he can increase the price without the risk (or with little risk) of being replaced by another operator.”37 The Report contains quantitative data about such price-setting frequencies. The second element concerns the volumes by the price-setting operator at the clearing price. Here again, the Report provides an overview of markets where the largest price setter controlled more than 50% of the electricity offers at a price around the clearing price. A table at paragraph 431 shows that the price setter in some markets accounts for 100% of electricity sold at the clearing price.

37

Ibid., para. 429.

(Zc) Ehlermann 08 van der Woude

634

6/11/08

11:55

Page 634

Session Three: Exploitative Conduct And The Remedies

Unfair pricing? The combination of these two elements overcomes some, but not all difficulties referred to above. The frequency element ensures that price-setting by the firm under scrutiny is not occasional and the volume conditions imply that the advantages of marginal prices also accrue to its base load operations at the left hand side of the merit curve: price-setting and profits coincide within the same firm. Still, this does not suffice to establish the existence of unfair prices. Marginal cost pricing is an intrinsic feature of electricity generation markets. Why would such pricing suddenly become abusive if it is applied by a frequent price setter that may be selling large volumes but which follows the same market conduct as that prevailing under competitive conditions? In addition, even if one were to assume that the price-setting firm makes additional profits on its plants at the left end side of the merit order, the argument that marginal cost pricing does not suffice to cover total long-term costs of and investments for these plants remains a valid one. Relying on the test developed under section I above, the application of Article 82(a) requires additional elements. Firstly, there should be no other means to intervene against the pricing problem. The sector inquiry Report refers, for example, to the possibility that some price-setting firms will push prices up by withdrawing capacity. In that event, Article 82 should apply to the withdrawal policy as a form of outputreducing behaviour. Preference should be given to fighting causes rather than symptoms. Secondly, price-setting should occur at a level that exceeds marginal cost pricing. The Report suggests that this situation can arise where the unique position of the price-setting firm has allowed it to gain specific knowledge about the demand curve and where it has used this knowledge to push prices above normal marginal cost prices. In other words, prices should be supramonopolistic in nature. Thirdly, the product should be indispensable, and there should be an inelastic demand curve for it. In the electricity sector, these two conditions are likely to be fulfilled Fourthly, one could also impose a requirement according to which the price-setting firm must be systematically pushing prices above monopoly levels without incurring the risk of stimulating market entry. The conduct should not be accidental. All these conditions are admittedly hard to meet and even harder to prove. A cocktail of cost/price analyses and various comparisons, comparable to the approach followed by the OFT in Napp,38 will probably offer the only means to meet that burden of proof. 38

See supra note 24 and accompanying discussion.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 635

Unfair and Excessive Prices in the Energy Sector 635

IV. National precedents a) Ex post network regulation In the absence of ex ante tariff regulation in the electricity and gas sectors by an energy-specific regulator, the German Federal Cartel Office (FCO) has dealt with tariff regulation on an ex post basis. It intervened in several instances against excessive transmission and distribution charges.39 In the Stadtwerke Mainz case,40 it compared the tariffs of the Stadtwerke Mainz to the tariffs of a comparable city distribution company, RWE Net, and assessed whether any differences could be explained by objective factors such as differences in costs, the structure of the grid, geographical features, the number of connected customers, economies of scale and reliability. On the basis of this comparison, the FCO found that the tariffs of Stadtwerke needed to be lowered by 19.9% to suppress their abusive nature. On 7 February 2006, Germany’s Supreme Civil Court confirmed the legality of this type of ex post tariff regulation in the Strom II plus case.41 The court noted that a voluntary framework agreement, called Strom II Plus, which lays down the cost calculation for network access pricing, did not exclude the possibility that the prices charged were excessive. It referred the case back to a lower court, requiring it to apply German sector-specific legislation and competition law. It also referred to Article 82 as a provision that the lower court should take into consideration. The FCO intervened in a similar manner against network operators on other aspects of their business, such as metering conditions and prices of balancing power.42 The Dutch authorities (NMa) intervened informally against a system of compensation for transformation losses applied by Essent in respect of windmill owners.43 Since the Dutch system is subject to detailed tariff regulation, the intervention of the NMa can be seen as complementary in nature. It is not clear whether the German precedents are relevant for unfair pricing issues under Article 82 EC. This is because the provisions in the German Act against Restraints of Competition (GWB) seem stricter than Article 82, at least insofar as unfair pricing is concerned, because they prohibit conduct 39

TEAG, FCO 19.02.2003. Decision of the Bundeskartellamt of 21 March 2003, Stadwerke Mainz AG. See the English-language press release at http://www.bundeskartellamt.de/wEnglisch/News/Archiv/ ArchivNews2003/2003_04_17.php. 41 Judgment of the Bundesgerichtshof of 7 February 2006, Strom II plus, KZR 8/05. 42 See http://www.bundeskartellamt.de/wEnglisch/News/Archiv/ArchivNews2003/2003_02_ 26_II.php. 43 See Nma, Press release 04-22 of 30 September 2004 (in Dutch). 40

(Zc) Ehlermann 08 van der Woude

636

6/11/08

11:55

Page 636

Session Three: Exploitative Conduct And The Remedies

that is not necessarily unfair.44 Article 19(4) sub 2 GWB prohibits a dominant firm from demanding “payments or business terms which differ from those which would very likely arise if competition existed; in that context, particularly the conduct of undertakings in comparable markets where effective competition prevails shall be taken into account”.

b) Wholesale markets Spanish temporary congestion cases The Spanish competition authorities decided that four electricity producers had abused the monopolistic positions which each of them held in their respective service areas for three consecutive days when “congestion problems” prevented a normal functioning of the Spanish electricity market.45 During these three days, each vertically integrated producer applied unusually high prices. These prices considerably exceeded variable costs, were significantly higher than the prices charged in comparable situations of technical constraints, and exceeded peak prices charged during the previous twelve months. Even if the Tribunal implicitly admitted that the prices corresponded to rational commercial conduct, it considered that each producer had abused its individual dominant position and imposed a fine of 900,000 euros on each of them. The ruling of the Tribunal primarily reflects a normative point of view; the four producers unduly took advantage of the temporary congestion problem on the Spanish grid. The producers should have priced as they would have done in the absence of that problem. The sanction therefore seems to focus more on an abuse of unexpected advantage rather than on the excessive nature of the prices as such.

44 Article 3(2) of Regulation 1/2003 (supra note 5) authorizes Member States to maintain stricter prohibitions for unilateral conduct. 45 Judgment of the Tribunal de Defensa de la Competencia of 7 July 2004, Expte. 552/02, Empresas eléctricas, available at http://www.noticias.info/archivo/2004/200407/20040710/ 20040710_28644.shtm. My Spanish is insufficient to understand all nuances of the case. The OMEL daily market does not accept prices which are too high. These prices are sold under a “technical restraints” regime. The markets resulting from these regimes are local in nature: due to capacity problems on the transport grid, local users must buy from local producers This gives rise to local dominant positions. See also Viesgo Generacion, 602/05, 28.12.2006.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 637

Unfair and Excessive Prices in the Energy Sector 637

The Elsam case On 30 November 2005, the Danish Competition Council took a decision against unfair prices charged by Elsam A/S.46 The reference period was considerably longer than the period examined by the Spanish authorities, i.e., from 1 July 2003 until 31 December 2004. During this period Elsam is supposed to have abused its position on the wholesale OTC market in Western Denmark. The Competition Council found that this market constituted the relevant market despite price interaction on the spot market between Western Denmark and other parts of the Nord Pool area. This is because the available price data were already affected by Elsam’s excessive prices. Bottlenecks on the electricity grid isolate Western Denmark from the rest of the country. Within this area Elsam enjoyed a dominant position, controlling 90% of the market. Decentralized production and willmills did not constrain its market power. As regards the assessment of Elsam’s prices, the Competition Council applied the United Brands test described above.47 It determined that these prices exceeded average total costs plus a mark up which was calculated as a function of the average rate of return in the industry. This part of the test revealed that prices were excessive. As regards the second part of the test, the Council compared the prices charged during the relevant period with prices charged during other periods. This showed that Elsam had been charging unfair prices for 900 hours. Elsam appealed against this decision before the Competition Appeals Court. On 14 November 2006, the Appeals Court confirmed the Council’s substantive assessment but annulled the price control order which the Council had imposed as a remedy.48 So far as I understand the case, the Danish authorities applied a very rigid interpretation of the United Brands test. Their objections concern monopoly pricing during a period of congestion. They do not address the question as to whether that monopoly pricing was unfair in itself or in comparison to competing products.

ENEL price manipulation The Italian energy regulator and competition authority intervened in 2005 and 2006 against ENEL, the electricity incumbent. They accused ENEL of price manipulation. The publicly available documents do not describe the nature of the practices in question, which seem to boil down to a cocktail of 46 Decision of the Danish Competition Council of 30 November 2005, Elsam A/S, available at http://www.ks.dk/english/competition/national-decisions/national-decisions-2005/2005-11-30elsam-as-abuse-of-dominant-position-by-charging-unfair-prices. 47 See supra note 10 and accompanying text. 48 Judgment of the Danish Competition Appeal Tribunal of 14 November 2006, Elsam A/S.

(Zc) Ehlermann 08 van der Woude

638

6/11/08

11:55

Page 638

Session Three: Exploitative Conduct And The Remedies

repression of and compensation for competitors in view of imposing ENEL’s market leadership. These practices do not include unfair pricing in the sense discussed above. Even so, the case is interesting because of the remedies offered by ENEL. The company and its subsidiary offered to auction 700 MW per year in order to increase liquidity in certain markets.49

CO2 prices In 2006 the FCO initiated proceedings against E.ON and RWE, because these companies had increased their prices by including a cost component relating to CO2 emission rights which these companies had obtained for free. The FCO considers that these two companies hold a collective dominant position on the national energy markets and that they abused that dominance in breach of Article 19(4) GWB, quoted above in Section IV a). Competitive conditions in other Member States do not allow electricity producers to pass on opportunity costs of freely obtained emission rights. Curiously, the FCO does not object to 25% of these costs being passed on, but it considers that any pass-on above that percentage is abusive. Since the case is in a preliminary stage, it is not clear whether the FCO will take a formal decision against E.ON and RWE. I am sceptical for two reasons. To my knowledge, market prices of freely obtained CO2 emission rights are also passed on in electricity prices in countries which are not necessarily characterized by dominant positions.50 Is it unfair for a dominant firm to apply the same pricing practice as the one which would have prevailed in competitive conditions? In addition, does the increase generate windfall profits or does it anticipate windfall costs? The pass-on some may seem unfair when assessed as an isolated event, but may be less so when assessed over a longer period, as suggested in Section I above.

German electricity prices Large, electricity-intensive industrial users are concerned throughout Europe about rising energy prices. These users often compete on worldwide markets, unlike electricity producers, which reside in local markets without effective competitive constraints. This mismatch may lead to relocations of energyintensive users outside Europe, i.e., in countries with regulated energy markets and with less CO2 concerns. These concerns also affect the EU’s manufacturing powerhouse. In an attempt to assist large users and to comfort 49 AGCM Decision 16250 of 20 December 2006, Case A366—Comportamenti restrittivi sulla borsa elettrica, Boll. 49/2006. 50 See Frontier Economics, “CO2 trading and its influence on electricity markets”, Final report for DTE, February 2006.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 639

Unfair and Excessive Prices in the Energy Sector 639 smaller one such as households, the German Ministry of Economic Affairs has presented a draft act amending the GWB as regards excessive prices in the energy industry.51 Article 29 of the proposed act prohibits an undertaking that holds, individually or collectively, a dominant position as supplier of electricity, gas or heat (supply company) from abusing that position, where it imposes tariffs or trading conditions that are more disadvantageous than those of other supply companies or companies operating on comparable markets, even if the difference is not considerable, or where it imposes tariffs that disproportionably exceed costs. The prohibition does not apply if the dominant supply firm is able to demonstrate that its prices are commercially justified. However, costs or cost components that would not have been incurred to the same extent under competitive conditions cannot be taken into consideration when determining the existence of an abuse within the meaning of the proposed Article 29. This prohibition as it is contemplated in the draft act is very strict and considerably lightens the burden of proof on the FCO. Under the proposal, prices can be abusive either when they exceed, even slightly, the price level of benchmark undertakings or when they are disproportionate in comparison to the costs of the dominant firm. In either scenario, the prohibition applies and the burden of proof shifts from the FCO to the dominant firm to justify those prices. In addition, inefficient monopoly costs of the type to which the ECJ alluded in the SACEM cases may not be taken into consideration to that effect.52 Moreover, the proposed prohibition considerably widens the choice for the FCO to find benchmark companies: these may be energy supply companies in other markets but they may even be supply companies in non-energy markets. The German Ministry of Economic Affairs has argued that the proposed prohibition corresponds to or specifies the prohibition laid down in Article 82(a). I do not think this is entirely correct. Article 29 as drafted is not only procedural in nature in that it eases the burden of proof for the regulator, but it also imposes a substantive norm that prohibits pricing that would not be caught by Article 82(a). Under the latter provision, charging higher prices than one’s competitors, or making monopoly profits, is not unfair as such. The test proposed above in Section I, and the case law of the ECJ, require a wider combination of elements to make the normative judgment that a dominant firm is pricing unfairly. More generally, the objectives pursued by the draft Article 29 are hard to reconcile with the subsidiary nature of Article 82(a) and may even be counterproductive.53 The proposed provision reflects 51 See, e.g., “Wettbewerb auf dem Strommarkt”, Orientierungen zur Wirtschafts- und Gesellschaftspolitik, Ludwig Erhard Stiftung (December 2006). 52 See supra note 12 and accompanying text. 53 The Monopolkommission issued a negative opinion on the proposed act.

(Zc) Ehlermann 08 van der Woude

640

6/11/08

11:55

Page 640

Session Three: Exploitative Conduct And The Remedies

distrust in the competitive process. It may discourage investments and, hence, may contribute to reducing liquidity. In the end, it may exacerbate the problem it seeks to cure.

V. Concluding remarks It was a widespread political belief that the opening of energy markets to competition would lead to lower prices. This is a fallacy: prices merely reflect the interplay between supply and demand. The market mechanism does not guarantee absolute price levels. Moreover, the sector inherited its monopolistic structure from the past. Legal liberalization cannot set economic realties aside. Still, confronted by political impatience and dissatisfaction, competition regulators and authorities are called upon to intervene. In this context, Article 82(a) and equivalent provisions of national law have attracted new attention. However, Article 82(a) should not be relied on too hastily. This prohibition is of a subsidiary nature and should be applied only if competition has ceased to function. Competition law interventions should primarily focus on practices harming the competitive process and on the causes of relatively high prices, rather than on the high prices themselves. It is only when the dominant firm is no longer subject to competitive or other market constraints that Article 82(a) should come into play. This is when norms and ethics offer the only recourse against the dominant firm. These exceptional circumstances may be present in the European energy sector, which is still characterized by the presence of quasi-monopolies in many markets and by relatively high entry barriers. Gas and electricity are indispensable in our modern societies. Demand for these products is often inelastic, particularly in the electricity sector. This inelastic demand allows incumbent monopolies to lift prices beyond the point where their marginal cost curve intersects the demand curve. Such supra-monopoly pricing is in my view unfair, within the meaning of Article 82(a), if it is a persistent feature of the dominant firm’s pricing conduct. It is not an easy task to find the evidence showing that all these conditions are met. There will rarely be one conclusive factor. The application of Article 82(a) will most probably require the convergence of various indicators, as shown by the CAT in Napp.54 There are few convincing precedents dealing with unfair pricing in the energy sector. A large part of these precedents concern ex post tariff regulation as a substitute for ex ante regulation. However, unfair pricing within the 54

Supra note 24 and accompanying discussion.

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 641

Unfair and Excessive Prices in the Energy Sector 641 meaning of Article 82(a) is a different concept than efficiency-driven tariff regulation. Article 82(a) does not allow the regulator to compress the costs of the network operator. The precedents relating to wholesale markets are not convincing either. The Spanish and Danish cases discussed in Section IV b) concern price increases limited in time. Moreover, the Danish authorities seem to consider that prices above a cost-plus level are excessive and therefore unfair. It follows from United Brands that such above-cost pricing is not necessarily unfair. Nor is the approach of the ECJ in United Brands accurately reflected in the proposed German act on abusive pricing in the energy sector. Article 82(a) is and will remain difficult to apply. It is also dangerous to apply. Overly rigorous enforcement can negatively affect the investment climate. Finally, I would like to conclude my contribution by expressing some general concerns about competition in the energy sector. I have doubts as to the welfare-creating role of competition in that industry. Is competition the best method to achieve Europe’s environmental and geopolitical objectives? What is the welfare-creating role of downstream competition when upstream gas prices are set by public authorities or companies as a function of political imperatives rather than of economic considerations? What’s the point of imposing competition in the electricity industry where demand is inelastic and where the supply curve reflects a rigid merit order?

(Zc) Ehlermann 08 van der Woude

6/11/08

11:55

Page 642

(Zd) Ehlermann 08 Siragusa

6/11/08

11:55

Page 643

II Mario Siragusa* Excessive Prices in Energy Markets: Some Unorthodox Thoughts August 2007

I. Introduction Rising prices are no less a challenge for the regulator than for consumers. In order to respond to high roaming charges, the European Parliament, on 23 May 2007, voted in favour of a Regulation to set roaming prices.1 I think that the Eurotariff represents one end in the spectrum of solutions available under competition law to deal with the problems posed by high prices. What I find less explored is the other end of the spectrum of solutions to the high prices themselves, i.e., structural remedies. In what follows below, I would like to offer a few thoughts on this issue in the hope of contributing to the discussion on excessive prices in energy markets.

II. The Mandate of Volvo/Veng Volvo/Veng is rarely cited, if at all, as a case of excessive pricing.2 Even in the very recent judgment of the Court of First Instance in Der Grüne Punkt,3 which deals with excessive pricing issues, the court confined itself to citing the “standard” line of cases on excessive pricing, i.e., General Motors, United Brands and British Leyland.4 * Partner, Cleary Gottlieb Steen Hamilton, Rome. 1 See Press Release IP/07/696 and text of the Regulation at www.europarl.europa.eu. The Regulation was approved, and it entered into force on 30 June 2007. See Regulation (EC) No 717/2007 of the European Parliament and of the Council of 27 June 2007 on roaming on public mobile telephone networks within the Community and amending Directive 2002/21/EC, [2007] OJ L171/32. 2 Case 238/87, AB Volvo v Erik Veng (UK) Ltd., [1988] ECR 6211. 3 Case T-151/01, Der Grüne Punkt—Duales System Deutschland GmbH v Commission, judgment of the Court of First Instance of 24 May 2007, not yet reported, para. 121. On appeal: Case C-385/07, not yet decided. 4 Case 26/75, General Motors Continental NV v Commission [1975] ECR 1367; Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207; Case 226/84, British Leyland Plc v Commission [1986] ECR 3263.

(Zd) Ehlermann 08 Siragusa

644

6/11/08

11:55

Page 644

Session Three: Exploitative Conduct And The Remedies

The omission of Volvo/Veng from the progeny of cases on excessive pricing can be explained by the fact that the case did not provide any concrete indication as to the issues that are more typical of excessive pricing cases, and in particular, how to determine an “excessive” price. Yet I find the lesson of Volvo/Veng particularly important, and I believe that its remit is of great significance to our discussion. In dealing with the right of a proprietor of a protected design to prevent third parties from manufacturing and selling or importing, without his consent, products incorporating the design, the Court stated that the exercise of that right: “may be prohibited by Article 86 if it involves, on the part of an undertaking holding a dominant position, certain abusive conduct such as the arbitrary refusal to supply spare parts to independent repairers, the fixing of prices for spare parts at an unfair level or a decision no longer to produce spare parts for a particular model even though many cars of that model are still in circulation . . .” (emphasis supplied).5

The force of the Court’s statement should not go unnoticed. In this case, the right in question is, by the admission of the Court, “the very subjectmatter of his exclusive right”.6 Yet the Court considered unfair pricing to be a violation of such gravity that is deserved what may be called “capital punishment”, i.e., the end of the exclusive right itself. In my view, the lesson of Volvo/Veng is very relevant to our debate because it shows that, first of all, unfair pricing holds such a high rank in the hierarchy of antitrust offences that it should not be so quickly written off in favour of exclusionary abuses.7 Indeed, the Court in Volvo/Veng listed it alongside the classic examples of exclusionary conduct. Second, Volvo/Veng also shows that unfair pricing can and should be addressed not so much by bringing the price to what can be considered a just or fair level, but rather by going to the root of the problem. In other words, unfair pricing should be dealt with by attacking the cause of the defendant’s market power, of which high prices are a symptom. Of course, I concur with what has been said before about the care that should be exercised in dealing with issues of high prices. Marc van der Woude has emphasized that no less than five conditions should be satisfied before high prices may be brought as cases of excessive pricing under the competition rules.8 Similarly, Emil Paulis has highlighted the importance of very high and long lasting barriers to entry and expansion.9 Once these conditions are 5

Volvo/Veng, supra note 2, para. 9. Ibid., para. 7. 7 See Bruce Lyons, “The Paradox of the Exclusion of Exploitative Abuse”, Centre for Competition Policy Newsletter, Spring 2007. 8 See van der Woude, this Volume. 9 See Paulis, this Volume. 6

(Zd) Ehlermann 08 Siragusa

6/11/08

11:55

Page 645

Excessive Prices in Energy Markets: Unorthodox Thoughts

645

present, I argue, the main issue becomes which remedial antitrust action to take. On this point, I must say that ever since Volvo/Veng, I rarely find arguments suggesting that one should go to the root of the problem. Like the Commission with its roaming Regulation, I have the impression that often there is a preference for a “quick fix” in the belief that all that can be expected under antitrust law is that the dominant firm will correct its pricing. Amelia Fletcher and Alina Jardine have started to lay the foundation for an alternative approach.10 They call for “demand side” solutions. I believe that this change of attitude is to be welcomed, but we need bolder solutions for the energy markets. I now turn to this issue.

III. High Prices and Structural Antitrust Remedies in the Energy Sector There is little reason to be apprehensive about structural remedies under antitrust law in the domain of regulated industries. Let us take a look at our colleagues across the Atlantic. Those who may think about Trinko, and a sort of reluctance in US antitrust law to tamper with regulated sectors, should be reminded that the most stunning revolution in US telecoms was mandated not by any act of Congress but instead by the AT&T consent decree.11 Thus, at the very origin of the industry, which in the wake of Trinko is now enjoying some sort of antitrust immunity, we find a conspicuous structural remedy imposed by a federal judge. The AT&T case should serve to reinvigorate competition authorities and prompt them to think about structural remedies for excessive prices. If all of the conditions are indeed met for excessive prices to be considered an abuse, and if therefore (and this requirement should be emphasized) one observes a persistent pattern of very high and long lasting barriers to entry and expansion in which a single firm charges prices that are conspicuously above costs for a long period of time, I would then argue that the firm should not simply be met with merely an order to return to “normal” pricing, with all the difficulties of determining whatever that may be. As has been a long-standing policy in merger cases, the use of structural remedies should be the preferred option for uprooting the problem of excessive pricing under antitrust law. The choice in favour of structural remedies would also be consistent with the respective roles of sector-specific regulators and competition authorities. One of the reasons for the reluctance to enforce unfair pricing under Article 10 11

See Fletcher and Jardine, this Volume. MCI Communications Corp v. AT&T, 708 F.2d 1081 (7th Cir. 1983).

(Zd) Ehlermann 08 Siragusa

646

6/11/08

11:55

Page 646

Session Three: Exploitative Conduct And The Remedies

82 is the fear that the competition authority may become entangled with the day-to-day monitoring of pricing behaviour. That fear is steeped in the conviction that the “right prices” are the remedy to excessive pricing offences. If one were to choose the structural remedies option, that fear would clearly be seen as being misguided. The regulator would remain in charge of price setting, while the antitrust authority would be responsible for reforming market structure. In fact, structural remedies may usefully supplement the arsenal of tools given to the regulator. One cannot forget that, while neither the electronic communications nor the energy regulatory framework provides for structural remedies, Article 7 of Regulation 1/2003 does so, and it contains terse language to that effect.12 Having said that, let me turn to two instances of high prices in the energy sector where structural remedies may be an appropriate antitrust response. At the risk of oversimplification, discussions about high prices in the energy sector normally concern one of the following two matters: (i) prices for access to networks, and/or (ii) prices in the wholesale markets.

1. High Prices for Access to Networks I do not think that I need to spend too much time dwelling on this issue. The problems of unbundling have been discussed at length by the Commission in connection with the sector inquiry and by stakeholders in its aftermath. I observe that some of the national cases discussed as excessive pricing often concern precisely network pricing (i.e., the price for access to a network component).13 As we all know, the Council has chosen to defer unbundling in the hope that energy markets will improve. That may also be a sensible solution from an antitrust point of view. As noted, excessive pricing cases should be brought only in the face of the most compelling evidence, including, inter alia, proof of a systematic pattern of behaviour over time, whereas in certain cases, access to energy networks is a recent development. Here, “the quick fix” may be more appropriate in the short term. 12 According to Article 7, “[w]here the Commission, acting on a complaint or on its own initiative, finds that there is an infringement of Article 81 or of Article 82 of the Treaty, it may by decision require the undertakings and associations of undertakings concerned to bring such infringement to an end. For this purpose, it may impose on them any behavioural or structural remedies which are proportionate to the infringement committed and necessary to bring the infringement effectively to an end. Structural remedies can only be imposed either where there is no equally effective behavioural remedy or where any equally effective behavioural remedy would be more burdensome for the undertaking concerned than the structural remedy. If the Commission has a legitimate interest in doing so, it may also find that an infringement has been committed in the past.” (emphasis supplied). 13 For instance, the 2003 Stadtwerke Mainz decision by the FCO and the Strom II judgment by Germany’s supreme civil court (see Marc van der Woude, “Unfair and Excessive Prices in the Energy Sector”, this Volume) concern transmission and distribution charges.

(Zd) Ehlermann 08 Siragusa

6/11/08

11:55

Page 647

Excessive Prices in Energy Markets: Unorthodox Thoughts

647

However, it is important to bear mind the confusion that excessive pricing cases—particularly those resulting in orders to apply “the right price”—create as regards the allocation of competences between competition authorities and sector-specific regulators. In the end, it may even be less costly for an undertaking to accept a structural remedy. Here I would recall the British experience of Openreach, which I think is instructive of how a structural solution can be embraced by the undertaking concerned in lieu of a complicated review process based on behavioural commitments. Thus, structural solutions should remain on the agenda for high network prices, even for the sake of protecting the interests of those undertakings which control the networks.

2. High wholesale prices The second type of high prices are high wholesale prices. Evidence of these types of shortcomings are documented in both the sector inquiry,14 and in the study published on DG Comp’s website on 20 April 2007.15 High wholesale prices very often translate into high prices for final customers. As such, they are a very visible item on the agenda of national governments. It is difficult for competition authorities to resist the temptation of going for the “quick fix” in excessive pricing cases and ordering companies to behave in a particular way. Indeed, who would trust the “invisible hand” with structural measures which may or may not work, which may take time to deliver, and which moreover may earn you a powerful enemy, i.e., the incumbent energy company? In spite of these concerns, however, I think that structural measures deserve credit as a remedy to high wholesale prices. First of all, the “quick fix” may work in the short term to appease customers, but there is no guarantee as to its effectiveness in the medium to long term. Second, “quick fix” measures may be quickly ordered, but they certainly take time to carry out. That is to say that antitrust action takes place after the damage to the market has already been done. This is compounded by the fact that, given the difficulty of proving an excessive pricing case, decisions may come only after a significant lag following the alleged offence.16 Third, I agree with Marc van 14

See http://ec.europa.eu/comm/competition/sectors/energy/inquiry/index.html. See Press Release IP/07/522: “The European Commission has published a detailed study carried out by an external consultant which finds that fuel costs have contributed to the increase of EU electricity prices since 2003, but that wholesale electricity prices are significantly higher than would be expected on perfectly competitive markets. The differences are highest when only a few generators with available capacity are needed to meet demand, especially at peak time. The results of the study broadly support the conclusions of the Commission’s Final Report of the Energy Sector Competition Inquiry (see IP/07/26 and MEMO/07/15), namely that competition in EU wholesale electricity markets is not yet functioning properly.” 16 In Italy, the Enel case on wholesale prices in the pool was initiated in April 2005 and was closed in December 2006 when the company offered commitments. A ”typical” case could last longer since, in this case, the Italian Antitrust Authority did not have to prove an infringement. See AGCM 15

(Zd) Ehlermann 08 Siragusa

648

6/11/08

11:55

Page 648

Session Three: Exploitative Conduct And The Remedies

der Woude and François Lévèque that it would be ironic if energy liberalization were to end in regulated prices. It is well known that the Commission has criticized Member States for using tariffs to protect energy intensive customers from high prices. It would seem odd if antitrust action resulted in much the same type of measure. By contrast, I think that structural remedies have advantages as a solution to excessive pricing. First of all, they can more happily co-exist with regulation. Structural remedies are one-off measures that leave the regulators’ remit intact. In fact, such remedies do not replace regulatory intervention but accompany it, and they can add to what a regulator can do.17 Second, structural remedies in the energy sector do not necessarily result in an irreversible situation. Through Virtual Power Plant (VPP) or gas release programmes, the incumbent’s market power can be temporarily reduced. The incumbent retains ownership of the assets concerned, and these assets can then be returned to him once competition has taken off. Third, structural remedies are more in tune with the liberalization effort, as they are based on the same market logic that is the premise on which liberalization legislation is built.

IV. Conclusion In my intervention, I have mentioned the Eurotariff several times with a hint of criticism. Let me be more clear about it. The Eurotariff may be odd in a liberalized market. But it certainly deserves praise. It takes no small dose of courage to tackle telecoms companies in such a head-on way. Thus, the Eurotariff shows that the Commission can muster the courage that it takes to pass bold measures in the face of strong opposition. What deserves less praise is inertia. We have seen too much of that in the energy sector.18 There are hints that change may be imminent. However, for Decision 16250 of 20 December 2006, Case A366—Comportamenti restrittivi sulla borsa elettrica, Boll. 49/2006. 17 See Otter Tail Power Co. v. United States, 410 U.S. 366 (1973). Although it did not entail a structural remedy, Otter Tail is a good example of an antitrust remedy in the energy sector which in fact enlarged the remit of the regulator. In that case, the US Supreme Court held that it was illegal monopolization under Section 2 of the Sherman Act for a large electricity utility to refuse to transmit power as part of an attempt by the utility to stop other providers from competing with it in the sale of electricity to final customers. The Court imposed an obligation to transmit such electricity, and it empowered the regulator to oversee compliance. 18 See, e.g., Concetta Cultrera, “Les décisions GDF—La Commission est formelle: les clauses de restriction territoriale dans les contrats de gaz violent l’article 81”, Competition Policy Newsletter, No. 1, 2005. As noted by Cultrera, the Commission’s decisions addressed to GDF and ENI “sont très intéressantes à maints égards: elles sont les premières décisions formelles adoptées par la Commission depuis une décennie dans le secteur de l’énergie et viennent confirmer, après un certain nombre d’affaires concernant les clauses de restriction territoriale clôturées par règlement amiable, que ces clauses violent l’article 81 du traité.” (emphasis supplied).

(Zd) Ehlermann 08 Siragusa

6/11/08

11:55

Page 649

Excessive Prices in Energy Markets: Unorthodox Thoughts

649

over a decade, the Commission has maintained conspicuous silence with respect to this sector, relying on informal action. That has left a significant enforcement gap. It would be a pity if, after all this time, all we could get was an energy version of the Eurotariff.

(Zd) Ehlermann 08 Siragusa

6/11/08

11:55

Page 650

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 651

III Pierre Régibeau* The (Complex?) Relationship between Art 82(a) and Intellectual Property Rights July 2007

I. Introduction Article 82(a) of the EC Treaty prohibits the abuse, on the part of one or more undertakings, of a dominant position by “directly or indirectly imposing unfair purchase or selling price or other unfair trading conditions”. In spite of the reference to possible “indirect” channels, the provision has mostly been interpreted quite narrowly as prohibiting exploitative pricing by dominant concerns. It does not cover constructive refusals to sell or any exclusionary practice that might result in high prices by affecting the structure of competition in the relevant markets. This standard view is the one that I will take. Throughout this note, then, applying Article 82(a) simply refers to penalizing dominant undertakings that are found to enjoy unacceptably high price-cost margins. This note, as with all other papers in this symposium, must be read in the context of the Commission’s ongoing efforts to modernize the application of Article 82. Although many influences are at play, one dominant concern behind this review is to put the application of Article 82 more in line with the current guidelines on the application of Article 81 by moving toward a more effects-based approach. In this light, it seems useful to ask what might, or should, become of competition policy under Art 82(a) once the new regime is in place. My purpose is, however, more modest as I limit myself to the specific issue of the relationship between Article 82(a) and intellectual property. This is of special interest given the broadly held view that technological progress is essential for sustained economic growth and because the encouragement of innovation and its diffusion are themselves important objectives of the Treaty. The relationship between Article 82(a) and intellectual property rights can be construed more or less broadly. At one extreme, one might discuss how Article 82(a) should be applied to any dominant firm when it prices any IP or IP-based products. Alternatively, one might look at all aspects of the pricing behaviour of dominant firms—IP-related or not—when dominance is rooted * University of Essex and CEPR.

(Ze) Ehlermann 08 Régibeau

652

6/11/08

11:55

Page 652

Session Three: Exploitative Conduct And The Remedies

in the ownership of intellectual property rights. Finally, one could restrict attention to the pricing of IP or IP-based products by firms for which the corresponding IP rights are a main source of dominance. I adopt this third, narrowest, perspective. There are two main reasons why the presence of intellectual property rights might be relevant to the application of Article 82(a). The first reason is that establishing IP-based dominance and/or identifying excessive pricing of IP or IP-based products runs into difficulties that either might not arise or might be less severe when IP is not involved. The second issue is that, since it imposes a direct constraint on the prices that a dominant firm might charge, Article 82(a) has often been seen as a potential source of conflict between competition law and intellectual property law. After all, the basic approach of patent law, for example, is to grant monopoly rights for a limited period of time in order to provide adequate incentives to invest in the pursuit of innovation, as well as incentives to disclose the fruits of such investigations. For this system to work, it must be that, in some cases at least, the intellectual property rights confer enough market power to support a price-cost margin that allows investors to justify their initial investments.1 If application of Article 82(a) limits the recoupment ability of inventors, then it would seem to make intellectual property law less effective in light of its basic purpose. The organization of this paper reflects these two sets of issues. Section II briefly reviews the economic arguments for and against a strict application of Article 82(a). Section III looks at market definition and discusses how IP-based dominance might best be established. Section IV then turns to some difficulties that might arise in assessing what constitutes an exploitative price for an IP-related transaction. The possible systemic conflict between Article 82(a) and IP law is addressed in section V. While this note is mainly written with patents in mind, copyright protection is briefly discussed in section VI. Section VI also considers some IP-specific aspects of Article 82(a) in the case of joint dominance, and section VII concludes.

