The Metaphysics of Market Power: The Zero-Sum Competition and Market Manipulation Approach 9781509928071, 9781509928101, 9781509928095

Citation preview

THE METAPHYSICS OF MARKET POWER Australian competition law has just emerged from a significant period of reform which has seen controversial changes to the legal test to distinguish between normal competitive conduct and conduct that should be condemned. The ­controversy continues, arguably because the traditional legal conception of market power does not provide a useful standard in real world markets. This important new book offers a radical interpretation of market power, based on the power to manipulate. Seeing it in this way allows for positive and normative standards within which to frame a legal theory of liability for misuse of that power. The book provides suggestions to improve the forensic assessment of conduct that should be condemned as misuse of market power. Volume 22 in the series Hart Studies in Competition Law

Hart Studies in Competition Law Sanctions in EU Competition Law: Principles and Practice Michael Frese Fairness in Antitrust: Protecting the Strong from the Weak Adi Ayal European Merger Remedies: Law and Policy Dorte Hoeg Media Ownership and Control: Law, Economics and Policy in an Indian and International Context Suzanne Rab and Alison Sprague The Interface between Competition and the Internal Market: Market Separation under Article 102 TFEU Vasiliki Brisimi Anti-Cartel Enforcement in a Contemporary Age: Leniency Religion Edited by Caron Beaton-Wells and Christopher Tran Public Procurement and the EU Competition Rules Albert Sánchez Graells The Concept of Abuse in EU Competition Law: Law and Economic Approaches Pınar Akman The Competitive Effects of Minority Shareholdings: Legal and Economic Issues Panagiotis Fotis and Nikolaos Zevgolis The More Economic Approach to EU Antitrust Law Anne C Witt Private Power, Online Information Flows and EU Law Angela Daly The Role of Competitors in the Enforcement of State Aid Law Fernando Pastor-Merchante The Legality of Bailouts and Buy Nationals: International Trade Law in a Crisis Kamala Dawar A Critical Account of Article 106(2) TFEU: Government Failure in Public Service Provision Jarleth Burke Dawn Raids Under Challenge Helene Andersson A Framework for European Competition Law: Co-ordinated Diversity Christopher Townley Evidence Standards in EU Competition Enforcement: The EU Approach Andriani Kalintiri

The Metaphysics of Market Power The Zero-Sum Competition and Market Manipulation Approach

George Raitt

HART PUBLISHING Bloomsbury Publishing Plc Kemp House, Chawley Park, Cumnor Hill, Oxford, OX2 9PH, UK HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2019 Copyright © George Raitt, 2019 George Raitt has asserted his right under the Copyright, Designs and Patents Act 1988 to be identified as Author of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. While every care has been taken to ensure the accuracy of this work, no responsibility for loss or damage occasioned to any person acting or refraining from action as a result of any statement in it can be accepted by the authors, editors or publishers. All UK Government legislation and other public sector information used in the work is Crown Copyright ©. All House of Lords and House of Commons information used in the work is Parliamentary Copyright ©. This information is reused under the terms of the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/ open-government-licence/version/3) except where otherwise stated. All Eur-lex material used in the work is © European Union, http://eur-lex.europa.eu/, 1998–2019. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication data Names: Raitt, George, author. Title: The metaphysics of market power : the zero-sum competition and market manipulation approach / George Raitt. Description: Oxford, UK ; Chicago, Illinois : Hart Publishing, 2019.  |  Series: Hart studies in competition law ; volume 22  |  Includes bibliographical references and index. Identifiers: LCCN 2018057167 (print)  |  LCCN 2018057923 (ebook)  |  ISBN 9781509928088 (EPub)  |  ISBN 9781509928071 (hardback : alk. paper) Subjects: LCSH: Antitrust law—Australia.  |  Industrial concentration. Classification: LCC KU979 (ebook)  |  LCC KU979 .R35 2019 (print)  |  DDC 343.9407/21—dc23 LC record available at https://lccn.loc.gov/2018057167 ISBN: HB: 978-1-50992-807-1 ePDF: 978-1-50992-809-5 ePub: 978-1-50992-808-8 Typeset by Compuscript Ltd, Shannon To find out more about our authors and books visit www.hartpublishing.co.uk. Here you will find extracts, author information, details of forthcoming events and the option to sign up for our newsletters.

FOREWORD THE HON ROBERT FRENCH AC This thoughtful and forward-looking book is about competition law and policy in relation to misuse of market power. It begins with the proposition that the law in Australia is informed by a Structure-Conduct-Performance approach which has been overtaken by economic theory. The author, Dr George Raitt, argues that current legal criteria of liability for misuse of market power do not meet legal determinacy requirements and rule of law concerns. The judicial application of the central provision relating to misuse of market power, s 46 of the Competition and Consumer Act 2010 (Cth) is said to have relied upon ‘commercial intuition’ with insufficient legal and economic reasoning to provide certainty and predictability for business. The recent amendment to the section has not delivered predictability and certainty in the law. The Final Report of the Competition Policy Review, released on 31 March 2015 (the Harper Review) included a recommendation that s 46 include mandatory factors to be considered by a court in determining whether conduct covered by the section had the prohibited purpose, effect or likely effect. One of those factors was: The extent to which the conduct has the purpose, or would have or be likely to have the effect, of increasing competition in the market including by enhancing efficiency, innovation, product quality or price competitiveness in the market;

In the event, s 46 as amended in 2017 did not include that requirement. The Senate Economics Legislation Committee which reported on the Competition and Consumer Amendment (Misuse of Market Power) Bill in February 2017 received a submission from the Australian Competition and Consumer Commission that the subsection ‘creates scope for judicial interpretation about the interaction between efficiencies and competition and innovation and competition.’ The Committee recommended that it be removed from the Bill. By way of contrast, the introduction of an authorisation process for conduct falling within s 46 and the application of a public benefit test in that connection allowed the ACCC to have regard to efficiencies. Dr Raitt contends that the legal concept of market power as it now stands is insufficiently determined and that efficiency would be an appropriate matter for consideration in the exercise of the judicial function. He draws, by way of analogy, on the concept of efficient markets in securities industry laws. He argues that the legal concept of market power and standards to determine its use or misuse should

vi  Foreword be based on a model of competition which is more reflective of the real world than that presently embodied in the legislation. On that model, market power is conceived as the power of the market which cannot be possessed but can be manipulated. Profit maximising conduct is justified not because it is rational or logical but because it may be efficient. Different kinds of efficiency need to be identified and forensic tests developed. A legal theory of market power based on this approach requires the adaptation of economic principles. The market manipulation approach is concerned with determining whether market forces are operating without being distorted by manipulative use of market power. The book directs attention to the argument that consideration of economic efficiency is not a matter upon which courts are competent to embark, having regard to the skills and experience of judges and the constitutional function of courts as determining disputed matters of fact and law. It is arguable that the weighing of anti-competitive effects against a range of public benefits in the context of authorisation, now available under the law through the ACCC, falls outside the notion of judicial power. It may be noted that ‘public benefit’ as a criterion is not unknown to the common law. In any event, it does not seem likely that the concept of judicial power would exclude the consideration of issues of efficiency to the extent that they are relevant to the characterisation of the purpose or effect or likely effect of conduct. The scheme of the book in Part I includes discussion of four concepts of market power deficiency. The concept of market manipulation in securities is drawn upon in support of what the author calls ‘an alternative legal model’. Changing approaches in Australia, the United States and the European Union in relation to reliance upon market structure and changes in market structure to infer competitive effects are considered. The book reviews and critiques market structure and direct approaches to assessing the impact of conduct on the process of competition. These approaches are contrasted with the market manipulation approach to affects. In Part II the author analyses the application of the most recent version of s 46 to Australian appellate decisions relating to refusal to deal, margin squeezes, predatory pricing, meeting competition, raising rivals’ costs, and bundling. He also looks to institutional limitations on the power of the courts to consider efficiencies. The final Part of the book argues the author’s conclusion that s 46 as it now stands may be applicable to large firms simply because they are large firms. It is contended that there is a risk that they will be exposed to liability for conduct which is efficient in the sense that it does not distort the operation of markets. The principal proposition for which the author contends is that a market manipulation approach is to be preferred and that efficiency can be adapted to become a centrepiece of a positive and normative legal theory of market power. No doubt the author’s arguments are contestable and will be contested. They do, however, provide the kind of high level reflection that is necessary for the informed development of competition law and policy in Australia.

PREFACE Australian competition legislation was amended in late 2017 following a ­controversial three-year law reform process. This monograph has been in preparation before, during and since that process was completed. Review of the Australian and overseas literature indicates enduring controversy about market power in Australia and overseas. The author suggests this arises because our legal concept of market power is insufficiently legally determined. The law and policy has been internationally criticised for lacking clear positive and normative standards that would enable us to distinguish between normal competitive conduct and conduct that should be condemned. The monograph covers mainly Australian and EU law, but also compares key points of US and Canadian law on market power. The issue of market power addressed in this monograph is of universal significance for competition laws around the world. The traditional legal concept of market power as the ability to give less and charge more (or the ability to behave differently than a hypothetical competitive market would enforce) is based on a paradigm of the hypothetical competi­ tive market in which no firm has market power. The author suggests that such a benchmark does not provide a useful standard of conduct in real-world markets. This monograph proposes a new approach to both our conception of market power and the paradigm of competition in which we assess the misuse of market power. The author proposes that market power can alternatively be related to the response of market demand to firms’ price/output decisions, and thus can be legally conceived as the power of the market, which cannot be possessed, but can be manipulated. All firms have this power to some degree. The author derives this concept of market power from the correlation between market power and elasticity of demand revealed by well-known analysis of the Lerner Index, which is used to measure market power. This approach is also supported by analogy with laws prohibiting the manipulation of securities markets, which is theorised without the overlay of conflicting welfare objectives and economic efficiency standards that has troubled competition law over a long period. The author critically examines the assumptions underlying the paradigm of the hypothetical competitive market and suggests that we can alternatively identify a paradigm of competition, more closely resembling real-world competition, in which to assess misuse of market power. This is derived from the conventional description of real-world competition as rivalry for the same object. The author calls this paradigm ‘zero-sum’, or survival, competition in which one firm’s gain is another firm’s loss, where firms cannot freely exit the market and so must compete for survival.

viii  Preface When we consider market power as power to manipulate the market, we can distinguish firms’ price/output decisions and other decisions which respond to market demand from conduct which manipulates the market, distorting the market’s efficient operation. It is proposed that by reconceptualising market power in the context of zero-sum competition and by developing the legal concept of an efficient market from the law prohibiting manipulation of securities markets, we can develop positive and normative standards to frame our legal theory of liability for misuse of market power. The recent Harper Review in Australia challenged conventional thinking that economic efficiency is not a matter that the courts are competent to assess. The Review recommended that the statutory provision be amended to require courts to consider the efficiency effects of conduct when applying the test to assess whether alleged misuse of market power causes harm. The current legislative reform in Australia is based on the traditional legal conception of market power and, despite the Review’s recommendation, rejects explicit consideration of economic efficiency, so is unlikely to quell the controversy. The author suggests that conventional concepts of economic efficiency cannot be forensically applied and so must be adapted and redefined if we are to apply them in our legal theory of liability. He also suggests that the concept of market manipulation, which distorts the efficient operation of the market, enables us to do this. The market manipulation theory is tested on the facts of leading Australian cases under past law, using mainly EU, but also US and Canadian law, as a context in order to show what difference the new and proposed laws would make. The monograph compares the application of the market manipulation theory with the Harper Review’s recommended approach, which would consider economic efficiency, and the new law as enacted by Parliament, which prohibits conduct of a dominant firm that has the purpose or likely effect of substantially lessening competition.

ACKNOWLEDGEMENTS My title is taken from remarks by Heerey J in Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (1999) 169 ALR 554, 562. His Honour likened the task of determining a firm’s purpose, when self-interested conduct at one and the same time harms rivals, to a metaphysical exercise. That observation sparked my continuing professional and academic interest in the problem of market power. I undertook my undergraduate studies in economics and law in the 1970s, when the ink was barely dry on the Trade Practice Act 1974 (Cth), now the Competition and Consumer Act 2010 (CCA). I was admitted to legal practice in the State of Victoria in 1980 and commenced practice as a principal in Blake Dawson Waldron (now Ashurst) in 1988. I have practised law at my current firm, Piper Alderman, Melbourne, since 2002. The process of developing competition law in Australia by judicial interpretation and legislative reform has been controversial since its inception. I suggest that we have a way to go before our ideas about the legal prohibition of misuse of market power can be regarded as settled. I hope this monograph will make a contribution to that process. I would like to record my gratitude to colleagues and students at the Monash University Faculty of Law, and Deakin University Law School, where I have lectured in various law subjects since 2013. I have lectured in competition law to undergraduate and graduate students at Monash over the past few years. I want to particularly acknowledge the encouragement and support provided by Professor Danuta Mendelson and Dr Andrew Torre, who I first met in 2013 at the Deakin University School of Law research hub on law and economics. I would like to thank the Honourable Robert French AC, immediate past Chief Justice of the High Court of Australia, for agreeing to write the Foreword. I  am grateful for helpful comments on a prior draft of this monograph from Dr  Liza Lovdahl Gormsen of the University of Manchester Law School and ­Associate Professor Hedvig Schmidt of the University of Southampton Law School. I also acknowledge helpful input from editors and anonymous reviewers in relation to my articles in peer-reviewed law journals, the details of which are set  out below. This monograph draws on and develops my ideas previously published in various articles and conference presentations, which are acknowledged: a. ‘Taking Advantage of Market Power: The Competitive Market Hypothesis’ (1999) 73 Law Institute Journal 70; b. ‘Misuse of Market Power: Why Policy Objectives Matter’ (2014) 22 Competition & Consumer Law Journal 1;

x  Acknowledgements c. ‘Changing Views Concerning “Competition Effects” in the US, EU and Australia: The Competition Policy Review’, presentation at the Australasian Law Teachers Association Conference, 16–18 July 2015; d. ‘Shifting the Goalposts: Current Issues in Australian Competition Law Affecting the Energy Sector’ (2016) 9 Journal of World Energy Law and Business 424; e. ‘Competition and Efficiency Effects in Europe, North America and Australia’ (2016) 24 Competition & Consumer Law Journal 187. This monograph sets out a legal theory of market power, but you cannot address competition law without considering economics and how the law has adapted and applied economic concepts. The cases on market power reveal substantial differences between the perspectives of lawyers and economists. In part, this derives from economic theory moving on while the law lags behind, since appellate decisions take a long time and are relatively infrequent. I have enjoyed many discussions with Dr Andrew Torre and his colleague Dr Ching-Jen Sun of the Deakin University Faculty of Business on game theory and mathematical models. I am also grateful for helpful comments from my colleague from the early 1970s, Vincent Martin, who pursued a career in economics, latterly as director of competition and regulation policy with the Department of Treasury and Finance of the State of Victoria. I have formed the view that the law cannot directly apply economics, but must adapt economic principles, eg, by modifying assumptions of a model, we can develop legal concepts that can be applied forensically. Thus, I suggest that the concept of an ‘efficient market’ in the law of securities market manipulation is a legal construct. Last but not least, I am grateful for the support of my family and the quiet encouragement of colleagues at my law firm, Piper Alderman, Melbourne, where I am now a Consultant. My daughter Emily has a keen eye and has helped me greatly with proofreading. All tables and figures in this monograph have been drawn by the author. Figures 1.2 and 1.3 have previously been published in the 2014 Competition & Consumer Law Journal article, and Figure 1.4 has been previously published in the 2016 Competition & Consumer Law Journal article noted above. Figure 7.1 is a schematic representation of Annexure D to the judgment in ACCC v Cement Australia Pty Ltd (2013) 310 ALR 165; [2013] FCA 909. The law is relevantly stated as at 31 July 2018. The ‘time of writing’ referred to in the monograph is late August 2018.

CONTENTS Foreword������������������������������������������������������������������������������������������������������������������������v Preface������������������������������������������������������������������������������������������������������������������������� vii Acknowledgements������������������������������������������������������������������������������������������������������ ix Abbreviations��������������������������������������������������������������������������������������������������������������xv Table of Cases����������������������������������������������������������������������������������������������������������� xvii Table of Legislation���������������������������������������������������������������������������������������������������� xxi List of Figures and Tables���������������������������������������������������������������������������������������� xxiii Introduction���������������������������������������������������������������������������������������������������������������������1 I. The Problem�������������������������������������������������������������������������������������������������3 II. Australia’s Reform Process�������������������������������������������������������������������������5 III. The Proposed Market Manipulation Approach��������������������������������������7 IV. Implications of the Study beyond the Present Scope���������������������������13 V. The Problematic Role of Efficiency���������������������������������������������������������15 VI. Differing Views of the World�������������������������������������������������������������������18 VII. Economic Models and the Paradigm of Competition�������������������������19 VIII. The Overseas Experience�������������������������������������������������������������������������22 IX. What Difference Will Market Manipulation Theory Make?���������������24 X. Application to the Digital Economy�������������������������������������������������������30 XI. Outline of the Chapters����������������������������������������������������������������������������31 XII. Effective Date of Law��������������������������������������������������������������������������������34 PART I THE THEORETICAL FRAMEWORK 1. Market Power in Economics and Law�����������������������������������������������������������������39 I. Introduction�����������������������������������������������������������������������������������������������39 II. Policy Objectives and Concepts of Efficiency���������������������������������������43 III. Economic Models and the Paradigm of Competition�������������������������49 IV. Market Manipulation Laws in Securities Markets��������������������������������59 V. Thought Experiment: Incumbent Response to New Entry�����������������61 VI. Norms in Zero-Sum Competition����������������������������������������������������������70 VII. Zero-Sum Competition, Market Manipulation and Efficiency����������75 VIII. Conclusion�������������������������������������������������������������������������������������������������76

xii  Contents 2. The Mischief and Australia’s Institutional Response�����������������������������������������78 I. Introduction�����������������������������������������������������������������������������������������������78 II. The Continuing Influence of the SCP Approach in Australia�������������78 III. Australia’s Institutional Division of Functions between Courts and Agencies����������������������������������������������������������������������������������������������81 IV. Institutional Arrangements in the EU and the US: Court-Centred or Agency-Centred?������������������������������������������������������86 V. Conclusion�������������������������������������������������������������������������������������������������87 3. Competition and Efficiency Effects in Europe, North America and Australia�����������������������������������������������������������������������������������������������������������89 I. Introduction�����������������������������������������������������������������������������������������������89 II. The Current Australian Effects Test Outside Section 46����������������������95 III. The Harper Review Effects Test and the Role of Efficiency Effects�������������������������������������������������������������������������������������������������������102 IV. EU Abuse of Dominance and Defences�����������������������������������������������107 V. US Monopolisation and Defences��������������������������������������������������������113 VI. Efficiency in the Law of Securities Market Manipulation�����������������117 VII. Adapting Efficiency as a Legal Concept�����������������������������������������������118 VIII. Zero-Sum Competition, Market Manipulation and Efficiency��������124 IX. Conclusion�����������������������������������������������������������������������������������������������126 PART II TESTING MARKET MANIPULATION AND EFFICIENCY APPROACHES Introduction���������������������������������������������������������������������������������������������������������129 4. Refusal to Deal and Margin Squeeze�����������������������������������������������������������������133 I. Introduction���������������������������������������������������������������������������������������������133 II. Queensland Wire: Factual Matrix����������������������������������������������������������136 III. Gaps in the Factual Matrix��������������������������������������������������������������������140 IV. Queensland Wire under Harper Section 46�����������������������������������������142 V. Queensland Wire under Market Manipulation�����������������������������������146 VI. Melway: Factual Matrix��������������������������������������������������������������������������149 VII. Melway under Harper Section 46����������������������������������������������������������151 VIII. Melway under Market Manipulation����������������������������������������������������154 IX. Conclusion�����������������������������������������������������������������������������������������������154 5. Predatory Pricing�������������������������������������������������������������������������������������������������155 I. Introduction���������������������������������������������������������������������������������������������155 II. Boral: Factual Matrix������������������������������������������������������������������������������158

Contents  xiii III. Gaps in the Factual Matrix��������������������������������������������������������������������160 IV. Boral under Harper Section 46��������������������������������������������������������������163 V. Boral under Market Manipulation��������������������������������������������������������166 VI. Conclusion�����������������������������������������������������������������������������������������������166 6. Meeting Competition������������������������������������������������������������������������������������������169 I. Introduction���������������������������������������������������������������������������������������������169 II. Rural Press: Factual Matrix��������������������������������������������������������������������170 III. Gaps in the Factual Matrix��������������������������������������������������������������������172 IV. Rural Press under Harper Section 46����������������������������������������������������174 V. Rural Press under Market Manipulation����������������������������������������������175 VI. Conclusion�����������������������������������������������������������������������������������������������176 7. Raising Rivals’ Costs��������������������������������������������������������������������������������������������178 I. Introduction���������������������������������������������������������������������������������������������178 II. Cement Australia: Factual Matrix���������������������������������������������������������182 III. Gaps in the Factual Matrix��������������������������������������������������������������������195 IV. Cement Australia under Harper Section 46�����������������������������������������196 V. Cement Australia under Market Manipulation�����������������������������������198 VI. Conclusion�����������������������������������������������������������������������������������������������198 8. Bundling����������������������������������������������������������������������������������������������������������������200 I. Introduction���������������������������������������������������������������������������������������������200 II. Baxter: Factual Matrix����������������������������������������������������������������������������202 III. Gaps in the Factual Matrix��������������������������������������������������������������������207 IV. Baxter under Harper Section 46������������������������������������������������������������208 V. Baxter under Market Manipulation������������������������������������������������������210 VI. The Implications of Pfizer����������������������������������������������������������������������211 VII. Conclusion�����������������������������������������������������������������������������������������������213 9. Institutional and Procedural Implications��������������������������������������������������������214 I. Introduction���������������������������������������������������������������������������������������������214 II. NT Power��������������������������������������������������������������������������������������������������214 III. The Burden of Proof�������������������������������������������������������������������������������217 IV. Conclusion�����������������������������������������������������������������������������������������������219 PART III CONCLUSION 10. Findings and Recommendations�����������������������������������������������������������������������223 I. Introduction���������������������������������������������������������������������������������������������223 II. The Concept of Market Power���������������������������������������������������������������225

xiv  Contents III. Paradigms of Competition and Norms of Conduct���������������������������225 IV. Forensic Assessment of Misuse of Market Power: Conclusions from the Case Studies�����������������������������������������������������������������������������227 V. Implications for Competition Law�������������������������������������������������������234 Appendix: Text of the Old and New Versions of Section 46 of the CCA�������������238 Bibliography���������������������������������������������������������������������������������������������������������������244 Index��������������������������������������������������������������������������������������������������������������������������255

ABBREVIATIONS ABA

American Bar Association

ACCC

Australian Competition and Consumer Commission (formerly the Trade Practices Commission)

ACT

Australian Competition Tribunal (formerly the Trade Practices Tribunal)

AMC

Antitrust Modernisation Commission

ASIC

Australian Securities and Investments Commission

CA

Cement Australia Pty Ltd (and related bodies corporate)

CCA

Competition and Consumer Act 2010 (Cth) (formerly the Trade Practices Act 1974)

CJEU

Court of Justice of the European Union

DPP

Director of Public Prosecutions

EC

European Commission

FCA

Federal Court of Australia

FCAFC

FCA Full Court

HCA

High Court of Australia

HD

Haemodialysis

PAWA

Power and Water Authority

PD

Sterile fluids for use in peritoneal dialysis

SCP

The Structure-Conduct-Performance approach to assessing competi­tion effects of business conduct

SF

Sterile fluids for clinical use in humans, other than PD products

SLC

substantial lessening of competition in a relevant market (the competition test in Part IV of the CCA)

SPA

State purchasing authority

TFEU

Treaty on the Functioning of the European Union

TPC

Trade Practices Commission (now the ACCC)

xvi

TABLE OF CASES Australian Tribunals Re Qantas Airways Ltd [2004] ACompT 9 ����������������������������������16, 41, 47, 110, 121 Re Application by Chime Communications Pty Ltd (No 3) (2009) ACompT 4 ���������������������������������������������������������������������������������������������������������������79 Re Ford Motor Company (1977) ATPR 40-043��������������������������������������������� 100, 115 Re Glencore Coal Pty Ltd and Port of Newcastle Pty Ltd [2016] ACompT 6��������������������������������������������������������������������������������������������������������������135 Re Application for Authorisation of Acquisition of Macquarie Generation by AGL Energy Ltd [2014] ACompT 1 ������������������������������������������������ 39, 47, 216 Re Queensland Co-operative Milling Association Ltd (1976) 8 ALR 481 �������������������������������������������������������������������������������������������39, 79, 95, 240 Re Tooth & Co Ltd and Tooheys Ltd (1979) 39 FLR 1���������������������������������������������97 Australian Courts ACCC v ACT [2017] FCAFC 150��������������������������������������������������������16, 47, 110, 121 ACCC v Australia and New Zealand Banking Group Ltd (2015) 236 FCR 78�������������������������������������������������������������������������������������������������������� 20, 22 ACCC v Australian Safeway Stores Ply Ltd [2003] FCAFC 149�����������������������������15 ACCC v Baxter Healthcare Pty Ltd (2007) 232 CLR 1; 237 ALR 512; [2007] HCA 8��������������������������������������������������������������������������������������������������������201 ACCC v Baxter Healthcare Pty Ltd [2008] FCAFC 141 �����������������35, 201, 205–07, 209, 212 ACCC v Baxter Healthcare Pty Ltd [2005] FCA 581 �������������������������������������� 201–09 ACCC v Cement Australia Pty Ltd (2013) 310 ALR 165; [2013] FCA 909 ���������������������������������������������������������������������� viii, 95, 179–98, 201, 209–10, 215, 219, 241 ACCC v Cement Australia Pty Ltd [2014] FCA 148 ���������������������������������������������181 ACCC v Cement Australia Pty Ltd [2016] FCA 453������������������������������������� 181, 192 ACCC v Cement Australia Pty Ltd [2017] FCAFC 159 ����������������������������������������181 ACCC v Coles Group Ltd [2014] FCA 363����������������������������������������������������������������40 ACCC v Colgate-Palmolive Pty Ltd & Others (No 4) [2017] FCA 1590���������������65 ACCC v Flight Centre Travel Group Ltd [2016] HCA 49�������������������22, 79, 90, 146 ACCC v Leahy Petroleum Pty Ltd [2007] FCA 794�������������������������������������������������65 ACCC v Metcash Trading Ltd (2011) 284 ALR 662�������������������������������������������������97

xviii  Table of Cases ACCC v Pfizer Australia Pty Ltd [2018] FCAFC 78 ����������� 22, 35, 96, 111, 200–01, 205–06, 211–13, 234, 241 ACCC v Woolworths Ltd [2014] FCA 364����������������������������������������������������������������40 Apco Service Stations Pty Ltd v ACCC [2005] FCAFC 161������������������������������������65 ASIC v National Australia Bank Ltd [2017] FCA 1338��������������������������������������������59 ASIC v Westpac Banking Corporation (No 2) [2018] FCA 751�������������������� 60, 117 ASX Operations Pty Ltd v Pont Data Australia Pty Ltd (1990) 97 ALR 513 ��������������������������������������������������������������������������������������������� 95, 140, 228 ASX Operations Pty Ltd v Pont Data Australia Pty Ltd (1990) 21 FCR 385 ������������������������������������������������������������������������������������������������������������228 Australian Energy Regulator v ACT (No 2) [2017] FCAFC 79��������������������������������2 Australian Gas Light Company v ACCC (No 3) (2003) 137 FCR 317; [2003] FCA 1525����������������������������������������������������������������������������15, 40, 80, 83, 97 ACCC v Australian Safeway Stores Pty Ltd (2003) 129 FCR 339���������������������������91 Boral Besser Masonry Ltd v ACCC (2003) 215 CLR 374;195 ALR 609; [2003] HCA 5 ��������������������������������������������������������� 46, 115, 121–22, 155–67, 175, 179–80, 187, 210, 230, 240 ACCC v Boral Ltd [1999] FCA 1318 ������������������������������������������������� 155–56, 158–61 Burge v Swarbrick (2007) 232 CLR 336��������������������������������������������������������������������126 Bryan v Maloney (1995) 182 CLR 609�����������������������������������������������������������������������71 Dandy Power Equipment Pty Ltd v Mercury Marine Pty Ltd [1982] FCA 178; (1982) 64 FLR 238�������������������������������������������������������������������������������151 D’Arcy v Myriad Genetics Inc (2015) 258 CLR 334; [2015] HCA 35�������������������128 DPP (Cth) v JM (2013) 250 CLR 135; [2013] HCA 30������������������������������� 8, 60, 126 Jones v Dunkel (1959) 101 CLR 298�������������������������������������������������������������������������219 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1; 178 ALR 253; [2001] HCA 13������������������������������������22, 41, 79, 90, 145, 149–154 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (1999) 169 ALR 554������������������������������������������������������������������������������� vii, 91, 96, 150, 152 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (1998) 42 IPR 627���������������������������������������������������������������������������������������������145, 150, 153 Murphy v Electoral Commissioner [2016] HCA 36�������������������������������������������������73 NT Power Generation Pty Ltd v Power and Water Authority (2004) 219 CLR 90; 210 ALR 312; [2004] HCA 48 �����������������������������������������������������99, 124, 126, 135, 214–17, 233 NT Power Generation Pty Ltd v Power and Water Authority (2002) 122 FCR 399�����������������������������������������������������������������������������������������������������������215 NT Power Generation Pty Ltd v Power and Water Authority (2001) 184 ALR 481���������������������������������������������������������������������������������������������������� 215–16 Outboard Marine Pty Ltd v Hecar Investments (No 6) Pty Ltd [1982] FCA 265; (1982) 66 FLR 120�������������������������������������������������������������������������������151 The Pilbara Infrastructure Pty Ltd v ACT (2012) 246 CLR 379; [2012] HCA 36 ���������������������������������������������������������������������������������������������������2, 83

Table of Cases  xix Port of Newcastle Operations Pty Ltd v Australian Competition Tribunal [2017] FCAFC 124������������������������������������������������������������������������������������������������135 Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd (1989) 167 CLR 177; 83 ALR 577 ����������������������������������������21–22, 39, 44, 63, 71, 74, 90, 97–98, 105, 110, 114, 136–39, 141–49, 185, 199, 233 Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd (1987) 17 FCR 211 ����������������������������������������������������������������������������������137 Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd (1987) 16 FCR 50 ������������������������������������������������������136–37, 139–42, 149 Rural Press Ltd v ACCC (2003) 216 CLR 53; ALR 217; [2003] HCA 75 ������������������������������������������������������������������� 22, 79, 90, 169–76, 240 Rural Press Ltd v ACCC (2002) FCAFC 213��������������������������������������������169–70, 173 ACCC v Rural Press Ltd [2001] FCA 116 �������������������������������������������������������� 169–70 Seven Network Ltd v News Ltd (2009) 262 ALR 160; 182 FCR 160���������������� 22, 96 Strong v Woolworths Ltd (2012) 285 ALR 420���������������������������������������������������������98 Sydney Airport Corporation v Australian Competition Tribunal (2006) 155 FCR 124����������������������������������������������������������������������������������������������135 Tillmans Butcheries Pty Ltd v Australasian Meat Industry Employees’ Union (1979) 27 ALR 367��������������������������������������������������������������������������������������������������97 TPC v Email Ltd (1980) 31 ALR 53 ���������������������������������������������������������������������������85 Universal Music Australia Pty Ltd v ACCC (2003) 201 ALR 636������������ 13, 23, 96, 98, 104, 144, 151 Zhu v The Treasurer of the State of New South Wales (2004) 218 CLR 530�������������������������������������������������������������������������������������������������������������74 Overseas Cases Aspen Skiing Co v Aspen Highlands Skiing Corporation 472 US 585 (1985)������������������������������������������������������������������������������������������� 3, 115, 218 Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223�������������������������������������������������������������������������������������������������������72 Barry Wright Corporation v ITT Grinnell Corporation 724 F2d 227 (1st Cir 1983)���������������������������������������������������������������������������������������������������������115 Commissioner of Competition v Canada Pipe Company Ltd (2006) FCA 233 (Canadian Federal Court) ����������������������������������������������������������� 105–06 Co-operative Insurance Society Ltd v Argyll Stores (Holdings) Ltd [1997] UKHL 17����������������������������������������������������������������������������������������������������������������176 Case C-453/99 Courage Ltd v Bernard Crehan (2001) (CJEU, 20 September 2001)������������������������������������������������������������������������������������������������72 In re Crude Oil Commodity Futures Litigation 913 F Supp 2d 41 (US District Court, SDNY, 2012)��������������������������������������������������������������������������93

xx  Table of Cases Case C280/08P Deutsche Telekom AG v European Commission (2010) (CJEU, 14 October 2010)�������������������������������������������������������������������������������������135 Director of Investigation and Research v NutraSweet Co No CT-89/2 (Canadian Competition Tribunal, 1990)�����������������������������������������������������������106 Case C-202/07P France Télécom SA v Commission (2009) (CJEU, 2 April 2009)�����������������������������������������������������������������������������������������111, 116, 155 Case T-340/03 France Télécom SA v Commission (2007) (General Court, 30 January 2007)����������������������������������������������������������������������������������������������������111 Illinois Corporate Travel Inc v American Airlines Inc 806 F 2d 722 (7th Cir 1986)����������������������������������������������������������������������������������������������������������22 Case C-413/14 P Intel Corporation Inc v EC [2017] (CJEU, 6 September 2017, rectified 19 September and 24 October 2017)����������� 35, 91, 107, 109, 111–12, 185 Case C-52/09 Konkurrensverket v TeliaSoneraSverige (2011) (CJEU, 17 February 2011) ��������������������������������������������������������������90–91, 107, 135 Matsushita Electric Industrial Co v Zenith Radio Corporation 475 US 574 (1986)��������������������������������������������������������������������������������������������������������115 Pacific Bell Telephone Co v Linkline Communications, Inc 555 US 1 (2009)����������������������������������������������������������������������������������������������135 Case C-209/10 Post Danmark A/S v Konkurrencerådet (2012) (CJEU, 27 March 2012) �������������������������������������������������������������������������������� 91, 107 Daniel Shak v JP Morgan Chase & Co (US District Court SDNY, 12 January 2016)������������������������������������������������������������������������������������������������������93 Telecom Corporation of New Zealand Ltd v Clear Communications Ltd [1995] 1 NZLR 385 (UK Privy Council on appeal from the New Zealand Court of Appeal)�����������������������������������������������������������������������������������������������������91 In re Term Commodities Cotton Futures Litig. No. 12 Civ. 5126 (ALC), 2013 WL 9815198 (US District Court, SDNY, 2013)�����������������������������������������93 US v Patten, 226 US 525 (1913)���������������������������������������������������������������������������������114 US v Aluminum Co of America 148 F2d 416 (2nd Cir 1945)���������������������� 115, 178 US v Griffith 334 US 100 (1948)��������������������������������������������������������������������������������115 Verizon Communications Inc v Law Offices of Curtis v Trinko, LLP 540 US 398 (2004)�������������������������������������������������������������������������������������������������135

TABLE OF LEGISLATION Australia Australia Act 1986 (Cth)����������������������������������������������������������������������������������������������91 Competition and Consumer Act 2010 (Cth) (formerly the Trade Practices Act 1974)������������������������������������������������������������vii, 2–4, 6, 8, 14–15, 22, 40–43, 53, 79–87, 89, 91–92, 96, 116–17, 127–28, 131, 135–36, 142, 145, 147, 156, 167, 169, 181, 188, 190, 201, 215, 228, 230, 238–43 Competition and Consumer Amendment (Competition Policy Review) Act 2017���������������������������������������������������������������������� 5–7, 41, 94, 98, 136 Corporations Act 2001 (Cth)����������������������������������������������������������������59, 71, 117, 157 Crimes Act 1958 (Vic)������������������������������������������������������������������������������������������������163 Patents Act 1990 (Cth) �����������������������������������������������������������������������������������������������128 Overseas Australia Act 1986 (UK)�����������������������������������������������������������������������������������������������91 Commerce Act 1986 (NZ)������������������������������������������������������������������������������������������3, 5 Commonwealth of Australia Constitution Act (UK)���������������������������������������������217 Competition Act 1985 (Canada)�������������������������������������������������������������������������������105 Treaty on the Functioning of the European Union (EU)����� 3, 90–91, 107, 112, 155 Sherman Antitrust Act 15 USC (US)��������������������������������������3, 16, 22, 47, 50, 53, 61, 80, 82, 87, 90, 93, 113–14

xxii

LIST OF FIGURES AND TABLES Figure 1.1 Figure 1.2 Figure 1.3 Figure 1.4 Figure 1.5 Figure 3.1

Monopoly equilibrium Monopoly price and output Allocative and productive efficiency Payoff matrix with strategies Demand curve, marginal revenue and total revenue Allocative, productive and dynamic efficiency in a two-product economy Figure 3.2 Effect of demand reduction (if there were a supply curve) Figure 7.1 Fly ash distribution in SE Queensland Figure 8.1 Market shares in PD products Figure 8.2 Market shares in HD products Figure 10.1 The inter-relationship between competition and efficiency Table 0.1 Comparison of legal theories of liability for misuse of market power

52 55 55 64 67 118 122 183 203 203 230 24

xxiv

Introduction Economists advocating for effective competition laws in Australia in the 1960s were troubled by the power of large firms in highly concentrated markets.1 Markets today appear to be even more highly concentrated, which continues to be of concern to competition regulators.2 According to Andrew Gavil, the aim of prohibiting misuse of market power is to prevent incumbent firms with substantial market power in highly concentrated industries insulating themselves from competitive challenges and thus perpetuating a concentrated market structure.3 The ‘structuralist approach’, known in the EU as ‘ordo-liberalism’ and in the US and Australia as the ‘Structure-Conduct-Performance’ (SCP) approach, is generally regarded as having been superseded by developments in economic theory.4 This view held that ‘the presence of dominant companies weakens the competitive process’.5 Economic theory now accepts that ‘competitive dynamics can function well even if a market has some very large suppliers’.6 In a speech in October 2016, Rod Sims, the Chairman of the Australian Competition and Consumer Commission (ACCC), noting that concentration has increased in the Australian economy, questioned the validity of economic research ‘being used to justify increases in market concentration’ and suggested that this

1 PH Karmel and M Brunt, The Structure of the Australian Economy (Melbourne, FW Cheshire, 1966). 2 R Sims, ‘Is Australia’s Economy Getting More Concentrated and Does This Matter?’, speech delivered at the RBB Economics Conference, Sydney, 27 October 2016, www.accc.gov.au/speech/ keynote-address-rbb-economics-conference-0. See also A Gavil, ‘Imagining a Counterfactual Section 36: Rebalancing New Zealand’s Competition Law Framework’ (2015) 46 Victoria University of Wellington Law Review 1043. 3 Gavil (n 2) 1048. 4 See, eg, R Smith and D Round, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5 Agenda 25; H Hovenkamp, Federal Antitrust Policy (St Paul, West Publishing, 2005) 42–47; G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford ­University Press, 2011) 180. L Lovdahl Gormsen, A Principled Approach to Abuse of Dominance (Cambridge, Cambridge University Press, 2010) 57–58 notes that there are some differences between ordo-liberalism and SCP, particularly regarding the significance attributed to economic efficiencies. 5 Niels, Jenkins and Kavanagh (n 4) 180. 6 ibid. See also JM Clark ‘Toward a Concept of Workable Competition’ (1940) 30 American Economic Review 241; Commonwealth of Australia, National Competition Policy (Canberra, ­Australian Government Publishing Service, 1993) (hereinafter ‘Hilmer Report’) 3; T Muris, ‘Improving the Economic Foundations of Competition Policy’ (2003) 12 George Mason Law Review 1; S Stroux, US and EC Oligopoly Control (The Hague, Kluwer Law International, 2004) 11–12; J Perloff, L Karp and A Golan, Estimating Market Power and Strategies (Cambridge, Cambridge University Press, 2007) 34–37; J Perloff, R Smith and D Round, Microeconomics (Frenchs Forest, Pearson Australia, 2014) 431–33.

2  Introduction should not replace ‘conventional wisdom’ that increasing concentration normally will harm economic efficiency.7 The speech refers to differing views about the objective of competition laws, or standard of efficiency, which are notorious in the field of competition law.8 As to the conventional kinds of economic efficiency, see for example, the discussion in the Hilmer Report, which outlines the conventional components of ‘economic efficiency’: productive efficiency, allocative efficiency and dynamic efficiency.9 Productive efficiency also embraces transaction cost efficiencies and economies of scale and scope. The components of economic efficiency are illustrated further in Chapter 1 by reference to Figure 1.3. For those of us who, like Sims, completed our undergraduate degrees in economics and law in Australia in the early 1970s, the contemporary debate creates a sense of déjà vu. At the time, we accepted SCP thinking as an article of faith. However, while we can acknowledge and respect the pivotal role of SCP thinking in the development of current laws, we now know that economic thought has moved on. I suggest that Australian law has lagged behind, for various reasons that are discussed in this monograph. In the decades that followed the introduction of effective competition law in Australia in 1974, the elegance of the SCP approach has been tested and found wanting in the harsh reality of business and our legal system. The speech by Sims raises important questions, which are addressed in this study, about the role of market structure, and differing views about economic efficiency and non-economic considerations, in the assessment of market power and its misuse. The debate in the literature concerns whether allocative or productive efficiency is to prevail when, under real-world conditions, one can only be achieved at the expense of the other. These differing views are addressed in ­Chapters  1 and 3. Conventional concepts of economic efficiency refer to social efficiency. However, we cannot measure the impact on social efficiency. We must therefore adapt economic concepts of efficiency for use in our legal theory. Economic efficiency has not been judicially defined in the present context.10 I propose to start with the legal concept of an efficient market from securities markets manipulation laws. This is a market in which undistorted forces of demand

7 Sims (n 2). 8 ibid. See also, eg, R Steinwall, ‘The Dawson Committee on Competition, Economic Efficiency, Universality and Public Policy’ (2003) 26 University of New South Wales Law Journal 226, who observes that views differ as to whether efficiency should be the sole objective of competition law and, indeed, that Australian law contemplates ‘broader equity and public interest goals including political freedom, protection of small business, equity in economic dealings and comity’. Further literature regarding the international debate will be referred to below. 9 Hilmer Report (n 6) 1–7. 10 Economic efficiency has been considered as an objective in the context of the statutory regime providing access to essential facilities in pt IIIA of the Competition and Consumer Act 2010 (Cth), but has not been without its difficulties in adapting from a social to industry application. See, eg, The Pilbara Infrastructure Pty Ltd v Australian Competition Tribunal (2012) 246 CLR 379, [2012] HCA 36; Australian Energy Regulator v Australian Competition Tribunal (No 2) [2017] FCAFC 79. The access regime is beyond the scope of this monograph.

The Problem  3 and supply ­operate. This study proposes a new legal theory concerning the extent to which the law can adapt and forensically apply the conventional concepts of economic efficiency to the conduct of a dominant firm. The subject matter of the monograph is therefore law and legal determinacy in developing legal concepts that adapt or make use of economic principles and concepts in a way that can be forensically tested. To develop the legal theory, we need to discuss economics to the extent necessary in order to identify the legal problem to be addressed, and to follow the process of reasoning that has led me to propose alternative legal concepts. We will see that when we depart from the idealised assumptions of an economic model, as we must to take account of real-world conditions which do not conform to those assumptions, we can open alternative ways of seeing the concepts and operative forces.

I.  The Problem Misuse of market power (or unilateral dominant firm conduct) is one of the most complex and controversial areas in competition policy, which is acknowledged by Australia’s recent Competition Policy Review (which I refer to as the ‘Harper Review’, or the ‘Review’ or simply ‘Harper’).11 There is a continuing divergence of opinion in Australia and overseas concerning the standard, or approach, to be applied in determining whether a dominant firm’s conduct in the market amounts to an unlawful misuse of market power or may be permitted as a normal incident of competition.12 I therefore suggest that our legal theory of liability for

11 See, eg, D Bernheim and R Heeb, ‘A Framework for the Economic Analysis of Exclusionary Conduct’ in R Blair and D Sokol (eds) The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 2, 3; M Schinkel and P Larouche, ‘Continental Drift in the Treatment of Dominant Firms: Article 102 TFEU in Contrast to Section 2 Sherman Act’, in Blair and Sokol, ibid, 154 (note: ‘Larouche’ is spelt as in the original Tilburg Law School research paper, but is rendered as ‘LaRocuhe’ in the Oxford publication). See also Competition Policy Review, Issues Paper, 14 April 2014, para 5.7. 12 See, eg, B Facey and D Assaf, ‘Monopolization and Abuse of Dominance in Canada, the United States, and the European Union: A Survey’ (2002) 70 Antitrust Law Journal 513; P Crampton, ‘“Abuse” of “Dominance” in Canada: Building on the International Experience’ (2005) 73 Antitrust Law ­Journal 803; W Reid, ‘Aspen Skiing in a Sunburnt Country’ (2005–06) 73 Antitrust Law Journal 209; G Hay and R Smith, ‘“Why Can’t a Woman Be More Like a Man?” – American and Australian Approaches to Exclusionary Conduct’ (2007) 31 Melbourne University Law Review 1099; A Meese, ‘Reframing Antitrust in Light of Scientific Revolution: Accounting for Transaction Costs in Rule of Reason Analysis’ (2010) 62 Hastings Law Journal 457; M Rosario, A Torre, D Mendelson and L Griggs, ‘Software Developer Policies: Economics and s 46 of the Competition and Consumer Act 2010’ (2011) 21 Journal of Law, Information and Science 37; P Akman, The Concept of Abuse in EU Competition Law (Oxford, Hart Publishing, 2012); H Schmidt, ‘Market Power – The Root of All Evil? A Comparative Analysis of the Concepts of Market Power, Dominance and Monopolisation’ in A Ezrachi (ed), Research Handbook on International Competition Law (Cheltenham, Edward Elgar, 2012); D ­Geradin, A Layne-Farrar and N  Petit, EU Competition Law and Economics (Oxford, Oxford University Press, 2012); J  Cross, D  ­Richards, M Stucke and S Waller, ‘Use of Dominance, Unlawful Conduct, and Causation under Section  36 of the New Zealand Commerce Act: A US Perspective’ (2013)

4  Introduction misuse of market power is insufficiently determined to satisfy the requirements of legal determinacy and rule of law concerns. Consequently, experience has shown that the law has lacked predictability and certainty, which has led to the reform process that took place in Australia, most notably between 2014 and 2017. Australia’s prohibition, set out in section 46 of the Competition and Consumer Act 2010 (Cth) (CCA), was, prior to amendments following the Harper Review, based on purpose rather than effect, and required the court to undertake a counterfactual assessment of how a dominant firm would be likely to conduct itself under circumstances of competition, ie, in the absence of its market power. According to Kathryn McMahon, Australian judicial reasoning in this area of competition law ‘mask[s] undisclosed policy decisions’13 and consequently there is a need for a ‘deeper theoretical understanding’ of the role of economic theory, legal precedent and public policy in the regulation of economic activity.14 Michael O’Bryan suggests that Australian judges have tended to rely on ‘commercial intuition’ and that there is a need for legal and economic reasoning that provides certainty and predictability for business.15 Even in the US context, where there is a more significant body of legislative and judicial experience in the field, according to Keith Hylton, ‘surprisingly few scholars have attempted to provide either a positive or normative theory of monopolisation law’16 and consequently ‘lack of clarity has been a long running problem’.17 Herbert Hovenkamp takes a more pragmatic view, suggesting that, although our understanding of market power and our ability to empirically identify it is ‘too crude for making fine judgments’, we can nevertheless adequately

New Zealand B ­usiness Law Quarterly, ssrn.com/abstract=2170538; L White, ‘Monopoly and Dominant Firms: Antitrust Economics and Policy Approaches’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 1, 313; A Merrett, ‘Understanding Market Power’ in J Duns, A Duke and B Sweeney (eds) Comparative Competition Law (Cheltenham, Edward Elgar, 2015) 109; R Miller, ‘The Ongoing Australian Struggle with Misuse of Marker Power’ (2016) 24 Australian Journal of Competition and Consumer Law 242; AC Witt, The More Economic Approach to EU Antitrust Law (Oxford, Hart Publishing, 2016); K Kemp, ‘The Big Chill? A Comparative Analysis of Effects-Based Tests for Misuse of Market Power’ (2017) 40 University of New South Wales Law Journal 493; L Kaplow, ‘On the Relevance of Market Power’ (2017) 130 Harvard Law Review 1303; J Clarke, ‘Section 46: Its Purpose and the Proposed New Effects Test’ (2017) Australian Business Law Review 364; H Schmidt, ‘Taming the Shrew: There’s No Need for a New Market Power Definition for the Digital Economy’, Academic Society for Competition Law conference paper, June 2017, ssrn.com/abstract=3048266; K Kemp, Misuse of Market Power: Rationale and Reform (Cambridge, Cambridge University Press, 2018). 13 K McMahon, ‘Competition Law, Adjudication and the High Court’ (2006) 30 Melbourne University Law Review 782, 784. 14 ibid 836. 15 M O’Bryan, ‘Section 46: Law or Economics? (1993) 1 Competition & Consumer Law Journal 64, 78; see also M O’Bryan, ‘Section 46: Legal and Economic Principles and Reasoning in Melway and Boral’ (2001) 8 Competition & Consumer Law Journal 1. 16 K Hylton, ‘The Law and Economics of Monopolization Standards’ in K Hylton (ed) Antitrust Law and Economics (Cheltenham, Edward Elgar, 2010) 82, 88. 17 ibid 107; see also J Baker ‘Exclusion as a Core Competition Concern’ (2013) Antitrust Law Journal 527.

Australia’s Reform Process  5 identify ‘the most flagrant abuses’.18 This lack of clarity appears to have contributed to what William Kovacic calls the ‘intervention scepticism’ of US courts.19

II.  Australia’s Reform Process The Australian legislature has accepted the Harper Review’s recommendation that the statutory prohibition should condemn conduct of a firm having substantial market power where the conduct has the purpose or likely effect of substantially lessening competition (SLC), but has rejected the recommendation that the court must have regard to any offsetting efficiency effects.20 An exposure draft Bill and Explanatory Memorandum adopting the Review’s recommendations for section 46 in all respects were first released in September 2016, together with an exposure draft guidance from the ACCC.21 I suggest that the Review’s proposed requirement, set out in proposed section 46(2), that efficiency effects be taken into account in judicial proceedings should be welcomed.22 However, neither 18 H Hovenkamp, Antitrust Enterprise: Principle and Execution (Cambridge, MA, Harvard University Press, 2005) 95. 19 W 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, 71–73. 20 The Competition and Consumer Amendment (Misuse of Market Power) Act 2017 was passed by the House of Representatives on 31 March 2017 after the Bill as first introduced was amended in the Senate to omit the Harper recommended section 46(2); compare Commonwealth of Australia, Competition Policy Review, March 2015 (hereinafter ‘Harper Review’), Appendix A, 513–14, setting out the recommended form of s 46 (extracts are set out in the Appendix to this book). The Bill was finally passed by both Houses of Parliament and received Royal Assent on 23 August 2017. The amended version of s 46 entered into force on 6 November 2017. The corresponding New Zealand provision is s 36 of the Commerce Act 1986. The responsible Minister has undertaken a targeted review of provisions including s 36 since 2014. At the time of writing, this review was incomplete. The New Zealand Cabinet on 28 March 2018 appears to have accepted the recommendation of the responsible Minister that officials should seek further empirical evidence of harm occurring and the costs and benefits of alternatives to the current formulation of s 36, with a view to further consideration by the Cabinet before the end of 2018: see Ministry of Business, Innovation and Employment, ‘Targeted Review of Commerce Act 1986’, www.mbie.govt.nz/info-services/business/competition-policy/targeted-reviewof-the-commerce-act/phase-one. At the time of writing, the New Zealand provision corresponds to Australia’s old s 46, ie, it retains the ‘taking advantage’ element repealed in Australia. I do not consider New Zealand law in this book. 21 Exposure Draft Competition and Consumer Amendment (Competition Policy Review) Bill 2016 and Explanatory Memorandum, September 2016; ACCC, consultation draft ‘Framework for Misuse of Market Power Guidelines’, September 2016. 22 The Harper proposed section 46(2) refers more broadly to the effects of conduct on efficiency and innovation: Harper Review (n 20). The inter-relationship between the economic models of perfect competition and monopoly respectively and innovation is a matter of some complexity; see, eg, J Church and R Ware, Industrial Organisation (Boston, Irwin McGraw-Hill, 2000) ch 18. While that relationship is outside the scope of this study, the paradigm of ‘zero-sum competition’ advanced in this monograph may well inform our understanding of it. It may be accepted that ‘innovation’ effects can be difficult to determine or predict; see, eg, H Hovenkamp, ‘Schumpeterian Competition and Antitrust’ (2008) University of Iowa College of Law, research paper, 5, ssrn.com/abstract=1275986. In the present study, I focus on ‘efficiency’, a core concern of neoclassical economics, which at least in some respects may be seen to emanate from and thus embrace innovation.

6  Introduction the  Review nor the 2016 exposure draft documents throw any light on how we should resolve differing views in the literature about the objective of competition laws, the standard of efficiency or the assessment of efficiency effects, and how these might be weighed against anticompetitive effects. Judicial experience in Australia is lacking regarding efficiency effects and other defences recognised overseas.23 This is partly a result of the institutional division of functions between the courts on the one hand (to determine breach of section 46 and other competition law provisions), and the ACCC and the Australian Competition Tribunal (ACT) on the other (to determine whether there are offsetting efficiency or other public interest benefits that justify authorising otherwise anticompetitive conduct).24 The ACCC submitted to the Review that formal authorisation by the ACCC/ACT would be the appropriate mechanism to deal with conduct that involves adverse competition effects and offsetting efficiency effects.25 The progress of the Harper Review recommendations regarding misuse of market power has been controversial. The independent review panel of four persons with experience in economics, business and law was appointed on 27  March 2014 with terms of reference to comprehensively review Australia’s competition policy and law.26 The Review consulted widely, circulating an issues paper and draft report before presenting its final report to the Government on 6 March 2015.27 The Government took some time to respond to the report, but on 18 November 2015 issued a response accepting many of the Review’s recommendations.28 With regard to the Review’s recommendations for reform of section 46, the Government decided to consult further with the community by issuing an options paper for reform of section 46.29 Following further consultation, the Government announced on 16 March 2016 that the Harper Review recommendations for reform of section 46 would be accepted.30 As noted above, proposed legislation based on the Harper Review was circulated in exposure draft form in September 2016.31 The Competition and Consumer Amendment (Misuse of Market Power) Bill was first introduced into Parliament in December 2016 and was referred to the Senate Economics Legislation Committee, which delivered its report on the Bill in February 2017.32 In submissions to the Committee, the ACCC argued 23 It should be noted that the ACCC gives qualified recognition to efficiencies in merger cases under the ‘substantial lessening of competition’ (SLC) test as it applies in s 50 (see ACCC Merger Guidelines, November 2008 [7.63–7.66]), though, as discussed further in chs 1 and 3, opinions differ on the matter. 24 See ss 88 and 90 of the CCA. 25 ACCC, ‘Submission to the Competition Policy Review – Response to the Draft Report’, 26 November 2014, 49. 26 Harper Review (n 20) Appendix C, Terms of Reference. 27 ibid 1. 28 Government Response to the Competition Policy Review, 18 November 2015. 29 ibid 2. 30 Joint press release by the Prime Minister, Treasurer and Minister for Small Business, ‘Competition Policy’, 16 March 2016. 31 Exposure Draft (n 21). 32 Senate Economics Legislation Committee, Report on the Competition and Consumer Amendment (Competition Policy Review) Bill 2016, February 2017.

The Proposed Market Manipulation Approach  7 against the Review’s proposed section 46(2) on the grounds that it ‘creates scope for judicial interpretation about the interaction between efficiencies and competition, and innovation and competition’.33 The Committee noted that the ACCC draft Framework document released in September 2016 states that ‘pro-­ competitive conduct such as enhancing efficiency and innovation would not cause [the ACCC] concern under s 46’.34 The Committee concluded accordingly that it is not necessary or desirable to adopt the Review’s proposed section 46(2) and by majority vote recommended it be removed from the Bill.35 As noted above, the amended Bill was subsequently passed by Parliament with section 46(2) removed.36 Thus, the reform process involved the Harper Review recommending, but Parliament rejecting, that reference be made to efficiency in the statutory text. The reform process did not include any detailed consideration of differing views in the literature about the objective(s) of competition laws, the standard of efficiency or the assessment of efficiency effects. Legislative amendments have also been made to implement the Harper Review recommendations that conduct falling within section 46 may be authorised by the ACCC if there is a sufficient public benefit to outweigh any anticompetitive detriment.37 This is not possible under previous law, which excludes conduct falling within section 46 from the authorisation process. However, the consequences for the institutional framework of the Harper section 46(2), which would have required the courts to consider efficiency effects and of Parliament’s rejection of that requirement, remain to be addressed.

III.  The Proposed Market Manipulation Approach I suggest that the legal concept of market power is insufficiently determined and that the ways in which legislatures in Australia and overseas, regulators and businesses, economics experts and the courts have conceptualised ‘market power’ and

33 ACCC Submission to the Senate Economics Legislation Committee, cited in ibid [2.43]. 34 ACCC (n 21) [4.7], cited in Senate Economics Legislation Committee (n 32) [2.44]. The ACCC subsequently issued ‘Interim Guidelines on Misuse of Market Power’, October 2017, which states at [4.2]: ‘Conduct that enhances efficiency, innovation, and product quality or price competitiveness is unlikely to substantially lessen competition.’ This statement is retained in the final version: ACCC, ‘Guidelines on Misuse of Market Power’, August 2018 [4.2]. 35 Senate Economics Legislation Committee (n 32) [2.79]. However, as the draft guidelines indicate that the ACCC’s proposed approach under the Harper recommended s 46, they do not support the Committee’s contention that section 46(2) is unnecessary. 36 For the sake of completeness, it is noted that a private member’s Bill proposed by Senator Gallagher, the Competition and Consumer Legislation Amendment (Small Business Access to Justice) Bill 2017, was tabled on 16 February 2017 and was being considered by Parliament during 2017. The Bill would, among other things, give courts the discretion to relieve small business plaintiffs from the risk of an adverse costs order if they unsuccessfully bring a private action under s 46. The Bill was withdrawn from Parliament on 19 June 2018 and at the time of writing is not proceeding. 37 Competition and Consumer Amendment (Competition Policy Review) Act 2017, sch 9, introduced into Parliament on 30 March 2017, extends the powers of the ACCC under ss 88 and 90.