II. Should We Apply Article 82(a)? Economists have never been wildly enthusiastic about Article 82(a). Should exploitative pricing be labelled and pursued as an independent form of abusive conduct? It is not, of course, that economists feel that high prices do not deserve antitrust scrutiny. The adverse welfare consequences of prices that far exceed competitive levels are, after all, one of the main economic bases for 1 The cost of investment must of course be computed so as to reflect the ex ante risk of failure faced by any innovator.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 653

Relationship between Article 82(a) and Intellectual Property Rights 653 competition laws. The two principal difficulties are the competition authority’s ability to correctly identify this kind of abuse and the type of remedy that might effectively deal with a recurrence of the abuse. Not much needs to be said about measurement issues. It is always perilous to embark on complex cost-allocation exercises, but in industries with significant sunk costs and where methods of production might vary across competitors, it is not even clear how one would define a sensible “competitive” benchmark. The second issue is how the abuse might best be prevented. This involves both deterrence and remedies. While large penalties could in principle deter firms from engaging in exploitative pricing, the very fact that the diagnosis is bound to remain imprecise limits the possibility of determining the optimal level of penalties. One is therefore faced with the problem of designing remedies that would prevent a recurrence of the abuse even though economic conditions are likely to change over time. It seems likely that designing such a remedy and monitoring its continued application would require regulators with significant industry-specific knowledge. Because of these obstacles, the dominant opinion among economists seems to be that exploitative pricing should be dealt with through a two-pronged approach. On the one hand, competition policy should concentrate on conduct that is likely to adversely affect market structure, eventually leading to high prices. This would involve both merger regulation and enforcement of the prohibition against non-exploitative abuses covered under Article 82. An extreme version of this view prevails in the US, where exploitative pricing is not in itself an abuse. On the other hand, direct regulation should be exercised by a specialist agency in the few industries where one feels that excessive pricing is likely to be a recurrent issue.2 My own opinion is slightly different. While the two-pronged approach makes a lot of sense, one should remember that merger control and the enforcement of the rest of Article 82 are themselves very uncertain exercises. In particular, establishing an abuse with the precision required to convince courts can be difficult. There is therefore some value in maintaining a rule against exploitative pricing as an additional tool to help restrain the behaviour of dominant undertakings. The Commission’s position on the future of Article 82(a) is hard to guess. While this provision has been used in the past,3 so far, it has been conspicuously absent from the debate on the revision of the application of Article 82.

2 Note, however, that—even in these industries—exploitative pricing is not necessarily the main issue (except for legal or natural monopolies). In practice, there tends to be much emphasis on excessive pricing of interfaces (access pricing) because of its potential anticompetitive (but not exploitative) effects. 3 See, e.g., Commission Decision 75/75 of 19 December 1974, General Motors [1975] OJ L29/14, upheld: Case 26/75, General Motors Continental NV v Commission [1975] ECR 1367; Commission Decision 76/353 of 17 December 1975, Chiquita [1976] OJ L95/1; upheld in part: Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207. See also Case 110/88, Lucazeau et al. v SACEM et al. [1989] ECR 2811.

(Ze) Ehlermann 08 Régibeau

654

6/11/08

11:55

Page 654

Session Three: Exploitative Conduct And The Remedies

This reflects the Commission’s decision to exclude exploitative abuses from the (already long) Discussion Paper. However, exploitative abuses will eventually be covered by a separate review by the Commission. For the sake of this paper, the first interpretation will be retained. It would not make much sense to be discussing the future of a legal dodo.

III. IP-Based Dominance Since Article 82(a) only applies to dominant (or jointly dominant) concerns, we first need to discuss how one would go about defining IP-based dominance. When discussing intellectual property from the point of view of competition law, it is traditional to distinguish between the three types of vertically related markets shown in the graph below.4 Clearly, this distinction is better adapted to patented material and know-how than to traditional copyrighted material.5 The following discussion should therefore be read as applying mostly to the first two types of intellectual property. While the definition of product markets is standard, the notion of technology and innovation markets needs to be clarified before we proceed. Technology markets are markets where technology itself is (potentially) traded. Such transactions can be loosely referred to as “licensing”, although it can cover a wide variety of agreements, from straight licensing to cross-

Innovation markets

Technology markets

Product markets Figure 1 4 See, e.g., Steven Anderman and John Kallaugher, Technology Transfer and the New EU Competition Rules, Oxford University Press, 2006. 5 Traditional copyrighted works would not include software, which, although protected under copyright law in the EU, is in many respects a hybrid case.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 655

Relationship between Article 82(a) and Intellectual Property Rights 655 licensing or even the formation of (research) joint ventures. As a first cut, the definition is functional: any piece of intellectual property that enables companies to “do the same thing”, e.g., to introduce a foreign gene into a host cell, or to produce any low-calorie substance that tastes like sugar, belongs in the same technology market. However, as is the case for any other input market, it is often hard to rely exclusively on such a functional approach without eventually tying the definition of the technology market to the downstream product markets in which the technology can be usefully deployed. A gene insertion technology might be more suitable for plant cells than for animal cells, or it might be more effective for introducing herbicide-resistant genes than genes modifying the nutritional content of plants. Similarly, not all sugar-replacement substances are equally soluble or fare equally well when baked. In the case of IPRs, the need to link the definition of the “input” market to specific downstream markets is also reinforced by the fact that IPR holders are typically allowed (and able) to offer different licensing terms for different fields of use. Hence, even if a given patent, for example, were equally useful to design surgical equipment or develop data-reading devices (think LASER), the patent holder will usually license its intellectual property to a given company for only a limited number of specific uses, and this will be reflected in the agreed upon price. This is rather different from the situation in other input markets. If, for example, ball bearings can be used in several different product markets, the possibility of resale would still lead to a single price across the various uses so that the relevant upstream market for ball bearings can be defined independently of the precise downstream markets for which they are destined. In the case of intellectual property, the lack of arbitrage makes it more reasonable to define different upstream technology markets for each field of use and, therefore, for each set of corresponding downstream markets. This means that the same piece of intellectual property, for example a given patent, can easily belong to several distinct technology markets. Supposing that technology markets have been properly defined, how would dominance be assessed? A widespread view is that, when it comes to technology markets, market shares are essentially useless.6 The general idea seems to be that the number of “units” of IP sold in the technology market is not an operational concept because the same unit of IP can be re-used indefinitely. This is not completely correct. Following our discussion of technology markets, one could—in principle at least—obtain useful estimates of a technology’s market share by computing the downstream market shares of the products that rely on the (licensed) technology.7 In other words, since 6 See, e.g., Steven Anderman, EC Competition law and Intellectual Property Rights, Oxford University Press, 1998, pp. 142–143. 7 In this sense, see the Commission’s Technology Transfer Guidelines, [2004] OJ C101/2, para. 23.

(Ze) Ehlermann 08 Régibeau

656

6/11/08

11:55

Page 656

Session Three: Exploitative Conduct And The Remedies

technology markets are closely tied to the downstream product markets that they serve, downstream market shares would be a good indicator of market power in the upstream market. Hence, the lack of measurable market share in the technology market itself is not an insurmountable issue. There are, however, some practical and conceptual difficulties with this approach. On the practical side, computing the market shares in a given technology market would involve tracking down not only who uses whose technology but also computing market shares for each of the users in each of the downstream markets relevant to a given field of use. The problem is compounded by the fact that, as licensing contracts do not always involve volume-related royalties, downstream usage cannot always be inferred from data on royalty payments. On the conceptual side, the main issue is that intellectual property is not always licensed in order to be used by the licensee. Licensing agreements are also reached in order to step out of the shadow of possible infringement and litigation, bringing the benefits of legal certainty.8 Such licensing does not translate into any downstream market share, but the fact that other companies feel the need to obtain such a licence anyway surely reflects some degree of dominance on the part of the licensor. Because of these difficulties, computing technology market shares is likely to be cumbersome and might not give an adequate picture of whether an IP right-holder is dominant. One might therefore prefer to adopt an essential facility-like approach, where dominance is assessed more directly. Since, in practice, exploitative pricing is only likely to be investigated—and punished—when it is fairly extreme, I believe that such an approach would make sense.9 Exploitative pricing can also, in principle at least, arise in innovation markets. While these “markets” are often just a convenient term to summarize competition policy’s concerns about dynamic efficiency, one can also think of them as places where actual transactions occur. In particular, ex ante licensing agreements or research joint ventures, where the parties agree on the sale or sharing of intellectual property assets that are yet to be developed can be regarded as trades occurring in innovation markets. One could then imagine that a firm that happens to be dominant in an innovation market could abuse this dominance by imposing unfair prices or conditions in order to enter into such trades. Still, both the very definition of dominance in innovation markets and the concept of excessive pricing of assets yet to be developed raise such difficulties that applying Article 82 in this context would seem unwise.10 I will therefore not consider this issue any further in this paper.

8 This corresponds to the notion of “design freedom” found in the Commission’s Guidelines, cited in previous footnote, e.g. at para. 148. 9 It would also have the merit of imposing some coherence between the treatment of excessive pricing and the determination of fair access prices as remedies in cases of refusal to supply. 10 I am unaware of any case ever having been brought on such a basis.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 657

Relationship between Article 82(a) and Intellectual Property Rights 657

IV. Assessing Exploitative Pricing of IP Let us now assume that dominance in a properly defined technology market has indeed been established. Let us also leave aside for now the fundamental question of whether pursuing exploitative pricing as an independent form of abuse is in fact desirable. Let us suppose that it is. Does the fact that the dominance and abuse are IP-related cause any special difficulty in diagnosing the abuse or in imposing adequate remedies? As background for this discussion, it is useful to establish a few stylized facts about technology licensing. Unfortunately, reliable data about licensing behaviour is not widely available. Probably the most extensive study to date is Anand and Khanna (2000).11 They rely on publicly available US data that indicate the industry where the licensing occurs (according to 2- or 3-digit SIC codes) and a number of characteristics of the contracts. These characteristics refer to the nature of the contract (licensing, cross-licensing or joint venture agreements) and to contractual clauses such as exclusivity or use restrictions. There is unfortunately no information about the structure of the payments involved. The data set includes 1612 licensing agreements signed between 1990 and 1993. The first striking feature is industry concentration. At the two-digit SIC level, three sectors, “Chemicals” “Industrial and Commercial Machinery” and “Electronic and Electrical Equipment and Components” account for more than 80% of the agreements. At the three-digit level, “Drugs”, “Computer and Office Equipment”, “Communications Equipment”, “Electronic Components” and “Surgical Equipment” jointly represent 72.7% of the total number of licensing deals. “Drugs” alone make up more than a third of the total. These shares are much larger than the corresponding shares of patents obtained by these sectors. One can therefore conclude that there are many innovative sectors where licensing is scarce. A second feature of interest is the prevalence of exclusivity. Clauses granting exclusivity were found in more than two-thirds of the contracts for which their presence could be checked. In a third of the cases, exclusivity was worldwide. There are also quite significant differences across industries, with exclusivity being granted much less frequently for the machinery and electronic sectors than for the rest of the data set, especially drugs and chemicals. Finally, about 12.6% of all licensing agreements were cross-licensing agreements. Two main points are relevant to our discussion of Article 82(a). Firstly, there are significant sectoral differences in the level of effective protection provided by IP law. The consequences of such differences will be discussed in 11 Bharat Anand and Tarun Khanna, “The Structure of Licensing Contracts”, in 48(1) Journal of Industrial Economics 103 (2000).

(Ze) Ehlermann 08 Régibeau

658

6/11/08

11:55

Page 658

Session Three: Exploitative Conduct And The Remedies

section V. Secondly, except for the drugs and chemicals sectors, licensing appears to be relatively scarce. Moreover, when licensing occurs, it often takes the form of exclusive agreements. This seems to suggest that there are two market structures that we are most likely to run into: one where the patent holder is the only one to exploit the patented technology in downstream markets, and another where the right is exclusively licensed to a third party.12 This immediately raises the question of where Article 82(a) should be enforced. Should the focus be on excessive pricing in the technology market itself or on excessive pricing in the related downstream markets? This choice is not neutral, at least when the right-holder also has the ability to operate on the downstream markets. Suppose, for example, that—due, perhaps, to some of the measuring difficulties to which I will refer shortly—enforcement of Article 82(a) is mostly focused on excessive pricing in downstream markets. This might lead the right-holder to choose the licensing route even when vertical integration would be more efficient. Conversely, shifting attention to technology markets might prevent efficient licensing, pushing the firm to exploit the intellectual property internally. In other words, giving undue weight to enforcement either upstream or downstream invites inefficient bypass. Let us now take the licensing arrangements—or lack thereof—as given. If the firm does not license its intellectual property and decides to develop it itself for downstream markets, Article 82(a) can only be meaningfully applied in those markets. On the other hand, if the right-holder does license his intellectual property, then one could in principle enforce rules against exploitative pricing in either—or both—upstream and downstream markets. The practicality of focusing on downstream markets depends on how the courts are likely to compute downstream margins. To see this, suppose that the downstream market is currently perfectly competitive and that the licensed technology lowers (constant) unit costs from co to c1 (see figure 2). Assume for simplicity that the innovation is not drastic in the sense that the pre-innovation downstream price po = co is lower than the monopoly price corresponding to the lower cost c1, i.e. pM(c1). The licensor maximizes its profits by offering the license, exclusively or not, at a per-unit royalty equal to the cost difference co – c1 imputable to the use of the protected IP. If downstream margins include all operating costs, including the royalty paid to the licensor, then there would be no abuse under Article 82(a) in the downstream 12 Given the nature of intellectual property, such a pattern should not be surprising. We know that an upstream supplier faces a potential commitment problem. If it cannot guarantee exclusivity, then buyers know that they will eventually be competing downstream with many other firms that have also obtained the same input. This commitment issue is especially serious for intellectual property because the upstream vendor’s ability to “keep on selling” the upstream input is not limited by any capacity constraint. Hence, we would expect an IP-holder to either avail itself of the fairly broad legal tolerance for exclusivity clauses or to avoid the commitment issue altogether by vertically integrating into the downstream market.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 659

Relationship between Article 82(a) and Intellectual Property Rights 659

Figure 2

market. If on the other hand, downstream margins do not factor in the cost of gaining access to the IP, then applying Article 82(a) downstream would have exactly the same effect as applying it directly to the level of the royalty in the technology market. Since it is hard to imagine that the courts would not consider royalty payments as part of the legitimate costs of downstream firms, this equivalence is unlikely to hold. Overall then, a good rule of thumb might be that Article 82(a) should be enforced in downstream markets when the right-holder does not license and in the technology market when he does.13 Except for the traditional concern about ensuring an appropriate reward to innovators, which is discussed in the next section, the case of vertical integration without any licensing does not present any IP-specific aspect. Let us therefore concentrate on the application of Article 82(a) to the terms of licensing contracts. The first, frustrating issue is what the right benchmark for assessing excessive pricing in technology markets is. Clearly, as the cost of actually delivering the IP to the licensee is usually quite low, enforcing a price that is close to marginal or variable unit costs would be inappropriate. In practice, however, we can safely assume that, as in downstream markets, the actual 13 Of course, both types of enforcement would be relevant if the right-holder were competing downstream with some of his licensees. However, in this case, the main concern is likely to be foreclosure, rather than exploitative pricing.

(Ze) Ehlermann 08 Régibeau

660

6/11/08

11:55

Page 660

Session Three: Exploitative Conduct And The Remedies

enforcement of Article 82(a) would allow for significantly higher margins than that meaningless competitive benchmark. Still, licensing contracts present some additional difficulties. As we learned from Anand and Khanna (2000), a majority of licensing contracts are exclusive. The terms of such contracts are not set unilaterally by the licensor but are negotiated on a bilateral basis. We should therefore think of the process in terms of bargaining theory, not in terms of models of posted prices. We would therefore expect that the parties would do their best to find contractual clauses that allow them to maximize the size of the total surplus to be shared between them. Given the pervasive moral hazard and adverse selection problems that plague IP transactions,14 the resulting contracts are complex not only in terms of their pricing structure but also because of the conditions on use, litigation or grant-backs that are often imposed. Evaluating the true value/price of any given licensing contract is therefore a fairly major and hazardous undertaking. Concretely, would an ad valorem royalty of 12% without grant-back requirements be considered excessive pricing while a royalty of 6% with grant-backs is not? Using royalties charged for similar products is unlikely to be much of a help either, as intellectual property that is significant enough to be the source of dominance is likely to be fairly unique. Moreover, since EU competition law has long recognized that patent-holders should generally be able to discriminate by field of use, comparisons with the conditions offered to other (exclusive) licensees of the right-holder are unlikely to be informative. A final difficulty is that, while barter is not common in other “input” markets, cross-licensing of intellectual property is not rare. In fact, there are fields where it is the norm rather than the exception. The problem is not simply that assessing excessive pricing in such cases appears to be nearly impossible. If one does give up on such assessments, it also provides incentives to use crosslicensing in lieu of more transparent arrangements, even if the latter would be more efficient. The bottom line of this section is quite simple: there are some reasons to believe that enforcing Article 82(a) in an appropriate manner is likely to be even more challenging when the basis of dominance is rooted in IP. This is especially true when the intellectual property is licensed rather than used by the innovator himself.

14 See Nancy Gallini and Brian Wright, “Technology Transfer under Asymmetric Information”, 21 (1) Rand Journal of Economics 147 (1990); Katharine Rockett, “The Quality of Licensed Technology”, 8(4) International Journal of Industrial Organization 559 (1990).

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 661

Relationship between Article 82(a) and Intellectual Property Rights 661

V. The Systemic Conflict between Article 82(a) and IP Law In this section, I assume that the difficulties discussed above can be overcome: dominance in a properly defined market has been established and the antitrust authority measures the relevant price-cost margin accurately. In this context, an effective application of Article 82(a) would limit the firm’s ability to charge high prices for the intellectual property that is the source of its dominance. As discussed in the previous section, this pricing constraint would in principle apply both to the pricing of the technology itself and to the prices of the products made by the firm using its own intellectual property. Since the intellectual property system works by using the rewards of exclusivity to induce effort and diffusion, one might legitimately wonder whether enforcing Article 82(a) would in fact make it difficult for IP law to achieve its objectives. To assess this systemic interaction between IP law and competition law, it is useful to review, however briefly, the economics literature on optimal intellectual property protection. My focus will be on patents (mostly) and copyrights. In both cases, the basic structure of the protection is that the right-holder can exclude others from use for a limited period of time.15 I should also mention that I only consider a “reasonable” application of Article 82(a), which, in line with the approach suggested above, would only catch situations of fairly extreme dominance. Clearly, using Article 82(a) to effectively force all firms to make zero operating profits would completely undermine the intellectual property rights system. The paucity of IP-related Article 82(a) cases suggests that this interpretation probably captures any likely level of Article 82(a) enforcement.

Article 82(a) and the Breadth of Patent Protection Consider first a very simplified description of the patent granting process, where the patent-granting authority can only determine two dimensions: the length of protection and its breadth. For now, following Gilbert and Shapiro (1990),16 let us simply define patent breadth as anything that makes it possible for the right-holder to profitably charge a higher price during the lifetime of the patent. Let us also assume that we know the level of “reward”, call it R, that we want to procure for the right-holder. The issue of optimal patent (or copyright) design then involves finding the combination of length and 15 In order to side-step IP law subtleties that are not especially relevant to the discussion in this paper, I will also assume that patent and copyright protection actually allow the right-holder not only to exclude others but also to use the asset himself. 16 Richard Gilbert and Carl Shapiro, “Optimal Patent Length and Breadth”, 21(1) Rand Journal of Economics 106 (1990).

(Ze) Ehlermann 08 Régibeau

662

6/11/08

11:55

Page 662

Session Three: Exploitative Conduct And The Remedies

breadth that provides the necessary reward to the innovator at the least possible social cost. What is relevant for our discussion is that Gilbert and Shapiro’s notion of patent breadth maps very easily onto the possible effects of implementing Article 82(a). After all, enforcing Article 82(a) amounts to imposing a ceiling on the prices (or price-cost margins) that the firm can earn while its IP-based dominance lasts.17 This is true irrespective of where Article 82(a) would actually be applied. If the effective price limit arises in the downstream markets, then the analogy with Gilbert and Shapiro (1990) is complete.18 If, on the other hand, the price limit is imposed in the technology market, then this is equivalent to imposing an upper bound on royalties, which is essentially equivalent to the situation analyzed by Tandon (1982).19 Interestingly, Gilbert and Shapiro (1990) and Tandon (1982) reach the same conclusion: the welfare-maximizing manner of ensuring a given reward to the patent-holder is to issue patents of infinite length but of a breadth that is just sufficient to obtain the required compensation. One could read this analysis as good news for the potential conflict between IP law and Article 82(a): except for cases where the necessary reward R is very high, patent breadth should be small enough that the resulting price levels are unlikely to be construed as exploitative. In other words, intellectual property protection designed around these principles should only rarely give rise to the kind of dominance and pricing behaviour that would warrant scrutiny under Article 82(a). This interpretation would in fact be consistent with the scarcity of IP-related Article 82(a) cases. However, upon closer examination, such a view proves to be rather superficial, for several reasons. The first issue is that actual patent systems are not designed according to the Gilbert-Shapiro-Tandon (hereinafter, “GST”) prescription. Most obviously, the length of protection is not infinite. The main reason for this is that, beyond a point, additional length does not help the patent-holder because new generations of innovations eventually arise that make the initial invention obsolete. To capture this aspect of cumulative research, economists distinguish between two notions of breadth. Lagging breadth protects the patent-holder against imitation, while leading breadth sets a minimum novelty standard that offers protection against improvements.20 If 17 As already mentioned, I regard Article 82(a) as essentially constraining operating profit margins, without explicit allowance for the resources expended in obtaining the intellectual property. The issue of whether such an allowance should be made is discussed separately later. 18 In fact, Gilbert and Shapiro analyze just such a special case, where limited breadth is modelled as a ceiling on the price of the patent-related product. 19 Pankaj Tandon, “Optimal Patents with Compulsory Licensing”, 90 (3) Journal of Political Economy 470 (1982). However, Tandon’s analysis is slightly narrower, as it treats the case of compulsory licensing to a perfectly competitive downstream industry. 20 For a more precise definition of these concepts, see Pierre Régibeau and Katharine E. Rockett, “The Relationship between Intellectual Property Law and Competition Law: An Economic Approach”, in Steven Anderman, ed., Intellectual Property Rights and Competition Policy, Cambridge University Press, 2007, pp. 505 et seq.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 663

Relationship between Article 82(a) and Intellectual Property Rights 663 leading breadth were infinite, then one could indeed use extra length in order to provide additional compensation to inventors. However, a very large leading breadth is undesirable: while it does help pioneering innovators to capture some of the returns from the follow-on innovations that they have effectively spawned, it also tends to slow down the rate of future innovation and can limit the diffusion of knowledge.21 Not surprisingly, then, actual patent systems only offer limited leading breadth and therefore can only offer a limited length of effective patent protection. If the length of patent protection cannot be infinite, then lagging breadth must be greater than in the GST scheme in order to ensure any given level of reward R. Since higher lagging breadth leads to higher prices, infringement of Article 82(a) becomes more likely than what GST’s analysis would suggest. This also means that, in order to preserve the same level of reward R, any enforcement of Article 82(a) that would effectively constrain price levels during the period of patent protection would require an extension of leading breadth, with possible adverse consequences for future innovations. The literature on optimal patent design also gives too stylized an image of how actual patent systems operate. Firstly, patent systems are not explicitly designed to implement different levels of “reward” for different (types of) inventions. The reward R must therefore be seen as the reward that one might want to assign on average. This means that it is only if that average level of reward is quite high that a well-designed patent system would implement patent breadths that are likely to trigger the application of Article 82(a). Secondly, what economists call lagging and leading breadth result in fact from the application of a series of technical requirements that determine both patentability and infringement.22 The application of similar criteria to different types of innovations can therefore result in widely varying “breadth” of patent protection in the economic sense. We have already evoked Anand and Khanna’s quite convincing argument that, under current technical standards, the protection granted in relation to drugs and chemical patents, for example, tends to be significantly stronger than for electronic component patents. Hence, a more appealing way to look at the patent system is that it aims to ensure what is judged to be, on average, a sufficient level of reward but that the application of the criteria meant to achieve this average still leads to very different levels of effective protection for different types of innovations. These variations do not arise by design, i.e., they do not reflect any judgement that some types of innovations are worthier of protection than others. An important implication of this view is that the differences observed in the amount of resources that are invested in the pursuit of (patentable) innovation are, to some extent at least, the result of the varying degrees of protection that are 21

See ibid. for a more extensive review of the relevant economic literature. For a detailed discussion of this issue, see Suzanne Scotchmer, Innovation and Incentives, MIT Press, 2004. 22

(Ze) Ehlermann 08 Régibeau

664

6/11/08

11:55

Page 664

Session Three: Exploitative Conduct And The Remedies

effectively available. In other words, the fact that R&D outlays in, say, the pharmaceutical industry tend to be high is not necessarily an exogenous feature of this industry that simply follows from its underlying technology. It can also result from the greater profitability of pharmaceutical patents due to the stronger patent protection enjoyed in that sector. A corollary is that the greater investment observed in sectors that de facto enjoy greater IPR protection does not necessarily reflect any inherently greater social desirability of innovation in these fields.

How “Directed” is Innovation? In order to disentangle the implications of this pragmatic conception of patent protection, we need to distinguish between different views of the innovation process. At one extreme, one might believe that the whole R&D process is essentially undirected. According to this view, while the likelihood of success (or speed) of innovation can be improved by investing more, the outcome of research is completely random. Not only might a given project succeed or not, but there is essentially complete uncertainty as to what might be produced in the end. An attempt to discover a new medically useful compound might as well produce aspartame or some mechanical innovation. In such a context, incentives to invest in any given project are driven by the potential rewards that could be expected from success across a broad section of fields. Hence only the expected returns averaged across fields matter. This means that the fact that the strength of patent protection might differ across types of innovations has no effect on incentives to invest. Similarly, the fact that Article 82(a) might constrain pricing behaviour in one sector more than in another is irrelevant: applying Article 82(a) would not bias the allocation of resources across innovation projects. Moreover, since the number of cases to which Article 82(a) is realistically likely to be applied is rather small, its effect on the average expected return to—and thus on R&D incentives—is likely to be very minor: applying Article 82(a) would not significantly affect IP law’s ability to provide innovation incentives. This view of the innovation process is of course extreme. More reasonably, one might think that, while there can indeed be a lot of uncertainty as to what a successful outcome of a given project might be, projects are directed towards specific fields of research and are more likely to yield positive outcomes in that field. In such a “semi-directed” context, it is important for the combined impact of patent policy and Article 82(a) to be neutral across sectors of innovative activities so that the relative private returns to R&D projects accurately reflect their relative social value. Any induced discrepancy between the ranking of private and social returns would result in a bad allocation of R&D effort. Unfortunately, the implications of such neutrality for IP law and Article 82(a) cannot be discussed in any great generality. Still, a simple example helps to make a few useful points.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 665

Relationship between Article 82(a) and Intellectual Property Rights 665 Imagine that all research projects lead (if successful) to the creation of new products. Further assume that the projects only differ in the size of the new markets that are created. This market size may be designated as the “size of the project”. Suppose that sectors of innovative activity are known to differ in (the distribution of) project sizes. For example, imagine that drug-related research projects are on average “larger”—and hence more socially worthwhile—than research projects in electronics. Abstracting from R&D competition between firms,23 let us assume that the only effect of investing more in a given project is to reduce the time required for the project to bear fruit. In a world without Article 82(a) and with infinite lagging breadth, each successful inventor would simply charge the monopoly price in the relevant market. Given our assumption that markets only differ in their size, this monopoly price is the same in all markets. Under these conditions, private investment will be allocated across R&D sectors in a manner that equalizes the marginal benefit of investment. This marginal benefit is just the gain from speeding up the project a little bit more. Since firms typically cannot extract the full social value from the new products that they sell, social surplus in each market is a multiple of the firm’s expected profits. Again, with markets that are identical except for size, this multiple is the same for all markets and hence for all sectors of research. The socially optimal allocation of investment across sectors is such that the marginal benefit derived from bringing the corresponding expected social value forward is equalized across sectors. But since social surplus is simply a multiple of expected profits, this optimality condition must be achieved because a private investment decision already equalizes private marginal benefits across fields: the privately optimal allocation is also socially optimal. At this stage, let us simply note that equilibrium investment will be higher in larger sectors, even though the operating profit margins are the same in all markets. Now let us impose a uniform upper limit on operating price-cost margins in all markets. This limit can be seen as resulting from the definition of patent breadth and/or the application of Article 82(a). Imposing this limit simply reduces the ratio between profits and social surplus in each market, but this lower proportion is still the same in all sectors. Hence, applying the same reasoning as above, private investment incentives still lead to a socially optimal allocation of resources across sectors. This in turn implies that, as long as the limit is applied to operating profit margins, Article 82(a) would not induce any inefficiency in the pattern of investment. In this context, giving more favourable treatment to sectors where R&D outlays are larger (for example by including R&D costs in the computation of the profit margins) would be 23 Considering R&D competition complicates matters a lot because it can break the link between greater private profitability and greater social desirability, e.g., due to greater duplication of effort for larger projects. The uncertain welfare properties of R&D competition are the main reason why this part of the argument cannot be made more general.

(Ze) Ehlermann 08 Régibeau

666

6/11/08

11:55

Page 666

Session Three: Exploitative Conduct And The Remedies

a mistake, as it would inefficiently channel more resources into sectors with larger projects. However, we have seen that IP law itself will, in practice, effectively grant broader patent protection in some areas of research. This has two main implications. Let us now, for simplicity, assume that the distribution of project size is the same in all sectors so that, in the absence of inequalities in patent protection, investment levels per project would be the same across fields. The first consequence of uneven patent protection is an undesirable bias in the allocation of resources: (relatively) too much investment will take place in the favoured sectors. The second implication concerns the application of Article 82(a). Since greater patent breadth leads to greater operating price-cost margins, innovations in favoured sectors would be more likely to be caught by Article 82(a). Application of Article 82(a) would then help reduce the abnormal expected returns in “high breadth” sectors, moving the allocation of innovative investments across sectors back towards the social optimum. Treating sectors with higher investments more favourably would be exactly the wrong policy, as it would work against this welfare-increasing effect. Finally, at the other extreme of the spectrum lies a view of perfectly targeted R&D investment, where innovators choose not only a field of research but a very specific project for which they know the expected size of the market. The analysis of this situation is almost the same as for semi-directed research. As above, implementation of Article 82(a) as a limit on operating profit margins would be neutral in terms of the allocation of resources across projects, so conditioning the application of Article 82(a) on the size of the investment undertaken would be counterproductive. The point about counteracting possible IP law-induced biases in this allocation also remains valid. These are strong conclusions: any reasonable application of Article 82(a) is unlikely to have large effects on the levels of investment in R&D, and it is in itself neutral in terms of the allocation of R&D resources across sectors or projects. It might even help to correct some of the biases created by IP law itself. How robust is this analysis? The simple example on which it is based has two key elements. Firstly, large projects do not require disproportionate incentives. This comes from our assumption that the proportion between private and social surplus is independent of project size. If this proportion were lower for larger projects, then—at least in the semi-directed and directed views of R&D—larger projects would need to receive a more-thanproportionate reward in order for the allocation of investment across projects (or sectors) to be efficient. Hence, higher operating price-cost margins would have to be guaranteed (on average) for large successful projects. Since larger projects also involve larger (equilibrium) outlays, there would be an argument for greater Article 82(a) leniency for firms that have expended significant resources. One should point out, however, that there is absolutely no convincing evidence that the ratio of private to social surplus is indeed negatively

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 667

Relationship between Article 82(a) and Intellectual Property Rights 667 related to project size. In the absence of any such evidence, our neutral benchmark seems a safer basis on which to develop policy guidelines. The second important feature of our example is that agents that invest in larger R&D projects do not need to enjoy higher operating profit margins in order to secure the same return on investment as agents that pursue smaller projects. The extra profit required to compensate them for the greater expense simply comes from applying the same profit margin to a larger market. This ensures that, provided that IP law is itself neutral, a given level of enforcement of Article 82(a) is not more likely to “catch” larger projects. Absent a convincing reason why large projects would systematically require a higher operating price-cost margin to ensure a socially efficient allocation of R&D investment, this does not seem like a bad benchmark.

Efficient Innovation The previous argument in favour of applying Article 82(a) (in moderation) without explicit regard for the resources invested was based on the idea that any bias would result in a bad allocation of resources across projects. Another, more familiar, reason for advocating an “investment-blind” policy has to do with the ex ante incentives to run any given project efficiently. Stated bluntly, a policy that treats right-holders who spend a lot more leniently than those who do not would not exactly reward efficiency in research and development. Such concerns have long been part of the debate on regulation, where schemes like rate of return regulation are known to provide incentives for “gold plating”.24

VI. Remarks Copyrights This paper has been written with patents as the running example of intellectual property. The main difference between patents and copyrights is that patent protection offers both greater lagging breadth and greater leading breadth, but it is of shorter duration.25 This means that, in principle, copyright protection is even less likely to result in the type of dominance and high 24 For an early discussion of the economics of “gold plating”, see E.E. Zajac, “Note on ‘Gold plating’ or ‘rate base padding’”, 3(1) Bell Journal of Economics and Management Science 311 (1972). 25 For a discussion of the economic rationale for these differences, see Régibeau and Rockett, cited supra note 20.

(Ze) Ehlermann 08 Régibeau

668

6/11/08

11:55

Page 668

Session Three: Exploitative Conduct And The Remedies

price-cost margins that would be likely to trigger the application of Art. 82(a). However, in other respects, the arguments presented above readily apply to cases of copyright-based dominance. In particular, the idea that the application of a given legal regime of protection to various sectors of activity might result in rather different degrees of effective protection is still appealing: copyright protection is unlikely to grant the same amount of market power when applied to movies as it does when applied to software.

Joint Dominance: Patent Pools and Collecting Societies A distinguishing feature of intellectual property rights is their uncertainty. In the case of patents, this often leads to situations where several patents might be mutually infringing, creating a so-called patent “thicket”.26 When the number of patentees involved is small, such thickets can be cleared through bilateral cross-licensing deals. However, this approach is not feasible when the number of parties concerned is large. In such cases, a natural response is the formation of patent pools. Typically, members of the pool have access to all of its patents on pre-established terms. Patent pools also set conditions for pool membership and for third party access to pooled patents. Patent pools raise a number of significant competition law issues ranging from the potentially anticompetitive exclusion of third parties to the fear that they facilitate tacit collusion.27 However, most of these relate either to Article 81 or to exclusionary abuses under Article 82. The only relevant aspect from the perspective of Article 82(a) is that patent pools (or collecting societies) can be a source of joint dominance that might lead to abusively high prices charged to third parties.28 In this respect, Lerner and Tirole (2004) offer significant reassurance: even if one allows for asymmetries between parties, blocking patents and a mixture of substitute and complementary rights, stipulating that individual right-holders must remain free to license their intellectual property outside of the patent pool arrangement is a sufficient condition to ensure that the pool will be welfare-enhancing.29 In that sense, in the presence of jurisprudence that requires that such independent transactions be allowed, application of Article 82(a) might be superfluous.