8  Introduction its use tell us more about the worldviews of those persons than they do about ‘market power’. By interrogating these differing worldviews, we may better understand the concept of market power and more effectively determine the standard, or approach, to distinguish between conduct that can and should be prohibited, and that which is a normal incident of competition (variously described as ‘competition on the merits’ or ‘rational business decisions’). As we will see in Chapter 1, economists may well consider that competition law, in our case embodied in Part IV of the CCA, directly adopts economic principles. With the greatest respect, I suggest that any law must adapt concepts from other disciplines in order to frame rules that are legally and forensically determinate, ie, so that the pertinent facts of the case can be ascertained and the application of the law can be determined by courts according to principle. I suggest that the product of this adaptation is a legal concept.38 As we will see, the concept of an ‘efficient market’ in securities industry laws appears to be a legal construct used by the court to explain and apply market manipulation laws. Australian courts have been at pains to distinguish competition law from economic theory. Of course, one cannot discuss competition law without addressing economic theory. However, my purpose in discussing economics in this monograph is not to propose any new theory of economics, but to critically assess assumptions of the economic models that are not satisfied under real-world conditions. I suggest that such assumptions may explain the conceptual and forensic problems that the law encounters in using or adapting economic principles in the legal theory of liability under the present and proposed versions of section 46 of the CCA. Thus, we will give greater consideration in this monograph to the separate roles and relationship between ‘law’ and ‘economics’ than the usage of the term in the literature on competition law might otherwise suggest. I propose the following positive and normative legal theory of liability for misuse of market power: a. The hypothetical competitive market of neoclassical economics (sometimes called ‘atomistic competition’) does not provide a satisfactory basis for our legal concept of market power because that paradigm of competition is one in which no firm has market power, ie, firms may sell more or less without taking sales away from rivals, and as there is assumed to be free entry and exit, no firm is harmed by competition. The same could be said of alternative hypotheticals, eg, workable competition, which, although not as rigorously theorised as perfect competition, assumes that no firm has market power to any material extent and that normal competition does not necessarily threaten the survival of firms.

38 See, eg, discussion in Director of Public Prosecutions (Cth) v JM (2013) 250 CLR 135, [2013] HCA 30 [38]–[39]. It is for the court to determine the meaning and legal application of technical terms used in legislation that derive from other disciplines.

The Proposed Market Manipulation Approach  9 b. We should base our legal concept of market power, and our standards to determine the use or misuse of market power, on a paradigm of competition that is more reflective of the real world. For the purposes of this study, I define this paradigm as ‘zero-sum competition’. I base this concept on the conventional definition of competition as rivalry for the same object. However, I suggest that conventional thinking has not fully considered the implications of this concept. In legal usage, ‘zero-sum’ simply indicates that one person’s gain is another’s loss. This might not be the way in which economists use the term technically in game theory. In the paradigm of zero-sum competition as I define it, firms compete for sales by taking sales away from each other or by seeking to grow sales faster than the market. All firms to some degree have the capacity to influence the market by price/output decisions, the impact of which depends on how market demand responds. Further, since exit is not free but carries exit costs, incumbent firms cannot easily exit the market and must stay and fight for survival in response to competitive threats such as new entry. I suggest that this paradigm provides a significant insight into market power and its misuse. c. Thus, we see that market power can alternatively be conceived as the power of the market, which cannot be possessed, but can be manipulated. Australia’s section 46 clearly assumes that a dominant firm may ‘have’ market power, but the law prohibits only misuse, not merely ‘having’ market power. As we will see in Chapter 1, although some economists have suggested that section 46 should prohibit a non-dominant firm from strategically ‘creating’ market power, the section has not been judicially interpreted that way. The Harper Review has not recommended any change to the concept of market power. We can see from the US and EU experience that ‘having’ market power is not itself prohibited because market power in the conventional sense can arise from success in business, which ought not to be condemned. I propose that our legal theory of misuse should condemn manipulation of the market. This creates a lower threshold than the conventional structuralist conception of market power, since all firms to some degree have the power to manipulate the market. Accordingly while I use the shorthand expression ‘dominant firm’ to mean a firm having substantial market power, in the context of the market manipulation approach, we must recognise that a ‘non-dominant firm’ in the traditional structuralist sense may well have the power to manipulate the market. The aim of the present study is to identify manipulative conduct that should be condemned. d. We can derive norms of behaviour from the paradigm of zero-sum competition and conventional concepts of economic efficiency to propose that conduct which is efficient can be considered not to be manipulative and should not be condemned. We may thus justify what is called profit-­maximising conduct not because it is rational or logical, but because it may be efficient, ie, in the sense that while we cannot identify the point of optimum social efficiency, we can say that increasing output beyond the profit-maximising

10  Introduction position may well be inefficient both privately for the firm and socially. The legal approach adopted by this reasoning is analogous to the competition test: we cannot identify the optimum amount or state of competition, but our legal theory is based on the assumption that we can identify when competition is substantially lessened. I use the term ‘efficient’ and ‘inefficient’ in this incremental sense. The Harper Review appears to make similar use of the concept in the proposed section 46(2), which refers to effects ‘enhancing’ efficiency.39 To implement this new theory, we need to consider different kinds of efficiency and develop forensic tests. The expression ‘efficiency’ is conventionally understood to refer to social efficiency, and I use it in this sense, unless otherwise indicated. As we will see, conventional components of economic efficiency may well be forensically indeterminate. Our legal theory will therefore require us to adapt economic principles. We will see that the market manipulation approach proposed in this monograph focuses on forensically ascertaining whether market forces are operating without being distorted by the manipulative use of market power.40 We will see, in the context of laws prohibiting manipulation of securities markets, that the courts use the term ‘efficient markets’ in this sense. That is, there is no welfare connotation as there is when economists refer to economic efficiency. e. We need to be particularly careful about the exclusion of equally efficient firms, since the market itself (ie, the undistorted operation of market forces) may exclude such firms where market demand can be satisfied without them. On the other hand, exclusion of a more efficient firm would appear to be manipulative, as we expect more efficient firms to enter and survive. The neoclassical model does not explain how firms self-select for entry to or exclusion from the market: the model assumes that firms have perfect information and may freely enter or leave the market; thus, firms know which producers are more cost-efficient and move efficiently in and out of the market. However, these assumptions are not satisfied under real-world conditions. We will consider further in Chapter 1 how equally efficient firms may self-select for entry/exit under zero-sum competition when what is at stake is survival. Thus, we need to find a means to determine whether equally efficient firms are excluded by the market or by strategies employed by the dominant firm, or possibly, as I argue in this monograph, as a consequence of conduct encouraged by the real-world competitive paradigm. f. We can see that firms locked in zero-sum competition may increase output to inefficient levels, eg, in price wars, or by increasing the price of inputs in

39 Harper Review (n 20) Appendix A, 513–14, setting out the recommended form of s 46 (extracts are set out in the Appendix to this book). 40 Indeed, the ACCC, ‘Interim Guidelines’ (n 34) [1.1] state that misuse of market power ‘undermines the effective operation of markets’ and at [4.4] state that the aim of s 46 is to ‘preserve the integrity of markets’. Those statements are retained in the final version: ACCC, ‘Guidelines’ (n 34) [1.1] and [4.4].

The Proposed Market Manipulation Approach  11 bidding wars reduce consumption to inefficient levels. Such conduct and outcome is a product of the paradigm of competition, and may be tolerated because economic surplus is transferred to consumers, or at least away from firms, and so should not be condemned. Therefore, we should not wrongly attribute inefficiency to the assumed manipulative conduct of dominant firms when it may be the natural consequence of the competitive paradigm. In the present study, we also test the recommendation made by the Harper Review to expressly refer to efficiency in the new section 46. As we will see in Chapter 2, this recommendation challenges over 60 years of economic and legal conventional thinking that suggests efficiency is either legally indeterminate or at least beyond the competence of the courts to assess. The Harper Review uses the term ‘efficiency’ and recommends making reference to efficiency in the text of the proposed law, without defining it or discussing differing views on the subject. Presumably the Review thought this to be unnecessary, since the conventional meaning of efficiency is well understood and differing views are notorious in the literature on competition law. Nevertheless, to give effect to the law recommended by the Review would require the court to judicially determine the ‘efficiency effects’ of conduct and to weigh them against any anticompetitive effect. As we have noted, the ACCC opposed that requirement in the Senate, and it was removed from the law as enacted, apparently because it was thought to be unnecessary. On either view, I suggest that Australia lacks judicial authority on what ‘efficiency’ means and how it could be forensically determined and weighed up by the court, as it appears the Harper Review intended. This monograph will address these issues. I suggest, with respect, that the Harper Review recommendation has much to commend it, but that more thought is required to effectively implement it: i) ‘Economic efficiency’ as conventionally understood to refer to social efficiency (ie, at the level of the whole economy) is legally indeterminate. The expression ‘efficiency’ in the Harper Review’s proposed section 46(2) requires not merely judicial interpretation, but, I suggest, adaptation to be legally and forensically determinate. This monograph proposes how ‘efficiency’ could be adapted, judicially defined and forensically tested under the Harper proposals. Element d) of the market manipulation theory outlined above addresses this question. ii) I suggest that the ‘weighing-up’ process contemplated by section 46(2) of the Harper Review is problematic since ‘competition’ and ‘efficiency’ are unlike concepts. Indeed, competition law uses ‘competition’ as a surrogate for ‘efficiency’. Thus, I suggest that by investigating the Harper Review’s proposed section 46(2), we are led ultimately to accord primacy to efficiency, as it can be adapted to become legally determinate. Thus, we return to the alternative legal theory proposed in this monograph. I suggest that Hovenkamp’s pragmatism, in believing that we can identify the most flagrant cases of misuse of market power, does not provide a satisfactory

12  Introduction answer to the legal theoretical and forensic problems we face with regard to market power. We may draw a parallel here between competition law and the doctrines of equity jurisprudence, which proponents find necessary to defend against charges that equitable doctrines involve ‘discretionary practical justice’ inconsistent with the rule of law.41 Equitable doctrines have, of course, been developed by the judges over more than 400 years, compared to just over 40 years of competition law in Australia. Nevertheless, it is not too early to acknowledge ‘rule of law concerns’ with the jurisprudence on misuse of market power.42 I suggest that the current reform of Australian law has failed to adequately address concerns that have been expressed internationally. Reference is made in this monograph to overseas laws, particularly Europe, and to a lesser extent Canada and the US, to note the international context and particularly issues that have not been addressed by the Australian law reform process. For e­xample, there are notorious differences in approach between the EU and the US in several key areas concerning misuse of market power. However, the focus of the monograph is Australian law. The Australian law under consideration has been newly enacted at the time of writing. The new law discards 40 years of Australian case law developed under the prior version(s) of section 46. See the Appendix to this monograph for a summary of the legislative development of section 46, which has occurred largely in response to judicial decisions which did not coincide with the ACCC’s preferred interpretation of the law. Thus, the task of this study is to consider the legal implications of new laws, and alternative laws proposed by this monograph, rather than to analyse principles developed by the courts under the old law that have now been superseded. Reference is nevertheless made where appropriate to Australian appellate decisions that have been influential in developing the old law. The proposed new laws are tested in Part II by applying them to the facts of leading cases under the old section 46 to see what difference the new laws might make. The case studies have been selected from leading Australian appellate decisions concerning categories of conduct internationally recognised to raise market power issues. Thus, the case studies can be considered in the context of approaches adopted in the US and the EU.

41 WMC Gummow, ‘Equity: Too Successful?’ (2003) 77 Australian Law Journal 30 rebuts those charges. 42 Akman (n 12) 2 states that the ‘uncertainty of the law [in the EU regarding unilateral dominant firm conduct] and the current [2011] state of economics in this area raise important questions of legitimacy’. Akman also suggests (ibid 290) that such questions arise because ‘economics suggests no universal rules at all which can be used to operationalise a prohibition of unilateral conduct’. Similarly, in 2012, Geradin, Layne-Farrar and Petit (n 12) [2.53] stated that differences between economic schools of thought give rise to ‘a serious problem of legal uncertainty’. I take these comments to reflect ‘rule of law concerns’. Witt (n 12) 309 concludes from her 2016 study of the EC’s ‘more economic approach’ that there are serious concerns in the EU for ‘legal certainty and the rule of law’.

Implications of the Study beyond the Present Scope  13

IV.  Implications of the Study beyond the Present Scope I suggest that consideration of the interaction between competition effects and efficiency effects has important implications for competition law generally, beyond the focus of this study on misuse of market power. The attention given by the Harper Review to economic efficiency in the context of section 46 may well lead us to question the focus of competition law on ‘competition effects’ and the institutional division of functions that accords the ACCC/ACT primacy regarding assessment of ‘efficiency effects’. In her 2008 study, Alexandra Merrett considered the existing institutional division of functions in Australia to be generally appropriate.43 However, while the Review did not question the institutional division of functions, it did propose to expressly provide in section 46(2) that the court must consider efficiency effects, which could well be seen as a step in that direction. According to Stephen Corones, the question whether efficiencies are relevant to the analysis of competition effects is one of the most vexed questions in Australian competition law.44 He concludes that, despite some uncertainty, it may be possible to argue that efficiencies are relevant to a limited extent.45 However, the Full Court of the Federal Court of Australia in Universal Music Australia Pty Ltd v ACCC considered that efficiency benefits cannot be balanced against anticompetitive effects in judicial proceedings.46 The ACCC/ACT clearly has the power to do so in authorisation proceedings. As Andrew Gavil notes, opinions differ on the objectives of competition law, which crucially affect the approach to assessing efficiency.47 Whatever the position may have been prior to the Harper recommended section 46(2), it would, if adopted, have been made clear that efficiency is relevant to the assessment of anticompetitive effect under the Harper section 46.

43 A Merrett, ‘The Assessment and Regulation of Market Power in Australia’ (PhD thesis, University of Melbourne, 2010) 278–79. 44 S Corones, Competition Law in Australia, 6th edn (Sydney, Lawbook Co, 2014) 39. 45 ibid 43 and 369–70, citing M Brunt, Economic Essays on Australian and New Zealand Competition Law (The Hague, Kluwer Law International, 2003) 30–35 and 334–36, to suggest that efficiency may be relevant to ‘taking advantage’ of market power under s 46 and the ‘purpose’ element of the substantial lessening of competition test in ss 45 and 47. See also F Hanks and P Williams, ‘The Treatment of Vertical Restraints’ (1987) 15 Australian Business Law Review 147, 150; O’Bryan, ‘Law or Economics?’ (n 15) 64. 46 Universal Music Australia Pty Ltd v ACCC (2003) 201 ALR 636 [270]–[273] (hereinafter ‘Universal Music’). Since the new s 46 as enacted omits the Harper Review’s proposed section 46(2), the ACCC ‘Interim Guidelines’ and ‘Guidelines’ (n 34) do not address the possibility of balancing enhanced efficiency effects against anticompetitive effects. ACCC, ‘Guidelines’ (n 34) [2.22] states that: ‘A firm’s commercial rationale may be relevant to understanding the conduct in question and assessing its purpose and/or effect on competition. However, it will not amount to a defence.’ 47 A Gavil, ‘Burden of Proof in U.S. Antitrust Law’ (2008) 1 Issues in Competition Law and Policy 125, 153. See also L Griggs and E Sharp, ‘The Homogeneity of Object in Part IV – Consumer Sovereignty as the Answer?’ (2003) 11 Competition and Consumer Law Journal 1.

14  Introduction To the extent that the matter under prior law is open to doubt, the parliamentary rejection of the Harper section 46(2) may well strengthen that doubt. The supplementary Explanatory Memorandum states that, in removing Harper section 46(2), ‘the new s 46 continues to focus on harm to the competitive process, rather than individual competitors, and does not shield inefficient competitors from the natural effects of strong competition’.48 Whereas the Review clearly proposed to modify the SLC test in the context of section 46, Parliament has evinced a different intention, and ultimately the courts will determine the meaning and effect of the new section 46 as enacted. Those recommendations would not alter the SLC test in other provisions of Part IV (which is outside the scope of this study), but may well influence our thinking on that topic. This study concerns dominant firm conduct; however, the literature on another area of competition law touches on the concept of market power – the prohibition of anticompetitive mergers (which is outside the scope of the study). While the specific legal test is whether the merger would be likely to substantially lessen competition, the underlying enquiry is often expressed to be whether the merger is likely to create or enhance market power.49 Again, the analysis of competition effects in the case of mergers has tended to be crude.50 Nevertheless, the literature provides further examples of the diversity of worldviews concerning market power. Refinement of our understanding of market power may well affect our thinking about mergers. We may infer that the ACCC’s intention in opposing the Harper Review section 46(2) is to preserve the current institutional division of functions, which gives the ACCC/ACT primacy in assessing efficiency effects. While it is of course a matter of interest to determine the effect of Parliament’s rejection of the Harper section  46(2), the focus of this monograph is the role that economic efficiency would have under the Harper section 46(2) and the alternative legal conceptualisation of market power as ‘market manipulation’ proposed in this monograph. This monograph concerns liability for contravention of section 46 of the CCA, which is punishable by the court awarding a pecuniary penalty payable to the Commonwealth under section 76 of the CCA. Liability is determined according to the civil standard of proof, not the criminal standard. There are, of course, a range of other remedies. However, this monograph does not address remedies, except insofar as the courts have found it problematic to apply the hypothetical

48 Supplementary Explanatory Memorandum, 28 March 2017, 9. 49 Hovenkamp (n 18) 210. 50 See, eg, ibid 210; G Sidak and D Teece, ‘Dynamic Competition in Antitrust Law’ (2009) Journal of Competition Law and Economics, ssrn.com/abstract=1479874 advocate reform of the US merger guidelines, subsequently addressed to some extent in the 2010 guidelines. see also R Feinstein, ‘2010 Revisions to the US Horizontal Merger Guidelines’ (2011) 7 Competition Law International 6; US Department of Justice and Federal Trade Commission, ‘Horizontal Merger Guidelines’, 19 August 2010; cf Australian Competition and Consumer Commission, ‘Merger Guidelines’, November 2008.

The Problematic Role of Efficiency  15 competitive market standard in assessing quantitatively the harm that is said to arise from conduct in question. The same principles of liability apply to a private action which section 82 of the CCA confers on any person who suffers loss or damage arising from a contravention of the CCA. Again, however, this monograph does not address remedies, such as the measure of damages, under a private action. Some of the leading cases discussed in Part II concern private actions but do not address the question of damages. This monograph focuses on the market power of firms as sellers. I do not discuss the majority decision of the Full Federal Court in ACCC v Australian Safeway Stores Ply Ltd.51 The case concerns the market power of the dominant firm as buyer. The application of the principles proposed in this monograph to ‘buyer power’ is beyond the scope of the present study.

V.  The Problematic Role of Efficiency Conventional wisdom in Australia has long held that, quite apart from the institutional division of functions under the CCA, consideration of economic efficiency is not a matter upon which courts are competent to embark, having regard to the skills and experience of judges and the constitutional authority of courts to determine disputed matters of fact and law.52 It is generally accepted that authorisation proceedings call for administrative (rather than judicial) decision-making, which may require weighing up and trading off competing objectives, eg, whether anticompetitive effects may be offset by efficiency benefits. This study reconsiders the competence of the courts with regard to ‘efficiency’ as the concept can be adapted for use in my proposed legal theory. We need to consider whether determining ‘efficiency’ is no less a fact-finding exercise than determining ‘competition effects’ and, if so, how it may forensically be determined. We need to consider the international debate concerning the possibility of weighing up competition effects against efficiency effects. I argue in this monograph that efficiency effects (as a higher-order objective of competition law) should prevail. The Harper Review approached reform of section 46 as primarily an Australian problem of redrafting the provision by legislative amendment to facilitate enforcement, informed to some degree by international comparisons, primarily Canada. The Review did not reconsider the legal concept of market

51 ACCC v Australian Safeway Stores Ply Ltd [2003] FCAFC 149 (Heerey and Sackville JJ; Emmett J dissenting). 52 See, eg, French J (prior to his appointment as Chief Justice of the High Court of Australia) in Australian Gas Light Company v ACCC (No 3) [2003] FCA 1525 [607], citing his own extra-judicial paper ‘The Role of the Courts in the Development of Australian Trade Practices Law’ in F Hanks and P Williams (eds) The Trade Practices Act: A Twenty-Five Year Stocktake (Sydney, Federation Press, 2001) 108.

16  Introduction power itself or engage with the international debate on unilateral dominant firm conduct, and so is unlikely to quell the international controversy about the misuse of market power. This study addresses those issues and contributes to that debate. Just as the Review did not engage with the long-running debate in Australia for and against an ‘effects’ test, this study passes over that debate to focus on an analysis of the consequences of adopting the Harper section 46 or the alternative legal conceptualisation of market power as ‘market manipulation’ proposed in this monograph. There are trenchantly opposed views about the purposes to be served by competition laws and the extent to which such purposes may guide the interpretation and application of provisions regulating dominant firm conduct.53 One view is that productive efficiency, ie, maximising production and income from society’s scarce resources (total welfare), should be the goal.54 This view is supported by influential US thinkers55 and is endorsed by the ACT in the Qantas decision.56 However, in this decision, the Tribunal was considering the meaning of ‘public benefit’ for the purpose of a merger authorisation and cited only one author in support.57 The ACCC opposed that interpretation58 and may still do so. The opposing view is that allocative efficiency, ie, maximising consumer choice and satisfaction, should be the goal. On this view, the purpose of competition law is to protect consumers (and small business) from big business.59 The US Antitrust Modernisation Commission in 2007 considered this debate, but did not attempt to resolve it because, in the Commission’s view, in many cases the application of the law would not be affected by differing views about the policy objective.60 However, as Gavil notes, the treatment of efficiencies as a possible defence to anticompetitive effects raises some issues about the applicable o ­ bjective,61 which will come to the fore when we consider the application of the Harper section 46(2).

53 See, eg, J Kirkwood, ‘The Essence of Antitrust: Protecting Consumers and Small Suppliers from Anticompetitive Conduct’ (2013) 81 Fordham Law Review 2425; see also J Wright, ‘The Antitrust/ Consumer Protection Paradox: Two Policies at War with Each Other’ (2012) 121 Yale Law Journal 2216; M Stucke, ‘Reconsidering Antitrust’s Goals’ (2012) 53 Boston College Law Review 551. 54 See, eg, R Bork, The Antitrust Paradox: A Policy at War with itself (New York, Perseus Books, 1978), quoted in D Crane and H Hovenkamp (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013) 408; Department of Justice (US), ‘Competition and Monopoly: Single-Firm Conduct under Section 2 of the Sherman Act’ (2008) 7. 55 H Hovenkamp, ‘Roger Blair and the Goals of Antitrust’ (2016) 61 Antitrust Bulletin 382. 56 Australian Competition Tribunal, Re Qantas Airways Ltd [2004] ACompT 9 [185] (hereinafter ‘Qantas’). 57 ibid [170]; MS Gal, Competition Policy for Small Market Economies (Cambridge, MA, Harvard University Press, 2003) 203–05. 58 Qantas (n 55) [169]. The Full Federal Court in ACCC v ACT [2017] FCFCA 150 [62]–[68], an appeal from the Australian Competition Tribunal in the matter of the Tabcorp merger, seemingly approves of the view taken by the Tribunal in Qantas. I will discuss this further in ch 1. 59 See, eg, Kirkwood (n 53) 2429. 60 Antitrust Modernisation Commission, Report and Recommendations (2007) 26, fn 22. 61 Gavil (n 47) 153–55.

The Problematic Role of Efficiency  17 This debate is exemplified by the generally accepted proposition, reiterated by the Harper Review, that competition law aims to protect not competitors, but the process of competition.62 On the other hand, when rejecting the Review’s proposed requirement to take efficiency effects into account, the Senate ­Economics ­Legislation Committee expressed the view that ‘the current section 46 has not provided adequate protection for non-dominant firms from the destructive actions of firms with substantial market power’.63 This does not sit easily with the statement in the supplementary Explanatory Memorandum quoted above.64 Robert French has noted the history of previous amendments to section 46 proposed by a Senate Committee in 2004 ‘for the purpose, not of protecting competition, but of protecting small business’ and suggests that this is an example of amendments ‘in response to particular political imperatives which will … generate confusion about [a statutory provision’s] policy objectives’.65 The legislative response to the Review’s recommended changes to section 46 may well fall into that category. The Harper Review refers to ‘consumer welfare’ without clarifying whether this refers to allocative efficiency (consumer welfare) or productive efficiency (total welfare) as the objective of competition law.66 As Herbert Hovenkamp notes, the term ‘consumer welfare’ is often used ambiguously in the US to refer in fact to total welfare.67 He also observes that in the EU, there has been a shift to focus on the narrow concept of consumer welfare.68 I suggest that the debate about policy objectives seems to have lost sight of what can and cannot be achieved through the competitive market mechanism. Policy choices should be informed by an understanding of the consequences, which may be elucidated by positive economic analysis and may be tested against legal ideals such as certainty and­ predictability.69 That is, we must ask questions such as the following: • ‘Is the objective achievable?’ • ‘If we pursue this objective, are we able to tolerate any adverse consequences?’

62 Harper Review (n 20) 339. 63 Senate Economics Legislation Committee (n 32) [2.75]. 64 Supplementary Explanatory Memorandum (n 48). 65 R French, ‘Dolores Umbridge and the Concept of Policy as Legal Magic’ [2007] Federal Judicial Scholarship 14 [16]–[17]. 66 See, eg, references to the ‘long-term interests of consumers’ and to competition resulting in reduced prices for consumers (Harper Review (n 20) 16), which may be inferred to refer to total welfare, though references to ensuring access and equity for vulnerable consumers (Harper Review (n 20) 20) may suggest a commitment to intervention in markets for reasons of distributional equity. 67 Hovenkamp (n 55) 383. 68 ibid 390. 69 Stucke (n 52) 574–75 calls these ‘rule of law concerns’. See also, eg, L Solum, ‘Indeterminacy’ in D Patterson (ed), A Companion to Philosophy of Law and Legal Theory (Oxford, Wiley-Blackwell, 2010) 479–92. Compare the Harper Review (n 20) 27, where it is said that: ‘Laws that are less predictable in their immediate application may nevertheless prove more reliable over time as they are adapted through the judicial process to encompass novel developments.’ We will see that experience with s 46 may not encourage such an expectation.

18  Introduction • ‘Does the hypothetical competitive market (ie, a hypothetical market in which no firm has appreciable market power) provide a standard of conduct to guide business decisions?’ • ‘Is such a law sufficiently certain and predictable to engender community respect for the law and legal process?’

VI.  Differing Views of the World The influence of ‘worldview’ has long been associated with the policy debate on competition law, and opposing worldviews are easily identified. Hovenkamp describes the opposing worldviews as: (a) on the one hand, confidence in the ability of markets to ‘self-correct’ (ie market power creates excess profits which encourage entry of new firms thus returning profits to normal levels) together with a corresponding lack of confidence in government agencies to intervene effectively; and (b) on the other hand, a diametrically opposed lack of confidence in markets and belief in the ability of regulators to intervene effectively.70 Alan Devlin and Michael Jacobs discern a greater confidence in Europe than the US in the ability of regulators to intervene effectively.71 Again, it should be possible, by positive economic analysis, to test the consequences of competition laws and accordingly demonstrate whether proposed laws are capable of achieving those objectives. That the unresolved debate is relevant to the regulation of market power has been previously demonstrated by the release in 2008 and subsequent withdrawal in 2009 of the US Department of Justice’s report on dominant firm conduct.72 In announcing the Department’s decision to withdraw the report, Assistant Attorney-General Christine Varney states that she does not agree with the view implicit in the report that markets are generally self-correcting, and in her view enforcement must proceed on the basis that the market alone cannot be relied on to ensure that competition and consumers will be protected.73 I suggest that

70 H Hovenkamp, ‘Post-Chicago Antitrust: A Review and Critique’ (2001) Columbia Business Review, reproduced in D Crane and H Hovenkamp (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013) 482. 71 A Devlin and M Jacobs, ‘Antitrust Divergence and the Limits of Economics’ (2010) 104 Northwestern University Law Review 270; see also J Vickers ‘Competition Law and Economics: A Mid-Atlantic Viewpoint’ (2007) 3 European Competition Journal 1, 6. 72 Department of Justice (US) (n 54). 73 Department of Justice (US), ‘Justice Department Withdraws Report on Antitrust Monopoly Law, Press Release 11 May 2009’.

Economic Models and the Paradigm of Competition  19 neither ­worldview is adequate to elucidate competition law or its application to dominant firm conduct, and it is necessary to interrogate these opposing worldviews in order to progress the debate. There is a critical distinction between removing impediments to the market mechanism to facilitate efficient markets and intervening in the market to ‘correct’ perceived undesirable market outcomes. We need to be realistic about what the market can achieve, and to critically consider when the term ‘market failure’ is used and whether in fact it reflects a normative view about how markets should operate, as opposed to how they do operate and what they can achieve. This study provides a better understanding of the legislative and judicial response to the regulation of market power in Australia and overseas, and it is hoped that it may contribute to the future development of the law and international harmonisation. The central problem is to elucidate the nature of ‘market power’, our purpose or purposes in regulating it, and the consequences when we focus on economic efficiency (as we would under the Harper Review section 46(2) or the alternative conceptualisation of market power proposed in this monograph of ‘market manipulation’). This study suggests developments of the Australian and international competition law framework with a view to achieving greater clarity on the distinction between what is permitted and what is proscribed. As Hedvig Schmidt observes, there is a diversity of national approaches, which ironically are based on the same underlying concept of market power.74 Thus, the legal conception of market power addressed in this monograph is of universal significance for competition laws around the world.

VII.  Economic Models and the Paradigm of Competition Our legal concept of market power, which has not been questioned by the Harper Review, derives from a spectrum on which the neoclassical economic models of perfect competition and monopoly are at opposite ends. Although ‘workable competition’ falls in between, it cannot be rigorously theorised or defined as can the opposite ends of the spectrum. As an undergraduate student of economics and law in the 1970s, I found it strange that, despite the unreality of the hypothetical perfectly competitive market of neoclassical economics and the lack of any obvious constraint on the long-run growth of firms under that model,75

74 Schmidt, ‘Market Power’ (n 12) 391–92. 75 The only limitation on the size of firms under that model applies in the short run due to the assumption that costs rise according to the ‘law of diminishing returns’ when the availability of one or more factors of production is constrained, which no longer applies in the long run. See, eg, R Marchionatti, ‘On the Methodological Foundations of Modern Microeconomics: Frank Knight and the “Cost Controversy” in the 1920s’ (2003) 35 History of Political Economy 49.

20  Introduction it informs the paradigm of competition that underpins Australian competition law and policy. At the same time, there was a satisfying elegance in the SCP approach, which despite those misgivings provided the unquestioned foundation for assessing competition effects in Australia.76 This continued to be the case through the 1980s (and it remains influential today in Australian competition law textbooks),77 despite being discredited in the US in the 1970s (as we will see in Chapter 2). In my experience, economics students in those days appeared to be more sceptical about the policy implications of economic models than lawyers and judges applying the new competition laws. Three principal reasons for this scepticism have been subsequently noted in the literature. First, the models of perfect competition and monopoly demonstrate optimality under such an unrealistic set of assumptions that the models cannot be used to explain the real world.78 Richard Posner does not consider this ‘lack of realism’ to be a problem, however, his argument that economic theory nevertheless has ‘predictive power’ is insufficiently developed and does not appear to be ­convincing.79 The Full Court of the Federal Court of Australia in ACCC v Australia and New Zealand Banking Group Ltd (2015)80 were not specifically addressing the models of perfect competition and monopoly; however, the following observation of the court captures the problem raised by unrealistic assumptions: Economists frequently construct economic models to analyse complex commercial or economic events or scenarios. But a model is unlikely to be a useful analytical tool if based on unrealistic assumptions that materially depart from the real world facts and circumstances involving commercial behaviour in which the events to be analysed occur.

It could be argued that such models, rather than being predictive, simply set out conditions (assumptions) under which certain outcomes would be possible. Thus, when the outcome is not achieved under real-world conditions, we should consider whether this may simply be the natural consequence of the assumptions not being satisfied. I find it surprising, therefore, that Joseph Schumpeter’s analysis of the inadequacies of the model of perfect competition as a benchmark for economic efficiency and a basis for competition law tends to be regarded

76 Staple texts used in Australia in the 1970s and 1980s were R Caves, American Industry: Structure Conduct and Performance, 2nd edn (Englewood Cliffs, Prentice Hall, 1967); and R Caves, I Ward, P Williams and C Wright, Australian Industry: Structure Conduct and Performance, 2nd edn (Sydney, Prentice Hall Australia, 1987). 77 See, eg, Corones (n 44) 22–26 and 46–48; J Duns and A Duke, Competition Law Cases and Materials, 4th edn (Chatswood, LexisNexis Butterworths, 2015), 37–38. 78 See, eg, R Lipsey, An Introduction to Positive Economics, 2nd edn (London, Weidenfeld & Nicolson, 1963) 375–79; R Lipsey and A Chrystal, An Introduction to Positive Economics, 8th edn (Oxford, Oxford University Press, 1995) 407–08; A Eichner and JA Kregel, ‘An Essay on Post-Keynesian Theory’ (1975) 13 Journal of Economic Literature 1293, 1310. 79 R Posner, Economic Analysis of Law, 7th edn (New York, Aspen Publishers, 2007) 16. 80 ACCC v Australia and New Zealand Banking Group Ltd [2015] FCAFC 103; 236 FCR 78; 324 ALR 392 [138] (Allsop CJ, Davies and Wigney JJ).

Economic Models and the Paradigm of Competition  21 as unconvincing (in part due to the macroeconomic focus of his analysis and in part because his argument lacks the mathematical rigour of the neoclassical model).81 It has long been accepted that when economists compare monopoly to perfect competition, the outcome of the comparison depends on the assumptions you make.82 In other words, we cannot predict by a process of a priori reasoning whether perfect c­ ompetition or monopoly will be the more efficient in any given case.83 Such a conclusion requires an empirical assessment. As we will see in Chapter 1, identification of the ‘competitive level’ of price and output, and the analysis of efficiency that is based on it, relies on assumptions that are not satisfied in real-world markets. This gives rise to forensic issues which, as we will see in Part II, have led judges to reject opinion evidence of economists attempting to determine the hypothetical ‘competitive level’. Second, according to Lipsey and Lancaster’s ‘theory of second-best’, strategies indicated by a model to achieve optimality cannot be reliably predicted to achieve it when all the assumptions are not satisfied, ie, in the real world we cannot predict whether such strategies will make things better or worse.84 Therefore, we should not be surprised to find that zero-sum competition does not necessarily produce efficient outcomes and may in fact produce inefficient outcomes. Third, the economic concepts of opportunity cost and marginal cost do not correspond with management or financial accounting measures that businesses use to make decisions and report to owners. Economists argue that this does not affect the predictive value of economic theory.85 Even if that is accepted, Landes and Posner describe the economic concept of ‘marginal cost’ as a ‘hypothetical construct’, and there appears to be no doubt that it is extremely difficult to measure empirically.86 The Australian cases on section 46 considered in Part II bear this out. My experience as a practising lawyer in the 1980s attempting to apply the SCP approach added to the misgivings noted above. Business clients often lack the data required to fully assess SCP factors when considering competition law

81 J Schumpeter, Capitalism, Socialism and Democracy (London, George Allen & Unwin, 1976) 106. See, eg, criticism by Hovenkamp (n 22); FM Scherer, ‘Joseph Schumpeter’ in H de Jong and W Shepherd (eds), Pioneers of Industrial Organisation (Cheltenham, Edward Elgar, 2007) 211–13; and a defence by D Evans and K Hylton, ‘The Lawful Acquisition and Exercise of Monopoly Power and its Implications for the Objectives of Antitrust’ (2008). 4 Competition Policy International 203, ssrn.com/ abstract=1275431. 82 Lipsey (n 78) 373. 83 See, eg, EAG Robinson, Monopoly (Cambridge, Cambridge University Press, 1941) 140–42; this is another staple text used in Australia in the 1970s, cited by Dawson J in Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd (1989) 167 CLR 177, 201 (hereinafter ‘Queensland Wire’). 84 See R Markovits ‘Second-Best Theory in Law and Economics’ (1998) Chicago-Kent Law Review 73, discussing RG Lipsey and K Lancaster, ‘The General Theory of Second Best’ (1956) 24 Review of Economic Studies 11. 85 Lipsey (n 78) 394. 86 W Landes and R Posner, ‘Market Power in Antitrust Cases’ (1981) 94 Harvard Law Review 937, 941; Hovenkamp (n 18) 163.

22  Introduction issues.87 Further, the assessment of SCP factors is in practice highly impressionistic. For this reason, as is well known, opinions on market definition and hence market power can reasonably differ.88 Thus, the Federal Court has stated that the concept of a ‘market’ is not a directly observable feature of the real world, but an analytic construct devised by economists.89 While economic models assume for the sake of simplicity that products are homogeneous, in reality they are not. Consequently, it is difficult to discern when substitution of products within a market ends and substitution with products outside the market begins.90 Experience also suggests that business persons, lawyers and judges, economists and regulators bring different perspectives to the analysis of business decisions. These perspectives may be influenced by differing institutional frameworks for evaluating rational business decisions, as well as differing views on the hypothetical competitive market standard, or paradigm of competition, that is implicit in our competition law. While there is no authoritative appellate decision on the point, observations of justices in the High Court and the Federal Court suggest that some justices regard ‘perfect competition’, or at least a hypothetical competitive market in which no firm has appreciable market power, as the relevant benchmark.91

VIII.  The Overseas Experience The last decade has seen controversial inquiries in Europe and North A ­ merica regarding ‘unilateral dominant firm conduct’, ie, misuse of market power. 87 Geradin, Layne-Farrar and Petit (n 12) [2.08], writing in 2012 state that ‘it is rarely the case that sufficient data are available to reach a foolproof conclusion’ regarding economic effects. 88 Queensland Wire (n 81) 187–88 (Mason CJ and Wilson J): ‘too narrow a description of the market will create the appearance of more market power than in fact exists; too broad a description will create the appearance of less market power than there is.’ And at 195–96 (Deane J): ‘The identification of relevant markets and the definition of market structures and boundaries … involves value judgments about which there is room for legitimate differences of opinion.’ 89 Seven Network Ltd v News Ltd (2009) 182 FCR 160, 350–51 (Dowsett and Lander JJ); cited in ACCC v Australia and New Zealand Banking Group Ltd (2015) 236 FCR 78 [136] and ACCC v Pfizer Australia Pty Ltd [2018] FCAFC 78 [265]. 90 Indeed, R Trindade and R Smith, ‘Modernising Australian Merger Analysis’ (2007) 35 Australian Business Law Review 358, 362–64 argue that too much emphasis on market definition can lead us to overlook important constraints on market power that come from outside the market. 91 The following are the only references I can find to perfect competition in appellate judgments of the Federal Court and High Court: the High Court in Rural Press Ltd v ACCC (2003) 216 CLR 53, ALR 217 (hereinafter ‘Rural Press’) [49] (Gummow, Hayne and Heydon JJ), noting with apparent approval the Full Federal Court’s reference to ‘perfect competition’ as the relevant benchmark; the High Court in ACCC v Flight Centre Travel Group Ltd [2016] HCA 49 [141] (Nettle J), observing that the Sherman Act’s ‘conception of competition appears to be less “atomistic” than that which applies under the Trade Practices Act [now the CCA]’ (Nettle J cited Illinois Corporate Travel Inc v American Airlines Inc 806 F 2d 722 (7th Cir 1986), 727, which relevantly states: ‘The antitrust laws do not adopt a model of atomistic competition …’); compare the High Court in Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1 (hereinafter ‘Melway’) [52] (Gleeson CJ, Gummow, Hayne and Callinan JJ), where the majority observe that nothing in s 46 assists us in identifying the ‘nature and structure of the competitive market’ and that an absence of market power does not require ‘perfect competition’, but merely a ‘sufficient level of competition’.

The Overseas Experience  23 In Europe, the European Commission (EC) chose to issue guidance on enforcement priorities, rather than legislative amendment, to promote a more economic approach instead of a formalistic approach.92 In the US, as a result of differing opinions, neither legislative amendment nor regulatory guidance was adopted, leaving further development of the law to judicial decisions.93 There is little evidence that Australian thinking has benefited from overseas controversies, to which this study will refer in more detail in Chapter 3. In Australia it seems that our courts are more influenced by close analysis of the statutory text than they may be by developments in economic theory. It may well be that Australian courts will be less likely than administrative bodies such as the ACCC and the ACT to depart from the presumed 1970s legislative intent to favour more atomistic competition, unless there was compelling evidence of contrary legislative intent in the amended legislation. The reasoning of the Harper Review is not abundantly clear on this issue. It will, however, be argued in this monograph that the real underlying problem identified by the Review and others, with the current requirement that the dominant firm ‘take advantage’ of its market power, arises from the implicit endorsement this appears to give to the hypothetical perfectly competitive market (or a hypothetical competitive market in which no firm has appreciable market power) as the relevant paradigm of competition. Having dispensed with that requirement, it will be left to the courts to determine what degree of causal nexus (if any) is required between market power, conduct and competition effects, especially if it is posited that the same conduct by a firm that lacks market power may not affect competition. Under the current competition test, the Full Court of the Federal Court of Australia has held that market power is irrelevant to the application of the SLC test.94 The context was exclusive dealing in contravention of section 47, and the court reasoned that section 47 is not drafted on the premise that market power could be relevant.95 In the context of the Harper section 46, it seems that market power is relevant to the SLC test, so this issue is open for reconsideration by the court. It is argued in this monograph that some causal nexus should be required under the Harper section 46. The paradigm of competition continues to be implicated in the Harper Review section 46: the concept of market power derives from the hypothetical perfectly competitive market; and economic analysis of efficiency is usually explained by reference to that hypothetical. I suggest that the Review is correct to direct our attention in the proposed section 46(2) to efficiency; however, the failure to address the legal concept of market power itself is likely to ensure that past problems recur. Further, the legislative response to the Review, by removing the proposed section  46(2), may well exacerbate those problems. Regrettably, the Australian

92 European Commission, ‘Guidance on the Commission’s Enforcement Priorities in Applying ­Article 82 EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings’ [2009] OJ C45/7. 93 Antitrust Modernisation Commission (n 60). 94 Universal Music (2003) 201 ALR 636 [268]. 95 ibid.

24  Introduction reform debate leaves us none the wiser regarding how we should resolve differing views in the international literature about the objective of competition laws, or the standard of efficiency or its forensic assessment. I suggest that the market manipulation approach should satisfactorily resolve past problems which many attribute to lack of positive and normative standards to assess dominant firm conduct.

IX.  What Difference Will Market Manipulation Theory Make? Table 0.1 below sets out a schematic outline of the legal conceptual and forensic differences of the market manipulation approach in comparison with other legal theories. Following the table is an illustrative example comparing the application of the different legal theories. These are intended only as an illustrative guide to the detailed discussions and legal argument in the ensuing pages of this monograph. In the table, ‘yes’ or ‘no’ indicates whether the criterion or element is adopted by the relevant legal theory; grey highlighted elements indicate the legal theory of market manipulation advanced in this monograph. Table 0.1  Comparison of legal theories of liability for misuse of market power Criteria or elements

Old section 46

New section 46

Harper section 46 with sub-section (2)

Market manipulation

(1) Theory of harm Social welfare standard (economic efficiency across the whole economy, conventionally productive, allocative and dynamic efficiency)

Legally indeterminate

Surrogate welfare No test (substantial lessening of the process of competition (SLC) in a market)

Legally Legally indeterminate indeterminate

Legally indeterminate

Yes

Insufficiently legally determined

Yes

(continued)

What Difference Will Market Manipulation Theory Make?  25 Table 0.1  (Continued) Criteria or elements

Old section 46

New section 46

Amount and allocation of producer/ consumer surplus

Hypothetical ‘competitive level’ of price/ output

Hypothetical ‘competitive level’ of price/ output

Firm’s profitmaximising price/output

Harper section 46 with sub-section (2)

Market manipulation

Private welfare standard (an industry or market) Hypothetical ‘competitive level’ of price/ output

Legally indeterminate

Yes, if achieved Yes, if on merit achieved on merit

Yes, if achieved on merit

Yes, as starting point

No

No

No

Yes

Hypothetical competitive market (no structural market power)

Yes

Yes

Yes

Legally indeterminate

Zero-sum (or survival) competition

Inchoate

Inchoate

Inchoate

Yes

Yes

Yes

Insufficiently legally determined

Efficient market standard (undistorted forces of demand and supply) Positive and normative standards

(2) Theory of liability Legal concept of market power Structural Yes economic theory (compared to hypothetical competitive market)

(continued)

26  Introduction Table 0.1  (Continued) Criteria or elements

Old section 46

Strategic No (an alternative economic theory) Market No manipulation (distorting forces of demand and supply) Threshold for market power Is creating market power prohibited? Is having and profiting from market power prohibited? Prohibited use of market power Substantial subjective exclusionary purpose Substantial subjective purpose of SLC Effect of SLC Predicted effect of SLC Purpose and effect of distorting forces of demand and supply Attempt to contravene

New section 46

Harper section 46 with sub-section (2)

Market manipulation

No

No

Not needed (see next)

No

No

Yes

Substantial

Substantial

Substantial

No threshold

No

No

No

Not an issue

No

No

No

No

Yes

No

No

No

No

Yes

Yes

No

No No

Yes Yes

Yes Yes

No No

No

No

No

Yes

Yes

Yes

Yes

Yes (continued)

What Difference Will Market Manipulation Theory Make?  27 Table 0.1  (Continued) Criteria or elements Dual purposes problem in zerosum competition (rivalry for same object harms loser)

Old section 46 Condemns substantial subjective exclusionary purpose

New section 46

Harper section 46 with sub-section (2)

Market manipulation

SLC test condemns self-interested conduct (effect without purpose)

SLC test condemns self-interested conduct (effect without purpose)

Condemns purpose and effect (no effect without purpose)

No

No

Yes

Taking Yes advantage of market power

No

No

No

Causal nexus required between power and effect

(Taking advantage covers)

Unclear

Unclear

Yes

Rational business decision

Probably

Only for purpose

Only for purpose

Not sufficient

Efficiency may offset harm to competition

Unclear

Unclear

Yes (untried)

No (efficiency and competition are unlike concepts)

No

No

Yes

Mere selfYes interested conduct should not be condemned Causation

Defences

Efficiency No (forensic tests below) to prevail

(continued)

28  Introduction Table 0.1  (Continued) Criteria or elements

Old section 46

New section 46

Harper section 46 with sub-section (2)

Market manipulation

Forensic tests for efficiency Undistorted forces of demand and supply, eg, unmet demand?

N/A

N/A

N/A

Yes

Private efficiency (revenue and profit effects; relative efficiency of firms)

No

No

No

Yes

Hypothetical ‘competitive level’ of price and output

Yes

Yes

Yes

Legally indeterminate

Efficient market; private efficiency

No

No

No

Yes

Qualitative

Qualitative

Qualitative

Qualitative

(3) Measure of harm Benchmark for quantification of harm

Assessment of penalty

This monograph examines the application of current and proposed legal theories to several different kinds of conduct internationally recognised to involve market power issues. The following example concerns only one scenario. For more detailed discussion, see Part II. Let us assume that there is a firm which is structurally dominant in a market. Supply and demand are in balance. A new entrant commences operations, hoping to win market share from the incumbent and that demand will respond positively to some initial price discounting. The incumbent responds by cutting prices and increasing its output. Demand does not respond positively and supply exceeds

What Difference Will Market Manipulation Theory Make?  29 demand. Prices fall to below avoidable cost. The new entrant fails financially. Prices in the market subsequently trend upward towards pre-entry prices. We can analyse the scenario under each of the legal theories summarised in Table 0.1: a) Old (or pre-Harper Review) section 46: market power would be conventionally assessed on largely structural factors. The court would assess whether a firm having substantial market power took advantage of that power for a substantial subjective exclusionary purpose. Possibly the conduct might not be impugned if the conduct was rational profit-making conduct. The outcome will depend on evidence of subjective purpose. If an incumbent floods the market without rational explanation, it could be held to have an exclusionary purpose. The dominant firm appears to have continued to expand output and cut prices even though revenue and profits fell. It is likely that the court would find an exclusionary purpose. However, this finding could go either way. It requires detailed examination of the evidence of the incumbent firm’s executives and depends crucially on the impression they make upon the trial judge. b) New section 46 as enacted: assessment of market power is unchanged. Now the enquiry concerns the purpose and effect of SLC. The purpose is still assessed as substantial subjective purpose. While the relevant purpose is different under the SLC test, it would appear on the facts that the dominant firm’s purpose cannot be justified by rational self-interest, so it would probably be liable under the purpose limb. Again, the trial judge’s impression of the witness evidence will be crucial. The effect of SLC is assessed on largely structural factors. It is a predictive test, but in this scenario we know the new entrant fails. It would seem that the process of competition is harmed by the new entrant leaving the market, though harm to a competitor is not necessarily harm to the competitive process. Even if the court finds there to be no substantial subjective purpose of SLC the incumbent can be held liable if there is a likely effect of SLC. c) Harper Review section 46 as recommended, ie, requiring consideration of efficiency: it seems that the incumbent priced below avoidable cost, so would have harmed an equally efficient rival. If the new entrant was more efficient, that would add weight to a finding of SLC effects. Perhaps if the incumbent made cost savings, ie, became more efficient to compete more effectively, that might weigh against a finding of SLC effects. The Harper section 46 gives no guidance as to how a court might weigh up these considerations. d) Market manipulation: all firms have the power to manipulate the market to varying degrees. The threshold of substantial structural power is irrelevant. Has either firm manipulated the market? We assess whether the market is efficient, ie, whether forces of supply and demand are operating undistorted. There does not appear to have been unmet demand when the new entrant commenced operations. We assess this by reference to whether the dominant

30  Introduction firm is meeting all demand that can profitably be supplied (ie, not by reference to the forensically indeterminate ‘competitive level’). Thus, we consider whether there are shortages or delays in delivery and whether the dominant firm may be capacity-constrained, eg, by allocating resources to higher value uses, it may not be able to meet all demand for the product. We consider how demand responded to new entry. It seems that industry output has expanded, but the extra revenue has not covered the extra costs, so would appear to be an inefficient use of society’s resources. We cannot measure the impact on social welfare, but equally we cannot assume it is neutral. How has demand for the incumbent firm’s product responded to its strategy? If the firm’s output expansion has been profitable, that would appear to be privately efficient. If that is not the case, the conduct is seemingly inefficient and could well be manipulative. But is this inefficiency the result of firms competing for the same object (zero-sum competition), possibly exacerbated by buyer power when supply exceeds demand? We consider the exit costs of the incumbent and whether these explain pricing below avoidable cost. We look at the relative efficiency of the incumbent and the new entrant, as it appears the Harper Review intended. If a more efficient new entrant has been driven out, that would suggest the market has been manipulated. But both firms it seems have been driven to inefficient conduct, ie, producing more than the market appears to want. We consider whether the size of the market or the strategies of the incumbent or the new entrant’s miscalculation of the risk of entry caused the new entrant’s failure. We might well conclude that there has been inefficient conduct, but that it results from the paradigm of zero-sum competition and consequently that there has been no manipulation, ie, no distortion of market forces of demand and supply.

X.  Application to the Digital Economy This study addresses Australian appellate court decisions concerning market power in conventional products. While there have been significant developments in the EU regarding digital platforms, inquiries in Australia at the time of writing are at an early stage.96 I suggest that the study in this monograph of theories

96 On 27 June 2017, the EC announced it had fined Google €2.42 billion for abusing its market dominance as a search engine by promoting its own comparison shopping service in its search results: press release, 27 June 2017, europa.eu/rapid/press-release_IP-17-1784_en.htm. There are controversial differences between the US and the EU in the assessment of this matter: see, eg, Kemp, Misuse of Market Power (n 12) 5–6. On 18 July 2018, the EC announced it had fined Google €4.34 billion for abusing its market dominance by placing restrictions on Android device manufacturers and network operators to cement the dominance of its search engine: press release, 18 July 2018, europa.eu/rapid/ press-release_IP-18-4581_en.htm. At the time of writing, the ACCC is investigating digital platform providers such as Facebook and Google to ascertain whether these platforms are exercising market

Outline of the Chapters  31 of liability for misuse of market power could potentially be applied to the digital economy. However, that task is beyond the scope of the present monograph. The paradigm of zero-sum competition proposed in this monograph is closely related to the idea of ‘competition for the market’, which is noted in the literature to be a feature of markets in the digital economy.97 Such markets apply to conventional physical products which can be marketed and sold over digital platforms, as well as products which once had a physical form, but now can be reproduced in digital form, and intangible services provided in digital form. The common features with zero-sum competition are that firms compete for the market, ie, the success of one is achieved at the expense of the other(s), and since firms invest heavily in technology, with consequently high exit costs, they cannot freely leave the market, so must compete for survival.98 The rise and fall of the BlackBerry company provides a powerful example of this kind of competition.99 Of course, examples of Schumpeterian dynamic competition, such as pharmaceuticals, pre-date digital technologies. We may thus readily accept that concepts of market power and misuse of market power derived from conventional products should be adaptable to the digital economy.100 However, I respectfully suggest that difficulties that have been encountered in the application of the law to markets for conventional products should preferably be addressed first.

XI.  Outline of the Chapters Chapter 1 (‘Market Power in Economics and Law’) discusses the core concepts of market power and efficiency, and aspects of the neoclassical positive and normative model that do not fit with the real world and so must be re-engineered if we are to understand market power and regulate its use. Descriptions of the competitive paradigm by economists and some judges in the 1990s suggest a shift from atomistic competition to zero-sum competition, yet the current observations of

power in commercial dealings to the detriment of consumers and users of their services. The investigation began in December 2017 and is expected to be completed by December 2018: ACCC, Digital Platforms Inquiry, www.accc.gov.au/focus-areas/inquiries/digital-platforms-inquiry. 97 See, eg, S Bishop and M Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, 3rd edn (London, Sweet & Maxwell, 2002) 36–37; Akman (n 12) 23; Rosario et al (n 12) 39; Schmidt, ‘Taming’ (n 12) 9. 98 Bishop and Walker (ibid 37) and Akman (n 12) cite Schumpeter (n 79) 84. While Schumpeter does not refer to competition for the market, it is clear that when he refers to firms competing ‘for their very lives’, he is alluding to survival competition. 99 See, eg, V Savov, ‘BlackBerry’s Success Led to its Failure’, The Verge, 30 September 2016, www. theverge.com/2016/9/30/13119924/blackberry-failure-success. 100 Thus, Schmidt, ‘Taming’ (n 12) 2 argues that the current legal concept of market power is flexible enough to adapt to the digital economy. I note that Schmidt (at 16) describes the conduct of Google for which it was fined by the EC in 2016 as ‘manipulating the market’ (though this appears to be no more than a colloquial description of conduct prohibited in the EU as abuse of dominance).

32  Introduction other judges suggest that the legislative intent continues unchanged in order to support more atomistic competition. The implications for our positive and normative legal theory of market power need to be considered. In this chapter we develop the concept of zero-sum competition, the proposed legal concept of market power as the power to manipulate the market and begin to derive norms of conduct that may be used to distinguish permitted from prohibited conduct. We review economic and legal analyses in the literature that relate market power to demand response, and draw upon the concept of market manipulation in securities markets in order to propose an alternative legal concept. We conduct a ‘thought experiment’ into how we would expect an incumbent dominant firm to respond to new entry if the paradigm of competition were zero-sum competition. This provokes inquiry into rational business strategies, and how lawyers and economists may differ regarding rational strategies under game theoretical models as a result of their differing frames of reference or views of the world (the legal system being essentially a zero-sum game). Our focus on these strategies is from the perspective of efficiency, ie, I first critique the assumptions underlying the ‘competitive level’ to remind us that it is known in the literature and Australian case law to be conceptually and forensically problematic, and hence I argue that it is not an appropriate benchmark to assess conduct in real-world markets. I then develop points noted in the literature to argue further that if we assess the efficiency effects of moving not from that hypothetical, but from the monopolist’s actual profit-maximising position, we can see that business decisions may be justified. This is not because they are logical to business persons (or rational to an economist), but because they may be economically efficient, ie, in the sense that while we cannot forensically identify the point of optimum social efficiency, we can say that increasing output beyond the profit-maximising position could well be inefficient both privately for the firm and socially. This approach builds on the conventional view that a monopolist is not to be deprived of the fruits of success, which leads to suggestions in the literature that the ‘competitive level’, where such profits are dissipated, cannot be the appropriate benchmark to assess dominant firm conduct. These views will be discussed in relation to leading Australian cases covered in Part  II. This monograph takes the argument further to propose that we should consider the efficiency implications of moving from the dominant firm’s actual position, rather than by comparison with the hypothetical competitive market equilibrium. Thus, we can generalise from the discussion of ‘monopoly’ in the literature and cases to consider the position of a dominant firm. Efficiency in the conventional sense is probably legally indeterminate. I leave further consideration of efficiency to Chapter 3, where I suggest how it may be adapted for use in my proposed legal theory, and how it may be forensically determined. Chapter 2 (‘The Mischief and Australia’s Institutional Response’) reviews changing views in the US, the EU and Australia concerning the SCP approach by which competition effects have been inferred and predicted from market

Outline of the Chapters  33 s­ tructure and changes in market structure. In this chapter we address the basis in the 1960s discourse in Australia regarding the institutional division of functions in ­administering competition law, which suggests the courts are ill-equipped to address and weigh up competition and offsetting efficiency effects. The Harper Review proposed section 46(2), if it had been passed by Parliament, would have clearly brought efficiency within the purview of the courts. Absent that modification to section 46, the ACCC/ACT remains the only institution having the unambiguous power to weigh up competition and offsetting efficiency effects. I suggest that it is undesirable for questions, which may be raised in judicial proceedings, to be potentially answered differently from the same questions when raised in administrative proceedings before the ACCC/ACT. I suggest that issues of efficiency (adapted and redefined for use in the proposed legal theory) should be no less capable of determination by the courts than competition effects under the existing SLC test. Chapter 3 (‘Competition and Efficiency Effects in Europe, North America and Australia’) reviews and critiques market structure and direct approaches to assessing the impact on the process of competition (ie, competition effects). We contrast these approaches with the focus of the market manipulation approach on effects on the market. We draw upon the concept of an efficient (undistorted) market from laws prohibiting manipulation in securities markets to propose how we can adapt the concept of efficiency for our legal theory in a way that can be forensically tested in Part II. The Harper Review’s proposed section 46 is discussed, particularly the idea that conduct, which might be innocuous if engaged in by a firm without market power, may have an anticompetitive effect if engaged in by a dominant firm. It is unclear whether and if so how it is to be shown that market power contributes to such an effect. While market power is not strictly relevant to the Australian SLC test, the presumption of the Harper section 46 is that market power may well exacerbate the effect. I suggest that we can solve the problem of weighing up competition and efficiency effects, which are unlike concepts, by giving primacy to efficiency, as a higher-order objective of competition law. This chapter develops approaches to adapting and redefining our legal concept of efficiency, and forensically assessing it, which can be tested in the following chapters. Part II (‘Testing Market Manipulation and Efficiency Approaches’) analyses the application of the Harper Review section 46 to leading Australian appellate cases, in contrast to the proposed market manipulation approach. The areas of business conduct considered in Chapters 4–9 are: refusal to deal and margin squeezes; predatory pricing; meeting competition; raising rivals’ costs; bundling; and, finally, institutional and procedural aspects, ie, current institutional limitations on the court’s power to consider efficiency, and the burden of proof. The case studies do not address exclusive dealing as this is prohibited under section  47, subject to the SLC test. Possibly for this reason, we lack appellate

34  Introduction ­ ecisions concerning loyalty rebates as a potential misuse of market power;101 d hence, the case studies do not address that topic. In this Part we see that firms use competition law for strategic reasons, as part of the framework in which competition occurs. By focusing on efficiency and adapting it to our legal theory, we can better assess the economic rather than private implications of competition. In this Part we see that zero-sum competition does not necessarily produce efficient outcomes and may in fact produce inefficient outcomes. For example, price wars and bidding wars for inputs are a product of the paradigm of competition, and may be tolerated because economic surplus is transferred to consumers, or at least away from dominant firms. This has contributed to what I call the ‘bipolarity’ of competition law, ie, our law endorses the paradigm of zero-sum competition, but then seeks to condemn harm that may well be caused or contributed to by the real-world competitive paradigm itself. Part III (‘Conclusion’) sets out the study’s conclusion that the market manipulation approach more effectually addresses the theory of harm underlying the concept of market power. I suggest that the Harper Review section 46 (particularly following Parliament’s removal of section 46(2), which would have mandated consideration of efficiency effects) may tend to facilitate the application of section 46 to large firms merely because they are large firms. In this way, section 46 may make it possible for competition law to be used towards an end that may well be unattainable, ie, to use markets and competition to constrain the power of big business and counter the trend to more concentrated markets. I suggest that there is a real chance that the Harper section 46 will expose large firms to the risk of potential liability for conduct which is efficient (ie, does not distort the operation of markets) or which may be a consequence of the competitive paradigm itself. Some recommendations are made as to how section 46 could be modified to ameliorate that risk. Nevertheless, that remains a second-best solution. I suggest that the market manipulation approach is to be preferred and that ‘efficiency’ can be adapted to provide the centrepiece of our positive and normative legal theory of market power. The study also has implications beyond the scope of section 46 which could inform further studies, eg, regarding the role of efficiency in the SLC test, and the institutional division of functions between the courts and the ACCC/ ACT, more generally in competition law.