26 For a recent discussion of patent thickets, see Mark Lemley and Carl Shapiro, “Probabilistic Patents”, 19(2) Journal of Economic Perspectives 75 (2005). 27 On this last point, see Régibeau and Rockett, supra note 20. 28 To make this a “pure” Article 82(a) issue, let us assume, somewhat unrealistically, that these third parties do not compete in the same downstream markets as members of the patent pool. 29 Josh Lerner and Jean Tirole, “Efficient Patent Pools”, American Economic Review, June 2004, pp. 691 et seq.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 669

Relationship between Article 82(a) and Intellectual Property Rights 669

VII. Conclusions The conclusions drawn from this paper are simple. Applying Article 82(a) in a fairly reliable manner is bound to be hard in any context, which suggests that use of the provision should be reserved for fairly extreme cases. The situation is even less pleasant when intellectual property assets are the source of dominance and (indirectly at least) the object of excessive pricing. There are two types of additional difficulties. Firstly, assessing dominance in technology markets might be harder than it is in the case of traditional input markets. This is mostly due to the practical difficulties in computing meaningful market shares. This first difficulty could be minimized by relying on a more direct “essential facility” type of approach. The second difficulty is that the public good aspect of intellectual property and the fuzzy nature of intellectual property rights make for rather complex licensing arrangements. Determining which arrangements amount to exploitative pricing and which ones do not is bound to be extremely challenging. On the other hand, I find little merit in the claim that applying Article 82(a) (reasonably) is undesirable because it undermines the socially desirable properties of the system of intellectual property rights protection. In particular, I argue that applying Article 82(a) more leniently when the intellectual property involved was obtained after significant investments is likely to be counterproductive, as it would interfere with the efficient allocation of investment across R&D projects and weaken incentives to conduct these projects efficiently. Overall, then, Article 82(a) is hard to love, especially in the context of IP, but not because of some supposed conflict between IP law and competition law. Still, as other approaches to abuse of dominance are also imperfect, Article 82(a) is probably worth keeping as part of the competition authority’s arsenal . . . provided that it is wielded with moderation.

(Ze) Ehlermann 08 Régibeau

6/11/08

11:55

Page 670

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 671

IV Damien Geradin* Abusive Pricing in an IP Licensing Context: An EC Competition Law Analysis June 2007

I. Introduction In the last two decades, the reliance upon “licensing” strategies as a source of revenue for intellectual property (“IP”) rights holders has seen a dramatic increase.1 Put simply, in return for adequate remuneration (typically a royalty, but there may be other forms of consideration2), innovators (licensors) grant to other firms (licensees) the right to use their proprietary technology to manufacture products for sale in downstream markets. IP licensing strategies are not only pursued by organizations without manufacturing capabilities (e.g., university research centres).3 IP holders active in downstream product markets (“vertically integrated” firms) may license their technologies to reap additional profits from their research and development (“R&D”) expenditures, but they may also do so to obtain access to other firms’ technologies through cross-licensing agreements. Licensing agreements typically benefit both the licensor and the licensee. The licensee gains access to new technologies, which can be used to improve its manufacturing operations or which the licensee can embed in its own products to increase their functionalities. The licensor accrues revenues from his initial R&D expenditures that can be invested in the development of new technologies, which will in turn lead to additional revenues, hence creating a virtuous circle of innovation. Licensing agreements are generally heavily negotiated between licensors and licensees, which in the vast majority of the cases reach mutually satisfactory agreements.

* The author is part of a team representing Qualcomm, Inc. in the context of a pending competition case before the European Commission. The views expressed in this paper are the author’s own and cannot be attributed to Qualcomm, Inc. or to any other client of Howrey LLP. 1 See Ashish Arora, Andrea Fosfuri and Alfonso Gambardella, Markets for Technology: Economics of Innovation and Corporate Strategy, MIT Press, 2001. 2 Licensees can also pay an upfront free, or they can cross-license some of their rights in exchange for the licensors’ rights, etc. 3 See Alfonso Gambardella, Science and Innovation, Cambridge University Press, 1995.

(Zf) Ehlermann 08 Geradin

672

6/11/08

11:56

Page 672

Session Three: Exploitative Conduct And The Remedies

Yet tensions may arise between licensors and licensees over the terms of their IP licensing deals. The diverging incentives of licensors (eager to obtain a fair level of compensation for the investments made in developing their IP) and licensees (eager to minimize the cost of acquiring proprietary technologies) may generate disputes over royalty levels and other forms of consideration.4 Such disputes are particularly likely to arise when licensing agreements have the potential to be worth hundreds of millions of euros and small variations in terms and conditions can be financially significant for both parties. Potential licensees may also insist on obtaining a licence on terms that are identical, or at least equivalent, to those obtained by licensees with which they compete. At the same time, licensors may resist such requests insofar as differing licensing terms are justified by the particular circumstances of each specific agreement. Additional tensions may arise when the IP in question is essential to a standard. Some have argued that once a proprietary technology has become part of a standard, its owners will be able to extract royalties in excess of those they could have charged before the adoption of such standard (the so-called “hold up” theory).5 Although this theory has clear limitations (as will be seen), it has contributed to the belief that royalty rates charged by IP holders are too high. Another claim that has been made is that in circumstances where a standard comprises essential IP held by numerous patent holders, the aggregation of the rates charged by such holders (even if individually reasonable) may lead to a royalty burden of a level such that the standard will be too costly to implement (the so-called “royalty stacking” theory).6 The proponents of such theories argue that some form of control should be placed on the royalties that can be charged by essential patent holders.7 While differences of views between licensors and licensees are generally ironed out through negotiations, there will be situations where licensees may be tempted to rely on competition rules to seek redress against what they perceive to be unfair licensing terms. Against this background, this paper explores the extent to which Article 82(a) and 82(c) of the EC Treaty—which respectively prohibit as abusive for dominant firms “directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions” on their customers, and “applying dissimilar conditions to equivalent trans4 See Damien Geradin, “Standardization and Technological Innovation: Some Reflections on Ex-ante Licensing, FRAND, and the Proper Means to Reward Innovators”, 29(4) World Competition 511 (2006). 5 See Mark Lemley and Carl Shapiro, “Patent Hold Up and Royalty Stacking”, Stanford Law and Economics Olin Working Paper No. 324 (July 2006), available at http://ssrn.com/abstract= 923468. 6 Ibid. But see Damien Geradin, Anne Layne-Farrar and Jorge Padilla, “Royalty Stacking In High Tech Industries: Separating Myth from Reality”, available at http://papers.ssrn.com/sol3/ papers.cfm?abstract_id=949599. 7 Most of these proposals seek to reduce the bargaining power of essential patent holders. But see Damien Geradin, Anne Layne-Farrar and Jorge Padilla, “The Ex Ante Auction Model for the Control of Market Power in Standard Setting Organizations,” available at http://papers.ssrn. com/sol3/papers.cfm?abstract_id=979393.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 673

Abusive Pricing in an IP Licensing Context 673 actions with other trading parties, thereby placing them at a competitive disadvantage”—can be relied on by licensees unhappy with the deals they have obtained from licensors. These issues are particularly important at a time when economic growth is increasingly dependent on innovation. This paper is divided into five parts. Part II discusses the specific challenges raised by market definition and the assessment of dominance in hightechnology markets with a specific focus on technology licensing. Part III discusses the application of Article 82(a) to licensing agreements. It explains the significant conceptual and practical difficulties of applying this provision of the Treaty in the field of technology licensing and argues that competition authorities should refrain from seeking to control prices or rates in dynamic industries. Part IV explores the issue of price/rate discrimination in IP licensing agreements. It argues that while non-vertically integrated licensors have no incentives to discriminate against their licensees, vertically-integrated firms have strong incentives to offer more favourable licensing terms to their downstream operations than they offer to other downstream firms with which they compete. The enforcement of Article 82 EC in this field, it is argued, should therefore focus on preventing vertically-integrated firms from raising their downstream rivals’ costs through discriminatory licensing fees. Part V contains a short conclusion.

II. Market Definition and Dominance in Technology Markets Prices will only be examined under Article 82 where they are imposed by dominant firms. Thus, the definition of one or several product and geographic market(s) and the determination of the presence of dominance on such market(s) is the first necessary step of any enquiry into abusive pricing. The developments which follow show that in high technology industries the assessment of market definition and dominance raises a number of complex issues which need to be considered carefully.

A. Market definition In the context of technology covered by IPRs incorporated into a standard, the primary relevant market consists of the market for the licensed technology and its substitutes. Such substitutes comprise other technologies which by reason of their characteristics, price (i.e., royalty rate) and intended use are regarded by licensees as interchangeable with or substitutable for the licensed technology. However, the key to ascertaining whether such technologies are

(Zf) Ehlermann 08 Geradin

674

6/11/08

11:56

Page 674

Session Three: Exploitative Conduct And The Remedies

substitutable for the licensed technology is to examine whether licensees could switch to them in response to a small but significant, permanent increase in the royalties charged by the IPR owner for its standardized technology.8 If licensees of the standardized technology can switch to alternative technologies, patented or otherwise, then these alternative technologies form part of the relevant product market. Although this conceptual framework does not appear to differ significantly from that employed to define more traditional product markets, market definition in technology markets is a more complex undertaking. The intricacy of the task is compounded when, as is often the case, the technology at issue forms part of a standard. A standard can be defined as a set of technical specifications which seeks to provide a common design for a product or process.9 The welfare benefits inherent to standardization are obvious. Standards increase consumer choice and convenience, and reduce costs by allowing complementary or component products from different manufacturers to be combined or used together.10 A variety of standards have been defined in fields as diverse as communications technology, computer manufacturing or the automotive industry, and these standards are constantly being improved by a large number of standard-setting organizations (“SSOs”).11 The first element that needs to be considered when attempting to define relevant markets for standardized technology is the fact that, in practice, the implementers of a standard generally take a licence covering a company’s entire portfolio of essential IPRs for a given standard that is needed for the products they intend to manufacture and sell. An implementer of a standard would not typically seek a licence for an individual essential IPR on a standalone basis unless that particular IPR is the only one needed for the implementer’s specific product. Second, in many circumstances multiple firms hold essential IPRs to a given standard, each therefore being a complementary input for those wishing to manufacture and sell standard-compliant products. Companies wishing to practice the standard must therefore obtain 8 The conceptual framework for defining such technology markets is set out, inter alia, in the European Commission’s “Guidelines on the application of Article 81 of the EC Treaty to Technology Transfer Agreements” (“Technology Transfer Guidelines”), [2004] OJ C101/2, at p. 22. 9 See Herbert Hovenkamp, Mark Janis and Mark Lemley, IP and Antitrust: An Analysis of Antitrust Principles Applied to Intellectual Property Law, Aspen Publishing (2003–04 Supplement) at § 35.1. 10 See Amy Marasco, “Standards-Setting Practices: Competition, Innovation and Consumer Welfare”, testimony before the Federal Trade Commission and Department of Justice, available at: http://www.ftc.gov/opp/intellect/020418marasco.pdf, p. 3 (“Standards do everything from solving issues of product compatibility to addressing consumer safety and health concerns. Standards also allow for the systemic elimination of non-value added product differences (thereby increasing a user’s ability to compare competing products), provide for interoperability, improve quality, reduce costs and often simplify product development. They also are a fundamental building block for international trade.”). 11 See Mark Lemley, “Intellectual Property Rights and Standard-Setting Organizations”, 90 California Law Review 1889 (2002).

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 675

Abusive Pricing in an IP Licensing Context 675 licences for those essential IPRs from all of these firms. As the IPRs of these companies will typically cover different aspects of the standard, such IPRs are complements, not substitutes. The existence of non-substitutable complements obviously has profound implications for market definition. Third, as will be seen below, holders of essential IPRs contained in a standard are subject to a number of vertical, horizontal and dynamic competitive constraints with substantial implications both for market definition and for the assessment of dominance. Moreover, these constraints will differ significantly according to the role played by the IPR owner in the standardization process, i.e., depending on whether the IPR owner is a vertically-integrated firm active in the product market or a pure licensor which does not supply the end-product. The identification of the vertical competitive constraint resulting from the ability of final consumers to switch between devices using different access technologies is fundamental to market definition in the context of technology licensing. In other words, the existence of a downstream market for the product incorporating the standardized technology is paramount to any appropriate definition of the relevant upstream technology market. The potential for demand side substitution by consumers of the final product is thus yet another element with significant implications for market definition. If a hypothetical monopolist licensing essential IPRs raises the price of those IPRs, i.e. the royalty, at least some of the increase in costs is likely to be passed on by the manufacturer to final consumers (assuming a competitive market) who could switch to products using alternative technologies.12 If there are sufficiently close substitute products, end-users will switch in response to an increase in prices, making the initial increase in royalties unprofitable for the IPR owner. The important role of downstream competition in constraining upstream market power in technology markets is well established.13 Furthermore, prices for the final product may be constrained even if alternative products are attractive to some but not all customers of the hypothetical monopolist. The European Commission’s Discussion Paper on Article 82 makes it clear that it is not necessary for all customers to regard the products as substitutable for them to belong to the same product market. What matters is that enough marginal customers would switch to alternatives 12 Economic theory and empirical analysis suggest that there is generally a pass-through of costs to at least some extent. 13 The Commission’s Technology Transfer Guidelines recognize this point, stating: “If the downstream product market is competitive, competition at this level may effectively constrain the licensor. An increase in royalties upstream affects the costs of the licensee, making him less competitive, causing him to lose sales.” See supra note 8 at para. 23. See also Daniel Swanson and William Baumol, “Reasonable and Nondiscriminatory (RAND) Royalties, Standards Selection, and Control of Market Power”, 73 Antitrust Law Journal 1 (2005), at note 17 (“There may be no market power in the technology market even if the alternative technology set is small if there is vigorous rivalry from substitute goods in the market for the final product that makes use of the technology”.).

(Zf) Ehlermann 08 Geradin

676

6/11/08

11:56

Page 676

Session Three: Exploitative Conduct And The Remedies

if the price of end-products were to increase by a small but significant amount, so as to make the price increase unprofitable.14 These vertical constraints must be thoroughly examined in order for the relevant market(s) to be correctly defined.

B. Dominance in Technology Markets Pursuant to the legal standard established by the Court of Justice (“ECJ”), dominance arises where a firm has the power to behave to an “appreciable extent independently of its competitors, its customers and ultimately of the consumers” allowing it to “prevent effective competition being maintained on the relevant market”.15 The identification of the competitive pressures to which a firm is subject is thus crucial for the assessment of the existence of a dominant position. Where firms face significant competitive constraints they cannot behave independently of their customers and therefore cannot be deemed to enjoy a dominant position. This holds true whether such firms have any competitors in the market for the goods or services provided to such customers or not. While it is generally recognized that the owner of an IPR is not automatically placed in a dominant position, it has been argued that holders of IPRs that are essential to practice a standard automatically enjoy significant market power.16 The claim is that once a given technology becomes part of a standard, competition between technologies for the essential parts of that standard ends. No longer constrained by such competition, each owner of an IPR that is essential to the standard would ipso facto enjoy market power akin to dominance in the market(s) for the licensing of those IPRs. It has been argued that this effect would be compounded by the “hold up” of potential licensees which may have made substantial investments for its implementation and are locked into the standardized technology. As will be seen below, these claims cannot be sustained because they ignore the different horizontal, vertical and dynamic competitive constraints to which owners of IPRs essential to a standard are subjected and which preclude an automatic finding of dominance.

14 DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (Dec. 2005), http://ec.europa.eu/comm/competition/antitrust/art82/discpaper2005.pdf, at para. 18. 15 See Case 85/76, Hoffmann-La Roche & Co. AG v Commission [1979] ECR 461. 16 See Marcus Glader and Sune Chabert Larsen, “Article 82: Excessive pricing—An outline of the legal principles relating to excessive pricing and their future application in the field of IP rights and industry standards”, Competition Law Insight, 4 July 2005, p.3. But see Damien Geradin and Miguel Rato, “Excessive prices: In reply”, Competition Law Insight, 10 October 2006.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 677

Abusive Pricing in an IP Licensing Context 677

1. Vertical Constraints Stemming from Competition between Rival Standards and Non-standardized Substitute Products The adoption of a standard by an SSO may bring an end to effective competition between rival technologies for inclusion in that specific iteration of the standard. However, it will not affect competition between rival standards, in the guise of either downstream competition between substitutable endproducts compliant with different standards or competition between standards at the upstream licensing level. As seen above, competitive constraints arising at either the upstream or downstream level will prevent an owner of essential IPRs from holding a dominant position in the technology licensing market(s). If licensees of the standardized technology can switch to alternative technologies, covered by IPRs or otherwise, the IPR owner will not be able to exercise monopoly power because it would lose sales if it tried to increase price. Similarly, if end-customers can easily switch to substitute products that do not use the licensed technology, such competition between end-products will represent a significant competitive constraint on the owner of an IPR that is essential to a standard. This will hold true regardless of whether the substitute products comply with any given standard.

2. Horizontal Constraints Stemming from the Complementary Nature of IPR Incorporated in a Standard As seen above, standards usually comprise complementary essential IPRs owned by numerous firms. In order to practice the standard, implementers must obtain licences from all such owners of complementary IPRs. If other complementary IPR owners charge high royalty rates, a given firm will not be able unilaterally to set a high royalty rate for its IPR. This will be the case even if the company in question holds a monopoly over a given technology. When individually setting their prices, owners of essential IPR will inherently take into account prices set by other owners of complementary IPRs because the market—i.e., prospective licensees—will only bear a certain overall price level. Owners of IPRs essential to standard are thus horizontally priceconstrained and this absence of pricing independence will preclude a finding of dominance under Article 82.

3. Dynamic Constraints Owners of IPRs essential to a standard are also constrained in their ability to price independently by the dynamic nature of standard-setting. As noted above, competition between members of SSOs usually takes place not only before those SSOs adopt a standard but also afterwards, i.e., for the inclusion

(Zf) Ehlermann 08 Geradin

678

6/11/08

11:56

Page 678

Session Three: Exploitative Conduct And The Remedies

of new releases and next generation technologies. If a firm’s technology is included in a standard, that firm will face constraints when pricing any associated IPRs because it will continue to depend on the SSO for its position as the standard evolves. The dynamic and evolving nature of standards gives participants in SSOs a number of opportunities to “punish” companies that have previously set what are considered to be excessive royalties. SSO members may be able to choose not to include a company’s contributions in evolutions of the standard.17 Moreover, SSO members may be able to choose not to include a company’s contributions in future generations of the standard (or in other unrelated standards).

4. The Role of Dynamic Competition The final element which must be addressed when assessing dominance in the standard-setting context is not specific to standardization but appears inextricably linked to it insofar as technology standards and licensing occupy a preponderant place in dynamically competitive markets such as the ICT sector. These industries are characterized by dynamic competition for the market whereby drastic innovation makes market leadership highly contestable.18 By contrast, in other industries, competition takes place primarily through traditional price competition and perhaps also via incremental innovations.19 Dynamic competition consists of a series of races for market dominance. Firms do not compete by slightly undercutting each other but engage instead in what economist Joseph Schumpeter described as a “perennial gale of creative destruction” that “strikes not at the margins of the profits of the existing firms but at their foundations and their very lives”.20 In these industries,

17 See David Teece and Edward Sherry, “Standards Setting and Antitrust”, 87 Minnesota Law Review 1913 (2003): “[I]n many industries in which standards play an important role, the fast pace of technological change drives the continual redesign and reengineering of products. For example, the product life cycle in the semiconductor industry is reported to be as low as ten months. Therefore, even if there may be some ‘lock-in’ of earlier designs, once the existence of the patent is disclosed, the SSO has the opportunity to revise the standards, and manufacturers have the opportunity to redesign their products to avoid incorporating the patented features. In other words, the extent of ‘lock-in’ may be limited by the pace of technological change.” 18 See David S. Evans and Richard Schmalensee, “Some Economic Aspects of Antitrust Analysis in Dynamically Competitive Industries,” in Josh Lerner and Scott Stern, eds., Innovation Policy and the Economy, vol. 2, MIT Press, 2002, pp. 1 et seq. 19 For a detailed analysis of the competition policy implications stemming from dynamically competitive industries, see Christian Ahlborn, Vincenzo Denicolò, Damien Geradin, and Jorge Padilla, “DG Comp’s Discussion Paper on Article 82: Implications of the Proposed Framework and Antitrust Rules for Dynamically Competitive Industries”, 31 March 2006, available from the Commission’s website at http://ec.europa.eu/comm/competition/antitrust/others/057.pdf. 20 See Joseph Schumpeter, Capitalism, Socialism and Democracy, Harper Collins Publishers 1942/1984, p. 84.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 679

Abusive Pricing in an IP Licensing Context 679 competition takes place for the market rather than in the market. Firms take part in a race for innovation, striving to introduce new and superior products that will win the market and achieve massive transfers of market shares. In other words, competition comes not from readily available substitutes but from new, innovative products not yet present in the marketplace. Once a market is won, the ensuing dominance will afford substantial benefits, but it will be fragile and temporary. It can only be maintained if the dominant firm continues to innovate, as the initial race is succeeded by a new wave of investment by rival firms to displace the leading technology with something superior. The implications of such dynamic competition for the assessment of dominance must be carefully considered. The competitive constraints faced by any incumbent stem not only from existing competitors but also from significant forces outside the market. The underlying analysis should thus be adapted to reflect the special characteristics of these industries. Given their fleeting nature, market shares should not be blindly used as relevant indicators of market power in those industries, and supply-side constraints should be carefully considered when dominance is assessed. A firm which prima facie appears to enjoy a dominant position could be found, upon careful consideration, not to possess any significant market power.

III. The application of Article 82(a) EC to licensing agreements It is only when a licensor has been found to be dominant on one or several relevant market(s) that the question may arise as to whether the royalties charged to its licensees are abusive. Article 82(a) prohibits dominant firms from imposing “unfair purchase or selling prices or other unfair trading conditions”.21 While this provision is generally invoked as a tool to prevent excessive pricing, its reference to “trading conditions” suggests that it can also be used to prevent the imposition of unfair terms and conditions by dominant firms.22 This observation is important in a licensing context since, as noted above, monetary payments (royalties) are generally not the only form of consideration a licensor may seek to obtain in return for granting access to its proprietary technology. However, this paper will focus on the issue of excessive royalties, leaving aside issues surrounding the imposition by dominant firms of unfair trading conditions.

21 On excessive pricing, see generally Robert O’Donoghue and Jorge Padilla, The Law and Economics of Article 82 EC, Hart Publishing, 2006, Chapter 12. 22 See, e.g., Case T-83/91, Tetra Pak International SA v Commission [1994] ECR II-755.

(Zf) Ehlermann 08 Geradin

680

6/11/08

11:56

Page 680

Session Three: Exploitative Conduct And The Remedies

While there is no doubt that DG Competition, the national competition authorities and the courts can prohibit excessive royalties under Article 82(a), this section addresses the issue of whether competition authorities should make use of Article 82(a) to place limits on the level of royalties charged by a dominant licensor to one or several licensees.23 To date, no decision of the Commission or judgment of the Community Courts has formally condemned a dominant firm for charging excessive royalties for a patent licence.24 However, this issue has become relevant due to the growing importance of IP licensing agreements in a knowledge economy and the presence of a highly publicized dispute over royalty rates in the mobile telephony sector.25 This part is divided into four sections. Section A shows that claims of excessive royalties are likely to negatively impact firms’ incentives to innovate. Section B underlines the fact that such claims also give rise to insuperable measurement problems. Section C questions whether the analysis carried out in Sections A and B should be different when proprietary technologies have become part of a standard and concludes that it should not. Finally, Section D examines which markets should be candidates for intervention to curb prices. It argues that high technology markets should not be subject to such intervention.

A. Claims of Excessive Royalty Pricing and Firms’ Incentives to Innovate Economic theory suggests that price regulation, including placing a cap on the royalties that can be charged by a licensor, will negatively impact a firm’s incentives to innovate.26 In high-tech industries, most R&D investments fail to generate marketable results, and incentives to innovate are directly linked to the prospect of generating significant profits. Hence, interventions aimed 23 See also David Evans and Jorge Padilla, “Excessive Prices: Using Economics to Define Administrable Legal Rules”, 1(1) Journal of Competition Law and Economics 97, 119–120 (2005). The authors suggest that Article 82 should not be used to sanction excessive prices on innovation markets. The Commission has on many occasions expressed its reluctance to apply Article 82 to excessive pricing claims. See Commission, Vth Annual Report on Competition Policy (1975), points 3 and 76; Commission, XXIVth Annual Report on Competition Policy (1994), point 207. Note that, since 2000, there has been only one decision in which the Commission has sanctioned a firm for excessive prices under Article 82. See Commission Decision 2001/892 of 25 July 2001, Deutsche Post AG (“Deutsche Post II”) [2001] OJ L331/40. For comment, see Peter Oliver, “The Concept of ‘Abuse’ of a Dominant Position under Article 82 EC: Recent Developments in Relation to Pricing”, 1(2) European Competition Journal 179 (2005). 24 As far as EC competition law is concerned, most IP licensing issues have arisen in the context of technology transfer agreements, which may fall within Article 81 EC. See Commission Regulation 772/2004 on the application of Article 81(3) of the Treaty to categories of technology transfer agreements, [2004] OJ L123/11. 25 See “Nokia hits back in Qualcomm dispute”, Financial Times (24 May 2007). 26 See Evans and Padilla, supra note 23.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 681

Abusive Pricing in an IP Licensing Context 681 at curbing profits affect incentives to invest. This is why the US Supreme Court stated that the ability for firms to charge supracompetitive prices is the “very essence” of the free market system,27 as it is the prospect of reaping large rewards that induces market actors to take risks, invest, innovate, and ultimately contribute to economic growth.28 Because it can affect the return on innovation and investments, competition policy may thus have a significant impact on the development of dynamically competitive industries in Europe.29 A stringent policy regarding excessive prices will have effects similar to the introduction of an upper limit on profits. Given that profits are ex ante uncertain, a firm would only be willing to invest if the expected return on its investment exceeds the cost of capital by a significant measure. The introduction of an upper bound to prices and hence to profits may thus cause a reduction in investment and a loss of dynamic competition. In addition, it could disproportionately alter firms’ ability to maintain R&D expenditures, as the borrowing capacity of a firm is generally proportional to its current earnings. Furthermore, placing a cap on dominant firms’ royalties is likely to disrupt potential competitors’ incentives to enter the market in question.30 The opportunity to charge supracompetitive prices signals to possible new entrants and investors that R&D expenditures will generate profits in the future.31 Imposing tight controls on the remuneration of innovation thus could prevent or at least significantly discourage market entry by new firms, and could hurt dynamic competition. On the other hand, it could be argued that, by virtue of “path dependence” effects, dominant firms are at any rate compelled to innovate in order to maintain their market position in the long run.32 On this view, the negative 27 See Steven Anderman, EC Competition Law and Intellectual Property Rights: The Regulation of Innovation, Oxford University Press, 1998, at p.224. 28 See Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 405 (2004): “[T]he mere position of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful, it is an important element of the free market system. The opportunity to charge monopoly prices—at least for a short period—is what attracts ‘business acumen’ in the first place; it induces risk taking that produces innovation and economic growth” (emphasis added). 29 See Ahlborn, Denicolò, Geradin and Padilla, supra note 19. 30 This argument is based on the conceptual framework provided by the Chicago school in the 1960s. The charging of high prices (and the achievement of substantial profits) at one point in time (short term) stimulates, in the following periods (mid-term) the entry of new firms into the market, and triggers a decline of the market price. The quantities supplied increase and the market price falls. In turn, the substantial profits enjoyed disappear. In fact, a high price may well be evidence of the lack of competition in the market, but it will trigger competition for the market. See Frank Easterbrook, “The Limits of Antitrust”, 63 Texas Law Review 1, 2 (1984); Harold Demsetz, “Barriers to Entry”, 72 American Economic Review 47 (1982). For a useful summary, see Richard Posner, Antitrust Law: An Economic Perspective, 2nd ed., University of Chicago Press, 2001, pp. 13–14. 31 See Feng Gu and Baruch Lev, “Markets in Intangibles: Patent Licensing” in Intangible Assets, Intellectual and Human Capital, University of Ottawa, 2003. 32 For a viewpoint, see Jonathan Baker, “Beyond Schumpeter vs. Arrow: How Antitrust Fosters Innovation” (June 2007), available at http://ssrn.com/abstract=962261, at p. 6.

(Zf) Ehlermann 08 Geradin

682

6/11/08

11:56

Page 682

Session Three: Exploitative Conduct And The Remedies

effects on firms’ incentives stemming from any such stringent application of Article 82(a) to royalty schemes could therefore be limited. However, this argument fails to take account of the fact that placing a cap on royalties may induce firms active in dynamic industries to watch and wait to see whether R&D investments made by other firms are successful and then seek to obtain access to these technologies at a controlled rate. As pointed out by Sidak, while the traditional view in microeconomic theory is that one should invest in any project that has a positive net present value of cash flows, real option theory shows that it may in fact be better to wait until some uncertainty about viability (of a newly developed technology) is resolved and cost reduction can be achieved.33 The granting of a free option on other firms’ R&D would reduce incentives to invest and decrease the level of innovation. Conversely, placing a cap on royalties may induce innovators to exploit their IP differently by, for instance, keeping their innovation for themselves as trade secrets and embedding it exclusively in their own products. The intrusion of competition law into royalty pricing could thus modify the terms of the tradeoff between producing and licensing and hence the choice made by innovators when deciding how to market their technology.34 Such choices should be driven by market factors rather than price controls. Moreover, such a reaction by licensors would have an effect opposite to the one sought by those calling for limits to be placed on royalties—allegedly to ensure wider technology transfer—since keeping technologies as trade secrets ensures that their use by third parties is prevented. More importantly, the imposition of caps on the remuneration of innovation would disproportionately affect firms without downstream operations for which royalties represent the main or only source of revenues. By interfering with the ability of firms to freely determine their royalties, competition authorities or courts could thus unwittingly contribute to eliminating firms that have legitimately opted for a licensing business model. Unlike firms operating under traditional models of vertical integration, the revenues and profits of licensing firms are not generated by the sale of products embedding new technologies, but by the licensing against royalties of such new technologies to other firms that are better able to incorporate those technologies into products. Such an undesirable outcome would deprive society of some of its most innovative companies. It would either result in their elimination or force them to vertically integrate despite the fact that their comparative strength may not 33 See Gregory Sidak, “Holdup, Royalty Stacking, and the Presumption of Injunctive Relief for Patent Infringement: A Reply to Lemley and Shapiro”, 92 Minnesota Law Review 714 (2008). 34 The decision to license rests on a tradeoff between two effects: the “revenue” effect and the “profit dissipation” effect. The revenue effect is the value of the flows of rents accruing to the innovator. The profit dissipation effect is the loss of revenues resulting from the activity of competing licensees on the downstream market when there is horizontal licensing (i.e., licensing between competitors). When prices are regulated on the basis of Article 82, the revenue effect diminishes, and the profit dissipation stays the same. On this distinction, see Arora, Fosfuri and Gambardella, supra note 1. Vertical licensing means licensing to non-rivals.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 683

Abusive Pricing in an IP Licensing Context 683 reside in manufacturing. Innovation and prices would be affected and consumer welfare would be impaired by such inefficient vertical integration.

B. Claims of Excessive Royalties and Measurement Issues The case law of the ECJ and of the Court of First Instance (“CFI”) provides some degree of guidance as to the principles applicable to measure whether a price is (or is not) excessive within the meaning of Article 82 (point 1 below). However, the principles established by the Community Courts are poorly tailored to the particular context of IP licensing (point 2).

1. Standards set by the ECJ’s case law for assessing whether the price charged by a dominant firm is excessive The criteria for assessing whether a price is “unfair” within the meaning of Article 82 were established in some of the first competition cases brought before the ECJ. In its seminal United Brands ruling, the Court held that a price is deemed “excessive” when “it has no reasonable relation to the economic value of the product supplied”.35 Importantly, the ECJ adopted the following two-step approach for determining whether a price is excessive. Specifically, one would have to: (i) “[Examine w]hether the difference between the costs actually incurred and the price actually charged is excessive”; and (ii) “[I]f the answer to this question is in the affirmative, [determine] whether a price has been imposed which is either unfair in itself or when compared to competing products”.36 In other words, a comparison between price and cost is first carried out to reveal the profit margin achieved by the dominant firm. If that profit margin is found to be “excessive”, the dominant firm’s pricing policy needs to be further analyzed, in order to determine whether the price is “unfair”. The Court’s judgment provided no further guidance on the application of this test. In particular, it did not clarify the basis on which to determine whether a price-cost difference is excessive. Similarly, it does not explain the notion of “unfair” when applying the second branch of the test. This is problematic 35 Case 27/76, United Brands Company and United Brands Continentaal BV v Commission [1978] ECR 207, at para. 250. See also para. 251: “This excess could, inter alia, be determined objectively if it were possible for it to be calculated by making a comparison between the selling price of the product in question and its cost of production, which would disclose the amount of the profit margin.” 36 Ibid. at para. 252.

(Zf) Ehlermann 08 Geradin

684

6/11/08

11:56

Page 684

Session Three: Exploitative Conduct And The Remedies

since terms such as “excessive” and “unfair” are inherently vague and devoid of meaning in the absence of a precise economic test to determine whether a given price falls under their scope.37 Unfortunately, subsequent cases referred to the ECJ only led to sporadic pronouncements on the methods applicable for establishing an excessive price within the meaning of Article 82. The Court even seemed to relinquish the United Brands two-stage method in favour of a more “integrated” benchmarking test. In a first strand of cases, the Court compared the pricing policy of a dominant firm with the prices of equivalent firms active on neighbouring geographic markets.38 In a second strand, the Court undertook to make comparisons between the prices charged by the same dominant firm (i) to different customers and (ii) over time.39 It has thus been difficult to find consistency in the standards promoted by the ECJ.40 The most recent decision of the Commission in this area suggests that the two-stage test set forth in United Brands remains the relevant analytical framework for assessing whether a price is excessive. In this case, Scandlines Sverige AB v Port of Helsingborg, the Commission recalled that evidence of an “excessive” profit margin was not sufficient in itself to establish an abuse.41 It 37 In addition, the Court introduced further complexity by indicating in obiter dicta that other methods could be devised to determine whether a price is unfair. Id. at para. 253. National courts and competition authorities could thus approach excessive price allegations through a variety of methods not necessarily mentioned by the Court in United Brands. 38 See Case 110/88, Lucazeau and others v SACEM and others [1989] ECR 2811, para. 25: “When an undertaking holding a dominant position imposes scales of fees for its services which are appreciably higher than those charged in other Member States and where a comparison of the fee levels has been made on a consistent basis, that difference must be regarded as indicative of an abuse of a dominant position”. See also Case 30/87, Corinne Bodson v SA Pompes funèbres des régions libérées [1988] ECR 2479. In Bodson, the Court stated that, in order to determine whether prices are unfair, “it must be possible to make a comparison between the prices charged by the group of undertakings which hold concessions and prices charged elsewhere”. This test had already been implicitly referred to in Case 78/70, Deutsche Grammophon Gesellschaft mbH v Metro-SB-Großmärkte GmbH & Co KG [1971] ECR 487. 39 See Case 226/84, British Leyland Plc v Commission [1986] ECR 3263, paras. 27–28, where the Court recalled (along the lines of the language used in United Brands) that a price is excessive where it is “disproportionate to the economic value of the service provided”. However, the Court concluded that the dominant firm’s prices were excessive because the price differential between the various services in question was not proportionate to the minimal cost differences between several services. A similar standard had already been applied in Case 26/75, General Motors Continental NV v Commission [1975] ECR 1367, para. 12. 40 The lack of clarity of the case law is further aggravated by isolated rulings applying a different methodology. See, e.g., Case T-89/98, National Association of Licensed Opencast Operators (NALOO) v Commission [2001] ECR II-515, para. 72. In this case, the CFI applied an “efficient demand” benchmark, i.e., it checked whether the dominant firm’s efficient customers could still achieve profits without suffering a competitive disadvantage. 41 Commission Decision of 23 July 2004, Case COMP/36.568, Scandlines Sverige v Port of Helsingborg. This decision arose from a complaint brought by Scandlines Sverige AB, a ferry operator active on the Helsingborg (Sweden)—Elsinore (Denmark) route, according to which the pricing policy of the port of Helsingborg infringed Article 82. At paragraph 158 of the decision, the Commission states: “In any event, even if it were to be assumed that the profit margin of HHAB [the dominant firm] is high (or even ‘excessive’), this would not be sufficient to conclude that the price charged bears no reasonable relation to the economic value of the services

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 685

Abusive Pricing in an IP Licensing Context 685 underlined that it was bound to prove the existence of an “unfair” price pursuant to the second limb of the United Brands principle. Arguably, this is where the “integrated” benchmarking approach becomes relevant.

2. The practical difficulties of applying the case law standards in an IP licensing context Excessive pricing is one of the most controversial issues in the field of EC competition law. In addition to the valid argument that competition authorities and courts should not engage in price control, one reason for the controversial nature of this area of the law lies in the insuperable practical difficulties encountered in ascertaining whether a price is excessive, and the potentially enormous consequences of an erroneous determination.42 The intricacy of ascertaining the “correct” or “competitive” price for a given product is exacerbated in the case of intangible assets such as IP.43 In substance, three main criticisms have been put forward by competition lawyers and economists.44 These are discussed below. a. Finding an adequate cost measure For the purposes of applying the first limb of the United Brands standard, a major difficulty lies in the determination of the dominant firms’ costs which provided. The Commission would have to proceed to the second question as set out by the Court in United Brands, in order to determine whether the prices charged to the ferry operators are unfair, either in themselves or when compared to other ports” (emphasis added). 42 These difficulties have been acknowledged by Philip Lowe, Director General of DG COMP: “On exploitative abuses, there is widespread criticism, some of which we concur with. For example, it is extremely difficult to measure what constitutes an unfair or excessive price”. Speech delivered by Philip Lowe at the Fordham Antitrust Conference in New York, 23 October 2003, available at http://ec.europa.eu/comm/competition/index_en.html. See also Emil Paulis, “Article 82 EC and exploitative conduct”, this Volume (“Determining whether a price is excessive may also involve difficult comparisons with whatever useful ‘benchmark’ prices can be identified. Some of the problems involved in these comparisons—for example, the issue of cost allocation in multi-product firms—are also present for other price-based abuses. However, when these problems are ‘solved’ for the other abuses, the price/cost question becomes relatively ‘simple’ in that the issue is whether the price is higher or lower than some well-defined cost measure. To determine whether excessive pricing has taken place, there is another layer of complication since it has to be decided whether a price—which may be higher than all relevant cost measures— is in fact too high. According to many commentators, such a decision will necessarily be somewhat arbitrary, unless one takes the rather draconian position that any price over some well-defined cost benchmark is excessive.”). 43 For surveys of the theoretical literature, see Morton Kamien, “Patent Licensing,” in Robert Aumann and Sergiu Hart, eds., Handbook of Game Theory with Economic Applications, Vol. 1, North-Holland, 1992, pp. 331 et seq. See also Susanne Scotchmer, “Licensing, Joint Ventures, and Competition Policy,” in Susanne Scotchmer, Innovation and Incentives, MIT Press, 2004. 44 For a full account of these criticisms, see David Evans and Jorge Padilla, supra note 23. See also Alexandre De Streel and Massimo Motta, “Excessive Pricing and Price Squeeze under EU Law” in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 91 et seq.

(Zf) Ehlermann 08 Geradin

686

6/11/08

11:56

Page 686

Session Three: Exploitative Conduct And The Remedies

need to be taken into consideration.45 Economic theory suggests that the appropriate cost measure is the dominant firms’ marginal cost (“MC”), or its average variable cost per unit (“AVC”).46 However, for obvious reasons it would be nonsensical to use these cost benchmarks in an IP licensing context. While innovation generates very high fixed costs, the MC and AVC of granting a single licence are indeed equal or close to zero. Alternative cost benchmarks must therefore be found. In that respect, the relevant cost measure should probably factor in the R&D expenditures of the dominant firm. But this again would raise considerable difficulties. First, there is the question of which R&D costs should be taken into account. Considering only the R&D costs directly linked to the development of a given technology would be under-inclusive, as innovative firms usually have to engage in dozens of research projects to develop one successful technology.47 The costs of failed projects would thus have to be taken into account.48 Another difficulty arises from the fact that R&D expenditures are typically “common costs” when the dominant firm is also active on downstream manufacturing markets.49 Hence, while only part of the R&D costs should be allocated to licensing activities, finding an adequate method of allocating between manufacturing and licensing activities may prove insuperable.50 Finally, on technology markets where “incremental innovations” (minor evolutions of existing technologies) are protected by IPRs, the question arises whether the R&D expenditures incurred for the existing technology should be factored into the analysis. b. Setting the level where a profit becomes “excessive” The definition of what constitutes an “excessive” profit in the meaning of the first limb of the United Brands standard also lacks clarity. The Commission 45 In addition, dominant firms often have difficulties in providing data regarding their costs. See, e.g., Manuel Martínez, “Some Views on Pricing and EC Competition Policy”, mimeo, available at http://ec.europa.eu/comm/competition/speeches/text/sp1998_060_en.html, at p. 6. 46 Economists also refer to AVC as the incremental cost of production. See O’Donoghue and Padilla, supra note 21, p. 614. The relevant questions, then, are: whether to allocate common costs (general expenditures), indirect costs, etc.; and what share to assign to each. 47 See Technology Transfer Guidelines, supra note 8. 48 See Paulis, supra note 42 (“[I]nvestment costs should be taken into account when determining whether prices are excessive”). 49 This point has also been made by Steven Anderman and John Kallaugher, Technology Transfer and the New EU Competition Rules—Intellectual Property Licensing after Modernisation, Oxford University Press, 2006. 50 See Michal Gal, “Monopoly Pricing as an Antitrust Offense in the U.S. and the EC: Two Systems of Belief About Monopoly?” 49 Antitrust Bulletin, 343–384 (2004). See also the practical difficulty underlined in Shigeki Kamiyama, Jerry Sheehan and Catalina Martinez Valuation and Exploitation of Intellectual Property, STI Working Paper 2006/5, at p. 13: “[A]ccounting guidelines and corporate disclosure rules do not require firms to break out IPR-related revenues from other sources of income”. The paper is available at http://www.oecd.org/dataoecd/62/52/ 37031481.pdf.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 687

Abusive Pricing in an IP Licensing Context 687 and the Courts have indeed (rightly) omitted to quantify a threshold above which profits become excessive. The case law nonetheless indicates that dominant firms will only be sanctioned when their profit margin is “grossly exorbitant”.51 A common thread to all the cases is that Article 82 has been applied only when prices exceeded costs by more than 100% of the value of the product/service in question.52 Yet, in the particular case of dynamic industries, such a margin in respect of the dominant firms’ profits is still overly restrictive.53 First, the innovation process is akin to a painful “trial and error” process. As noted above, firms generally experience a number of setbacks prior to obtaining a successful patent which can be licensed. Worse, innovators often incur huge R&D investments which never lead to the award of a patent and even when a patent is granted there is no guarantee that it will be commercially significant.54 The upshot of this is that when firms hold successful patents, setting royalties well in excess of R&D costs is a perfectly rational and efficient pricing policy, which compensates for failed R&D investments and provides, in turn, incentives for further risky investment.55 Moreover, determining whether a price is “excessive” would make it necessary to forecast the market evolution when appraising dominant firms’ profits. It is when a market grows fast that concerns about excessive profits typically arise (as royalties are often generally computed ad valorem on the basis of sales achieved by the licensee).56 However, the final level of profit is very often more limited than initially expected. The erosion of profits due to a fall in market demand is also compounded by the limited lifetime of innovations, which may be shorter than the life of the relevant patents due, for instance, to rapid technological obsolescence, the entry of new firms on the technology market, etc. In other words, the royalties charged by licensors may generate substantial profits, but only for a period of time, which will be limited and of uncertain length, and thus hard to evaluate by competition

51 See Eric Pijnhacker Hordijk, “Excessive Pricing under EC Competition Law; An Update in the Light of ‘Dutch Developments’ ”, in Barry Hawk, ed., Fordham Corporate Law Institute, Juris Publishing, 2002, pp. 463 et seq. See also John Temple Lang and Robert O’Donoghue, “The Concept of an Exclusionary Abuse under Article 82 EC”, GCLC Research Papers on Article 82, July 2005, mimeo. Temple Lang and O’Donoghue explain at page 39 that Article 82 EC “arguably applies only in cases where there are significant barriers to entry that cannot be overcome by investments in anticipation of monopoly rents”. Their paper is available at http://gclc.coleurop.be. 52 See Pijnhacker Hordijk, cited in previous footnote. 53 This of course assumes that an adequate cost measure can be found. 54 This ties in to the difference between rents and quasi-rents. As pointed out by Sidak, supra note 33, “the latter is the risk-adjusted return to sunk investment made in risky activities; it may look excessive ex post, but only because one already has turned the cards over and knows with certainty what was unknown at the time that bets had to be laid”. 55 For a similar argument, see Anderman and Kallaugher, supra note 49, at § 10.17, p. 273. 56 See Gu and Lev, supra note 31 at p. 4. The authors note that markets for patents are expected to grow fast.

(Zf) Ehlermann 08 Geradin

688

6/11/08

11:56

Page 688

Session Three: Exploitative Conduct And The Remedies

authorities and courts seeking to determine whether a royalty is excessive within the meaning of United Brands. c. Identifying the appropriate benchmarks In the context of IP licensing transactions, the various benchmarks that have been applied by the Commission and the EC Courts to determine whether a price is “unfair”, pursuant to the second limb of the United Brands standard, are seriously flawed. The deficiencies of these benchmarks are discussed below. The historical costs benchmark. In British Leyland, the ECJ undertook a comparison between the prices of the dominant firm and the prices it had charged in the past.57 The Court found that the fees had increased 600% during the relevant period, and on this basis it considered the fees to be abusive. The application of this principle to dynamic markets could prove dangerous. First, the availability of an equivalent comparator in the past is not guaranteed, as an IPR is ex hypothesi unique. A comparator could arguably be found in expired licence agreements for a similar technology which have been replaced by new licence agreements with different royalty provisions. However, in such case, competition authorities and courts run the risk of comparing apples and oranges, i.e., they risk comparing licence terms and conditions negotiated in the context of different market situations. In a number of dynamic industries, IP holders interested in fostering the uptake of their technology may initially opt for a low-royalty policy (a strategy called “penetration pricing”). At a later stage, when the technology is well implanted and mature, licensors may then seek to increase their royalty rates in order to recoup part of the low prices charged in the past. The geographic benchmark. In United Brands and Bodson, the ECJ compared the prices of a given product over different neighboring markets. However, it is unclear whether a similar methodology should be applied to IP licensing transactions. While geographic benchmarking assumes the identification of distinct geographic markets, technology markets will often be EUwide or worldwide, thereby rendering the identification of separate geographic markets impossible.58 In addition, if local markets are delineated, the crux of the problem lies in finding two (or more) equivalent or at least comparable markets, a situation which is unlikely to occur in practice. Finally, if the royalty price on the compared market is itself excessive, bringing evidence of an abuse will simply become impossible.59 The competitors benchmark. There are both conceptual and practical objections to comparing the royalty charged by dominant firms with the royalties 57 See British Leyland Plc v Commission, supra note 39. See also Frank Fine, The EC Competition Law on Technology Licensing, Thomson—Sweet & Maxwell, 2006, at p. 124 § 6.13. 58 See Anderman and Kallaugher, supra note 49, at p. 273 § 10.17. 59 See O’Donoghue and Padilla, supra note 21, at p. 617.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 689

Abusive Pricing in an IP Licensing Context 689 charged by its competitors. As seen above, IP rights, unlike the bananas at stake in United Brands, present ex hypothesi unique features. It thus seems difficult to identify one or several IP holders with a comparable patent or set of patents for the purpose of determining the excessive character of the dominant firm’s pricing policy.60 But even assuming that two firms had comparable IP, the differences in the royalty rate charged to their respective licensees would not signal that the firm with the higher rate has committed an abuse. Indeed, as will be discussed below, because each instance of licensing negotiations is unique and typically provides for other forms of consideration in addition to a royalty (e.g., cross licences, etc.), no inferences can be drawn from differences in royalty rates without an in-depth look at the other terms and conditions in the licensing contract.

C. Excessive royalties in the context of standardization One interesting question is whether the above analysis should be different when proprietary technologies have become part of a standard. As we have seen above, by ensuring compatibility between products, standardization generates significant welfare benefits. However, achieving product compatibility through standardization usually entails making choices, the effects of which will represent a cost. While standards increase downstream competition between implementers, they may also constrain the choice between technological options and reduce competition between technology developers.61 Moreover, when the technologies involved are covered by IPRs, the adoption of standards may also raise issues related to access.62 As standards often include proprietary technologies, those wishing to implement a standard should obtain licences from all the essential patent holders. Given the significant stakes frequently involved, the outcome of the discussions over which technologies should be incorporated into any given standard has occasionally strained the standard-adoption process.63 Some tension is inevitable as each firm desires to promote its own solutions as part of the standard but also needs to work together with other SSO members to develop, establish, endorse, and promote the standard.64 Another factor 60 See Maurits Dolmans, “Standards for Standards”, 26 Fordham International Law Journal 163, 202 (2002). 61 On the other hand, standardization promotes competition within a standard, i.e., between products implementing the standard. See Teece and Sherry, supra note 17, at 1915. 62 See Carl Shapiro, “Setting Compatibility Standards: Cooperation or Collusion?” in Rochelle Dreyfuss, Diane Zimmerman and Harry First, eds., Expanding The Bounds Of Intellectual Property, Oxford University Press, 2001, at Section III. 63 For case studies, see Brian DeLacey, Kerry Herman, David Kiron and Josh Lerner, “Strategic Behavior in Standard-Setting Organizations” (2006), at http://papers.ssrn.com/sol3/ papers.cfm?abstract_id=903214. 64 See Shapiro, supra note 62, at 1–2.

(Zf) Ehlermann 08 Geradin

690

6/11/08

11:56

Page 690

Session Three: Exploitative Conduct And The Remedies

contributing to SSO tensions relates to the fact that firms involved in standard-setting often wear different “hats” corresponding to the fundamentally different business models they adopt.65 As Geradin and Layne-Farrar have shown elsewhere, firms participating in standardization activities do not necessarily share symmetrical incentives.66 While, for instance, pure innovators (e.g., firms which do not engage in manufacturing activities) are entirely dependent on licensing revenues to continue their operations, verticallyintegrated operators may be more interested in protecting their downstream manufacturing operations through cross-licensing than they are in collecting royalties on their essential IP. In light of these widely acknowledged tensions, most formal SSOs have written IPR policies whose primary goal is to ensure adequate disclosure and subsequent availability through licensing of IP rights incorporated into a standard.67 Although their scope may vary significantly across SSOs, the procedures put in place usually seek to encourage essential IP owners to make their proprietary inventions known and available to other SSO members and/or other implementers of the standard.68 To this effect, most SSOs encourage IP owners involved in standardization to disclose upfront, i.e., prior to the adoption of a standard, the IPRs that they consider may be “essential” for its implementation.69 Once disclosure is made, or contemporaneously with disclosure, IP owners are typically asked to provide an assurance or undertaking that, should their IP turn out in fact to be essential for the final standard, they will make licences to that IP available on fair, reasonable and non-discriminatory (FRAND) terms and conditions to other members of the SSO and, as is often the case, to outsiders.70 Most SSOs do not require such commitments—which could be interpreted as compulsory licensing—but if the owner of potentially essential IP seeks to have its technology included in a standard it has a strong incentive to provide the SSO with the assurance that it will license on FRAND terms and conditions. A FRAND commitment is intended to prevent an outright refusal to license or the setting of royalty rates and other terms and conditions so unusually high as to suggest an intent by the IPR owner to do indirectly what it has committed not to do directly, i.e., refuse to license its essential IP to other 65

See Teece and Sherry, supra note 17, at 1929. Damien Geradin and Anne Layne-Farrar, “The Logic and Limits of Ex Ante Competition in a Standard-Setting Environment”, 3(1) Competition Policy International 78 (2007). 67 See Lemley, supra note 11 at 20–21. 68 Id. 69 ETSI defines “Essential IPR” as meaning “that it is not possible on technical (but not commercial) grounds, taking into account normal technical practice and the state of the art generally available at the time of standardization, . . . [to] comply with a standard without infringing that IPR.” ETSI IPR Policy (version of 23 November 2005), Article 15. 70 See Lemley, supra note 11, at 26. The ETSI IPR Policy, for example, provides that IPR holders should be rewarded properly, explicitly recognizing that they “should be adequately and fairly rewarded for the use of their IPR”. See ETSI IPR Policy, cited in previous footnote, Article 3.2. 66

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 691

Abusive Pricing in an IP Licensing Context 691 firms (a constructive refusal to license). The FRAND commitment therefore entails a promise by the IP owner that it is prepared to engage in good faith negotiations with any company wishing to implement the standard with a view to reaching a licensing agreement that will be defined in light of all circumstances prevailing between the two parties at the time of the negotiations. While SSOs have significantly contributed to the development of, and the growing competition within, high technology sectors, some commentators nonetheless believe that the current disclosure and FRAND licensing commitments are inadequate or ill-tailored to meet current needs.71 They argue that standardization allows essential IP holders to act opportunistically and that commitments to license on FRAND terms are not sufficient to prevent such opportunistic behaviour (point 1 below). This has led some scholars and firms to reinterpret FRAND as imposing some constraints on the ability of patent holders to monetize their essential IP (point 2).

1. The hold up problem One of the criticized pitfalls of the current FRAND regime is the alleged risk that owners of IP essential to a standard will be able to unduly capture some of the economic value that may be attributable not to the intrinsic value of those rights but to standardization itself.72 It is argued that if members of an SSO had known, prior to a standard being set, the terms under which essential IP owners would license their rights, they might have chosen an alternative technology (provided, of course, such alternative technology existed—which is not a given73). But once the standard has been adopted and implemented, switching to an alternative technology may have become too onerous for those practicing it. The argument continues that the bargaining power of the owner of essential IPRs will have thus increased and that it may be able to extract more favourable licensing terms following standardization than would otherwise have been the case.74 This phenomenon, which can be 71 See, e.g., Gil Ohana, Marc Hansen and Omar Shah, “Disclosure and Negotiation of Licensing Terms Prior to Adoption of Industry Standards: Preventing Another Patent Ambush”, 24 European Competition Law Review 644 (2003); Robert Skitol, “Concerted Buying Power: Its Potential for Addressing the Patent Holdup Problem in Standard Setting, 72 Antitrust Law Journal 727 (2005). 72 See Lemley and Shapiro, supra note 5. 73 See Teece and Sherry, supra note 17 at 1938–39 (“Whether the SSO would have in fact adopted another alternative had it known of the patent claims raises a complex counterfactual question: ‘What would the SSO have done if the world had been different?’ The answer is likely to be hotly debated, and depends on the particular facts of the standard at issue. The greater the advantages of the (patented) standard over the alternatives that were considered and rejected at the time the standard was originally set, the less likely it is that an alternative would, in fact, have been chosen.”) 74 Doug Lichtman, “Patent Holdouts in the Standard-Setting Process”, University of Chicago, Law and Economics Olin Working Paper No. 292 (May 2006), available at http://papers.ssrn. com/sol3/papers.cfm?abstract_id=902646; Shapiro, supra note 62 at 19–20.

(Zf) Ehlermann 08 Geradin

692

6/11/08

11:56

Page 692

Session Three: Exploitative Conduct And The Remedies

described as ex post opportunism, would lend credence to the need to control the level of royalties charged by holders of essential IP. Attractive at first blush, the theory of ex post opportunism overlooks several critical issues. The first is that this theory is based on the premise that alternative technologies existed at the time of adoption of a particular standard and that the successful technology would not have been chosen due to the licensing disparity.75 In many instances of standard development, however, no sufficiently attractive alternative technology exists. In the absence of substitute technologies, it cannot be argued that the standard-setting process gives additional market power to the IP holder: the technology had no competition either before or after the vote on standards is held. Such market power already existed prior to adoption of the standard and is due to the uniqueness of the technology in question. Fundamental economics maintains that firms with a unique product or IP will be in a stronger position than those with products or IP for which alternatives exist. The fact that the IP is embedded in a standard does not confer additional market power. Instead, what standardization might do is increase the value of the IP by allowing its holder to collect royalties on larger volumes of products complying with the standard, but this is a direct consequence of the adoption of a standard rather than of any opportunistic behaviour on the part of the patent holder. As noted above, firms holding patents relevant for a standard also face a number of important constraints. Regardless of whether the patented technology faces viable substitutes, its licensing price is constrained by the prices commanded by complementary patents within the standard.76 That is, patent prices are limited by their context. In addition, patent holders without any downstream operations (upstream firms) are constrained by the elasticity of demand for the standard-compliant product in the end market.77 While verticallyintegrated firms can have incentives to raise the prices paid by rival downstream firms through their licensing terms, they may also be open to cross-licensing agreements with other integrated companies, which can hold down royalty rates as well.78 And lastly, all firms face dynamic constraints inherent to the formal standard-setting process. Because standards evolve over time, and because many high technology standards pass through multiple versions (for instance, mobile telecommunications is currently passing through its “third generation” (3G), with successive generations already under development), any unreasonable pricing or abuse of market power can be punished in

75

See Teece and Sherry, supra note 17 at 1939. See Geradin, Layne-Farrar and Padilla, supra note 6. 77 Klaus Schmidt, “Licensing Complementary Patents and Vertical Integration” (November 2006), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=944169. 78 Note, however, that royalty free cross-licences between vertically-integrated firms do not ensure lower prices to consumers, as those firms will have to recover their R&D costs in addition to the other costs they incur (e.g., from manufacturing) in their prices to consumers. 76

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 693

Abusive Pricing in an IP Licensing Context 693 future iterations of the standard.79 Firms that act opportunistically in today’s version of a standard may find their technologies excluded, avoided, or at least minimized in votes on tomorrow’s version of the standard. Another overlooked issue relates to the question why, if standardization increased the value of a given IP, the essential patent holder should not capture part of that value. The implicit assumption in the ex post opportunism claim is that all of the additional value created by the standardization process improperly accrues to patent licensors. But formal standardization is a costly cooperative effort that requires both innovators and implementers. There is no reason to assign all of the rents to one or the other. Thus, while owners of IP may benefit from a broader adoption of their technologies, implementers—as well as consumers—also benefit from the opportunity to gain access to and use innovative superior technologies. This sharing of benefits helps to ensure participation incentives. Leaving the above considerations aside, for a royalty rate to be considered excessive under Article 82(a), the test established by the ECJ in United Brands must be met and the fact that the price of essential IP may have somewhat increased as a result of standardization plays no part in this test. If at all, the fact that royalties may have been increased as a result of standardization would add a further element of complexity in the application of the United Brands test, as it would require a determination of what would have been the “appropriate” level of a given royalty “but for” the fact that the technology in question had benefited from standardization.

2. Royalty stacking Royalty stacking as a theoretical concept can be explained simply. A firm wishing to produce a good, especially one embodying a technical standard, typically needs to acquire rights to the intellectual property underlying the good. When that good is comprised of multiple complementary components, each of which is necessary for production and each of which is covered by patents held by separate firms, the aggregate royalty fees for licensing all of the required pieces can, it is sometimes suggested, add up to a very large amount—perhaps so large that it is no longer economical for the manufacturing firm to make the good.80 This can allegedly happen even if each component’s patent is offered on “reasonable” terms. Stacking up so many reasonable terms could lead to an unreasonable sum. Advocates of the royalty-stacking theory have thus made various proposals to tackle it, such as placing a cap on the aggregate royalty rates that could be charged by

79 For a discussion of such dynamic and institutional constraints, see DeLacey et al., supra note 63. 80 See Lemley and Shapiro, supra note 5.

(Zf) Ehlermann 08 Geradin

694

6/11/08

11:56

Page 694

Session Three: Exploitative Conduct And The Remedies

essential patent holders, as well as introducing mechanisms to apportion those royalty rates among essential patent holders.81 In a recent paper, Anne Layne-Farrar, Jorge Padilla and I have shown that the royalty stacking theory as developed by Lemley and Shapiro was not based on serious empirical evidence and rests on assumptions that limit its applicability.82 The 3G mobile industry, which was presented by these authors as an example of a sector where royalty stacking prevailed, was and is not in fact characterized by excessive cumulative royalty rates. Our conclusion was that royalty stacking is far less prevalent than assumed. Moreover, as I have also shown in a separate paper, the proposed mechanisms to cap aggregate royalty rates and apportion royalties among patent holders find no basis in law or economics, and their main objective is in fact to hurt firms operating under a licensing business model.83 Many firms—such as pure manufacturers (which do not have significant IP but need to have access to technologies developed by others) or verticallyintegrated operators (which do have significant IP but may make their profits downstream) share an interest in seeing a decrease in royalties. This would save them costs and, in the case of vertically-integrated operators, would eliminate competitors in future innovation races. However, while there is no evidence that any savings on royalties they could achieve would necessarily translate into lower customer prices (as this depends on a number of factors such as the level of competition on downstream product markets), drastic cuts on royalties would effectively eliminate firms whose innovation is mainly funded by licensing revenues. But even if royalty stacking were an issue, no rational interpretation of Article 82(a) could force a firm to reduce its royalty rates on the ground that these rates combined with the rates charged by producers of complementary inputs (i.e., other licensors) would make the price of the product for which these inputs were needed unreasonable. If, for instance, an automobile maker decided to construct a car whose components (Daimler Benz chassis, Ferrari engine, and Rolls-Royce interior design) were so expensive that its price would make it un-sellable, there would be no legal basis for it to claim that the makers of these components should cut their respective prices to a level that would make the car sellable. Mutatis mutandis, the fact that—for good or bad reasons—a standard is based on such a wide range of proprietary technologies that it is too expensive to implement does not give implementers—most of which participated in the creation of the standard—a claim under Article 82(a) that essential patent holders should reduce their rates to a level that will make the standard less costly to implement. Of course, it is in the standard members’ best interests to see the standard become commercially successful, 81 82 83

See Geradin, supra note 4. See Geradin, Layne-Farrar and Padilla, supra note 67. See Geradin, supra note 4.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 695

Abusive Pricing in an IP Licensing Context 695 which provides incentives for the IP holders to take other royalty rates into consideration. As pointed out by Bekkers, one of the most significant problems facing standards is their over-inclusiveness.84 The efforts of many firms participating in standardization to force their technology into a standard have the effect of making that standard more costly and thus hard to implement. Ensuring greater discipline in terms of what goes in or stays out of a standard offers a more promising solution than placing artificial caps on royalties (and thus profits) of firms that contribute to the value of standards.

3. The implications of a FRAND commitment In light of the discussion above, an interesting question is whether the making of a FRAND commitment by a dominant firm should modify the assessment competition authorities or courts could be called upon to make under Article 82(a). The answer can only be negative. The test to determine whether a price is excessive was developed by the ECJ in United Brands and it still represents good law. The test requires a demonstration that “the difference between the costs actually incurred and the price actually charged is excessive” and if this is the case, it must further be shown that the price that has been imposed is “unfair”. The fact that the dominant firm in question has committed to license its essential IP on FRAND terms does not assist in this enquiry, especially since the terms “fair” and “reasonable” are no more specific than the concepts referred to by the ECJ in United Brands. In case of disagreement between an essential IP holder and a potential licensee over whether an offer made by the former is in compliance with its FRAND commitment, the latter is free to seek a contractual remedy. Note, however, that even in this case, the fact that a potential licensee is unhappy with the royalty rate (and/or other licensing terms) proposed by the essential IP holder does not in itself amount to a breach of FRAND. Indeed, a FRAND commitment cannot mean an obligation on the essential patent holder to license its IP at the rate preferred by the potential licensee. Otherwise, claims of unreasonable licensing terms would merely reflect a desire by the prospective licensee to avoid having to take a licence on terms it simply does not like. Puzzlingly, some authors have argued that the failure of an essential patent holder to make a FRAND offer to a potential licensee could amount to a violation of Article 81.85 The reasoning, on this view, is that standard 84 See Rudi Bekkers, “Patent drag and stacking IPR fees—Are the IPR policies of standards bodies failing or should we better address excessive technology inclusion?” Position paper for the High-Level Workshop on standardization, IP licensing and antitrust organized by the Tilburg Law and Economic Center (TILEC), Tilburg University at Chateau du Lac, Brussels (January 2007), available at http://home.tm.tue.nl/rbekkers. 85 See Dolmans, supra note 60.

(Zf) Ehlermann 08 Geradin

696

6/11/08

11:56

Page 696

Session Three: Exploitative Conduct And The Remedies

agreements between competitors would (i) fall under Article 81(1)—a position which can be criticized in itself given the pro-competitive features of standardization agreements—and (ii) could only be justified under Article 81(3) provided that essential patent holders make a FRAND commitment as this would be the only means to prevent anticompetitive hold-up. This approach obviously fails to convince. An essential patent holder’s refusal to offer a licence on FRAND terms would plainly be a unilateral act and it is a fundamental tenet of EC competition law that unilateral acts do not fall within the scope of Article 81.86

D. Which markets are candidates for intervention, and should high-tech markets be among them? “Which markets are candidates for intervention?” is the title of one of the sections of the paper prepared by Emil Paulis for this Workshop.87 This is of course a key question, which is explored in this section. The focus of the Commission’s recent efforts to modernize the application of Article 82 is on exclusionary abuses. By contrast, exploitative abuses are entirely left out of the Discussion Paper. This tends to suggest, and this has been confirmed by Commissioner Kroes,88 that the Commission’s primary concern is with the prevention of exclusionary abuses and the need to adapt current thinking on such abuses to bring it more in line with economic theory. Yet, as clearly expressed by Mr Paulis, the Commission wants to retain the ability to apply Article 82(a) in some markets.89 But which markets? Mr Paulis correctly observes that the primary candidates for intervention against excessive prices are markets with “very high and long-lasting barriers to entry and expansion”.90 There is no doubt that markets characterized by natural monopolies are obvious candidates for intervention, although in most cases price control will be carried out by sector-specific regulators. But where does this leave us with respect to high technology markets such as the ones where IP licensing tends to prosper? Mr Paulis notes that, in “many markets with considerable investment and innovation, barriers to entry may be high, but not necessarily long-lasting”.91 86 If an act is unilateral, there is no agreement or meeting of minds between two (or more) parties within the meaning of Article 81. See Case 41/96, Bayer AG v Commission [2000] ECR II-3383. 87 See Paulis, supra note 42. 88 See Neelie Kroes, “Tackling Exclusionary Practices to Avoid Exploitation of Market Power: Some Preliminary Thoughts on the Policy Review of Article 82”, 29 Fordham International Law Journal 593 (2005). 89 See Paulis, supra note 42. 90 Ibid. 91 Ibid.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 697

Abusive Pricing in an IP Licensing Context 697 The way high technology markets have evolved these last twenty years amply illustrates this point. The video game industry, for instance, has witnessed cut-throat competition between firms, such as Nintendo, Sega, Sony, and more recently Microsoft, whose market shares and profits fluctuated depending on which of them had the “must have” consoles and games at any given time.92 The mobile telephony industry has similarly gone through three generations of standards since the arrival of the first handsets and fourth generation standards are about to emerge. While such market evolutions create opportunities for some firms, they may also threaten others. Market shares and profits are unstable. Firms licensing proprietary technologies may draw substantial rents, but these rents are always temporary not only because they may end with the expiry of their patents, but also because such technologies will inevitably be made redundant due to technological innovation (or, in the case of standardization, when old standards are replaced by new ones to reflect such innovation). It is also interesting to observe that DG Competition has so far never adopted an Article 82(a) decision in a high technology industry. Even in the market for computer operating systems, which is characterized by high rents, the Commission has not sought to apply Article 82(a)—and fortunately so. Its recent efforts to curb Microsoft’s proposed royalties for the licensing of interoperability information do not seem to be motivated by a willingness to control rates on the ground that they would be exploitative, but by a desire to prevent allegedly exclusionary behaviour from occurring, and to ensure compliance with a prior decision.93 Yet even with regard to markets which would prima facie appear to be candidates for intervention, it is suggested that competition authorities should take a number of factors into consideration before launching an investigation of alleged excessive pricing, including the availability of: • Adequate benchmarks allowing the assessment of whether a price (or for that matter a rate) is excessive. Short of such benchmarks, determining the excessiveness of a price amounts to guesswork; • An adequate and administrable remedy. As pointed out by Mr Paulis, some price remedies may force a competition authority “to come back time and again to the pricing of the dominant firm when cost or other conditions change in the industry”, thereby “finding itself in the situation of a semipermanent quasi-regulator”;94 and 92 See, e.g., Daniel Rubinfeld, “Competition, Innovation, and Antitrust Enforcement in Dynamic Network Industries” Address Before the Software Publishers Association (1998 Spring Symposium) San Jose, California, 24 March 1998, available at http://www.usdoj.gov/atr/public/ speeches/1611.htm. 93 See Commission Press Release IP/07/269 of 1 March 2007, “Commission warns Microsoft of further penalties over unreasonable pricing as interoperability information lacks significant innovation”. 94 See Paulis, supra note 42.

(Zf) Ehlermann 08 Geradin

698

6/11/08

11:56

Page 698

Session Three: Exploitative Conduct And The Remedies

• Sufficient human resources to properly staff such an investigation. Experience with sector-specific regulators shows that price control investigations may take years to complete and absorb considerable resources. Given the limited resources of competition authorities and the conceptual and practical difficulties raised by the application of Article 82(a), a central question is whether it is good policy for an authority to invest its scarce resources in this type of investigation.