XII.  Effective Date of Law The law is relevantly stated as at 31 July 2018. The ‘time of writing’ referred to in the monograph is late August 2018. 101 ACCC (n 21) [4.5] does not mention ‘loyalty rebates’ as an example of types of conduct that may involve a contravention of s 46; however, subsequent versions of that guidance document do refer to loyalty rebates as an example, ie, ACCC ‘Interim Guidelines’ (n 34) [3.2] and [3.11]–[3.13], which are repeated in the final ACCC ‘Guidelines’ (n 34). Geradin, Layne-Farrar and Petit (n 12) [4.159] state in relation to EU case law on the subject that ‘the assessment of rebates is one of the most complex and unsettled areas of competition law’.

Effective Date of Law  35 This monograph therefore covers Intel Corporation Inc v EC102 and the longawaited decision of the Full Court of the Federal Court of Australia delivered on 25 May 2018 in ACCC v Pfizer Australia Pty Ltd.103 Pfizer is an ‘originator’ of patented medications. In the few months before the expiration of a patent on one of Pfizer’s ‘blockbuster’ drugs, it sought to launch its own generic version of the product and to bundle it with supply of the patented product to gain a market presence ahead of anticipated entry by manufacturers of generics. The court dismissed the ACCC’s appeal, so the ACCC’s claim against Pfizer under section 46 failed. Importantly, the court at [559] recognises, in circumstances akin to zero-sum (or survival) competition, that a dominant firm may permissibly use its market power to protect its own sales, without it necessarily following that the firm has a substantial subjective purpose of excluding competitors or substantially lessening competition. Thus, the case may well inform our understanding of how the SLC test might apply under the Harper Review section 46. The case also invites us to reconsider the competition analysis of the majority of the court in ACCC v Baxter Healthcare Pty Ltd,104 which is discussed in Part II.

102 Case C 413/14 P Intel Corporation Inc v EC [2017] (CJEU, 6 September 2017, rectified 19 ­September and 24 October 2017) (hereinafter ‘Intel’). 103 ACCC v Pfizer Australia Pty Ltd [2018] FCAFC 78 (hereinafter ‘Pfizer’). On 19 October 2018, the High Court of Australia dismissed the application by the ACCC for special leave to appeal the decision in Pfizer. 104 ACCC v Baxter Healthcare Pty Ltd [2008] FCAFC 141.

36

part i The Theoretical Framework

38

1 Market Power in Economics and Law I. Introduction It appears that despite the attention given to ‘law and economics’ in the US and elsewhere since the 1970s, key questions remain unsettled. Australian thinking about market power appears largely untouched by these developments. Thus, our conception of market power is that traditionally derived from US thinking of the 1950s and is defined as the ‘ability to give less and charge more’ or the ‘­ability to behave differently than a competitive market would enforce’. For ­example, in its expert’s report in the 2014 merger authorisation case Re  ­Application for A ­ uthorisation of Acquisition of Macquarie Generation by AGL Energy Ltd,1 ­Frontier Economics uses the first definition, which dates from 1955 and was adopted by the ACT in Re Queensland Co-operative Milling Association Ltd.2 The second definition, which dates from 1959, has also been adopted by the High Court in Queensland Wire Industries Pty Ltd v The Broken Hill Proprietary Company Ltd.3 Alternative economic theories of market power, eg, ‘strategic market power’, have been canvassed in the literature in Australia, but have not been adopted by the courts.4 Since the Harper Review did not raise any issue with the conventional conception of market power, the traditional definitions remain unquestioned in Australia. This monograph does not enter into

1 Re Application for Authorisation of Acquisition of Macquarie Generation by AGL Energy Ltd [2014] ACompT 1 (hereinafter ‘AGL/MacGen’). 2 Re Queensland Co-operative Milling Association Ltd (1976) 8 ALR 481 (hereinafter ‘QCMA’); Frontier Economics, ‘Competition Issues’ 26 March 2014, Expert’s Report prepared for AGL, filed in the Australian Competition Tribunal, AGL/MacGen (n 1) [6], referring to the US Attorney-General’s National Committee to Study the Antitrust Law, Report of 1955. 3 Queensland Wire Industries Pty Ltd v The Broken Hill Proprietary Company Ltd (1989) 167 CLR 177 (hereinafter ‘Queensland Wire’), 200 (Dawson J), citing C Kaysen and D Turner, Antitrust Policy (Cambridge, MA, Harvard University Press, 1959) 75. 4 A Niblett, J Gans and S King, ‘Structural and Behavioural Market Power under the Trade Practices Act: An Application to Predatory Pricing’ (2004) 32 Australian Business Law Review 83 argue that firms which might not have market power in the conventional structuralist sense may nevertheless have strategic power to create it, which they argue should be captured by s 46 and that it is open to the courts to so interpret the section. For a more detailed discussion of ‘strategic market power’, which has not been judicially accepted in Australia, see: A Merrett, ‘The Assessment and Regulation of Market Power in Australia’ (PhD thesis, University of Melbourne, 2010); S Corones, Competition Law in Australia, 6th edn (Sydney, Lawbook Co, 2014) 49 concluded that ‘there do not appear to be any signs that the strategic behaviour approach is likely to be adopted by the courts in Australia for the foreseeable future’.

40  Market Power in Economics and Law the debate concerning alternative economic theories. In any event, since there is no threshold for the power to manipulate the market, ie, all firms have it to some degree, the market ­manipulation approach proposed in this monograph may well address the concerns raised by proponents of alternative economic theories of ‘strategic market power’. The first edition of Richard Posner’s book Economic Analysis of Law5 was published only shortly before the enactment of the Trade Practices Act 1974 (Cth) and so did not enter into Australian thinking at the time. In a 1981 article, William Landes and Richard Posner argue that a key determinant of market power is the responsiveness of demand to price changes (‘elasticity of demand’, which is not directly measurable) and suggest improvements to US judicial approaches of inferring market power from market shares.6 Landes and Posner approach the issue as one of evidence and proof. Frontier Economics states that elasticity can be estimated if there is historical data available.7 This is a significant proviso, which has so far proved problematic in Australia for the purposes of litigation8 and for studies of the economic effects of, eg, bundling in the petrol and grocery markets.9 Roger Blair and Celeste Carruthers in 2010 published a similar study to Landes and Posner, showing that elasticity of demand has a key role to play in the analysis of market power.10 However, I suggest that the implications for the legal ­conceptualisation of ‘market power’ have not been fully explored. US thinking since the 1970s regarding ‘law and economics’ does not appear to have made a significant impact on Australian jurisprudence. Indeed, former Chief Justice Anthony Mason has expressed the view that economic theory has a limited role in the judicial application of statute law in Australia, ie, ­determining

5 R Posner, Economic Analysis of Law, 7th edn (New York, Aspen Publishers, 2007). 6 W Landes and R Posner, ‘Market Power in Antitrust Cases’ (1981) 94 Harvard Law Review 937, 941–43 and 979. 7 Frontier Economics (n 2) [61]. 8 See, eg, Australian Gas Light Company v ACCC [2003] FCA 1525 [494]–[569] (French J), rejecting econometric evidence concerning elasticity of the residual demand curve in a merger case. 9 See, eg, Z Wang, ‘Supermarkets and Gasoline: An Empirical Study of Bundled Discounts’ (2011) 5th Annual Conference on Empirical Legal Studies, ssrn.com/abstract=1628770. In 2013, the ACCC reached voluntary settlements with supermarket chains regarding ‘shopper docket’ loyalty arrangements. The ACCC required the supermarkets to cap the petrol price discounts they offered on grocery purchases: ACCC v Coles Group Ltd [2014] FCA 363; ACCC v Woolworths Ltd [2014] FCA 364. The litigation concerned the interpretation of the enforceable undertakings, and the potential application of provisions of pt IV of the CCA has not been adjudicated. Accordingly, these matters are not addressed in this monograph. See further, eg, B Kobayashi, ‘The Economics of Loyalty Discounts and Antitrust Law in the United States’ (2005) George Mason University, ssrn.com./abstract_id=794944; J Gans and S  King ‘Paying for Loyalty: Product Bundling in ­Oligopoly’ (2006) 54 Journal of Industrial Economics 43; J  Wright ‘Simple But Wrong or Complex and More Accurate? The Case for an Exclusive Dealing Approach to Evaluating Loyalty Discounts’ (2013) Bates White 10th Antitrust Conference, Washington DC; A  Merrett, ‘Are Shopper Dockets ­Anti-competitive? Good Luck Proving it’, The Conversation, 5 ­September 2013, theconversation.com/ are-shopper-dockets-anti-competitive-good-luck-proving-it-17499. 10 R Blair and C Carruthers, ‘The Economics of Monopoly Power in Antitrust’ in K Hylton (ed), Antitrust Law and Economics (Cheltenham, Edward Elgar, 2010) 64.

Introduction  41 rights and liabilities of parties by reference to past events.11 Mason considers that a deep understanding of economics is not necessary to interpret competition legislation, though an understanding of economics may be of assistance in the application of the legislation, which requires consideration of economic data, expert opinion and economic analysis.12 With respect, I suggest that Mason is correct: any law must adapt concepts from other disciplines in order to frame rules that are legally and forensically determinate, ie, so that the application of the law to the facts of the case can be determined by courts according to principle. Thus, the outcome should be certain and predictable. Outcomes that are matters on which opinions can reasonably differ could be perceived to be capricious and thus undermine confidence in the rule of law. However, some economists may well consider that the law directly adopts economic principles. The Hilmer Report of the early 1990s noted ‘a degree of dissatisfaction with the current court procedures for the utilisation of economic material’13 and considered it worthwhile that further consideration be given to measures to address the situation, including ‘arrangements for increasing the specialisation of judges involved in competition matters’.14 The High Court in Melway stated that economic analysis may be consistent with the purposes of section 46 ‘if it can be undertaken with sufficient cogency’ and observed that ‘in some cases, a process of inference, based upon economic analysis, may be unnecessary’.15 The Antitrust Modernisation Commission in 2007 said much the same thing when it suggested that ‘brightline legal rules’ that business can easily follow are to be preferred over rules that depend on the often-conflicting views of economists.16 However, we do not yet have ‘bright-line rules’. The ACT in Qantas Airways Ltd expressed some dissatisfaction with economics experts who see their role as advocates for an outcome rather than as truly independent experts informing the tribunal about the applicable economic principles.17 We will see in Part II of this monograph that the 11 A Mason, ‘Law and Economics’ (1991) 17 Monash University Law Review 167, 173; see also M Trebilcock, ‘The Value and Limits of Law and Economics’ in M Richardson and G Hadfield (eds), The Second Wave of Law and Economics (Sydney, Federation Press, 1999) 12. Compare A Fels, ‘The Competition Review’, speech at the University of Melbourne, 19 May 2014 (submitted to the Harper Review, 25 June 2014) 4, who argues that the CCA is an ‘economic statute’ and that the purpose test in s 46 has been wrong in principle since 1976. 12 Mason (n 11) 178. 13 Commonwealth of Australia, National Competition Policy (Canberra, Australian Government Publishing Service, 1993) (hereinafter ‘Hilmer Report’) 170–71. 14 ibid 178. See also Fels (n 11) 3 and 6, who submitted to the Harper Review that courts make ‘heavy going’ of s 46 cases and that enforcement is difficult for the ACCC, in part because ‘well heeled’ defendants tenaciously resist cases. We saw in the Introduction that the ACCC opposed the Harper s 46(2) before the Senate subcommittee because ‘efficiency’ was thought to provide too much scope for judicial interpretation: ACCC Submission to the Senate Economics Legislation Committee, cited in the Committee Report, Report on the Competition and Consumer Amendment (Competition Policy Review) Bill 2016, February 2017 [2.43]. 15 The majority in the High Court in Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1 (hereinafter ‘Melway’), 24; 178 ALR 253, 266. 16 Antitrust Modernisation Commission, Report and Recommendations (2007) 87. 17 Re Qantas Airways Ltd [2004] ACompT 9 [227].

42  Market Power in Economics and Law courts are often faced with conflicting opinions proffered by economic experts, which ultimately are of little assistance in deciding contested cases. Judicial attitudes towards economics, or more particularly the opinion evidence of economists, may demonstrate McMahon’s point that judicial decisions mask underlying worldviews.18 The opposed perspectives of lawyers and economists are probably based on mutual misunderstanding. Neither law nor economics is a value-free science: both require a framework, or view of the world, within which to make sense of both the positive and the normative. Of particular interest to this study is the way in which legal and policy thinking has used economics. FM Scherer suggests that the choices made by regulators and courts in reading ‘what economics has to say; that is, on which among conflicting propositions they have placed emphasis and which ones they have downplayed … depend importantly upon the values [of] the decision-makers’,19 ie, their view of the world. By interrogating such choices, we may elucidate underlying worldviews. This is not merely a matter of identifying such worldviews, which Robert French considers to include the ‘vision of a better world’ achieved through ‘social engineering of commercial behaviour’.20 Thus, when economic models set out assumptions (which are not satisfied in the real world) under which certain outcomes would be possible, we could infer that the models express a view about how the world should be, ie, a normative view. The approach in this monograph is to critically analyse the assumptions of the models to see if an alternative set of assumptions assists us in developing norms more relevant to real-world conditions. I suggest that the paradigm of zero-sum competition as defined in this monograph embodies assumptions more reflective of the real world. We also need to consider whether competition law can achieve the ‘better world’ to which some theorists aspire (and, if not, we need to look for other more effective policy levers that would have to be engaged in order to pursue that objective should the community wish to do so). Accordingly, this study will consider how competition law has used neoclassical economic models in developing the legal theory of liability, in our case set out in section 46 of the CCA. However, this study does not critique the models themselves or consider in detail the comparative merits of other approaches,

18 K McMahon, ‘Competition Law, Adjudication and the High Court’ (2006) 30 Melbourne University Law Review 782. 19 FM Scherer, ‘Conservative Economics and Antitrust: A Variety of Influences’ (2008), originally published in R Pitofsky (ed), How the Chicago School Overshot the Mark (Oxford, Oxford University Press, 2008), reprinted in D Crane and H Hovenkamp (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013) 469. In the present study, I critically assess how the law ‘uses’ economics. By contrast, in her 2008 study, Merrett ((n 4) Appendix C) simply notes conventional economic analyses of monopoly and efficiency. I suggest that we need to go further to bridge the gulf that appears to exist between law and economics. 20 R French, ‘The Role of the Courts in the Development of Australian Trade Practices Law’ in F Hanks and P Williams (eds), Trade Practices Act: A Twenty-Five Year Stocktake (Sydney, Federation Press, 2001).

Policy Objectives and Concepts of Efficiency  43 eg, the Schumpeterian dynamic approach. However, we will see that zero-sum competition, and its implications for efficiency, may elucidate Schumpeter’s concept of ‘creative destruction’ and its application to dominant firm conduct.21 Tony Freyer argues that firms adopt a variety of strategies to ameliorate the risk of creative destruction, from diversification through mergers to various anticompetitive strategies.22 I suggest that we need to find a benchmark for conduct that is not determined by the assumptions of a model, and that economic efficiency could be adapted to provide a legal benchmark. An economic model’s assumptions simplify the real world, and if we are to understand the implications of the model for the real world, the assumptions must be investigated (eg, by selectively relaxing assumptions, the effects on the model can be better understood). If we treat a model’s assumptions as an acceptable approximation to the real world, such assumptions may become a ‘blind spot’ in our thinking. A key blind spot, or gap, in the neoclassical models that will be pursued in this study is how firms selfselect for entry to or exclusion from the market. Before turning to this, with a thought experiment regarding incumbent response to new entry, it is useful to briefly review the controversy about the objectives of competition law, which will be relevant to our analysis of efficiency and its relationship to misuse of market power. Differing views about the objectives of competition law affect the extent to which efficiency may be regarded as a defence to misuse of market power.23 My purpose in this monograph, and in particular in this chapter, is not to propose any new theory of economics, but to critically assess conceptual and forensic problems that the law encounters in ‘using’ or adapting economic principles in the legal theory of liability in present and proposed versions of section 46 of the CCA. This discussion is used as a point of departure to develop a new legal theory, not any new economic theory.

II.  Policy Objectives and Concepts of Efficiency Hovenkamp argues that the policy objective of antitrust is to ‘make the economy work better’ and to do this, one must not only be able to identify ‘wrongs’, but also to produce ‘suitable remedies’.24 By the latter we may infer that he means remedies

21 J Schumpeter, Capitalism, Socialism and Democracy (London, George Allen & Unwin, 1976) ch VII. See further, eg, S Bishop and M Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, 3rd edn (London, Sweet and Maxwell, 2002) 36–37. 22 T Freyer, Antitrust and Global Capitalism, 1930–2004 (Cambridge, Cambridge University Press, 2006) 394–95. 23 See A Gavil, ‘Burden of Proof in U.S. Antitrust Law’ (2008) 1 Issues in Competition Law and Policy 125, 153. See also L Griggs and E Sharp, ‘The Homogeneity of Object in Part IV: Consumer Sovereignty as the Answer? (2003) 11 Competition and Consumer Law Journal 1, 17–20. 24 H Hovenkamp, Antitrust Enterprise: Principle and Execution (Cambridge, MA, Harvard University Press, 2005) 110.

44  Market Power in Economics and Law which improve the working of the economy. Stating this policy objective another way, Hovenkamp argues that ‘intervention is justified only in the relatively few cases where the judiciary can fix the problem more reliably, more cheaply, or more quickly than the market can fix itself ’.25 As will be apparent in the discussion of his analysis below, it is doubtful in many cases that the ‘wrong’ and ‘remedy’ are sufficiently determined in economics and law. Hovenkamp describes capitalism as a free market ideology, ie, it adopts the thesis that ‘the uncontrolled, apparently chaotic, and completely self-interested behaviour of businesspersons actually produces more welfare than governmental command and control’.26 It is right to describe this as a thesis (or view of the world) because it has not been (and probably cannot be) demonstrated by empirical testing. Hovenkamp states that economic theory does not give us ‘normative ideas about what the optimal price and output should be’.27 Despite this, he states that ‘the indicia of antitrust harm’ are ‘reduced output and higher prices in a ­properly defined relevant market’.28 In this monograph we will leave aside problems of market definition. I suggest that the key point, which has been made in the Introduction to this volume, is that the concept of a market is an analytic construct  – that is, a market (and hence market power) is not directly observable, but is a matter of analysis and opinion. If Hovenkamp’s earlier statement is correct, it would not be possible to establish whether higher prices depart from or approach the indeterminate optimum. Hovenkamp goes some way towards acknowledging this when he says that in order to provide a remedy, the court would have to determine the ‘correct’ price.29 The decision of the High Court in Queensland Wire has been criticised for evading this issue.30 However, the statement does demonstrate that reduced output and higher prices are merely surrogates for the public harm to economic efficiency. This brings us to the economic analysis of monopoly. Hovenkamp concludes that, because a monopolist curtails output to the level that maximises prices and profits, ‘consumers are either poorer or else they make inefficient substitutions’, ie, they choose products ‘that they would regard as “inferior” if their first choice were competitively priced’.31 In any market situation, there will be a body of consumers who choose not to buy because the price does not yield them the benefits they desire. For the reasons set out below, I suggest that we cannot conclude that they necessarily suffer a loss because of their choice.

25 ibid 124. 26 ibid 15. 27 ibid. 28 ibid 101. 29 ibid 102. 30 See eg, W Pengilley, ‘Queensland Wire and its Progeny Decisions: How Competent are the Courts to Determine Supply Prices and Trading Conditions?’ (1991) 21 Western Australian Law Review 225; cf F Hanks and P Williams, ‘Implications of the Decision of the High Court in Queensland Wire’ (1990) 17 Melbourne University Law Review 437, 460–61. 31 Hovenkamp (n 24) 19.

Policy Objectives and Concepts of Efficiency  45 I suggest that it is impossible to generalise and conclude that the alternative choices are inefficient. Blair and Carruthers discuss consumer and producer surplus under market equilibrium: consumer surplus is the benefit to consumers who would have paid more than the equilibrium price; producer surplus is the benefit to producers who would have supplied at less than the equilibrium price.32 In the model of perfect competition, the equilibrium price tends towards the producer’s break-even point (ie, it covers normal profit). When we critically examine the assumptions of the model which are required for this outcome, we find no conclusive explanation in principle for above-normal profits (other than that the assumptions are not satisfied under real-world conditions).33 In other words, it cannot be inferred that market power is being misused, and so empirical analysis is required. For example, there may be shortages of inputs, some firms may be more successful due to innovation, economies of scale etc and thus may be more efficient than others, with the result that above-normal profits may be earned.34 Nevertheless, the way in which we expect supply to respond to demand, and our analysis of consumer surplus, in real-world markets is typically based on the assumptions of the economic models. Thus, it is said that at prices below the break-even point, producers will make a loss and will stop producing if prices do not cover their variable costs.35 The supply curve at prices above the equilibrium indicates the rising marginal costs of firms already in the market (entry or exit of firms would move the short-run supply curve to the right or left respectively). However, it is important to note that the ‘supply curve’ can be defined in this way only under circumstances of perfect competition (when firms’ marginal cost curves can be horizontally added to derive it, because the firmfacing, or residual, demand curve is horizontal). Thus, no such supply curve can be defined in a monopoly36 or oligopoly, because supply increases until marginal cost equals marginal revenue (not price, ie, average revenue). The market supply curve under which supply responds to price (average revenue) is a critical construct in the supply and demand model of a perfectly competitive market.

32 Blair and Carruthers (n 10) 66. 33 Assumptions include: all markets in the economy outside the one in question are perfectly competitive; firms may freely enter and exit the market; all firms in the market have the same costs; in the long run, there are no constraints on inputs (factors of production), so there are no fixed costs. See eg, R  Lipsey, An Introduction to Positive Economics, 2nd edn (London, Weidenfeld & Nicolson, 1963) 354; J Perloff, R Smith and D Round, Microeconomics (Frenchs Forest, Pearson Australia, 2014) 243; J Gans, S King, M Byford and NG Mankiw, Principles of Microeconomics, 7th edn (South Melbourne, Cengage Learning, 2018) 317–18, Figure 14.8, which demonstrates that under the assumptions of the model the long run supply curve is horizontal, ie, supply adjusts to demand by firms entering or leaving the market. 34 See, eg, RH Bork and JG Sidak, ‘The Misuse of Profit Margins to Infer Market Power’ (2013) 9 Journal of Competition Law and Economics 511, 513–14 and 521–22. 35 P Dooley, Elementary Price Theory (New York, Appleton-Century-Crofts, 1967) 70–71. 36 R Lipsey and A Chrystal, An Introduction to Positive Economics, 8th edn (Oxford, Oxford ­University Press, 1995) 241.

46  Market Power in Economics and Law We ­therefore cannot use traditional demand and supply analysis to explain conduct of firms in r­ eal-world markets where the assumptions of the economic model are not ­satisfied.37 I suggest, with respect, that it was this that led Kirby J into error in Boral,38 which will be discussed further in Chapters 3 and 5. In the present chapter we will begin to consider how markets work when the familiar market mechanism of perfect competition (where both supply and demand respond to price) does not operate. Continuing our discussion of producer surplus, economists do not apparently take into account the ‘detriment’ to producers who are excluded from the market simply because at the equilibrium price, the market is not large enough for more producers. Similarly, in assessing consumer surplus, no account is taken of the ‘detriment’ to consumers who are excluded from the market because they are unwilling to pay at least the equilibrium price. There could be several reasons for this. First, it may be presumed that the economy would be better served by equally or less efficient firms applying their resources elsewhere, and it may also be presumed that consumers who consider themselves better off by not buying would be better served by spending their money elsewhere. Second, the purpose of the market is to ration scarce goods: there must be a sufficient benefit to both producers and consumers to use this mechanism. The model of perfect competition demonstrates an elegant community of interest if all markets in an economy are at equilibrium: (a) consumers who are prepared to pay get the quantities and mix of goods and services they desire (a feature of allocative efficiency); and (b) ­producers in each market produce the volume that minimises costs and provides them with normal profits (a feature of productive efficiency). Some argue that the sum of producer surplus and consumer surplus indicates the welfare outcomes from a market.39 However, these benefits are internal to the market, sometimes described as the gains from trade, ie, without mutual gains to producers and consumers, there would be no trade and no market. It is inaccurate to describe these surpluses as indicators of social welfare outcomes because producers and consumers excluded from the market apply their resources elsewhere. Thus, unless we assume that such movements of resources to or from other markets are welfare-neutral, the sum of producer surplus and consumer surplus in a market is suggestive only of private welfare of those producers and­ consumers.40 There is argument for and against an objective of competition law being to ensure an equitable distribution of surplus between producers and

37 See, eg, Perloff, Smith and Round (n 33) 39. Other models discussed by Perloff, Smith and Round (ibid ch 13), such as Cournot and Bertrand games, are likewise based on assumptions, many of which are not satisfied in the real world; these are discussed further below. 38 Boral Besser Masonry Ltd v ACCC (2003) 215 CLR 374 (hereinafter ‘Boral’). 39 See, eg, J Kirkwood, ‘The Essence of Antitrust: Protecting Consumers and Small Suppliers from Anticompetitive Conduct’ (2013) 81 Fordham Law Review 2425, 2453–54. 40 See, eg, M Stucke, ‘Reconsidering Antitrust’s Goals’ (2012) 53 Boston College Law Review 551, 575.

Policy Objectives and Concepts of Efficiency  47 consumers.41 It is common in the economics of public finance (ie, taxation) to speak of distributional equity, and taxation often has clear (though controversial) redistributive objectives.42 The ACT in Qantas appears to reject redistribution as an objective of competition law;43 however, the debate continues. The ACT subsequently addressed the issue of public benefits from efficiency gains in AGL/ MacGen and took the same view, which was opposed by the ACCC.44 This difference of opinion between the ACCC and leading economists in Australia appears to go back to the 1980s.45 The redistribution issue also manifests itself in the question whether a firm with substantial market power is permitted to profit from it. It is accepted in the US that the Sherman Act does not ‘prohibit the setting of a monopoly price or the monopoly quantity’.46 The Hilmer Report firmly recommends that competition law should not attempt to regulate monopoly prices.47 There is nevertheless a popular belief that competition law pursues a redistributive agenda in favour of consumers48 and small business.49 We may also draw on Cass Sunstein’s ideas about background norms, ie, ‘equality before the law’ and ‘antipathy to naked interest-group transfers’, to suggest that the legislature should hesitate to intervene in distributive issues between interest groups such as producers and consumers (or big business and small business).50 As we have seen in the Introduction, these views continue to influence reform of section 46. Redistributive objectives would appear to run counter to the project of competition law, ie, far from protecting the market mechanism to work unimpeded, such an objective would require a judgement that the unimpeded mechanism would

41 See discussion by Kirkwood ((n 39) 2434) of the argument in favour; and J Farrell and M Katz, ‘The Economics of Welfare Standards in Antitrust’ (2006) Competition Policy International, available at escholarship.org/uc/item/1tw2d426, who argue against. 42 See, eg, H Rosen, Public Finance (Boston, Irwin McGraw Hill, 1995). 43 Qantas (n 17) [170]. 44 AGL/MacGen (n 1) [252]–[261]. 45 M Brunt, ‘The Australian Antitrust Law after 20 Years: A Stocktake’ (1994) 9 Review of Industrial Organisation 483, 507 is critical of the ACCC view that efficiency gains must be passed on to consumers in order to be considered in merger cases; she cites similar criticism by RR Officer in 1987 in an unpublished paper. The Full Federal Court in ACCC v ACT [2017] FCFCA 150 [62]–[68], on appeal from the Australian Competition Tribunal in the matter of the Tabcorp merger, approved of the view taken by the Tribunal in Qantas (n 17), ie, that benefits to the merged firm such as cost efficiencies can be taken into account as a public benefit, but may be given less weight if consumers do not benefit (though the Full Court held that the Tribunal is not required to identify the weight it gives to such consideration). 46 Hovenkamp (n 24) 112; K Hylton, ‘The Law and Economics of Monopolization Standards’ in Hylton (n 10) 85. 47 Hilmer Report (n 13) 269. 48 Corones (n 4) 35. 49 R French, ‘Dolores Umbridge and the Concept of Policy as Legal Magic’ [2007] Federal Judicial Scholarship 14 [16]–[17]. 50 C Sunstein, ‘Interpreting Statutes in the Regulatory State’ (1989) 103 Harvard Law Review 405, 487 suggests that these norms explain why courts describe the Sherman Act prohibition of monopolisation ‘as an attempt to promote consumer welfare rather than as protection of small business as such’, but does not analyse the conflicting interests of producers and consumers as ‘interest groups’.

48  Market Power in Economics and Law produce excessively high prices and excessive profit for producers, and should for social justice reasons be corrected. How competition law could do this is unclear, but assuming that the law could, eg, by discouraging the exit of firms through mergers or by protecting smaller less efficient firms from competition by ­dominant firms, thereby encouraging excess capacity, it will nevertheless be the markets (importantly the capital markets) which decide whether sufficient profits can be made to justify investment of resources in producing the increased volume. As we will see, distributional equity is an objective that underpins the notions of ‘consumer harm’ and ‘market power’ derived from the neoclassical economic model of monopoly. Another policy objective of competition law is said to be ‘dynamic efficiency’, which refers to the ability of firms to initiate or respond to changes in technology, consumer choice and the global economy (and is therefore difficult to m ­ easure).51 The neoclassical model does not explain how firms self-select for entry to or exclusion from the market: the model assumes that firms and consumers have perfect information and may freely enter or leave the market, that transformation of production between goods happens effortlessly and that technological change simply ‘happens’ – thus, firms know which producers are more cost-efficient and move efficiently in and out of the market. Nothing, of course, could be further from the truth in the real world, where these processes are both destructive and ­imperfectly understood. For example, if there are many more equally efficient firms than are required to satisfy market demand, could we anticipate that those firms whose managers are more aggressive will stay or enter, forcing the meek to exit? It will be useful to examine dynamic gaps in the neoclassical economic models, eg, concerning how markets exclude producers, so we can distinguish harmful conduct of dominant firms from normal incidents of competition. Accordingly, conventional economic thinking does not elucidate the key concern of section 46: to distinguish situations where the market itself excludes firms from situations where a dominant firm uses its power to exclude rivals. We will return to this point later in the present chapter with a thought experiment concerning how equally efficient firms may self-select for entry/exit under zero-sum competition when what is at stake is survival. Turning to more traditional concepts of economic efficiency, ‘allocative efficiency’ refers to the allocation of resources in the economy to produce the mix of goods and services that satisfies consumer preferences. ‘Productive efficiency’ refers to the maximisation of production in the economy from given resources by a process in which each firm is operating at the minimum point on its average cost curve, which only occurs with certainty in the long run in perfect competition,

51 See eg, Hilmer Report (n 13) 4–5; Stucke (n 40) 577; P Akman, The Concept of Abuse in EU Competition Law (Oxford, Hart Publishing, 2012) 317.

Economic Models and the Paradigm of Competition  49 ie, it would be inefficient to apply more resources to produce goods and services beyond the equilibrium due to increasing costs. As we will see in the present ­chapter, in zero-sum competition we can address market efficiency (ie, whether market forces are undistorted) by focusing not merely on producers’ cost functions, but also on how market demand responds to price and how this determines the size of markets. In perfect competition, productive efficiency serves consumer preferences, ie,  ‘equilibrium’ achieves both allocative and productive efficiency. As we have seen Hovenkamp observe, economic theory does not provide a normative framework to assess consumer preferences. That is, there is no means of valuing (or forensically determining) any particular resource allocation that may be driven by consumer preferences; what is valued is freedom of choice according to one’s means. I suggest that we can take support from this observation for an ‘efficient market’ as a means to ensure that supply responds effectively to demand so that market forces operate without distortion. In an imperfect world, allocative efficiency and productive efficiency are inconsistent objectives: the pursuit of one can only be achieved at the expense of the other. It may well be impossible to resolve the necessary trade-off in general terms, as the nature and circumstances of markets for particular goods and services will differ, and socially desirable trade-offs between competing objectives will differ. Thus, for example, in a pandemic, with a scarce supply of medicines, the market mechanism would determine a price to ration supplies among consumers based on consumer preferences (ie, willingness to pay) and would ensure that efficient producers would enter the market to profitably supply the required medicines. Access to medicines is, however, an area where society has been prepared to sacrifice allocative efficiency in the interests of equity. ­Nevertheless, in a pandemic, the government would ration supplies to give priority to workers in essential services, ie, the government is prepared to sacrifice equity to keep institutions working in the interests of society as a whole. We may doubt that competition law, and our institutions which administer competition law, provide an appropriate mechanism to manage such trade-offs. In the next section of this chapter, we will see how the use of economic models in the discourse on ‘law and economics’ reflects underlying views of the world.

III.  Economic Models and the Paradigm of Competition In this section I examine the role which the paradigm of the hypothetical competitive market plays in our conception of market power. I propose an alternative conception of market power, ie, power to manipulate the market, which is not

50  Market Power in Economics and Law dependent on the assumptions of that paradigm. In the subsequent sections of this chapter, I will propose an alternative paradigm of competition more reflective of real-world markets, ie, zero-sum competition, in which to assess the misuse of market power. Hovenkamp acknowledges that while section 2 of the Sherman Act is aimed at preventing practices that ‘keep other firms from entering the market’, it is ­empirically difficult to distinguish practices that are a normal part of competition.52 As previously noted, he goes on to state that ‘the empirical and legal machinery we use for measuring market power and dealing with it is both costly and too crude for making fine adjustments’, so ‘we need to focus on the most flagrant abuses’.53 This would be legitimate if ‘we’ referred to a community of opinion shared by ­regulators and industry, economics experts and the courts. ‘Market power’ is defined by Hovenkamp as the ‘ability to profit by reducing output and raising price above the competitive level’.54 The reference to the competitive level is to a hypothetical that may or (as Hovenkamp acknowledges) more likely may not be known.55 It would be incorrect to describe the monopoly price as being above the competitive level (if that were a reference to the level of price and output that could be expected under a perfectly competitive market structure), as the conventional comparison of the two theoretical models depends on restrictive assumptions. It is sometimes said that a dominant firm ‘artificially’ restricts output in order to increase price and derive above-normal profits.56 This is not so, as the ‘profit maximising’ output is determined by market demand response (elasticity of demand). That is, increasing output beyond that point (or restricting output below it) will not be profitable and (as we will see) may well be socially inefficient. We will consider the ways in which theorists compare the two models, as this tells us more about their view of the world than it does about the models. The above definition of ‘market power’ leaves us uncertain about the source and nature of the power. As Hovenkamp observes, in perfect competition, a supplier ‘is said to have no market power’.57 Such a firm also has the ability to increase prices and reduce output, but because the firm-facing (or residual) demand curve for the individual firm’s product is horizontal (ie, perfectly elastic), an increase in the firm’s price will result in the loss of all business, and a reduction in output will not affect the price received and so would merely diminish the firm’s revenue with no compensating benefit. I suggest that firms do not ‘compete’ in a perfectly competitive market. Contrary to the views of

52 Hovenkamp (n 24) 23–24. 53 ibid 95. 54 ibid. 55 ibid 102. 56 See eg, G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011) 13–14. 57 ibid 97.

Economic Models and the Paradigm of Competition  51 William Redmond, when firms are ‘price-takers’, they do not compete, ie, there is no competition between suppliers in a perfectly competitive market; nor is there competition (ie, bidding or negotiation) between suppliers and consumers, which, as Redmond observes, is another mode of competition.58 Neri Salvadori and Rodolfo Signorino examine the absence of competition in the perfectly competitive equilibrium, where firms are price-takers.59 They conclude that ‘competition’ occurs only in disequilibrium, ie, when demand exceeds supply, buyers compete to bid up prices, and when supply exceeds demand, firms compete to reduce prices.60 As noted above, the downward sloping demand curve indicates that consumers compete with each other for scarce products, and those who are willing to pay the equilibrium price obtain supplies, while those who are outbid do not.61 Redmond observes that competition between consumers ‘rarely springs to mind when market competition is discussed’.62 In the discourse on law and economics, ‘competition’ is typically conceived of as competition between producers which is necessary to keep prices down. This view considers only one side of the market – competition between consumers is part of the market mechanism envisaged by classical authors.63 I suggest that the responsiveness of consumers to changes in the price of goods, relative to both close (if any) and distant substitutes, is the source of ‘market power’. When Hovenkamp says that ‘substantial market power is not something that a firm acting alone can easily create’,64 I would go further to suggest that the firm cannot create market power at all. Posner concludes that market power ‘depends entirely on the elasticity of demand’65 and asks rhetorically ‘whether such power can be obtained or enlarged through the efforts of one firm’.66 I suggest it cannot, but that the power of the market can be manipulated by firms. I have noted earlier in this chapter that the market manipulation approach can be reconciled with the ‘strategic market power’ theory advocated by some economists. As the firm’s ability to set its own prices and output are present regardless of whether a firm can be said to have market power or not, we must focus on the consequences, which are a function of how others respond to the firm’s unilateral conduct. Let us first consider the responsiveness of others to price and output ­decisions in monopoly, indicated by the traditional diagram which assumes rising cost curves due to the law of diminishing returns.

58 W Redmond, ‘Three Modes of Competition in the Marketplace’ (2013) 72 American Journal of Economics and Sociology 423, 425 and 427. 59 N Salvadori and R Signorino, ‘The Classical Notion of Competition Revisited’ (2013) 45 History of Political Economy 149. 60 ibid 159. 61 Redmond (n 58) 432. 62 ibid 435. 63 Salvadori and Signorino (n 59) 159 (discussing Adam Smith) and 163 (Karl Marx). 64 Hovenkamp (n 24) 15. 65 Posner (n 5) 312; see also Landes and Posner (n 6) 938. 66 Hovenkamp (n 24) 317.

52  Market Power in Economics and Law Figure 1.1  Monopoly equilibrium Price

P1

MC Monopoly profit

AC

AR MR Q1

Output

In a monopoly, the source of ‘market power’ is the sloping demand curve (ie, the responsiveness of consumer demand to changes in price) and the inability or reluctance of would-be suppliers to enter the market. To consider the impact of competition between suppliers, we need to consider less than perfectly competitive markets. We may expect that, should the firm seek to cut prices, it will not realise the benefit of selling as much product as it could make (as it would under perfect competition) because it can be anticipated that other suppliers would follow suit.67 This follows from responses of competing suppliers as well as consumers.68 Thus, market power can be seen to be a function of the responsiveness of consumers, but can be constrained by the response of competing suppliers. The argument that ‘market power’ is a function of the responses of others, rather than any inherent capacity of the firm, is supported by Blair and ­Carruthers’ discussion of the ‘Lerner Index’. They define ‘monopoly power’ as the ‘degree to which one firm can exclude others from competing’.69 Although they use the term ‘monopoly power’, it is clear from their context (including the reference in US law to ‘monopolisation’) that they are discussing ‘market power’. The Lerner Index, which is used to measure market power, is the margin by which a dominant firm’s price exceeds its marginal cost.70 In other words, the Lerner Index uses the ‘competitive level’ as the benchmark to measure market power. Blair and ­Carruthers show that the index correlates with elasticity of demand. I infer from their analysis that, in legal terms, market power as measured by the Lerner Index is ‘caused or contributed to by’ elasticity of demand. I do not consider it to be n ­ ecessary for the present purposes to empirically test or prove a functional



67 Dooley

(n 35) 114. and Chrystal (n 36) 259, fn 1. 69 Blair and Carruthers (n 10) 69. 70 ibid. 68 Lipsey

Economic Models and the Paradigm of Competition  53 r­ elationship as a matter of economics. Of course, market share is also relevant, but is thought not to be legally determinative of market power. I acknowledge that above-normal economic profits can be explained by factors other than misuse of market power.71 Nevertheless, I use the article by Blair and Carruthers, supported by the earlier article by Landes and Posner, as a provocation to propose the­ alternative legal conception of market power set out in this monograph, ie, market power can be conceived of as the power of the market (demand response) which cannot be possessed, but can be manipulated by firms. As will become apparent, the market manipulation approach has the advantage that we may avoid forensic difficulties that arise under the current legal theory, which require us to benchmark conduct against a legally indeterminate hypothetical competitive level of price and output. As noted above, the market itself excludes less efficient rivals and equally ­efficient rivals for whom there is no place in the market (ie, because additional output would depress prices below the level required for profitable production). Blair and Carruthers conclude that, given the role of elasticity of demand, ‘market share is not the only determinant of market power’.72 As noted above, Posner discusses the role of market share: a higher market share enables a firm to have a greater capacity to influence the market, but the result depends entirely on elasticity of demand.73 It is proposed that the response of competitors may detract from market power, but that the source of it can be viewed as the responsiveness of consumers, ie, elasticity of demand.74 As Redmond observes, non-price competition such as product differentiation is a resultant strategy of suppliers under such market conditions.75 Though this may be regarded as an element of competition between suppliers, Redmond regards it as a means by which suppliers negotiate higher prices from consumers, and hence of competition between suppliers and consumers (which he regards as an undesirable contribution to suppliers’ market power).76 Redmond considers that the object of antitrust laws is ‘to hold prices down for consumers’.77 However, section 2 of the Sherman Act does not ‘prohibit the setting of a monopoly price or the monopoly quantity’.78 The Hilmer Report states the benefits from competition include ‘lower prices’.79 However, the context of the report was recommended legislative reform to expose publicly owned enterprises in Australia to the rigours

71 Bork and Sidak (n 34) 513–14 and 521–22. 72 ibid 73–74. 73 Posner (n 5) 312. 74 Compare s 46(3) of the CCA, which assumes that competitive constraints are central to the ­assessment of market power (and is unchanged by the Harper Review amendments). 75 Redmond (n 58) 427. 76 ibid 428–29. 77 ibid 439. 78 Hovenkamp (n 24) 112; Hylton (n 46) 85. 79 Hilmer Report (n 13) 1.

54  Market Power in Economics and Law of competition law, and the report firmly recommends that competition law should not attempt to regulate monopoly prices.80 What is said to be inefficient about monopoly equilibrium is the reduced output. The inefficiency cannot be explained by reference to the level of prices in an assumed competitive market (as there is no way of determining them), but can only be ‘internal’ to the model of monopoly. When Blair and Carruthers refer to ‘competitive prices’ in the context of monopoly, they mean the price that would apply if the monopolist continued to produce (at an increasing loss for each additional unit produced) until profit were reduced to some acceptable level. Landes and Posner appear to use the term ‘competitive price’ in the same way.81 This is ­indicated by Figure 1.2 below,82 at the output (Q1) at which marginal cost equals price, but exceeds average cost (so the monopolist still earns excess profits). The reason for choosing output Q1 (and P1 as the competitive price) is apparently that the sum of producer surplus and consumer surplus (and the share allocated to consumer surplus) is maximised when output is expanded until marginal cost equals price. It could be inferred that, defining the ‘competitive level’ in this way, the monopolist’s marginal cost curve is a surrogate for the industry supply curve in a competitive market. However, this would require an assumption that there are no relevant economies of scale, and in fact no a priori comparison can be made.83 In the model of perfect competition, the aggregate supply curve is the horizontal sum of each firm’s marginal cost curve. However, horizontal summation only applies because the individual demand curve facing each firm is horizontal; as Lipsey and Chrystal point out, there is no supply curve for a monopolist.84 In perfect competition, the firm’s equilibrium output is the point where marginal cost equals average cost, as production becomes increasingly inefficient for the firm if increased beyond that. In Figure 1.2 below, this point is to the left of Q2. The conventional analysis of deadweight loss depicted in Figure 1.2 also assumes that the ‘competitive level’ is reached before the point at which demand response becomes negative, ie, where marginal revenue becomes equal to or less than zero. However, the relationship depicted between cost and revenue functions is simply an arbitrary assumption. When we come to address efficiency below, we will ask whether increasing output from Q2 to Q1 is an efficient use of society’s resources. This requires an empirical analysis depending on the circumstances of the case. 80 ibid 269. 81 Landes and Posner (n 6) 941. 82 Based on Blair and Carruthers (n 10) 68. The concept of deadweight loss can be diagrammatically represented in slightly different ways (eg, J Perloff, R Smith and D Round, Microeconomics (Frenchs Forest, Pearson Australia, 2014) 343, Figure 11.5); however, they share the underlying assumptions discussed in this monograph. 83 See eg, Lipsey (n 33) 373; EAG Robinson, Monopoly (Cambridge, Cambridge University Press, 1941) 140–42; Akman, Concept of Abuse (n 51) 21. 84 Lipsey and Chrystal (n 36) 241.

Economic Models and the Paradigm of Competition  55 Figure 1.2  Monopoly price and output Monopoly profit Deadweight loss (area abc) a

P2

MC b

P1

AC

c AR Q2

MR

Q1

Allocative efficiency and productive efficiency coincide in perfect competition because the firm’s demand curve is flat, ie, price = marginal revenue = marginal cost. When the demand curve facing a firm is downward-sloping, the conditions required by the model are not satisfied, and allocative efficiency and productive efficiency no longer coincide.85 This is demonstrated graphically in Figure 1.3 below by the point (Ep) at which the production possibilities frontier touches the consumers’ indifference curve,86 representing all combinations of two goods that maximise consumer satisfaction. The tangent line at Ep represents relative prices (ie, the marginal rate of substitution (MRS)). Figure 1.3  Allocative and productive efficiency Y EM MRS MRT

EP

MRS X

85 ibid 297. 86 ibid 406, fn 6, observing that this requires constructing a community indifference curve (which may or may not be possible).

56  Market Power in Economics and Law At equilibrium in perfect competition, the substitution of goods will occur until consumer satisfaction is maximised, at Ep. Because each producer’s marginal cost equals price, the transposition between production of the two  goods will continue until profit is maximised. In perfect competition, both profit and consumer ­satisfaction would be maximised at point Ep, so the production possibilities curve and indifference curve would touch, sharing the same tangent. In any model other than perfect competition, the demand curve facing the firm will be sloping, ie, price ≠ marginal cost. Therefore, the marginal rate of transformation (MRT) (the ratio of marginal cost of producing the two goods) will differ from the ratio of prices at the optimum, Em. Assuming Em to represent monopoly equilibrium, the monopolist’s profit maximisation will prevail over consumer choice, so there will be other possible production mixes of the two goods (to the right of Em in the above figure) that would increase consumer satisfaction and approach allocative efficiency, but this would come at the cost of productive efficiency. To show this, refer back to Figure 1.2: if we induce the firm to increase production until price equals marginal cost, we could theoretically move production to Q1, but because marginal cost is greater than average cost, production becomes increasingly inefficient for the firm, since the firm incurs a loss on the increased output, and profit is not maximised. It is only under the assumptions of perfect competition that it can be argued that social welfare is improved by causing the firm to act against its own interests by increasing output when marginal costs exceed marginal revenue. However, in real-world markets, where these assumptions do not hold, we cannot assume that drawing resources away from other uses in the economy has no effect on social welfare. Thus, it may well be that increasing output beyond the firm’s profit-maximising output would have a negative effect on social welfare and thus it would be inefficient to draw resources from other uses in the economy simply to redistribute surplus from producers to consumers in a particular market. Accordingly, in an imperfect world, we cannot achieve both allocative and productive efficiency, and must trade one off against the other. If we can never align marginal cost and price, some economists argue that allocative efficiency is not an achievable objective of competitive markets and that the more reliable strategy to increase consumer welfare is to increase productive efficiency, ie, to move the production possibilities frontier outwards.87 The process by which productive efficiency is increased by technological innovation is referred to as ‘dynamic ­efficiency’.88 Increasing productive efficiency in this way will not necessarily result in an allocatively efficient equilibrium, but consumer welfare will be increased by the greater income produced. Allocative efficiency may have even

87 ibid 410. 88 See, eg, S Bishop and M Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, 3rd edn (London, Sweet & Maxwell, 2002) 36–37.

Economic Models and the Paradigm of Competition  57 less policy appeal if we doubt that consumer preferences are truly the product of rational self-interest.89 Further, as noted above, in an open economy, the preferences of domestic consumers are not necessary or sufficient to drive productive efficiency of the economy’s resources. Accordingly, I suggest that one cannot refer to allocative and productive efficiency as a single achievable goal of competition law.90 Returning to Figure 1.2, economists have differing approaches to the possible regulation of monopoly prices. As noted above, some suggest that output should be increased until price equals marginal cost. Others suggest that the acceptable output is where price equals average cost, so excess profits are eliminated.91 In Figure 1.2, this point is to the right of Q1 (where average cost equals average revenue). Since the assumptions of the economic models under which such outcomes would be possible are not satisfied in the real world, I suggest that these strategies require intervention in the market. That is, we may doubt whether idealised outcomes can be achieved by competition laws which aim to facilitate the undistorted operation of the market mechanism.92 Further, there is no way of comparing the equilibrium price under the models of monopoly and perfect competition in order to determine whether equilibrium prices would be higher or lower under monopoly. What Blair and Carruthers describe as a ‘deadweight loss’ reflects a value judgement in which lower prices for consumers are achieved at the cost of increasingly inefficient production. In other words, this intervention redistributes welfare from producers to consumers. Intervention also reduces incentives for would-be competitors to invent around the monopolist’s advantage. ‘Excess profits’ are thought to be required in order for the market mechanism to be self-correcting, ie, would-be competitors need the incentive of monopoly profits to encourage them to enter the marker or invent around the monopolist’s advantage. By down-regulating a monopolist’s prices, this incentive to self-correct is removed. This mechanism will be considered further in the next section of this chapter when we model incumbent response to new entry. When theorists refer to the ‘competitive’ output or price in the model of monopoly, they fail to address the incomparability of the two models. As noted above, Hovenkamp, Blair and Carruthers do not expressly state the assumed correspondence. Posner mentions the assumption, but does not consider the implications.93 As Lipsey points out, the conclusion that monopoly equilibrium is at a lower output and higher price than competitive equilibrium depends entirely

89 See, eg, M Huffman, ‘Marrying Neo-Chicago with Behavioural Antitrust’ (2012) 78 Antitrust Law Journal 105, 117. 90 As apparently does, eg, J Wright, ‘The Antitrust/Consumer Protection Paradox: Two Policies at War with Each Other’ (2012) 121 Yale Law Journal 2216, 2239. 91 Dooley (n 35) 87–90. 92 Contra: eg, Perloff, Smith and Round (n 33) 358: ‘Encouraging competition is an alternative to regulation as a means of reducing the harms of monopoly.’ 93 Posner (n 5) 281 and 291; see also Landes and Posner (n 6) 941.

58  Market Power in Economics and Law on the assumption that the monopolist has the same marginal cost as producers in a competitive market.94 The conclusion will not hold where a monopolist could produce more efficiently, ie, if there are economies of scale or scope.95 The purpose of this discussion is to critique the ‘competitive level’ as a relevant conceptual benchmark for our legal theory of liability. I will suggest in Chapter 3 that we should start our analysis from the actual position in which the firm finds itself rather than from an hypothetical ‘competitive level’. Thus, we approach from the opposite direction to that if we were to consider a hypothetical move from the ‘competitive level’ as our starting point.96 As it is impossible to determine whether or how far price may have been raised above the hypothetical competitive level, it seems that the attainment of that level must remain an objective which can only be achieved, if at all, by intervening in the market, ie, it is not sufficient merely to ensure there are competitive constraints on market power. Hovenkamp states that ‘if we could confidently predict that new firms would enter a market whenever price exceeded the competitive level, ­antitrust would have little to do’.97 However, he acknowledges controversies about the definition of entry barriers, which adds to the indeterminacy of market power, as ‘the presence or absence of entry barriers is always essential to estimating market power’ and, further, conduct that excludes rivals is ‘often ambiguous and difficult to assess’.98 For example, capital expenditure may be efficient and desirable, but it may also raise barriers to entry. Hovenkamp observes that ‘an overly aggressive antitrust rule would chill innovative conduct, so we frequently give the firm acting unilaterally the benefit of the doubt’.99 This appears to explain the scepticism of US courts to intervention in cases of dominant firm conduct and could well be a factor affecting judicial decisions in Australia which may appear to be based on commercial intuition. Either way, such a law lacks certainty and predictability, and will inevitably lose the confidence of the business community, though it may provide comfort for consumer interests fearful of the perceived power of business. The pursuit of allocative efficiency is not achievable in imperfectly competitive, ie, real-world, markets and (if achievable at all) involves a consequential trade-off at the expense of production and income. To value allocative efficiency as synonymous with freedom of choice overlooks the fact that markets give effect to a qualified freedom of choice according to willingness and ability to pay. If we are realistic about what markets can achieve, we would focus on letting the 94 Lipsey (n 33) 373; Lipsey and Chrystal (n 36) 408. 95 Lipsey (n 33) 379–80. 96 For a discussion of that approach, see, eg, Oxford Economic Research Associates, Costs and Benefits of Market Regulators, Part 2: Practical Application, report for Dutch Ministry of Economic Affairs, October 2004, 72–73, where the authors discuss O Williamson, ‘Economies as an Antitrust Defense: The Welfare Tradeoffs’ (1968) 58 American Economic Review 18. 97 Hovenkamp (n 24) 102. 98 ibid 103. 99 ibid 109.

Market Manipulation Laws in Securities Markets  59 market achieve what it can undistorted, and leave equity or social justice objectives (eg, achieving ‘fair and reasonable’ prices) to measures other than competition law. To allow the market to work undistorted, we should consider ‘abuse of dominance’ in the broader context of preventing market manipulation.100

IV.  Market Manipulation Laws in Securities Markets It is recognised in securities and derivatives markets such as commodity futures that sellers can manipulate the market by reducing their volume of sales (‘cornering’) or increasing their volume of sales (‘squeezing’) for the purpose of profiting from resulting price effects.101 The problem with ‘market manipulation’ is distinguishing prohibited price manipulation from normal market price volatility due to undistorted forces of supply and demand,102 a similar problem to that we encounter with misuse of market power. We can thus see that we do not need to determine hypothetical competitive standards of price and output when we adopt a market manipulation approach. The issue of market manipulation in energy markets raises a similar problem to that we encounter with misuse of market power in distinguishing between policy purposes of allowing markets to do what they do unimpeded, or alternatively intervening in markets to achieve what we consider to be equitable or social justice outcomes. Kelliher discusses the statutory duty of the US federal energy regulator to ensure that energy prices are ‘just and reasonable’, and the dual strategies by which it has sought to achieve that objective – promoting competition and prohibiting ‘market manipulation’.103 There would seem to be no difficulty condemning manipulation as it can be conceived of as conduct intended to distort the market mechanism. In early 2016, the Australian Securities and Investments Commission (ASIC) commenced litigation against several banks for manipulating the Bank Bill Swap Reference Rate. The case concerned a number of statutory provisions, notably sections 1041A and 1041B of the Corporations Act 2001 (Cth), and did not address competition laws. For the sake of completeness, the subsequent outcome is noted, but will not be discussed further in this book. In November 2017, two banks, ANZ and NAB, reached a settlement with the ASIC admitting liability and accepting agreed penalties which were subsequently approved by the court.104

100 See eg, C Pirrong, ‘Commodity Market Manipulation Law: A (Very) Critical Analysis and a Proposed Alternative’ (1994) 51 Washington & Lee Law Review 945, 951–52, who analyses market power and market manipulation in commodity futures markets. 101 ibid 950–51; J Kelliher, ‘Market Manipulation, Market Power, and the Authority of the Federal Energy Regulatory Commission’ (2005) 26 Energy Law Journal 1, 15. 102 Pirrong (n 100) 960–61. 103 Kelliher (n 101) 17. 104 See eg, ASIC v National Australia Bank Ltd [2017] FCA 1338.