IV. The application of Article 82(c) to licensing agreements Price discrimination seems to be ubiquitous in technology licensing. This is due in particular to the fact that many IP licensing agreements have an element of cross-licensing and the size of the portfolios of potential licensees tends to vary considerably. Moreover, as already noted, a great variety of factors will generally be taken into consideration to set the royalty rate applicable to a given licensee. Thus, forcing licensors to offer identical royalties (or, more generally, identical licensing terms) to their licensees would prevent efficient discrimination and discourage innovation, as licensees would be unable to extract proper value for their own portfolios. Worse, it would introduce undue rigidity in IP licensing and reduce opportunities for licensors and licensees to reach mutually acceptable deals, thus negatively affecting technology transfer and entry into downstream markets. Against this background, it is nevertheless interesting to explore how Article 82(c) could apply to IP licensing agreements. An immediate difficulty with Article 82(c) is that it does not provide a definition of price discrimination.95 It simply states that it is an abuse for one or several firms holding a dominant position to apply “dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage”. The ECJ has extended this notion of abuse to the converse situation of the application of similar conditions to unequal transactions.96 Article 82(c) as interpreted by the ECJ thus means that some forms of price discrimination may be regarded as abuses of a dominant position. 95 Scholars have, however, provided economic tests helping to identify price discrimination. For instance, in his famed antitrust book, Judge Posner explains that: “Price discrimination is a term that economists use to describe the practice of selling the same product to different customers at different prices even though the cost of sale is the same to each of them. More precisely, it is selling at a price or prices such that the ratio of price to marginal costs is different in different sales [. . .]”. Richard Posner, Antitrust Law, 2nd ed., University of Chicago Press, 2001, at 79–80. This definition is helpful in that it provides an objective criterion, i.e., the presence of different ratios of price to marginal costs (i.e., rates of return), to identify instances of price discrimination. It also suggests that different prices for the same product do not necessarily amount to price discrimination, as such a difference may be justified by cost variations. 96 In the context of the ECSC Treaty, see Case 13/63, Italy v Commission [1963] ECR 165.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 699

Abusive Pricing in an IP Licensing Context 699 The wording of the provision suggests that two conditions must be met for Article 82(c) to apply to a dominant firm’s prices. First it must be shown that the firm under investigation has applied dissimilar prices to “equivalent transactions”. The evaluation of the equivalence of two transactions is not an easy undertaking, as there are myriad factors that can be invoked to justify differences between two transactions. This is particularly true in the context of IP licensing, where potential licensors and licensees will usually take a wide range of factors into consideration (the level and the mode of calculation of the royalty, the presence or absence of an upfront fee payment, the size of their respective portfolios and the possibility to cross-license, the scope and territorial coverage of the licence, etc.) in their licensing negotiations. Thus, as most IP deals do not amount to “equivalent transactions”, differences in prices (royalties) are usually perfectly legitimate. By implication, forcing licensors to offer the same royalty level to all licensees would in fact amount to price discrimination within the meaning of Article 82(c). Second, Article 82(c) indicates that, as a result of such dissimilar treatment of equivalent transactions, some of the dominant firm’s trading parties must have been placed at a competitive disadvantage vis-à-vis others.97 Scholarly discussions regarding price discrimination often draw a distinction between “primary line” injury, which affects the dominant firm’s competitors, and “secondary line” injury, which affects one of several customers of the dominant firm by comparison to one or several other customers.98 The reference to the placing of the dominant firm’s “trading parties at a competitive disadvantage” clearly indicates that Article 82(c) seeks to prevent “secondary line” injury.99 All legal scholars seem to agree on this point.100 The requirement 97 In fact, the rationale behind Article 82(c) is similar to the primary rationale behind the Robinson-Patman Act, a controversial piece of legislation adopted in the US in 1936. See Paul, London, The Competition Solution, American Enterprise Institute, 2005, at p. 136; Domenick Armentano, Antitrust and Monopoly—Anatomy of a Policy Failure, 2nd ed., The Independent Institute, 1999, at 167. In the years following the Great Depression, small US retailers argued that large retailers were using their bargaining power with manufacturers to obtain better prices, as a result of which smaller players could not compete. The retailers claimed that, once they were excluded, large retailers would have the power to increase downstream prices at the expense of consumers. To protect small retailers, the US federal legislator adopted the Robinson-Patman Act, whose terms compelled manufacturers to justify or eliminate discounts to large retailers through a prohibition of price discrimination practices. Since then, however, the US Supreme Court has shown reluctance to enforce the Robinson-Patman Act. For a recent illustration, see Volvo Trucks North America, Inc. v. Reeder-Simco GMC, Inc., 546 U.S. 164 (2006). 98 See, e.g., Alison Jones and Brenda Sufrin, EC Competition Law, 2nd ed., Oxford University Press, 2004, at 411; Jonathan Faull and Ali Nikpay, eds., The EC Law of Competition, Oxford University Press, 1999, at § 3.235. 99 See René Joliet, Monopolization and Abuse of a Dominant Position, Collection scientifique de la Faculté de droit de l’Université de Liège, 1970, at 247. 100 See Santiago Martínez Lage and Rafael Allendesalazar, “Community Policy on Discriminatory Pricing: A Practitioner’s Perspective”, paper presented at the 2003 Annual EU Competition Law and Policy Workshop—What is an Abuse of a Dominant Position?, Florence, at 14; Van Bael & Bellis, Competition Law of the European Communities, 4th ed., Kluwer Law International, 2005, at 915; Richard Whish, Competition Law, 5th ed., LexisNexis Butterworths, 2003, at 716 and 710.

(Zf) Ehlermann 08 Geradin

700

6/11/08

11:56

Page 700

Session Three: Exploitative Conduct And The Remedies

that competitive disadvantage should occur also suggests that, for Article 82(c) to apply, the dominant firm’s customers should be in competition with each other.101 As a result, for differences in licensing conditions to fall under Article 82(c) they must affect licensees active on the same downstream product market. The combination of these two conditions strongly suggests that Article 82(c) will be applicable to licensing agreements only in very limited circumstances. The first condition indicated above will not be met in most instances due to the fact that potential licensees will generally not be similarly situated (there will, for instance, be differences in the licensees’ patent portfolios and their ability to offer a cross-licence, differences in the scope and geographical coverage of the licences they request, etc.). The second condition, which is more likely to be met in practice since licensees may compete on downstream markets, is only relevant when the first condition has been met, which, as just noted, will not often be the case. But more generally one may question why dominant licensors would seek to discriminate against similarly situated potential licensees that compete with each other on a downstream market. A key distinction has to be drawn here between vertically-integrated licensors (licensors which are active both in the upstream licensing market(s) and in the downstream product market(s)) and non-vertically integrated licensors (licensors active on the upstream licensing market(s) only).102 Non-vertically integrated IP licensors generally have no incentive to price discriminate so as to place one of their licensees at a competitive disadvantage vis-à-vis one or several others. Upstream licensors benefit from competition between their licensees because strong downstream competition will usually expand output. As royalties are typically calculated on the basis of downstream sales (e.g., X% of the sale prices of the products manufactured by the licensee(s)), this should benefit the licensor unless it can be shown that fierce downstream competition would depress the prices of downstream products to such an extent that the licensor’s royalty revenues would shrink. But even in that case, discriminating between licensees by giving a lower rate (i.e., a cost advantage) to one of them could allow that firm to expand its output and further depress prices. Prices would subsequently increase if the cost advantage in question had the effect of inducing the exit of licensees that have been discriminated against but, again, this scenario would eventually play against the interest of the licensor because it would increase concentration on the downstream market(s). This would in turn enhance the countervailing buying 101 See John Temple Lang and Robert O’Donoghue, “Defining Legitimate Competition: How to Clarify Pricing Abuses under Article 82 EC”, 26 Fordham International Law Journal 83, 115 (2002). 102 See Damien Geradin and Nicolas Petit, “Price Discrimination Under EC Competition Law: Another Antitrust Doctrine in Search of Limiting Principles?”, 2(3) Journal of Competition Law and Economics 479 (2006).

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 701

Abusive Pricing in an IP Licensing Context 701 power of the remaining licensee(s) and accordingly constrain the licensor’s own market power. This may explain why the number of secondary line price discrimination cases involving non-vertically integrated dominant firms is very limited.103 And the source of most of these cases is not the actions of dominant firms but rather protectionist Member State measures designed to maintain or strengthen the dominant position of domestic firms.104 By contrast, market structures where vertically-integrated firms control essential inputs are prone to “secondary line” injury price discrimination.105 Indeed, vertically-integrated operators generally have a strong incentive to charge a lower price to their downstream operations than to the operations of their competitors. The decisional practice of the Commission and the case law of the Community Courts contain many examples of such discrimination.106 This may be the reason why one of the leading US treatises on antitrust and IP observes that “[t]he only plausible anticompetitive explanation for [discriminatory license pricing] is as an act of foreclosure by a vertically integrated monopolist”.107 Much in the same vein, Swanson and Baumol observe that: “[While discriminatory license fees will generally not raise significant concern,] [t]here is a subset of cases . . . where potentially valid reasons exist for concern about discrimination in license fees for intellectual property: those instances when the owner of the IP uses it as an input in a downstream market where competitors also require the IP for the same purpose. A licensor exercising bottleneck market power that discriminates in licensing in order to handicap its competitors and favor its 103 See, e.g., Case C-18/93, Corsica Ferries Italia Srl v Corpo dei Piloti del Porto di Genova [1994] ECR I-1783; Commission Decision 95/364 of 28 June 1995, Brussels National Airport [1995] OJ L216/8; Commission Decision 1999/199 of 10 February 1999, Portuguese Airports [1999] OJ L69/31; Commission Decision 1999/98 of 10 February 1999, Ilmailulaitos/ Luftfarsverket [1999] OJ L69/24; Commission Decision, 98/153 of 11 June 1998, Alpha Flight Services/Aéroports de Paris [1998] OJ L230/10, upheld: Case T-128/98, Aéroports de Paris v Commission [2000] ECR II-3929, upheld on further appeal: Case C-82/01 P, Aéroports de Paris v Commission [2002] ECR I-9297. 104 See Geradin and Petit, supra note 102. 105 Ibid. 106 For instance, in the Deutsche Bahn case, Transfracht, a subsidiary of the German railway operator, was active in the carriage of maritime containers to or from Germany passing through German ports. Intercontainer was active in the carriage of maritime containers to or from Germany passing through western ports (i.e., ports in Belgium and The Netherlands). Although they provided a similar service (the carriage of maritime containers to and from Germany), both firms had been charged different prices by Deutsche Bahn for access to the rail infrastructure. The facts revealed, for instance, that the price differences in the case of transport of empty containers ranged from 2% to 77% in favour of Transfracht. The Commission and the CFI thus considered that Deutsche Bahn had infringed Article 82(c) by applying dissimilar conditions to equivalent services. The discrimination had the effect of placing the parties operating from western ports at a competitive disadvantage vis-à-vis Deutsche Bahn and its subsidiary. See Case T-229/94, Deutsche Bahn AG v Commission [1997] ECR II-1689, at para. 93. In support of this, the Commission had gathered evidence that Deutsche Bahn’s price discrimination had substantially limited the carriage of containers between the western ports and Germany in favour of imports and exports to and from Germany through the port of Hamburg. See Commission Decision 94/210 of 29 March 1994, HOV-SVZ/MCN [1994] OJ L104/34, recital 254. 107 See Hovenkamp, Janis and Lemley, supra note 9.

(Zf) Ehlermann 08 Geradin

702

6/11/08

11:56

Page 702

Session Three: Exploitative Conduct And The Remedies

own downstream sales can create or enhance market power in downstream markets for standard-compliant products and services. By contrast, a pure licensor (even one with monopoly power) will ordinarily lack anticompetitive reasons for engaging in discrimination.”108

Interestingly, Swanson and Baumol also consider that the risk of foreclosure presented by a vertically-integrated licensor “is (or should be taken to be) the principal justification for the RAND nondiscrimination requirement”.109 The non-discrimination element of FRAND would thus be designed to create a blanket prohibition on royalty rates (or other forms of consideration) leading to discrimination between licensees, but more narrowly to prevent vertical foreclosure by firms active in upstream and downstream markets. These authors then explain that the economics of price regulation provide a pricing principle that can be relied on to determine an efficient, non-discriminatory licensing fee for technology. According to this principle, which has been referred to as the efficient component-pricing rule (ECPR) or as the parity principle, “the price that the IP-holder firm charges itself for the use of its own innovation input equals the price the firm charges customers for a final product using that IP, minus the incremental cost to the IP-holding firm of all other inputs, including capital, used to produce the final product”.110 Swanson and Baumol argue that compliance with this principle is “necessary and sufficient for a license fee to be competitively neutral in downstream markets and, therefore, at least on that basis, a necessary condition for that fee to be nondiscriminatory”.111 This strongly suggests that, while price discrimination in IP licensing is usually perfectly legitimate and pro-competitive, particular attention must nevertheless be paid by competition authorities and courts to attempts by vertically-integrated licensors to raise their downstream rivals’ costs by giving more favourable treatment to their own operations.

V. Conclusion This paper has aimed to show that competition authorities and courts should proceed with extreme caution when facing claims that an IP licensee charges excessive royalties or abusively discriminates between its licensees. Controlling royalties involves significant theoretical and practical difficulties which should not be underestimated. Placing caps on rates may also 108 109 110 111

See Swanson and Baumol, supra note 13. Ibid. Ibid. Ibid.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 703

Abusive Pricing in an IP Licensing Context 703 produce a range of unintended negative consequences: it may harm innovation (by reducing the profits of firms that make risky investments) and may impede dynamic competition (by decreasing incentives for new firms to enter into licensing markets subject to price control). In addition, as abundantly illustrated by the large number of acrimonious disputes generated by the introduction of price control in regulated sectors, controlling rates is likely to turn competition authorities into quasi-permanent regulators even though they lack the resources to do a good job. This may lead to mistakes with tragic consequences for economic welfare. In parallel with recent efforts by some firms to force the application of competition rules in the field of IP licensing, significant manufacturing interests have also been funding scholarly papers which propose patent law reforms designed to reduce the protection and the bargaining power of licensors with a view to reducing their ability to obtain adequate compensation for their technologies. One may of course hold the view that society should benefit from lower royalties (as it generally benefits from lower prices). But this would be too simple. No convincing case has been made that lower royalties will automatically translate into lower consumer prices and wider dissemination of valuable technologies. The primary effect of any such proposals is more likely to be a transfer of rents from innovators to manufacturers. Whether this is a desirable industrial policy outcome is highly dubious, especially at a time where governments unanimously describe privately-funded innovation as the primary force driving economic growth.

(Zf) Ehlermann 08 Geradin

6/11/08

11:56

Page 704

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 705

V Mark R. Patterson* The Peculiar “New Product” Requirement in European Refusal to License Cases: A US Perspective June 2007

I. Introduction In this paper, I take issue with one aspect of the “new product” requirement set out by the Court of Justice in Magill 1 and reaffirmed, or perhaps reinterpreted, by the Court in IMS Health.2 I do not question the basic thrust of the requirement, which I take to be an insistence that a challenger, to succeed in forcing a dominant firm to license its intellectual property, must itself contribute something new to the market. But in my view the ECJ has gone beyond that goal in stating that a refusal to license “may be regarded as abusive only where the undertaking which requested the licence does not intend to limit itself essentially to duplicating the goods or services already offered on the secondary market by the owner of the intellectual property right”.3 In effect, this requirement holds that the greater the degree of competition between the parties in the downstream market, the less likely it is that competition law will require a license. Although many commentators have addressed the “new product” requirement,4 they have not generally focused * Professor of Law, Fordham University School of Law; Visiting Fellow, European University Institute, 2006–2007. 1 Joined Cases C-241/91 P and C-242/91 P, RTE and ITP v Commission (“Magill”) [1995] ECR I-743. 2 Case C-418/01, IMS Health GmbH & Co. OHG and NDC Health GmbH & Co. KG [2004] ECR I-5039 (“IMS Health”). There are several other decisions and opinions related to the IMS Health case. See Commission Decision 2002/165 of 3 July 2001, NDC Health/IMS Health: Interim Measures [2002] OJ L59/18, withdrawn: Commission Decision 2003/741 of 13 August 2003, NDC Health/IMS Health: Interim Measures, [2003] OJ L268/69; Order of the President of the Court of First Instance, Case T-184/01 R, IMS Health Inc. v Commission [2001] ECR II-3193; Order of the President of the Court, Case C-481/01 P(R), NDC Health Corp. [2002] ECR I-3401; Opinion of Advocate General Tizzano of 2 October 2003, Case C-418/01, IMS Health GmbH & Co. OHG and NDC Health GmbH & Co. KG [2004] ECR I-5039. 3 Case C-418/01, IMS Health, cited previous footnote, para. 49. 4 See, e.g., James S. Venit, “Article 82 EC: Exceptional Circumstances: The IP/Antitrust Interface after IMS Health”, available at http://www.iue.it/RSCAS/Research/Competition/2005/ 200510-CompVenit.pdf, at 16–18; Christian Ahlborn, David S. Evans and A. Jorge Padilla, “The

(Zg) Ehlermann 08 Patterson

706

6/11/08

11:56

Page 706

Session Three: Exploitative Conduct And The Remedies

on this issue. The focus of much of the commentary has been on how “new” the “new product” must be without distinguishing whether its novelty should be measured against the dominant firm’s own downstream product (if any) or against its upstream intellectual property.5 It is that issue on which I will focus. Although I accept that a challenger, to gain access to another’s intellectual property, must make an additional contribution in a downstream market, I argue that the presence of the intellectual property owner in the downstream market should not make it more difficult for the firm seeking access to the intellectual property to prevail. The flaws of the IMS Health approach on its own terms will be discussed below, but at this point it will just be noted that a recent US case took the opposite approach. Whereas under the ECJ’s “new product” requirement a case for access is weakened by direct competition between the two parties, the Federal Circuit Court of Appeals held in Intergraph Corp. v. Intel Corp.6 that the success of a case requires such competition: “Although the viability and scope of the essential facility theory has occasioned much scholarly commentary, no court has taken it beyond the situation of competition with the controller of the facility, whether the competition is in the field of the facility itself or in a vertically related market that is controlled by the facility. That is, there must be a market in which plaintiff and defendant compete, such that a monopolist extends its monopoly to the downstream market by refusing access to the facility it controls.”7

The discussion of this issue, and of the difference between the relevant European and US approaches, will proceed in several steps. First, I will discuss the distinction between this issue and the separate question of whether there must be two-market leveraging. Second, I will discuss the shortcomings in the European test’s treatment of the presence of the intellectual property

Logic & Limits of the ‘Exceptional Circumstances Test’ in Magill and IMS Health”, 28 Fordham Journal of International Law 1109 (2005); Josef Drexl, “Abuse of Dominance in Licensing and Refusal to License: A ‘More Economic Approach’ to Competition by Imitation and to Competition by Substitution”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2005: The Interaction between Competition Law and Intellectual Property Law, Hart Publishing, 2007; Heike Schweitzer, “Controlling the unilateral exercise of intellectual property rights: A multitude of approaches but no way ahead?”, EUI Working Paper LAW 2007/31, available at http://cadmus.iue.it/dspace/handle/1814/7625. 5 See, e.g., Ahlborn, Evans and Padilla, supra note 4. 6 Intergraph Corp. v. Intel Corp., 195 F.3d 1346 (Fed. Cir. 1999). 7 195 F.3d at 1357 (citations omitted). See also Aquatherm Industries, Inc, v. Florida Power & Light Co., 145 F.3d 1258, 1261, 1262 (11th Cir. 1998) (“By stating a monopoly leveraging claim under these facts, Aquatherm in effect asks this court to extend Berkey Photo to a situation in which a monopolist projects its power into a market it not only does not seek to monopolize, but in which it does not even seek to compete. There is no support for such an extension in either the language of § 2 or the case law interpreting it.”), quoted in Intergraph, cited in previous footnote, 195 F.3d at 1361. A similar view was taken by the Court of First Instance in Case T-504/93, Tiercé Ladbroke SA v Commission [1997] ECR II-923. See infra text accompanying notes 23–24.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 707

“New Product” Requirement in European Refusal to License Cases 707 owner in the downstream market. Third, I will discuss the very different weaknesses of the US approach. I will conclude with some general observations.

II. The “Market Leveraging” Issue Because this paper is primarily about market relationships, it is useful at the outset to distinguish the discussion here from another market-relationship issue that has been discussed in the IMS Health context.8 That is the issue of market leveraging: whether a case for access to intellectual property requires that the intellectual property itself constitute, or at least be a part of, a separate relevant antitrust market, or if it is sufficient for the intellectual property to be an important (or indispensable, or essential) input for products supplied in a downstream market. This issue is important in this context because, as I will argue later, the significance that the ECJ attributes to the presence of the intellectual property owner in the downstream market is inappropriate if that market is distinct from the upstream one. In IMS Health the ECJ held that “it is determinative that two different stages of production may be identified and that they are interconnected, inasmuch as the upstream product is indispensable for the supply of the downstream product”.9 This stops short of requiring that the upstream product be in or constitute a relevant market. On the other hand, the Court also said that “it is sufficient that a potential market or even hypothetical market can be identified”.10 This indicates that the intellectual property at issue should be capable of being defined as a relevant market, even if no such active market yet exists. However, that such a market definition will generally be possible seems evident, given the other requirements of the IMS Health test. If the intellectual property is truly indispensable or essential, it has no substitutes, and thus must constitute a relevant antitrust market of its own. The relevant market test is one of demand, and for those sellers that require “indispensable” intellectual property as an input there is no “sufficient substitutability” or “reasonable interchangeability” with other products. As a result, it seems clear that intellectual property that is indispensable for production of another product would define a relevant market. 8 This issue has been discussed in Venit, supra note 4; Robert Pitofsky, Donna Patterson and Jonathan Hooks, “The Essential Facilities Doctrine Under U.S. Antitrust Law”, 70 Antitrust Law Journal 443 (2002); Paul D. Marquardt and Mark Leddy, “The Essential Facilities Doctrine and Intellectual Property Rights: A Response to Pitofsky, Patterson, and Hooks”, 70 Antitrust Law Journal 847 (2003). 9 Case C-418/01, IMS Health, para. 45. 10 Ibid., para. 44.

(Zg) Ehlermann 08 Patterson

708

6/11/08

11:56

Page 708

Session Three: Exploitative Conduct And The Remedies

This point is obscured in the cases because in some of them the upstream product had never been sold separately. However, that can just as easily be explained by the monopolist’s decision to keep the downstream market to itself as by any economic factor that would keep the upstream product from being a relevant antitrust market. That is, the decision of a supplier not to provide a particular product does not mean that there is no demand for the product, and demand is the factor that defines a relevant market. Indeed, those commentators who insist that market leveraging must be present do not actually argue, as a factual matter, that the upstream products in these cases would not be relevant antitrust markets. Instead, they generally argue only that there should be a requirement that the plaintiff establish the existence of such a market.11 In the intellectual property area, though, there is ample precedent for potential markets. The US antitrust agencies’ Guidelines for the Licensing of Intellectual Property describe the use of “technology markets” and “innovation markets,”12 either of which could involve technologies that have not yet been marketed. Indeed, the example that the Guidelines provide for the analysis of a technology market describes the analysis in a context “[b]efore the firms use their technologies internally or license them to third parties”.13 Several challenges to pharmaceutical company mergers have also involved products that had not yet been marketed. But to say that the intellectual property in these cases constitutes a relevant market is not necessarily to say that there are two separate markets. It could be that the intellectual property market is effectively the only market at issue. Thus, although the commentary on this issue has focused on whether the upstream intellectual property is a relevant market,14 the better question would seem to be whether the downstream product is in fact distinct from the upstream one. In that respect, it seems, IMS Health directs attention to the

11 Marquardt and Leddy, supra note 8, at 849–55. Conversely, an article arguing that there is no requirement that the plaintiff define two separate markets does not really address whether in the cases it cites, the plaintiff could have defined two markets, although that is the relevant question. See Pitofsky et al., supra note 8, at 458–461. 12 U.S. Department of Justice and Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property 8–13 (April 1995). The Guidelines are available at the DOJ’s website: http://www.usdoj.gov/atr/public/guidelines/0558.htm. 13 Ibid., Example 2, at 8–9. 14 This has been true even when the context is the essential facility doctrine, which has not generally emphasized the need to demonstrate an upstream market:

“ ‘Market’ is here used in its normal sense of a demand for and a supply of a product or service. It does not include the mere possibility of granting licences of the intellectual property right. If there is such a right, it is always in principle possible to license it. But that does not create a market for the purposes of the essential facility principle, and it could not do so, without making every intellectual property right subject to compulsory licensing.” John Temple Lang, “The Application of the Essential Facility Doctrine to Intellectual Property Rights under European Competition Law”, conference on Antitrust, Patent and Copyright, Ecole des Mines, Paris, 15–16 January 2004, at 9 n. 19. The paper is available at http://www. cerna.ensmp.fr/cerna_regulation/Documents/?ColloqueAntitrust2004/?TempleLang.pd.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 709

“New Product” Requirement in European Refusal to License Cases 709 correct question.15 As described by the Court, a key issue in these cases is that of “distinguish[ing] an upstream market, constituted by the product or service [to which access is sought], and a (secondary) downstream market, on which the product or service in question is used for the production of another product or the supply of another service”.16 It is true, though, that even if the upstream intellectual property is analytically a separate market, there may be practical considerations that counsel against requiring the licensing of a product that has never been licensed before. We might decline to require licensing because the absence of any previous licensing arrangement would make it difficult to establish the terms on which a license should be made.17 Marquardt and Leddy, for example, argue that “there is no reason to believe that antitrust enforcement agencies or courts are equipped to determine the ‘correct’ price of intellectual property”.18 In fact, however, this same concern can arise even where the intellectual property owner has previously licensed the intellectual property. The owner’s willingness to license its property in some circumstances at particular prices does not mean that the same price is appropriate for all circumstances. It is generally accepted that an intellectual property owner is entitled to price discriminate among various licenses of its property.19 John Temple Lang argues that in these cases it might be appropriate to require the grant at better terms than the intellectual property owner has given to others.20 The reverse is also true: it might be that the appropriate terms for a particular compelled license should be more favorable to the owner than were those used for previous licensing arrangements. To sum up, it seems that the question of defining an upstream intellectual property market should not pose an obstacle to a refusal to license case. The key question, instead, is whether the downstream market in which the license is sought is sufficiently independent of the intellectual property itself. “Community law allows a dominant company to use intellectual property rights in the market in which they primarily apply, but it may be an abuse to cut off an input which is essential on another market, if that would otherwise be unlawful”.21 The remainder of this paper proceeds in light of this conclusion,

15 Cf. Temple Lang, cited in previous footnote, at 13 (“The more clearly distinct the two markets are, and the greater the restriction of competition on the second market which results from the refusal to supply, the stronger the argument for compulsory licensing.”); Venit, supra note 4, at 16 (“[R]etention of the requirement in Magill that the product the would-be licensee seeks to introduce be a novel one that is not offered by the owner of the IPR has the potential to compensate for elimination of the leveraging requirement if this novelty requirement, which embraces the key goal of the IP system, is strictly applied.”). 16 Case C-418/01, IMS Health, para. 42. 17 Marquardt and Leddy, supra note 8, at 855–856 and n. 30. 18 Ibid., at 856 n. 30. 19 Temple Lang, supra note 14. 20 Ibid., at 5 and n. 11. 21 Ibid., at 13.

(Zg) Ehlermann 08 Patterson

710

6/11/08

11:56

Page 710

Session Three: Exploitative Conduct And The Remedies

focusing on the distinction between the downstream market and the intellectual property.

III. The ECJ’s “New Product” Approach The conceptual problems with the IMS Health test can be seen if one considers either of two sorts of cases: those in which the party seeking access to a monopolist’s intellectual property provides a product very different from that of the monopolist, and those in which the party seeking access provides a product identical to that of the monopolist. The first type of case can be illustrated by reference to the figure below, in which the dotted lines represent possible dividing lines between relevant antitrust markets U, D1, and D2, and the curved line represents the boundary of the monopolist firm:

The Court’s requirement that the party seeking access must be providing a “new product” could mean, as shown in the figure, that downstream markets D1 and D2 are distinct. If so, then the monopolist is not actually present in the relevant downstream market, i.e., market D1, in which access is being sought. This in fact seems to be the situation, or one of the situations, contemplated by the ECJ, in that in the Magill case the copyright owners were not in the market for comprehensive weekly television listings, but provided only single-channel listings, which the court appeared to view as products in a separate market.22 (One might reasonably wonder, though, why an intellectual property owner would refuse to grant a license in a market where its own products did not compete.) 22 The Court referred to the CFI’s finding that there was “no actual or potential substitute for a [comprehensive] weekly television guide”. Joined Cases C-241/91 P and C-242/91 P, Magill, cited supra note 1, para. 52.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 711

“New Product” Requirement in European Refusal to License Cases 711 But if the monopolist is not present in the relevant downstream market, then the case is one that, as was described above and will be discussed further below, the Federal Circuit said in Intergraph should not succeed. The Court of First Instance has also taken this position in Tiercé Ladbroke.23 In that case, the arguably jointly dominant firms, French sociétés de courses, or race courses, refused to grant a license covering their copyrighted broadcast signals of horse races in France. The potential licensee sought to have access to those signals in Belgium, where it operated, but where the sociétés did not. The CFI found this factor to be decisive: “[N]either the absence of technical barriers to the transmission of French sound and pictures in Belgium nor the fact that the applicant might be regarded, from an overall perspective, as a potential competitor of the sociétés de courses is sufficient for the refusal to supply sound and pictures to be regarded as constituting an abuse of a dominant position since the sociétés de courses themselves are not present in the separate geographical market on which the applicant operates and, secondly, they have not granted any licence to other operators on that market.”24

The ECJ in IMS Health did not reconcile this apparent conflict between its approach and that of Tiercé Ladbroke. In any event, perhaps the Court in IMS Health did not mean to require that D1 and D2 be separate relevant markets but rather intended also to include circumstances in which the products of the monopolist and the challenger comprise a single relevant market in which the two parties sell differentiated products. That seems somewhat inconsistent with the Court’s language, which referred to a “new product for which there is a potential consumer demand”,25 but it is certainly a possible reading of the case.26 At the least, it appears that the Court’s view is that the more similar to the monopolist’s product is the one offered by the party seeking access, the less compelling is the case for a license. That seems incorrect, though. Even if the two products offered in the downstream market were identical, that would not necessarily show that no license is required. Suppose, for example, that a firm has intellectual property which is an essential element of a downstream product, but which makes up only 10% of the downstream product’s value. Suppose also that the monopolist and the challenger offer (or, in the challenger’s case, that it seeks to offer) identical products in the downstream market. This can be represented in the figure as follows:

23

Case T-504/93, Tiercé Ladbroke SA v Commission, cited supra note 7. Ibid., para. 129 25 IMS Health, para. 38 26 The Court might have meant only that the new product generated additional demand that could be met, though less well, by the dominant firm’s own product. 24

(Zg) Ehlermann 08 Patterson

712

6/11/08

11:56

Page 712

Session Three: Exploitative Conduct And The Remedies

This does not seem an implausible interpretation of many factual circumstances. Although one might question whether any intellectual property that is “essential” or “indispensable” to a downstream product can really be only a small part of that product, one might consider, for example, interoperability information, as in the Microsoft case.27 Although Microsoft argued that its interface specifications were themselves the field of competition, the Commission rejected that argument, pointing to evidence of the importance of other product characteristics.28 In a recent article, Eleanor Fox discussed another example (see below). In such a case, it does not appear that the possibility of a compelled license should be excluded. How, though, should the importance of the intellectual property be assessed? Hovenkamp, Janis, and Lemley argue that, in the US, “it seems fair to characterize the law as distinguishing between cases in which the intellectual property right itself is that facility to which the plaintiff wants access and cases in which intellectual property rights exist but are incidental to the control of the facility itself ”.29 However, this does not seem to capture the situation in the figure above. The examples mentioned by these authors are those in which either the intellectual property owner did not rely on its intellectual property rights or the right is only an “incidental” part of the upstream essential facility at issue.30 In contrast, in Magill and IMS 27 See Commission Decision of 24 March 2004, Microsoft, Case COMP/C-3/37.792; on appeal: Case T-201/04, not yet decided as of this writing. 28 Microsoft, cited previous footnote, § 5.3.1.3.1. 29 Herbert Hovenkamp, Mark D. Janis and Mark A. Lemley, “Unilateral Refusals to License in the U.S.” (April 2005), available at http://ssrn.com/abstract=703161, at p. 26. 30 Ibid. As an example of the latter type of case, they suggest MCI Communications Corp. v. AT&T, 708 F.2d 1081 (7th Cir. 1983), where the plaintiff sought access to local telephone distribution lines, i.e., the “last mile”. The authors say that access in such a case would presumably not turn on the existence of patented electronic switches in those lines.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 713

“New Product” Requirement in European Refusal to License Cases 713 Health, the intellectual property right was coextensive with the claimed facility.31 Another possibility would be to consider the investment involved in creating the intellectual property, perhaps in comparison to the effort made in the downstream market. In a recent article,32 Eleanor Fox discussed a US case, New York Mercantile Exchange, Inc. v. Intercontinental Exchange, Inc.,33 which illustrates the possibility of a disproportionate relationship between the two. The parties were commodity exchanges, and customers demanded that their contracts employ settlement prices set, allegedly by formula, by the incumbent. The incumbent argued that its settlement prices were copyrighted, and it refused to provide them in a timely manner to the challenger. By this means, as Fox describes, the incumbent was able to deny “a necessary input which was merely mechanistic, not creative” in order to harm competition in “the principal market”.34 This issue is taken up further below, but the point here is that it is not, or at least not only, the distance between the two downstream products that is important, but the distance between the downstream products and the upstream intellectual property. Generally speaking, one can ask whether the downstream market is within the scope of the intellectual property owner’s rights. Although that is a difficult conceptual question, there are a variety of ways to answer it, at least in certain circumstances.35 It is important to recognize that the scope of the intellectual property right is defined by the limits of those rights, not by the definition of infringement. Referring to Volvo v Veng, James Venit has stated that one of the forms of abuse identified in that case—“the refusal to supply protected body panels to independent repair shops—would also come squarely within the scope of the

31 In Magill, however, the facility was information, rather than the sort of creative contribution to which copyright protection is usually limited. See Temple Lang, supra note 14, at 2–3. Indeed, access to the facility could perhaps have been denied without reliance on the copyright, if the television stations could have withheld the information and released it only when it was too late to be useful in preparing the composite weekly program guide. 32 Eleanor M. Fox, “A Tale of Two Jurisdictions and an Orphan Case: Antitrust, Intellectual Property, and Refusals to Deal”, 28 Fordham International Law Journal 952 (2005). 33 New York Mercantile Exchange, Inc. v. Intercontinental Exchange, Inc., 323 F. Supp. 2d 559 (S.D.N.Y. 2004). 34 Fox, supra note 32, at 963–64. The court rejected the plaintiff’s Section 2 claims under the Sherman Act. Although it did not focus on the intellectual property aspect of the case, it did note that “NYMEX has a legitimate business interest in preventing its competitor, ICE, from free-riding on NYMEX’s settlement prices”. 323 F. Supp. 2d at 571. It is not clear, however, whether it was referring to intellectual property in its reference to free-riding, because it then said “NYMEX’s settlement prices have value because they are viewed as proxies for market prices, and NYMEX has a legitimate interest in preventing rivals from free-riding on this reputation”. Ibid. 35 I have previously proposed tests for this issue in specific contexts. See Mark R. Patterson, “When Is Property Intellectual? The Leveraging Problem”, 73 Southern California Law Review 1133 (2000); Mark R. Patterson, “Inventions, Industry Standards, and Intellectual Property”, 17 Berkeley Technology Law Journal 1043 (2002).

(Zg) Ehlermann 08 Patterson

714

6/11/08

11:56

Page 714

Session Three: Exploitative Conduct And The Remedies

IP monopoly”.36 But that is true only if the scope of the monopoly is viewed as encompassing any refusal to license, without regard to the market where the effects of that refusal are felt or to the purpose of the refusal.37 That makes the scope of the intellectual property right turn on the general definition of infringement, not on the particular patent or copyright at issue. For the scope inquiry to have independent content, it must have reference to the limits of protection provided by the intellectual property right at issue. In the case of patent rights, that protection is defined by the patent claims, and in copyright, by the protected expression. Where it is competition in a downstream market that is at issue, the downstream market will almost inevitably involve innovation and competition that is not encompassed within the definition of the intellectual property rights in the upstream market. Hence, as John Temple Lang has said, “[t]he more clearly distinct the two markets are, and the greater the restriction of competition on the second market which results from the refusal to supply, the stronger the argument for compulsory licensing”.38 Actually, the IMS Health test might serve to address this issue, but it is a very under-inclusive means of doing so. IMS Health asks, as discussed above, whether the two products in the downstream market are significantly different. If those two downstream products in fact differ significantly, it seems likely that each of them, or at least one of them, must also differ significantly from the intellectual property itself. That is, it seems that it would be difficult for the products to differ from each other while being, at the same time, very similar to the intellectual property. But the IMS Health approach ignores the possibility that the downstream products could be very different from the upstream intellectual property even if they are quite similar vis-à-vis each other. Specific problems with using a scope inquiry to address the refusal to license issue are discussed in the concluding section below. The argument here is simply that if the downstream market is outside the scope of the dominant firm’s intellectual property, the fact that the dominant firm itself is in that downstream market should not in itself defeat a challenger’s effort to compel a license of the intellectual property, even if the products of the dominant firm and the challenger are identical in the downstream market.

IV. A US Approach: Intergraph v. Intel There are also problems with the test set out by the Federal Circuit Court of Appeals in Intergraph Corp. v. Intel Corp., which requires that the monopolist 36 37 38

Venit, supra note 4, at 9. Venit would apparently confine the scope inquiry to look at acts otherwise forbidden. Temple Lang, supra note 14, at 13.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 715

“New Product” Requirement in European Refusal to License Cases 715 be a competitor in the market in which access to the facility is sought. Although the court in Intergraph was not entirely clear on the rationale for this holding, it said that “the presence of a competitive relationship is fundamental to invoking the Sherman Act to force access to the property of another”.39 But in Intergraph, and in the only case that it cited as relying on the same principle,40 there was indeed a competitive relationship—the competitive relationship was simply in another market. This presents the question, then, of whether the owner of an essential facility should be permitted to deny that facility in order to harm a competitor in a market other than the one in which the two compete. That is, the circumstances are as shown in the following figure, in which again the dashed lines represent divisions between markets and the curved shapes represent firm boundaries:

The question is whether it should be permissible for the intellectual property owner, operating in the upstream market U and in another market O, to deny its competitor in market O access to its intellectual property in market D, even if the two are not competitors in market D. On the one hand, such a denial certainly does not appear to be competition on the merits in market O. On the other hand, though, it is not clear that a loss of potential profits to the competitor in market D should hinder its ability to compete in market O. However, the problem may be not so much the ability of the competitor to compete in market O as its incentive to do so. If the competitor anticipates significant profits in market D, which could be lost through denial of the intellectual property if it seeks to compete in market O, it may choose not to compete in market O at all. That is, the problem is not that being denied intellectual property in market D harms the challenger’s ability to compete in 39

Intergraph Corp. v. Intel Corp., cited supra note 6, at 1357 (Fed. Cir. 1999) (citations omitted). See Caribbean Broadcasting System, Ltd. v. Cable & Wireless plc, 148 F.3d 1080 (D.C. Cir. 1998). In Caribbean Broadcasting, the plaintiff competed with one defendant that was partly owned by the other defendant. However, the court determined that it did not have jurisdiction over the plaintiff’s competitor (ibid. at 1089–1091), which left the part-owner as the only available defendant. The court rejected the essential facility claim because that part-owner did not itself compete in the relevant market. Ibid. at 1087–1089. 40

(Zg) Ehlermann 08 Patterson

716

6/11/08

11:56

Page 716

Session Three: Exploitative Conduct And The Remedies

market O. The problem instead is that competing in market O destroys its ability to compete in market D. In fact, this appears to describe very well the facts of Intergraph. In the case, market D was a computer graphics workstation market, market U was “various special benefits, including proprietary information and products”, which Intel provided to its “strategic customers”,41 and market O was a market for “high-end microprocessors”, in which Intergraph claimed that it competed through its ownership of relevant patents in that market.42 When Intergraph sued Intel for patent infringement based on Intel’s conduct in market O, Intel cut off the supply of its proprietary information from market U to Intergraph, allegedly making Intergraph unable to compete in market D. Intel had apparently been quite willing to provide the proprietary information to Intergraph prior to Intergraph’s initiation of the patent infringement suit. It also describes some of the conduct in the US Microsoft case, although there the conduct was not treated as raising intellectual property issues.43 The court in Microsoft described the company’s threat, made to Apple Computer, to cease producing Microsoft’s Mac Office program, which would have been a severe blow to Apple. The reason for the threat was Microsoft’s dissatisfaction with Apple’s unwillingness to enter into an exclusive deal to include only Microsoft’s Internet Explorer on Apple computers. Thus, Mac Office (or office productivity software) was market U, personal computers were market D, and web browsers were market O. (Although Apple did not itself produce a web browser, it could be viewed as a participant in that market to the extent that it was compensated by Netscape for including Netscape’s Navigator browser on its computers.) The court pointed to the effect of the subsequent Microsoft-Apple exclusive deal as an anticompetitive one of restricting distribution of competing web browsers, and it was unwilling to accept Microsoft’s argument that the deal was pro-competitive because it was part of a package of arrangements between the two companies: “Microsoft offers no procompetitive justification for the exclusive dealing arrangement. It makes only the irrelevant claim that the IE-for-Mac Office deal is part of a multifaceted set of agreements between itself and Apple; that does not mean it has any procompetitive justification. Accordingly, we hold that the exclusive deal with Apple is exclusionary, in violation of § 2 of the Sherman Act.”44

Although the court’s holding here was not the result of a claim by Apple for access to continued production of Mac Office, the court devotes a considerable portion of its discussion to the threat to cut off Mac Office. It is possible that the court would have found the exclusive arrangement to be a 41 42 43 44

Intergraph, 195 F.3d at 1349–1350. Ibid. at 1355. See United States v. Microsoft Corp., 253 F.3d 34, 72–74 (D.C. Cir. 2001). Ibid. at 74 (citation and quotation of Microsoft’s brief omitted).

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 717

“New Product” Requirement in European Refusal to License Cases 717 violation even without the threat, but the discussion gives the impression that the court relied on that threat in finding a violation. The circumstances in the figure above also echo the issue in EC law of what sort of “associative links” must exist among the markets in which a firm has a dominant position, in which it abuses that dominance, and in which the effects of the abuse are felt.45 The relationships represented in the figure are somewhat similar to those in AKZO,46 where AKZO, a dominant firm in the organic peroxides market, responded to a challenge by ECS in the sales of organic peroxides in the plastics market—market O?—by charging “unreasonably low prices” in the flour additives market—market D?—a market that was “vital” to ECS. Furthermore, as the ECJ said, “AKZO did not adopt its behaviour in order to strengthen its position in the flour additives sector, but to preserve its position in the plastics sector by preventing ECS from extending its activities to that sector”. In these cases the argument for a finding of abuse, and in Intergraph for mandated access, seems strong. As the Microsoft court’s unwillingness to accept Microsoft’s justification shows, it is difficult for the defendants in these cases to point to an objective justification or legitimate business purpose for their actions. The very fact that the effects of those actions take place in other markets makes their legitimacy questionable. One could perhaps imagine a case in which there was no competition between the monopolist and the challenger in any market, and in that case the Intergraph rule might make sense. But in that case it would be hard to imagine why the monopolist would engage in anticompetitive conduct at all. The more likely scenario is the one seen in these cases, and there it makes little sense to give monopolists carte blanche to do harm whenever they can find a way to link two markets in which they face competition. Moreover, this argument is even stronger in the intellectual property context. As described above, intellectual property rights, arguably unlike other property rights, have a defined scope. Because a case like this involves the use of intellectual property to gain an advantage in an entirely unrelated market—and thus a market that is outside the scope of the intellectual property right—there seems little justification for allowing the intellectual property right to prevent the application of normal antitrust principles.

V. Observations on the Scope Inquiry The typical case at issue here, if the intellectual property and the product in which it is embodied are in different markets, looks like this: 45 46

See Case C-333/94 P, Tetra Pak International SA v Commission [1996] ECR I-5951. Case C-62/86, AKZO Chemie BV v Commission [1991] ECR I-3359.

(Zg) Ehlermann 08 Patterson

718

6/11/08

11:56

Page 718

Session Three: Exploitative Conduct And The Remedies

This figure makes clear the scope inquiry. The question is why, as a result of its intellectual property contribution in the upstream market, the monopolist should also be able to exclude competition in the downstream market, if that market is distinct. Generally speaking, an intellectual property right in one market gives its owner no right to injure competition in another market. Moreover, the contributions, including innovative contributions, in the secondary, downstream markets can be as significant as those in the upstream one, so there is no a priori reason to favor the upstream market. It would perhaps be possible to conduct a balancing inquiry, weighing the anticompetitive effects in the downstream market against the pro-competitive innovative effects in the upstream market. This in fact is what the European Commission appears to have done in Microsoft. But to do this is to apply an antitrust approach to an intellectual property problem. The goal of intellectual property law is not to strike such a balance in an individual case, but to create a framework that provides a satisfactory balance overall. And that framework contemplates that the intellectual property owner will receive protection against competition for its own innovation, but not that it will be permitted to extend that protection beyond the scope of that innovation. Thus, as stated above, what is required is an inquiry into whether the downstream product is within the scope of the claims of the patent, or the expression protected by copyright. That is, what is required is not an economic inquiry but a technical one, as the nature of intellectual property would suggest. James Venit has questioned such an approach, noting that “[t]o the extent that novelty is a more difficult concept to apply than leveraging, or involves a subjective element, the erosion of IPRs resulting from IMS Health may be greater than appears”.47 But it seems unlikely that the scope inquiry could be more difficult than a relevant market one, which is notoriously 47

Venit, supra note 4, at 18.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 719

“New Product” Requirement in European Refusal to License Cases 719 difficult. Particularly for new or hypothetical markets, where economic data may be lacking, it seems that a focus on the nature of products, in a manner akin to an infringement inquiry, is likely to prove more tractable than a focus on market definition and market power. Moreover, the intellectual property owner will often have available to it options to obviate the inquiry entirely. If the owner believes that its contribution to the downstream market is significant enough to justify a monopoly in that market, the owner may be able to seek intellectual property protection in that market as well as in the upstream one. For example, suppose the IMS Health case had arisen in the US. In that scenario, if IMS Health believed that its 1860 brick structure provided special benefits in the provision of pharmaceutical data, it could have sought business method patent protection for that technique. In that case, with intellectual property protection in the market at issue, there would be no question about the owner’s right to exclude in that market. Of course, the owner of intellectual property in an upstream market might be denied such protection in a downstream market, but that in itself is significant. If the relevant patent authorities refuse to grant protection for the intellectual property owner’s contribution in the downstream market for a reason of patentability, such as obviousness, that would be evidence that the owner should not be able to control access to that market through its upstream intellectual property.48 The same may be true even if the reason for the inability to obtain downstream protection is the fact that patent protection is unavailable for the type of product or service at issue, as in jurisdictions where business-method protection is not available. The unavailability of such protection would suggest that the balance has been struck in favor of competition for such products or services, and it would be inappropriate for the intellectual property owner to use upstream protection of a different product or service to circumvent that limitation.49 It is true that, if the monopolist has never offered its upstream intellectual property separately, mandated access raises the specter of allowing others to demand licenses for any portion of a product incorporating intellectual property: “The implications of the ‘potential market’ concept, if adopted without limiting principles, would be far-reaching. Every process patent is an input. New functions can always be added to software. A compound medicine is always a product 48 That would be so, at least, if the contribution in the downstream market were viewed as obvious in light of prior art that was not the patentee’s. If the contribution in the downstream market were made obvious as a result of the patentee’s own contribution in the upstream market, the question would be more difficult. Whether the inquiry in this context should incorporate all the details of patent law is not clear. 49 Cf. Fashion Originators’ Guild of America v. FTC, 312 U.S. 457 (1941). This point is related to the fact that in some of these cases, like Magill, the intellectual property protection at issue was not really necessary to accomplish the dominant firm’s goals, but appears simply to have been a convenient means of furthering those goals.

(Zg) Ehlermann 08 Patterson

720

6/11/08

11:56

Page 720

Session Three: Exploitative Conduct And The Remedies

different from its component medicines. Many patents are for ‘follow-on’ inventions. The law could hardly impose a duty to share important internally-generated competitive advantages with direct competitors, on demand, merely on the basis of their intention to offer a product with some new characteristics. It would be unsatisfactory if a dominant company had to decide whether it was free to refuse a licence merely on the basis of a competitor’s allegations about the degree of novelty of a product which the competitor was not yet in a position to produce, and which it would certainly be unwilling to describe in detail.”50

These concerns are valid ones, but it is not clear that they should always be decisive. It is worth noting that several of the “downstream” examples given in the quotation above would themselves likely be patented or copyrighted. That is true, for example, of software with new functions, of compound medicines, and of follow-on inventions. It will also often be true of products produced with new processes. In such instances, the intellectual property owner will be protected not only by its upstream intellectual property protection (which may, however, be limited as described here) but also by independent downstream protection. Furthermore, the essential facility doctrine itself, without regard to the intellectual property context, has safeguards that could in many cases address the concerns expressed above. For example, for access to be compelled, the upstream intellectual property has to be truly essential and impractical to duplicate, which will exclude many claims. And it must also be feasible for the intellectual property owner to license the intellectual property. That means, at the least, that the intellectual property owner is able practically to provide access to the intellectual property without prohibitive expense.51 The feasibility requirement could also be interpreted to require that access by others to the intellectual property not hinder the owner’s ability to capture the benefits of its innovation. This factor could also be considered in Europe under the rubric of the intellectual property owner’s ability to justify its refusal to license by showing an objective justification for the refusal.52 It is 50

Temple Lang, supra note 14, at 18 (footnote omitted). For example, it could perhaps be prohibitively expensive to write and enforce contractual terms that would prevent disclosure of the intellectual property beyond the party receiving access. 52 Whether an ability to reap sufficient profits from intellectual property would constitute an objective justification is not entirely clear: 51

“The status of the second condition identified by the ECJ [in Magill] depends on whether ‘objective justification’ includes the reward for, and incentive to, innovation that underlies the grant of a monopoly under IP law. The EU Courts have not to date elaborated on the meaning of this condition in the IP context so that it remains something of a black hole. To the extent the ECJ would accept that the refusal of the IP owner to grant licenses can be ‘objectively’ justified by the preservation of the incentive to innovate through the reaping of its monopoly reward, this condition would also be consistent with IP law.” Venit, supra note 4, at 11. Cf. the Commission’s Decision in Microsoft, cited supra note 27; judgment of the CFI of 27 September 2006, Case T-168/01, GlaxoSmithKline Services Unlimited v Commission [2006] ECR II-2969, on appeal: Cases C-501/06 P, C-513/06 P, C-515/06 P and C-519/06 P, not yet decided.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 721

“New Product” Requirement in European Refusal to License Cases 721 not clear that it will always be easy to design licensing arrangements that will adequately compensate the owner. For example, it is generally accepted that an intellectual property owner is entitled to price discriminate in its licensing practices. It may not be easy, however, to compel licenses in such a way that the intellectual property owner is able to maintain its discriminatory pricing.53 If not, it seems appropriate to view compelled licensing as infeasible. One might view the “new product” requirement’s attention to the presence of the intellectual property owner in the downstream market as addressing this issue.54 If the intellectual property owner is not present in the downstream market, a compelled license in that market is not likely to interfere with its licensing approach, even in other markets. Although arbitrage between markets would theoretically be possible, the intellectual property owner could presumably make use of field-of-use licensing to prevent arbitrage.

IV. Conclusion The purpose of this paper is to question the IMS Health rule’s award to an intellectual property owner of any downstream market that relies on its intellectual property, so long as the intellectual property owner serves the downstream market itself. As Temple Lang puts it, “one competitor in a downstream market must not be able to get control over the only source of supply of an input which is essential in that market, and monopolise the market by shutting off supply to its rivals”.55 Only if the upstream intellectual property owner has also made a protectable contribution in the downstream market should it be entitled to monopolize that market as well.

53 This might be true, for example, for innovations that improve production processes. Although in theory one might be able to structure payments for the use of such processes, in practice such arrangements might be difficult. Cf. Ioannis Lianos, “Competition Law and Intellectual Property Rights: Is the Property Rights Approach Right?”, in John Bell and Claire Kilpatrick, eds., Cambridge Yearbook of European Legal Studies, Hart Publishing, 2006, 153 et seq., at 163 (asking why the “new product” test appears to allow compelled licensing for new products, but not for new production processes). 54 Other commentators have suggested, on the other hand, that a likely source of that aspect of the test was a focus on the intellectual property owner’s failure to exploit its intellectual property. See Venit, supra note 4, at 2, 9; Temple Lang, supra note 14. 55 Temple Lang, supra note 14, at 6.

(Zg) Ehlermann 08 Patterson

6/11/08

11:56

Page 722

(Zh) Ehlermann 08 Lowe

6/11/08

11:56

Page 723

Philip Lowe Concluding Remarks June 2007

Colleagues, Friends, It would be presumptuous to try in fifteen minutes to summarize the debates we have had in six panels, preceded by the submission of so many papers of high quality, which provided a rich and firm basis for the discussions of the last two days. I think we have to thank Claus Ehlermann and the staff of the Institute sincerely for that. After the review of Article 82 initiated by the Commission and the hearings on Section 2 in the US, the issues which we have been addressing are now relatively familiar to all of us. At the start of the programme, the first panel tackled the fundamental objectives of competition policy with respect, in particular, to unilateral conduct. I would dare to say that we have reached a substantial degree of transatlantic convergence and consensus on objectives. Within Europe, there is broad agreement that we are protecting competition, not competitors. But there is still some outstanding debate about the ultimate objective of competition policy: principally German traditions retain the concept of the freedom to compete as such; on the other side the majority of other European jurisdictions have, similar to the US tradition, established consumer welfare as the ultimate standard, and they assess unilateral conduct in relation to its effects on consumers. However, the debate around these two opposing concepts appears, in the end, to be more at the level of theory than of practical investigation and decision-making. Generally speaking, all European jurisdictions are pursuing cases with the same concerns in mind: the acquisition and/or use or abuse of substantial market power which endangers the competition process and ultimately harms consumers. In the longer run, one can argue that even the theoretical divergences between the concept of Wettbewerbsfreiheit and welfare standards are reconcilable. It is interesting to observe how even the most traditional and ordoliberal authorities devote so much of their investigation time to the examination of effects. Maybe, as Frédéric Jenny reported, they have no time to look at efficiencies. But in general, effects, long run and short run, draw their close attention. At the same time, in terms of case investigation, I believe we had good advice from David Gerber and others that economics should be a tool of analysis of the facts, not a normative discipline. There may be many cases

(Zh) Ehlermann 08 Lowe

724

6/11/08

11:56

Page 724

Overall conclusions

where you don’t need to model anything to establish a theory of harm. In addition, legal practitioners too often assume that a more economic approach has by definition to be a more complex and sophisticated one. Surely the objective is to ensure both that ex ante rules make broad economic sense and that the story we tell in any decision on a specific case makes broad economic sense. So the messages of economics can be simple. On the next question, as to precisely how you pursue a policy on unilateral conduct and how you communicate that policy, I think we recognize, as far as Europe is concerned, that the case law we have on unilateral conduct does not provide us with a comprehensive or consistent body of principles and tests which could respectably be called a policy. In any event, in all other areas of EU competition policy—in merger control and in the field of horizontal and vertical agreements—there has been a perceived need for clarity and predictability of policy. This has been met, with or without a substantial body of case law, by providing upfront guidance, generally in the form of guidelines. This is not to say that more enforcement and more case law would not assist and consolidate policy. But the initiation and prosecution of cases is, unlike policy, not something which is entirely in the hands of a competition authority or court. It is very much dependent on the specific circumstances of the defending and other parties and may or may not offer scope for the identification of principles and tests of more general application. So there is an argument for guidance, in addition to case law. But guidance for whom and on what? First of all, I would argue, guidance is necessary for all those working within competition authorities in the jurisdiction where the substantive rules on unilateral conduct are to be applied. This means, within the EU, that we need to address guidance to our staff and case teams within the Commission and within all national competition authorities who are applying Article 82. Otherwise, we have little prospect of ensuring the coherent and consistent application of the law, which the business and legal communities in Europe expect us to develop. Imagine too a situation where the Commission alone establishes guidance, without reference to national Article 82 policy and case law, and then is left to ensure EU-wide coherence by intervening to take competence for national cases or stay national decisions. So what guidance is produced must be of general application by all competition authorities applying EU law and its drafting must involve all of them. But secondly, guidance must surely be given to courts and other review bodies. How are the courts to review the consistency of policy, without a clear indication from the responsible administrative authorities as to what policy they are trying to pursue? If a court is itself an investigating and decisionmaking body, the situation would obviously be different. But, in particular with European administrative systems, courts must rely on administrative authorities, not only for detailed investigation and analysis, but also for

(Zh) Ehlermann 08 Lowe

6/11/08

11:56

Page 725

Concluding Remarks

725

policy development. At the same time, administrative authorities must obviously defer to courts in the application of legal principles to the execution of policy. There is an inevitable, and usually constructive, interplay of administrative and judicial activity, provided both sides do what they are expected to do. And on the side of administrative authorities, policy development seems to be one of their essential functions as instruments of wider public policy. So guidance cannot remain within the confines of authorities. It should be accountable to courts and it needs to be adapted to what courts decide. Can policy guidance contradict what some call “established case law”? In principle, no, but does that mean that policy cannot be educated either by market developments or by better definition of objectives or superior investigation techniques? Surely not. And in addition, we should not be shocked by the inevitable lag which occurs between the initial decisions by an administrative authority and the judgments of courts on them, on the one hand, and the parallel development of policy by the authority on the other. But I would argue here that the provision of regularly updated policy guidance by the administrative authority should diminish the significance of this lag rather than increase it. Even more importantly, alongside active enforcement for the effectiveness of a policy, policy guidance has its greatest value in relation to those by whom it may eventually be applied—the business and legal community. Guidance should in principle allow businesses and their legal advisers to distinguish situations where conduct could be judged to be anticompetitive from those which are pro-competitive. In that sense, it should enable competition authorities to deter relatively clear-cut anticompetitive behaviour and, as a result, to concentrate enforcement action on those cases where there is a genuine need for investigation of situations where the effects on competition and consumers are not so obvious. Will published guidance bind the competition authority which issues it? This must be the case, otherwise the authority is not delivering on its obligation to apply the law in a consistent and coherent way, with the necessary level of predictability for firms. But guidance cannot be assumed to provide the same predictability as law (whether in the form of legislation or case law). The facts of any specific case—in particular the market characteristics around it— may not have been foreseen by the guidance. And in these specific circumstances, it would be normal for a court not to condemn an authority for deviating from its guidelines, but to expect from the authority a reasoned argumentation where it finds it necessary to deviate from the guidance originally given. And, of course, subject to any subsequent court decision, it would be normal for guidance to be updated in the light of market developments and of specific case decisions by the authority. Another objection made to providing upfront guidance is that it is supposedly adding further “regulation” to markets at a time where everyone is

(Zh) Ehlermann 08 Lowe

726

6/11/08

11:56

Page 726

Overall conclusions

looking for opportunities for better, if not less, regulation. This reaction does not reflect the widespread perception in the business and legal community that the lack of guidance from the Commission on what unilateral conduct will be regarded as anticompetitive creates an unacceptable degree of uncertainty. This uncertainty engages the time and resources of firms in the Kremlinology of working out what conduct may or may not be acceptable to European competition authorities. Far from increasing the administrative burden on companies, useful guidance on unilateral conduct can have no other effect than to reduce it, while avoiding further regulation. Formal guidance offers an adequate degree of legal certainty. On the other hand, it is flexible enough to allow the administrative authority to adapt it to market developments and specific case situations. Turning now to the issue of precisely what level of guidance can and should be given in the area of unilateral conduct, I believe we have reached the conclusion that it is possible to draft useful guidance on the overall policy objectives and the general framework for analysis of substantial and durable market power (dominance) and of the parameters which may determine whether conduct has the actual or likely effect of anticompetitive foreclosure (that is, foreclosure that brings harm to consumers). To provide useful guidance, I think we can go on to say that, where possible, guidance should provide safe harbours, rebuttable presumptions and perhaps some tests. This simplifies compliance issues. It also reduces the administrative burden of investigation and wider scope for more cases to be brought—which would certainly please Mario Siragusa and Ian Forrester, after the remarks they made about our enforcement record. However we have to ensure that any test is going to allow us to arrive at the correct result on a repetitive basis to a high degree of probability. That is not always easy. It may be more advisable, in particular when looking at specific practices, to refrain from being too presumptive. In that respect, we can probably be more confident of specific bright-line tests, in the area of, for example, predatory pricing, and some forms of rebates. But we should be much more cautious establishing upfront tests and rules for practices such as tying and bundling or refusal to deal. In the development of formal policy guidance, we would perhaps be well advised to adopt a progressive approach, starting off with an emphasis on principles and methodology and then developing more precise precepts on the basis of case experience. However, we should equally be conscious not to confuse the need for caution with some implicit or explicit presumption of per se legality, just as in establishing bright lines we should be careful to avoid a presumption of illegality without any empirical evidence to justify it. In making transatlantic comparisons, a number of interventions reminded us that public enforcement of Section 2 in the US is strongly complemented by private action. So it is probably incorrect to jump to the conclusion that the level of overall enforcement is very different between Europe and the US.

(Zh) Ehlermann 08 Lowe

6/11/08

11:56

Page 727

Concluding Remarks

727

We also need to take into account the fact that a great deal of national enforcement within Europe in relation to unilateral conduct relates to issues of vertical economic dependence and to unfair trade practices, and not to an exclusive application of Article 82. In addition, another significant aspect of national European enforcement relates to exploitative abuses. In that context, we had an illuminating discussion on the advantages and disadvantages of using competition law instruments or relying on the intervention of regulators. There was a strong consensus on the relatively limited conditions under which Article 82 could be used for exploitative conduct. That would seem to offer a solid basis for inclusion of policy towards exploitative conduct in any guidance. There was equally a recognition on both the EU and the US side that sectoral regulation can sometimes be quicker and more effective than long, drawn-out antitrust investigations. Regulators are often perceived as doing something somehow “better” than competition authorities for consumers, especially on price. Maybe the overall conclusion here is that competition authorities cannot function too independently from the market and the public policy environment around them. If there is a perceived competition problem, we should surely be looking at what is the appropriate instrument to tackle it (legislation, sector regulation, action of regulators, application of Article 82) rather than leaping to use the instrument which competition authorities have in their own competence. In our last Panel VI of this Workshop, we looked at price regulation in the energy sector. I think our speeches rightly warned against the dangers of competition authorities attempting to control market outcomes, especially when it is done on the basis of some false economic premises. Finally, we all agreed that the protection of intellectual property rights on the one hand, and competition on the other could be mutually reinforcing if there was a better dialogue between competition authorities and patent offices.

(Zh) Ehlermann 08 Lowe

6/11/08

11:56

Page 728

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 729

BIBLIOGRAPHY

I. Publications, speeches and presentations Adams, Walter, “Antitrust, Laissez-faire and Economic Power”, in Fritz Neumark et al., eds., Wettbewerb, Konzentration, und wirtschaftliche Macht, Duncker & Humboldt, 1976 Afferni, Giorgio, “Case: ECJ—Manfredi v Lloyd Adriatico”, 3 European Review of Contract Law 179 (2007); Aghion, Philippe, Nick Bloom, Richard Blundell, Rachel Griffith and Peter Howitt, “Competition and Innovation: An Inverted-U Relationship”, 120 Quarterly Journal of Economics 701 (2005) Aghion, Philippe and Patrick Bolton, “Contracts as a Barrier to Entry”, 77 American Economic Review 388 (1987) Ahlborn, Christian, David Evans and Jorge Padilla, “The Antitrust Economics of Tying: A Farewell to Per Se Illegality”, 49 Antitrust Bulletin 287 (2003) Ahlborn, Christian, David Evans and Jorge Padilla, “The Logic and Limits of the ‘Exceptional Circumstances’ Test in Magill and IMS Health”, 28 Fordham International Law Journal 1109 (2005) Ahlborn, Christian, David Bailey and Helen Crossley, “An Antitrust Analysis of Tying: Position Paper”, GCLC Research Papers on Article 82 EC 166 (2005), http://ec.europa.eu/comm/competition/mergers/legislation/draft_nonhorizontal_ mergers.pdf Ahlborn, Christian, Vincenzo Denicolò, Damien Geradin and Jorge Padilla, “DG Comp’s Discussion Paper on Article 82: Implications of the Proposed Framework and Antitrust Rules for Dynamically Competitive Industries” (March 2006), http://ec.europa.eu/comm/competition/antitrust/others/057.pdf Akman, Pinar, “Searching for the Long-Lost Soul of Art. 82”, CCP Working Paper No. 07-5 (March 2005), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=977221 Alese, Femi, “(Fore)closing the Gap: The Commission’s Non-Horizontal Merger Guidelines—A Response to Simon Bishop”, 29 European Competition Law Review 196 (2008) Alesina, Alberto and Guido Tabellini, “Bureaucrats or Politicians? A Single Policy Task, Part I”, 97 American Economic Review 169 (2007) Amato, Giuliano, Antitrust and the Bounds of Power, Hart Publishing, 1997 Amato, Giuliano and Claus-Dieter Ehlermann, eds., EC Competition Law: A Critical Assessment, Hart Publishing, 2007 Anand, Bharat and Tarun Khanna, “The Structure of Licensing Contracts”, in 48 Journal of Industrial Economics 103 (2000) Anderman, Steven, EC Competition Law and Intellectual Property Rights: The Regulation of Innovation, Oxford University Press, 1998

(Zi) Ehlermann 08 Biblio

730

6/11/08

11:57

Page 730

Bibliography

Anderman, Steven and John Kallaugher, Technology Transfer and the New EU Competition Rules—Intellectual Property Licensing after Modernisation, Oxford University Press, 2006 Areeda, Phillip, Antitrust Law: An Analysis of Antitrust Principles and Their Application (1986) Areeda, Phillip, “Essential Facilities: An Epithet in Need of Limiting Principles”, 58 Antitrust Law Journal 841 (1989) Areeda, Phillip and Donald Turner, “Scherer on Predatory Pricing: A Reply”, 89 Harvard Law Review 891 (1976) Areeda, Phillip and Herbert Hovenkamp, Fundamentals of Antitrust Law, Aspen Publishing, 2003 Armentano, Domenick, Antitrust and Monopoly—Anatomy of a Policy Failure, 2nd ed., The Independent Institute, 1999 Armstrong, Mark and John Vickers, “Price Discrimination, Competition, and Regulation”, 41 Journal of Industrial Economics 335 (1993) Arora, Ashish, Andrea Fosfuri and Alfonso Gambardella, Markets for Technology: Economics of Innovation and Corporate Strategy, MIT Press, 2001 Bacon, Kelyn, “European Court of Justice Upholds Judgment of the European Court of First Instance in the British Airways/Virgin Saga”, 3(2) Competition Policy International 227 (2007) Baker, Jonathan, “Recent Developments in Economics that Challenge Chicago School Views”, 58 Antitrust Law Journal 645 (1989) Baker, Jonathan, “Beyond Schumpeter vs. Arrow: How Antitrust Fosters Innovation” (June 2007), http://ssrn.com/abstract=962261 Barnett, Thomas, “The Gales of Creative Destruction: The Need for Clear and Objective Standards for Enforcing Section 2 of the Sherman Act”, Opening Remarks for the Antitrust Division and Federal Trade Commission Hearings Regarding Section 2 of the Sherman Act, Washington, D.C., 20 June 2006 Barnett, Thomas, “Recent Developments in Antitrust and Intellectual Property”, speech delivered in New York, 16 May 2007, http://www.justice.gov/atr/public/ speeches/223390.htm Baumol, William and Daniel Swanson, “The New Economy and Ubiquitous Competitive Price Discrimination: Identifying Defensible Criteria of Market Power,” 70 Antitrust Law Journal 661 (2003) Baumol, William, The Free-Market Innovation Machine: Analyzing the Growth Miracle of Capitalism, Princeton University Press, 2004 Bekkers, Rudi, “Patent drag and stacking IPR fees—Are the IPR policies of standards bodies failing or should we better address excessive technology inclusion?” Position paper for the High-Level Workshop on standardization, IP licensing and antitrust organized by the Tilburg Law and Economic Center (TILEC), Tilburg University at Chateau du Lac, Brussels (January 2007), http://home.tm.tue.nl/rbekkers Bellamy, Christopher and Graham Child, European Community Law of Competition, 2nd ed., Sweet & Maxwell, 1978 Bishop, Simon, “(Fore)closing the Gap: The Commission’s Non-Horizontal Merger Guidelines”, 29 European Competition Law Review 1 (2008) Bishop, Simon, “The Elephant in the Room: Re-Thinking the Role of Dominance in Article 82 Investigations”, European Competition Journal, forthcoming