60  Market Power in Economics and Law The  case against the third bank, Westpac, proceeded to judgment, which was delivered on 24 May 2018 (ASIC v Westpac Banking Corporation (No 2)). The judgment of Beach J contains a useful analysis of the legislative history and issues involved in the forensic application of the Australian securities market manipulation laws.105 Of particular note is the discussion of market manipulation involving market power, such as ‘corners’ and ‘squeezes’.106 Detailed discussion of securities markets manipulation laws is beyond the scope of this monograph. Nevertheless, I suggest that we can draw on concepts from such laws, eg, the concept of market power as power to manipulate the market and the concept of an ‘efficient market’. Judicial usage of the term ‘efficient market’ in the context of securities markets manipulation laws simply means that prices reflect the undistorted interplay of forces of supply and demand.107 I adapt this concept in the present context, where we focus on whether supply is responding to demand, eg, whether there is unmet demand. To consider whether the market is working efficiently, we need to understand how the market mechanism works under conditions of imperfect competition, ie, when supply responds not to price (average revenue), but to changes in revenue (marginal revenue). The market will not respond to demand shifts in the same way as it would under perfect competition. We will discuss this further in Chapter 3 when we develop forensically determinable measures of efficiency. The commonly accepted definition of market power as the ability to reduce output and raise prices is based on assumptions concerning the comparison between neoclassical models of perfect competition and monopoly which are not comparable. Thus, our concept of market power is built on idealised foundations. Several consequences follow from this. I suggest that it is legally indeterminate to benchmark a firm’s profits against the hypothetical ‘competitive level’, whether this be conceived as the ‘perfectly competitive’ or ‘workably competitive’ level.108 I suggest that it is similarly legally indeterminate to benchmark a firm’s conduct against what it may be expected to do in a competitive (or workably competitive) market. However, it may be possible to compare the effect on the market of conduct which manipulates the market. The discussion of economics in this monograph is not intended to prove or disprove any theories or principles of economics, but rather to show that there are many underlying assumptions that are not reflective of the real world and 105 ASIC v Westpac Banking Corporation (No 2) [2018] FCA 751 [1886]–[2126]. 106 ibid [1902]. The court at [1895] acknowledges the classification of various categories of securities market manipulation by E Avgouleas, The Mechanics and Regulation of Market Abuse: A Legal and Economic Analysis (Oxford, Oxford University Press, 2005) 118–53. See also, eg, AF Bergþórsson, What is Market Manipulation? An Analysis of the Concept in a European and Nordic Context (Leiden, Brill, 2018). 107 ibid [1925]–[1926], citing the High Court of Australia in DPP (Cth) v JM (2013) 250 CLR 135, [2013] HCA 30 [72]. The High Court at [77] made it clear that the concept of market manipulation is not limited to ‘corners’ and ‘squeezes’ or by concepts from US jurisprudence. 108 As to ‘workable competition’, see JM Clark, ‘Toward a Concept of Workable Competition’ (1940) 30 American Economic Review 241.

Thought Experiment: Incumbent Response to New Entry  61 many empirical matters that we must regard as legally indeterminate. Accordingly, I propose an alternative legal conception of market power (it is not something that can be possessed or taken advantage of): the power is that of the market, which can be manipulated. When market prices rise, consumers reduce the quantity they buy. Depending on the elasticity of demand, total revenue may rise or fall. The same holds true for a reduction in market prices: total revenue may rise or fall depending on the elasticity of demand. We may thus begin to see a way to theorise market power without the overlay of conflicting welfare objectives and efficiency standards that has troubled competition law over a long period. The recent Australian law reform debate has concluded that the concept of ‘taking advantage’ of market power is inadequate and should not form part of our legal theory of liability. The implicit assumption underlying section 46 that a firm may ‘have’ market power has not been questioned. Nevertheless, it seems clear that ‘having’ market power, eg, by creating or acquiring it in some way, is not prohibited by section 46. Despite suggestions in the Australian literature that it is not unlawful to have market power or profit from it, there is no authoritative pronouncement from Australian courts on the question. We must look to the literature and decided cases in the US and the EU for consideration of the question, which is pursued in Chapter 3 and Part II. Suffice to say here that there is a significant inconsistency in the conventional legal approach to misuse of market power arising from the following generally accepted propositions: (a) it is not unlawful for a firm to have or profit from market power; (b) misuse of market power can be assessed by reference to a hypothetical in which the firm would not have or profit from market power. Having proposed another way of legally conceiving market power and its misuse, we can now proceed to examine a gap in the economic models concerning how firms self-select for entry to and exit from markets.

V.  Thought Experiment: Incumbent Response to New Entry As noted above, section 2 of the Sherman Act does not ‘prohibit the setting of a monopoly price or the monopoly quantity’.109 This statement invites several questions. First, it appears to acknowledge that, given the differences of approach by economists to the regulation of monopolies, the regulation of price and output in the public interest is not a matter suitable for judicial determination. In Australia, the Hilmer Report appears to have adopted a similar approach, recommending that competition laws should not seek to regulate monopoly pricing, but the



109 Hylton

(n 46) 85.

62  Market Power in Economics and Law competitive process itself.110 Second, it seems to support my conclusion that the equilibrium price and output can alternatively be said to arise not from the exercise of market power, but from the response of consumers, ie, the elasticity of demand. Third, if the objective of antitrust is not to regulate the monopolist’s profit, we can conclude it is to ensure that market forces will be allowed to operate to test the efficiency of the monopolist’s position (ie, if an equal or more efficient producer does not enter the market, we may conclude that the monopoly price and output is most efficient).111 As Blair and Carruthers observe, the market mechanism excludes less efficient firms.112 We can also state that the market excludes equally efficient firms not required to satisfy market demand. Posner considers that the focus of the legal prohibition should be conduct that is likely to exclude equally or more efficient firms.113 The economic model of perfect competition does not tell us how the market excludes equally efficient firms. Do they self-select or is there a process akin to ‘survival of the fittest’, whereby equally efficient rivals enter the market by knocking out incumbents? This is the paradigm which I call ‘zero-sum ­competition’. If zero-sum competition applies, it would seem difficult to make value judgements about the conduct of equally efficient competitors. The model of perfect competition assumes that firms have perfect knowledge and can enter and leave the market without cost. As such, the fact that some firms are excluded and quietly deploy their resources in other productive activities is of no policy concern in the model world. In the model world, a monopolist would know, when confronted by a new entrant with new technology, which firm is more efficient. In the real world, these assumptions are not borne out and so we encounter barriers both to entry and exit.114 Thus, for example, BlackBerry did not simply cease production, realise assets and return funds to its shareholders when the Apple iPhone came on to the market, but went into a long decline in which shareholder value was greatly diminished, until it exited the smartphone market in 2016.115 The Harper Review notes that barriers to exit impede the efficient operation of markets, resulting in transitional costs which can be painful to those involved, and consequently that such barriers should be reduced ‘wherever ­possible’.116 It would seem unlikely that such barriers (which exist by reference to the simplified assumptions of the model of perfect competition) can be meaning-

110 Hilmer Report (n 13) 269. 111 Hovenkamp (n 24) 156. 112 Blair and Carruthers (n 10) 74. 113 R Posner, Antitrust Law, 2nd edn (Chicago, University of Chicago Press, 2001) 194–95, cited in Hovenkamp (n 24) 153. 114 See, eg, S Nargundkar, F Karakaya and M Stahl, ‘Barriers to Market Exit’ (1996) 13 Journal of Managerial Issues 239, 241. 115 V Savov, ‘BlackBerry’s Success Led to its Failure’, The Verge, 30 September 2016, www.theverge. com/2016/9/30/13119924/blackberry-failure-success. 116 Commonwealth of Australia, Competition Policy Review, March 2015 (hereinafter ‘Harper Review’) 12 and 16.

Thought Experiment: Incumbent Response to New Entry  63 fully reduced. This, I suggest, leads us to a critical aspect of real-world competition, that firms cannot freely exit the market and so must compete for survival. Accordingly, we need to consider how this aspect of zero-sum competition may affect economic efficiency and norms conduct. In Australia, the judicially accepted definition of real-world competition is rivalry between firms for the same object,117 which in this study is called ‘zero-sum competition’ because conduct having the purpose of advancing the commercial interests of the firm at one and the same time harms competitors. For example, if demand is finite and a firm wants to increase its sales, or wants to increase its sales growth faster than the market, it can only do so by taking away sales from its competitors, thus threatening their survival. I use the term ‘zero-sum competition’ as a defined term. I do not believe that it has been applied previously in the context of market power. Courts in a variety of contexts use the expression ‘zero-sum’ as I do, ie, to indicate that one person’s gain is another’s loss.118 This kind of competition can be distinguished from the economic model of perfect competition, in which firms may sell more or less without affecting competitors or the market, ie, there is no real inter-firm competition (and no firm has market power). On the other hand, economists use non-zero-sum games to analyse competitive strategies in oligopoly competition.119 This is appropriate because demand may increase in response to falling prices or markets may be growing, so competitive outcomes are not necessarily zero-sum in the strict game-theoretical sense. Nevertheless, game theory acknowledges ‘survival games’ as a special case.120 Smith and Round suggest that an oligopoly in some respects resembles a survival game.121 When economists use game theoretical models to inform our understanding of strategies in an oligopoly, they commonly assume that peaceful coexistence is possible, ie, that the market will absorb the impact of output or price competition and that non-zero-sum games are appropriate.122 Having thus defined ‘rational’

117 Hilmer Report (n 13) 2; Queensland Wire (1989) 83 ALR 577, 584. 118 See, eg, NIB Health Funds Ltd v Private Health Insurance Administration Council [2002] FCA 40, [10] and [23]; Randall Pty Ltd v Willoughby City Council [2005] NSWCA 205 [14]. This might not be the same way that economists use the term technically in game theory. However, some usage of the term in economics may suggest that ‘zero-sum competition’ is reflective of the real world; see, eg, D Besanko, D Dranove, M Shanley and S Schaefer, Economics of Strategy, 7th edn (Hoboken, Wiley, 2016) 253, who say, in relation to Michael Porter’s ‘five forces’ framework of industry analysis, that Porter ‘tends to view all other firms – be they competitors, suppliers or buyers – as threats to profitability, as if business is a zero or even negative sum game’. 119 See, eg, S Stroux, US and EC Oligopoly Control (The Hague, Kluwer Law International, 2004) 13, fn 52. 120 FM Scherer and DR Ross, Industrial Market Structure and Economic Performance, 3rd edn (Boston, Houghton Mifflin, 1990) 209, fn 26. It may have been this that Schumpeter (n 21) 80 had in mind when he refers to ‘cutthroat’ competition in contradistinction to ‘beneficial’ competition of the neoclassical model. 121 R Smith and D Round, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5 Agenda 25, 35. 122 See Scherer and Ross (n 120) 223, from whose discussion I infer that these are common assumptions of Cournot, Bertrand and Stackelberg games. We could use those games to model outcomes, as a

64  Market Power in Economics and Law conduct by reference to, eg, a Nash equilibrium as the norm, it is easy for economists to condemn ‘win-lose’ strategies. However, I suggest that lawyers and regulators may bring different views of the world to bear. I would argue that there are distinct norms of conduct reflecting different views of the world which economists, lawyers and regulators bring to the competitive paradigm. We can illustrate this by applying the overlay set out in Figure 1.4 below to the payoff matrix from the ‘Prisoner’s Dilemma’ game.123 Figure 1.4  Payoff matrix with strategies Lose-lose Nash (best outcome given competitor strategy; defensive, not destructive) ‘Conscious parallelism’ (rational and legal)

Win-lose (eg, exclusionary strategy; irrational) (illegal for more powerful firm)

Win-lose (eg, exclusionary strategy; irrational) (legal for powerless firm?)

Win-win Cartel (best outcome under collusive strategy) (rational but illegal)

‘Win-lose’ is the norm in our legal system, which forms the frame of reference lawyers bring to the adjudication of disputes. Experience suggests that when lawyers participate in games based on the Prisoner’s Dilemma they prefer winlose strategies even when win-win strategies are possible.124 Such games are used to train lawyers in negotiating and resolving disputes where mutual agreement, or win-win, is valued. In our present context, of course, agreements between competitors would generally be unlawful. I suggest that we can anticipate that win-lose strategies will be normal conduct in the case of ‘zero-sum’ (or survival) competition. We might even find that regulators may suspect a Nash equilibrium to indicate possible collusion, or at least to suggest an absence of active competition

basis for deriving a norm of behaviour, though I do not do so because they may not necessarily coincide with ‘zero-sum competition’ as I have defined it. I am more interested in how firms self-select for exit when the market is not large enough to allow them all to attain efficient scale and normal profitability. No doubt, survival competition could be modelled, as Scherer indicates in a footnote ((n 120) 209, fn 26), but I have not encountered it in the literature on market power. J Perloff, L Karp and A Golan, Estimating Market Power and Strategies (Cambridge, Cambridge University Press, 2007) 34–39 assess the outcome of Cournot and Bertrand games by reference to the ‘competitive level’. 123 After Stroux (n 119) 14; see also Scherer and Ross (n 120) 210. 124 My experience is similar to that reported by C Maughan, M Maughan and A Thornhill, ‘Confronting Adversarial Attitudes to Negotiation: What the Red/Blue Exercise Can Teach Us and Our Students’ (1998) 32 Law Teacher 79.

Thought Experiment: Incumbent Response to New Entry  65 which would be better for consumers. Certainly, a Nash equilibrium appears to be a defensive strategy against a more destructive ‘lose-lose’ outcome, ie, ‘mutually assured destruction’. We can see that it could be forensically difficult to distinguish a Nash equilibrium (conscious parallelism) from collusion (cartel conduct).125 I suggest that it would be desirable if we can develop norms of conduct that do not wholly depend on one’s view of the world, given that economists, lawyers and regulators may have differing views concerning ‘normal’ competitive strategies. It is therefore pertinent to make some observations about zero-sum competition and efficiency, which may assist us in developing relevant norms. While a Nash e­ quilibrium minimises loss to firms, it can be unstable when anticipated behaviour of other firms changes, which may lead to more destructive ‘lose-lose’ outcomes, eg, a price war.126 Such an outcome may well be an effective mechanism, when there is excess capacity in the market, to cause firms to exit. Perhaps this is inevitable when there are barriers to exit. When competing firms find themselves engaged in destructive ‘lose-lose’ situations, it would be natural for a firm to adopt a ‘­win-lose’ strategy in an endeavour to ensure that it is the survivor, ie, to destroy one or more other firms. However, this mechanism would appear (certainly by comparison with perfect competition) to provide an inefficient way for firms to exit the market, ie, society’s scarce resources are expended in producing goods or services below cost (though this creates a private benefit for consumers in the market in question).127 Thus, we may appreciate the ‘Achilles’ heel’ of zero-sum competition: the value created by new entry is not guaranteed to exceed the value destroyed. We may illustrate this by referring to the traditional analysis of new entry in the case of monopoly. Figure 1.2 above sets out the commonly assumed

125 See eg, Apco Service Stations Pty Ltd v ACCC [2005] FCAFC 161 and ACCC v Leahy Petroleum Pty Ltd [2007] FCA 794, which indicate the courts are unwilling to infer collusion from conscious parallelism. See also G Hay, ‘Practices that Facilitate Cooperation’ in J Kwoka and L White (eds), The Antitrust Revolution (Glenview, Scott, Foresman & Co, 1989) 183–90. In a rare Australian case discussing the Cournot and Bertrand games, the court in a cartel case, ACCC v Colgate-Palmolive Pty Ltd and Others (No 4) [2017] FCA 1590 [656]–[660], accepted Professor Hay’s expert evidence that conscious parallelism can be distinguished from cartel conduct. The court rejected the evidence of the ACCC’s expert, Professor Williams, that the competitive benchmark requires ‘unilateral conduct’, which seems to have been defined in a manner reminiscent of perfect competition, where firms neither know nor care what other firms may do. It appears that in Professor Williams’ view, the competitive benchmark would rule out ‘interdependence’, ie, would not allow decisions to be made with regard to the conjectured response of competing firms. The court considered this to be unrealistic and flawed. Further discussion of cartels is beyond the scope of the present monograph. See, eg, C Beaton-Wells and B Fisse, Australian Cartel Regulation: Law, Policy and Practice in an International Context (Cambridge, Cambridge University Press, 2011). 126 See, eg, A Krämer, M Jung and T Burgatz, ‘A Small Step from Price Competition to Price War: Understanding Causes, Effects and Possible Countermeasures’ (2016) 9 International Business Research 1. 127 ibid.

66  Market Power in Economics and Law r­elationship between cost and revenue for a monopolist. The monopolist in Figure 1.2 achieves efficient scale and maximises profit at around output Q2. Let us assume that an equally efficient new entrant appears. If new entrant and incumbent compete on price, and let the market determine resulting output, it is likely that neither would be viable because both cannot achieve efficient scale without market demand depressing prices to a level that dissipates previous monopoly profits. Further, monopoly profits will be dissipated if the new entrant must invest capital so that additional fixed costs have to be absorbed. Thus, we have a situation where the market may allow only one firm to achieve efficient scale. Perloff, Karp and Golan show that industry concentration, ie, the number of firms in a market, is determined in the first instance by the size of the market,128 so we need to ask what is the size of the market indicated by a downwardsloping demand curve? I suggest an answer to this question can be derived from Figure  1.5 below. Once we are in possession of that empirical information, we can then enquire whether the size of the market limits the number of firms or whether other causes such as anticompetitive strategies of the dominant firm might limit the number of firms by excluding equally or more efficient firms from entering the market. While monopoly profits are said to provide an incentive for new entry, a rational new entrant will anticipate that by entering the market, such profits will be dissipated. This is the ‘Bertrand Paradox’, and conventional wisdom considers that the size of the market will provide an insurmountable barrier to entry.129 However, I suggest that in zero-sum competition, the ‘Bertrand Paradox’ may be overcome if the new entrant can appropriate the incumbent’s market to itself in order to be viable, ie, the new entrant has an incentive to commercially destroy the incumbent. This is a rational strategy and can be anticipated, especially if we may expect that the new entrant will overestimate its chances of success.130 ­ Furthermore, it would be inefficient to increase production beyond the point at which marginal revenue becomes negative, because falling revenue would not justify the additional resources (including capital investment by the new entrant) that would be diverted from other productive uses in the economy. This would cause a detriment to society beyond the private gains to participants in the market in question from the transfer of producer surplus to consumers.

128 Perloff, Karp and Golan (n 122) 34–39 discuss and critique studies by John Sutton, Sunk Costs and Market Structure (Cambridge, MA, MIT Press, 1991), which indicate that endogenous sunk costs and competitive strategies, as well as market size, can affect concentration. 129 See, eg, J Church and R Ware, Industrial Organisation (Boston, Irwin McGraw-Hill, 2000) 473–80. 130 Suggested by the research of FM Scherer, ‘The Innovation Lottery’ in RC Dreyfuss, DL Z ­ immerman and H First (eds), Expanding the Boundaries of Intellectual Property (Oxford, Oxford University Press, 2001) 3–21, who concludes that skewed rewards in creative industries indicate that participants overestimate their chances of success. This appears to have been anticipated by Schumpeter (n 21) 73–74.

Thought Experiment: Incumbent Response to New Entry  67 In this regard, we are assuming that economies of scale exist, and the monopolist has succeeded on its merits, so the firm is efficient (and we have no reason to believe the demand curve will change, or shift, when and if new entry occurs). In the preceding section of this chapter, I critiqued the hypothetical ‘competitive level’ as a basis to assess the efficiency of monopoly and suggested instead that we should start from the actual position in which the monopolist finds itself, ie, the profit-maximising price and output, and assess whether it would be efficient to increase output beyond that point. In real-world markets, we cannot assume that drawing resources away from other uses has no effect on social welfare. It is only in the assumed conditions of the perfectly competitive equilibrium that productive and allocative efficiency are optimised in all markets simultaneously, ie, both privately and socially. Thus, it may well be that increasing output beyond the profit-maximising point would have a negative effect on social welfare and that it would be inefficient to draw resources from other uses in the economy simply to redistribute surplus from producers to consumers in a particular market. It is for this reason that Chapter 3 will propose that a monopolist’s profit-maximising output can be justified not because it is logical or rational, but because it can be seen to be efficient in this sense. We can also see that increasing output beyond the point at which marginal revenue becomes negative would be even more inefficient because market demand response would not justify the increased output. We can illustrate this schematically in Figure 1.5 below by examining the properties of the straight-line demand curve commonly used in the discussion of monopoly.131 In the upper part of Figure 1.5, we see the typical demand curve, AR, relating price to quantity. As price (average revenue) is falling, marginal revenue is lower than price and is also falling, and becomes negative at some point. In other words, the total revenue generated by sales in this market rises at a diminishing rate, reaches a peak where MR becomes zero, and falls thereafter as output increases. It is suggested that the ‘size of the market’ is indicated by the output at which MR = 0. While the price is still positive beyond that point, revenue is falling and it follows that no expenditure of resources can be socially justified to increase output beyond that point. I will consider further in Chapter 3 the forensic issues associated with testing the efficiency of increasing output beyond a monopolist’s profit-maximising output, including whether market demand response justifies expanding production. We know that the profit-maximising output relies on assessing marginal cost, which is not directly observable, and consequently raises difficult forensic issues. On the other hand, there may be forensic advantages in assessing demand response, ie, whether market revenue increases or decreases in response to price/output decisions.

131 See, eg, L Bumas, Intermediate Microeconomics: Neoclassical and Factually-Oriented Models (New York, Routledge, 2015) 39–40.

68  Market Power in Economics and Law Figure 1.5  Demand curve, marginal revenue and total revenue Price

AR MR Quantity Total revenue

Quantity

It is clearly geometrically possible to construct demand curves where marginal revenue remains positive (ie, sales increase) over a wider range as output increases in response to falling prices.132 We might expect that the range of such performance would be limited by consumers’ income and capacity to consume any single product. The problem faced by empirical attempts to estimate these characteristics of demand curves is that economic theory requires us to hold constant all other influences on demand, such as income, consumer preferences, advertising expenditure, prices of substitute goods and other unobservable factors, all of which are difficult to determine.133 In fact, demand curves must be constantly in flux as population grows and income and consumer preferences change. Nevertheless, we can draw some conclusions regarding ‘efficiency’ from

132 ibid 40–41. 133 See, eg, WB Allen, N Doherty, K Weigelt and E Mansfield, Managerial Economics, 6th edn (New York, WW Norton, 2005) ch 5, ‘Estimating Demand Functions’; S Berry and A Pakes, ‘Classical Empirical Models of Static Equilibrium’ (2003) Yale University, 2–3, available at pdfs.semanticscholar.org.

Thought Experiment: Incumbent Response to New Entry  69 the analysis in Figure 1.5 without the need to speculate on whether demand curves in reality have these properties. First, we might ask whether there is unmet demand in the market, ie, that could profitably be satisfied. In the case of a monopoly, where the firm is the market, the firm could be expected to have data that might help us answer this question. We need to ask here whether the market is working efficiently, ie, whether supply responds to demand shifts to bring the market into equilibrium, given we know that under imperfect competition, supply responds not to price, but to marginal revenue. We may infer that the efficient working of the market is somehow related to the size of the market and, through marginal revenue, to the demand response function. Second, where there are competing firms, we can envisage the possibility that there might not be unmet demand, but that zero-sum competition may lead firms to seek increased output at the expense of other firms in order to achieve sales growth and economies of scale. Given the realities of barriers to exit, firms do not have the option of cost-free exit, so must fight for survival. This may be inefficient if as a result output is increased beyond the point at which marginal revenue across the market becomes negative. Third, under zero-sum competition, if an equally or less efficient firm is forced out of the market, we may conclude that the success of the surviving firm, even with such a consequence, is efficient. However, there is a social cost to causing the exit of equal or less efficient firms in this way, ie, business failures cause knock-on losses to employees, suppliers and customers etc which are external to the firm. We may wonder whether the gains from survival of the more efficient firm outweigh the losses from the failure of the less efficient. We also need to be cautious as to whether the market would have excluded an equally efficient firm or whether the exclusion results from strategies of the dominant firm that can be condemned because such strategies are not a normal incident of competition, or whether exclusion is a consequence of the competitive paradigm itself. On the other hand, the exclusion of a more efficient firm is a clear indicator of harm to economic efficiency. The implications of zero-sum competition for new entry apply to every firm at the margin and are not limited to the monopolist and the new entrant. Of course, under perfect competition, firms enter and exit without costs, and since each possesses perfect information, incumbents can exit without incurring losses. In the real world, these assumptions are not borne out and so we encounter barriers to both entry and exit. The incumbent may not be able to match the technical features of the new product or to sell the business and deploy resources elsewhere without loss to shareholders. An incumbent facing its own failure and loss might have an incentive, even a duty to shareholders, to protect the value of its assets. These problems do not arise in the perfect world. So we can expect in the real world a contest for market share, increased costs of advertising and product differentiation which might raise the stakes for both and ensure pyrrhic victory (ie, if average cost curves rise as a result of competition).

70  Market Power in Economics and Law If we accept that there are no policy issues arising from the exclusion of less efficient firms or equally efficient firms that are not required to satisfy market demand, should we conclude that we are indifferent as to whether market power is used or misused in the above scenario? Is there market power at play at all? Should we attempt to determine empirically whether the economic gains from survival of the more efficient firm outweigh the losses from the failure of the less efficient firms? It seems evident that consumer welfare will be increased by the entry of an equally efficient firm, as output may rise and price may fall. However, our purpose is to consider: first, whether the incumbent has market power and, if so, whether it has used or misused its market power; and, second, whether the increased output is efficient (ie, whether the cost of applying resources to the increased production is less than the value created). It is proposed that the mere fact that the firm faces a downward-sloping demand curve does not mean that the incumbent has market power, and, further, the fact that a new entrant is emerging may suggest that any market power evaporates at that point. However, assuming the incumbent does have market power, can we distinguish permitted use from misuse? In this regard, Daniel Simon observes that ‘theories of entry deterrence and incumbent response to entry … [yield] a variety of predictions’ and the results of empirical studies are ‘very inconsistent’.134 Based on our thought experiment concerning new entry and the expected response of an incumbent monopolist set out in this chapter, we may conclude that a new entrant would be induced by the incumbent’s excess profits to enter the market only if the new entrant expects to be able to destroy the incumbent. Given there are barriers to exit in the real world, it is not unreasonable to expect the incumbent to fight for survival. It is hard to discern a normative case for competition law to condemn either the defensive conduct of the incumbent or the predatory conduct of the new entrant in this scenario. We can now consider whether norms of conduct from other areas of law – eg, self-defence in criminal law or voluntary assumption of risk in tort law – offer any assistance when competitors in zero-sum, or survival, competition seek to destroy each other.

VI.  Norms in Zero-Sum Competition Our focus is to consider the norms involved when the concept of competition shifts from ‘non-zero-sum’ competition in the model of perfect competition. In the latter, no competitor is harmed when one firm increases its sales, whereas in ‘zero-sum’ competition, a firm increases its sales relative to competitors only by taking away sales from its competitors and hence by harming them. As has been

134 D

Simon, ‘Incumbent Pricing Responses to Entry’ (2005) 26 Strategic Management Journal 1229.

Norms in Zero-Sum Competition  71 observed, the kind of injury arising from zero-sum competition has never been a tort.135 Similarly, in Bryan v Maloney, Mason CJ and Deane and Gaudron JJ observed that: [I]n a competitive world where one person’s economic gain is commonly another’s loss, a duty to take reasonable care to avoid causing mere economic loss to another … may be inconsistent with community standards in relation to what is ordinarily legitimate in the pursuit of personal advantage.136

We can draw on Neil MacCormick to propose that law is simply a system of normative rules.137 In many cases, laws prescribe standards of behaviour (eg, tort law), but in the case of competition law (and in many other areas of law from criminal law to fiduciary duties), certain conduct is proscribed, but norms of permitted conduct are not explicit, and so remain in the background. In the case of misuse of market power, we are searching for norms of conduct that may, in addition to clarifying the concept of market power and the proscribed use of market power, reveal defences that ought to be recognised explicitly by the law. Clearly, Sunstein is correct that ‘private autonomy’ cannot be a relevant norm by which to interpret regulatory statutes, since by definition the purpose of regulation is to limit private autonomy.138 However, some of the ‘constitutional norms’ Sunstein proposes demonstrate that conceptions of the relationship between citizen and state remain relevant and can be distinguished from the concept of ‘private autonomy’, which he considers to be anachronistic.139 He argues similarly that ‘the market’ cannot be a norm where regulations promote ‘non-market values’.140 I suggest that Sunstein’s insight may explain the ‘bipolarity’ in competition law that is demonstrated by this study, ie, we wish to promote the undistorted operation of markets, yet we appear to be applying non-market values when we respond to resulting harm. The ‘business judgement’ rule that applies to directors’ duties in corporations law provides an example of another relevant norm. Here the law limits the extent to which courts should second-guess business decisions. Company directors must carry out their duties with the degree of care and diligence that a reasonable person would exercise.141 Subject to important requirements including good faith, directors are taken to comply with their duties in relation to business judgements where they ‘rationally believe’ that the judgement in question ‘is in the best interests of the corporation’.142 Such a belief is defined to be rational ‘unless the belief is



135 Queensland

Wire (n 3) 191 (Mason CJ and Wilson J). v Maloney (1995) 182 CLR 609, 618. 137 N MacCormick, Rhetoric and the Rule of Law (Oxford, Oxford University Press, 2005) 47. 138 Sunstein (n 50) 444; see also MacCormick (n 137) 47. 139 Sunstein (n 50) 504. 140 ibid 453 and 485–86. 141 See s 180 of the Corporations Act 2001 (Cth). 142 ibid s 180(2). 136 Bryan

72  Market Power in Economics and Law one that no reasonable person in their position would hold’.143 Thus, there is an ­objective norm applied in this case by reference to the reasonable company director, but the norm is stated in such a way as to suggest that the court should not substitute the court’s own views.144 The legislative purpose is to ensure that company ­directors act in the interests of the companies they serve. The legislative purpose in the case of competition law is more complex. Sunstein also correctly points out that the concept of legislative purpose may be incoherent, since purposes are often multiple and conflicting.145 This is particularly true in the case of competition law, as we have seen in the Introduction in relation to Australia’s process of reforming section 46. Just as statutory purposes can be multiple and conflicting, Sunstein concedes that there may be disagreement regarding norms, though he considers that there may be some that should be widely shared.146 We will consider especially those that may manifest themselves in certain legal doctrines outside the field of competition law. Phillip Landolt mentions possible defences derived from other areas of law (ie,  self-defence in criminal law and voluntary assumption of risk in tort law), noting that they are not yet accepted in EU competition law.147 In fact, the two defences are related, in that a person who is attacked is generally thought to be entitled to take reasonable steps to protect themselves, and an attacker is by ­breaking the law assuming a risk of retaliation.148 These areas of law may be seen as attempts by the state to regulate dealings between citizens, in which the state should not take sides, all other things being equal. The criminal law in my jurisdiction, the State of Victoria, on the subject of self-defence has undergone controversial changes in recent years, which are not necessary to consider in detail.149 It seems that the initial threat to the person

143 ibid. 144 See also, eg, A Mason, ‘The Pursuit of Excellence in Tribunal Decision-Making in Australia, the United Kingdom and Canada’ (2016) 29 Canadian Journal of Administrative Law and Practice 235, 238, discussing the test in Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223, 230 for judicial review of administrative decisions on grounds of unreasonableness. 145 Sunstein (n 50) 446 and 448. 146 ibid 466. 147 P Landolt, Modernised EC Competition Law in International Arbitration (The Hague, Kluwer Law International, 2006) 344, noting the 2001 ECJ decision in Case C-453/99 Courage v Crehan, 20  ­September 2001, where it was held that the plaintiff ’s participation in an anticompetitive (and hence unlawful) agreement did not disqualify the plaintiff from recovering under EU law, but neither did it preclude the national law from providing a defence where the plaintiff ’s own unlawful actions bore significant responsibility for any anticompetitive harm. I do not consider that the EU ‘passing on’ defence to a private action for damages, eg, for cartel pricing (Directive 2014/104/EU [2014] OJ L349/1), assists the present discussion of possible defences to liability for misuse of market power, because even if a private plaintiff passes on excessive prices, there would still ultimately be economic harm which would expose the firm to liability for a pecuniary penalty for contravening the law. Further, in Australia, unlike the EU, s 46 does not currently prohibit excessive pricing. 148 O Bakircioglu ‘The Contours of the Right to Self-Defence’ (2008) 72 Journal of Criminal Law 131, 132. 149 See, eg, L Finlay and T Kirchengast, Criminal Law in Australia (Sydney, LexisNexis Butterworths, 2015) 366–71.

Norms in Zero-Sum Competition  73 need not be unlawful and that the test for the defence is part subjective and part objective: the belief in the need to respond to the threat is subjective, so a very nervous person may hold a belief that others might not; and, while at common law there was no requirement that the defensive action be proportionate, ie, o ­ bjectively reasonable, in States adopting the Model Criminal Code, the action must be ‘reasonable’.150 In the context of competition law, Hovenkamp, for example, argues that abuse of market power involves conduct which limits the opportunities of rivals and produces harm ‘seriously disproportionate to the resulting benefits’.151 In the context of judicial review of administrative decisions, Australian courts appear to have avoided assessing whether a disproportionate decision could be regarded as objectively unreasonable.152 However, in that context, the issue concerns the institutional constraint that courts should not engage in a merits review of administrative decisions.153 Similar questions arise when the courts consider the constitutionality of legislation limiting personal freedoms, where the court may consider the rational connection between the limitation and its purpose, the availability of reasonably practical alternatives, and the balance between the limitation and its purpose.154 In that context, the issue concerns the institutional constraint that courts should not ‘enter into the realm of the legislature’.155 Accordingly, in the context of competition law, the relevant constraint on implementing Hovenkamp’s suggestion would be the practicalities of the court second-guessing business decisions. In zero-sum, or survival, competition where one firm’s gain is another’s loss, we may doubt that ‘proportionality’ can assist us if, for example, we accept that competition may be expected to threaten the firm’s survival and that it is necessary to take sales away from its competitor (which may necessarily reduce its scale and impact more than proportionately on profitability). To ask whether the firm’s strategy was objectively reasonable, or proportionate, would require finding that ‘peaceful coexistence’ is possible, ie, that survival was not threatened and/or that sales could be increased without taking sales away from rivals. I propose that it would be better to focus on the relative efficiency of the rivals, ie, to ask if the firm that exits was more efficient than the survivor, rather than to ask if the conduct of the dominant firm was objectively reasonable or proportionate. The two protagonists in our thought experiment about incumbent response to new entry are like persons engaged in contact sports, where there is some voluntary acceptance of risk in the context of rules of the sport which aim to



150 ibid.

151 Hovenkamp 152 Mason 153 ibid. 154 See,

(n 24) 152. (n 144) 243.

eg, Murphy v Electoral Commissioner [2016] HCA 36 [36]–[38] (French CJ and Bell J). v Electoral Commissioner [2016] HCA 36 [39] (French CJ and Bell J).

155 Murphy

74  Market Power in Economics and Law protect the players from serious injury. In the context of competition, as the High Court recognises, injuries such as loss of profits and catastrophic business failure are the inevitable consequence of conduct which deliberately seeks to advantage one party at the expense of the other.156 Each party accepts the risk of failure – in the real world, where there is imperfect information about competitors and market factors, it is likely that issues of consent would be even more problematic to define than, eg, in tort law.157 Consent by sportspersons would usually be limited to lawful conduct within the rules of the game (or incidental infringements of the rules), though not to deliberate violence unconnected with the play.158 This may be appropriate and determinable where the ‘rules of the game’ clearly define the relevant norms of conduct. In Australia, the tort of inducing breach of contract recognises a defence of justification. This requires more than mere self-interest. To qualify, the ­wrongful act must be ‘no more than reasonably necessary for the protection of some actually existing superior legal right’, since this indicates against the intentional element of the tort.159 Thus, while there are analogies with issues encountered in other areas of law, we do not find approaches that can readily be adapted to our present context. However, the preceding examples suggest that defences in other areas of law implicitly acknowledge that conduct captured by the legal theory of liability may not necessarily be blameworthy. This would seem to be the case especially where the relevant legal theory of liability adopts a proxy, or surrogate, test, since the surrogate and that which it represents may well not coincide. As we will see in Chapter 3, the competition test (or SLC test) is a proxy, or surrogate, for efficiency and in some cases it is accepted that ‘competition’ is a poor surrogate for efficiency. Further, the problem we are addressing in competition law is that section 46 does not explicitly identify ‘rules of the game’ or norms of conduct. Sunstein’s analysis of ‘naked transfers’ suggests that regulatory intervention requires something more than private gain and loss, which suggests the public interest as something beyond the private interests of the protagonists – which may lead us to the most efficient, or productive, use of society’s resources as a measure of the public interest. That is, we have no basis to intervene in the case of transfers of surplus between producers and consumers unless the transfer causes a social gain or loss outside the market in question, eg, it would not be justified to draw resources away from other productive uses to subsidise consumers in the market in question. Thus, under the market manipulation approach, we want to be ­satisfied

156 Queensland Wire (n 3) 191 (Mason CJ and Wilson J). 157 B Richards, M de Zwart and K Ludlow, Tort Law Principles (Melbourne, Lawbook Co, 2013) 148–50. 158 ibid. 159 See Zhu v The Treasurer of the State of New South Wales (2004) 218 CLR 530 [105], [114] and [138]–[144].

Zero-Sum Competition, Market Manipulation and Efficiency  75 that the market has not been distorted by supply being restricted and price raised. This brings us to the dominant firm’s right to profit from its success, which appears to be generally regarded as a fundamental norm. We will return to the question of norms of conduct when we consider efficiency and other defences recognised in Europe and North America in Chapter 3.

VII.  Zero-Sum Competition, Market Manipulation and Efficiency In the present chapter, we have developed our paradigm of zero-sum or survival competition. In summary, zero-sum competition entails: a. firms competing for sales by taking sales away from each other or seeking to grow sales faster than the market (which gives rise to the ‘dual purposes’ problem, ie, it is virtually impossible to distinguish between self-interested and malicious conduct); b. entry and exit is not free, ie, incumbent firms incur costs on exit that encourage them to fight for survival rather than exit the market; c. new entrants may overestimate their chances of entering the market and successfully taking sales away from the incumbent; d. all firms, including new entrants, have some degree of power, through their price/output decisions, to instigate a demand response. I suggest that the proposed paradigm of competition affects how we conceptualise market power and assess its misuse. I suggest that market power can be legally conceived as power to manipulate the market. We want to ensure that the process by which demand and supply adjust to each other is undistorted. In the present chapter, I suggest that ‘efficiency’ can be adapted to provide us with a basis to assess whether conduct is manipulative, ie, inefficiency interferes with the operation of the market and so may be condemned. Our central forensic enquiry is to ask whether there is ‘unmet demand’ which could be profitably supplied by new entry. This concerns: first, whether it would be profitable for the dominant firm to increase output (presumably not if it is producing profitmaximising output); and, second, what is the size of the market, ie, the level of output beyond which demand response fails to justify the expenditure of additional resources (illustrated in Figure 1.5). We will develop in Chapter 3 the ideas we need concerning market manipulation and efficiency to distinguish conduct which is manipulative, and how this can be forensically determined. Some key principles we can take forward from the present chapter are as follows: (1) A monopolist may be expected to meet market demand. This is driven by profit-maximising behaviour. We can justify such conduct not because it is rational, but because it is efficient in the sense described earlier.

76  Market Power in Economics and Law (2) When demand is being satisfied by the incumbent, new entry may not be efficient, ie, the size of the market may preclude new entry. (3) Zero-sum competition will not necessarily produce efficient outcomes, so we need a normative system to assess whether we should condemn conduct of firms that leads to inefficient outcomes. (4) Exclusion of a more efficient firm would appear to be manipulative, as we expect more efficient firms to enter and survive. However, we anticipate that more efficient firms can fail through underestimating risks. We anticipate that incumbent firms face exit costs that may justify a decision to stay in the market and fight to survive, eg, by becoming more efficient.

VIII. Conclusion The generally accepted legal concept of market power, as the ability to reduce output and increase price, is derived from a comparison of the neoclassical economic models of perfect competition and monopoly, which assumes that one firm is no more efficient, or productive, than many smaller firms under the model of perfect competition. Accordingly, it may well be that our current understanding of market power is the product of a value judgement or worldview. Further, the economic models are static analyses of equilibria and do not answer our questions about how firms enter and leave the market under dynamic real-world conditions. Nor does the neoclassical market mechanism explain how markets work efficiently to bring about equilibrium under imperfect competition when supply responds not to price, but to marginal revenue. I suggest that zerosum competition, as exemplified in this chapter, is a useful paradigm in which to develop a positive and normative legal theory of market power. Under zerosum competition, where growth in revenue and profits becomes the measure of success, it can be posited that normal business conduct will involve strategies that increase revenue and profits at the expense of rivals. I suggest that such strategies can be justified not because they are rational, but because they may be an efficient response to market demand. Other strategies may well involve manipulation of the market and thus require explanation. The market itself excludes firms for reasons of efficiency: if there is oversupply, firms leave the market so that those that remain cover their costs and earn normal profits. This is a natural part of competitive markets, which makes it difficult to distinguish situations where a competitor has been excluded otherwise than due to market forces. The paradigm of perfect competition, in which no firm has market power and competitors are not harmed by competition, would appear to have little to offer us in our efforts to understand normal conduct in real-world markets. I suggest that the paradigm of zero-sum competition offers us a positive basis to reconceptualise market power as market manipulation, and a normative basis to assess conduct which is manipulative, ie, interferes with the efficient

Conclusion  77 operation of the market, and so should be condemned. We can begin to see that demand response plays a key role both in our legal conception of market power and our legal conception of efficiency. The debate about the policy objectives of competition law is notable for two reasons. First, we see not only that redistribution is largely rejected as an ­appropriate objective, but also that competition laws are not capable of achieving the objectives of equity and social justice. The policy implications are that where policy-makers wish to pursue these objectives transparently, there is a need for policy measures other than competition laws. Second, when we focus on efficiency as the ultimate goal of competition law, it can be seen that it will become necessary to gain a better understanding of how we may forensically determine efficiency. In particular, we want to know how the market mechanism works in zero-sum competition, ie, how supply responds to demand shifts by firms entering or leaving the market. Our review of the key cases in Part II will help us develop this analysis. Under the Harper Review section 46, we still need to be able to distinguish between the normal operation of the market, which may exclude or harm competitors, from situations where harm flows from the impugned conduct of the dominant firm, whether it be conventional ‘misuse of market power’ or ‘market manipulation’. I suggest that the use of static economic models of perfect competition and monopoly, which are not comparable, to identify market power and conduct that would not be expected in a hypothetical competitive market, as a basis to determine whether an effect does or does not flow from the conduct, would continue to be a significant defect in Australian competition law. In Chapter 3 we will take up a detailed discussion of the Harper section 46 and how competition and efficiency effects may be determined. The discussion of the literature on law and economics in this chapter enables us to propose an alternative legal concept of market power and to propose an alternative paradigm of competition as a framework to understand market power and regulate its misuse. The next chapter will examine the foundational role that ‘structuralist’ thinking has played in Australian competition law, ie, the idea that competition effects can be inferred from market structure and changes in market structure. As this approach has fallen out of favour around the world, attention has turned to ‘direct assessment’ of competition effects. The literature indicates that no clear method of direct assessment has emerged. Our main focus, however, will be on how structuralist thinking has promoted ‘competition’ as a surrogate for ‘efficiency’ and the institutional division of functions has generally limited the courts to determining competition issues, whereas the assessment of broader questions of efficiency and the public interest, and weighing up adverse competition effects and offsetting efficiency effects, has been reserved to the ACCC and the ACT. We will consider whether this division of functions is necessary or desirable, having regard to the institutional and constitutional capabilities of the courts.

2 The Mischief and Australia’s Institutional Response I. Introduction Economic teaching of the 1960s and the following decades in Australia was dominated by the neoclassical model and policy strictures against high industry concentration (ie, few sellers).1 As we have seen, high industry concentration still seems to be of concern to regulators, and some argue that the purpose of regulating market power is to prevent the maintenance of high concentration.2 We have also seen that such concerns are associated with the SCP approach, which held that the presence of dominant firms weakens the competitive process, but that economic theory now accepts that competition can function well even with large firms.3 In this chapter we examine how SCP thinking still affects the way in which Australian competition law views market power and its misuse, and the division of functions between the courts and the administrative agencies, the ACCC and the ACT.

II.  The Continuing Influence of the SCP Approach in Australia As we will see in Chapter 3, the courts in the US moved away from the position in the 1960s that atomistic competition is to be promoted even at the expense 1 See, eg, PH Karmel and M Brunt, The Structure of the Australian Economy (Melbourne, FW  ­Cheshire, 1966); R Caves, American Industry: Structure Conduct and Performance (Englewood Cliffs, Prentice Hall, 1967). 2 See R Sims, ‘Is Australia’s Economy Getting More Concentrated and Does This Matter?’, speech delivered at the RBB Economics Conference, Sydney, 27 October 2016, www.accc.gov.au/speech/ keynote-address-rbb-economics-conference-0; and A Gavil, ‘Imagining a Counterfactual Section 36: Rebalancing New Zealand’s Competition Law Framework’ (2015) 46 Victoria University of Wellington Law Review 1043. 3 See G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011) 180; MS Gal, Competition Policy for Small Market Economies (Cambridge, MA, Harvard University Press, 2003) 55.

The Continuing Influence of the SCP Approach in Australia  79 of ­efficiency.4 At the time of writing, the position in Australia seems to be that, while there is no authoritative appellate decision on the point, observations of influential appellate judges suggest they consider that the CCA endorses a more atomistic view of competition.5 We have seen that the SCP approach fell out of favour in the US and elsewhere about the time it was being applied in Australian competition law in the 1970s.6 According to Corones, our courts and the ACCC/ACT still apply a modified SCP approach.7 For the purposes of this study, structuralist thinking is relevant in four respects. First, under the SCP approach, ‘market structure is the primary source of market power’.8 I suggest that this concept of market power derives from the paradigm of the hypothetical competitive market and can be reconceptualised in the context of zero-sum competition. Second, the SCP approach provides a chain of causation (under which market structure determines market conduct which determines market performance, ie, efficiency outcomes), which it is said enables inferences to be made about market performance from observations of market structure and conduct.9 Thus, the ACT in 1976 explained that the ‘process of competition’ is very much a matter of market structure.10 However, the ACT has more recently said that this approach has been overtaken by developments in economic theory: market structure does not determine conduct and performance; regard must be had to strategic­ behaviour.11 Nevertheless, we still have a situation where the ‘process of competition’ is a proxy, or surrogate test, for efficiency and market structure is considered to be a key determinant of the process of competition. Third, we see that the mischief that SCP addresses with respect to market power is that market power is said to maintain high industry concentration and by that means to produce inefficient outcomes. I suggest that we ought to focus on efficiency, as harm to efficiency is the principal mischief which the SCP approach seeks to address. Carl Kaysen and Donald Turner, while advocating the SCP approach, are clear that ‘in so far as reduction of market power is incompatible with efficiency’, efficiency should prevail.12 4 See, eg, D Evans and K Hylton, ‘The Lawful Acquisition and Exercise of Monopoly Power and its Implications for the Objectives of Antitrust’ (2008) 4 Competition Policy International 203, ssrn.com/ abstract=1275431, discussing the famous opinion of Learned Hand J in US v Alcoa 148 F 2d 416 (2d Cir 1945). 5 See Rural Press (2003) 216 CLR 53 [49] (Gummow, Hayne and Heydon JJ); ACCC v Flight Centre Travel Group Ltd [2016] HCA 49 [141] (Nettle J); compare the observations of the majority in Melway (2001) 205 CLR 1 [52]. 6 H Hovenkamp, Federal Antitrust Policy (St Paul, West Publishing, 2005) 42–47. See also S Stroux, US and EC Oligopoly Control (The Hague, Kluwer Law International, 2004) 10–12; T Muris, ‘Improving the Economic Foundations of Competition Policy’ (2003) 12 George Mason Law Review 1, 3–6. 7 S Corones, Competition Law in Australia, 6th edn (Sydney, Lawbook Co, 2014) 41. 8 ibid 23. 9 ibid 23–26. 10 QCMA (1976) 8 ALR 481, 516. 11 Application by Chime Communications Pty Ltd (No 3) (2009) ACompT 4 (hereinafter ‘Chime’) [11]. 12 C Kaysen and D Turner, Antitrust Policy: An Economic and Legal Analysis (Cambridge, MA, Harvard University Press, 1959) 45.

80  The Mischief and Australia’s Institutional Response Fourth, the SCP approach has implications for the forensic determination of competition and efficiency effects. Proponents of SCP consider that market structure and conduct (which the ACT and courts traditionally equate with the ‘process of competition’) provides the only test which is capable of being administered by the courts.13 We know the resulting test as the SLC test. Indeed, Kaysen and Turner suggest that neither ‘efficiency’ nor ‘competition’ is sufficiently determinate and prefer market structure as the appropriate surrogate test.14 As we will see in Chapter 3, many theorists believe there is a need for direct measures of competition effects that do not rely on structuralist reasoning.15 Nevertheless, these doubts about the capability of the courts are shared by influential ­Australian judges.16 Thus, the view that courts are not capable of determining economic effects such as efficiency appears to be entrenched in the CCA by virtue of the division of functions between the courts and the administrative agencies, the ACCC and the ACT. The CCA reserves to those agencies the power to authorise conduct which fails the SLC test, but which results in overriding public benefits such as efficiency.17 The authorisation process has not previously applied to section 46; however, the legislative changes being introduced to implement the Harper Review recommendations will allow it. We will consider efficiency, and how we may forensically determine it, in more detail in Chapter 3. For the present purposes, we may regard the debate about the objectives of competition law, or the original intention of the legislatures, in Australia and overseas as somewhat unproductive. For example, Howard Shelanski notes attempts in the US to resolve the debate about objectives by reference to the presumed legislative intention behind the Sherman Act in the 1890s, and that while efficiency is increasingly recognised in the US as a defence, debate continues over which kinds of efficiencies may be considered.18 The concern of this monograph is with the consequences of the law, and law reform initiatives. It is recognised that conduct which lessens competition may nevertheless be justified because it has efficiency or other benefits. I suggest it is equally important to ask whether our real-world paradigm of competition, zero-sum competition, may produce inefficient outcomes. We may suspect that the key SCP thinkers mentioned in the present chapter would respond to the effect that, if this were the case, we need to re-adjust the way we apply the SLC test. 13 Corones (n 7) 23, citing E Mason, ‘The Current Status of the Monopoly Problem in the United States (1949) 62 Harvard Law Review 1280. 14 Kaysen and Turner (n 12) 59. 15 See, eg, Muris (n 6) 14. 16 See, eg, French J in Australian Gas Light Company v ACCC (No 3) [2003] FCA 1525 [607]. 17 M Brunt, ‘The Australian Antitrust Law after 20 Years: A Stocktake’ (1994) 9 Review of Industrial Organisation 483, 499. Corones (n 7) 41 says the same in more qualified terms: ‘the courts, for the most part, are required to assess the effect or likely effect of conduct or the competitive process without regard to efficiency, except where the efficiencies will have a bearing on the future state of competition in the market’. We will discuss in ch 3 differing views on the latter issue. 18 Howard Shelanski, ‘Efficiency Claims and Antitrust Enforcement’, in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 2, 452 at 457.

Australia’s Institutional Division of Functions between Courts and Agencies  81 The remainder of this chapter addresses the institutional division of functions. I suggest that ‘efficiency’ as we can adapt it for the proposed legal theory is capable of being determined by the courts or at least is no less determinate than the ‘process of competition’ that is addressed by the SLC test.

III.  Australia’s Institutional Division of Functions between Courts and Agencies The precursor to the current CCA was the Trade Practices Act 1965 (Cth). Under that scheme, restrictive trade practices would be reviewed by the regulator and if held by a tribunal to be contrary to the public interest, the tribunal could make restraining orders that were enforceable through judicial process.19 In a paper assessing the success of Australia’s competition law after the first 20 years, Maureen Brunt explains the reasoning behind that approach: in the early days of competition law in Australia, it was unclear to legislators how and why restrictive trade practices might be justified.20 Clearly, this presumes there may be some justification. Brunt notes that the public interest criterion was retained in the 1974 version of the CCA, deliberately casting a wider net than ‘efficiency’.21 Brunt appears to favour excluding efficiency from the jurisdiction of the courts on the grounds that the standard of legal liability ought to be ‘positive’, whereas efficiency raises ‘normative’ issues.22 Yet Brunt considers that efficiency may be relevant to section 46 in either or both of the ‘taking advantage’ and ‘exclusionary purpose’ requirements.23 Now, of course, these aspects of section 46 have been removed by the Harper Review amendments. Brunt considers that ‘legitimate business justification’, if read as ‘production efficiency … may well have a role in Australian antitrust, despite the presence of authorisation’.24 As we will see in Chapter 3, Australian courts have subsequently given some credence to legitimate business rationale, without making the connection with efficiency that Brunt contemplates. It may well be that the Harper amendments, absent any reference to efficiency following that aspect being removed by Parliament, may raise further obstacles to courts interpreting section 46 in a way that would make efficiency relevant. However, I will suggest in Chapter 3 that there should still be some

19 R French ‘The Role of the Courts in the Development of Australian Trade Practices Law’ in F Hanks and P Williams (eds), Trade Practices Act: A Twenty-Five Year Stocktake (Sydney, Federation Press, 2001) 103. 20 Brunt (n 17) 506. 21 ibid 507. 22 ibid 510–11. 23 ibid 512. 24 ibid. This point is also made by E Mason, ‘Market Power and Business Conduct: Some Comments’ (1956) 46 American Economic Review 471, 479.

82  The Mischief and Australia’s Institutional Response causal connection between market power and SLC under the Harper section 46, which may open the door to efficiency, though this is not free from doubt. Brunt goes on to discuss the limits of court-centred competition law, quoting Edward Mason to suggest that courts can only apply competition law if ‘valid inferences can be made from relatively simple statements of fact’ and that ‘an attempt to push enquiry into effects very far is clearly an invitation to non-enforcement’.25 The reference to valid inferences may well suggest the kind of inferences facilitated by the SCP approach. He observes that how far any enquiry into ‘economic consequences’ needs to go depends on ‘what confidence courts should have in the process of economic analysis’.26 It is now over half a century since Mason expressed that concern, and Brunt expressed concerns in relation to the Trade Practices Act 1965 regarding the lack of experience with business practices that may be justified. I suggest that by now, we should have greater confidence in the ability of economics and the courts to address the economic consequences of business conduct. I will suggest in Chapter 3 that economic efficiency can be adapted to become the centrepiece of both a positive and normative legal theory of market power. It seems reasonably clear from Mason’s discussion of ‘effects’ that he uses the term to embrace both competition and efficiency effects which can be revealed by ‘economic enquiry’.27 However, he makes it clear that, in his view, ‘we expect from the competitive system a set of effective limitations to the growth of private economic power’.28 This objective is one that the words of Australia’s section 46 (and apparently section 2 of the Sherman Act) manifestly do not contain. Mason acknowledges that ‘the importance of efficiency as a desideratum inevitably condemns any purely “limitist” interpretation of antitrust policy’.29 A few years after Mason published that paper, Carl Kaysen and Donald Turner, while advocating the SCP approach, stated that ‘we would discard the general limiting of big business power as an independent goal of antitrust policy’.30 I suggest it is time to let go of the remaining vestiges of the SCP approach embedded in the CCA that impede consideration of economic efficiency.

A.  The Role of the Court At this point, we need to revisit the views of Robert French (before he became Chief Justice of the High Court) about the objectives of competition law and the role of the courts. He refers to a dimension of competition law being ‘the potential



25 Brunt

(n 17) 516–17; Mason (n 24) 478 and 475. (n 24) 477. 27 ibid 474–76. 28 ibid. 29 ibid 479. 30 Kaysen and Turner (n 12) 49. 26 Mason

Australia’s Institutional Division of Functions between Courts and Agencies  83 for social engineering of commercial behaviour’ which appeals to ‘aesthetic sensibilities with the vision of a better world’.31 Clearly, economists of the SCP school have been greatly concerned by the growth in economic power of big business, though with the rise of globalisation, expressions of concern from the 1940s to the 1960s regarding, eg, the impact of big business on ‘village life’ may seem quaint today.32 French explains the normative role of judges in applying ‘standards’, eg, due care in torts law, as opposed to ‘rules’, and suggests that Part IV of the CCA sets out standards that require ‘normative and evaluative judgments in their ­application’.33 Further, the judicial function is constitutionally limited to finding facts and applying law to determine legal rights and obligations, to be distinguished from executive arms of government such as the ACCC and the ACT.34 French argues that the kinds of decisions involved in authorisation matters, such as public benefit or efficiency issues, are beyond the constitutional or institutional competence of the courts.35 In a subsequent judicial role on the Federal Court, French J describes authorisation proceedings as involving ‘polycentric decision-making of a kind which the Court is not institutionally competent nor authorised by statute or the Constitution to undertake’.36 Polycentric decision-making connotes discretionary decision-making, which may require trading off competing objectives. This is ordinarily associated with public interest functions of the executive branch of government as distinct from judicial functions of the courts under the rule of law.37 French also considers that neither the courts nor the regulators are equipped to micro-manage business decisions of market participants, eg, terms of supply in refusal to supply cases under section 46.38 French’s remarks relate specifically to authorisation proceedings, ie, concerning the administrative discretion of the ACCC/ACT to authorise conduct that would otherwise contravene Part IV of the CCA. These do appear to involve polycentric decision-making. However, we might ask why this is so. I suggest it may be because conventional components of economic efficiency such as productive and allocative efficiency are not congruent objectives under conditions of imperfect competition. Further, as we will see, authorisation proceedings permit consideration of broader public interest factors than economic efficiency.

31 French (n 19) 98. 32 See, eg, EAG Robinson, Monopoly (Cambridge, Cambridge University Press, 1941) 183–90; Karmel and Brunt (n 1) 55. 33 French (n 19) 105. 34 ibid 107. 35 ibid 108. 36 Australian Gas Light Company v ACCC (No 3) (2003) 137 FCR 317 [607]. 37 See, eg, C Finn, ‘The Justifiability of Administrative Decisions: A Redundant Concept?’ (2002) 2 Federal Law Review 239. An example would be a ministerial declaration under the access regime in pt IIIA of the CCA: see, eg, The Pilbara Infrastructure Pty Ltd v ACT (2012) 246 CLR 379, [2012] HCA 36. 38 French (n 19) 108.