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 731

Bibliography

731

Bishop, Simon and Simon Baker, “The Role of Market Definition in Dominance and Monopoly Enquiries”, Office of Fair Trading Economic Discussion Paper 2 (November 2001) Bishop, Simon and Mike Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, Sweet and Maxwell, 2002 Blumenthal, William, “Clear Agency Guidelines: Lessons from 1982”, 68 Antitrust Law Journal 5 (2000) Blumenthal, William, “The Relationship between Competition Agencies and Other Units of Government¸ before International Seminar: Review of AntiMonopoly Law” (19 May 2006), http://www.ftc.gov/speeches/blumenthal/ 20060519Mofcom-ADBFinal.pdf Böhm, Franz, Wettbewerb und Monopolkampf. Eine Untersuchung zur Frage des wirtschaftlichen Kampfrechts und sur Frage der rechlichen Struktur der geltenden Wirtschaftsordnung, Carl Heymann Verlag, 1933/1964 Böhm, Franz, “The Non-State (‘Natural’) Laws Inherent in a Competitive Economy”, in Wolfgang Stützel et al., eds., Standard Texts on the Social Market Economy, Gustav Fischer, 1982 Boland, Lawrence, The Methodology of Economic Model Building, Routledge, 1989 Bolton, Patrick, Joseph Brodley and Michael Riordan, “Predatory Pricing: Strategic Theory and Legal Policy”, 88 Georgetown Law Journal 2239 (2000) Bork, Robert, “The Rule of Reason and the Per Se Concept: Price Fixing and Market Division (part I)”, 74 Yale Law Journal 775 Bork, Robert, “Legislative Intent and the Policy of the Sherman Act”, 9 Journal of Law and Economics 7 (1966) Bork, Robert, The Antitrust Paradox: A Policy at War with Itself, Free Press, 1978; second edition, Free Press, 1993 Bowman, Ward, “Tying Arrangements and the Leverage Problem”, 67 Yale Law Journal 19 (1957) Breyer, Stephen, “Analyzing Regulatory Failure: Mismatches, Less Restrictive Alternatives, and Reform”, 92 Harvard Law Review 547 (1979) Breyer, Stephen, Regulation and its Reform (1982) Brittan, Leon, European Competition Law: Keeping the Playing-Field Level, Brassey’s, 1992 Brodley, Joseph and George A. Hay, “Predatory Pricing: Competing Economic Theories and the Evolution of Legal Standards”, 66 Cornell Law Review 738 (1981) Burstein, Mayer, “The Economics of Tie-In Sales,” 42 Review of Economics and Statistics 68 (1960) Carlton, Dennis and Michael Waldman, “The Strategic Use of Tying to Preserve and Create Market Power in Evolving Industries”, 33 RAND Journal of Economics 194 (2002) Carlton, Dennis and Michael Waldman, “Tying, Upgrades and Switching Costs in Durable-Goods Markets,” NBER Working Paper 11407 (2005) Chamberlin, Edward, Theory of Monopolistic Competition: A Reorientation of the Theory of Value, 8th ed., Harvard University Press (1933/1962) Chen Zhiqi and Thomas Ross, “Refusals to Deal, Price Discrimination and Independent Service Organizations,” 2 Journal of Economics and Management Strategy 593 (1983)

(Zi) Ehlermann 08 Biblio

732

6/11/08

11:57

Page 732

Bibliography

Choi, Jay Pil and Christodoulos Stefanadis, “Tying, Investment and the Dynamic Leverage Theory,” 32 RAND Journal of Economics 52 (2001) Clark, John, “Toward a Concept of Workable Competition”, 30 American Economic Review 241 (1940) Clark, John, Competition as a Dynamic Process, The Brookings Institution, 1961 Coase, Ronald, “Adam Smith’s View on Man”, in 19 Journal of Law and Economics 529 (1976) Comanor, William, “Vertical Market Restrictions, and the New Antitrust Policy”, 98 Harvard Law Review 983 (1985) Connelly, Mark, “Exclusive Dealing and Tied Selling Under the Amended Combines Investigation Act”, 14 Osgoode Hall Law Journal 521 (1976) Cournot, Augustin, Reserches sur les Principes Mathématiques de la Théorie des Richesses, Hachette, 1838 Coyle, Diane, The Soulful Science: What Economists Do and Why it Matters, Princeton University Press, 2007 Cultrera, Concetta, “Les décisions GDF—La Commission est formelle : les clauses de restriction territoriale dans les contrats de gaz violent l’article 81”, Competition Policy Newsletter, No. 1, 2005 De la Mano, Miguel, “The dominance concept: new wine in old bottles”, oral remarks at the FTC/DOJ’s Sherman Act Section 2 Joint Hearing on Understanding Single-Firm Behavior: Monopoly Power Session, Washington, D.C., 7 March 2007 De Smijter, Eddy and Lars Kjølbye, “The Enforcement System under Regulation 1/2003”, in Jonathan Faull and Ali Nikpay (eds.), The EC Law of Competition, 2d edition, Oxford University Press, 2007, chapter 2 Dibadj, Reza, “Article 82: Gestalt, Myths, Questions”, 23 Santa Clara Computer and High Tech Journal 615 (2007) DeLacey, Brian, Kerry Herman, David Kiron and Josh Lerner, “Strategic Behavior in Standard-Setting Organizations”, Harvard NOM Working Paper No 903214 (September 2006), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=903214 Demsetz, Harold, “Barriers to Entry”, 72 American Economic Review 47 (1982) Dewatripont, Mathias and Jean Tirole, “Advocates”, 99 Journal of Political Economy 1 (1999) Director, Aaron and Edward Levi, “Law and the Future: Trade Regulation”, 51 Northwestern University Law Review 281 (1956) Dolmans, Mauritz, “Standards for Standards”, 26 Fordham International Law Journal 163 (2002) Dolmans, Maurits, “Efficiency Defences Under Article 82 EC Seeking Profits Or Proportionality? The EC 2004 Microsoft Case in the Context of Trinko,” 24th Annual Antitrust And Trade Regulation Seminar, NERA, Santa Fe, New Mexico 8 July 2004 Drexl, Josef, “Intellectual Property and Antitrust Law-IMS Health and TrinkoAntitrust Placebo for Consumers Instead of Sound Economics in Refusal-to-Deal Cases”, 35 ICC: International Review of Intellectual Property and Competition Law 788 (2004) Drexl, Josef, “Abuse of Dominance in Licensing and Refusal to License: A ‘More Economic Approach’ to Competition by Imitation and to Competition by Substitution”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 733

Bibliography

733

Competition Law Annual 2005: The Interaction between Competition Law and Intellectual Property Law, Hart Publishing, 2007 Easterbrook, Frank, “The Limits of Antitrust”, 63 University of Texas Law Review 1 (1984) Eccles, Richard, “Parallel Exports in the Pharmaceutical Sector: Take Nothing for Granted”, 28 European Competition Law Review 134 Edlin Aaron, “Stopping Above-Cost Predatory Pricing”, 111 Yale Law Journal 941 (2002) Ehlermann, Claus-Dieter and Laraine Laudati, eds., European Competition Law Annual 1997: Objectives of Competition Policy, Hart Publishing, 1998 Ehlermann, Claus-Dieter and Louisa Gosling, eds., European Competition Law Annual 1998: Regulating Communications Markets, Hart Publishing, 2000 Ehlermann, Claus-Dieter and Isabela Atanasiu, eds., European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law, Hart Publishing, 2003 Ehlermann, Claus-Dieter and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006 Ehlermann, Claus-Dieter and Isabela Atanasiu, eds., European Competition Law Annual 2005: The Interaction between Competition Law and Intellectual Property Law, Hart Publishing, 2007 Ehlermann, Claus-Dieter and Isabela Atanasiu, eds., European Competition Law Annual 2006: Enforcement of Prohibition of Cartels, Hart Publishing, 2007 Eilmansberger, Thomas, “How to Distinguish Good from Bad Competition under Article 82 EC: In Search of Clearer and More Coherent Standards for AntiCompetitive Abuses”, 42 Common Market Law Review 129 (2005) Elhauge, Einer, “Defining Better Monopolization Standards”, 56 Stanford Law Review 253 (2003) Elhauge, Einer, “Why Above-Cost Price Cuts to Drive out Entrants are Not Predatory—and the Implications for Defining Costs and Market Power”, 112 Yale Law Journal 681 (2003) Elhauge, Einer and Damien Geradin, Global Competition Law and Economics, Foundation Press (US case book version), 2007; Hart Publishing (global version), 2007 Engel, Cristoph, “Imposed liberty and its limits: The EC treaty as an economic constitution for the Member States”, in Talia Einhorn (ed.), Spontaneous Order, Organization and the Law. Roads to a European Civil Society. Liber Amicorum Ernst-Joachim Mestmäcker, Asser Press (2003), pp. 429 et seq. Epstein, Richard, “Monopoly Dominance or Level Playing Field—the New Antitrust Paradox”, 72 University of Chicago Law Review 49 (2005) Esteva Mosso, Carles, Stephen Ryan, Svend Albaek and María Luisa Tierno Centella, “Article 82”, in Jonathan Faull and Ali Nikpay, eds., The EC Law of Competition, 2d edition, Oxford University Press, 2007, chapter 4 Eucken, Walter, “Überlegungen zum Monopolproblem“, in Wirtschaftsmacht und Wirtschaftsordnung, Walter Eucken Archiv (LIT), 2001 Eucken, Walter, Grundsätze der Wirtschaftspolitik, 7th ed., Mohr Siebeck, 2004 Eucken, Walter, “The Competitive Order and its Implementation”, 1949, reprinted in 2(2) Competition Policy International 222 (2006)

(Zi) Ehlermann 08 Biblio

734

6/11/08

11:57

Page 734

Bibliography

Evans, David, “How Economists Can Help Courts Design Competition Rules: An EU and US Perspective”, 28 World Competition 93 (2005) Evans, David and Michael Salinger, “Why Do Firms Bundle and Tie? Evidence from Competitive Markets and Implications for Tying Law”, 22 Yale Journal on Regulation 37 (2005) Evans, David and Jorge Padilla, “Excessive Prices: From Economic Theory to Administrable Rules”, 1 Journal of Competition Law and Economics 97 (2005) Evans, David and Jorge Padilla, “Designing Antitrust Rules for the Assessment of Unilateral Practices: A Neo-Chicago Approach”, 72 University of Chicago Law Review 73 (2005) Evans, David, Jorge Padilla and Michael Salinger, “A Pragmatic Approach to Identifying and Analyzing Legitimate Tying Cases”, in Claus-Dieter Ehlermman and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, at pp. 557 et seq. Evans, David and Michael Salinger, “Why do Firms Bundle and Tie? Evidence from Competitive Markets and Implications for Tying Law”, 22 Yale Journal on Regulation 37 (2005) Evans, David and Richard Schmalensee, “Some Economic Aspects of Antitrust Analysis in Dynamically Competitive Industries,” in Josh Lerner and Scott Stern, eds., Innovation Policy and the Economy, vol. 2, MIT Press, 2002, pp. 1 et seq. Facey, Brian and Roger Ware, “Predatory Pricing in Canada, the United States and Europe: Crouching Tiger or Hidden Dragon?”, 26 World Competition 625 (2003) Faull, Jonathan and Ali Nikpay, eds., The EC Law of Competition, Oxford University Press, 1999 Forrester, Ian, “The Modernisation of EC Antitrust Policy: Compatibility, Efficiency, Legal Security”, in Claus-Dieter Ehlermann and Isabela Atanasiu, European Competition Law Annual 2000: The Modernisation of EC Antitrust Policy, Hart Publishing, 2001, pp. 115 et seq. Forrester, Ian, “Article 82: Remedies in Search of Theories?”, 28 Fordham International Law Journal 919 (2005) Fox, Eleanor, “The Modernization of Antitrust—A New Equilibrium”, 66 Cornell Law Review 1140 (1981) Fox, Eleanor, “Monopolization and Dominance in the United States and the European Community: Efficiency, Opportunity, and Fairness”, 61 Notre Dame Law Review 981 (1986) Fox, Eleanor, “What is Harm to Competition?—Exclusionary Practices and Anticompetitive Effect”, 70 Antitrust Law Journal 371 (2002) Fox, Eleanor, “We Protect Competition, You Protect Competitors”, 26 World Competition 155 (2003) Fox, Eleanor, “A Tale of Two Jurisdictions and an Orphan Case: Antitrust, Intellectual Property, and Refusals to Deal”, 28 Fordham International Law Journal 952 (2005) Fox, Eleanor, “Is There Life in Aspen After Trinko?: The Silent Revolution of Section 2 of the Sherman Act”, 73 Antitrust Law Journal 153 (2005) Fox, Eleanor, “Monopolization, Abuse of Dominance, and the Indeterminacy of Economics. The US/EU Divide”, 2006 Utah Law Review 799 (2006) Fox, Eleanor, “Abuse of Dominance and Monopolisation: How to Protect Competition Without Protecting Competitors”, in Claus-Dieter Ehlermann and

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 735

Bibliography

735

Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 69 et seq. Franklin, Benjamin, Writings, Library of America, 1987. Friedman, Benjamin, The Moral Consequences of Economic Growth, Knopf, 2005 Fumagalli, Chiara and Massimo Motta, “Exclusive Dealing and Entry, When Buyers Compete”, 96 American Economic Review 785 (2006) Gal, Michal, “Monopoly pricing as an antitrust offense in the U.S. and the EC: Two systems of belief about monopoly?”, 49 Antitrust Bulletin 343 (2004) Gallini, Nancy and Brian Wright, “Technology Transfer under Asymmetric Information”, 21 Rand Journal of Economics 147 (1990) Gambardella, Alfonso, Science and Innovation, Cambridge University Press, 1995 Gavil, Andrew, “Exclusionary Distribution Strategies by Dominant Firms: Striking a Better Balance”, 72 Antitrust Law Journal 3 (2004) Geradin, Damien, “Limiting the scope of the Article 82 EC: What can the EU learn from the U.S. Supreme Court’s judgment in Trinko in the wake of Microsoft, IMS, and Deutsche Telekom?”, in 41 Common Market Law Review 1519 (2004) Geradin, Damien, “Standardization and Technological Innovation: Some Reflections on Ex-ante Licensing, FRAND, and the Proper Means to Reward Innovators”, 29 World Competition 511 (2006) Geradin, Damien, “The Necessary Limits to the Control of ‘Excessive’ Prices by Competition Authorities—A View from Europe”, Tilburg University Legal Studies Working Paper No 2007-032 (October 2007), http://www.tilburguniversity.nl/tilec/ publications/discussionpapers/2007-032.pdf Geradin, Damien, Anne Layne-Farrar and Jorge Padilla, “Royalty Stacking In High Tech Industries: Separating Myth from Reality”, http://papers.ssrn.com/sol3/ papers.cfm?abstract_id=949599 Geradin, Damien, Anne Layne-Farrar and Jorge Padilla, “The Ex Ante Auction Model for the Control of Market Power in Standard Setting Organizations,” http://papers.ssrn.com/sol3/papers.cfm?abstract_id=979393 Geradin, Damien and Anne Layne-Farrar, “The Logic and Limits of Ex Ante Competition in a Standard-Setting Environment”, 3(1) Competition Policy International 78 (2007) Geradin, Damien and Nicolas Petit, “Price Discrimination Under EC Competition Law: Another Antitrust Doctrine in Search of Limiting Principles?”, 2 Journal of Competition Law and Economics 479 (2006) Geradin, Damien and Miguel Rato, “Excessive prices: In reply”, Competition Law Insight (10 October 2006) Gerber, David J., “Constitutionalizing the Economy: German Neo-liberalism, Competition Law, and the ‘New Europe’”, 42 American Journal of Comparative Law 25 (1994) Gerber, David, Law and Competition in Twentieth Century Europe: Protecting Prometheus, Oxford University Press, 1998 (paperback 2001) Gerber, David, “Competition Law”, in Peter Cane and Mark Tushnet, eds., Oxford Handbook of Legal Studies, Oxford University Press, 2004, pp. 510 et seq. Gerber, David, “Courts as Experts in European Merger Law”, in Barry Hawk, ed., Fordham Corporate Law Institute 2003: International Antitrust Law and Policy, Juris Publishing, 2004, pp. 475 et seq.

(Zi) Ehlermann 08 Biblio

736

6/11/08

11:57

Page 736

Bibliography

Gerber, David, “Comparative Antitrust Law”, in Mathias Reimann and Reinhard Zimmermann, eds., Oxford Handbook of Comparative Law, Oxford University Press, 2007, pp. 1193 et seq. Gerber, David, “Private Antitrust Enforcement in the U.S. and Europe: A Comparative Perspective”, in Thomas Möllers and Andreas Heinemann, eds., The Enforcement of Competition Law in Europe, Cambridge University Press, 2007 Gerber, David, “Competition Law and the Institutional Embeddedness of Economics”, in Josef Drexl, Laurence Idot and Joël Monéger, eds., The Crisis in Competition Law: A Comparative Perspective (approximate title), forthcoming 2008 Ghidini, Gustavo, Intellectual Property and Competition Law: The Innovation Nexus, Edward Elgar, 2006 Gilbert, Richard and Carl Shapiro, “Optimal Patent Length and Breadth”, 21 Rand Journal of Economics 106 (1990) Gilbert, Richard, “Holding Innovation to an Antitrust Standard”, 3(1) Competition Policy International 48 (2007) Glader, Marcus and Sune Chabert Larsen, “Article 82: Excessive pricing—An outline of the legal principles relating to excessive pricing and their future application in the field of IP rights and industry standards”, Competition Law Insight (4 July 2005) Goldman, Calvin and John Bodrug, eds., Competition Law of Canada, Juris Publishing, 2004 Grant, Jeremy and Damien Neven, “The Attempted Merger Between General Electric and Honeywell: A Case Study of Transatlantic Conflict”, 1 Journal of Competition Law and Economics 595 (2005) Gregor, Mary, Laws of Freedom, Basil Blackwell, 1963 Groeben, Hans v.d., Deutschland und Europa in einem unruhigen Jahrhundert, Nomos, 1995 Gu, Feng and Baruch Lev, “Markets in Intangibles: Patent Licensing” in Intangible Assets, Intellectual and Human Capital, University of Ottawa, 2003 Gyselen, Luc, “Rebates-Competition on the Merits or Exclusionary Practice?”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., 2003 European Competition Law Annual—What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 287 et seq. Hallett, Andrew, “Econometrics and the theory of economic policy: The TinbergenTheil Contributions 40 Years On”, 41 Oxford Economic Papers 189 (1989) Hart, Oliver and Jean Tirole, “Vertical Integration and Market Foreclosure”, Brookings Papers on Economic Activity (Microeconomics) 205 (1990) Hatje, Armin, “The Economic Constitution”, in Armin von Bogdandy and Jürgen Bast, eds., Principles of European Constitutional Law, Hart Publishing, 2006 Hawk, Barry, “Article 82 and Section 2: Abuse and Monopolizing Conduct”, in Wayne D. Collins, ed., Issues in Competition Law and Policy, ABA Section of Antitrust Law, forthcoming Hayek, F.A., “The Use of Knowledge in Society”, 35 American Economic Review 519 (1945) Hayek, F.A., Individualism and Economic Order, University of Chicago Press, 1948

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 737

Bibliography

737

Hekscher, Eli, Der Merkantilismus, volume 1, Gustav Fischer, 1932, pp. 448 et seq. Hellwig, Martin, “Effizienz oder Wettbewerbsfreiheit? Zur normativen Grundlegung der Wettbewerbspolitik”, in Christoph Engel and Wernhard Möschel, eds., Recht und spontane Ordnung, FS für Ernst-Joachim Mestmäcker zum 80. Geburtstag, Nomos, 2006, pp. 233 et seq. Heyer, Ken, “A World of Uncertainty: Economics and the Globalization of Antitrust”, 72 Antitrust Law Journal 375 (2005) Heyer, Ken, “Welfare Standards and Merger Analysis: Why Not the Best?”, 2(2) Competition Policy International (2006) Hovenkamp, Herbert, “Antitrust Policy after Chicago”, 84 Michigan Law Review 213 (1985) Hovenkamp, Herbert “Post-Chicago Antitrust: A Review and Critique”, 2001 Columbia Business Law Review 257 (2001) Hovenkamp, Herbert, “Exclusion and the Sherman Act”, 72 University of Chicago Law Review 147 (2005) Hovenkamp, Herbert, Federal Antitrust Policy: The Law of Competition and its Practice, 3rd ed., 2005 Hovenkamp, Herbert, The Antitrust Enterprise: Principle and Execution, Harvard University Press, 2006 Hovenkamp, Herbert, “Signposts of Anticompetitive Exclusion: Restraints on Innovation and Economies of Scale”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007, chapter 18 Hovenkamp, Herbert, Mark Janis and Mark Lemley, IP and Antitrust: An Analysis of Antitrust Principles Applied to Intellectual Property Law, Aspen Law & Business, 2002 (looseleaf) Hovenkamp, Herbert, Mark Janis and Mark Lemley, “Unilateral Refusals to License in the U.S.”, John M. Olin Program in Law and Economics Working Paper No. 303, Stanford Law School Law (April 2005), http://papers.ssrn.com/sol3/papers.cfm? abstract_id=703161 Howarth, David, “The Court of First Instance in GE/Honeywell”, 27 European Competition Law Review 485 (2006) Howarth, David and Kathryn McMahon, “’Windows has performed an illegal operation’: The Court of First Instance’s Judgment in Microsoft v Commission”, 29 European Competition Law Review 117 (2008) Hume, Christophe and Cyril Ritter, “Refusal to Deal”, College of Europe Working Paper (July 2005) Hylton, Keith, Antitrust Law, Economic Theory and Common Law Evolution, Cambridge University Press, 2003 Hylton, Keith, “Section 2 and Article 82: A Comparison of American and European Approaches to Monopolization Law”, Boston University Law Working Paper No. 06-11, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=902655 Jacobson, Jonathan and Scott Sher, “The No Economic Sense Test Makes No Sense for Exclusive Dealing”, 73 Antitrust Law Journal 779 (2006) Joerges, Christian, “The Market Without a State? States Without a Market? Two Essays on the Law of the European Economy”, EUI Working Paper LAW No. 96/2 Joerges, Christian, “What is Left of the European Economic Constitution? A Meloncholic Eulogy”, 30 European Law Review 461 (2005)

(Zi) Ehlermann 08 Biblio

738

6/11/08

11:57

Page 738

Bibliography

Joliet, René, Monopolization and Abuse of Dominant Position. A Comparative Study of the American and European Approaches to the Control of Economic Power, Martinus Nijhoff, 1970 Jones, Alison and Brenda Sufrin, EC Competition Law, Oxford University Press, 2004; third edition, OUP, 2007 Joskow, Paul, “Transaction Cost Economics, Antitrust Rules and Remedies”, 18 Journal of Law, Economics and Organization 95 (2002) Junod, Valérie, “An End to Parallel Imports of Medicines? Comments on the Judgment of the Court of First Instance in GlaxoWellcome, 30 World Competition 291 (2007) Kallaugher, John and Brian Sher, “Rebates Revisited: Anti-Competitive Effects and Exclusionary Abuse under Article 82”, 25 European Competition Law Review 263 (2004) Kamien, Morton, “Patent Licensing,” in Robert Aumann and Sergiu Hart, eds., Handbook of Game Theory with Economic Applications, Vol. 1, North-Holland, 1992, pp. 331 et seq. Kamiyama, Shigeki, Jerry Sheehan and Catalina Martinez, “Valuation and Exploitation of Intellectual Property”, STI Working Paper 2006/5, http://www.oecd.org/dataoecd/ 62/52/37031481.pdf Kant, Immanuel, The Doctrine of Virtue (Mary Gregor, trans.), Harper & Row, 1964 Kant, Immanuel, The Metaphysical Elements of Justice (John Ladd, trans.), BobbsMerrill, 1965; second edition, Hackett Publishing, 1999 Kant, Immanuel, On the Old Saw: That May Be Right in Theory But It Won’t Work in Practice (E.B. Ashton, trans.), University of Pennsylvania Press, 1974 Kaplow, Louis, “The Patent-Antitrust Intersection: A Reappraisal”, 97 Harvard Law Review 1813 (1984) Kaplow, Louis and Steven Shavell, Fairness versus Welfare, Harvard University Press, 2002 Kaplow, Louis and Steven Shavell, “Fairness versus Welfare: Notes on the Pareto Principle, Preferences, and Distributive Injustice”, 32 The Journal of Legal Studies 331 (2003) Katz, Michael and Howard Shelanski, “Mergers and Innovation”, 74 Antitrust Law Journal 1 (2007) Kenney, Roy and Benjamin Klein, “The Economics of Block Booking”, 26 Journal of Law and Economics 497 (1983) Klees, Andreas, “Breaking the Habits: The German Competition Law after the 7th Amendment to the Act against Restraints of Competition (GWB)”, 7(4) German Law Journal 399 (2006) Kobayashi, Bruce, “The economics of loyalty discounts and antitrust law in the United States”, 1 Competition Policy International 115 (Autumn 2005) Kokkoris Ioannis, “Buyer Power Assessment in Competition Law: A Boon or a Menace?”, 29 World Competition 139 (2006) Kolasky, William, “What is Competition? A Comparison of U.S. and EU Perspectives”, 49 Antitrust Bulletin 29 (2004) Komninos, Assimakis, “New Prospects for Private Enforcement of EC Competition Law: Courage v. Crehan and the Community Right to Damages”, 39 Common Market Law Review 447 (2002)

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 739

Bibliography

739

Komninos, Assimakis, “Enter the White Paper: A First Selective Appraisal”, Global Competition Policy (online paper of 3 April 2008) Kovacic, William, “The Intellectual DNA of Modern U.S. Competition Law for Dominant Firm Conduct: The Chicago/Harvard Double Helix”, 2007 Columbia Business Law Review 1 (2007) Kovacic, William, “The Importance of History to the Design of Competition Policy Strategy: The Federal Trade Commission and Intellectual Property Law”, 30 Seattle University Law Review 319 (2007) Krattenmaker, Thomas and Steven Salop, “Anticompetitive Exclusion: Raising Rivals’ Costs to Achieve Power over Price”, 96 Yale Law Journal 209 (1986) Kroes, Neelie, “Some Preliminary Thoughts on the Policy Review of Article 82”, SPEECH/05/537 delivered at Fordham Law School, New York, 23 September 2005 and published in Barry Hawk, ed., International Antitrust Law & Policy: Fordham Corporate Law Institute 2005, Juris Publishing, 2006, chapter 16 Kroes, Neelie, “European Competition Policy—Delivering Better Markets and Better Choices”, SPEECH 05/512, London, 15 September 2005 Kroes, Neelie, “The Law and Economics of State aid control—a Commission Perspective”, SPEECH/07/601, Berlin, 8 October 2007 Kühn, Kai-Uwe, “Reforming European merger review: targeting problem areas in policy outcomes”, 2 Journal of Industry, Competition and Trade 311 (2002) Küsters, Hanns Jürgen, Die Gründung der Europäischen Wirtschaftsgemeinschaft, Nomos, 1982 Kymlicka, Will, Contemporary Political Philosophy, 2nd ed., Oxford University Press, 2002, chapter 2 Lande, Robert, “The Rise and (Coming) Fall of Efficiency as the Rule of Antitrust”, 33 Antitrust Bulletin 429 (1988) Lande, Robert, “Wealth Transfers and the Original and Primary Concern of Antitrust: The Efficiency Interpretation Challenged”, 34 Hastings Law Journal 65 Lande, Robert, “Proving the Obvious: The Antitrust Laws Were Passed to Protect Consumers (Not Just to Increase Efficiency)”, 50 Hastings Law Journal 959 (1999) Landes, William and Richard Posner, “Market Power in Antitrust Cases”, 94 Harvard Law Review 937 (1981) Langenfeld, James, “Intellectual Property and Antitrust: Steps Toward Striking a Balance”, 52 Case Western Reserve Law Review 91 (2001) Lao, Marina, “Defining Exclusionary Conduct Under Section 2: The Case for NonUniversal Standards”, in Barry Hawk, ed., 2006 Fordham Competition Law Institute, Juris Publishing, 2007 Lemley, Mark, “Intellectual Property Rights and Standard-Setting Organizations”, 90 California Law Review 1889 (2002) Lemley, Mark and Carl Shapiro, “Probabilistic Patents”, 19 Journal of Economic Perspectives 75 (2005) Lemley, Mark and Carl Shapiro, “Patent Hold Up and Royalty Stacking”, 85 Texas Law Review 1991 (2007) Lévêque, François, “Innovation, Leveraging and Essential Facilities: Interoperability Licensing in the Microsoft Case”, 28 World Competition 71 (2005)

(Zi) Ehlermann 08 Biblio

740

6/11/08

11:57

Page 740

Bibliography

Lévêque, François, “Antitrust Enforcement in the Electricity and Gas Industries: Problems and Solutions for the EU”, 19 The Electricity Journal 27 (2006) Levi, Edward, “The Antitrust Laws and Monopoly”, 14 University of Chicago Law Review 153 (1947) Levine, Michael, “Airline Competition in Deregulated Markets: Theory, Firm Strategy and Public Policy”, 4 Yale Law Journal on Regulation 393 (1987) Lianos, Ioannis “Competition Law and Intellectual Property Rights: Is the Property Rights Approach Right?”, in John Bell and Claire Kilpatrick, eds., Cambridge Yearbook of European Legal Studies, Hart Publishing, 2006, pp. 153 et seq. Lichtman, Doug, “Patent Holdouts in the Standard-Setting Process”, University of Chicago, Law and Economics Olin Working Paper No. 292 (May 2006), http://papers. ssrn.com/sol3/papers.cfm?abstract_id=902646 London, Paul, The Competition Solution, American Enterprise Institute, 2005 Lowe, Philip, “DG Competition’s Review of the Policy on Abuse of Dominance” in Barry Hawk, ed., International Antitrust Law & Policy: Fordham Corporate Law 2003, Juris Publishing, 2004, chapter 10 Lowe, Philip, “Enforcement Authorities”, in Barry Hawk, ed., International Antitrust Law and Policy: Fordham Competition Law 2005, Juris Publishing, 2006, chapter 3 Lyons, Bruce, “The Paradox of the Exclusion of Exploitative Abuse”, Centre for Competition Policy Newsletter (Spring 2007) Maduro, Miguel Poiares, “Reforming the Market or the State? Article 30 and the European Constitution: Economic Freedom and Political Rights”, 3 European Law Journal 55 (1997) Maier-Rigaud, Frank, “Article 82 Rebates: Four Common Fallacies”, 2 European Competition Law Journal 85 (2006) Maier-Rigaud, Frank and Dovile Vaigauskaite, “Prokent/Tomra, a textbook case? Abuse of dominance under perfect information”, Competition Policy Newsletter No. 2 (2006)) Mankiw, Gregory and Michael Whinston, “Free Entry and Social Inefficiency”, 17(1) Rand Journal of Economics 48 (1986) Marasco, Amy, “Standards-Setting Practices: Competition, Innovation and Consumer Welfare”, testimony before the FTC and DOJ (18 April 2002), http://www.ftc. gov/opp/intellect/020418marasco.pdf Marquardt, Paul and Mark Leddy, “The Essential Facilities Doctrine and Intellectual Property Rights: A Response to Pitofsky, Patterson, and Hooks”, 70 Antitrust Law Journal 847 (2003) Marquis, Mel, “O2 (Germany) v Commission and the exotic mysteries of Article 81(1)”, 32 European Law Review 29 (2007) Marquis, Mel, “Tutela del consumatore e disciplina della concorrenza: in margine alla sentenza del Tribunale di prima istanza nel caso GlaxoSmithKline”, 12(2) Il Diritto dell’Unione Europea 369 (2007) Martínez Lage, Santiago and Rafael Allendesalazar, “Community Policy on Discriminatory Pricing: A Practitioner’s Perspective”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 325 et seq. Massoudi, Gerald, “Antitrust Enforcement and Standard Setting: The VITA and IEEE Letters and the ‘IP2’ Report”, speech delivered in Boston, 10 May 2007, http://www.justice.gov/atr/public/speeches/223363.htm

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 741

Bibliography

741

McDonald, J. Bruce, “The Struggle For Standards”, presentation at the Spring Meeting of the American Bar Association Section of Antitrust Law, Washington, D.C., 1 April 2004 McGee, John, “Predatory Price Cutting: The Standard Oil (N.J.) Case”, 1 Journal of Law and Economics 137 (1958) Melamed, Douglas, “Exclusionary Conduct Under the Antitrust Laws: Balancing, Sacrifice, and Refusals to Deal”, 20 Berkeley Technology Law Journal 1247 (2005) Melamed, Douglas, “Exclusive Dealing Agreements and Other Exclusionary Conduct—Are There Unifying Principles?, 73 Antitrust Law Journal 375 (2006) Mercier, Pierre, Olivier Mach, Hubert Gilliéron and Simon Affolter, Grands Principes du Droit de la Concurrence: Droit Communautaire: Droit Suisse, Dossiers de Droit Europeen No. 7, Helbing Lichtenhahn Verlag/Editions Bruylant, 1999 Mestmäcker, Ernst-Joachim, Das Marktbeherrschende Unternehmen im Recht der Wettbewerbsbeschränkungen, Tübingen, 1959 Mestmäcker, Ernst-Joachim, “Unternehmenszusammenschlüsse nach Artikel 86 des EWG-Vertrages“, in Ernst von Caemmerer, Hans-Jürgen Schlochauer and Ernst Steindorff, eds., Festschrift für Walter Hallstein, 1966, pp. 322 et seq. Mestmäcker, Ernst-Joachim, “Verpflichtet § 22 GWB die Kartellbehörde, marktbeherrschenden Unternehmen ein Verhalten aufzuerlegen, als ob Wettbewerb bestünde?”, DB, 1968, 1800 et seq. Mestmäcker, Ernst-Joachim, Der verwaltete Wettbewerb, Mohr Siebeck, 1984 Mestmäcker, Ernst-Joachim, “Bausteine zu einer Wirtschaftsverfassung—Franz Böhm in Jena”, in ders., Wirtschaft und Verfassung in der Europäischen Union, 2 ed., 2003 Mestmäcker, Ernst-Joachim, “Die Interdependenz von Recht und Ökonomie in der Wettbewerbspolitik”, in Monopolkommission, ed., Zukunftsperspektiven der Wettbewerbspolitik, Nomos, 2005 Mestmäcker, Ernst-Joachim, “Die Beurteilung von Unternehmenszusammenschlüssen nach Artikel 86 des EWG-Vertrags”, reprinted in Mestmäcker, Wirtschaft und Verfassung in der Europäischen Union, 2nd ed., Nomos, 2006, pp. 597 et seq. Mestmäcker, Ernst-Joachim, A legal theory without law: Posner v. Hayek on economic analysis of law, Mohr Siebeck, 2007 Miksch, Leonard, Wettbewerb als Aufgabe. Grundsätze einer Wettbewerbsordnung, Kohlhammer, 1937; 2nd ed., Küpper, 1947 Milgrim, Roger, “Commission Proposed Capital Punishment—By Definition—for Trade Secrets, A Uniquely Valuable IP Right”, 88 Journal of the Patent and Trademark Office Society, No. 11 (November 2006), at 919 et seq. Mirrlees, James, “An Exploration in the Theory of Optimum Income Taxation”, 38 Review of Economic Studies 175 (1971) Monti, Giorgio, “Article 81 EC and Public Policy”, 39 Common Market Law Review 1057 (2002) Monti, Giorgio, EC Competition Law, Cambridge University Press, 2007 Monti, Mario, “EU Competition Policy After May 2004”, SPEECH/03/489, delivered at Fordham Law School, New York, 24 October 2003 and published in Barry Hawk, ed., International Antitrust Law & Policy: Fordham Corporate Law 2003, Juris Publishing, 2004, chapter 16 Monti, Mario, “Introductory Exposé”, in Claus-Dieter Ehlermann and Isabela Atanasiu, European Competition Law Annual 2004: The Relationship Between Competition Law and the (Liberal) Professions, Hart Publishing, 2006, pp. 3 et seq.