84  The Mischief and Australia’s Institutional Response Clearly, in some legal contexts, courts are familiar with the type of evaluative process that involves weighing up factors that may point in different directions. To give one example, the statutory prohibition of unconscionable conduct spells out a number of factors that the court must consider when it determines whether conduct is unconscionable (formerly in Part IVA of the CCA, now found in s 22 of the Australian Consumer Law). Thus, while Brunt considers efficiency to have normative or evaluative implications, the courts are able to undertake such tasks in some legal contexts. However, there may well be a limit to which a theory of legal liability can embrace polycentricity. We might speculate whether the problems that have been encountered with the theory of liability for misuse of market power under past law arise from undisclosed issues of polycentricity, such as differing views about the economic and non-economic objectives of competition law or differing views about conflicting standards of welfare or efficiency. The same issues could well apply to the new section 46 as recently enacted. Suffice to say that the criticisms made by French J of polycentric decision-making as a judicial function would not appear to raise any impediment to a unicentric legal theory of liability, which is the focus of this monograph. I will argue in Chapter 3 that the determination of issues of efficiency, as we can adapt and redefine it for the purposes of our legal theory, can be seen to be no more than traditional fact-finding and need not involve weighing up offsetting competition and efficiency effects, if we accord primacy to efficiency as a higherorder objective of competition law.

B.  The Role of Administrative Agencies We do not address here questions of constitutional law which concern the validity of laws conferring broad, even polycentric, decision-making powers on administrative agencies.39 Nor do we address questions of constitutional and administrative law concerning the valid exercise of such powers. It seems that there is no constitutional barrier to Parliament conferring power on agencies to make polycentric decisions, even though (by definition) the outcome of such decisions cannot be predicted with certainty.40 However, it is recognised that regulatory uncertainty creates risk for business and adds to the costs of regulation.41 For the purposes of the present study, we will assume, without further consideration, that the authorisation process set out in the CCA is constitutional. Our enquiry considers whether a broad public interest test, which does not clarify the priority to be given to efficiency, may have undesirable consequences because it is unpredictable. 39 See, eg, L Crawford, ‘Can Parliament Confer Plenary Executive Power? The Limitations Imposed by Sections 51 and 52 of the Australian Constitution’ (2016) 44 Federal Law Review 287. 40 ibid 296–97. 41 See, eg, Oxford Economic Research Associates, Costs and Benefits of Market Regulators, Part 1: Conceptual Framework, report for Dutch Ministry of Economic Affairs, October 2004, 16.

Australia’s Institutional Division of Functions between Courts and Agencies  85 To pursue this enquiry, we can draw on the study by Vijaya Nagarajan of authorisation decisions by the ACCC/ACT under the CCA in the period from 1974 to 2010.42 Nagarajan finds that authorisation decisions have been justified on wide-ranging public benefit grounds, and consequently principles have not yet emerged to elucidate the role of efficiency.43 Clearly, there are issues about assessing efficiencies, which remain to be resolved, though we may conclude that mere transfers of surplus do not generate value for society.44 It appears that protection of small business has been accepted as a public benefit, though this may not be consistent with efficiency.45 I suggest that the authorisation process recognises that efficiency is a higherorder objective than the pursuit of competition. Thus, conduct which would be anticompetitive and a contravention of the law can be excused by the administrative agencies (but not the courts) if efficiency benefits result. However, it is not clear whether the CCA gives priority to efficiency over other public interest factors, eg, equity. Presumably by not mentioning efficiency, the CCA fails to give it priority.46 We saw in Chapter 1 that it may be doubted whether competition law is a suitable vehicle to address equity, since markets work by excluding producers and consumers based on cost-efficiency and ability to pay respectively. We may thus question whether authorisation should be possible on noneconomic grounds (or at least whether they should be allowed to prevail over efficiency) and whether the competition agencies are appropriately placed to undertake that assessment. Does the CCA authorise administrative agencies to, eg, sacrifice productive efficiency (national income) in order to redistribute surplus to consumers (or small business) in a particular market? That is, does the CCA contemplate authorising conduct that may be inefficient yet arguably justified on public interest grounds, such as redistributive equity? Those who argue that the objectives of competition law include distributive equity would presumably answer in the affirmative. However, I suggest that it is not obvious that the

42 V Nagarajan, ‘The Paradox of Australian Competition Policy: Contextualising the Coexistence of Economic Efficiency and Public Benefit’ (2013) 36 World Competition 133. 43 ibid 161–62. Curiously, Nagarajan (at 144) classifies ‘promotion of competition in industry’ as an efficiency justification for anticompetitive conduct. 44 ibid 142. 45 ibid 148. Gal (n 3) 49 rebuts the argument sometimes put forward that even inefficient competitors provide increased consumer choice; see also TPC v Email Ltd (1980) 31 ALR 53, from which we might infer that an inefficient competitor provides no effective competition or real choice. However, the ­European Commission, ‘Guidance on the Commission’s Enforcement priorities in Applying ­Article 82 EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings’ [2009] OJ C45/7 [24] expresses the view (which is not elaborated) that ‘in certain circumstances a less efficient competitor may … exert a constraint’ on dominant firms. D Geradin, A Layne-Farrar and N Petit, EU Competition Law and Economics (Oxford, Oxford University Press, 2012) [4.149]–[4.150] describe this approach as ‘regrettable’, since it appears to protect less efficient competitors and thus creates uncertainty. 46 Brunt (n 17) 507 notes that the Commonwealth of Australia, National Competition Policy (­Australian Government Publishing Service, 1993) (hereinafter ‘Hilmer Report’) 121 recommended that efficiency should be recognised as the primary consideration in authorisation matters, though the availability of other public benefits was not opposed. However, this has not been implemented.

86  The Mischief and Australia’s Institutional Response CCA does or should confer that authority on the competition agencies. Greater certainty would be provided if the legislation expressly gave guidance on priorities. We might also ask whether the CCA contemplates authorising conduct which is competitive (ie, there is no anticompetitive harm) but inefficient. We will see in Chapter 3 that this is not something that competition law generally contemplates;47 nevertheless, we must confront inefficiency when we analyse the consequences of zero-sum competition. On the basis of Nagarajan’s study, I suggest that the administrative processes contemplated by the CCA have not tended to develop principles of economic efficiency by which we can predict with reasonable certainty whether conduct which is anticompetitive may be excused from the consequences of contravening the law. This issue has not previously been addressed in Australia with respect to market power, but the Harper Review amendments have made it a live issue.

IV.  Institutional Arrangements in the EU and the US: Court-Centred or Agency-Centred? We can note here some views from the EU and the US that are pertinent to our institutional framework. Dirk Schroeder argues that: Predictability is a key factor if competition law is not to restrain unduly legitimate economic activity. Predictability is created through effective judicial control and the development of transparent and generally applicable tests.48

We will discuss the trend in the EU towards a more economic evaluation of dominant firm conduct in Chapter 3. Schroeder argues that ‘a shift from a formbased to an effects-based approach’ means that we must ‘openly admit that economic assessment is part of applying the legal rule’.49 I agree, and suggest further that experience indicates that, over time, administrative functions must develop transparent and generally applicable principles in order to gain community respect. For example, the modern equitable jurisdiction of our courts began as an administrative exercise of discretionary practical justice, which over a period of 500 years morphed into a court with settled principles and precedents.50 This experience also demonstrates that the divided jurisdiction of the common law courts and courts of chancery was a continual problem until the fusion of jurisdictions, which occurred in Victoria in 1883.51

47 The possibility is briefly noted by Nagarajan (n 42) 159. 48 Dirk Schroeder, ‘Normative and Institutional Limitations to a More Economic Approach’ in J Drexl, W Kerber and R Podszun (eds), Competition Policy and the Economic Approach: Foundations and Limitations (Cheltenham, Edward Elgar, 2011) 279, 280. 49 ibid 283. 50 See, eg, M Evans, Equity and Trusts (Chatswood, LexisNexis Butterworths, 2012) 4–10. 51 ibid 11.

Conclusion  87 Hylton discusses the US ‘court-centred’ process, which he contrasts with the EU ‘agency-centred process’.52 He argues that prosecutorial and adjudicative functions should be separated, as is the case in the US system, because it is only in the court-centred process that ‘evidence and arguments [adopted by agencies] are subjected to a rigorous and independent evaluation of merit’. We will see in Part II that this argument may well be applicable to the CCA’s dual prosecutorial and administrative ‘authorisation’ roles. Hylton also argues that the development of monopolisation law in the US by common law method is a strength of the US system and is to be preferred to the EU system of agency-led development.53 This may be contrasted with the way in which section 46 has developed in Australia through a cycle of enforcement failure and subsequent legislative amendment (briefly summarised in the Appendix to this monograph addressing the legislative history of section 46). However, the Australian legislation is very detailed and prescriptive, unlike the US equivalent, which means that there is not the same scope for Australian courts to develop section 46 of the CCA by common law method as there may have been in the US. In any event, I suggest that the development of the law by judicial law-making using the common law method is a slow and erratic process, and that legislative amendment to clarify the role of efficiency would be preferred.

V. Conclusion In the 40 years since competition law was enacted in Australia, there has been significant change in the fundamental ideas of economics and law around the world. Policy debates continue in Australia and overseas about the interpretation and application of section 46 and its corresponding provisions. However, the implications for our concepts of ‘market power’, ‘competition effects’ and ‘efficiency effects’ (which are still influenced by SCP thinking) remain to be addressed. Certainly, the Harper Review did not address the implications. This study addresses these issues and proposes that the courts should be capable of determining issues of efficiency as it can be adapted and redefined for the purposes of the market manipulation approach proposed by this study. I suggest that our institutional framework is a legacy of the 1960s, when the legal approach in Australia regarded the essential issue as one of policy, based on weighing up public interest factors, ie, an executive or administrative function rather than a judicial function. Under our system, actions for penalties and civil damages are prosecuted in the courts (as they must be for constitutional reasons).

52 K Hylton, ‘Antitrust Enforcement Regimes’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 1, 17 at 29. 53 ibid 27–28, pointing out that the Sherman Act says relatively little, and the law has effectively been developed by the courts.

88  The Mischief and Australia’s Institutional Response I suggest that it would be undesirable for the courts to continue to be excluded from determining efficiency issues that are recognised internationally to be relevant to misuse of market power and that could be raised in authorisation proceedings. The recommendation of the Harper Review to include express reference to efficiency in the Review’s proposed section 46(2) is to be welcomed. However, had it been enacted, it would have raised difficult conceptual and forensic issues for the courts. These issues are addressed in the present study. In the next chapter, drawing on experience in the EU, the US and Canada, we will develop our understanding of the relationship between ‘competition effects’ and ‘efficiency effects’ in connection with market power under the paradigm of zero-sum competition. We will contrast the current section 46 ‘purpose’ test, the Harper Review section 46 modified SLC test and the market ­manipulation approach which looks to actual efficiency effects in the market. We will see how economic concepts of efficiency and legal concepts of an ‘efficient market’ derived form securities markets manipulation laws can be used or adapted for the purpose of the legal theories examined in this monograph.

3 Competition and Efficiency Effects in Europe, North America and Australia I. Introduction This chapter discusses abuse of dominance provisions in Canada, the US and the EU (with particular attention to the EU) to provide a context for Australia’s new and proposed laws. We consider the revised section 46 of the CCA proposed by the Harper Review (ie, including section 46(2), which requires consideration of efficiency effects) and the alternative prohibition of ‘market manipulation’ proposed by the present study. The focus of the discussion is on the nature of ‘effects’, how these might be determined, and how institutional aspects of competition law regimes affect the determination and application of the concept to the conduct of dominant firms. As we saw in Chapter 2, in Australia, ‘competition effects’ fall within the jurisdiction of the courts; however, ‘efficiency’ and other public interest effects are generally considered to be beyond the court’s competence, and more clearly fall within the administrative functions of the ACCC and the ACT. The Harper Review notes seven previous inquiries that recommended against adopting an ‘effects’ test in section 46 and describes the enduring debate in Australia over the issue as ‘somewhat unproductive’.1 The Review states that, internationally, competition laws examine both purpose and effect.2 It briefly describes the European Commission’s 2008 guidance as one ‘that is viewed as advocating an effects test’ and describes the US test as ‘conduct-based’.3 This glosses over what is generally regarded as a highly debated area in competition economics and law internationally.4 It is not intended here to engage with the unproductive Australian 1 Commonwealth of Australia, Competition Policy Review, March 2015 (hereinafter ‘Harper Review’, ‘Review’ or ‘Harper’) 336. 2 ibid. 3 ibid 524, referring to the European Commission, ‘Guidance on the Commission’s Enforcement Priorities in Applying Article 82 EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings’ (December 2008) [2009] OJ C45/7 (hereinafter ‘Guidance’); see also the discussion in Harper Review (n 1) 340. 4 D Bernheim and R Heeb, ‘A Framework for the Economic Analysis of Exclusionary Conduct’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford

90  Competition and Efficiency Effects in Europe, North America and Australia debate for and against an effects test (or, for that matter, the equally unproductive debate in Parliament that resulted in the proposed Harper section 46(2) being removed from the new law). Instead, the focus of this chapter is on the nature of ‘competition effects’, the consequences of basing the prohibition of dominant firm conduct on them as an alternative to the existing ‘purpose’ test, and their relationship with ‘efficiency effects’ and other defences recognised internationally. Thus, the focus is on principles that provide a legally determinate framework to examine in Part II certain unilateral practices of dominant firms known internationally to be problematic, but for which there is little uniformity in international competition law analysis. We have seen in previous chapters that the SCP approach to assessing competition effects has been discredited overseas and to some extent in Australia (at least so far as the ACT is concerned). Some suggest that the underlying changes in economic theory create a need for ‘direct’ measures of ‘competition effects’ and ‘efficiencies’ that do not rely on using structural analysis as a surrogate.5 The ACCC in its submission to the Harper Review appears to accept a broader view as a matter of economics: In the ACCC’s view, the SLC test analyses the conduct in question beyond the static limits of structural characteristics which affect competition in a market (such as effect upon market share, concentration and barriers to entry) to also include strategy and the dynamic competitive constraints upon conduct within a market.6

However, as previously noted, observations by appellate court judges suggest that some still regard more atomistic competition as the relevant benchmark,7 despite strong arguments that suggest the hypothetical competitive market cannot be the appropriate benchmark to assess the conduct of dominant firms in real-world competition.8 Hanks and Williams argue that, since competition law does not prohibit a monopolist deriving monopoly profits, the assumed ‘competitive level’ of price and output cannot be the benchmark to assess dominant firm conduct.9

University Press, 2015) vol 2, 3; M Schinkel and P Larouche, ‘Continental Drift in the Treatment of Dominant Firms: Article 102 TFEU in Contrast to Section 2 Sherman Act’ in Blair and Sokol ibid 154. 5 See, eg, T Muris, ‘Improving the Economic Foundations of Competition Policy’ (2003) 12 George Mason Law Review 1, 14; R Cudahy and A Devlin, ‘Anticompetitive Effect’ (2010) Minnesota Law Review 59, 75–78; D Crane, ‘Market Power without Market Definition’ (2014) 90 Notre Dame Law Review 31, 69. 6 ACCC, ‘Submission to the Competition Policy Review – Response to the Draft Report’, 26 November 2014, 51. 7 See the Introduction to this monograph, n 91 for detail of observations by: Gummow, Hayne and Heydon JJ of the High Court in Rural Press Ltd v ACCC (2003) 216 CLR 53; ALR 217 [49]; Nettle J in the High Court in ACCC v Flight Centre Travel Group Ltd [2016] HCA 49 [141]; compare the observations of the majority in Melway (2001) 205 CLR 1 [52]. 8 F Hanks and P Williams, ‘Implications of the Decision of the High Court in Queensland Wire’ (1990) 17 Melbourne University Law Review 437, 460–61. 9 ibid 448. See also D Evans and K Hylton, ‘The Lawful Acquisition and Exercise of Monopoly Power and its Implications for the Objectives of Antitrust’ (2008) 4 Competition Policy International 203, ssrn.com/abstract=1275431. In the EU, it is acknowledged that dominance per se is not ­illegal: see, eg, Case C52/09 Konkurrensverket v TeliaSoneraSverige [2011] (CJEU, 17 February 2011)

Introduction  91 This chapter takes the point further by proposing that the appropriate benchmark should be what we might reasonably expect a dominant firm to do in pursuit of its own interests in the real-world market in which it finds itself, ie, conduct which is rational for the dominant firm. This benchmark provides a basis to identify conduct that may have harmful consequences or, as this study proposes, may manipulate the market, and thus requires some explanation. In the New Zealand context, where the law corresponds with the old (or preHarper Review) version of the Australian section 46, Andrew Gavil suggests that the benchmark should be the actual market, not a hypothetical competitive market.10 This leads him to propose that ‘merits-based strategies’ and conduct that excludes rivals on the basis of efficiency should be acceptable as they do not rely on market power.11 I draw some support for my proposed approach from his arguments. These issues have been touched upon in Australia and have been a source of controversy as our understanding of the ‘purpose test’ in section 46 has developed. What remains to be seen is how these arguments may be applied to an effects test. Hanks and Williams argued in 1990 that cases under section 46 would in future revolve around questions of economic efficiency, as decisions of a dominant firm made on that basis would not rely on market power.12 That prediction has not been borne out, though Heerey J in the Federal Court adopted Hanks and Williams’ argument that rational business decisions do not indicate a purpose that relies on market power.13 However, Katherine Kemp argues that rational business (hereinafter ‘TeliaSonera’) [24]; Case C209/10 Post Danmark A/S v Konkurrencerådet [2012] (CJEU, 27 March 2012) (hereinafter ‘Post Danmark’) [21]; and Case C 413/14 P Intel Corporation Inc v EC [2017] (CJEU, 6 September 2017, rectified 19 September and 24 October 2017) (hereinafter ‘Intel’) [133]. However, s 102 TFEU prohibits excessive pricing. 10 A Gavil, ‘Imagining a Counterfactual Section 36: Rebalancing New Zealand’s Competition Law Framework’ (2015) 46 Victoria University of Wellington Law Review 1043, 1048 and 1053. Although the New Zealand equivalent of s 46 is similar, I do not discuss New Zealand cases. The decision of the UK Privy Council in Telecom Corporation of New Zealand Ltd v Clear Communications Ltd [1995] 1 NZLR 385 was controversial, and literature concerning that controversy, such as Gavil’s ­article, informed the reform debate in Australia. Since the passage of the Australia Acts in the 1980s, the High Court of Australia is our final court of appeal, and Privy Council decisions are not binding on it. Accordingly, I have focused on Australian appellate decisions in this monograph. 11 ibid 1058. 12 Hanks and Williams (n 8) 445–46. Hanks and Williams do not define ‘efficiency’; however, the examples they give of possible efficiencies that they consider might justify conduct of a dominant firm are productive efficiencies, eg, transaction cost efficiencies and economies of scope in production (at 449–52). See also L Griggs, ‘Can a Corporation No Longer Maximise Profits: Misuse of Market Power and the Decision in Melway v Hicks’ (2000) 19 University of Tasmania Law Review 145. 13 See Heerey J’s dissenting opinion in Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (1999) 169 ALR 554 [22]; S Quo, ‘Interpretation and Application of the Purpose Test in s 46 of the Competition and Consumer Act 2010 – Part 1’ (2011) 19 Competition and Consumer Law Journal 90, 108 concludes that it is unclear whether efficiency is imported by virtue of the ‘purpose’ element or the ‘taking advantage’ element of s 46 (the latter element being removed under the Harper proposals); K Kemp, ‘Uncovering the Roots of Australia’s Misuse of Market Power Provision: Is it Time to Reconsider?’ (2014) 42 Australian Business Law Review 329, 348–9 and ‘Taking Advantage of Substantial Market Power, and Other Profit-Focussed Tests for Unilateral Anticompetitive Conduct’ (2015) 41 Monash University Law Review 655, 681–82 discusses decisions involving Heerey J, including ACCC v Boral Ltd (1999) 166 ALR 410 [158], ACCC v Australian Safeway Stores Pty Ltd (2003) 129 FCR 339 [329] and a further decision influenced by this reasoning, Cement Australia v ACCC (2013) 310 ALR 165 [1896]–[1900], [1904].

92  Competition and Efficiency Effects in Europe, North America and Australia ­ ecisions of a dominant firm may have harmful competition effects, which need d to be assessed.14 This chapter considers how we might do so, drawing on experience principally in the US and the EU (but also with reference to Canada, which the Harper Review cites as an example of an effects tests similar to that proposed by the Review). As noted in Chapter 1, it is sometimes said that a dominant firm ‘artificially’ restricts output to increase price and derive above-normal profits.15 This is not so, as the ‘profit-maximising’ output is determined by elasticity of demand and cost functions, ie, the marginal revenue from increasing output beyond that point does not cover the marginal costs, so reduces profits. Similarly, reducing output below that point would reduce revenue and profit. Thus, it is possible to examine business decisions against a benchmark of what is rational for the dominant firm in the market in which it finds itself. I argue that such conduct can be justified not because it is logical (ie, rational), but because it may be efficient, that is, in the sense that while we cannot forensically identify the point of optimum social efficiency, we can say that increasing output beyond that point may well be inefficient both privately for the firm and socially. We will see that we cannot measure the impact on social efficiency, but equally we cannot assume it to be neutral. By critically analysing the assumptions underpinning the conventional benchmark for social efficiency (ie, the presumed competitive level), we can propose an alternative paradigm of competition in which to assess the efficiency of the dominant firm’s conduct. As we will see when considering the Australian ‘effects test’ in more detail below, Australian courts have interpreted the ‘purpose’ test to focus on ‘subjective purpose’, which suggests we should exercise a degree of caution when exposing business decisions to scrutiny by economists and lawyers whose frames of reference and subjective opinions may reasonably differ. On the other hand, adding an overlay requiring a purely objective analysis of ‘effects’ without defences of the kind recognised overseas may well create a ‘no fault’ or strict liability regime, ie, conduct of a dominant firm would be condemned without a need to demonstrate a fault element such as intent or purpose. I suggest this too requires a degree of caution. While Part IVA of the CCA provides a strict liability regime for manufacturers and importers of defective consumer goods, it appears that attempts to create a ‘no fault’ regime for monopolisation in the US in the 1970s were rejected and such a regime is still out of favour as a basis for legal liability.16 The Harper Review does 14 K Kemp, ‘Taking Advantage’ (n 13) 682. 15 See, eg, G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011) 13–14. 16 See, eg, S Stroux, US and EC Oligopoly Control (The Hague, Kluwer Law International, 2004) 57–61, who discusses the 1959 proposal by C Kaysen and D Turner, Antitrust Policy: An Economic and Legal Analysis (Cambridge, MA, Harvard University Press, 1959) 91–92 to dispense with the fault requirement and base liability for monopolisation ‘primarily but not entirely, in terms of … effect’, and the US discourse of the 1960s and 1970s on the proposed ‘no fault’ prohibition of monopolisation; FM Scherer and DR Ross, Industrial Market Structure and Economic Performance, 3rd edn (Boston,

Introduction  93 not appear to engage with the US discourse, so we need not address it in detail. We will, however, consider the consequences of the Harper section 46, eg, whether it could operate as a ‘no fault’ prohibition with possible adverse consequences. In dispensing with the hypothetical competitive market as a relevant benchmark to determine whether market power has been exercised, there is a further implication that remains to be addressed. Market power is defined, and assumed to exist, by reference to the same benchmark, ie, a hypothetical competitive market in which no firm has appreciable market power. I suggest that we need to reconsider the concept of market power itself and our theory of harm that market power is said to cause. I suggest that this ‘power’ derives from elasticity of demand and can alternatively be regarded as the power of the market, which cannot be possessed by the dominant firm, but can be manipulated. That is, what competition law regards as ‘market power’ looks very much like ‘market manipulation’, which is addressed in the literature on commodity market regulation.17 We referred to Australian securities markets laws in Chapter 1 when discussing the concept of market power. We will return to them later in this chapter when considering how such laws adapt the concept of ‘efficiency’. A ‘market manipulation’ approach might also help address the problem of directly assessing ‘competition effects’ by examining the effects of conduct in question on market performance. Recent US cases confirm that the US monopolisation provision (section 2 of the Sherman Act) can apply to alleged manipulation in the cotton, crude oil and silver futures markets.18 In Daniel Shak v JP Morgan Chase & Co, the court held that direct evidence of price effects could suffice to demonstrate market power – in contrast to usual methods relying on market definition and market share, which are generally acknowledged to be notoriously unreliable.19 It is suggested that we need to address the implications of changes in economic theory and engage with international debate on the subject in order to be better able to assess the effects of dominant firm conduct in Australia. We must bear in mind that our legal objective is to ‘demarcate the boundaries of acceptable conduct’.20 A focus on rational business decision-making by a dominant firm in the real-world market lends itself to this.21 In Chapter 1 we examined a­ ssumptions

Houghton Mifflin, 1990) 481; H Hovenkamp, Federal Antitrust Policy (St Paul, West Publishing, 2005), 243–47; MS Gal, Competition Policy for Small Market Economies (Cambridge, MA, Harvard University Press, 2003) 89. 17 See, eg, C Pirrong, ‘Commodity Market Manipulation Law: A (Very) Critical Analysis and a Proposed Alternative’ (1994) 51 Washington & Lee Law Review 945, 951–52; and J Kelliher, ‘Market Manipulation, Market Power, and the Authority of the Federal Energy Regulatory Commission’ (2005) 26 Energy Law Journal 1, 15. 18 Daniel Shak v JP Morgan Chase & Co US District Court SDNY 12 January 2016; Cotton Futures SDNY 2013; Crude Oil SDNY 2012, available at law.justia.com/cases/federal/district-courts/new-york/ nysdce/1:2015cv00992/438330/43. 19 R Smith, ‘Market Definition: Going, Going, Gone? Developments in the United States’ (2010) 18 Competition and Consumer Law Journal 119. 20 Cudahy and Devlin (n 5) 59. 21 As Kemp, ‘Taking Advantage’ (n 13) 689 notes.

94  Competition and Efficiency Effects in Europe, North America and Australia about ‘norms of conduct’ underlying economic models and competitive ­paradigms, and noted important differences in the worldviews of economists and lawyers. In the present chapter, we examine how these differing perspectives affect the ­analysis in Australia and overseas of competition and efficiency effects, and how the latter may mitigate the former. We will start with a review of the current Australian effects test used in prohibitions other than section 46, the relevance of ‘efficiency effects’ that may mitigate anticompetitive effects, and the institutional arrangements that affect such determinations. The chapter then examines the Harper Review proposal for introducing an ‘effects test’ into section 46 based on the Canadian experience, with an important modification to the current Australian effects test making it necessary to consider offsetting efficiency gains.22 The ACCC’s draft ‘Framework’ document,23 intended to elucidate the Harper proposed section 46, does not address how courts may balance negative competition effects against positive efficiency effects. This issue is taken up in the present chapter. As noted in the Introduction, the requirement to consider efficiency effects has been removed from the amending legislation by the Parliament. However, I propose that efficiency should be the key issue to be addressed in regulating misuse of market power, either under the original Harper proposals or the alternative of market manipulation proposed in this study. After discussing the original Harper Review proposals, this chapter then addresses EU and US approaches to the prohibition of dominant firm conduct and applicable defences, before considering economic concepts of efficiency that may guide the courts in the application of the Harper modified effects test. Australian courts have not previously had to address efficiency in section 46 or SLC cases or how it may forensically be assessed. We have seen in Chapter 1 that the conventional approach to assessing economic efficiency effects within a market by reference to the hypothetical ‘competitive level’ relies on assumptions that are not satisfied in the real world. The Harper Review did not discuss economic efficiency or elucidate how it would be forensically assessed for the purposes of the Harper recommended section 46(2). We might speculate that the Review anticipated that this would be done by reference to the hypothetical ‘competitive level’. We might also speculate that the ACCC opposed the recommended mandate to consider efficiency because of the well-known forensic problems associated with ascertaining the ‘competitive level’. Thus, I suggest that we have some work to do if we are to implement a legal theory focused on efficiency. This chapter concludes 22 The Explanatory Memorandum to the Exposure Draft Competition and Consumer Amendment (Competition Policy Review) Bill 2016 refers, at [7.8], to the view of responsible Ministers that the ‘effects test’ proposed by the Harper Review ‘uses existing legal concepts from within the competition law’. As will appear, this under-states the impact of the Review’s proposed changes. The ACCC consultation draft ‘Framework for Misuse of Market Power Guidelines’, September 2016, [4.4] states that ‘subject to the import of the two new statutory factors [in s 46(2)], the proposed law will not change the ACCC’s established approach’ to the competition test (emphasis added). The ‘Framework’ document was subsequently published as ACCC, ‘Interim Guidelines on Misuse of Market Power’, October 2017. 23 ACCC, ‘Framework’ (n 22).

The Current Australian Effects Test Outside Section 46  95 by ­proposing how conventional economic concepts of efficiency, and the legal concept of an ‘efficient market’ from securities markets manipulation laws, can be adapted and redefined for use in the proposed legal theory of misuse of market power as market manipulation.

II.  The Current Australian Effects Test Outside Section 46 The Harper Review proposes that section 46 be subject to the same SLC test that applies in some other provisions of our competition law. That is, conduct covered by, eg, sections 45, 47 and 50 is prohibited if it has the purpose, effect or likely effect of substantially lessening competition in a relevant market.24 It is generally understood that by ‘competition’ we mean the process of competition.25 The expression of harm as a ‘lessening’ of this process does not fit semantically – though it is too late now to revisit this language, it might be better to think of SLC as ‘weakening the competitive process’.26 Similarly, the expression ‘pro-competitive’ is semantically inadequate, as it suggests a binary opposite, that which strengthens the competitive process. The Harper proposed section 46(2) assumes a binary opposition between conduct that ‘lessens’ or ‘increases’ competition. In fact, there must be much conduct that simply reflects the normal working of the competitive process and thus neither weakens nor strengthens the process. The current section 46 prohibits conduct having the ‘purpose’ of excluding competition (we will return below to the precise drafting of the proscribed exclusionary purposes). At this stage, it is worth noting that ‘purpose’ has been judicially interpreted to be ‘subjective purpose’, apparently in order to give it a meaning that adds something to ‘effect’.27 A subjective purpose may be inferred where the effect is a commonly understood consequence of the conduct.28 Nevertheless, subjective purpose requires a forensic examination of the process of reasoning of management and directors of a corporation. It appears that the courts have, by this interpretation, solved the ‘dual purposes’ problem, which arises in zero-sum competition, ie, normal self-interested conduct

24 The Harper Review s 46 as enacted exhaustively defines the relevant markets in a slightly more restrictive way than the current or Harper recommended versions, which refer to the market in which the firm has market power ‘and any other market’ (see the extracts set out in the Appendix to this monograph). The more restrictive drafting appears to be intentional: see Explanatory Memorandum (n 22) [1.40]–[1.45]. I will not address possible issues of interpretation that may arise under the new provision, which should not affect the outcome of cases considered in ch 5. 25 QCMA (1976) 8 ALR 481, 515–16. 26 Niels, Jenkins and Kavanagh (n 15) 180. 27 ASX Operations Pty Ltd v Pont Data Australia Pty Ltd (1990) 97 ALR 513, 526 (Lockhart, Gummow and von Doussa JJ). 28 See, eg, ACCC v Cement Australia Pty Ltd [2013] FCA 909 [3006]. Note also that s 46(7) was inserted in 1986 to provide that purpose could be inferred from conduct or other relevant circumstances.

96  Competition and Efficiency Effects in Europe, North America and Australia in competing for sales at one and the same time harms competitors.29 Despite this, section 4F of the CCA appears intended to solve the dual purposes problem against a firm in circumstances of zero-sum competition by providing that the ‘purpose’ test is satisfied if one purpose among many is itself a substantial purpose. So the judicial interpretation of ‘purpose’ enables us to distinguish between permitted conduct (that is, which in the interests of the firm, but incidentally harms competitors) from prohibited conduct (which has a substantial subjective purpose of excluding competition). However, this in itself does not help us distinguish conduct which harms the process of competition from conduct which is part of the normal process of competition, under which competitors will be harmed and excluded from the market. As we will see, legislative and judicial approaches to this issue vary internationally. I suggest that the solution to this problem lies not in assessing ‘competition effects’, but in the analysis of ‘efficiency effects’, which this chapter addresses. The judicial interpretation of ‘purpose’ in Australia raises the question whether a firm may be found to have a prohibited subjective purpose, notwithstanding that there is no effect on competition (either because the conduct was not carried out or because the conduct simply could not have the contemplated effect). While conduct of the latter nature would appear to be delusional rather than blameworthy, the courts have held this to be prohibited conduct under the SLC test (again, in order to give ‘purpose’ a meaning that adds something to ‘effect’).30 Given that ‘attempts’ attract ancillary liability under section 76(1)(b), it appears that this judicial interpretation adds nothing to the law. We will see that this issue, of ‘purpose without effect’, has been addressed internationally in different ways, and will arise under the Harper Review proposed section 46 for consideration in the Australian context. The decision of the Federal Court of Australia delivered on 25 May 2018 in ACCC v Pfizer Australia Pty Ltd31 is a case of ‘purpose without effect’, ie, no anticompetitive effect was alleged by the ACCC under section 47, so the case turned on ‘purpose’ under the SLC test in section 47 and proscribed purposes under section 46 (which at the time did not contain an effects test). We will discuss this case further in Part II. A further point to note about our current effects test (in provisions other than section 46) is that it is predictive – that is, it does not require forensic examination of actual effects (though of course that would be sufficient). The prohibition

29 It is this problem that Heerey J in Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (1999) 169 ALR 554, 562 suggests makes an assessment of ‘purpose’ akin to a metaphysical exercise. The dissenting judgment of Heerey J in that case was preferred by the High Court on appeal: (2001) 178 ALR 253, 268. My discussion of the ‘dual purposes’ problem was first published in my article ‘Taking Advantage of Market Power: The Competitive Market Hypothesis’ (1999) 73 Law Institute Journal 70. 30 Universal Music (2003) 201 ALR 636, [274], largely because the court put aside hindsight knowledge that the effect was not achieved [247]; Seven Network Ltd v News Ltd (2009) 262 ALR 160 [894] and [899]–[900]. 31 ACCC v Pfizer Australia Pty Ltd [2018] FCAFC 78 (hereinafter ‘Pfizer’).

The Current Australian Effects Test Outside Section 46  97 captures conduct that ‘would have the effect’ or even more tenuously ‘would be likely’ to have the proscribed effect. While the meaning of ‘likely’ has not been decisively pronounced upon by the appellate courts, influential judgments hold that it requires only that there be a ‘real chance’ that a proscribed effect may occur, not that it be ‘more likely than not’.32 This interpretation has been criticised.33 While Gavil does not address the Australian debate on this subject, he notes that the focus of effects tests internationally is on ‘actual or probable effects’.34 As we will see, the debate in the EU about whether effects should be inferred from the nature of conduct has been resolved in favour of requiring an actual competition analysis. We can only speculate whether the approach of the courts in Australia to the predictive SLC test will follow suit. Suffice to say that judicial authority in Australia to date does not require it, and the legislative amendments to the proposed Harper Review section 46(2) discussed earlier militate against the courts modifying or adapting the SLC test in the context of section 46. We will see in Part II how the SLC test has been applied in section 46 cases which have also raised SLC issues under section 45 or section 47. This may give some indication how the test will be applied under the new section 46. I suggest that even a predictive test based on ‘probability’ would push respect for the law to breaking point simply because human beings have a poor record in predicting the future. Parties and judges in competition cases are no exception. For example, in the foundation competition law case of Re Tooth & Co Ltd and Tooheys Ltd,35 the Competition Tribunal in 1979 found that breweries had market power because beer was not subject to close competition from other alcoholic beverages and hotels selling draught beer were not subject to close competition from other outlets. The Tribunal acknowledged that consumer demand was changing, but did not foresee the growth that subsequently occurred in licensed restaurants, supermarket sales of alcohol and the growth in consumption of wine and spirits.36 Another example is the High Court decision in Queensland Wire, where it was

32 See Tillmans Butcheries Pty Ltd v Australasian Meat Industry Employees’ Union (1979) 27 ALR 367, 380–82 (Deane J); Australian Gas Light Company v ACCC (No 3) (2003) 137 FCR 317, 343 (French J prior to his elevation as Chief Justice of the High Court). 33 See, eg, P Heerey, ‘Shootout at the Real Chance Café: Aiming at s 50’ (2011) 39 Australian ­Business Law Review 457; ACCC v Metcash Trading Ltd (2011) 284 ALR 662 [89] (Buchanan J) obiter. See AF Bergþórsson, What is Market Manipulation? An Analysis of the Concept in a European and Nordic Context (Leiden, Brill, 2018) 246–47; it seems that the concept of ‘likely effect’ in the context of EU securities markets manipulation laws is thought to be similar to that described above, though an accurate legal definition is lacking. Bergþórsson (at 279–80) discusses the inter-relationship between ‘effect’ and ‘intent’ in EU securities market manipulation law, further discussion of which is beyond our present scope. 34 Gavil (n 10) 1066 and 1072; see also Gal (n 16) 98. P Akman, The Concept of Abuse in EU Competition Law (Oxford, Hart Publishing, 2012) 144 argues that prohibiting the ‘mere likelihood of anticompetitive effects’ would be inconsistent with the fact that the law does not prohibit dominance (ie, market power) itself. 35 Re Tooth & Co Ltd and Tooheys Ltd (1979) 39 FLR 1 (hereinafter ‘Tooth and Tooheys’). 36 M Livingston, ‘Liquor Licensing: Research and Policy’, Turning Point Drug and Alcohol Centre, presentation to Drug Policy Modelling Program Symposium, 16 March 2012.

98  Competition and Efficiency Effects in Europe, North America and Australia found that imports of steel were insignificant.37 Thus, although the parties were well aware in the 1980s of the threat of steel imports from Asia,38 the implications for the litigation were not foreseen. Too great a focus on structural factors tends to distract attention from underlying dynamic forces, particularly those just outside the market at the time in question, but which may impact in the foreseeable future,39 such as were evident or acknowledged in the above cases. It would seem obvious that events subsequent to the conduct in question could assist in the application of a predictive test (particularly when litigation and appeals may take many years to conclude). However, courts are scrupulous to avoid predictions being affected by hindsight, even when applying the predictive test, and so do not admit evidence about subsequent events.40 The rationale appears to be that the courts are determining a state of affairs at the time that the conduct in question occurred, ie, are determining a historical state of affairs rather than predicting the future. The same generally applies in the case of appeals to the ACT from the ACCC. However, the Harper Review recommended that the ACT should not be confined to the material before the ACCC, and should be able to receive any further information it considers material.41 Even so, the underlying assumption seems to be that the state of affairs thus determined can be expected to continue.42 However, given that markets appear to be in a constant state of flux, this may well be a questionable assumption.43 It is the case, for example, that torts law and contract law require ‘foreseeability’ as an element of liability, defining the duty of care in negligence and the determination of damages for breach of contract. While these issues are determined finally by the courts in retrospect, compliance with the law requires a party to use reasonable foresight before acting, thus imposing a behavioural standard. So, for example, the High Court has held that a supermarket proprietor could reasonably foresee that if staff do not check the floors for spills at no more than 20-minute intervals, a shopper may slip and fall, sustaining an injury.44 Legal compliance in this situation presumably does not deter supermarkets from trading, but may add 37 Queensland Wire (1989) 83 ALR 577. 38 A Clark, ‘The Giant of the Miner League’, The Bulletin, 29 January 1980, 387 and 396. See also Anti-Dumping Commission, Australia, ‘Analysis of Steel and Aluminium Markets’, August 2016, 21–24, particularly Figure 2.12 charting the rise of steel imports in the period 1988 to 2015. 39 R Trindade and R Smith, ‘Modernising Australian Merger Analysis’ (2007) 35 Australian Business Law Review 358. 40 Universal Music (n 30) [247]. 41 Harper Review (n 1) 479–80. The Competition and Consumer Amendment (Competition Policy Review) Bill 2017, sch 9, pt 2, item 128 inserts a new s 102(9), which allows the Tribunal in merger cases (ie, apparently not in s 46 authorisation cases) to consider information that was not in existence when the ACCC made its original decision. 42 See, eg, ACCC Merger Guidelines, November 2008 [3.17], which in the context of the merger SLC test states that the future generally would involve a continuation of the past. 43 See, eg, J Schumpeter, Capitalism, Socialism and Democracy (London, George Allen & Unwin, 1976) 83, who considers that it is pointless to attempt to observe a process of change, which may unfold over decades, at one point in time. 44 Strong v Woolworths Ltd (2012) 285 ALR 420 [58].

The Current Australian Effects Test Outside Section 46  99 to costs of supplying goods and services to consumers. In the context of competition law, a predictive test requires that firms possess the ability to foresee the economic consequences of their conduct and to alter their behaviour accordingly. We must ask whether a firm has sufficient information and ability to analyse economic effects to fulfil such an obligation. For example, the firm will not know whether a competitor or potential new entrant is more or less efficient than itself. Perhaps it is sufficient if the firm asks itself whether the conduct would deter an equally efficient competitor and thus distort the process of competition. In NT Power, the court held that a dominant firm must not stand in the way of a more efficient entrant.45 This raises many questions not answered in the case, eg, how one assesses efficiency and its relevance to the purpose test. We will return to these questions shortly. It is unusual in law that legal liability is attracted by the risk or possibility of harm rather than actual harm. One area requiring legal prediction is the test for the courts granting interlocutory injunctions. What is being predicted here is the plaintiff ’s prospects of success in the asserted cause of action. The test requires assessing the balance of convenience, ie, the harm to either party if the conduct is not restrained. The courts require the plaintiff to give an undertaking to compensate the defendant for any loss arising from the injunction if the case does not succeed. It is suggested that justice would be served by repealing the predictive element of the SLC test and applying ordinary rules for courts granting an injunction to restrain conduct that has not had any effect, but should be restrained in order to prevent reasonably anticipated harm. We saw in Chapter 2 that the prohibitions in our competition law focus on the impact of conduct on the process of competition. This has been framed as a matter that is to be determined by the courts. However, it is recognised that harm to competition may be justified by offsetting public benefits, eg, efficiency gains, so an administrative process is provided for conduct (which would otherwise be prohibited) to be authorised by the ACCC or the ACT. It has been generally thought that the weighing-up of detriments and benefits is not suited to judicial determination, and the courts have no power to consider offsetting public benefits when determining whether the law has been contravened.46 This leads Stephen Corones to state that the question whether efficiency gains are relevant to assessing ‘competition effects’ is ‘one of the most vexed questions’ in Australian competition law.47 The ACCC submission to the Review appears to understate this issue: Efficiency enhancing conduct which is pro-competitive would not generally substantially lessen competition in a market. For any matters involving ‘offsetting’ efficiencies, 45 NT Power Generation Pty Ltd v Power and Water Authority (2004) 210 ALR 312 (hereinafter ‘NT Power’) [138]. We will also discuss NT Power in Part II, where we may wonder whether (given there are barriers to exit) courts should hasten the exit of an alleged inefficient firm or whether it should be allowed the opportunity to become more efficient (thus benefiting society) or to fail through the competitive process rather than through the intervention of the court. 46 ibid [137]. 47 S Corones, Competition Law in Australia, 6th edn (Sydney, Lawbook Co, 2014) 39.

100  Competition and Efficiency Effects in Europe, North America and Australia authorisation would be a suitable mechanism to allow potentially efficient conduct that may otherwise breach the provision to proceed.48

If a firm introduces changes that make it more cost-efficient and profitable, the consequence could be argued either way – it could be argued that the firm becoming more efficient must be good for the process of competition.49 However, the conduct may raise barriers to new entry or weaken rivals and so may arguably harm the process of competition,50 or if the firm is able to keep the profit for itself, this might indicate that the process of competition has been weakened (­otherwise we might expect such gains to benefit consumers in some way). The latter ­example suggests that harm to the ‘process of competition’ may be equated with ­‘efficiency’ losses when the effect is confined to one market. As we will see, excess profits resulting from market power legitimately gained, eg, by success, are not regarded as contrary to competition law in the US and Australia, but are condemned in the EU. However, in a 2016 review of east coast gas markets the ACCC found that, while the exercise of market power to derive above-normal profits in one market had no impact on competition in that market, in the ACCC’s view, it causes efficiency losses in other markets.51 The ACCC has recommended changes to the competition law’s access regime to constrain market power and prices where access is already being provided to gas pipelines, but at prices the ACCC considers to be excessive as a result of the exercise of market power.52 That is, while the ACCC acknowledges that excess prices and profits are not objectionable under section 46,53 the perceived harm in this case is a loss of efficiency in upstream and downstream markets.54 This example reminds us that ‘competition’ is not an end in itself, but is merely a surrogate for ‘efficiency’.55 Indeed, the ACCC and the Productivity Commission acknowledge that ‘competition can be an imprecise proxy for efficiency in some circumstances’.56 This suggests that competition law, and our 48 ACCC (n 6) 49. 49 This is the view of RBB Economics, cited by Harper Review (n 1) 343. 50 R Smith and D Round, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5 Agenda 25, 30–31, discussing the Competition Tribunal decision in Ford Motor Company (1977) ATPR 40-043 [17,498], where exclusive distribution arrangements enhanced Ford’s ability to compete, but raised barriers to new entry, and at [17,499] the Tribunal stated that if a dominant firm increased its competitive strength by weakening rivals, that would in the Tribunal’s view reduce competition. 51 ACCC, Inquiry into the East Coast Gas Market, April 2016. 52 ibid. For more detailed discussion, see G Raitt, ‘Shifting the Goalposts: Current Issues in Australian Competition Law Affecting the Energy Sector’ (2016) 9 Journal of World Energy Law and Business 424, 426–28. 53 ACCC (n 51) 92. 54 ibid 18. 55 Akman (n 34) 47–48 acknowledges that competition should not be protected for its own sake, but argues that, due in part to differing welfare objectives, harm to competition should be a necessary condition for abuse of dominance. Akman (at 316) also argues that the ‘lack of ’ an efficiency justification should also be a necessary condition. 56 ACCC (n 51) 131, quoting from the Productivity Commission, Final Report – National Access Regime, 25 October 2013, 173.

The Current Australian Effects Test Outside Section 46  101 institutional arrangements, should clearly provide for courts to consider efficiency effects when assessing harm from ‘competition effects’.57 One reason why the question may be described as ‘vexed’, as noted above, is that the process of competition and ‘efficiency’ are conceptually separate. We may question whether they can be successfully combined in the way suggested, eg, by Gavil, who apparently would assess whether ‘anticompetitive effects are more substantial than any realised efficiencies’.58 However, he acknowledges that pro- and anticompetitive effects cannot be weighed up and balanced in any quantifiable sense.59 This appears to be what the Harper Review proposal invites the court to do.60 Accordingly, we need to consider whether (and how) we should recognise efficiency defences to uses of market power that fail either the purpose or effects tests. Consideration of such defences has implications for other SLC provisions of our competition law. We might seek to distinguish ‘defences’ which negate the elements of the contravention (and so need no statutory recognition) from ‘affirmative defences’ which acknowledge harm to competition, but seek to justify it.61 As we will see, in the US, efficiency gains are regarded as a defence to claims of competitive harm, but do not fit neatly into the framework of defence or affirmative defence.62 Thus, the discussion of competition effects in the US context proceeds from a different legal and conceptual framework than our own, where the weighing-up of adverse competition effects and public interest factors such as efficiency has been regarded as something not suitable for judicial determination (so the courts have worked around the issue in other ways). It is proposed that the solution to the ­problem may be found in the hierarchical relationship between ‘competition effects’ and ‘efficiency effects’, ie, competition is a surrogate for efficiency, so while the two effects cannot be combined or weighed up to achieve a ‘net’ effect, I suggest that positive ‘efficiency effects’ should prevail over negative ‘competition effects’. This adds to the force of arguments noted above that rational business decisions which may be justified on grounds of efficiency, or that do not rely on market power, simply do not cause adverse ‘competition effects’. We can now consider how the Harper proposal treats the issue of causation. 57 Kemp, ‘Uncovering the Roots’ (n 13) 335 argues that there should be an efficiency defence in s 46 and that conduct covered by s 46 should be capable of authorisation, which the Harper Review has now proposed, but the question of the institutional framework remains to be addressed. 58 Gavil (n 10) 1067. 59 ibid 1066. See also Akman (n 34) 317, who considers that there are difficulties with such a ‘balancing exercise’. Compare Kemp, ‘Uncovering the Roots’ (n 13) 349, who suggests that weighing up is possible, or at least it is possible to assess whether ‘the means adopted by the dominant firm were proportional to the claimed efficiencies’; Kemp ‘Taking Advantage’ (n 13) 681 notes arguments that it is near-impossible to distinguish the gains the dominant firm makes from efficiencies and the gains it makes from anti-competitive effects. 60 Proposed s 46(2), Harper Review (n 1) 513–14. 61 A Gavil, ‘Burden of Proof in U.S. Antitrust Law’ (2008) 1 Issues in Competition Law and Policy 125, 152–55. Akman (n 34) 324 argues that efficiency should be an element of the contravention rather than a defence. 62 ibid 154.

102  Competition and Efficiency Effects in Europe, North America and Australia

III. The Harper Review Effects Test and the Role of Efficiency Effects The Harper Review states that the forensic problem of proving ‘subjective purpose’ is one of two principal reasons why the Review recommends adopting an effects test.63 The objection to ‘subjective purpose’ overlooks the reasons why the courts interpreted the SLC test in this way and reopens the problem of ‘dual purposes’. It is suggested that underlying different approaches to the ‘dual purposes’ problem are two conflicting worldviews, or paradigms, of competition. It has been suggested above that the model of ‘perfect competition’, in which there is no interfirm competition for sales and no market power, cannot inform our benchmark to assess competition and market power in the real world. Further, it appears that courts accept that competition is zero-sum. We need to acknowledge the influence of these paradigms when we examine ‘competition effects’ (and, as we will see, the same applies to ‘efficiency effects’). The other principal reason given is that the Harper Review considers competition effects to be a better indicator of harm to efficiency, which the Review refers to as ‘consumer welfare’.64 This statement may imply that we do not need to address ‘efficiency’ itself, which is curious since, as mentioned above, ‘competition’ is a surrogate, or proxy, for efficiency and the Harper proposed section 46(2) requires consideration of efficiency effects. As previously noted, the Review does not clarify what it means by ‘consumer welfare’, which in the US is often used ambiguously to refer in fact to ‘total welfare’.65 The treatment of efficiencies as a defence to anticompetitive effects raises some issues about the applicable objective.66 Thus, we need to acknowledge that the process of competition is not necessarily harmed when equally or less efficient competitors are excluded because that is how competition works. While this could conceivably provide a negative defence under current law, it does not appear to have been raised in the cases, presumably because information is lacking to assess relative efficiency of competitors. In any event, we do not have judicial experience of dealing with this issue, which will come to the fore if we engineer the prohibition to find an ‘effect’ without a substantial subjective purpose. This exposes rational self-interested conduct to potential liability. It is proposed that ‘rational self-interested conduct’ is not only relevant to ‘purpose’, but can also be seen to be a surrogate ‘efficiency defence’ that should prevail over adverse ‘competition effects’. As we will see, there are a variety of ways used internationally for dealing with this problem, eg, basing the prohibition on ‘purpose and effect’ (Canada), no prohibition of ‘purpose without effect’ (the EU) or catching ‘purpose without effect’ as a prohibited ‘attempt’ (the US).



63 Harper 64 ibid. 65 H

Review (n 1) 335.

Hovenkamp, ‘Roger Blair and the Goals of Antitrust’ (2016) 61 Antitrust Bulletin 382, 383. (n 61) 153–55.

66 Gavil

The Harper Review Effects Test and the Role of Efficiency Effects  103 The Harper Review expresses the view that the current ‘purpose’ test is focused on protecting competitors rather than the process of competition, which is regarded as the proper focus of competition law.67 Despite reference to ‘a competitor’ or ‘a person’ being a new entrant, the drafting of proscribed purposes in the current section 46 may be considered to be a reasonable effort to capture what is internationally regarded as exclusionary conduct, ie, conduct which excludes actual or potential competitors that would not otherwise be excluded by the market. As Philip Williams observes, the acid test of exclusionary conduct is evidence that an actual competitor or new entrant has been excluded.68 As we will see from the EU experience, if the law works to protect competitors rather than the process of competition, we may expect a tendency for cases under section 46 to succeed when the process of competition is not actually harmed. In fact, the opposite criticism has been raised in Australia, ie, that there have been too few successful prosecutions.69 The Harper Review not only recommends adopting an effects test, but also recommends changes to the SLC test in the context of section 46. The proposal requires the court to consider, when assessing the impact on competition, whether conduct has the purpose or effect of ‘increasing competition in the market including by enhancing efficiency, innovation, product quality or price competitiveness in the market’.70 There is no corresponding provision in the SLC test in other provisions of our competition law. It is suggested that this change proposed by the Review is desirable insofar as it expressly recognises the need to consider efficiency effects, and has important implications for competition law. Here we see a mixing of the two distinct concepts of ‘process of competition’ and ‘efficiency’. The ‘vexed question’ of determining whether this merely acknowledges a negative ‘defence’ or creates an ‘affirmative defence’ for the first time in Australian competition law will be left to the courts to determine, since the proposed Harper section 46(2) has been removed by Parliament. If the courts are capable of assessing efficiency effects in the context of section 46, it should follow that the dividing line will become blurred between the jurisdiction of the courts and the administrative powers of the ACCC and the ACT to authorise on public benefit grounds conduct that would fail the SLC test. Under the current law, authorisation is not available for conduct falling within section 46. The Review recommends that the authorisation process be extended to apply to section 46.71 Gavil argues that if ‘efficiencies associated with the challenged conduct are likely to counteract the anticompetitive effects, it should be 67 Harper Review (n 1) 339. 68 P Williams, ‘The Counterfactual Test in s 46’ (2013) 41 Australian Business Law Review 93, 102. 69 In calling for amendments in 2004, A Fels, ‘The Past and Future of Competition Law’ (2004) 23 Economic Papers 3, 8 observes that the ACCC at that time had only one success in litigating abuse of market power, and asks: ‘Does anyone believe that there has been only one instance in which Australian business with substantial market power has misused it to harm competition in that period?’ The draft Explanatory Memorandum (n 22) [7.1] makes it clear that the purpose of the legislative amendments to s 46 is to ‘strengthen the prohibition’. 70 Harper Review (n 1) 513. 71 ibid 348.

104  Competition and Efficiency Effects in Europe, North America and Australia treated as a defense’, but that if ‘efficiency, more specifically producer surplus, has independent value’, then it should be treated as an affirmative defence.72 It would be undesirable if judicial interpretation of the Harper recommended that section 46 were to result in some efficiency ‘defences’ being recognised to be within the jurisdiction of the courts, and other ‘affirmative defences’ based on efficiency were to be held to fall within the administrative power of the ACCC and the Competition Tribunal. Further, it is hard in principle to see why there should be any difference between the SLC test in section 46 and other provisions which use it. I suggest that an affirmative defence should be recognised by express statutory amendment to section 46 and other provisions which use the SLC test. Preferably, however, we would adopt the market manipulation approach recommended by this monograph. The Harper Review advances a further reason to adopt the modified SLC test in section 46: it is said that conduct that might be innocuous if engaged in by a firm without market power may well have harmful effects if engaged in by a firm with market power.73 Thus, under the current law, the conduct would not be condemned because anyone could do it. At one level, the Harper proposition is no more than a restatement of the premise underlying section 46 that firms with market power should be constrained from misusing their power because we assume they can.74 However, the proposition is not a principle of economics or a presumption to be read into the SLC test. It can be seen from Gavil’s detailed analysis of the New Zealand equivalent of section 46 that the real objection is to the benchmark of the hypothetical competitive market, which is used to establish the corollary (against which he objects), ie, that if a firm without market power could engage in certain conduct, then when a firm with market power engages in that conduct, it does not use its market power.75 Gavil’s point is that both purpose and effect need to be analysed in the real-world market in which market power exists, including having regard to efficiencies such as scale economies that might be available to the firm with market power.76 In other words, the proposition requires that all conduct of dominant firms should be subjected to legal scrutiny to ensure their conduct (in respect of both purpose and effect) can be justified. Interestingly, Gavil seems to have no objection to the current ‘taking advantage’ requirement insofar as it suggests a causal nexus between the market power and the benefits obtained by the dominant firm from its conduct, ie, the effect.77 However, the ‘taking advantage’ requirement is the basis for judicial interpretation 72 Gavil (n 61) 154. 73 Harper Review (n 1) 338. Contra: Akman (n 34) 319. Compare Universal Music (n 30) [268], where the Full Court of the Federal Court, in the context of s 47, rejected the view of some authors (eg, F Hanks and P Williams, ‘The Treatment of Vertical Restraints’ (1987) 15 Australian Business Law Review 147) that exclusive dealing can only be effective if the perpetrator has substantial market power. 74 Gavil (n 10) 1044–45. 75 ibid 1056. 76 ibid 1057. 77 ibid 1050.

The Harper Review Effects Test and the Role of Efficiency Effects  105 that reads into section 46 the benchmark of the hypothetical competitive market, to which Gavil and others do object.78 The Harper Review proposal removes the ‘taking advantage’ requirement, thereby raising a doubt as to whether the provision is a ‘no fault’ or strict liability provision, or whether causation must still be shown to prove an ‘effect’.79 It is suggested that a causal nexus should still be required between market power and effect, ie, it should be demonstrated that market power has contributed to the effect.80 Nevertheless, it is troubling that a competitor without market power may engage in conduct which the dominant firm may not match, as that would seem to create a discriminatory law. When we approach ‘market power’ as the power to manipulate the market, we can see that any firm, whether the incumbent, new entrant or a smaller rival of the dominant firm, has the capacity to instigate a demand response (ie, price effect) by its output decisions.81 The Harper Review notes that Canada ‘prohibits anticompetitive conduct by a dominant firm that has the effect or likely effect of substantially lessening competition’.82 It is worth giving a brief review of Canada’s provision, which was first adopted in 1986.83 There are three elements of the prohibition, which may be summarised as follows: the firm must have substantial market power; the firm must engage in anticompetitive conduct; and such conduct must have or be likely to have the effect of SLC.84 We can immediately make several observations based on this. First, the provision requires ‘purpose and effect’, so the question of ‘purpose without effect’ does not arise, and the Australian debate concerning ‘real chance’ is avoided. Prior to 2009, the only remedy was an injunction to restrain prohibited conduct – monetary civil penalties were first introduced in that year.85 Thus, the normal rules concerning the award of injunctions should be adequate to address threatened conduct. Further, by requiring both purpose and effect, the Canadian provision also seems to avoid the problem of prohibiting ‘effect without purpose’, which would appear to impose liability regardless of any legitimate business justification, ie, to create a ‘no fault’ or strict liability regime.86 78 The Harper Review (n 1) 537–38 discusses decisions of the High Court in Queensland Wire and other cases. 79 The ACCC consultation draft ‘Framework’ (n 22) throws no light on this issue. 80 See also Akman (n 34) 324. 81 Implicitly recognised by R Smith and D Round, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5 Agenda 25, 29: ‘even firms with quite a small market share may exercise unilateral market power through strategic behaviour’. 82 Harper Review (n 1) 340. 83 See, eg, G Addy, J Bodrug and C Tingley, ‘Abuse of Dominance in Canada: Reflections on 25 Years of Section 79 Enforcement’ (2012) 25 Canadian Competition Law Review 276, 279. See also K Kemp, ‘Misuse of Market Power in Australia and Abuse of Dominance in Canada: Two Legislated Effects Tests for Unilateral Conduct’ (2018) 26 Australian Journal of Competition and Consumer Law 174. 84 Addy, Bodrug and Tingley (n 83) 279; s 79 of the Competition Act (Canada). 85 Addy, Bodrug and Tingley (n 83) 294. 86 In Commissioner of Competition v Canada Pipe Ltd [2006] FCA 233 (hereinafter ‘Canada Pipe’) [73], the Federal Court of Appeal recognised that a ‘valid business justification’, eg, ‘credible efficiency or pro-competitive rationale’, may counter-balance purpose and effect. That might not be the case under the Australian ‘purpose or effect’ SLC test. See further, eg, Kemp (n 83) 188–92.

106  Competition and Efficiency Effects in Europe, North America and Australia Second, the circularity between the requirement for ‘anticompetitive conduct’ and ‘anticompetitive effect’ may suggest that there is no causal nexus required between market power and effect. This also means that the provision is not limited to ‘unilateral’ conduct. In fact, it seems that most cases have concerned bilateral agreements that in Australia would fall under section 45 or 47.87 ‘Market power’ has been defined in Canada as the ‘ability to set prices above competitive levels for a considerable period of time’, though the difficulties of this determination are recognised and indirect proof by reference to structural factors is ­permitted.88 Anticompetitive conduct is defined non-exhaustively and has been held to connote intentional exclusionary conduct, ie, conduct aimed at a competitor, though this intent can be overcome by a business justification such as a credible efficiency or pro-competitive rationale.89 However, the focus of judicial interpretation appears to be harm to competitors rather than harm to competition, which has been c­ riticised.90 The criticism affects the interpretation of SLC because the focus appears to be less on the process of competition and more on the effect of the impugned conduct on the dominant firm’s competitors, which does not necessarily increase market power or make the market less competitive.91 There remains no right of civil action by the parties affected for damages.92 If we wish to ensure that we protect competition rather than competitors, I suggest that providing a private remedy for competing firms which may be said to suffer loss or damage as a result of a contravention of competition law creates an insurmountable conflict between public policy interests and the private interests of firms (and too much temptation for firms to ‘game the system’). In summary, it appears that Canadian courts and tribunals are still working through many of the issues that have been addressed in Australia and, as we will see, have been addressed over a longer period in the EU and the US. I suggest that the consequence of bringing effects within section 46 will be twofold: first, creating a need for better analytical tools than judicially accepted structural analysis provides to identify competition effects and efficiency effects caused or contributed to by market power; and, second, requiring a clearer understanding of efficiency and other defences which may be developed by the courts or, preferably (since judicial development may require several decades to play out), by legislative enactment. We can now turn to the position in the EU and the US to see what their experience can offer us in this regard.

87 Addy, Bodrug and Tingley (n 83) 281. 88 ibid 279, citing the 1990 decision of the Competition Tribunal in Nutrasweet, CT-89/2; ibid 287–88, citing the Federal Court of Appeal decision in Canada Pipe (n 86). 89 ibid 280 and 288–89, citing the Federal Court of Appeal decision in Canada Pipe (n 86). 90 ibid. 91 ibid 290–91. 92 ibid 302.

EU Abuse of Dominance and Defences  107

IV.  EU Abuse of Dominance and Defences In the EU, the abuse of dominance provision, Article 102 of the Treaty on the Functioning of the European Union (TFEU) (formerly Article 82 of the E ­ uropean Community Treaty), was regarded as virtually a per se prohibition because certain forms of conduct were presumed to be abusive if engaged in by a dominant firm, ie, without the need to assess the economic effects.93 A frequent complaint about enforcement of Article 102 was that the EC and the courts were protecting competitors, not the competitive process, leading to condemnation of conduct that did not have adverse effects, ie, actually discouraging competition.94 In 2008, the EC attempted to reform the approach to Article 102 by issuing a guidance note, ie, without legislative or judicial backing.95 In other words, the intent of the proposed more economic approach was to reduce interventions based on presumptions about the harm associated with certain forms of conduct.96 In summary, the EC’s approach is that the commercial conduct of dominant firms should not be proscribed merely because it harms competitors – the test should focus on harm to competition, which may be justified by offsetting efficiency gains.97 The EU position regarding efficiencies is somewhat qualified, in that the efficiencies must benefit the consumer, the exclusionary effect must not go beyond what is necessary to secure the benefits and the conduct does not eliminate effective competition.98 The EC attempts to articulate the normal working of the competitive process with the concept of ‘competition on the merits’.99 The CJEU in Intel uses the concept in a way that suggests ‘competition on the merits’ excludes firms that we expect the market and competition to exclude, so we must look to exclusionary effects to see if these are unexpected and, if so, seek an explanation.100 This seems to be an indirect approach that, ultimately, focuses on efficiency. Some suggest that ‘competition on the merits’ is an empty concept because it provides no basis to determine what is or is not ‘competition on the merits’.101 Nevertheless, the 93 Niels, Jenkins and Kavanagh (n 15) 180. 94 R Whish, ‘National Competition Law Goals and the Commission’s Guidance on Article 82 EC: The UK Experience’ in L Pace (ed), Competition Law: The Impact of the Commissions’ Guidance on Article 102 (Cheltenham, Edward Elgar, 2011) 154. 95 ‘Guidance’ (n 3). 96 Niels, Jenkins and Kavanagh (n 15) 180. 97 ibid 181. 98 See TeliaSonera (n 9) [75]–[76]; Post Danmark (n 9) [41]–[42]; Intel (n 9) [140]. 99 Niels, Jenkins and Kavanagh (n 15) 182. 100 Intel [2017] (n 9) [134]–[136]. 101 See, eg, Kemp, ‘Taking Advantage’ (n 13) 658; D Geradin, A Layne-Farrar and N Petit, EU Competition Law and Economics (Oxford, Oxford University Press, 2012) [4.136]. G Hay and R Smith, ‘“Why Can’t a Woman Be More Like a Man?”: American and Australian Approaches to Exclusionary Conduct’ (2007) 31 Melbourne University Law Review 1099, 1132 advocate competition on the merits as a benchmark to assess dominant firm conduct. The ACCC (n 6) 50 uses the term in its submission to the Harper Review, though it has no usage in Australian competition law. The ACCC refers to the concept in its ‘Interim Guidelines’ (n 22) and expresses tentative conclusions as to whether certain examples of conduct are or are not competition on the merits, though principles and standards are not clarified.

108  Competition and Efficiency Effects in Europe, North America and Australia fact that markets exclude less efficient competitors suggests several bases to distinguish ‘abusive’ conduct from ‘competition on the merits’.102 First, we may conclude that market power is not influential in excluding less efficient firms, since competition does that in any case, ie, there may be no causal nexus between the impugned conduct and the alleged effect. As noted above, the dominant firm may not know whether a competitor is more or less efficient, so it seems this can be tested by hypothetically examining the effect of the firm’s conduct on an equally efficient competitor.103 This then raises a number of issues to do with identifying cost benchmarks (all of which are problematic) in order to determine whether conduct would exclude an equally efficient competitor from the market: the economic concept of marginal cost can only be approximated; efficiency requires that average costs be minimised, which raises problems comparing the dominant firm with a new entrant not yet at efficient scale.104 The neoclassical focus on marginal costs and the assumed law of diminishing returns tends to divert attention away from average costs (which are minimised in perfect competition equilibrium) and, importantly, the cost of capital. As noted in the Introduction, neoclassical theory acknowledges that the law of diminishing returns does not apply in the ‘long run’, ie, the period in which inputs of all factors of production can be varied. The cost of capital depends on the mix of equity and debt capital used to finance the firm’s assets, and thus raises questions of financial efficiency as well as operational efficiency. These aspects of the firm’s productive efficiency tend to be overlooked if we focus only on marginal costs. Second, we may examine the firm’s decision-making process to determine whether it engaged in rational profit-maximising conduct that does not rely on harming competitors more than the competitive process contemplates.105 Here we are addressing the ‘dual purposes’ problem, which creates some difficulty distinguishing conduct which incidentally harms competitors, from conduct the profitability of which relies on harming competitors. These poles ultimately mingle if profitability is contributed to by both outcomes, which can be ‘difficult to separate’.106 For example, Bernheim and Heeb argue that if the competitor is weakened – eg, by losing a sale to the dominant firm, the competitor may be denied economies of scale, such that its ability to compete in other transactions is impaired – then it could be argued that the process of competition is harmed.107 This argument is troubling since it seems to focus on harm to competitors as a surrogate for harm to competition. However, more importantly, as Niels et al observe, the size of the market and economies of scale exclude firms simply

102 Niels, Jenkins and Kavanagh (n 15) 182. 103 ibid 187. 104 ibid 189–97. 105 Kemp, ‘Taking Advantage’ (n 13) 658; Bernheim and Heeb (n 4) 20. 106 Schinkel and Larouche (n 4) 185. 107 Bernheim and Heeb (n 4) 13. Note that their context is US antitrust law, but the point has general relevance.

EU Abuse of Dominance and Defences  109 because entry is not required to satisfy demand.108 The relationship between market size and industry concentration (the number of firms in the market) is of course more complex than the discussion by Niels et al suggests. We know that endogenous sunk costs and competitive strategies, as well as market size, can affect concentration.109 Thus, we would want to know whether firms are excluded as a consequence of the size of the market, normal competitive strategies or strategies that misuse market power and so should be condemned. The CJEU in the Intel case implicitly acknowledged that the market can exclude an equally efficient firm when it stated that it is necessary to ‘assess the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking’.110 While this might suggest a ‘purpose’ test, the CJEU also makes it clear that it is necessary to establish actual effects where this is contested and that effects cannot merely be inferred from the nature of the conduct.111 The preceding discussion mainly uses the language of ‘competition’. We must bear in mind that ‘competition’ is a surrogate or proxy for ‘efficiency’. The preceding discussion can be reframed in the language of efficiency: the market excludes equally and less efficient firms because it would be an inefficient use of resources for them to participate in the market. As we will see when efficiency is examined in the last section of this chapter, zero-sum competition can only be regarded as efficient in the Hicks-Kaldor sense,112 ie, if value created exceeds value destroyed. Returning, then, to the difficulty of distinguishing conduct which incidentally harms competitors from conduct that weakens the process of competition, we can see that the principles developed overseas in an effort to do so have application both to the Australian ‘purpose’ and ‘effects’ tests. It appears that in the EU, ‘purpose’ is only a staging point on the road to determine ultimate ‘effects’. Thus, it is said that the EU (unlike the US) has no prohibition on purpose without effect, ie, ‘attempts’.113 In the EU, it appears that the SLC test is a narrower test than has applied in Australia and is more a ‘guiding principle’ for the tests described above than an independent test.114 Thus, in EU merger control, the ‘consumer welfare’ standard seeks to keep prices close to costs and excludes from consideration efficiencies that are not passed on to consumers.115 That appears to be the approach of the ACCC in merger cases, but the ACT takes a broader view of efficiencies that may

108 Niels, Jenkins and Kavanagh (n 15) 12. 109 See, eg, J Perloff, L Karp and A Golan, Estimating Market Power and Strategies (Cambridge, Cambridge University Press, 2007) 34–39, who discuss and critique studies by J Sutton, Sunk Costs and Market Structure (Boston, MIT Press, 1991). 110 Intel (n 9) [139]. 111 ibid [139]–[142]. 112 See Hovenkamp (n 65) 384. 113 Larouche and Schinkel (n 4) 156. 114 Niels, Jenkins and Kavanagh (n 15) 187. 115 ibid 337.

110  Competition and Efficiency Effects in Europe, North America and Australia be taken into consideration.116 According to Niels et al, the EU version of the SLC test, based on the consumer welfare standard, is not preferred as an independent test for effects of dominant firm conduct: The test would therefore ask whether the conduct reduces competition to the detriment of consumers, without creating efficiencies that are sufficient to offset this detriment. This would be similar to the substantial lessening of competition criterion used in merger control … It has the downside that judging net effects on consumer welfare is complex. For courts or competition authorities to evaluate all possible anti-competitive effects and offsetting efficiencies would be a laborious exercise, although not impossible (it is done for mergers).117

Institutional arrangements in Australia have excluded courts from entering into the assessment of merger efficiencies and, as noted above, differences of opinion exist as to the efficiencies that may be taken into consideration. What then remains of structural analysis, which is the foundation of the Australian SLC test? It appears that Hovenkamp’s assessment of structural factors subsequent to the discrediting of the SCP approach in the 1970s is an accurate statement of the current position in the EU: ‘market structure, entry barriers, and the market position of the strategizing firms are all important in determining whether a particular anticompetitive strategy is plausible’.118 It will be apparent that barriers to entry and exit exist by reference to the neoclassical model of perfect competition (where it is assumed that there are none). As has been judicially noted, the concept of barriers to entry in imperfectly competitive markets is a matter of ‘vigorous debate’ among economists.119 In the real world, assets have value as part of a going concern, where they generate an income stream. It is an inescapable reality that investment in new assets requires a risk assessment of expected returns and that the value of an asset plummets when the firm ceases to be a going concern. Thus, in zero-sum competition, an incumbent firm would be encouraged to compete for survival to avoid exit costs. In Part II we will consider the implications of this in the context of predatory pricing. In the present chapter we will consider the implications of departing from the hypothetical competitive market standard when we address efficient entry to and exit from markets, ie, how the market mechanism works in zero-sum competition to adjust supply in response to shifts in demand. Conduct that harms competition can be justified in the EU on the grounds of necessity, meeting competition or efficiency.120 It seems that the equally ­efficient 116 The ACT in Qantas [2004] ACompT 9 [185] endorses the view that economic efficiency (total welfare), ie, maximising production and income from society’s scarce resources should be the goal. This view was opposed by the ACCC at [170]. See also ACCC v ACT [2017] FCFCA 150, which is noted in the Introduction. 117 Niels, Jenkins and Kavanagh (n 15) 185. Note that in the merger context, a structuralist SLC test will hold against increasing concentration per se, which may account for changing focus of the effects test in merger cases towards unilateral and coordinated effects. 118 Hovenkamp (n 16) 46; Niels, Jenkins and Kavanagh (n 15) 202–03. 119 Queensland Wire (1989) 167 CLR 177, 201 (Dawson J) and also 189–90 (Mason CJ and Wilson J); see also Hanks and Williams (n 73) 150. 120 Niels, Jenkins and Kavanagh (n 15) 182.

EU Abuse of Dominance and Defences  111 competitor and rational business decision tests discussed above are used to address ‘efficiency’ as the ultimate end of competition law. Nevertheless, as ­Richard Whish points out, the efficiency defence remains controversial because the EC is reading into Article 102 a defence which it ‘manifestly does not contain’.121 Thus, we see the difficulties for lawyers of reinterpreting statutory texts to accommodate changes in economic theory. There are limits on the ‘meeting competition’ defence, which recognise that the incumbent should not be allowed to prevent a more efficient new entrant from gaining market share.122 Thus, the CJEU has held that a dominant firm would have to show that, in meeting lower prices offered by a competitor, the dominant firm is doing no more than protecting its own interests, ie, is not also exercising its dominant position.123 In reality, this test is subsumed in the others discussed above. The EU position is of interest particularly in the light of the 2018 Australian decision in the Pfizer case. The General Court in France Télécom observed, according to established case law, that a dominant firm could be permitted to align its prices to those of a competitor where there was a competing offer for a particular customer.124 This defence would not be available if the dominant firm’s ‘actual purpose is to strengthen its dominant position and abuse it’.125 In Pfizer, the trial judge found that below-cost pricing was a strategy of generic pharmaceutical manufacturers during the launch stage of products for a relatively short period, so that s­ imilar conduct by the dominant firm would be a normal incident of competition at the time its patent expires.126 However, the issue was not pursued on appeal to the Full Court.127 There are two other aspects of the EU position that should be mentioned, though they will not be discussed further as they are not directly relevant to Australia. The first is the peculiar European concern to foster the common market, which has led to the doctrine that dominant firms have a ‘special responsibility’ not to use their power to impede the development of that market.128 The second is that Article 102 prohibits excessive pricing, unlike the US or ­Australian provisions, but according to Niels et al, writing in 2011, ‘it has no clear criteria and is not often enforced’.129

121 Whish (n 94) 160. The CJEU endorses efficiency in Intel (n 9) [140]. 122 Niels, Jenkins and Kavanagh (n 15) 211. 123 ibid, referencing the decision in Case T-340/03 France Télécom SA v Commission (General Court, 30 January 2007) [176] and [187]; affirmed on appeal, in Case C-202/07P France Télécom SA v ­Commission [2009] (CJEU, 2 April 2009) [45]–[47]. 124 Case T-340/03 France Télécom SA v Commission (General Court, 30 January 2007) [179]–[181] (affirmed on appeal). 125 ibid [185]. 126 Pfizer (n 31) [241]. 127 ibid [370]. 128 See, eg, Larouche and Schinkel (n 4) 158. 129 Niels, Jenkins and Kavanagh (n 15) 268–82; see also D Gerardin, A Layne-Farrar and N Petit, EU Competition Law and Economics (Oxford, Oxford University Press, 2012) 269–79. That assessment may well be dated, given current developments in Europe and Australia. The EC announced on 15 May 2017 that it will investigate possible excessive pricing of pharmaceuticals:

112  Competition and Efficiency Effects in Europe, North America and Australia It will be seen that the EC Guidance uses tests rather like the current Australian ‘purpose’ test to find a causal nexus between conduct and harm, and by focusing on rational business conduct to distinguish harm to competitors from harm to the competitive process. It will also be seen that the EU version of the SLC test is quite different from that of Australia and does not comprehensively take account of efficiency gains. These aspects have been criticised, ie, the Guidance might promote a more economic approach, but is said to be less clearly an ‘effects’-based approach. For example, Liza Gormsen points out that the ‘equally efficient competitor test is based on an inference of harmful effects’ and that the European courts had not systematically adopted it.130 Further, the approach in the EU prior to the Guidance was to infer effects from the form of the conduct, so it was said that the Guidance does not change the inferential nature of the effects test (such criticism is now superseded following the CJEU decision in Intel).131 Thus, the Intel decision makes it clear that it is necessary to establish actual effects where this is contested and that effects cannot merely be inferred from the nature of the conduct. I suggest that the meaning of purpose and effect are inter-linked and inextricable, whereas in Australia the courts have attempted to give each an independent meaning and, as we will see in Part II, have tended to infer effects from the nature of the conduct in question. On the other hand, Whish welcomes the EC Guidance because he is critical of European courts which ‘have failed, to date, to provide a definition of the concept of abuse that successfully distinguishes conduct that should be tolerated on the part of dominant firms from behaviour that should be prohibited’.132 He notes that there is substantial difficulty in determining whether market power exists.133 The Guidance describes market power as freedom from constraint by competitors or customers, in terms broadly similar to Australia’s section 46.134 The Guidance sets out the traditional structural approach to assessing market power.135 I suggest ‘Antitrust: C ­ ommission Opens Formal Investigation into Aspen Pharma’s Pricing Practices for Cancer Medicines’, press release, 15 May 2017, europa.eu/rapid/press-release_IP-17-1323_en.htm. Nevertheless, the difficulties of benchmarking prices and profits against the ‘competitive level’ are demonstrated by the decision of the Competition Appeal Tribunal (UK) on 7 June 2018 to set aside the decision of the Competition and Markets Authority imposing fines on two pharmaceutical companies for alleged excessive pricing practices: Flynn Pharma Ltd and Pfizer Inc v Competition and Markets Authority [2018] CAT 11 [443]. Australia’s prohibition on the misuse of market power has not to date been interpreted to prohibit excessive pricing, nor is there case law in Australia on the issue. Accordingly, this monograph does not address the subject further. 130 L Lovdahl Gormsen, ‘Are Anti-competitive Effects Necessary for an Analysis under Article 102 TFEU?’ (2013) 36 World Competition 223, 225–26. Gormsen was writing prior to the CJEU decision in Case C 413/14 P Intel Corporation Inc v EC [2017] (CJEU, 6 September 2017, rectified 19 September and 24 October 2017). 131 Gormsen (n 130) 245. The Intel decision of the CJEU makes it clear that it is necessary to establish actual effects where this is contested and that effects cannot merely be inferred from the nature of the conduct: Intel (n 9) [139]–[142]. 132 Whish (n 94) 154. 133 ibid 153, referring to ‘Guidance’ (n 3) [10]. 134 ‘Guidance’ (n 3) [10]. 135 ibid [12]–[14].

US Monopolisation and Defences  113 that our concept of market power needs to be reconsidered, since it exists by reference to the hypothetical competitive market, in which there is no market power. Finally, Whish addresses the controversial question that has arisen in the UK as to whether successful prosecutions of abuse of market power have been unacceptably low.136 He concludes that a significant factor is the belief that Article 102 ‘should be applied only in cases where conduct is likely to have seriously anticompetitive, exclusionary effects on the market’.137 This reflects a realisation that to do otherwise is likely to err in favour of protecting competitors rather than the process of competition.138 We need to be aware that competitors will use competition laws, and specifically provisions such as Article 102 and Australia’s section 46, strategically as part of the competitive process. We will see this in the Australian cases discussed in Part II. The pre-1986 Canadian law therefore seems to have some merit: prohibited conduct can be prevented by injunction, but there was no pecuniary liability at the suit of the regulator or civil liability to competitors who may make strategic use of claims. We may draw some conclusions from the EU for our analysis of the consequences of implementing the Harper proposals; however, before doing so, we should consider the US experience.

V.  US Monopolisation and Defences The Harper Review concludes that Australia’s ‘subjective purpose’ test is inconsistent with the corresponding US law, section 2 of the Sherman Act, citing a summary in a submission received from the American Bar Association (ABA): ‘Modern US decisions hold that it is not subjective intent but objective intent that is relevant, and that intent can be inferred from conduct and effect.’139 However, the ABA saw no practical difference between the Australian and US positions regarding ‘purpose’, and recommended against adopting an effects test, preferring a focus on conduct that has exclusionary results.140 It is useful to look in more detail at the US approach to ‘specific intent’, which requires that the sole motivation behind the firm’s conduct be the elimination of competition.141 This seems directed at the same end as the Australian concept of ‘substantial subjective intent’, which addresses the dual purposes problem and works around section 4F.

136 Whish (n 94) 171–74 notes eight successful and 46 unsuccessful prosecutions in nine years. 137 ibid 173. 138 ibid. 139 Harper Review (n 1) 340, citing a submission received from the American Bar Association, ‘Joint Comments of the American Bar Association Section of Antitrust Law and Section of International Law on the Australian Competition Policy Review Issues Paper’, 27 June 2014. 140 American Bar Association (n 139) 10. 141 See, eg, K Hylton, Antitrust Law: Economic Theory and Common Law Evolution (Cambridge, Cambridge University Press, 2003) 228.

114  Competition and Efficiency Effects in Europe, North America and Australia In fact, some US lawyers argue that section 2 of the Sherman Act, which is a criminal provision, employs a subjective intent test, though intent can sometimes be inferred.142 Hovenkamp observes that judicial decisions on the point have been inconsistent and that modern decisions have not addressed it, since intent can be inferred from conduct.143 Putting this another way, the necessary intent can be inferred when conduct is not that which would be expected of a rational profitmaximising firm.144 Hylton states that intent can be inferred from the ‘absence of credible efficiency justifications’.145 He traces the well-known trajectory of the US courts from a view that the Sherman Act promotes atomistic competition, even if that may entail inefficiencies, to one in which credible efficiency justifications can negate antitrust harm.146 Similarly, Hovenkamp traces the trajectory of economic and antitrust structuralism from one in which market definition, though inevitably arbitrary, is determinative of market power to one in which market definition is ‘no more than a second-best way to assess market power’.147 The arbitrariness of market definition and its implications for findings of market power is well known in Australia.148 Why then does the law seem to lag behind developments in economic theory? The answer proposed by Hovenkamp for the US could apply equally in Australia: Economists generally find it much easier to accommodate … conceptual shifts than lawyers and judges do, largely because of the strong value that law places on precedent and the relative infrequency of Supreme Court decision making in specific antitrust areas.149

Much US law is influenced by the role that juries play in the US system, unlike the Australian system. So we find that the US approach to intent in antitrust cases is influenced by a belief that juries have tended to exaggerate the role of evidence of hostile intent, with the danger that intent alone may be condemned without proof of exclusionary conduct.150 This is a similar issue to that of ‘purpose without effect’, which, as noted above, Australian courts have held should be condemned. Concern for jury fallibility may likewise help explain the ‘objective reasonableness

142 See H Hovenkamp, ‘Reimagining Antitrust: The Revisionist Work of Richard Markovits’ (2016) 94 Texas Law Review 1221, 1224; Hylton (n 141) 202 argues that ‘modern case law leans toward requiring proof of specific intent to monopolize’. 143 Hovenkamp (n 142) 1225, citing the well-known proposition that a person can be taken to have intended ‘the necessary and direct consequences of their acts’ (per Hand J in US v Patten, 226 US 525 (1913), at 543). 144 ibid 1226. 145 Hylton (n 141) 187. 146 ibid 186–205. 147 Hovenkamp (n 142) 1229–33. 148 See, eg, the quotation from Queensland Wire (n 119) 187–88 (Mason CJ and Wilson J) and at 195–96 (Deane J) in the Introduction, n 88. 149 Hovenkamp (n 142) 1237. 150 ibid 1224; Hylton (n 141) 192.

US Monopolisation and Defences  115 test’ in predation claims.151 This test may also be explained by the reluctance of US courts to second-guess business decisions, which some argue is beyond the court’s competence in any event.152 It seems that Hylton argues that it is undesirable for this reason to empower the court to balance, or weigh up, anticompetitive effects against efficiency gains.153 Hylton argues that apparent judicial rejection of ‘specific intent’ suggests that the court may undertake this balancing exercise.154 He argues that subsequent case law impliedly rejects this theory.155 I agree with Hylton’s conclusion that the court does not contemplate weighing up anticompetitive effects against efficiency gains. He also makes a strong policy argument that resonates with the EU debate to suggest that the weighing-up exercise is difficult if not impossible. Let us return now to the argument by Bernheim and Heeb that if the competitor is weakened (eg, by losing a sale to the dominant firm, the competitor may be denied economies of scale, such that its ability to compete in other transactions is impaired), then it could be argued that the process of competition is harmed.156 The paradigm of zero-sum (or survival) competition suggests both the incumbent and an equally efficient new entrant considered in the new entry scenario in Chapter 1 will seek to obtain a level of sales that will enable the successful firm to achieve scale economies and dominate the market, with the result that the other will not be viable and will exit the market. Should competition law take sides in this situation (where we cannot use efficiency as a tie-breaker and neither prices below cost)? This problem is a close analogy to situations analysed in the US. It is possible to distinguish commercial efforts to exploit scale economies from intent to harm the competitor, so in this case, evidence of intent to destroy the competitor would be unreliable, and what this study calls the dual purposes problem is avoided by asking whether the firm’s conduct is efficient.157 The situation is analogous to what is called ‘limit pricing’, which keeps the incumbent’s output high and

151 Hylton (n 141) 217. 152 ibid 191 and 228. 153 ibid 191–93. 154 ibid 193–96, discussing US v Aluminum Co of America 148 F2d 416 (2nd Cir 1945); US v Griffith 334 US 100 (1948). 155 ibid 203–05, discussing Aspen Skiing Co v Aspen Highlands Skiing Corporation 472 US 585 (1985). The argument appears to proceed from the observation that the jury finding that the dominant firm intended to destroy the rival was, strictly, irrelevant and that the key reasoning of the court was that the dominant firm ‘failed to bring forth any credible evidence of justification whatsoever’. From this, it seems that Hylton infers that the court does not favour the balancing exercise. I agree; however, it seems to me that the court does not say at 472 US 585, 602–03 that specific intent is irrelevant, but simply that it is not determinative and, by contrast, exclusionary effects caused by superior efficiency raise no issue of intent. 156 Bernheim and Heeb (n 4) 13. Their argument appears to be consistent with the reasoning of the ACT in Ford Motor Company (n 50) [17,499]. 157 Hylton (n 141) 215–16 asks which firm’s conduct is ‘objectively reasonable’; see his discussion of Barry Wright Corp v ITT Grinnell Corp 724 F2d 227 (1st Cir 1983) and Matsushita Electric Industrial Co v Zenith Radio Corp 475 US 574 (1986).

116  Competition and Efficiency Effects in Europe, North America and Australia prices low to deny a new entrant viable scale. Hovenkamp says that this is not condemned in the US as predatory behaviour, and observes: The principal obstacle to using antitrust to pursue above-cost limit pricing is thought to be limitations on the fact-finding power of courts, which generally lack the tools to identify when above-cost prices are anticompetitive.158

It is suggested that the Australian approach to ‘subjective purpose’ could solve the dual purposes problem, resulting in an outcome consistent with the US. Further, since the new entrant has the power to manipulate the market by increasing output in a race to achieve scale economies, I suggest that the incumbent under attack cannot be said to have market power – or, to put it another way, unless there is unmet demand, new entry may not be efficient.159 However, we will see in Part II that Australian courts may be substantially influenced by evidence of hostile intent and, consequently, that the Harper section 46 may well condemn limit pricing. Finally, we can consider the US law relating to ‘attempted monopolisation’. The test requires specific intent to monopolise (ie, a sole purpose test) plus an objectively dangerous probability of success.160 In other words, ‘purpose without effect’ is not enough.161 Absent credible pro-competitive justification, it seems specific intent may be inferred.162 We might then ask whether the new entry scenario considered in Chapter 1 involves an ‘attempt to monopolise’ by either the incumbent or the new entrant. Can the court determine which one involves the more dangerous probability of success? Should we wait and see which firm survives and then determine whether the survivor is the more efficient or whether the value created exceeds the value destroyed? It would seem that there is a strong argument that might gain traction in the US that the court should not get involved.163 Since the Harper Review proposes to repeal a series of statutory provisions dealing with predatory pricing, which were controversial and basically overrule the US approach to predatory pricing, we need not specifically discuss the US position on predatory pricing.164 It will be apparent that the US takes account of ‘efficiency gains’ through surrogate tests rather similar to our purpose test, and apparently eschews the need 158 Hovenkamp (n 142) 1229. Note, however, that in the EU, prices below average total cost but above average variable cost can be considered to be predatory, but only where they are fixed in the context of a plan having the purpose of eliminating a competitor: Case C-202/07P France Télécom SA v Commission [2009] (CJEU, 2 April 2009) [109]. 159 See also the discussion in ch 1 of the Bertrand Paradox (n 129) and inefficiencies from duplicated capital investment by new entrants. 160 See, eg, Hylton (n 141) 244–51. 161 Hylton (ibid 248) gives the example of a person trying to carry out an armed robbery with a banana as a case of purpose without possible effect. 162 ibid 246–47. 163 See also, eg, C Sunstein, ‘Interpreting Statutes in the Regulatory State’ (1989) 103 Harvard Law Review 405, 487 (referring to antipathy to ‘naked interest-group transfers’, which he defines as ‘transfers which lack public value and are an exercise of political power of private groups’, eg, the tension between big business and small business). 164 Sections 46(1AA), (1B), (4A) of the CCA inserted in 2007; section 46(1AAA) inserted in 2008 (see the legislative history to s 46 in the Appendix; for the US cases, see, eg, Hylton (n 141) 212–28.

Efficiency in the Law of Securities Market Manipulation  117 to weigh up, or balance, anticompetitive effects against efficiency effects. So we gain no assistance in working out how we might do this in Australia if the Harper proposal to require the court to consider efficiency had been adopted. Some thoughts on efficiency follow, deriving what assistance we can from the EU and US position and Australian laws prohibiting manipulation of securities markets.

VI.  Efficiency in the Law of Securities Market Manipulation We noted earlier judicial usage of the term ‘efficient market’ in the context of securities markets manipulation laws, where it is used to refer simply to the undistorted interplay of forces of supply and demand.165 I adapt this concept in the present context, where we focus on whether supply is responding to demand, eg, whether there is unmet demand.166 Thus, our focus is different from securities markets manipulation laws which typically focus on manipulation that creates an artificial price. I suggest that the latter focus may explain the problems encountered in that area of law with the concepts of ‘purpose’ and ‘effect’, which largely mirror the problems encountered by the courts with ‘purpose’ and ‘effect’ under Part IV of the CCA. Sections 1041A and 1041B of the Corporations Act 2001 (Cth) create an ‘effects’ test for market manipulation. However, the courts accept that ‘purpose’ may nevertheless ‘loom large’ in assessing effects.167 In ASIC v Westpac, it was contended by the ASIC that effects could be inferred from evidence of purpose, but that claim was rejected on the facts of the case.168 We will see in Part II that there is a similar tendency in section 46 cases that also involve competition test issues under, eg, section 45 or 47 for the ACCC to rely on evidence of purpose rather than to seek forensic evidence of ‘effects’. I suggest that ‘purpose without effect’ can be regarded as an ‘attempt’ which can be dealt with as such, and is of little assistance when we forensically examine the efficient operation of the market. We can develop this legal concept in our present context by considering how conventional concepts of economic efficiency can be adapted to elucidate the efficient working of markets. We may thus be able to sidestep the longstanding debate in the literature about conflicting objectives, or efficiency standards, in competition law. 165 See, eg, ASIC v Westpac Banking Corporation (No 2) [2018] FCA 751 [1925]–[1926]. 166 Niels, Jenkins and Kavanagh (n 15) 483 discuss ‘unmet demand’ as an indicator of competitive harm, though they do not define it. We can take this further as an indicator of inefficiency. I define ‘unmet demand’ to mean demand that could profitably be supplied by the firm but is not being supplied, eg, because the firm is capacity-constrained. Given the critique in this monograph of the assumptions which underlie the ‘competitive level’, I suggest that it would not be appropriate to define ‘unmet demand’ by reference to that hypothetical. 167 ibid [1922]. 168 ibid [1951]. However, Beach J agreed at [1955] that ‘effect without purpose’ would infringe the prohibition.

118  Competition and Efficiency Effects in Europe, North America and Australia

VII.  Adapting Efficiency as a Legal Concept As we have seen, there are frequent references to ‘efficiency’ in the context of unilateral dominant firm conduct, but the discussions generally avoid getting to grips with the concept.169 We have also seen that ‘consumer welfare’ is an ambiguous term. This is because the interests of consumers are varied and some conflict, ie, consumers have an interest as producers (income earners) to maximise their production (income) and as consumers to buy the mix of goods and services that is most useful to them at the best price. Accordingly, in this section, we explore the subject of efficiency and how it may be assessed when we move from the level of economic activity in the whole economy to economic activity in a market (ie, industry) or in a firm. Just as our concepts of market power and harm from market power derive from the model of perfect competition, despite the generally accepted problems this creates for competition law, it is to perfect competition that we turn to understand ‘efficiency’. I suggest that this approach can provide only limited guidance on efficiency in real-world markets. Chapter 1 discusses efficiency as traditionally conceived under the paradigm of perfect competition. After a further brief discussion of these concepts, we will consider how we can assess efficiency when we shift our paradigm to zero-sum competition. This discussion suggests that we will have to adapt economic concepts for use in our legal theory of liability. The discussion of various kinds of efficiency as conventionally understood is assisted by referring to the ‘production possibilities frontier’ depicted in Figure 3.1. Figure 3.1  Allocative, productive and dynamic efficiency in a two-product economy Y

EP

MRS = MRT X

169 See, eg, G Gundlach and D Moss, ‘The Role of Efficiencies in Antitrust Law: Introduction and Overview’ (2015) 60 Antitrust Bulletin 91, 97–78, who suggest the concept of efficiency is clouded by differing views as to the objective(s) of competition law.

Adapting Efficiency as a Legal Concept  119 The ‘frontier’ indicates the maximum possible production (and income) in the economy from available resources. Of course, in reality, such a frontier would be in constant motion, but for our purposes, it is at rest. At the frontier, production is efficient in the sense that average costs are minimised. We can thus infer that the costs of capital are minimised and that transaction costs are minimised.170 It is not possible to produce more of any product without diminishing output of another. Under perfect competition, not only will output be maximised, but the allocation of resources between production of product X and product Y in Figure 3.1 will also yield the maximum community welfare because consumers as a whole choose the relative quantities of the products they desire (equilibrium Ep). In the case of allocative efficiency, efficiency is relative, ie, efficiency in this sense does not exist in one market in isolation. The process by which productive efficiency is increased by technological innovation is referred to as ‘dynamic efficiency’.171 This cannot be depicted by a static analysis. Nevertheless, at point Ep, efficiency may also be dynamic because as consumer preferences change, consumers substitute one product for the other according to relative prices, and producers transform output of one into output of the other according to relative marginal costs – yet only under perfect competition are these ratios the same (and only under perfect competition can firms freely move from one market to another). Important assumptions of perfect competition include: all goods and services in the economy are produced in perfectly competitive markets; there are no externalities, ie, all social costs or benefits are captured by the price of goods and services; consumers and producers have perfect information and perfect knowledge about products and choices; and firms may enter and leave the market without transaction costs. Thus, no firms fail or become insolvent under perfect competition because a firm knows when it will be outclassed and leaves the market before it becomes necessary to compete for survival. We saw in Chapter 1 that economists use non-zero-sum games to analyse strategic behaviour in imperfectly competitive markets and that this involves a competitive paradigm not unlike perfect competition, which may be described as ‘peaceful coexistence’. Thus, behaviour consistent with this paradigm, directed towards a Nash equilibrium, may be considered rational. This, however, does not explain how firms self-select for entry to or exit from markets under zero-sum competition, where firms wishing to enter may find that in order to do so, they must destroy the incumbent, and firms which cannot practicably exit the market due to exit costs must fight for survival. We also saw that lawyers operate within the framework of the legal system, which is one of zero-sum competition. It is suggested that these conflicting views of the world create far-reaching problems for competition law. 170 In economics, normal profits on equity capital are regarded as ‘costs’; as to transaction cost efficiency, see, eg, ACCC Merger Guidelines (n 42) [7.63]; A Meese, ‘Reframing Antitrust in Light of Scientific Revolution: Accounting for Transaction Costs in Rule of Reason Analysis’ (2010) 62 Hastings Law Journal 457. 171 See, eg, S Bishop and M Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, 3rd edn (London, Sweet & Maxwell, 2002) 36–37.

120  Competition and Efficiency Effects in Europe, North America and Australia There is of course no unanimity in the field of competition law and economics as to the proper objective of competition law.172 In the present context, this affects the standard of efficiency to be adopted in order to assess the effects of dominant firm conduct. One would intuitively conclude that it is better in the event of inconsistency between the objectives of productive and allocative efficiency for the economy to be as productive as possible (because that maximises incomes and produces lowest-cost goods and services) than it would be to pursue allocative efficiency if that may produce suboptimal output. Influential thinkers argue that productive efficiency (total welfare) should be the standard.173 In attempting to apply considerations of efficiency to actual cases, we make an important shift in focus from productive and allocative efficiency at the level of the whole economy to the level of a single market or industry. In so doing, our focus turns to ‘producer surplus’ (ie, some producers would have supplied at less than the market price and so are better off from trade) and ‘consumer surplus’ (ie, some consumers would have paid more than the market price and so are better off from trade).174 I suggest that this is of limited use for an enquiry into efficiency at the level of the market, because it ignores the ‘relative’ nature of allocative efficiency and effects on productive efficiency outside the market. We saw in Chapter 1 that the assumptions underlying the economic models enable the conventional analysis that the ‘competitive level’ is an indicator of not merely private welfare within the market, but of social welfare across the economy. Thus, if the necessary assumptions are not satisfied under real-world conditions, we cannot assume that increasing output and reducing price towards the assumed competitive level within a market (or industry) will have no effect outside the market. That is, maximising the sum of producer surplus and consumer surplus (and consumer surplus itself) may reduce efficiency by drawing resources from other productive uses in the economy, ie, might simply redistribute surplus without any benefit to society as a whole. This may especially be the case if we do not enquire whether the dominant firm has achieved its position through success in business. The forensic difficulties with ascertaining the ‘competitive level’ (ie, even accepting that the assumptions can be relied upon) compound the already problematic nature of the concept. We know that markets are agnostic about the interests of producers and consumers, ie, markets work by excluding both producers and consumers for whom there is no benefit from trade. We have noted that the market excludes firms when market size does not allow firms to operate at efficient scale.175 Thus, markets can be said to exclude firms if they are not needed for the efficient operation of the

172 See, eg, Hovenkamp (n 65) 383. 173 ibid. See also Gal (n 16) 55. 174 These concepts were discussed in more detail in ch 1. 175 What is ‘market size’ when we know demand responds positively to decreases in price? Arguably, market size in the case of straight-line demand curves typically used to illustrate monopoly is the point where marginal revenue changes from positive to negative, ie, revenue peaks and will decline if price falls below that point.

Adapting Efficiency as a Legal Concept  121 market. Similarly, consumers who would rather not buy at the market price may decide to buy something else (ie, another mix of products) or to save. I suggest that it cannot be said that they are worse off from that choice alone, because community welfare derives from the total mix of goods and services consumed in the economy, ie, allocative efficiency is relative and any assessment requires assumptions to be made about the situation in all other markets.176 One might argue, by analogy with exclusion of producers, that it is more efficient for such consumers to apply their resources elsewhere. Further, the better view appears to be that ‘equity’ or redistributional objectives are not the proper province of competition law.177 This is not a value judgement about equity as a social objective, but rather simply an observation that markets and competition do not and cannot address it. In other words, any attempt by regulators to pursue equity is mere regulation rather than an application of competition policy. Nevertheless, we may discern a concern for equity in the EC’s guidance on dominant firm conduct and the approach of the ACCC to merger efficiencies.178 When we make judgements about the balance of producer surplus and consumer surplus in real-world conditions where the assumptions underlying the economic models are not satisfied, I suggest that we are concerned with equity, not efficiency. This is particularly the case if we object to excessive pricing by a dominant firm and wish to redress market power by reducing prices. This concern exacerbates the problem of drawing conclusions from forensic evidence of market prices relative to some hypothetical ‘competitive level’,179 a problem acutely experienced in the Australian cases we will examine in Part II. However, I suggest that we can conceptually distinguish a situation where the market has been manipulated, as this distorts the efficient operation of the market mechanism. We thus need to consider the way in which markets work to bring supply and demand into balance. As we saw in Chapter 1, the familiar market mechanism of perfect competition (where both traditional supply and demand curves respond to price) does not operate under imperfect competition because supply responds to marginal revenue, not average revenue (ie, price). We can illustrate the problem by reference to the remarks of Kirby J (dissenting) in Boral, which was a case where market demand dropped due to an economic downturn, but firms maintained or 176 Contra: H Hovenkamp, Antitrust Enterprise: Principle and Execution (Cambridge, MA, Harvard University Press, 2005) 19, who argues that consumers who are excluded from the market derive less utility from their second preference, ie, make ‘inefficient substitutions’. This might be the case under perfect competition, where it is assumed that all other markets in the economy are perfectly competitive. If the necessary assumptions are not satisfied under real-world conditions, this conclusion may not be justified. Further, if all markets in the economy are imperfect, we cannot know whether price relativities that govern substitutions may effectively facilitate substitution. 177 See Qantas (n 116) [185]. See also Gal (n 16) 48. See also ACCC v ACT [2017] FCAFC 150, which is noted in the Introduction. 178 The view of the ACT was opposed by the ACCC in Qantas (n 116) [170]. See also the Harper Review’s suggestion that competition reduces prices and ensures access for vulnerable ­consumers: (n 1) 20. 179 See, eg, Hovenkamp (n 176) 102.

122  Competition and Efficiency Effects in Europe, North America and Australia increased supply, which would be unexpected under perfect competition. Kirby J observed that under such circumstances, he would expect that ‘one rational or logical response would have been for [firms] to use their capacity less intensively’ (emphasis in original).180 We can consider how the market mechanism works under perfect competition and alternatively zero-sum competition by reference to Figure 3.2 below. A simplistic (but incorrect) analysis might suggest that if the market is in equilibrium at price P1 and demand shifts downward from D1 to D2, firms would reduce output down the supply curve to re-establish equilibrium at price P2. This appears to reproduce the rational or logical response anticipated by Kirby J. Figure 3.2  Effect of demand reduction (if there were a supply curve) Supply1

Price Supply2 P1 P2 P3

Demand2 Q3

Q2 Q1

Demand1 Quantity

However, if the market were in equilibrium at price P1 under perfect competition, we could infer that average costs (including normal profit) would be equal to P1, ie, firms could not profitably produce at price P2 and so some firms would selfselect to exit the market (which they could do freely without cost). Consequently, the supply curve would move to the left, restoring price P1 at a reduced output to correspond to the reduced demand.181 Thus, free exit of firms is part of the working of the market mechanism under perfect competition. We may suspect that if firms maintain (ie, do not reduce) output, there is some other operative factor in their decision. Clearly, the ACCC and Kirby J thought that the dominant firm in the case in question was seeking to force out rival firms, though the majority of the High Court disagreed. But is this ‘manipulation’ or is it a rational business decision (and so may well be both privately and socially efficient) in the circumstances? We know that under imperfect competition, there is no defined supply curve, and supply responds to marginal revenue (this is discussed in more detail in

180 Boral (2003) 215 CLR 374, 508. 181 See, eg, J Gans, S King, M Byford and NG Mankiw, Principles of Microeconomics, 7th edn (Melbourne, Cengage Learning, 2018) 317–18, Figure 14.8, which illustrates how supply responds to demand shifts.

Adapting Efficiency as a Legal Concept  123 Chapter 1). As such, we may rule out both the analysis under perfect competition and the analysis of Kirby J above. We know that there are barriers to exit in the real world and that under zero-sum competition, new entrants seek to destroy incumbents, and incumbents seek to survive. We can infer in the above figure that the marginal revenue curve associated with D2 has also moved downwards to the left, with the result that output Q1 could well be an inefficient use of society’s resources for a further reason: if output exceeds that required to satisfy demand (ie, if marginal revenue at that point has become negative). Can socially inefficient production be justified? It seems to be the case that firms have a greater incentive to become more efficient (or innovate) under zero-sum competition (when exit is a costly and unpalatable option) than they do under perfect competition (when free exit removes the need to reduce costs). Thus, we may see zero-sum, or survival, competition as the normal working of the market mechanism under real-world conditions, where we may tolerate some socially inefficient production as the price we have to pay for ultimately increasing efficiency. In zero-sum competition, a firm can maintain or increase its revenue by taking sales away from competitors (ie, regardless of overall market demand response). That is clearly a rational strategy when demand is constant and may well be a compelling strategy to protect revenue when there is a downturn in demand. This brings us to the last level at which we might assess efficiency, ie, the level of the firm (noting that a monopolist is both the market and the firm). Clearly, profitmaximising behaviour can be regarded as privately efficient if the firm’s revenue is maintained or increased relative to its competitors. This can be assessed forensically either by examining the subjective decision-making process or the objective indicators of revenue, costs and profit. Conduct which cannot be explained on efficiency grounds (such as maintaining or expanding output despite a downturn in market demand) may well be ‘market manipulation’. In the situation depicted in Figure 3.2, when we apply the market manipulation approach, we must ask: first, whether any firm has power to manipulate the market when there is excess capacity and competition for survival; and, second, whether conduct is efficient from the firm’s perspective or may be condemned as manipulative (ie, to distort the efficient working of the market). It is well known that elasticity of demand is difficult to predict and the best laid plans often go awry, so perhaps subjective decision-making is for this reason an appropriate starting point. Here we begin to confront a startling proposition: competition may propel firms to expand market output beyond the level required to meet demand by offering the successful firm the opportunity to expand sales at the expense of rivals; that is, zero-sum competition may produce outcomes that are inefficient for the market as a whole. We must therefore consider efficiency in the context of zero-sum competition. Before turning to that subject, we should consider conditions for optimising efficiency. The production possibilities frontier embodies the proposition that (at the frontier) output of any product cannot be increased without reducing output of another (Pareto efficiency). Productive efficiency (like allocative efficiency) is

124  Competition and Efficiency Effects in Europe, North America and Australia relative. Thus, we can say that requiring a monopolist to increase output at the expense of increasing costs (thus dissipating excess profits) may well be inefficient because increasing output comes at the expense of drawing resources from other uses, ie, diminishing output of other products. If, as some suggest, we apply the Hicks-Kaldor test for optimising efficiency, we might ask if the value of increased production outweighs the lost output of other products in other markets.182 We cannot apply these tests to a market or a firm in isolation without knowing the impact on other products and firms. Thus, social efficiencies will rarely if ever be legally determinate in competition law cases. When one firm survives at the expense of another, it would be problematic to assess the value created compared to the value destroyed. Perhaps assessing the relative efficiency of the two firms may be an appropriate surrogate test. However, we must be mindful that the apparent justification for potentially inefficient outcomes under zero-sum competition is the dynamic potential for innovation and increasing efficiency, which may well be lost if the law intervenes to penalise an apparently less efficient incumbent firm. We will take this further in Chapter 9 when we reconsider the NT Power case under the market manipulation and efficiency approaches. Given that perfect competition and congruent optimisation of productive and allocative efficiency is not achievable, I suggest that our legal benchmark for efficiency should be what is achievable in the real world, having regard to the paradigm of zero-sum competition. We can only ask whether the conduct in question promotes or reduces efficiency, having regard to the position of the firm and the market as it exists in reality. This is a different approach from the structuralist approach to the SLC test, which benchmarks SLC against the structuralist ideals of a hypothetical competitive market. Further, when we focus on the efficient operation of markets, we can see how the relative efficiency of firms and issues concerning entry and exit of firms can affect market efficiency.

VIII.  Zero-Sum Competition, Market Manipulation and Efficiency In Chapter 1 we saw that the paradigm we adopt to reflect real-world competition, which I call zero-sum competition, affects how we legally conceptualise market power and assess its misuse. In summary, zero-sum competition entails: (a) firms competing for sales by taking sales away from each other or seeking to grow sales faster than the market (which gives rise to the ‘dual purposes’ problem); (b) entry and exit is not free – rather, incumbent firms incur costs on exit that encourage them to fight for survival rather than exit the market;

182 See

Hovenkamp (n 65) 384.

Zero-Sum Competition, Market Manipulation and Efficiency  125 (c) new entrants may overestimate their chances of entering the market and successfully taking sales away from the incumbent; (d) all firms, including new entrants, have some degree of power, by their price/ output decisions, to instigate a demand response. I suggest that market power can be legally conceptualised as power to manipulate the market. We want to ensure that the process by which supply and demand adjust to each other is undistorted. We derive this from the legal concept of an ‘efficient market’ in securities markets manipulation laws. In this chapter, I suggest that economic efficiency can be adapted to provide us with a legal concept which we can use forensically to assess whether conduct is manipulative, because inefficiency interferes with the operation of the market and so may be condemned. Our central forensic enquiry is to ask whether there is ‘unmet demand’ which could be efficiently supplied by new entry. This concerns: first, whether it would be profitable for the dominant firm to increase output (presumably not if it is producing profit-maximising output); and, second, what is the size of the market, ie, the level of output beyond which demand response fails to justify the expenditure of additional resources (illustrated in Figure 1.5 in Chapter 1). We will consider in Part II how this can be forensically determined. Suffice to say here that unmet demand would be indicated by, eg, shortages or delays in delivery. We can now identify several propositions concerning the relationship between manipulative conduct and efficiency. These propositions emerge not only from the analysis of market power and efficiency by the process of reasoning set out in this monograph so far, but, importantly, they are also derived from testing the market manipulation approach in Part II against the factual matrices of key Australian cases concerning misuse of market power. Accordingly, I suggest that: (1) A monopolist may be expected to meet market demand. This is driven by profit-maximising behaviour. We can justify such conduct not because it is rational, but because it may well be both privately and socially efficient. Thus, when there is unmet demand, ie, demand that could be satisfied without reducing profit, we would want to know why. For example, is the firm capacity-constrained and unwilling to invest in further capacity because investment is not justified by current demand or anticipated demand growth? (2) When demand is being satisfied, new entry may not be efficient, ie, the size of the market may preclude new entry. We should not condemn an incumbent dominant firm for, eg, refusing supply or competing to retain sales where new entry would be inefficient. We assess this by reference to the dominant firm’s ability to meet demand, the relative efficiency of the new entrant and the ability of the new entrant to satisfy unmet demand. Where the incumbent is meeting demand, the effect of new entry by an equally efficient firm may well be simply to redistribute surplus from the incumbent to the new entrant with no benefit to society. (3) Firms locked in zero-sum competition may manipulate the market by reducing price and increasing output to inefficient levels, eg, price wars.

126  Competition and Efficiency Effects in Europe, North America and Australia Similarly, by engaging in bidding wars for inputs, they may increase price to inefficient levels. Where no firm can be blamed, eg, a demand shift reduction, or a new entrant creating excess capacity, which creates disequilibrium, we may have to tolerate inefficient outcomes because these may be seen to result from the paradigm of zero-sum competition itself. Additional grounds to tolerate such outcomes would be the resulting transfer of surplus to consumers or (in the case of bidding wars for inputs) transfer of surplus away from the dominant firm. Such transfers may mitigate harm. (4) The exclusion of a more efficient firm would appear to be manipulative, as we expect more efficient firms to enter and survive. We need to be particularly careful about the exclusion of equally efficient firms, since the market itself may exclude such firms where market demand can be satisfied without them. We know that efficient firms can fail through underestimating risks and that less efficient incumbent firms can improve their efficiency and survival chances. We can assess a firm’s technical and financial efficiency by forensic enquiry concerning, eg, its ability to operate assets productively and to cover costs of capital on a long-term basis. We know that incumbent firms face exit costs that may justify a decision to stay in the market and fight to survive, eg, by becoming more efficient. We cannot condemn that kind of response to new entry and should not prevent it, since increasing efficiency benefits society. I suggest we now have a basis for forensic enquiry into the conduct of firms having power to manipulate the market, by adapting norms of economically efficient conduct for use under the real-world paradigm of zero-sum competition. In doing so, we have drawn on the legal concept of an ‘efficient market’ from the law concerning the manipulation of securities markets. In other words, we now have a positive and normative legal theory of liability for misuse of market power. We can now form some conclusions on the ability of competition law, specifically the Harper Review section 46 as enacted, to deal with ‘efficiency effects’ and assess different legal and judicial approaches to weigh up adverse ‘competition effects’ against positive ‘efficiency effects’.

IX. Conclusion It appears to be true in Australia no less than overseas that economists are more ready to accept developments in economic theory than are lawyers and judges who must deal with binding precedents. The courts are well aware of the problems of selecting isolated principles from overseas laws for adoption in Australia without full regard to the contextual framework in which they operate.183 183 See, eg, Burge v Swarbrick (2007) 232 CLR 336 [62]; NT Power (n 45) [121]; DPP (Cth) v JM (2013) 250 CLR 135 [77].

Conclusion  127 The Harper Review may well have focused on isolated issues without detailed analysis of the overseas frameworks. Thus, when we analyse in more detail the legal context in Canada, the EU and the US, we see that there is no internationally accepted ‘effects test’ – the approach in those jurisdictions is very much a product of the legal framework in which it is located. One consequence of this is that extending the Australian SLC effects test found in other sections of the CCA to section 46 will not result in uniformity to any meaningful extent with the laws of those countries. While the express recognition of efficiency in the proposed Harper section 46(2) should be welcomed, it raises new issues for Australian courts on questions that are regarded as controversial overseas. However, the removal of that provision by Parliament does not solve the problem of how we should address defences recognised internationally. For reasons of certainty and predictability, it would be better to avoid or clarify these issues in the legislation than to wait for appellate courts to attempt to do so. The passage of the amending legislation by Parliament now commits us to the latter course. We see that the key reasoning used by the Harper Review to advocate change – that one should protect competition, not competitors – has been used in Canada, the EU and the US for different (even opposite) purposes. We also saw in Chapter 1 that the reasoning of the Senate Economics Legislation Committee in rejecting the Harper section 46(2) appears to suggest that the law should protect competitors. It is necessary to inquire more deeply into the common issues that abuse of dominance laws address so that we can understand the methods adopted in other jurisdictions. Thus, we find that subjective intent or purpose has been used internationally to distinguish legitimate self-interested conduct that incidentally harms competitors from conduct that may harm competition. Purpose and effect are integrally linked, and we see that ‘purpose’ is a useful screen to isolate harmful effects, and effects indicate a relevant purpose, which suggests that our test should require both purpose and effect. In fact, we see that ‘purpose’ is used as a surrogate for efficiency in Canada, the EU and the US. Another key part of the Harper Review’s reasoning is the proposition that conduct that may be innocuous if carried out by a firm without market power may have the potential to SLC if carried out by a firm with market power. When we analyse this proposition, we see that the real concern is more with the hypothetical competitive market standard than it is with the requirement that there be some causal nexus between the market power and the conduct. It is unclear whether and, if so, how it is to be shown that market power contributes to the SLC ‘effect’ under the Harper proposal. If the provision is construed as a ‘no fault’ or strict liability provision, it would discriminate between powerful and less powerful firms, and may well discourage competition on the merits by preventing dominant firms from meeting competitive challenges. I suggest that the requirement in the previous section 46 that there be some causal nexus should be found to be present in the Harper section 46 and thus retained. This, of course, is a matter of interpretation for the courts.

128  Competition and Efficiency Effects in Europe, North America and Australia There is a tendency for impact on consumer surplus to be used as an indicator of efficiency effects in the EU, but less so in the US, and in Australia opinions differ as to the distinction between ‘equity’ and ‘efficiency’. The Harper Review does not clarify the preferred approach to efficiency in its recommended section 46. It would be desirable to do so lest differences of opinion that currently exist should continue to cloud the application and interpretation of the law by the courts. Bearing in mind that competition laws seek to address behaviour, I suggest that a ‘purpose’ test that adopts the norms of rational business decision-making as the standard has advantages. It would provide a useful surrogate test for efficiency and a behavioural standard to guide compliance. We have seen that in the EU, it is considered that attempts to weigh up and offset negative ‘competition effects’ against positive ‘efficiency effects’ are difficult (even if agreement could be reached on the appropriate efficiency standard – consumer welfare or total welfare). In the US, it is considered that courts do not have the expertise to do this. The Harper Review proposed that section 46(2) is based on the assumption that courts can do so, though how the courts may do so would be left to judicial interpretation. I suggest that this ‘weighing-up’ approach is not supported by experience in the US and the EU. To avoid legal indeterminacy under current laws, I propose that it would be better to proscribe conduct by reference to both ‘purpose and effect’ and that affirmative defences should be provided by legislative prescription to enable a dominant firm to justify its conduct by reference to efficiency or meeting competition. In the absence of such legislative prescription, it is possible that the courts may over time develop the law by common law method.184 Nevertheless, despite repeated calls in the literature in Australia and internationally over the past 30 years to give economic efficiency a central role in assessing misuse of market power, the chances of Australian courts taking this course cannot be predicted. For these reasons, I suggest that the market manipulation approach is worthy of consideration. In Part II we will test the proposed market manipulation approach, and the Harper Review recommended efficiency approach set out in the Harper section 46(2), by reference to the facts of key section 46 cases. We may contrast these approaches with the section 46 SLC test as passed by Parliament, which omits the Harper requirement to consider innovation and efficiency. While we do not expect there to be sufficient data to determine efficiency issues, we will be able to examine forensic issues and test whether ‘efficiency’ as adapted for our legal theory can be forensically determined by the courts carrying out traditional judicial functions of fact-finding and application of legal principles. We can then assess conditions of possibility for outcomes of the cases under different legal approaches to regulating the use of market power.