(Zi) Ehlermann 08 Biblio

742

6/11/08

11:57

Page 742

Bibliography

Monti, Mario, “Intervening against Government restraints on Competition: Reflections from the EU expertise”, SPEECH/03/493, delivered in Washington, D.C., 27 October 2003 Möschel, Wernhard, “Competition Policy from an Ordo Point of View”, in Alan Peacock and Hans Wilgerodt, eds., German Neo-Liberals and the Social Market Economy, St. Martin’s Press, 1989, pp. 143 et seq. Motta, Massimo, Competition Policy: Theory and Practice, Cambridge University Press, 2004 Motta, Massimo and Alexandre de Streel, “Excessive Pricing and Price Squeeze under EU Law”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 91 et seq. Motta, Massimo and Alexandre de Streel, “Excessive Pricing in Competition Law: Never Say Never?”, in Konkurrensverket, The Pros and Cons of High Prices, Konkurrensverket (Swedish Competition Authority), 2007, Chapter 2 Myrdal, Gunnar, Objectivity in Social Research, Duckworth, 1970 Nalebuff, Barry, “Bundling”, Yale International Centre for Finance (ICF) Working Paper No. 99-14 (November 1999), http://papers.ssrn.com/sol3/papers.cfm?abstract_ id=185193 Nalebuff, Barry, “Bundling, Tying and Portfolio Effects: Part 1—Conceptual Issues”, DTI Economics Paper No.1 (February 2003) Nalebuff, Barry and David Majerus, “Bundling, Tying and Portfolio Effects: Part 2-Case Studies”, DTI Economics Paper (February 2003) Nalebuff, Barry, “Bundling as an Entry Deterrent”, 119 Quarterly Journal of Economics 159 (2004) Nalebuff, Barry, “Exclusionary Bundling”, 50 Antitrust Bulletin 321 (2005) Nalebuff, Barry, “Tied and True Exclusion: Comment on Jean Tirole’s ‘The Analysis of Tying Cases: A Primer’”, 1(1) Competition Policy International 41 (2005) Nebbia, Paolisa, “Unfair Contract Terms in European Law: a Study in Comparative and EC Law”, in Modern Studies in European Law, Volume 15, Hart Publishing, 2007, pp. 171 et seq. Nebbia, Paolisa, “So What Happened to Mr Manfredi? The Italian Decision Following the Ruling of the European Court of Justice”, 28 European Competition Law Review 591 (2007) Negrinotti, Matteo, “The AstraZeneca Case”, College of Europe Research Paper in Law 4/2007 Neven, Damien, “Competition economics and antitrust in Europe”, 48 Economic Policy 741 (2006) Neven, Damien and Lars-Hendrik Röller, “Consumer surplus vs. welfare standard in a political economy model of merger control”, 23 International Journal of Industrial Organization 829 (2005) Neven, Damien, Lars-Hendrk Röller and Zhentang Zhang, “Endogenous Costs and Price-Cost Margins: an Application to the European Airline Industry”, 54 Journal of Industrial Economics 351 (2006) Newbery, David, Privatization, Restructuring, and Regulation of Network Industries, MIT Press, 1999 Nietsche, Friedrich, Untimely Meditations: On the Uses and Disadvantages of History for Life (R.J. Hollingdale, trans.), Cambridge University Press, 1983.

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 743

Bibliography

743

O’Donoghue, Robert and Jorge Padilla, The Law and Economics of Article 82, Hart Publishing, 2006 O’Donoghue, Robert and David Parker, “The Final Piece In The Jigsaw: An Analysis Of The Draft European Commission Guidelines On Non-Horizontal Mergers”, cCCP Viewpoint (March 2007) Odudu, Ogheohene, Note on the British Airways judgment of the ECJ, 44 Common Market Law Review 1781 (2007) Ohana, Gil, Marc Hansen and Omar Shah, “Disclosure and Negotiation of Licensing Terms Prior to Adoption of Industry Standards: Preventing Another Patent Ambush”, 24 European Competition Law Review 644 (2003) Oliver, Peter, “The Concept of ‘Abuse’ of a Dominant Position under Article 82 EC: Recent Developments in Relation to Pricing”, 1(2) European Competition Journal 179 (2005) Ordover, Janusz and Robert Willig, “An Economic Definition of Predation: Pricing and Product Innovation”, 91 Yale Law Journal 8 (1981) Ordover, Janusz, “Dominant Position and Monopoly Power: A View from the Cathedral of Economics”, ABA Antitrust Section Spring Meeting (slides) (April 19, 2007) Pallotta, Oreste, “Consumatori e concorrenza: le questioni irrisolte nella causa Manfredi”, 12(2) Il Diritto dell’Unione Europea 305 (2007) Pate, Hewitt, “The Common Law Approach and Improving Standards for Analyzing Single Firm Conduct,” Fordham International Antitrust Conference, New York, 23 October 2003 Patterson, Mark R., “When Is Property Intellectual? The Leveraging Problem”, 73 Southern California Law Review 1133 (2000) Patterson, Mark R., “Inventions, Industry Standards, and Intellectual Property” 17 Berkeley Technology Law Journal 1043 (2002) Patterson, Mark R., “The Sacrifice of Profits in Non-Price Predation”, 18 Antitrust 37 (2003) Paulis, Emil, “The Burden of Proof in Article 82 Cases”, in Barry Hawk, ed., 2006 Fordham Corporate Law Institute, Juris Publishing, 2007 Peeperkorn, Luc and Vincent Verouden, “The Economics of Competition”, in Jonathan Faull and Ali Nikpay, The EC Law of Competition, 2nd ed., Oxford University Press, 2007, chapter 1 Pegg, Carl, The Evolution of the European Idea, 1914–1932, University of North Carolina Press, 1983 Peritz, Rudolph, Competition Policy in America. History, Rhetoric, Law, Oxford University Press, 1996 Petrasincu, Alex, “The European Commission’s New Guidelines on the Assessment of Non-Horizontal Mergers—Great Expectations Disappointed”, 29 European Competition Law Review 221 (2008) Pijnhacker Hordijk, Eric, “Excessive Pricing under EC Competition Law; An Update in the Light of ‘Dutch Developments’”, in Barry Hawk, ed., Fordham Corporate Law Institute, Juris Publishing, 2002, pp. 463 et seq. Pijnacker Hordijk, Erik and Yvo De Vries, “Onbillijk hoge prijzen als vorm van misbruik van een economische machtspositie onder het Europese en het Nederlandse mededingingsrecht”, 2002 SEW 430

(Zi) Ehlermann 08 Biblio

744

6/11/08

11:57

Page 744

Bibliography

Pitofsky, Robert, Donna Patterson and Jonathan Hooks, “The Essential Facilities Doctrine Under U.S. Antitrust Law”, 70 Antitrust Law Journal 443 (2002) Popofsky, Mark, “Defining Exclusionary Conduct: Section 2, the Rule of Reason, and the Unifying Principle Underlying Antitrust Rules”, 73 Antitrust Law Journal 435 (2006) Posner, Richard, “An Economic Approach To Legal Procedure And Judicial Administration” 2 Journal Of Legal Studies 399 (1973) Posner, Richard, “Theories of Economic Regulation”, 5 Bell Journal of Economics and Management Science 335 (1974) Posner, Richard, Antitrust Law: An Economic Perspective, University of Chicago Press, 1976 Posner, Richard, “The Chicago School of Antitrust Analysis”, 127 University of Pennsylvania Law Review 925 (1979) Posner, Richard, Antitrust Law, 2nd ed., University of Chicago Press, 2001 Posner, Richard, “Vertical Restraints and Antitrust Policy”, 72 University of Chicago Law Review 229 (2005) Putnam, Hilary, The Collapse of the Fact/value Dichotomy and Other Essays, Harvard University Press, 2002 Ramirez-Perez, Sigfrido, “Public policies, European integration and multinational corporations in the automobile sector: the French and Italian cases in a comparative perspective 1945–1973”, Ph.D dissertation, EUI, 2007 Ramsey, Frank, “A Contribution to the Theory of Taxation”, 37 Economic Journal 47 (1927) Ranci, Pippo, “Greatness and smallness of a European energy regulator” (2 October 2007), http://www.energypolicyblog.com/?p=69 Rasumussen, Eric, Mark Ramseyer and John Wiley, “Naked Exclusion”, in 81 American Economic Review 1137 (1991) Régibeau, Pierre and Katharine E. Rockett, “The Relationship between Intellectual Property Law and Competition Law: An Economic Approach”, in Steven Anderman, ed., Intellectual Property Rights and Competition Policy, Cambridge University Press, 2007 Rey, Patrick, Paul Seabright and Jean Tirole, “The Activities of a Monopoly Firm in Adjacent Competitive Markets: Economic Consequences and Implications for Competition Policy”, IDEI mimeo, 2001, http://idei.fr/doc/wp/2001/activities_abs.pdf Rey, Patrick and Jean Tirole, “A Primer on Foreclosure”, in Mark Armstrong and Rob Porter, eds., Handbook of Industrial Organization, Volume III, North Holland, 2007 Riley, Alan, “The EU Reform Treaty and the Competition Protocol: Undermining EC Competition Law”, 12 European Competition Law Review 703 Ritter, Cyril, “Does the Law of Predatory Pricing and Cross-Subsidisation Need a Radical Rethink?”, 27 World Competition 613 (2004) Ritter, Cyril, “Refusal to Deal and ‘Essential Facilities’: Does Intellectual Property Require Special Deference Compared to Tangible Property?”, 28 World Competition 281 (2005) Ritter, Lenaert, David Braun and Francis Rawlinson, EC Competition Law—A Practitioner’s Guide, 2nd ed., Kluwer, 2000 Robertson, Beverley, “What is a Restriction of Competition? The Implications of the CFI’s judgment in O2 Germany and the Rule of Reason”, 28 European Competition Law Review 252 (2007)

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 745

Bibliography

745

Robinson, Joan, The Economics of Imperfect Competition, 2nd revised ed., Palgrave MacMillan, 1933/1969 Robinson, Joan, “What is Perfect Competition?”, 49 Quarterly Journal of Economics 104 (1934) Rockett, Katharine, “The Quality of Licensed Technology”, 8 International Journal of Industrial Organization 559 (1990) Röller, Lars-Hendrik, “Economic Analysis and Competition Policy Enforcement in Europe”, in Peter A.G. van Bergeijk. and Erik Kloosterhuis, Modeling European Mergers: Theory, Competition Policy and Case Studies, Edward Elgar, 2005, pp. 11 et seq. Rosch, J. Thomas, “FTC Litigation at the Antitrust/Intellectual Property Interface”, speech delivered in Washington, D.C., 26 April 2007 Rosch, J. Thomas, “The Challenges of Non-Horizontal Merger Enforcement”, speech delivered at Fordham Law School, New York, 27 September 2007, www.ftc.gov/ speeches/rosch/070927-28non-horizontalmerger.pdf Rousseva, Ekaterina, “Modernizing by Eradicating: How the Commission’s New Approach to Article 81 EC Dispenses with the Need to Apply Article 82 EC to Vertical Restraints”, 42 Common Market Law Review 587 (2005) Rousseva, Ekaterina, The Application of Article 82 EC to Exclusionary Abuses: Evolution or Revolution, Ph.D dissertation, EUI, Florence, 2008 Rowe, Frederick, “The Decline of Antitrust and the Delusions of Models: The Faustian Pact of Law and Economics”, 72 Georgetown Law Journal 1511 (1984) Rubinfeld, Daniel, “Competition, Innovation, and Antitrust Enforcement in Dynamic Network Industries” Address Before the Software Publishers Association (1998 Spring Symposium) San Jose, California, 24 March 1998, http://www.usdoj. gov/atr/public/speeches/1611.htm Sally, Razeen, Classical Liberalism and International Economic Order: Studies in Theory and Intellectual History, Routledge, 1998 Salinger, Michael, “Moneyball and Price Gouging”, address to Antitrust Committee of Boston Bar Association, Boston, 27 February 2006 Salop, Steven, “Monopolistic Competition with Outside Goods”, 10 Bell Journal of Economics 141 (1979) Salop, Steven, “Exclusionary Conduct, Effect on Consumers, and the Flawed ProfitSacrifice Standard”, 73 Antitrust Law Journal 311 (2006) Salop, Steven, “The Controversy Over the Proper Antitrust Standard for Anticompetitive Exclusionary Conduct”, in Barry Hawk, ed., International Antitrust Law and Policy: Fordham Competition Law 2006, Juris Publishing, 2007, chapter 21 Salop, Steven and Craig Romaine, “Preserving Monopoly: Economic Analysis, Legal Standards, and Microsoft”, 7 George Mason Law Review 617 (1999) Scherer, F.M., “Predatory Pricing and the Sherman Act: A Comment”, 89 Harvard Law Review 869 (1976) Schmidt, Klaus, “Licensing Complementary Patents and Vertical Integration” (November 2006), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=944169 Schulze, Reiner and Thomas Hoeren, eds., Dokumente zum Europäischen Recht, Band 3: Kartellrecht (bis 1957), Springer, 2000 Schumpeter, Joseph, Capitalism, Socialism and Democracy, Harper & Row, 1942/ 1975

(Zi) Ehlermann 08 Biblio

746

6/11/08

11:57

Page 746

Bibliography

Schwalbe, Ulrich and Zimmer, Daniel, Kartellrecht und Ökonomie—Moderne ökonomische Ansätze in der Europäischen und Deutschen Zusammenschlusskontrolle, Verlag Recht und Wirtschaft, 2006 Schwarz, Hans-Gunter and Cristoph Lang, “The Rise in German Wholesale Electricity Prices: Fundamental Factors, Exercise of Market Power, or Both?”, IWE Working Paper No. 02/2006, http://www.economics.phil.uni-erlangen.de/get. php?page=forschung/workingpapers/workingpapers Schweitzer, Heike, “Competition Law and Public Policy: Reconsidering an Uneasy Relationship. The Example of Art. 81”, EUI Working Paper LAW 2007/30, http://cadmus.iue.it/dspace/handle/1814/7623 Schweitzer, Heike, “Controlling the Unilateral Exercise of Intellectual Property Rights: A Multitude of Approaches but No Way Ahead? The Transatlantic Search for a New Approach”, EUI Working Paper LAW No. 2007/31 Scotchmer, Susanne, “Licensing, Joint Ventures, and Competition Policy,” in Susanne Scotchmer, Innovation and Incentives, MIT Press, 2004 Segal, Ilya and Michael Whinston, “Exclusive Contracts and Protection of Investment”, 31 Rand Journal of Economics 603 (2000) Segal, Ilya and Michael Whinston, “Naked Exclusion: Comment”, 90 American Economic Review 296 (2000) Sen, Amartya, Rationality and Freedom, Harvard University Press, 2003 Shakespeare, William, Romeo and Juliet, First Folio edition, 1623, http://historyofideas. org/toc/modeng/public/ShaRJF.html Shapiro, Carl, “Aftermarkets and Consumer Welfare: Making Sense of Kodak”, 63 Antitrust Law Journal 483 (1995) Shapiro, Carl, “Navigating the Patent Thicket: Cross-Licenses, Patent Pools and Standard-Setting”, 2001, http://faculty.haas.berkeley.edu/shapiro/thicket.pdf Shapiro, Carl, “Setting Compatibility Standards: Cooperation or Collusion?” in Rochelle Dreyfuss, Diane Zimmerman and Harry First, eds., Expanding The Bounds Of Intellectual Property, Oxford University Press, 2001 Sidak, Gregory, “Holdup, Royalty Stacking, and the Presumption of Injunctive Relief for Patent Infringement: A Reply to Lemley and Shapiro”, 92 Minnesota Law Review 714 (2008) Sinclair, Duncan, “Abuse of Dominance at a Crossroads—Potential Effect, Object and Appreciability under Article 82”, 25 European Competition Law Review 491 (2004) Siragusa, Mario, “Application of Article 86: tying arrangements, refusals to deal, discrimination and other cases of abuse”, in Regulating the behaviour of monopolies and dominant undertakings in Community law, 1977 Bruges Week, Cahiers de Bruges, N.S. 36, De Tempel, Bruges (1977) Skitol, Robert, “Concerted Buying Power: Its Potential for Addressing the Patent Holdup Problem in Standard Setting, 72 Antitrust Law Journal, 727 (2005) Slade, Margaret, “The Leverage Theory of Tying Revisited: Evidence from Newspaper Advertising”, 65 Southern Economic Journal 204 (1998) Smith, Adam, The Wealth of Nations, 1776 Spector, David, “Loyalty Rebates and Related Pricing Practices: When Should Competition Authorities Worry”, World Competition, forthcoming Stigler, George, “United States v. Loew’s, Inc.: A Note on Block Booking”, Supreme Court Review 152 (1963)

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 747

Bibliography

747

Stratakis, Alexandros, “Comparative Analysis of the US and EU Approach and Enforcement of the Essential Facilities Doctrine”, 27 European Competition Law Review 434 (2006) Streit, Manfred, “Economic Order, Private Law and Public Policy: The Freiburg School of Law and Economics in Perspective”, in 148 Journal of Institutional and Theoretical Economics 675 (1992) Streit, Manfred, “Cognition, Competition and Catallaxy”, 4(2) Constitutional Political Economy 223 (1993) Streit, Manfred and Michael Wohlgemuth, “The Market Economy and the State”, Diskussionsbeitrag 06-97, paper prepared for the Fifth Annual SEEP-Conference on Economics, Ethics and Philosophy, Marienrode, 29 October–November 1997 Sullivan, Lawrence and Warren Grimes, The Law of Antitrust: An Integrated Handbook, 2nd ed., Thomson West, 2006 Svetlicinii, Alexandr, “Draft Commission Guidelines on the Assessment of NonHorizontal Mergers: Do They ‘Defend’ the Efficiency Defence?”, 30 World Competition 403 (2007) Svetlicinii, Alexandr, “Assessment of the Non-Horizontal Mergers: Is There a Chance for the Efficiency Defence in EC Merger Control?”, 28 European Competition Law Review 529 (2007) Swaine, Edward, “‘Competition, Not Competitors,’ Nor Canards: Ways of Criticizing the Commission”, 23 University of Pennsylvania Journal of International Economic Law 597 (2002) Swanson, Daniel and William Baumol, “Reasonable and Nondiscriminatory (RAND) Royalties, Standards Selection, and Control of Market Power”, 73 Antitrust Law Journal 1 (2005) Tandon, Pankaj, “Optimal Patents with Compulsory Licensing”, 90 Journal of Political Economy 470 (1982) Teece, David and Edward Sherry, “Standards Setting and Antitrust”, 87 Minnesota Law Review 1913 (2003) Telser, Lester, “A Theory of Monopoly of Complementary Goods,” 52 Journal of Business 211 (1979) Temple Lang, John, “Abuse of Dominant Positions in European Community Law, Present and Future: Some Aspects”, in Barry Hawk, ed., Fifth Fordham Corporate Law Institute, Law & Business, 1979, pp. 52 et seq. Temple Lang, John, “Legal Certainty and Legitimate Expectations as General Principles of Law”, in Ulf Bernitz and Joakim Nergelius, eds., General Principles of European Community Law, Kluwer Law International, 2000, pp. 163 et seq. Temple Lang, John, “Anticompetitive Non-Pricing Abuses Under European and National Antitrust Law”, in Barry Hawk, ed., 2003 Fordham Corporate Law Institute, Juris Publication Inc., 2004, pp. 235 et seq. Temple Lang, John, “The Application of the Essential Facility Doctrine to Intellectual Property Rights under European Competition Law”, conference on Antitrust, Patent and Copyright, Ecole des Mines, Paris, 15–16 January 2004, http://www.cerna.ensmp.fr/cerna_regulation/Documents/?ColloqueAntitrust2004/ ?TempleLang.pd Temple Lang, John and Robert O’Donoghue, “Defining Legitimate Competition: How to Clarify Pricing Abuses Under Article 82 EC”, 26 Fordham International Law Journal 84 (2002)

(Zi) Ehlermann 08 Biblio

748

6/11/08

11:57

Page 748

Bibliography

Temple Lang, John and Robert O’Donoghue, “The Concept Of An Exclusionary Abuse”, GCLC Research Papers on Article 82 EC, College of Europe (July 2005), pp. 38 et seq. http://www.coleurop.be/content/gclc/documents/GCLC% 20Research %20Papers%20on%20Article%2082%20EC.pdf Thorelli, Hans, The Federal Antitrust Policy: Origination of an American Tradition, Johns Hopkins Press, 1954 Tirole, Jean, “The Analysis of Tying Cases: A Primer”, 1(1) Competition Policy International 1 (2005) Trebilcock, Michael, Ralph Winter, Paul Collins and Edward Iacobucci, The Law and Economics of Canadian Competition Policy, University of Toronto Press, 2002 Tridimas, Takis, The General Principles of EC Law, Oxford University Press, 1999, second edition, OUP, 2006 Turner, Donald, “Antitrust Policy and the Cellophane Case”, 70 Harvard Law Review 281 (1956) Turner, Frederick Jackson, “The Significance of the Frontier in American History”, paper delivered in Chicago, 12 July 1893, http://xroads.virginia.edu/~HYPER/ TURNER/chapter1.html Van Bael & Bellis, Competition Law of the European Community, 4th edition, Kluwer, 2005 Vanberg, Viktor, “The Freiburg School: Walter Eucken and Ordoliberalism”, Freiburg Discussion Paper 04/11, http://ies.fsv.cuni.cz/storage/sylab/138_vanberg2004.pdf Venit, James, “Article 82: The Last Frontier—Fighting Wire with Fire?”, 28 Fordham International Law Journal 1157 (2005) Venit, James, “Article 82 EC: Exceptional Circumstances: The IP/Antitrust Interface after IMS Health”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2005: The Interaction between Competition Law and Intellectual Property Law, Hart Publishing, 2007, pp. 609 et seq. Vickers, John, “Abuse of Market Power”, 115 The Economic Journal 244 (2005) Vickers, John, “How Does the Prohibition of Abuse of Dominance Fit with the Rest of Competition Policy?”,in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2003: What Is an Abuse of a Dominant Position?, Hart Publishing, 2006, pp. 147 et seq. Vickers, John, “Market Power in Competition Cases”, 2 European Competition Journal 3, special issue (2006) Viscusi, Kip, Joseph Harrington and John Vernon, Economics of Regulation and Antitrust, 4th ed., MIT Press, 2005 Von der Groeben, Hans, Deutschland und Europa in einem unruhigen Jahrhundert, Nomos, 1995 Waelbroeck, Denis, “Michelin II: A Per Se Rule Against Rebates by Dominant Companies?” 1 Journal of Competition Law and Economics 149 (2005) Waelbroeck, Michel and Aldo Frignani, European Competition Law (trans. Françoise Gamet-Pol), Transnational Publishers Inc., 1999 Weber, Max, The Methodology of the Social Sciences, Free Press, 1949 Werden, Gregory, “Identifying exclusionary conduct under Section 2: The ‘No Economic Sense’ Test”, 73 Antitrust Law Journal 413 (2006) Werden, Gregory, “The ‘No Economic Sense’ Test for Exclusionary Conduct”, 31 Journal of Corporation Law 293 (2006)

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 749

Bibliography

749

Werden, Gregory, “Identifying Single-Firm Exclusionary Conduct: From Vague Concepts to Administrable Rules”, Barry Hawk, ed., International Antitrust Law & Policy: Fordham Competition Law 2006, Juris Publishing, 2007, chapter 22 Whinston, Michael, “Tying, Foreclosure, and Exclusion”, 80 American Economic Review 837 (1990) Whinston, Michael, “Exclusivity and Tying in U.S. v. Microsoft: What We Know and What We Don’t Know”, 15 Journal of Economic Perspectives 63 (2001) Whish, Richard, Competition Law, 5th ed., Butterworths, 2003 Whish, Richard, “Oligopoly Theory and Economic Evidence”, in Claus-Dieter Ehlermann and Isabela Atanasiu, eds., European Competition Law Annual 2006: Enforcement of Prohibition of Cartels, Hart Publishing, 2007, pp. 105 et seq. Williamson, Oliver, “Assessing Vertical Market Restrictions: Antitrust Ramifications of the Transaction Cost Approach,” 127 University of Pennsylvania Law Review 953 (1979) Willig, Robert, “(Allegedly) Monopolizing Tying via Product Innovation”, oral remarks at the FTC/DOJ’s Sherman Act Section 2 Joint Hearing on Understanding Single-Firm Behavior: Tying Session, 1 November 2006 Wils, Wouter, “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 (2004) Wisking, Stephen, Craig Pouncey and Jacques Buhart, “European Union: Competition—Refusal to Supply”, 30 European Intellectual Property Review N1 (2008) Wurmnest, Wolfgang, “The Reform of Article 82 EC in the Light of the ‘Economic Approach’ ”, in Josef Drexl, Beatriz Conde Gallego, Stefan Enchelmaier and MarkOliver Mackenrodt, eds., Art. 82 EC: New Interpretation, New Enforcement Mechanisms?, Springer, forthcoming Zajac, E.E., “Note on ‘Gold plating’ or ‘rate base padding’”, 3 Bell Journal of Economics and Management Science 311 (1972) Ziller, Jacques, The European Constitution (Mel Marquis, trans.), Kluwer Law International, 2005

II. Legislation, proposed legislation and soft law European legislation Directive 91/440/EEC of 29 July 1991 on the development of the Community’s railways, [1991] OJ L237/25 Commission Regulation 2790/99/EC of 29 December 1999 on the application of Article 81(3) of the Treaty to categories of vertical agreements and concerted practices, [1999] OJ L336/21 Directive 2002/19/EC of the European Parliament and of the Council of 7 March 2002 on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive), [2002] OJ L108/7

(Zi) Ehlermann 08 Biblio

750

6/11/08

11:57

Page 750

Bibliography

Directive 2002/20/EC of the European Parliament and of the Council of 7 March 2002 on the authorisation of electronic communications networks and services (Authorisation Directive), [2002] OJ L108/21 Directive 2002/21/EC of the European Parliament and of the Council of 7 March 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive), [2002] L108/33 Directive 2002/22/EC of the European Parliament and of the Council of 7 March 2002 on universal service and users’ rights relating to electronic communications networks and services (Universal Service Directive), [2002] L108/51 Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector, [2002] OJ L201/37 Directive 2002/77/EC of 16 September 2002 on competition in the markets for electronic communications networks and services, [2002] L249/21 Directive 2003/87/EC of the European Parliament and of the Council of 13 October 2003 establishing a scheme for greenhouse gas emission allowance trading within the Community and amending Council Directive 96/61/EC, [2003] OJ L275/32, as amended by Directive 2004/101/EC of the European Parliament and of the Council of 27 October 2004, [2004] OJ L338/18 Commission Regulation 772/2004 on the application of Article 81(3) of the Treaty to categories of technology transfer agreements, [2004] OJ L123/11 Regulation (EC) No 717/2007 of the European Parliament and of the Council of 27 June 2007 on roaming on public mobile telephone networks within the Community and amending Directive 2002/21/EC, [2007] OJ L171/32

European proposed legislation Proposal for a Regulation of the European Parliament and of the Council establishing an Agency for the Cooperation of Energy Regulators, COM(2007) 530 of 19 September 2007, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri= CELEX:52007PC0530:EN: NOT

European soft law Guidelines Commission Notice—Guidelines on Vertical Restraints, [2000] OJ C291/1 Commission’s Guidelines on Horizontal Cooperation Agreements, [2001] OJ C3/2 Commission guidelines on market analysis and the assessment of significant market power under the Community regulatory framework for electronic communications networks and services (“SMP Guidelines”), [2002] OJ C165/6 Commission, Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, [2004] OJ C31/03 Commission Guidelines on the Application of Article 81(3) of the Treaty, [2004] OJ C101/97 Commission, Guidelines on the application of Article 81 of the EC Treaty to Technology Transfer Agreements, [2004] OJ C101/2 Commission, Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 751

Bibliography

751

(28 November 2007), available at http://ec.europa.eu/comm/competition/ mergers/legislation/notices_on_substance.html Draft guidelines Commission, Draft Notice on the Appraisal of Horizontal Mergers under the Council Regulation on the Control of Concentrations Between Undertakings (11 December 2002), available at ec.europa.eu/comm/competition/mergers/review/final_draft_ en.pdf Commission, draft Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings (13 February 2007), http://ec.europa.eu/comm/competition/mergers/legislation/ non_horizontal_consultation.html Recommendation Commission Recommendation on Relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC, [2003] OJ L114/45 Other jurisdictions Canadian Competition Bureau, Enforcement Guidelines on the Abuse of Dominance Provisions, http://www.competitionbureau.gc.ca/epic/site/cb-bc.nsf/en/01760e.html U.S. Department of Justice and Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property (April 1995), http://www.usdoj.gov/atr/public/ guidelines/0558.htm U.S. Department of Justice and Federal Trade Commission, Horizontal Merger Guidelines (issued April 1992; revised, April 1997), http://www.usdoj.gov/atr/ public/guidelines/hmg.htm

III. Official and unofficial reports, studies, discussion papers, etc. Antitrust Modernization Commission (US) Antitrust Modernization Commission, Report and Recommendations (April 2007), http://www.amc.gov/report_recommendation/amc_final_report.pdf Commission, Annual Reports on Competition Policy Commission, Vth Annual Report on Competition Policy (1975) Commission, Fifteenth Annual Report on Competition Policy (1986) Commission, XXIVth Report on Competition Policy (1994) Commission, XXVIIth Report on Competition Policy (1997) Commission, Communications Commission, Communication—Towards a new framework for Electronic Communications infrastructure and associated services (The 1999 Communications Review), COM(1999) 539, http://ec.europa.eu/archives/ISPO/infosoc/telecompolicy/review99/review99en.pdf

(Zi) Ehlermann 08 Biblio

752

6/11/08

11:57

Page 752

Bibliography

Commission, Implementation of the Regulatory Framework in the Member States— 10th Report of 2 December 2004, http://eur-lex.europa.eu/smartapi/cgi/sga_doc? smartapi!celexplus!prod!DocNumber&lg=en&type_doc=COMfinal&an_doc= 2004&nu_doc=759 Commission, Communication on Market Reviews under the EU Regulatory Framework: Consolidating the internal market for electronic communications, COM(2006) 28 final of 6 February 2006, http://ec.europa.eu/information_society/policy/ ecomm/doc/implementation_enforcement/article_7/com_2006_0028_en.pdf Commission, Implementation of the Regulatory Framework in the Member States 2005—11th Report of 20 February 2006, http://ec.europa.eu/information_society/ policy/ecomm/doc/implementation_enforcement/annualreports/11threport/com_ 2006_68_en_final.pdf Commission Communication on the Review of the EU Regulatory Framework for electronic communications networks and services (COM(2006)334) of 29 June 2006, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52006DC0334: EN:NOT Commission, Communication—Inquiry pursuant to Article 17 of Regulation (EC) No 1/2003 into the European gas and electricity sectors (Final Report) of 10 January 2007, http://eur-lex.europa.eu/LexUriServ/site/en/com/2006/com2006_ 0851en01.pdf Commission Communication on European electronic communications regulation and markets 2006 (12th report), COM(2007) 155 of 29 March 2007, http://ec. europa.eu/information_society/policy/ecomm/implementation_enforcement/ annualreports/12threport/index_en.htm Commission, Discussion Paper on Exclusionary Abuses DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses (December 2005), http://ec.europa.eu/comm/competition/ antitrust/art82/discpaper2005.pdf Commission, private enforcement Commission, Green Paper on Damages Actions for breach of the EC antitrust rules, COM(2005)672 of 19 December 2005, http://eur-lex.europa.eu/LexUriServ/ LexUriServ.do?uri=CELEX:52005DC0672:EN:NOT Commission, White Paper on damages actions for breach of the EC antitrust rules, COM(2008)165 final of 2 April 2008, http://ec.europa.eu/comm/competition/ antitrust/actionsdamages/files_white_paper/whitepaper_en.pdf Commission, Staff working paper accompanying the White Paper on damages actions, SEC(2008) 404 of 2 April 2008 http://ec.europa.eu/comm/competition/antitrust/actionsdamages/files_white_paper/w orking_paper.pdf Commission, Impact Assessment Report accompanying the White Paper on damages actions, SEC(2008) 405 of 2 April 2008 http://ec.europa.eu/comm/competition/antitrust/actionsdamages/files_white_paper/i mpact_report.pdf

(Zi) Ehlermann 08 Biblio

6/11/08

11:57

Page 753

Bibliography

753

Economic Advisory Group for Competition Policy (EAGCP) Gual, Hellwig, Perrot, Polo, Rey, Schmidt and Stenbacka, Economic Advisory Group for Competition Policy (EAGCP), “An economic approach to Article 82” (July 2005), http://ec.europa.eu/comm/competition/publications/studies/eagcp_july_21_05.pdf European Competition Network ECN Working Group on Cooperation Issues, “Results of the questionnaire on the reform of Member States (MS) national competition laws after EC Regulation No. 1/2003”, http://ec.europa.eu/comm/competition/ecn/ecn_ convergencequest_18052007.pdf International Competition Network ICN, Unilateral Conduct Working Group, “Report on the Objectives of Unilateral Conduct Laws, Assessment of Dominance/Substantial Market Power, and StateCreated Monopolies” (May 2007), http://www.internationalcompetitionnetwork. org/index.php/en/library/working-group/18 Organisation for Economic Co-operation and Development OECD, “Canada—Report on Competition Law and Institutions (2004)”, DAF/COMP(2005)4 (January 2005), http://www.olis.oecd.org/olis/2005doc.nsf/ ENGDATCORPLOOK/NT000009C2/$FILE/JT00177068.PDF OECD, “Competition on the Merits”, DAF/COMP(2005)27 (updated 30 March 2006), http://www.oecd.org/dataoecd/7/13/35911017.pdf OECD, “Barriers to Entry”, DAF/COMP(2005)42 (6 March 2006), www.oecd.org/ dataoecd/43/49/36344429.pdf Office of Fair Trading (UK) OFT, “Competition Prioritisation Framework” (October 2006), http://www.oft.gov. uk/shared_oft/press_release_attachments/compcriteria.pdf Simon Bishop and Simon Baker, “The role of market definition in monopoly and dominance inquiries”, Economic Discussion Paper 2, prepared for the Office of Fair Trading (July 2001), http://www.oft.gov.uk/shared_oft/reports/comp_policy/ oft342.pdf RBB Economics, “Selective Price Cuts and Fidelity Rebates”, Economic Discussion paper prepared for the Office of Fair Trading (July 2005), http://www.oft.gov.uk/ shared_oft/reports/comp_policy/oft804.pdf The Spaak Report Intergovernmental Committee of the Messina Conference, Report by the Heads of Delegations to the Foreign Ministers, 21 April 1956 (in French), http://aei.pitt.edu/ 996/01/Spaak_report_french.pdf U.S. Agencies FTC/DOJ, Report on Promoting Innovation and Competition, http://www.ftc.gov/ reports/innovation/P040101PromotingInnovationandCompetitionrpt0704.pdf FTC/DOJ Commentary on the US Merger Guidelines, http://www.ftc.gov/os/ 2006/03/CommentaryontheHorizontalMergerGuidelinesMarch2006.pdf