184 See High Court discussion of this methodology in the context of statutory patent law, where the peculiar drafting of the Patents Act 1990 (Cth) invites the court to do so: D’Arcy v Myriad Genetics Inc [2015] HCA 35. It is not obvious that the CCA contains a corresponding invitation.

part ii Testing Market Manipulation and Efficiency Approaches Introduction In this Part we will test the proposed market manipulation approach and the Harper Review section 46 (with and without the Harper section 46(2), which expressly requires consideration of efficiency effects) against the facts of leading section 46 appellate cases regarding practices recognised internationally to raise market power issues. We want to contrast the legal theories of liability to see what difference the proposed new theories would make. This Part will be organised into the following chapters: Chapter 4,

‘Refusal to Deal and Margin Squeeze’

Chapter 5,

‘Predatory Pricing’

Chapter 6,

‘Meeting Competition’

Chapter 7,

‘Raising Rivals’ Costs’

Chapter 8,

‘Bundling’

In Chapter 9 we will consider whether the current institutional division of functions between the court and the ACCC/ACT remains appropriate under the Harper section 46 or the market manipulation approach. In that chapter we will also address procedural issues such as the burden of proof. Each of the chapters noted above sets out chapter-specific conclusions. Part III sets out more generally the conclusions to be drawn from the study and the forensic implications of applying proposed new legal approaches to practices of the above kinds. It is not our purpose in Part II to critique the application of past and present law by the courts. Most of these cases have been controversial and much has been written about them, to which we need not refer in detail. The Harper Review section 46 does not alter current jurisprudence concerning market power, so our focus is upon assessing competition and efficiency effects which become relevant under the Harper proposals. The amending Bill to implement the Harper Review changes has been passed; however, Parliament removed the Harper recommended requirement that the court must consider efficiency. Thus, doubts remain as to whether efficiency is a relevant mitigating factor where conduct would appear to

130  Testing Market Manipulation and Efficiency Approaches substantially lessen competition (SLC). This study will nevertheless examine how efficiency effects may be forensically assessed and the relationship between such effects and the SLC test under the Harper proposals. Under the new legislation, the ACCC/ACT will for the first time be able to ‘authorise’ conduct, which would otherwise contravene section 46, if there is an overriding efficiency or other public benefit. As we saw in Chapter 2, the authorisation provisions do not refer to efficiency, and while efficiency has been considered in authorisation matters, there is no body of precedent providing direct guidance as to how the ACCC/ACT will take it into consideration in section 46 cases. By way of contrast with the Harper proposals, the market manipulation approach invites us to reconsider the legal concepts of market power and efficiency. By focusing on the power of the market, we are able to identify conduct of a firm that responds to market demand without distorting the market. We adapt the legal concept of an ‘efficient market’ from the law on securities market manipulation. This approach proposes that such conduct should be permitted not because it is rational to an economist (or logical to a business person), but because it may be both privately and socially efficient. We cannot measure the impact on social efficiency, but equally we cannot assume it to be neutral. We must therefore adapt economic concepts of efficiency for use in our legal theory. In the present Part, we test how our legal concept of efficiency can be forensically determined. We consider whether there is unmet demand, ie, whether an expansion of output would be profitable and also whether the demand response would justify drawing resources from other socially valuable uses in the economy. We look at the relative efficiency of firms to help determine whether the market is operating without distortion. Our discussion in Chapter 1 of the Prisoner’s Dilemma game shows that economists, lawyers and regulators bring different views of the world (and associated norms) to the task of distinguishing normal competitive conduct from that which should be condemned. I suggest that norms which simply reflect differing views of the world are not helpful and that we may transcend differing worldviews by adapting and redefining efficiency as the relevant benchmark. I propose that conduct which is efficient can be considered not to be manipulative, and conduct which is inefficient may be manipulative but requires further examination – ie, has the conduct interfered with the undistorted operation of the market? Some inefficient conduct may not be blameworthy; for example, strategies may be efficient in their conception, but may fail and so have inefficient outcomes. If we adopt a ‘no fault’ (or strict liability) prohibition, will firms be less inclined to attempt efficient conduct that carries a risk of failure? Similarly, when firms which are locked in zero-sum, or survival, competition increase output beyond the point where market revenue begins to decline (such as in a price war), it may be impossible to attribute blame to one firm alone. Should we accept this as a necessary consequence of zero-sum competition? Can we tolerate social inefficiency in this case because it may be mitigated by a resulting transfer of surplus to consumers? We must also confront the opposite situation, where firms engage in a bidding war for

Introduction  131 inputs, forcing the price above the presumed competitive level. In this case, we ask whether we can tolerate inefficiency because it may be mitigated by a transfer of surplus upstream, ie, away from the dominant firm. Accordingly, this Part examines past cases for their factual matrix so that we can apply the new or proposed laws. As will become apparent, the factual matrix presented to, and found by, the court at first instance depends largely on the strategic and tactical choices of the parties in running their case and on the court’s view of legal issues that will be determinative of the case. Such issues present themselves quite narrowly to appellate courts. Thus, there will be gaps in the factual matrix that would have to be filled in order to reach an outcome under the new or proposed laws. As the present study is not an empirical study, these gaps will be addressed by suggesting conditions of possibility for the case to fall either way, ie, alternative factual circumstances which would be required to uphold or dismiss the alleged contravention. While Australian courts appear to conscientiously avoid hindsight when determining competition law cases, we may usefully fill gaps in the factual matrix by reference to subsequent market experience. By so doing, we find that market forces and market conditions are constantly in flux. We will see that the time that typically elapses between events occurring which give rise to litigation and final appellate disposition of the case is usually so great that the competitive implications of the case may be overtaken by subsequent events. This has significant implications for the usefulness of the predictive SLC test that will be imported from other provisions of Part IV of the CCA into section 46 under the Harper Review amendment. In the first chapter in the present Part, we will examine the category of conduct known as ‘refusal to deal and margin squeeze’.

132

4 Refusal to Deal and Margin Squeeze I. Introduction The key issue that the Australian cases have not addressed, but which is addressed in the US and the EU, is the limits of the proposition that competition law does not prohibit a monopolist from deriving monopoly profits.1 As we will see, it is not clear from the judgments that the High Court subscribes to this theory, given that it adopts the hypothetical competitive market as the benchmark to assess conduct of the dominant firm. We have noted suggestions by, eg, Hanks and Williams,2 and Gavil3 that the hypothetical competitive market is not the appropriate benchmark. I propose that the conduct of the dominant firm should be assessed by reference to what is often called ‘rational profit-maximising’ conduct. I propose this not because it is rational or logical, but because it may well respond to market demand in a way that is both privately and socially efficient. We may recall from Chapter 1 that (as a result of the gradient of the demand curve typically assumed in analysis of monopoly outcomes) the monopolist’s profit-maximising output can thus be seen to be efficient, and increasing output beyond that point can be seen to be inefficient, because the revenue from increased output does not cover the additional costs (and sales revenue declines beyond the output at which marginal revenue becomes negative). Clearly, we cannot draw general conclusions from these typical assumptions. We need to make empirical enquiry as to whether there is unmet demand, and actual cases must be decided on the basis of evidence of the demand response in connection with the conduct in question. It may well be easier to obtain probative evidence when actual effects are viewed in hindsight than when cases are sought to be determined in advance of such experience, eg, when applying the predictive SLC test.

1 F Hanks and P Williams, ‘Implications of the Decision of the High Court in Queensland Wire’ (1990) 17 Melbourne University Law Review 437, 448. See also G Hay and R Smith, ‘“Why Can’t a Woman Be More Like a Man?” – American and Australian Approaches to Exclusionary Conduct’ (2007) 31 Melbourne University Law Review 1099, 1115–16. See also ACCC, Inquiry into the East Coast Gas Market, April 2016, 92, which accepts the proposition. 2 Hanks and Williams (n 1). 3 A Gavil, ‘Imagining a Counterfactual Section 36: Rebalancing New Zealand’s Competition Law Framework’ (2015) 46 Victoria University of Wellington Law Review 1043, 1048 and 1053.

134  Refusal to Deal and Margin Squeeze As we will see, applying the benchmark of the hypothetical competitive market, the High Court has held that a dominant firm’s refusal to deal, or setting a price that is tantamount to a refusal to deal, involves the dominant firm ‘taking advantage’ of its market power. The Court has avoided the question of deciding at what point the price imposed by the dominant firm crosses the line and becomes a misuse of market power. Perhaps it is unimportant, as Hanks and Williams suggest,4 and it can be left to the parties to work out once it is clear the law mandates supply. Pengilley and others argue otherwise.5 However, I would argue that it should be possible in principle to identify the borderline price. Since it is inextricably linked to the proposition that market power has been misused, if we cannot in principle identify the borderline price, this would cast doubt upon the conclusion that market power has been misused. If we accept that the dominant firm should be entitled to profit from success by deriving above-normal profits, we can accept that the firm should be allowed to decide on the price and output that maximises its profit.6 I suggest that we can justify this not because it is rational, but because it may be inefficient to expand output beyond this point. As we have seen, the dominant firm’s power is conventionally assessed by reference to its ability to charge prices above the ‘competitive level’, and this is taken to provide a measure of the harm that is addressed when we prohibit misuse of market power. In this Part, particularly in Chapters 7 and 8, we will test whether such a price can be determined in principle or forensically in practice. Even if we can determine the price above which there is a constructive refusal to supply, we must confront the implications of an ongoing supply relationship, namely where market risk lies. Ordinarily, we might expect that market participants fully bear the risk that the market price at which they sell their product may fluctuate, with a consequential impact on their profit margin. Thus, even if we accept that there should be an obligation to supply, we must consider whether this obligation should extend to relieving a party of this normal risk. We must also confront the question whether final judicial determination of the existence of market power at a point in time enables any useful prediction as to how long such power might endure.

4 Hanks and Williams (n 1) 460–61. 5 W Pengilley, ‘Queensland Wire and its Progeny Decisions: How Competent are the Courts to Determine Supply Prices and Trading Conditions?’ (1991) 21 Western Australian Law Review 225. See also Hay and Smith (n 1) 1108–10 regarding the US position; R French ‘The Role of the Courts in the Development of Australian Trade Practices Law’, in F Hanks and P Williams (eds), Trade Practices Act: A Twenty-Five Year Stocktake (Sydney, Federation Press, 2001) 108. 6 Hay and Smith (n 1) 1108–10. Contra: J Drexl ‘On the (A)political Character of the Economic Approach’ in J Drexl, W Kerber and R Podszun (eds), Competition Policy and the Economic Approach: Foundations and Limitations (Cheltenham, Edward Elgar, 2011) 323 argues that permitting a dominant firm to derive above-normal profits is inconsistent with the prohibition of ‘horizontal price cartels’. I suggest, however, on the reasoning set out in ch 3, that monopoly price and output may well be an efficient use of resources to satisfy market demand, whereas a cartel of two or more firms producing the same output would clearly involve an inefficient duplication of fixed capital and other resources.

Introduction  135 There is controversy between the US and EU positions on the question of ‘margin squeeze’, ie, whether the dominant firm that has an obligation to supply a competitor must do so at a price that would ensure the competitor a reasonable margin. In the US, Trinko and Linkline indicate that a dominant firm has no obligation to assist its rival, but if there is a regulatory duty to deal, the dominant firm has no obligation to supply at a price that ensures a reasonable margin for the rival.7 The US position indicates judicial reluctance to supervise commercial transactions in the market, whereas, as noted in Chapter 3, the EU position supports the regulatory role of the EC to do so. There has been controversy concerning the access regime in Part IIIA of the CCA that is relevant to the question whether competition law does or should regulate prices and margins. The Hilmer Report, which proposed the introduction in the 1990s of a regime to provide access to essential facilities, endorsed the proposition that competition law does not prohibit a monopolist deriving monopoly profits, ie, section 46 does not ensure access on reasonable terms.8 The focus of the Hilmer proposal was not to control excessive pricing, but to allow access to essential facilities to promote competition in other markets.9 There have been several controversial decisions under the access regime which grapple with the issue of perceived excessive pricing where there is no impact on competition in other markets, eg, where access is being provided, but at alleged excessive prices, and facility prices do not affect prices or competition in downstream markets.10 The Harper Review recommended the adoption of previous proposals by the Productivity Commission that the access regime in Part IIIA be amended to facilitate ‘access on reasonable terms’ where this would be expected to provide efficiency gains through a ‘substantial increase’ in competition in a dependent market that is nationally significant; however, the requirement that the increase be ‘substantial’ has not been accepted.11 The Review is focused on the ultimate 7 Verizon Communications Inc v Law Offices of Curtis V Trinko, LLP 540 US 398 (2004) (hereinafter ‘Verizon’); Pacific Bell Telephone Co v Linkline Communications, Inc 555 US 1 (2009). See also Hay and Smith (n 1); G Sidak, ‘Abolishing the Price Squeeze as a Theory of Antitrust Liability’ (2008) 4 J­ournal of Competition Law and Economics 279; C Rudaz, ‘Did Trinko Really Kill Antitrust Price Squeeze Claims? A Critical Approach to the Linkline Decision through a Comparison of EU and US Case Law’ (2010) 43 Vanderbilt Journal of Transnational Law 1077. The majority of the High Court in NT Power Generation Pty Ltd v Power and Water Authority (2004) 219 CLR 90, 134 express the view that ­Verizon does not assist in the interpretation of s 46 of the CCA. For the position in the EU, see, eg, Case C280/08P Deutsche Telekom AG v European Commission (2010) (CJEU, 14 October 2010) [140]–[185]; Case  C52/09 TeliaSonera (2011) [31]–[33], [75]. The analysis is similar to predatory pricing, in that if the dominant firm prices below retail distribution costs, it can be inferred that an equally efficient competitor could not compete. 8 Commonwealth of Australia, National Competition Policy (Australian Government Publishing Service, 1993) (hereinafter ‘Hilmer Report’) 243. 9 See, eg, ACCC, Inquiry into the East Coast Gas Market, April 2016, 101. 10 See, eg, Australian Competition Tribunal, Re Glencore Coal Pty Ltd and Port of Newcastle Pty Ltd [2016] ACompT 6; Sydney Airport Corporation v Australian Competition Tribunal (2006) 155 FCR 124; Port of Newcastle Operations Pty Ltd v Australian Competition Tribunal [2017] FCAFC 124. 11 Commonwealth of Australia, Competition Policy Review, March 2015 (hereinafter ‘Harper Review’) 73–74, recommending a higher bar than the Productivity Commission, which suggested that only a’

136  Refusal to Deal and Margin Squeeze goal of efficiency; yet, direct consideration of efficiency is not required by the declaration criteria in Part IIIA.12 While the High Court has ruled under section 46 that the dominant firm may be under a duty to supply, we have no judicial guidance on the issue of pricing and margin. We will test the assumption that section 46 in its current form ensures the competitor a reasonable, or any, margin in cases of refusal to supply. We will also consider how the Harper Review section 46 and the market manipulation approach might apply in this circumstance.

II.  Queensland Wire: Factual Matrix When Queensland Wire First Instance was decided by Pincus J in September 1987, roughly 13 years after the enactment of section 46, there had been few cases brought under the provision and no final relief had ever been granted.13 As we will see, the first instance decision dismissing the claim was reversed by the High Court on appeal and remitted to the trial judge to determine remedies. However, it does not appear that final relief was granted by the court, from which we may infer that the case was settled on confidential terms. Queensland Wire Industries Pty Ltd (QWI) was a fabricator which obtained supplies of steel rod from Broken Hill Pty Co Ltd (BHP), a vertically integrated iron ore miner and steel producer. QWI produced wire for rural fencing. BHP supplied a range of steel products to fabricators, but did not supply an extrusion called ‘Y-bar’ to anyone outside the BHP corporate group. BHP used this product to make and sell ‘star pickets’, a steel fencing post for rural or agricultural use. QWI claimed to be at a competitive disadvantage in selling fencing wire because BHP sold posts and wire in conjunction. It appears that QWI could have purchased star pickets from BHP at the same price as BHP’s distributors, but would not make a margin when it sold those posts and its own wire to those distributors.14 Thus, when Pincus J observed that QWI ‘cannot offer fence posts at a price which competes’ he meant that QWI could sell star pickets in conjunction with its own wire, but would not make any margin on the posts, and would incur a loss if its competitors reduced their prices and margins.15 It seems that, material increase’ in competition should be required (as currently provided in s 44H(4) of the CCA). See also Productivity Commission, National Access Regime, report of inquiry, 25 October 2013, 17. The Competition and Consumer Amendment (Competition Policy Review) Act 2017 does not adopt the Harper recommendation to require a ‘substantial’ increase in competition. The accompanying Explanatory Memorandum [12.21] states that ‘reasonable terms’ are not defined, and does not clarify whether this contemplates regulation of perceived excess profits. 12 However, the Explanatory Memorandum [12.4] to the Competition and Consumer Amendment (Competition Policy Review) Act 2017 states that the purpose of promoting competition is to enhance efficiency. 13 Queensland Wire Industries Pty Ltd v Broken Hill Pty Co Ltd (1987) 16 FCR 50, 52. 14 ibid 61. 15 ibid.

Queensland Wire: Factual Matrix  137 to address this s­ ituation, QWI wished to fabricate star pickets, and in the period from January 1984 to June 1987,16 QWI asked BHP to supply Y-bar; however, while BHP was prepared to supply it, the supply prices which BHP offered were such that QWI would not have been able to compete effectively.17 We do not know whether the prices offered by BHP allowed QWI any margin to fabricate and sell finished posts. Pincus J states that the price offered by BHP for Y-bar was ‘uncompetitive’ and ‘relatively to BHP’s other rolled products, is excessively high’.18 QWI sought damages and an injunction requiring BHP to supply it with Y-bar. Pincus J noted the difficulties with such remedies, which would require an assessment of the appropriate supply price, and expressed an inclination towards granting a declaration (if a remedy were required).19 In any event, the case was dismissed at first instance on the ground that a vital element was not proven, as Pincus J held that BHP had not ‘taken advantage’ of its market power. On appeal to the Full Court of the Federal Court, the appeal was dismissed on the ground that there was no market for Y-bar as BHP had never sold it outside the BHP corporate group.20 On appeal to the High Court, it was held in Queensland Wire that BHP had taken advantage of its market power since it was only because of that power that it could afford to refuse supply, and (per Deane, Dawson and Toohey JJ) that there could be a market even though no sales had taken place (or at least on the basis BHP had offered to supply to QWI, albeit at prohibitive prices). The events occurred, and the case was tried, under the version of section 46 that applied prior to its amendment in 1986, ie, a corporation ‘in a position substantially to control a market’ must not take advantage of that power for a proscribed purpose. The 1986 amended (pre-Harper Review) form of the provision prohibits a corporation ‘that has a substantial degree of power in a market’ from taking advantage of that power for a proscribed purpose. Nothing turned on that change, as the trial judge held that both would be satisfied.21 The definition of relevant markets was disputed and Pincus J observed that ‘not as much information as I would have liked was available’.22 BHP produced about 97 per cent of steel made in Australia, and supplied about 85 per cent of the domestic consumption of steel (the only other domestic producer supplied three  per cent of domestic requirements).23 Accordingly, we can infer that

16 ibid 52. 17 ibid 54. 18 ibid 61. 19 ibid 54. The limitations of declaratory orders, which pronounce upon a legal state of affairs in a way that is binding in further proceedings and are not in themselves enforceable, are usefully discussed by R French, ‘Declarations – Homer Simpson’s Remedy – Is There Anything They Cannot Do?’ (2007) Federal Judicial Scholarship 24. 20 Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd (1987) 17 FCR 211. 21 Queensland Wire First Instance (1987) 16 FCR 50, 52 and 56. 22 ibid 54. 23 ibid 54 and 67.

138  Refusal to Deal and Margin Squeeze 12  per  cent of domestic consumption was imported. Some star pickets were imported, but amounted to only one per cent of the market.24 In 1983, the Australian government introduced a Steel Industry Plan to provide subsidies for the domestic production of certain steel products and to establish specialised antidumping procedures to ensure that domestic producers’ share of the Australian market did not fall below 80 per cent.25 No evidence was provided regarding the subsidy at the end of the period in issue; however, it appeared to have been falling, as the devaluation of the Australian dollar in 1983 made imports more expensive.26 BHP had excess capacity and ‘could undoubtedly make sufficient Y-bar to satisfy all requirements’.27 The above description understates the situation of the Australian steel industry fighting for survival in the 1980s. Clearly, imports were expected to increase above the 12 per cent figure mentioned in the judgment to 20 per cent or more; however, the underlying market forces are not addressed in the judgment (nor was the fact that BHP was making losses on its steel division). Peter Demura vividly conveys the economic circumstances of the times: [I]n the early 1980s … a confluence of forces threatened the existence of one or more of BHP’s steelworks. During the early 1980s the world recession, falling demand and an increase of excess world steel capacity, estimated at 30 times Australian domestic consumption, resulted in a substantial increase in imports of cheap steel into the Australian domestic market. This occurred at the end of the resources boom, and BHP Steel found itself with excess capacity, both plant and labour, and a declining market share. BHP’s immediate response was a downsizing of operations. Between 1982 and early 1983 plant capacity was reduced by one third to 6.0 million tonnes and the workforce declined by almost 10,000 either through voluntary retirement or retrenchment.28 The crisis in the steel industry focused BHP’s attention on the need to increase productivity and modernise its steel-making operations. After a succession of industry inquiries and a change of government, BHP together with unions and Federal and State governments agreed to the Steel Industry Plan. Under the Plan, BHP gave an undertaking to maintain its steel-making capacity, invest $800 million over five years and refrain from further compulsory retrenchments. Governments, on the other hand, promised up to $71.2 million in bounties (not required), and committed to reduce indirect taxes, speedup anti-dumping provisions and review assistance measures if required. The unions agreed to productivity targets, to follow dispute settlement procedures and refrain from seeking increases in wages in excess of community standards.29

24 ibid 56. 25 ibid 55. 26 ibid 55 and 60. 27 ibid 67. 28 P Demura, ‘Productivity Change in the Australian Steel Industry: BHP Steel 1982–1995’ in Productivity and Growth (Canberra, Reserve Bank of Australia, 1995) 178 (references omitted), available at www.rba.gov.au/publications/confs/1995/pdf/demura.pdf. 29 ibid (references omitted).

Queensland Wire: Factual Matrix  139 By 1984, productivity at the steelworks had reached 243 tonnes per year and the net operating loss was substantially reduced. There is some debate about the role of the Steel Industry Plan in turning BHP Steel around. For example, the Bureau of Industry Economics argued that the turnaround had more to do with currency depreciation, recovery from the 1982/83 recession, cost cutting and retrenchments implemented prior to the Plan, and an improved industrial relations climate under the Accord. However, what the Plan did do was allow BHP Steel to restructure its operations, increase training and development and give it confidence to undertake substantial capital investments with the aim of improving business performance. Rather than investing $800 million, it spent almost $2 billion. This represented the ‘… largest capital expenditure program, on a per tonne of capacity basis, of any developed steel-making country – surpassing Japan, doubling EEC spending, and quadrupling that in the US’.30

According to a report by the Anti-Dumping Authority in 2016, steel imports have steadily increased since the late 1980s, and domestic production has diminished, such that imports in 2015 were just over 50 per cent of domestic consumption.31 BHP divested its steel-making businesses in around 2000.32 It seems clear that there were dynamic market forces at play in the 1980s that a static structural analysis of market shares would not address. We can also see that the position of the Australian steel industry in the 1980s and since provides an example of zero-sum, or survival, competition. The trial judge criticised the evidence presented by BHP on import competition, describing it as ‘scrappy and incomplete’, though he acknowledged that BHP was ‘anxious’ about the possibility of import competition and was prepared to reduce prices by 2–15 per cent to counter import competition.33 However, in the end, the trial judge concluded: Although the evidence … shows that the possibility of import competition is a constraint as far as BHP is concerned, the hard fact is that there has never been any substantial importation of star pickets.34

Similarly, with regard to potential competition from the other Australian steel producer (Smorgon), the trial judge acknowledged BHP’s belief that Smorgon would manufacture Y-bar and supply either Y-bar or star pickets, but concluded: ‘That has not happened yet, and there is no suggestion on the evidence that it is about to.’35 These findings demonstrate the difficulty of deciding future matters based on past experience, ie, implicitly finding that the past is likely to continue despite the apprehensions of industry and government.

30 ibid 178 (references omitted). 31 Anti-Dumping Commission, Analysis of Steel and Aluminium Markets, Report to the Commissioner of the Anti-Dumping Commission (31 August 2016) 22–23. 32 ibid. 33 Queensland Wire First Instance (n 21) 60. 34 ibid. 35 ibid 58.

140  Refusal to Deal and Margin Squeeze

III.  Gaps in the Factual Matrix We can now address gaps in the factual matrix. The first is that we do not have details of BHP’s purpose or reasoning in refusing to supply QWI except at ­prohibitive prices, since BHP chose not to adduce any evidence on the issue.36 This made it relatively easy for the court to infer, in accordance with section 4F, that at least one substantial purpose of BHP was to prevent QWI entering the market for star pickets and possibly to deter or prevent QWI from competing in the market for steel-fencing products.37 This finding was not articulated with specificity because the trial judge held that BHP had not ‘taken advantage’ of its market power. We may wonder why BHP did not proffer evidence as to its reasons. We should recall that the courts had not at that time fully developed the interpretation of ‘purpose’ to mean ‘subjective purpose’.38 In our thought experiment below, we may consider how a different approach might affect the outcome under the Harper Review section 46. The second gap is that since QWI was a fabricator of wire, not star ­pickets, there was no evidence regarding its production costs, ie, its efficiency as a potential producer of star pickets. We might wonder whether in fact QWI really wanted supply of Y-bar or star pickets. At the time, although there was no clear authority in Australia, it must be doubted that a dominant firm had a duty to supply finished product to a competitor. This may still be the case, but we must consider whether the Harper Review section 46 may change the position. We may speculate that QWI might have really wanted supply of finished star pickets, but may well have considered its case would have been weaker if it simply sought supply of the finished product. From common knowledge, it is evident that the value added by processing Y-bar is minimal, ie, it is cut to length, holes are made for fencing wire to pass through, a point is made at one end to facilitate driving into the ground and generally some form of bituminous rust-proofing is applied. Supply of finished star pickets would have made more sense, which must have occurred to Pincus J since he felt it necessary to conclude that QWI was ‘genuine’ in its desire to fabricate star pickets from Y-bar.39 Nor was there any satisfactory evidence regarding QWI’s efficiency as a wire producer competing with BHP. However, it was noted that QWI ‘competed fairly effectively with BHP in relation to wire sales’ and had around 28 per cent of the rural fencing market in Queensland.40 That is apparently the case despite BHP having ‘great advantages’ in being able to offer star pickets and fencing wire, and achieving cost benefits regarding transport of products.41

36 ibid

51. 52. 38 See ASX Operations Pty Ltd v Pont Data Australia Pty Ltd (1990) 97 ALR 513. 39 Queensland Wire First Instance (n 21) 61. 40 ibid 55. 41 ibid 57: ‘No doubt it would in reality be satisfied if it were sold fence posts cheaply enough.’ 37 ibid

Gaps in the Factual Matrix  141 The third gap concerns the inferences drawn by the trial judge regarding BHP’s profitability and prices. This may be accounted for by confidentiality, which prevented the publication of details in the court’s reasons, or it may be attributed to the court’s process of reasoning being less clearly developed as to what might be a permissible use of market power. Thus, we do not have facts from which to assess whether BHP’s output, prices and profit were consistent with efficiency. The key factors here are not whether BHP was deriving a monopoly profit (since it is generally thought that provisions such as section 46 do not condemn that), but whether there was unmet demand, ie, whether output could have been increased with marginal revenue exceeding marginal costs, or whether marginal revenue would have been negative, and hence increasing output may well have been both privately and socially inefficient. Pincus J accepted that Y-bar is a low margin product of value to BHP for its connection with wire sales; however, he found that BHP would be prepared to drop its prices substantially to meet import competition, apparently inferring that BHP had derived substantial excess ­profits.42 There is nevertheless some recognition in the first instance judgment that section 46 requires a distinction between permitted ‘use’ of market power and ‘misuse’. Regrettably, Pincus J expressed the latter to involve some element that is ‘reprehensible’ or ‘unfair’.43 This was overruled by the High Court, which applied an abstract principle, ie, we ask whether the dominant firm could afford to refuse the sale if it lacked market power44 rather than asking whether it would have been inefficient for the firm to supply. As we have seen in Chapter 1, in a monopoly, the market excludes equally efficient entrants if, having regard to the size of the market, it is not efficient to expand output when each additional unit is produced at an increasing loss (especially when revenue is declining). I argue that it is for that reason – ie, that increasing output is not only privately inefficient, but that drawing resources from other uses beneficial to society may well be socially inefficient, not merely that profit-maximising behaviour is rational – that the monopolist’s above-normal profit can be thought to be justified in legal theory. The fourth gap concerns the basis for reasoning that the lack of new entry was caused by BHP’s use of its market power. Pincus J concluded that: ‘In the end it is not really very clear by what means BHP has managed to preserve its monopoly for such a long period of time.’45 The judgment notes various attempts by other firms to import star pickets, which failed either because imported prices were not competitive or because BHP engaged in price competition.46 That could equally

42 ibid 58. 43 ibid 68. 44 Queensland Wire (1989) 167 CLR 177, 192, per Mason CJ and Wilson J; 202, per Dawson J; 216, per Toohey J. 45 Queensland Wire First Instance (n 21) 58. 46 ibid 60.

142  Refusal to Deal and Margin Squeeze be explained on the basis that new entry was not efficient or on the basis that BHP created strategic barriers to entry. Pincus J noted that BHP had entered into arrangements with major customers requiring some to obtain all their requirements from BHP, sometimes in return for a two per cent rebate, which suggested that BHP’s market power in respect of rural fencing is considerable, but not such as to enable it to dictate to its customers.47 This could equally suggest that BHP’s power was to some extent constrained by its customers, which tended to be large pastoral or trading houses. Despite the confidence of Hanks and Williams that the heart of section 46 cases after Queensland Wire would concern efficiency,48 that is not required by the judgments in Queensland Wire. In fact, Pincus J at first instance, while apparently reaching towards some concept of permitted use, rejected efficiency as a basis for concluding that a dominant firm could be said not to be taking advantage of its market power.49 In the High Court, Deane J said that market power could be ‘taken advantage of ’ within the meaning of section 46 even if it was acquired by lawful means such as efficiency. This observation clouds the issue as to whether conduct that is ‘efficient’ would not amount to ‘taking advantage’. The other judgments in the High Court, by referring to a hypothetical competitive benchmark, do not consider the potential efficiency of a monopolist’s profit-maximising output. It is suggested that the Harper Review recommended section 46 brings efficiency squarely into focus by mandating the court to take efficiency effects into account when applying the SLC test, which we can now consider. While efficiency is arguably relevant to the old (or pre-Harper) purpose test in section 46, it has not generally been considered relevant in Australia to the SLC effects test in, eg, sections 45 and 47 of the CCA (though it would be relevant in the US).50 The Parliament’s rejection of the Harper mandate to consider efficiency may well have a significant substantive effect. Thus, when we consider the Harper section 46, we will consider the standard SLC test and will also consider how overt ­consideration of efficiency might alter the outcome of the test.

IV.  Queensland Wire under Harper Section 46 The Harper Review did not address the concept of market power, so it would appear that traditional concepts from 1950s US thinking still apply.51 As we have seen, the assessment of the constraints upon the dominant firm’s conduct provided by customers and competitors (required by section 46(3)) involves



47 ibid

59–60. and Williams (n 1) 445–46. 49 Queensland Wire First Instance (n 21) 62. 50 Hay and Smith (n 1) 1123. 51 Queensland Wire (n 44) 200, per Dawson J. 48 Hanks

Queensland Wire under Harper Section 46  143 ­ uestions of degree and inference. It is possible that better evidence of those q matters might result in a different assessment. While we might infer from the abandonment of the ‘taking advantage’ element that the hypothetical competitive market should no longer be the relevant benchmark, it is unclear whether Parliament’s qualified adoption of the Harper Review section 46 manifests a legislative intention to alter the benchmark. So, for the present purposes, it may be safe to assume that under the Harper section 46, BHP would be held to have substantial market power. What is clear, however, is that the purpose test limb of SLC requires evidence of substantial subjective intent. It does not appear that there is any legislative intent, by adopting the SLC test, to change the generally accepted principle that a monopolist is not to be deprived of the fruits of success (in this case BHP’s risky investment in steel-making and fabrication capacity). I propose that we should judge BHP’s conduct not by reference to the hypothetical competitive market, but by reference to its circumstances, ie, a virtual monopolist (albeit inefficient by global standards) making substantial capital investment to improve productivity and survive in a changing market, seeking to turn around losses into profits to generate a return on that investment. As we will see in the subsequent chapters of this Part, courts tend to resolve the ‘dual purposes’ problem against the dominant firm. That is, were it BHP’s intent to maximise sales by taking sales away from QWI, it is likely that BHP would be held to have a substantial subjective purpose of SLC. While harm to competitors is distinguished from harm to the ‘process of competition’, harm to a new entrant or potential new entrant can be taken to represent harm to competition. It is also clear that in applying the SLC effects test, the Harper Review wishes the court to consider efficiency. Arguably this requires consideration of market size, ie, has there been no new entry because that would not be efficient or is there unmet demand that could be supplied profitably? The Harper Review did not elucidate what it meant by ‘efficiency’ or how it should be forensically assessed. We might speculate that the Review intended this to be assessed by reference to the presumed ‘competitive level’. However, for several reasons canvassed in Part I of this monograph, I suggest that such an approach would not assist. First, the ‘competitive level’ is not forensically determinate. Second, it is based on unrealistic assumptions which should be empirically tested in order to enable meaningful conclusions to be drawn. Third, by definition, a dominant firm would not be expected to increase output beyond its profit-maximising level, so it follows that a dominant firm would always produce less than the presumed ‘competitive level’, ie, there would always be ‘unmet demand’ if this benchmark were used. Accordingly, I propose that ‘unmet demand’ be defined by reference to the dominant firm’s profit-maximising output. It is thus neither logical nor efficient that unmet demand exists (ie, additional output that could be supplied profitably, even at reduced prices, and without marginal revenue becoming negative so that the demand response would reduce revenue). That is, we expect the dominant firm to satisfy demand by increasing output as far as is possible to maximise profits.

144  Refusal to Deal and Margin Squeeze We  would not expect there to be unmet demand unless the firm is capacityconstrained and the investment required to expand capacity is not justified by current demand or projected demand growth. Unmet demand could be indicated, eg, by delays in ordering and delivery, low stock levels having regard to production lead times, or rationing of supplies. Thus, we would need some explanation if it were suggested that BHP was withholding supply to the market, eg, was it capacity-constrained, was it satisfying higher value uses of its capacity or was it manipulating the market for some non-obvious purpose? It does not appear that QWI proposes to reach new customers, merely to be better placed to take sales away from BHP that BHP has the capacity to make itself and presumably does make itself. This is classic zero-sum competition. It is generally accepted that the process of competition is not harmed unless the excluded party, QWI, is a more efficient competitor, but there is no data to assess this, either with respect to fencing wire or star pickets. Suppose QWI were able to produce star pickets at lower cost than BHP – would the additional investment of capital to produce output that BHP could already produce be socially efficient? Would it be necessary to show that efficiencies would benefit consumers through lower prices, as opposed to merely transferring surplus from BHP to QWI? While it would be speculative to pursue these issues further here, there is no reason to believe that these forensic enquiries could not be made, and we may turn to the question of remedy having regard to pragmatic considerations. We recall that SLC is a predictive test. Typically, the ACCC assumes that the past will continue.52 Experience suggests that this is unlikely ever to be the case in reality. Courts apply the test without the benefit of hindsight,53 so in Queensland Wire, the court would not generally receive evidence about developments in the market after the period of the events in question, which was between January 1984 and June 1987. In this case, the appeal proceeded quickly and was heard by the High Court in June 1988, with judgment given in February 1989. Nevertheless, there may well have been ample evidence to suggest that the steel industry and the Government of the day had good grounds to consider that import competition was likely to increase. This is a subtly different issue from that considered by the trial judge – he was concerned only with assessing constraints on BHP’s power as at mid-1987 and did not consider the possibility of import competition to be sufficiently proximate at the time. Assuming the trial judge correctly assessed BHP’s market power in mid-1987, we must ask at what point would market circumstances have changed sufficiently to justify a finding that BHP no longer had market power as traditionally understood? If the harm which flows from BHP’s refusal to supply is that QWI cannot compete effectively with respect to fencing wire, this could be solved by supplying

52 ACCC ‘Merger Guidelines’, November 2008, [3.17]–[3.18] in the context of the s 50 SLC test (mergers). 53 Universal Music (2003) 201 ALR 636 [247].

Queensland Wire under Harper Section 46  145 star pickets at no margin, ie, QWI need not profit from star pickets if its aim is to achieve transaction cost efficiencies (and economies of scope) by supplying posts and wire. Thus, if efficiency were our sole concern, it might have been possible to conclude that QWI should have simply accepted BHP’s offer to supply fence posts at the same price as to other distributors. If QWI seeks to take sales of star pickets away from BHP (by obtaining unfinished Y-bar from BHP and fabricating the final product with apparently minimal value-adding), that would be pure zero-sum competition that would seem to be economically inefficient (because increased capital and other resources would be applied to produce an output that could have been produced without any additional resources being diverted from other productive uses in the economy). An efficient outcome would require that QWI’s entry to the market would result in increased output where marginal revenue exceeds marginal costs. That depends critically on unmet demand and a demand response involving positive marginal revenue. There is no reason to believe that supply by QWI would alter the demand response, ie, the result would be ‘mere redistribution of, rather than a net increase in, sales’.54 We must ask whether the case would have been different had QWI simply sought supply of finished star pickets. Around 10 years after Queensland Wire, when Merkel J at first instance decided Melway First Instance,55 it was thought that the same outcome should apply to a refusal to supply finished products. However, as we will see, while the High Court did not reconsider Queensland Wire, deciding Melway on its facts, it can be argued that the underlying factual enquiry about market response and economic efficiency is the same. Returning to the practicability of determining a remedy, let us assume that BHP’s refusal would be held under the Harper Review section 46 to have the purpose or likely effect of SLC in either or both of the markets for star pickets and fencing wire. If we assume that QWI will be equally efficient as BHP, the price should be BHP’s internal transfer price for Y-bar, calculated as a ‘net back’ price based on the selling price of star pickets, allowing for the costs of fabricating star pickets and normal margin as derived by BHP. That is, BHP would be entitled to above-normal profits on the intermediate product (Y-bar) and QWI should have the same opportunity as BHP to profit in the downstream market, assuming they are equally efficient. As in all supply arrangements, the terms of dealing will be critical to whether QWI continues to receive a reasonable, or any, margin. It may be unrealistic to consider that the court can supervise ongoing terms of dealing, and it must be questioned whether Part IV of the CCA even allows competitors to reach any understanding that would or might ensure QWI such a margin. For example, will QWI be supplied at spot prices or a fixed price? If it is a fixed price, how long will it endure and will it be indexed for cost increases? If competition

54 Hay and Smith (n 1) 1122. 55 Robert Hicks Pty Ltd v Melway Publishing Pty Ltd (1998) 42 IPR 627 (hereinafter ‘Melway First Instance’).

146  Refusal to Deal and Margin Squeeze in the market, eg, from imports, should depress prices for star pickets, would or could QWI be given price support to protect its margin? Pincus J correctly recognised that there would be some difficulties regulating the terms of supply by court order. Further, where a supplier and customer are competitors, they risk contravening the price-fixing (or cartel conduct) provisions.56 In principle, where the supplier competes with its customer in the downstream market, any agreement to protect the customer’s margin could be seen to unlawfully control or maintain prices. These difficulties suggest, as Pincus J appears to be well aware, that the pragmatic solution would be for QWI to obtain supplies of star pickets at the same prices as other wholesalers, ie, when QWI sells at wholesale, it would obtain no margin, but would enjoy the same transaction cost advantages (and economies of scope) as BHP in selling fencing posts and wire. For QWI to obtain court-enforced supply at ongoing prices protecting its profit margin (or ensuring its margin would be comparable to BHP) would result in QWI receiving protection from competitive pressures on its margin that would not be available to BHP. Accordingly, applying the Harper Review section 46 with the emphasis on efficiency that the Review recommends, we can see the need for forensic enquiries which would be required to determine whether QWI would be a more efficient producer and whether output could be expanded profitably to satisfy unmet demand. In a case where there appears to be no product or process innovation or benefit to consumers, but simply a transfer of sales and profits from BHP to QWI, this would appear to offer no efficiency benefits to society. In fact, it would appear to demonstrate how competitors may seek to use competition law as a substitute for actual competition. Under the Harper Review section 46 as enacted, absent any explicit reference to efficiency, we may doubt the enquiry would proceed in the same way. The focus of the SLC test on the ‘process of competition’ informed by traditional structural analysis may well overlook enquiry into the issues addressed above. That is, the presumption that more competitors should always be good for the process of competition may cause us to overlook the risk that a court-enforced duty to deal may simply redistribute surplus between firms without any innovation or efficiency benefits to consumers in the relevant market or to society.

V.  Queensland Wire under Market Manipulation In applying this approach, we bypass structuralist conceptions of market power and SLC, and go straight to considerations of economic efficiency along the lines discussed above. Thus, we avoid two aspects of the traditional concept of



56 See,

eg, ACCC v Flight Centre Travel Group Ltd [2016] HCA 49.

Queensland Wire under Market Manipulation  147 market power that the Harper Review did not address. First, if BHP’s market power depends substantially upon the constraint provided by import competition, we may need to constantly monitor, eg, exchange rates and other factors affecting imports, as market forces seem to be in a constant state of flux, and there were obvious trends suggesting continual erosion of any market power. Second, the highly qualitative nature of the traditional concept of market power, as the absence of constraint, leads to the strange provision in the current and proposed section  46 which provides that two or more firms may have substantial market power, ie, s­ eemingly, each may have the power to act substantially unconstrained by the others.57 This seems implicitly to acknowledge that the traditional concept of market power is inadequate and points towards the concept of ‘power to manipulate the market’ which is advocated in this monograph. As we will see, the concept of ‘constraint’ provided by competitors is of little or no assistance when we consider in Chapter 5 price-cutting as a possible exercise of market power. Under the market manipulation approach, it may be easier to conclude that BHP has the power to manipulate the market. However, potentially QWI also has that power, eg, by expanding output and thus depressing prices, and the key issue will be whether BHP’s conduct can be justified on efficiency grounds. The question is whether it would be inefficient to expand market output (ie, output could not be profitably increased because marginal revenue would not cover marginal costs or to do so would actually reduce revenue) and whether BHP’s refusal to supply QWI is efficient in this sense. These matters may be proved by reference to the sales, price and output data of BHP and QWI. If BHP was satisfying the market and QWI could not claim to be a more efficient producer, BHP’s refusal to supply could well be efficient in the sense indicated above. We can see that whether new entry and output expansion is efficient depends crucially on the size of the market and the demand response. We can say that conduct which is efficient is not manipulative. On the other hand, conduct which is inefficient may be manipulative, but requires further examination, ie, has the conduct interfered with the undistorted operation of the market? We would want to ascertain by forensic enquiry whether supply was responding to shifts in demand (which we might expect to be cyclical in the case of durable steel fencing products).58 It will be seen that we face a number of difficulties that the factual matrix from Queensland Wire does not assist us with. First, we may be called upon to second-guess business decisions as they are being made, before data emerges as to the outcome of the decisions. This is akin to the problem of assessing ‘future representations’ under section 51A of the CCA, now found in section 4 of the

57 Old s 46(3D) of the CCA; and Harper s 46(7)(c). 58 M Motta, Competition Policy: Theory and Practice (Cambridge, Cambridge University Press, 2004) 71–73 argues that the solution of the ‘durable goods monopolist’ problem proposed by R Coase, ‘Durability and Monopoly’ (1972) 15 Journal of Law and Economics 143 is unlikely to apply in practice if users do not have the discretion to delay commitment, which would appear to be the case with rural fencing in Australia for livestock management.

148  Refusal to Deal and Margin Squeeze Australian Consumer Law. That is, the maker of a representation as to a future matter must show reasonable grounds to make it. Thus, we might require the firm to demonstrate that it has reasonable grounds to decide as it did and so to embark upon the course of conduct. Should it be open to the court to conclude otherwise? Like the ‘business judgement’ rule in corporations law, we might suggest that the court should not conclude otherwise unless no reasonable person could have so decided. Second, we may be forced to consider decisions of the firm that would appear not to be dominant, eg, the new entrant which wishes to increase output. Can this decision be justified on efficiency grounds or is it manipulative, ie, seeking a demand response that will cause another firm to exit the market? Thus, we may have two firms locked in zero-sum competition. We presumably have no problem if a less efficient firm is forced out, but this may never be known, and we may consider it manipulative to force out an equally efficient firm by this means. I suggest that the preferred policy response is to hold that neither should be permitted to manipulate the market. Should we allow a firm to defend itself if its competitor does manipulate the market, ie, allow a defence of ‘meeting competition’? Should the law intervene when two firms are locked in zero-sum competition, eg, by assessing which firm is more efficient and declaring that firm to be the ‘winner’? I suggest that unless we are able to discern a conclusive answer on the question of efficiency, the law should not get involved in the competitive process. Third, to shift our focus from social efficiency to efficiency within a market (ie, industry) raises new and different issues. We focus on market efficiency because  we cannot assess efficiency gains and losses at the level of the whole economy, which would require assessing gains and losses in other markets. However, we can ask if the market is operating efficiently, ie, whether supply is responding to demand, though we know that in an oligopoly, supply responds to marginal revenue rather than price, so the mechanism does not work exactly as it does in perfect competition. Can a market be operating undistorted if a more cost-efficient competitor is denied entry (or excluded from the market)? It may well be the case, eg, if a potential new entrant has a process that enables it to produce the same goods more cost-effectively, that the firm will not attract investment capital because success depends upon taking sales away from competitors and is perceived by c­ apital markets to be too risky. Experience suggests that more efficient processes that offer only incremental benefits will be less likely to be adopted than those involving paradigm shifts. We might consider that private investment decisions may be regarded as efficient if the social benefits from zerosum competition do not exceed the costs. Acknowledging that efficiency at the level of the whole economy may thus require a wider Hicks-Kaldor assessment that  is beyond the practicalities of forensic determination, we could nevertheless examine efficiency at the level of the market by reviewing firms’ internal cost structures to determine which is more efficient (and thus which firm we expect to be able to enter the market and survive under zero-sum competition). I suggest that this would require a broader focus than marginal cost alone (for which we

Melway: Factual Matrix  149 have only the surrogate of average variable cost, or avoidable cost). I suggest that we would need to consider average cost and return on assets (as measures of efficient use of resources including fixed capital) and return on equity (as methods of financing assets through debt/equity leveraging affect the cost of capital, business risk and ability to withstand market shocks). Fourth, we may need to consider whether (if a predictive test is too uncertain and the outcome of business decisions in any event are short-lived in a changing world) it would be better to allow the consequences of business decisions to play out in the market. This would permit probative data concerning actual consequences to become available, and remedies could be provided to strip ill-gotten gains. The finding by Pincus J that BHP had made excessive profits over a long period to the detriment of consumers is, as the judge acknowledged, unexplained.59 We might expect that such an observation could not be made in the absence of probative evidence (which might possibly have been put before the court on a confidential basis). The market manipulation approach would not regard above-normal profits as an indicator of harm – in fact, only a departure from profit-maximising output (which suggests lower profitability) would indicate manipulative conduct which requires an explanation. Possible explanations for BHP’s excess profits might be: BHP has profited from industry protection policies that have kept out cheaper imports in order to protect Australian jobs; BHP has legitimately profited from its investment and had been protected by Australia’s isolation from import competition; or BHP’s profits did not result from business success, but from monopolistic practices. Without further data, we can only speculate as to which explanation might apply. We thus cannot discern whether the conclusion apparently reached by Pincus J might represent judicial rejection of the proposition that a monopolist may profit from success. We can now test the Harper Review section 46 and market manipulation approaches against the next landmark case concerning refusal to supply.

VI.  Melway: Factual Matrix The Melway street directory covering the greater Melbourne metropolitan area was first published in 1966. The first Melway and other directories published at the time were drawn and coloured by hand; however, the second and subsequent Melway editions used phototypesetting technology, which produced a superior, more readable product.60 The Melway directory was produced in this manner until 2000, when the maps were produced for the first time using computers,61 and apparently became available over the internet.62

59 Queensland 60 Melway 61 ibid. 62 See

Wire First Instance (n 21) 58. History, www.melway.com.au/melway-history.

www.street-directory.com.au/sd_new/home.cgi.

150  Refusal to Deal and Margin Squeeze In the same year, the US military made the ‘global positioning system’ (GPS) fully available for civilian use (this having been foreshadowed in 1998), and companies began marketing consumer GPS devices.63 Google Maps was launched in 2005 and was made available on the iPhone in 2007 (until Apple introduced Apple Maps in 2012, which was then made available on the iPhone).64 In 2012, Melway made its maps available through an iPhone app.65 The Melway section 46 case was heard during 1998 concerning events that occurred in 1996. The judgment of Merkel J records that the Melway street directory was at that time considered to be a superior product to that of its competitors.66 The reasoning of Merkel J with regard to market power is somewhat circular, in that the absence of competitive constraints leads to an inference of insurmountable barriers to entry.67 Barriers to entry are identified as primarily copyright law and the expense of preparing a competing directory.68 Merkel J concluded that: ‘Absent some form of new technology, it is neither rational or likely that a new entrant would enter [the] market.’69 This may have been a veiled reference to GPS technology, which, as noted above, was at the time expected to become available imminently for civilian use. By the time of the High Court judgment in Melway,70 this technological revolution had begun to transform the market. Merkel J at first instance held that Melway acted for an exclusionary purpose in terminating Hicks as a distributor of Melway street directories and had taken advantage of its market power, and thus contravened section 46.71 The Full Federal Court on appeal in Melway FFC72 upheld the decision by majority (Heerey J dissenting). The High Court reversed this decision, holding that while Merkel J’s finding of exclusionary purpose should not be disturbed, Melway’s conduct did not rely on market power since there was no reason to believe that it would not maintain its distribution system, even if it lacked market power.73 However, it seems that under the Harper Review section 46, vestiges of this issue remain, as the Review apparently considers that market power may contribute to anticompetitive effects.

63 See, eg, M Sullivan, ‘A Brief History of GPS’ PC World (9 August 2012), www.pcworld.com/ article/2000276/a-brief-history-of-gps.html. 64 See, eg, S Gibbs, ‘Google Maps: A Decade of Transforming the Mapping Landscape’ The Guardian (8 February 2015), www.theguardian.com/technology/2015/feb/08/google-maps-10anniversary-iphone-android-street-view. 65 See, eg, A Stephens, ‘The Melways is Turning 50. But Will Melbourne’s Back-Seat Bible Survive the GPS?’ Sydney Morning Herald (22 April 2016), www.smh.com.au/entertainment/the-melways-isturning-50-but-will-melbournes-backseat-bible-survive-the-gps-20160413-go5cvn.html. 66 Melway First Instance (n 55) 630. 67 ibid 639. 68 ibid 637–38. 69 ibid 639. 70 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1 (hereinafter ‘Melway’). 71 Melway First Instance (n 55). 72 Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (1999) 169 ALR 554 (hereinafter ‘Melway FFC’). 73 Melway (n 70).

Melway under Harper Section 46  151 The facts of the case place it in the category of a ‘terminated distributor’, ie, a supplier (Melway) terminates a distributor (Hicks) and replaces it with another, often with respect to an exclusive distribution territory or channel (market segment). This conduct falls within section 47 and would be prohibited only if the conduct (having regard to similar instances of exclusive dealing which can be aggregated) has the purpose or would have or be likely to have the effect of SLC in a relevant market. Courts have held that such conduct may harm the terminated distributor, but does not affect the process of competition.74 However, in Melway, we see the terminated distributor making a strategic decision to use the purpose test in section 46 rather than the SLC effects test in section 47 (which could have been used to impugn the Melway exclusive distribution system as a whole). This strategy is of interest because it may indicate a lack of confidence in the SLC test where market power may be implicated, and raises the question as to whether market power in some way enhances anticompetitive effects. As was noted in ­Chapter  3, the Full Federal Court in Universal Music rejected this view, on the basis of textual interpretation of section 47 (which makes no reference to market power).75 I suggest that the prevailing structuralist understanding of the SLC test already has within it an apprehension of the anticompetitive consequences of a highly concentrated market structure and exclusive dealing arrangements. Thus, I suggest that the Full Federal Court in Universal Music is correct in holding that ‘market power’ does not add a factor to the SLC test that is not already taken into account by that test. It is for this reason that I suggest we can conclude that the position accepted by the Harper Review (that conduct which may be innocuous if carried out by a firm without market power may have the potential to SLC if carried out by a firm with market power) is not a principle of economics or a presumption to be read into the SLC test. However, I argue that ‘substantial market power’ should be more than a discriminatory identifier like, eg, singling out persons with red hair or blue eyes for special attention. I suggest that the Harper test should still require a causal nexus to be demonstrated between market power and the apprehended SLC.

VII.  Melway under Harper Section 46 On the face of it, the SLC test in the Harper Review section 46 should not lead to any different result to the section 47 SLC argument that the plaintiff in Melway could have raised, but apparently chose not to run. However, Melway is of interest with regard to the dissenting opinion of Heerey J in the Full Federal Court, which

74 Dandy Power Equipment Pty Ltd v Mercury Marine Pty Ltd [1982] FCA 178, (1982) 64 FLR 238; Outboard Marine Pty Ltd v Hecar Investments (No 6) Pty Ltd [1982] FCA 265, (1982) 66 FLR 120. 75 Universal Music (n 53) [268].

152  Refusal to Deal and Margin Squeeze suggests that market power is not implicated in rational business decisions.76 I suggest that this proposition should be reframed to recognise that business decisions may be justified on efficiency grounds. While Heerey J refers to the article by Hanks and Williams concerning economic efficiency,77 it is clear that he uses the term ‘efficiency’ in a number of different senses, not all of which coincide with economic efficiency.78 For example, Heerey J refers to a firm’s ‘business efficiency’, ‘legitimate business reasons’ and, referring to US cases, ‘business judgment’.79 As I argue in Chapter 3, conduct which is logical and reasonable in terms of business judgement should not be justified unless it could be regarded as both privately and socially efficient (though we accept that the latter is legally indeterminate). Thus, we must ask whether the demand response justifies the expenditure of additional resources, ie, whether a reduction in price or increase in output would increase market revenue more than costs. If revenue falls (ie, marginal revenue is negative), an increase in output would appear to be socially inefficient. In zero-sum competition, we know that a firm may take sales away from a competitor. A reduction in price or increase in output by one firm may increase that firm’s revenue, as a result of taking sales away from its competitor, even though the demand response in the market as a whole may suggest that the conduct is inefficient. This would be an example of conduct that is rational in zero-sum competition, but that may not be efficient. This possibility should not surprise us since zero-sum competition is not synonymous with efficiency and may well lead to inefficient outcomes. Our problem is how to determine whether the conduct of a dominant firm under zero-sum competition is blameworthy. We can hardly blame firms if they are doing what we expect them to do under the paradigm of zero-sum competition. I propose that we should enquire whether the conduct was manipulative. Thus, if the firm increases its own sales and the conduct is profitable for it, notwithstanding an incidental inefficient market expansion of output, we should conclude that the conduct is not manipulative. The situation would be different if the firm does not increase its own sales, ie, the strategy is unprofitable, which may well be manipulative, absent some explanation. For example, the strategy may have been good in concept but failed in execution, or the firms may have been locked in inefficient price/output strategies and thus were meeting competition. As we will see when we consider predatory pricing, in a price war it may not be possible to blame one firm, particularly if the price war is caused by a demand shift reduction. Having explored the difference between efficiency and rational conduct, we can return to the Melway case, which does not really assist us in addressing this issue, since it concerned distribution of a product rather than competition



76 Melway

FFC (n 72). and Williams (n 1). 78 Melway FFC (n 72) [22]–[32]. 79 ibid [22]–[29]. 77 Hanks

Melway under Harper Section 46  153 between products. That is, the case primarily concerns a contest between aspiring ­distributors to determine who may make existing sales in the market. This raises no efficiency issues unless it can be shown that there is unmet demand. Merkel J found as a fact that Melway believed its distribution system enabled it to maximise sales, but concluded that: ‘The evidence does not enable me to form any view as to whether the dismantling of the current system would be likely to be harmful or beneficial to Melway’s business.’80 This is a significant gap in the factual matrix. In a case where one distributor is terminated and replaced with another, there would ordinarily be no reason to believe either that the process of competition would be harmed or that there would be any impact on market sales. In the High Court, Kirby J dissented, in part due to an inference of fact that His Honour was able to draw from the evidence that Hicks ‘proposed to sell the … directory to new retail segments not serviced by the existing distribution ­arrangements’.81 We should not confuse market segments (which in the present context refer to distribution channels) with end users. The relevant factual enquiry should concern the impact on sales in the market to end users. There is no reason to believe that there was unmet demand, ie, that Hicks would discover a new channel to end users who were not served by the existing distribution network. It appears that Hicks was not able to put evidence to the trial judge to support such a finding of fact. If Melway was satisfying the market and Hicks could not claim to be able to satisfy unmet demand from end users, Melway’s refusal to supply could well be efficient, or at least have no adverse efficiency implications or interfere with the efficient operation of the market (in which Melway’s product must compete with other products and technologies). The plaintiff deliberately chose not to run its case based on anticompetitive effect, so it may be safe to conclude that such a case would be weak and thus it should make no difference whether we apply the Harper section 46 test with mandatory consideration of efficiency or the test as enacted without that mandate. The area of uncertainty under the Harper Review section 46 concerns the suggestion accepted by the Review that market power may contribute to an anticompetitive effect, and how the courts may deal with this issue. I have argued that the structuralist approach already factors market structure into the SLC test. The section 47 SLC test provides that other exclusive dealing arrangements (eg, between Melway and its other distributors) may be taken into account when considering anticompetitive effect. Accordingly, as the opportunity was not taken to argue the section 47 SLC test, it would be surprising if the Harper section 46 as enacted would lead to any different outcome on the facts of the Melway case.



80 Melway 81 Melway

First Instance (n 55) 635. (n 70) 35.

154  Refusal to Deal and Margin Squeeze

VIII.  Melway under Market Manipulation Again, in applying this approach, we bypass structuralist conceptions of market power and SLC, and go straight to consideration of efficiency as we have adapted and redefined it along the lines discussed above. As the case concerns a finished product, with no suggestion that the plaintiff would be a more efficient distributor or would credibly satisfy unmet demand, the case does not seem to raise any efficiency issues.

IX. Conclusion We know that competition under the paradigm of zero-sum, or survival, competition can result in inefficient outcomes. The preceding cases suggest that private enforcement encourages firms to ‘game the system’, which inevitably distracts attention from the public interest in using ‘competition’ as a surrogate for efficiency and may exacerbate the risk of inefficient outcomes. This is particularly the case with refusal to deal under section 46 (and access under Part IIIA) where the legislation does not explicitly address the conventional understanding that competition law does not deprive a dominant firm of the fruits of success, ie, regulate prices and profits. Unless we bring efficiency explicitly into consideration, there would appear to be a serious risk that a court-enforced duty to deal based predominantly on competition effects may simply facilitate redistribution of surplus between firms without any innovation or efficiency benefits to consumers in the relevant market or to society. We can now consider the next (and possibly the most) problematic area to which section 46 has been applied. In Chapter 5 we consider conduct known as ‘predatory pricing’.

5 Predatory Pricing I. Introduction I do not propose to retrace the legislative amendments to section 46 regarding predatory pricing following the High Court decision in Boral,1 in which the ACCC failed in an action under section 46 alleging predatory pricing conduct by Boral. The position is amply summarised by the Harper Review: [T]he amendments themselves are cast in language that is difficult to interpret and apply in practice (while the amendments seek to prohibit pricing below cost, the expression ‘cost’ is not defined and there are circumstances in which pricing below certain measures of cost might be an ordinary business strategy in a competitive market).2

The Review considers that the recommended revised section 46 makes all the predatory pricing amendments unnecessary and that they can be repealed.3 The amendments basically overrule the US position that requires a dangerous probability that the dominant firm will recoup the losses it incurs pricing below cost.4 This was applied at first instance by the judge Heerey J in Boral First Instance.5 We may well conclude that the repeal of the post-Boral amendments suggests that the US recoupment approach will again be influential (though of course Heerey J was applying that reasoning in the context of the current section 46, which requires taking advantage of market power for a proscribed purpose). His Honour found that Boral had ‘no prospect of being able’ to recoup its losses from price-cutting,6 but nevertheless the ‘purpose’ element of section 46 was satisfied (though the existence of market power and ‘taking advantage’ were not).7 The ACCC argued in

1 Boral Besser Masonry Ltd v ACCC (2003) 215 CLR 374 (hereinafter ‘Boral’). 2 Commonwealth of Australia, Competition Policy Review, March 2015, 337. Sections 46(1AA), (1B) and (4A) inserted in 2007; section 46(1AAA) inserted in 2008. Refer to the extracts in the Appendix from current and proposed versions of section 46 and the appended summary of the legislative history. 3 ibid 345. 4 See, eg, K Hylton, Antitrust Law: Economic Theory and Common Law Evolution (Cambridge, Cambridge University Press, 2003) 212–28. L Lovdahl Gormsen, ‘Are Anti-competitive Effects Necessary for an Analysis under Article 102 TFEU?’ (2013) 36 World Competition 223, 230 notes that in the EU, recoupment is not a necessary element of predatory pricing. See Case C-202/07P France Télécom SA v Commission [2009] (CJEU, 2 April 2009) [110]–[113]. 5 ACCC v Boral Ltd [1999] FCA 1318 (hereinafter ‘Boral First Instance’) [159]–[167]. 6 ibid [169]. 7 ibid [190]–[194].

156  Predatory Pricing the High Court that the US recoupment approach is not a necessary condition for predatory pricing (ie, price-cutting can be predatory even without the possibility of recoupment) and in any event recoupment had no place in section 46 as it then stood, because there was no effects test.8 It is therefore likely, since the latter objection no longer applies, that the reasoning regarding recoupment accepted by the High Court in Boral would apply under the Harper section 46 effects test. The ACCC argued (in the context of the ‘purpose’ test) that it should be sufficient that the defendant expected to be able to restore prices to ‘normal’ levels.9 The High Court did not endorse the ACCC approach. Observations by McHugh J indicate a preference for the view that an ability to recoup at least part of the losses by pricing above normal levels is a necessary indication of market power.10 This appears to be the better view. As we saw in Chapter 3, under perfect competition, supply responds to a demand shift reduction by firms leaving the market so that prices and profits can be expected to return to pre-disturbance levels.11 Nevertheless, I suggest that our analysis should be based on what we expect to occur under zero-sum competition. Several judges in the High Court cautioned that it is dangerous to proceed from a finding about purpose to a conclusion that market power is implicated.12 We may recall the US approach discussed in Chapter 3, which cautions that evidence of hostile intent should not be taken to indicate that conduct is necessarily explained by, or solely predicated on, excluding competitors. The reason for this is that in zero-sum competition, self-preservation may well have the necessary consequence that competitors will be fatally harmed. I suggested in Chapter 3 that the judges have worked around section 4F (which provides that one substantial purpose among many is sufficient to satisfy the purpose test) by interpreting the required purpose as a substantial subjective purpose. Heerey J found that Boral considered that it was necessary for one or two firms to leave the market and thought their new more efficient plant would enable them to drive at least one competitor out, and they thought a long-term solution would be for the more efficient new entrant, C&M, to fail.13 Should it make any difference to the outcome of the case if Boral’s intent had been better expressed as ‘some firms must fail or exit the market and we want to be among the survivors’? It is on the basis of those findings that His Honour considered that there was sufficient evidence that Boral had a proscribed purpose of excluding competitors, though he acknowledged at [194] that the CCA is silent ‘about the reduction in number of competitors in

8 Boral (n 1) 384–89. 9 ibid 389; see also R Smith and D Round, ‘Section 46: Oligopoly and Predatory Pricing’ (1998) 6 Competition & Consumer Law Journal 1. 10 Boral (n 1) 464 [278]–[279]. 11 See, eg, D Besanko, D Dranove, M Shanley and S Schaefer, Economics of Strategy, 7th edn (Hoboken, Wiley, 2016) 29, Figure P.15. 12 Boral (n 1) 420 (Gleeson CJ and Callinan J) and 441 (Gaudron, Gummow and Hayne JJ). 13 Boral First Instance (n 5) [190]–[194].

Introduction  157 a genuinely competitive market’. Accordingly, we should examine the economic process under zero-sum competition by which supply in the market adjusts to a demand shift such as the economic downturn experienced in the Boral case. As we have seen in Chapter 3, when discussing Kirby J’s dissenting reasoning in Boral, in perfect competition firms would leave the market freely, without cost, so that supply would adjust to the reduced demand. However, free entry and exit does not apply in the real world, where there are barriers to exit, so firms must fight for survival. This is the paradigm I call zero-sum competition. In the High Court, Gleeson CJ and Callinan J recognised that sunk costs (which had not been considered in the ACCC’s analysis of pricing conduct) should be considered in an appropriate case.14 While sunk costs are regarded by economists as not ­relevant to decision-making, on exit certain sunk costs may be recoverable;15 thus, I suggest that the extent of recovery under different exit scenarios could be relevant. Experience suggests that if exit occurs and assets are disposed of at liquidation values, as opposed to going concern values, this typically gives rise to costs or losses that may cause firms to become insolvent and thus fail. The issue of exit costs would be familiar to most directors and officers of corporations, who face personal liability if the corporation incurs debts which the directors do not have reasonable grounds to believe can be satisfied.16 Accordingly, I suggest that exit costs should be considered as they may well represent a barrier to exit. We may readily accept, as Smith and Round argue, that cost-based tests for predatory pricing have their limitations.17 Given that exit costs may be hard to quantify, it could be difficult to attempt to assess Boral’s conduct by reference to the ‘shut down rule’ (ie, if price does not cover avoidable costs, it is said that the only rational decision is to shut down production).18 Perloff, Smith and Round argue that in the circumstances, Boral could be expected to have shut down, but did not do so, and that none of the exceptions to the shutdown rule appear to explain its decision; relevantly, it does not appear that the costs of shutting down exceed the losses from continued operation.19 However, it does not appear that the ACCC analysed those costs (and so we have no data on which to reassess Boral’s decision). We might infer that Boral itself was not able to undertake a quantitative evaluation of those costs, though it could be expected to have assessed the consequences of shutting down. Thus, loss of asset value, holding costs, and costs of reopening may well justify a decision to continue operations. There would be significant costs of reopening (eg, winning back customers whose business had 14 Boral (n 1) 407 [70], noting that sunk costs are generally not relevant, but that the issue could require consideration. 15 See, eg, Besanko et al (n 11) 16–17, where it is acknowledged that sunk costs may be recoverable on exit. 16 See ss 588G–588H of the Corporations Act 2001 (Cth). 17 Smith and Round (n 9) 15. 18 J Perloff, R Smith and D Round, Microeconomics (Frenchs Forest, Pearson Australia, 2014) 221. 19 ibid 223, noting as an example of relevant costs employee redundancy costs and rehiring and retraining costs if the firm later re-enters the market.

158  Predatory Pricing been lost) that would be avoided by continuing operations and so would affect the decision to shut down. Such factors could be even more significant on the ACCC view that entry barriers were higher than determined by the trial judge. We may well expect that the barriers to exit we find in zero-sum competition could explain the decision to stay in the market and to become more efficient by investing in additional capacity. Zero-sum competition may well provide a greater incentive for innovation and efficiency than perfect competition because less efficient firms would be forced to change their processes to become more efficient if they wish to survive. Our challenge, as Smith and Round note, is to determine whether reduced prices are predatory or a ‘socially desirable competitive response to changed market conditions of demand (a new entrant) or supply (commissioning of a new, larger, more efficient plant)’.20 We could infer that Smith and Round consider efficiency to be the test of social desirability and that, eg, new entry may be efficient if there is unmet demand. In the High Court, Gaudron, Gummow and Hayne JJ suggest that defensive price-cutting is a legitimate survival strategy in response to changed market circumstances, such as a drop in demand.21 I suggest that zero-sum competition and efficiency considerations enable us to understand and evaluate this process. In doing so, we need to inquire more deeply into the efficiency of survival strategies than their Honours suggest. How do we determine whether more efficient firms survive and can we intervene to ensure that they do? In Chapter 1, I likened this question to asking, in a criminal case where selfdefence is raised, which person is the more socially desirable, so that we could favour that person. For obvious reasons to do with equality before the law, it would not be acceptable for judges to make that kind of assessment in a criminal law case. As we will see, there is a belief in competition law that it is acceptable to make such assessments in relation to competing corporations on the basis of economic efficiency.

II.  Boral: Factual Matrix The Boral case concerned events in the Melbourne market for concrete masonry products, such as blocks, bricks and pavers, in the period from April 1994 to October 1996.22 These products are used in the building industry, which suffered a downturn as a result of the general economic recession in the State from about 1991 to 1995, though conditions in the industry remained depressed until about 1998.23 20 Smith and Round (n 9) 22. 21 Boral (n 1) 434 [171], citing D Heydon, Trade Practices Law (Sydney, Lawbook Co, 1989) vol 1 [5.760]. 22 Boral First Instance (n 5) [1]. 23 ibid [150].

Boral: Factual Matrix  159 At the relevant times, Boral had a market share of 25–30 per cent, Pioneer had 25 per cent, Rocla had 22 per cent (but exited the market during the recession), Budget Bricks had seven per cent (but likewise exited the market) and there were other manufacturers.24 Boral, Pioneer and Rocla were all vertically integrated producers of materials used in the manufacture of concrete masonry products, and producers of the final products.25 The case focused on the conduct of Boral’s division responsible for the manufacture of the final concrete masonry products. Boral argued that the profit element of transfer prices of raw materials sourced within the Boral corporate group ought to be excluded when determining whether it had priced below avoidable cost, but the trial judge accepted the ACCC’s submission that this would not be appropriate (however, the trial judge accepted that the contribution to group earnings was a legitimate factor for Boral to consider when making pricing decisions and deciding whether to exit the market).26 In the High Court, Gleeson CJ and Callinan J observed that ‘there is merit in [Boral’s] argument’.27 Considering traditional structural factors, Heerey J concluded that Boral did not possess substantial market power: the products were commodity products with available substitutes; production plant could be bought off the shelf in the US or Europe; there were low barriers to new entry (falling demand and excess capacity does not constitute a barrier); Boral’s market share stayed consistently between 25 per cent and 30 per cent throughout the period; and there were strong competitors.28 The Full Federal Court on appeal held that Boral did in fact have market power. The ACCC argued this in the High Court, essentially on the basis that Boral’s conduct, eg, pricing below avoidable cost and investing in increased plant capacity (by acquiring another Besser machine) to lower costs, was indicative or a manifestation of market power (and that the trial judge had erred in finding that barriers to entry were low).29 The ACCC’s arguments were not successful in the High Court. C&M was a manufacturer of concrete masonry products operating in the regional city of Bendigo. In 1993, it established a new operation in Melbourne with a state-of-the-art Hess machine sourced from Europe.30 This was thought to be a more efficient plant that would give C&M a cost advantage over Boral.31 24 ibid [149]; Rocla [53], [89]–[90]; Budget Bricks [101]. 25 See, eg, Kirby J’s discussion in Boral (n 1) 493 [365]–[367] of vertical integration, reflecting the structuralist view that it is anticompetitive. 26 ibid [105]–[107]. 27 Boral (n 1) 407 [68]. 28 Boral First Instance (n 5) [138]–[155]. 29 Boral (n 1) 383–84. 30 Boral First Instance (n 5) [13]. 31 ibid [46]–[47]. Perloff, Smith and Round (n 18) 162 make some observations on the technical efficiency of Besser and Hess machines for manufacturing concrete masonry products, which would take us into issues that are not obvious from the case law reports, ie, if Besser machines were technically uncompetitive compared to Hess machines, then everyone would get one, so we may need to know more about the capacity and minimum efficient scale of the particular machines used by C&M and Boral.

160  Predatory Pricing Detailed evidence of Boral’s pricing was reviewed on major tenders throughout the period. Overall, revenue exceeded variable costs in the years in question and in 22 out of the 30 months in question.32 However, Heerey J concluded that Boral priced below avoidable cost ‘for important parts of [its] product range for a significant part of the relevant period’.33 His Honour considered that a business rationale would point against a finding that a dominant firm was taking advantage of market power,34 and considered that ruthless competition designed to injure competitors was not necessarily unlawful, but that conduct to force competitors out of the market so that the firm could recoup losses later would be a proscribed purpose.35 Boral considered leaving the market, as a number of competitors had done, but decided to stay in and to ‘cut prices to win business, and upgrade its plant to improve efficiency’; these were held to be legitimate business decisions.36 The ACCC’s claim was therefore dismissed at first instance, and though the ACCC succeeded on appeal to the Full Federal Court in ACCC v Boral37 on the issues of market definition, barriers to entry and market power, the High Court on appeal in Boral reinstated the first instance decision. The High Court decision was handed down in February 2003. In June 2003, Adelaide Brighton Cement Ltd, an Australian Stock Exchange listed company, announced it would acquire C&M and Rocla for an initial consideration of A$50.5 million cash plus the assumption of debt of approximately A$10 million, making it a vertically integrated cement and concrete masonry products producer, second only in size in the Australian market to Boral.38 In 2004, Boral attempted to make a takeover of Adelaide Brighton, but withdrew when this was opposed by the ACCC.39 Without further data, we can only speculate whether C&M was a financially efficient producer able to cover costs of capital in the long run and whether vertical integration is driven by efficiency or competitive considerations.

III.  Gaps in the Factual Matrix Before we proceed to apply the Harper Review section 46 and market manipulation approach to the facts of the case, we can make some observations about 32 Boral First Instance (n 5) [110]–[112]. Perloff, Smith and Round (n 18) 223 state that Boral ‘sold blocks below avoidable costs for all but eight months’ in that period. It is unclear whether this is a misinterpretation or whether the authors had access to exhibits in the case in graphic form that are only described in outline in the judgment. 33 Boral First Instance (n 5) [119]. 34 ibid [158]. 35 ibid [167]. 36 ibid [179]–[180]. 37 ACCC v Boral [2001] FCA 30. 38 Adelaide Brighton Ltd, ‘Adelaide Brighton Acquires a Number Two Position in the Australian Concrete Masonry Products Market’, press release, 19 June 2003, www.adbri.com.au/-/adbri/lib/pdfsarchived/cmrpmaquisition.pdf. 39 ACCC, ‘Boral Limited Withdraws Takeover Bid, Provides Court Enforceable Undertakings’, press release, 8 October 2004, https://www.accc.gov.au/media-release/ boral-limited-withdraws-takeover-bid-provides-court-enforceable-undertakings.

Gaps in the Factual Matrix  161 gaps in the factual matrix, and strategic decisions made by the ACCC in running its case. First, the evidence concentrated on pricing for particular transactions, and aggregate sales revenue and variable costs during the period alleged. The published reasons state only the contribution margin, ie, the amount by which sales revenue exceeded variable costs.40 We would like to know sales revenue and quantity data, which is presumably available, to indicate whether Boral’s strategy enabled it to maintain revenue despite falling prices.41 We would like to know whether, following the demand shift reduction, industry output continued at inefficiently high levels (ie, to exceed demand) or whether industry output reduced. If, for example, the dominant firm flooded the market to depress price artificially, we could easily conclude that such conduct would be inefficient and should be regarded as manipulative.42 The situation might be more difficult to assess if the new entrant, C&M, had likewise flooded the market on entry, as that too would seem to be manipulative. It is hard to see how output could be increased artificially given a demand shift reduction, since construction activity is unlikely to be stimulated by a reduction in price for one building material. We may doubt that builders would stockpile material for future use, and if manufacturers were to build up stocks of unsold product, that would not affect the market price.43 For the present purposes, we might assume that output in the short term continued at inefficient levels simply due to market forces and our enquiry therefore focuses on the way in which firms in zero-sum competition compete to take sales away from each other and self-select for exit from the market. We may be inclined to favour more efficient firms, as we expect these to survive (and may presume foul play if they do not). But, having regard to exit costs, should we deny less efficient incumbent firms the opportunity to become more efficient, and if Boral has become more efficient, should it be penalised for doing so? Second, we may feel we need to know more about C&M and its conduct and efficiency. As noted above, we do not know whether C&M flooded the market to depress price (the same conduct which on the part of Boral would be considered blameworthy). We know that C&M seriously considered selling out and had to raise capital to survive,44 and subsequently was taken over, possibly debt-laden. I suggest that ‘efficiency’ requires us to consider more than technical efficiency of the productive plant and labour productivity, ie, requires consideration of the

40 Boral First Instance (n 5) [110]. 41 Boral’s volume of annual sales over the period is set out by Beaumont J in ACCC v Boral (n 30) [121], but we cannot infer whether Boral’s total sales increased or diminished, or whether the increased output affected output in the market (or represented sales taken away from competitors). 42 Perloff, Smith and Round (n 18) 223 appear to assume this to be the case in their analysis of the Boral decision. However, I cannot see evidence of this in the judgments, though Kirby J in his dissenting opinion in the High Court (Boral (n 1) 496 [374]) discusses a hypothetical strategy of flooding the market to depress prices. 43 Gaudron, Gummow and Hayne JJ in Boral (n 1) 441 [194] discuss the possibility of Boral building up stocks of unsold product as a strategy to keep its plant working. 44 Kirby J in Boral (n 1) 491 [356].

162  Predatory Pricing costs of capital (in turn a function of debt/equity leveraging which crucially affects business risk). We do not know the minimum efficient scale or capacity of C&M’s Hess machine (or whether Boral’s investment in a second Besser machine to create additional capacity enabled it to achieve comparable efficiencies). The wisdom of C&M’s decision to enter the market in the middle of a recession seems questionable. It appears to have taken a calculated risk of being able to take sales away from incumbent firms, knowing that its entry would increase pressures on some to exit or fail. In theory, if C&M was more efficient and had the capacity to produce a substantial part of market demand, no objection could be raised to C&M enjoying that success. If it fails, through miscalculating the risks of incumbents ­improving their efficiency to compete with it (anyone can buy a Hess machine or a larger-scale Besser machine), should C&M be able to call on competition law to protect it? Third, we know that it is pointless to ask squabbling children ‘who started it?’ Should it be relevant to enquire who started the price war? We may be satisfied to conclude that the price war was started by the demand shift reduction. However, this issue became pertinent because the ACCC argued in the appeals that Boral started the price war. This was comprehensively rejected in the High Court because Heerey J had made no relevant finding of fact to support the contention, and because the obvious cause of the price war was the demand shift reduction and the power it conferred on customers to force down prices (and, in fact, the evidence showed that Boral had been undercut by its competitors).45 Fourth, the ACCC made a strategic choice to allege predatory pricing in the period from April 1994 to October 1996. This was the subject of some curiosity in the High Court.46 Perhaps the evidence of below-cost pricing was stronger in that period. The price war broke out in mid-1993, when Boral discovered it had been undercut by Pioneer on a major project tender, and that one competitor, Rocla, decided it could not match those prices and decided to exit the market.47 However, in support of its allegation that Boral started the price war, the ACCC submitted that Boral’s market share increased from 18 per cent in December 1992 to more than 30 per cent in December 1993 due to an aggressive marketing campaign by Boral.48 This of course is outside the period in issue in the case. Recalling that C&M entered the market in mid-1993, we could probably conclude that C&M’s strategy in a depressed market was to take sales away from Boral and others. Fifth, we can see that competition law seeks to prevent conduct from giving rise to harm, ie, by intervening before harm occurs. The predictive recoupment test and likewise the predictive SLC test appear to be manifestations of this objective, which I suggest is problematic. In the area of criminal law, there is no doubt



45 Boral

(n 1) 405–06. 405–06. 47 ibid 400. 48 ibid 405. 46 ibid

Boral under Harper Section 46  163 that murder is reprehensible, yet, apart from punishing an ‘attempt’, the law has found no satisfactory way to prohibit anticipatory conduct, so we must wait to see if the victim dies.49 Can competition law’s objective of preventing anticipated harm be successful in the paradigm of zero-sum competition when firms must compete to survive? We may recall from Chapter 2 that Edward Mason was doubtful that competition effects could ever be analysed satisfactorily, so the surrogate structuralist competition test appeared to solve the problem of seeking to anticipate and prevent harm before it happens.50 In Boral, Kirby J acknowledged that ultimate effects ‘depend on too many imponderables’, but nevertheless it is unsatisfactory that we ‘must wait to see how things pan out’.51 In that passage, Kirby J argued that the purpose test in section 46 solves the problem because we do not have to wait to see if a more efficient competitor is forced out or a more concentrated market structure emerges. This overlooks the fact that the market was in the process of adjusting the number of firms to a more concentrated market based on the size of the market. McHugh J was prepared to advocate a more sophisticated analysis (albeit a structural analysis) than the cost-based approach to predatory pricing, taking into account events ‘before and after’ the conduct in question.52 I suggest that predictive tests create an unacceptable risk of error and that a more sophisticated analysis is required. I suggest that considerations of efficiency should be our focus.

IV.  Boral under Harper Section 46 We can now apply the Harper Review section 46 to the facts of the case. Where the relevant facts are unknown, we can identify issues and enquiries that would affect the outcome of the case applying the alternative tests. We know that the ACCC vigorously contested the trial judge’s findings that Boral did not have market power, and that the High Court overruled the ACCC and the Full Federal Court on the issue. It is troubling that views in the litigation differed so widely on basic issues of market definition, barriers to entry and market power. It may be safe to assume that Boral would not be held to have market power under the Harper section 46, which utilises the same concept of market power. That would be the end of the case. However, for the purposes of this discussion, let us assume that Boral in the circumstances might be found to have market power in the traditional sense, so that we can consider the SLC test, the question

49 At common law, the offence required that the victim die within ‘a year and a day’, but this rule was abolished in the author’s jurisdiction in 1991: s 9AA of the Crimes Act 1958 (Victoria). Nevertheless, we must wait for the victim to die for the offence to be constituted. 50 E Mason, ‘Market Power and Business Conduct: Some Comments’ (1956) 46 American Economic Review 471, 479. 51 Boral (n 1) 516 [442]–[443]. 52 ibid 462 [273] and 470 [292].

164  Predatory Pricing of causal link between anticompetitive effect and market power, the significance of recoupment, and the implications for innovation and efficiency. The ACCC accepts that there must be a causal link between market power and the impugned conduct under the old section 46.53 I have argued that there should still be a requirement inferred in the Harper section 46 that any anticompetitive effect must be caused or contributed to by the defendant’s market power. However, the Harper section 46 could be interpreted to use market power merely as an identifier, such as blue eyes or red hair, to distinguish persons of interest, in which case causation would not be required. We have seen that a key determinant of the number of firms in a market is the size of the market and that the way in which the number of firms responds to a demand shift reduction does not occur in the way Kirby J (dissenting) analysed it in the High Court in Boral. As we noted in Chapter 3, His Honour suggested that firms should respond by using their capacity less intensively, whereas we would expect they would be constrained by cost from doing so and it is more likely that supply would adjust to a major downturn in demand by firms leaving the market. Focusing on efficiency, as the Review recommends, we see that industry concentration will be likely to rise in response to a demand reduction shift, and we expect prices to return to pre-disturbance levels. Nevertheless, these effects are not caused or contributed to by Boral. We may conclude that the market is operating efficiently when supply responds to demand shift by firms exiting the market, and that competition law should not be concerned unless a more efficient firm was excluded which the market would not otherwise exclude. Although C&M was financially weakened and ultimately taken over, it remains a more efficient competitor than those that exited the market, and it seems that Boral improved its production processes to become a more efficient lower cost producer. We must consider whether the removal by Parliament of the reference to efficiency in the Harper Review section 46 affects the application of the SLC test. If the outdated view of the US courts during the 1950s were adopted – that we should sacrifice efficiency in the interests of less concentrated markets – we might conclude that there is likely to be a substantial lessening of competition on a largely structuralist view of the SLC test (ie, we would suspect that a more concentrated market structure would be likely to result in higher prices as a result of the ‘interdependent conduct’ of competing firms). If the predictive SLC test were applied, it might be argued that there is a real chance that prices could be increased and thus that the more concentrated market is likely to be substantially less competitive. If a causal link is still required under the Harper Review section 46 as enacted, it would seem that the ‘effect’ was not caused or contributed to by Boral, but by the contraction in the size of the market. We might ask whether Boral’s conduct had an anticompetitive effect that the same conduct by another firm without market power, eg, C&M, would not have had. We observe that C&M did engage in the

53 ibid

384.

Boral under Harper Section 46  165 same conduct, ie, it entered a market knowing it would have to force other firms out. Thus, a firm without market power could have provoked the same market shake-out, which would suggest the mischief being addressed by the Review is not present on these facts. On the other hand, if the Harper Review section 46 could be interpreted as a ‘no fault’ or strict liability provision, it would not be necessary to demonstrate a causal link between the firm’s market power and the competitive effect. All that would have to be proven is that Boral has market power and its conduct has an anticompetitive effect. Boral could be liable if pricing below avoidable cost could be said to impede entry. Since it appears that equally or less efficient firms exited due to demand reduction, I suggest that this conclusion should not hold unless it impedes entry by a more efficient competitor. If Parliament’s removal of reference to efficiency in the Harper section 46 were held to rule out consideration of efficiency, this could result in section 46 protecting equally or less efficient competitors which market size does not accommodate. The potential for section 46 to be discriminatory is evident here: a firm lacking market power could enter the market and trigger a price war, but it is possible that the Harper section 46 as enacted by Parliament could be interpreted to prevent the incumbent from meeting that competition. We might now ask whether the cost-based analysis of price, and the recoupment issue, has any role in the analysis of competition effects. I suggest that it does not when the market mechanism has enforced a reduction in the number of firms due to a demand shift. The situation would be quite different if the market had been in equilibrium prior to the conduct in question, because firms would not be responding to market forces. This situation is analysed by Smith and Round, and typically involves a larger firm cutting prices to force out a smaller firm when both are already in the market, ie, the market is in equilibrium.54 Here the cost-based analysis of price and the recoupment test are generally applicable. However, Smith and Round point out that the situation is different when a new entrant seeks to enter the market, anticipating that in order to do so, it must take sales away from existing firms and will need extra capital to cover costs of operating below efficient scale during the establishment period.55 When the new entrant is thus vulnerable, the incumbent may not need to price below avoidable cost to force it out (so limit pricing may come under scrutiny). I suggest that, by setting itself the task of taking sales away from incumbent firms, the new entrant may have to force one or more incumbents out of the market. This is zero-sum competition no different in kind from that which ensues from a demand shift reduction such as we saw in Boral. Should we want to analyse below-cost pricing, I suggest the judgments in Boral support a more rigorous analysis of exit costs, since we know that barriers to exit are a hallmark of zero-sum competition which encourage firms to ­tolerate continuing losses to a greater extent than the ‘shut down’ principle in perfect

54 Smith 55 ibid

and Round (n 9) 15–16. 16.

166  Predatory Pricing competition might otherwise suggest. With regard to recoupment, I suggest that barriers to entry are critical. On the majority findings in Boral that barriers to entry are low, it could not reasonably be contemplated that recoupment would be possible. However, if the ACCC view were to prevail (and in general we can see that when market power is found to exist, barriers to entry are likely to be high), then recoupment could be possible, though I suggest that in reality this adds very little to the SLC test itself.

V.  Boral under Market Manipulation Here we see that incumbents and new entrants all have the power to manipulate the market by increasing output beyond the point where marginal revenue ceases to cover marginal costs, and possibly also the point where demand ceases to respond to reduced prices by increasing revenue from the market. We would regard that conduct as inefficient. Nevertheless, firms lack information to know where these points lie, and firms in zero-sum competition may increase their own sales by taking sales away from other firms, so strategies of successful firms may appear efficient to them. We can see that there are some inefficiencies we may be prepared to tolerate because there is an offsetting transfer of surplus to consumers and some we may be prepared to tolerate because that is the price we pay for encouraging zero-sum competition. The first category occurs if all firms engaging in a price war expand output inefficiently, ie, beyond that required to profitably satisfy demand. Because we cannot blame any firm more than another and because consumers benefit from lower prices, I suggest that we cannot condemn any of Boral, Pioneer or C&M on efficiency grounds if this were the case. The second category relates to exit costs which firms incur when they leave the market (or unsuccessfully enter and fail). There may or may not be innovation and efficiencies initiated as a result of zero-sum competition sufficient to outweigh the losses, but this we cannot ever know. We justify zero-sum competition in the belief that the benefits will be positive in the long run.

VI. Conclusion The Boral case is an unfortunate case of little precedential value because it concerns a market in disequilibrium due to a demand shift reduction. The Harper Review section 46 as recommended (mandating consideration of efficiency effects), or as enacted with the standard SLC test, does not help us resolve the controversies in the Boral case about the fundamental concept of market power and barriers to entry. The High Court decision may well be authoritative on those issues and we would not expect the outcome to be any different under the Harper section 46.

Conclusion  167 Based on the ACCC’s submissions in the High Court, we may expect the recoupment principle to be applied under the SLC effects test. Despite reservations with the recoupment principle and difficulties applying the cost-based test in practice, the reasoning of the majority of the High Court in Boral on these issues may well be authoritative, though the decision was made in the context of the ‘purpose’ test under section 46 as it then stood. If, contrary to the views expressed in the High Court, the ACCC view that barriers to entry were high and Boral had market power were to prevail, it is suggested that the application of the Harper section 46, requiring consideration of efficiency effects, should not result in a different outcome than the High Court decision. By contrast, the market manipulation approach enables us to avoid the controversies of the Boral litigation and go straight to considerations of efficiency, since in a price war we may, in general, presume that incumbents and new entrants attempt to manipulate the market to destroy each other. This can be considered to be inefficient, and entry and exit of firms under zero-sum competition may well be inefficient, but appears to be a consequence of the paradigm of competition that our law endorses. We might question at this point, as Albert Foer does, whether the statutory policy precludes us from suggesting that ‘competition’ itself may have inefficient effects.56 As we noted in Chapter 2, it is possible under the CCA to obtain administrative authorisation from the ACCC/ACT to permit anticompetitive conduct when that would have overriding efficiency or other benefits. However, in the context of enforcement proceedings or a private action, it is not open to a defendant to argue that inefficiency (ie, harm) is an inescapable consequence of the process of competition. Nevertheless, we find this to be the case in a price war which may be triggered by market forces or by new entry, rather than a predatory incumbent, ie, zero-sum competition is not necessarily efficient. Accordingly, we may wonder how the statutory policy can tolerate such inefficiency. I suggest that we may tolerate inefficient price wars because surplus is transferred to consumers as a result. However, the contrasting situation, a bidding war for inputs, appears to be inefficient without any compensating benefit. We will examine bidding wars in Chapter 7 below in relation to ‘raising rivals’ costs’. I will argue that there may be a compensating benefit, in that bidding wars for inputs may transfer surplus upstream, ie, away from the dominant firm. We may thus begin to see an element of bipolarity in competition law: the ­paradigm of zero-sum competition is endorsed, but competition law may seek to lay blame on a dominant firm for inefficiencies which may be caused by the paradigm itself. Thus, attempts to restrain uses of market power which can be characterised as ‘predatory pricing’ and ‘raising rivals’ costs’ may be opposite

56 A Foer, ‘On the Inefficiencies of Efficiency as the Single-Minded Goal of Antitrust’ (2015) 60 Antitrust Bulletin 103, 115.

168  Predatory Pricing ends of the spectrum on which competition law seeks to ameliorate the downside of zero-sum competition. I suggest that this use of competition law may well be misguided insofar as we seek to penalise firms for behaving as our paradigm of competition requires. Our approach to efficiency in the present section has focused on the size of the market, which constrains entry and encourages exit. We still need to understand how firms self-select for entry and exit when this is not free (as it is assumed to be under perfect competition). We should not be concerned if an equally or less efficient firm were excluded, if that were a consequence of market size (as opposed to exclusionary conduct of the dominant firm). We should be concerned if a more efficient firm were excluded, as that would be unexpected. To determine relative efficiency of firms, we would inquire into the firm’s technical and financial efficiency. We anticipate some forensic difficulties with this enquiry. Firms have records of costs and sales, and costs of capital, but these are accounting, not economic measures. Confidentiality is not a bar to courts having access to the necessary evidence in competition law cases, but judgments will generally be redacted to protect confidentiality, so we may lack the data necessary for this study to apply the proposed principles with specificity. Nevertheless, we can test conditions of possibility for different outcomes applying the principles of the Harper section 46 and the market manipulation approach, and can assess whether either has advantages or disadvantages for forensic application. Having assessed the relative efficiency of the firms, we would then enquire whether the exclusion were the result of manipulation of the market by another firm, large or small. If exclusion has not occurred, we must question the ability of the court to predict the outcome of the competitive process, knowing that circumstances change, inefficient firms may innovate and seemingly efficient firms may fail to succeed due to underestimating a variety of commercial risks. I would suggest that a predictive test entails a significant risk of error and that the court’s power to grant injunctions to prevent threatened harm should be adequate to protect the public interest. We can now proceed to the next key decision to examine further the implications of the paradigm of zero-sum competition which concern the extent to which a dominant firm may defend itself against a competitive threat. In Chapter 6 we consider how section 46 might apply when a dominant firm responds to new entry.

6 Meeting Competition I. Introduction In this chapter we examine ACCC v Rural Press Ltd (hereinafter ‘Rural Press First Instance’). This is a case where the section 46 claim failed on appeal, but, by virtue of communications between the incumbent and the new entrant, the subsequent withdrawal of the new entrant was held to be the consequence of a market sharing (ie, cartel) arrangement that is per se prohibited.1 Our interest concerns unilateral conduct, so we will examine the facts of the case assuming that each party acts independently, ie, we want to consider how the incumbent may be expected to respond to new entry and thus what conduct should be permissible. At first instance, Mansfield J held that Rural Press possessed substantial market power in its home market and took advantage of it by threatening to retaliate against the new entrant by entering the new entrant’s home market in an adjoining geographical region.2 His Honour recognised that, in general, there is no reason why a firm such as Rural Press may not engage in competitive conduct of that kind, but held that in the circumstances of the case, the conduct of Rural Press in threatening retaliation of that nature could only be explained by its market power.3 On appeal, the finding of market power was not challenged, but both the Full Court of the Federal Court in Rural Press v ACCC4 and the High Court in Rural Press5 upheld the appeal by Rural Press on the section 46 aspect on the grounds that Rural Press had not taken advantage of its market power because the threatened conduct concerned entry to the adjoining market in which it had no market power, ie, such conduct would not rely on market power. The case demonstrates the consequences of the ACCC’s strategic decision to ‘plead and argue for a very limited market, with the obvious advantages that

1 The case concerned the prohibition of ‘exclusionary provisions’ under ss 4D and 45. These provisions overlap the cartel conduct provisions in Division 1 of pt IV of the CCA. Upon amending legislation being passed to give effect to the Harper Review recommendations, the per se prohibition of exclusionary provisions will be repealed and subsumed in the cartel conduct provisions. 2 Rural Press First Instance [2001] FCA 116 [123] and [131]. 3 ibid [132]. 4 Rural Press v ACCC [2002] FCAFC 213 [142]–[143]. 5 Rural Press v ACCC (2003) 216 CLR 53 (hereinafter ‘Rural Press’) [52]–[53].

170  Meeting Competition accrued to it in relation to the existence of market power’.6 The trial judge accepted the ACCC’s submissions that the relevant market is the market for the provision of services to readers and advertisers through the publication of a regional newspaper in the Murray Bridge district of South Australia.7 That is, electronic media and regional newspapers in adjoining districts are not sufficiently close substitutes to form part of the market. We can see that attracting readers is a means to an end: offering advertisers access to a sufficiently large body of potential buyers. The price (if any) that readers are charged does not in itself have any competitive implications.8 We would probably expect a more sophisticated analysis today of market definition in the case of two-sided platform markets.9 The analysis of platform markets has implications not only for market definition and market size, but also for the analysis of the profitability of pricing decisions and strategies on both sides of the platform.10 We do not have enough data in Rural Press to make specific assessments about the profitability of action taken by the dominant firm to meet competition, but can observe some general principles to assess that conduct. For our analysis, we will assume that the incumbent, Rural Press, actually responds to new entry by taking steps that were only contemplated or threatened in the circumstances considered by the court. We will also assume that there are no communications between the parties that could give rise to cartel conduct, ie, that each party acts independently. We will consider how the Harper Review section 46 and the market manipulation approach might apply in such circumstances.

II.  Rural Press: Factual Matrix Rural Press is a publisher of regional newspapers in South Australia. It operates two printing facilities, one at Whyalla and one at Murray Bridge. From the Murray Bridge facility, it prints and distributes newspapers for five regions in southeast South Australia.11 One of these newspapers is the Murray Valley Standard, which is distributed in the region surrounding Murray Bridge, including the township of Mannum, 30 kilometres away. This newspaper is published twice a week with a daily circulation of about 4,500 copies sold to readers at a price of 90 cents.12 6 Observation of the Full Court of the Federal Court in Rural Press v ACCC (n 4) [151]. 7 Rural Press First Instance (n 2) [97]–[108]. 8 See, eg, G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011) 89–90 for a discussion of ‘two-sided platforms’. 9 See, eg, D Evans and R Schmalensee, ‘The Antitrust Analysis of Multisided Platform Businesses’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 1, 406. 10 ibid 424. 11 Rural Press First Instance (n 2) [97]–[108]. 12 ibid [5].

Rural Press: Factual Matrix  171 Regional newspapers from adjacent regional areas are sold in the Murray Bridge district, but copies are few and they do not provide effective competition for Rural Press.13 By contrast, in the adjacent Barossa Valley region, Rural Press publishes a newspaper which competes with Leader Newspapers and, due to that competition, Rural Press distributes its newspaper in that district free to readers.14 About 220 kilometres upriver from Mannum is the town of Renmark. Waikerie Printing and its associates have a facility there from which three regional newspapers are published and distributed in the adjacent regions.15 Waikerie Printing publishes a newspaper distributed in the region centred on Waikerie, which is about 150 kilometres upriver from Mannum. This newspaper is published once a week and has a circulation of about 2,500 copies sold to readers at a price of 60 cents. Waikerie Printing sold a few copies in Mannum, but this was not regarded as part of its prime circulation area.16 In July 1997, Waikerie Printing arranged for two successive editions of its newspaper to be mailed to all households in Mannum free of charge and announced that copies could thereafter be obtained from the newsagency in Mannum.17 It appears that this was the first step in a 12-month trial to see if it would be financially viable.18 However, it was not until September 1997 that Waikerie Printing actively sought to obtain advertisers and readers in and around Mannum to expand its circulation.19 The trial judge found it ‘difficult to conclude with precision the extent of the increased circulation’, but believed it to be ‘in the order of 100 copies per week’, though Waikerie Printing claimed higher circulation.20 It appears that Waikerie Printing did not gain advertising until about January 1998 and that the advertising rates it offered were significantly lower than those offered by Rural Press.21 Initially, Rural Press responded by increasing its coverage of community news relevant to Mannum.22 However, the Rural Press manager responsible for the newspaper, Ms Beryl Price, became increasingly concerned that Rural Press should take action to protect its revenue. She proposed distributing their local newspaper in Waikerie Printing’s district or distributing a free bulletin there,23 and in February 1998 projected that advertising revenue could make the venture profitable.24 During March 1998, Ms Price’s urging to senior management became 13 ibid [8]. 14 ibid [23]. 15 ibid [10]. 16 ibid [9]. 17 ibid [21]. 18 ibid [20]. 19 ibid [29]. 20 ibid [18]. 21 ibid [33] and [64]. 22 ibid [30]. 23 ibid [28] and [31]. 24 ibid [38]. This is a critical issue for the game theoretical analysis of the case undertaken by J Gans, R Sood and P Williams, ‘The Decision of the High Court in Rural Press: How the Literature on Credible

172  Meeting Competition more strident, suggesting that they needed to be seen to respond to the competitive threat to be taken seriously by Waikerie Printing, lest they look like ‘wimps’.25 For our purposes, we will assume that Rural Press actually commences distributing its local newspaper, or a free bulletin, in the Waikerie district and that advertising revenue exceeds the avoidable cost of so doing. However, in about April 1998, Waikerie Printing withdrew from Mannum, ie, ceased to seek increased circulation in or advertising revenue from Mannum, and reverted to its prime circulation area.26 Waikerie Printing submitted that its reasons were that the expansion into Mannum had not been profitable; however, the trial judge rejected this explanation, finding that documents had been prepared after the event to mask the true reason – that Waikerie Printing feared (due to the communications that had taken place between the parties) an incursion into its home market by Rural Press.27 Absent the communications and threats that took place, we can consider whether that decision might be rational or efficient, and whether a court would compel Waikerie Printing to compete with Rural Press in the Murray Bridge district or restrain Rural Press from competing with Waikerie Printing in the Waikerie district. Rural Press was a company listed on the Australian Stock Exchange and was subsequently acquired by Fairfax Media Ltd in 2006. At the time of writing, the circulation of Rural Press’ local newspaper is 2,948 and each issue sells for A$1.50.28 The circulation of Waikerie Printing’s local newspaper is 2,250 and each issue sells for A$1.40.29 Both appear now to distribute their newspapers over the internet. Other than noting that the definition of the relevant market (and hence market size) might not be the same today, we will proceed on the basis of the original findings.

III.  Gaps in the Factual Matrix We can now address some gaps in the factual matrix. The first is that we lack data concerning the profitability of either party expanding its newspaper publishing into the territory of the other. If that would be profitable, ie, additional revenue would cover avoidable costs in a reasonable timeframe, that would appear to be

Threats May Have Materially Facilitated a Better Decision’ (2004) 32 Australian Business Law Review 337, 342, where they argue that Rural Press would not be relying on market power unless entry to the Waikerie district was not profitable in itself, but would be justified by monopoly profits in the Murray Bridge district which the strategy might restore. 25 ibid [43]. 26 ibid [62]. 27 ibid [32], [63]–[72]. Gans, Sood and Williams (n 24) 343 similarly assume that it was profitable for Waikerie Printing to operate in the Murray Bridge district and that the market would support both firms deriving normal profits in that market. 28 Country Press SA, www.sacountrypress.com.au/newspaper-lists/murray-bridge. 29 Country Press SA, www.sacountrypress.com.au/newspaper-lists/the-river-news.

Gaps in the Factual Matrix  173 a normal competitive response. In that event, we may even doubt the ACCC’s narrow geographical definition of the market (ie, the market would appear to be wider). There does not appear to have been any rigorous analysis by Waikerie Printing prior to embarking on the Mannum venture, which may suggest it was not well conceived.30 If the intention was to review profitably after 12 months, we might expect some analysis of the expected profitability, including the capacity to sustain losses during an establishment phase. Arguably, the failure to properly analyse the ‘shut down’ decision does not establish that the venture was profitable (though the trial judge infers it was).31 We would like to see more evidence of the demand response, ie, did Waikerie Printing satisfy unmet demand or did it, by significantly reducing advertising rates, merely take advertising revenue away from Rural Press? We do not know whether the viability of the exercise relied on Waikerie Printing attracting advertisers away from the Murray Bride district or whether advertisers in the Waikerie district could feasibly supply residents of Mannum, though it seems clear that Waikerie Printing needed to solicit advertising revenue from Mannum. Second, we lack data about the dynamic conditions of demand and supply in the market. It is evident from the first instance judgment that competitive pressures on Rural Press and Leader in the neighbouring Barossa Valley district were intense. The Full Court made some observations suggesting that regional print media may have had excess capacity,32 possibly suggesting that supply may have exceeded demand and indicating against market power. Further, common knowledge suggests competitive pressures on print media from, eg, internet platforms might cause us to doubt the ACCC’s narrow market definition confining the market to print media. Third, other than some qualitative response, in the form of increased community news of interest to readers in Mannum, we do not know what competitive steps Rural Press might have taken. Did it reduce advertising rates to retain sales? Could it have approached advertisers in Waikerie Printing’s district who might have wanted to reach consumers in the Murray Bridge district (ie, without necessarily distributing Rural Press’s newspaper in the Waikerie district)? The trial judge was at pains to point out that: ‘It is not a consequence of my conclusion that Rural Press … may not engage in competitive conduct by publishing a new regional newspaper in a regional market in competition with another publisher.’33 We would like to know under what circumstances this might contravene section  46 and under what circumstances it might be regarded as normal competitive conduct.

30 Kirby J in his dissent argued that section 46 was infringed, though he described Waikerie Printing’s managers as starry-eyed dreamers: Rural Press (n 5) 99–100. 31 ibid [64]. 32 Rural Press v ACCC (n 4) [143]. 33 ibid [132].

174  Meeting Competition Fourth, the significance, if any, of ownership and location of printing facilities receives little attention in the judgments. Clearly, a publisher does not need to own a printing press or locate it in the area in which a newspaper is circulated, though these factors will affect the marginal costs of additional output. The case therefore gives us no real guidance as to how we might assess ‘normal competitive conduct’ by an incumbent firm when a new entrant tries to take sales away from it, ie, under zero-sum competition. We can now consider how the Harper Review section 46 might apply to the factual matrix assuming that Rural Press were to take actual steps to protect its revenue.

IV.  Rural Press under Harper Section 46 While we may have some doubts about market definition, market size and the finding of market power, we will assume that Rural Press would be held to have market power under the Harper Review section 46. Despite the noted gaps in the factual matrix, we can identify issues and enquiries that would affect the outcome under the SLC test. Where firms from neighbouring markets may profitably enter any given regional market, we may doubt that there are barriers to entry that are sufficient to indicate market power. However, assuming there is market power, we can infer that the Harper Review wishes us to consider the size of the market, ie, is the market big enough to support two regional newspapers or are strategies of the dominant firm excluding equally efficient firms from the market? Assuming that the firms are equally efficient, what do we anticipate to be the consequence of price ­competition? If our focus under the SLC test is the process by which any firm may enter and secure a viable share of the market rather than to protect the ‘right’ of Waikerie Printing to do so, should we consider whether other neighbouring firms, such as Leader, could still do so? I suggest that in order to apply the SLC test, we would need to form a view as to whether Waikerie Printing exited the market because the demand response did not support viable entry. The High Court emphasised in Rural Press that causal connection between market power and conduct is essential under the old section 46.34 Presuming this still to be a requirement in the Harper Review section 46, we would ask, if Waikerie Printing is viable, whether conduct of Rural Press forced it out of the market, eg, pricing below avoidable cost. We can see that two equally efficient firms will compete to gain economies of scale, and that one may succeed and one may fail. We can see that there are advantages in a firm publishing a cluster of local newspapers. There may be barriers to entry, but they are not created by the



34 Rural

Press (n 5) [53]–[56].

Rural Press under Market Manipulation  175 incumbent. The fact that the incumbent will reduce its prices to compete or seek to expand its revenue base to adjoining geographical areas simply suggests the process of competition is working. Unless it is suggested that the incumbent has priced below avoidable cost, we do not have an issue of predation (if we did, the analysis would be as set out above in relation to Boral). We can see that a structuralist SLC analysis might emphasise barriers to entry and market power rather than the efficiency of market entry and reasons for exit. In theory, we know that harm to Waikerie Printing is not synonymous with harm to competition. However, if we do not consider efficiency, we can easily conclude that they are synonymous. As we will see when we consider ‘raising rivals’ costs’ in Chapter 7 below, there is a risk in the case of a bidding war for inputs that ‘harm to competitors’ can be treated as synonymous with ‘harm to the process of competition’. The predictive SLC test poses some difficulties. If, as we have assumed, we interpose ourselves into the market in April 1998 when Rural Press and Waikerie Printing are cutting prices, soliciting each other’s readers and advertisers across adjoining districts, we will find it hard to predict the end result. Should we attempt to predict which firm is more efficient, and hence likely to survive, or should we wait and see, allowing the competitive process to play itself out? Would the court restrain competitive tactics of the dominant firm, such as price-cutting above avoidable cost or entry into the rival’s home market, which might be profitable on that measure? Would the court have to consider the capacity of management of either firm to successfully carry out its strategy? We can see that it would be desirable to consider the relationship between ‘competition effects’ and ‘efficiency effects’ when, based on the demand response, it appears that a competitive market structure consisting of two firms may not be efficient. This would become even more important if it were held that causation is no longer required, ie, if the Harper Review section 46 were interpreted as a ‘no fault’ or strict liability provision.

V.  Rural Press under Market Manipulation We see again here that both the new entrant and the incumbent have the power to manipulate the market, and so can immediately direct our attention to the efficiency of the conduct (ie, if price-cutting fails to cover avoidable costs or if output is expanded beyond the level indicated by demand response) and the relative efficiency of the firms. Under zero-sum competition, we expect the process of entry and exit, by which supply responds to demand, to generate losses because the incumbent and the new entrant cannot accurately predict how demand will respond or which of them may be more efficient. We can be sure that zero-sum competition will rigorously answer these questions in due course.

176  Meeting Competition If Waikerie Printing were to desist from its attempt to take revenue away from Rural Press, it would seem strange that the court might be asked to grant a mandatory injunction requiring it to continue to trade35 or to penalise Waikerie Printing for not doing so. It would be strange also if, after testing the water in a market, the would-be entrant withdrew and by doing so renders itself liable to an allegation of cartel conduct, which does not require any analysis as to whether the withdrawal is due to normal competition or could be explained on efficiency grounds.

VI. Conclusion It is unfortunate that the Rural Press case concerns only threatened conduct, and communication between competitors, since the court did not need to consider what kind of competitive conduct the incumbent dominant firm would be permitted to undertake to meet competition from a new entrant. A structuralist SLC test, which presumes that the conduct of a firm lacking market power may take on significant anticompetitive consequences if engaged in by a firm with market power, has the potential to favour a less efficient firm. The Harper Review recommendation that efficiency must be considered would do much to alleviate this risk. However, I suggest that the traditional concept of market power on which the Harper section 46 relies tends to favour a presumption that the incumbent is there because of market power, when the incumbent’s position may be explained by its business success or the size of the market. An enquiry into market structure and the current state of the competitive process may not provide the explanation we need, whereas an enquiry into considerations of efficiency should do so. In the next chapter, we examine another case where there was held to be market power but no ‘taking advantage’, so the section 46 claim failed. However, the case succeeded under the section 45 SLC test and thus we may anticipate the possible application of the Harper Review section 46 based on the same SLC test. Cement Australia sought to secure exclusive access to an unprocessed raw material input which was held to exclude rivals from the market for the processed raw material. This strategy might be described as ‘cornering’ the market. The case is interesting because it succeeded under the ‘purpose’ limb and partly succeeded under the ‘effects’ limb of the SLC test. In some respects, ‘effects’ were held not to be established because market circumstances intervened. Despite this, the analysis of ‘effects’ proceeded as a matter of inference, facilitated by the predictive SLC test. In the phase of the case in which the penalty is determined, the problem of assessing actual harmful effects, eg, the effect on market price relative to the ‘competitive

35 We may doubt whether such an injunction would ever be contemplated; see, eg, the reasoning in Co-operative Insurance Society Ltd v Argyll Stores (Holdings) Ltd [1997] UKHL 17.

Conclusion  177 level’, is confronted and found to be imponderable. Despite this, the penalty is ultimately determined as a qualitative exercise of judicial discretion. I suggest that the highly qualitative nature of the SLC test would do little to address concerns that judicial reasoning in market power cases masks undisclosed policy decisions. However, we can examine what difference the Harper Review section 46 may make, and can consider how ‘effects’ can and should be proved in cases concerning market power. Since the ‘process of competition’ is a surrogate test for economic efficiency, I suggest that we cannot afford to ignore efficiency effects when we assess harm. Further, since ‘cornering’ is a classic example of market manipulation, we can examine how the market manipulation approach might be applied and how it might affect the outcome.

7 Raising Rivals’ Costs I. Introduction The concept of a dominant firm adopting a strategy to raise rivals’ costs is intuitively straightforward. Church and Ware cite the example of the 1945 US antitrust case against Alcoa, which overbought bauxite to lock up low-cost sources and force competitors to seek higher-cost reserves.1 However, David Scheffman and Richard Higgins point out that overbuying also raises the dominant firm’s own costs and that in order to determine whether there is an anticompetitive effect, we must ask whether the dominant firm as a result is able to raise prices above the competitive level in the relevant market.2 I suggested in ­Chapter 1 that the ‘competitive level’ is conceptually problematic as a benchmark, since it relies on many assumptions that we know are not satisfied under realworld conditions. It seems to be accepted in any event that the ‘competitive level’ is indeterminate.3 Hence, I suggest that the benchmark could alternatively be the actual position in which firms find themselves in the real-world market. Scheffman and Higgins suggest that the relevant market in which to consider anticompetitive effect is the downstream market and accordingly the analysis usually assumes that the dominant firm has significant market power in the downstream market ‘independent of any cost-raising strategies’.4 I infer the reason is that a firm does not need market power to corner a market (though having done so gains transitory market power), but it would not be able to raise prices to recoup its own additional costs arising from the strategy unless it has market power in the

1 J Church and R Ware, Industrial Organisation (Boston, Irwin McGraw-Hill, 2000) 639, citing US v Aluminum Co of America (1945) 148 F 2d 416. 2 D Scheffman and R Higgins, ‘Raising Rivals’ Costs’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 2, 62 and 65; see also Church and Ware (n 1) 634–35, who similarly note that the net effect on prices is hard to determine, since the strategy raises both the dominant firm’s costs and rivals’ costs. 3 See, eg, R Smith and A Merrett, ‘Unwrapping a Fallacy: Market Definition, Market Power & Cellophane’ (2014) 18 Competition 1, 4, available at thestateofcompetition.com.au/newsletter/issue18-the-cellophane-fallacy, citing Office of Fair Trading (report prepared by National Economic Research Associates), ‘The Role of Market Definition in Monopoly and Dominance Enquiries’, Economic Discussion Paper 2, July 2001 [3.8]: ‘in practice it is extremely difficult and in most cases impossible to determine the competitive price level’. 4 Scheffman and Higgins (n 2) 66.

Introduction  179 downstream market for the finished product. In Alcoa’s case, that would be the market for aluminium, where, presumably, it already has market power. Scheffman and Higgins also point out that the effect on competition of strategies that raise rivals’ costs cannot be predicted and must be assessed empirically, eg, the dominant firm may improve quality or efficiency through scale economies, which is what we expect competition to deliver, and that the effect of the strategy on downstream prices depends crucially on how demand responds.5 We can thus see that ‘overbuying’ has some aspects in common with ‘refusal to deal and margin squeeze’, in that the dominant firm may refuse to supply the input to its competitors or may do so at prices that squeeze the downstream margin of competitors. We can also see that to apply ‘raising rivals’ cost’ theory to a firm that does not have independent market power in a downstream market (ie, separate from the cornered market) raises some questions that the competition analysis in Cement Australia6 fails to confront. I suggest that the paradigm of zero-sum competition enables us to address these questions about the nature of the harm from overbuying an input. We have seen in our discussion of Boral that zero-sum competition may produce inefficient outcomes, eg, in a price war, but this may be tolerated in that context because there is a transfer of surplus to consumers. We can now consider the opposite situation of a bidding war for inputs. Here there can be no transfer of surplus to consumers. If our concern were to prevent a bidding war resulting in a transfer of surplus from consumers in the downstream market for the final product, we would enquire whether there is market power in the downstream market such that the extra costs of the strategy would affect consumers or even be exacerbated by the ‘double marginalisation’ problem.7 Thus, the correct identification of the downstream market and analysis of effects becomes crucial. We can pause here to note the ongoing debate about vertical integration and vertical foreclosure which typically arises in the context of mergers. If there is power in both upstream and downstream markets, a case could be made that vertical integration is more efficient (and beneficial to consumers) than separate firms all charging above normal margins.8 If there are efficiency benefits in a firm backward integrating the upstream market, and rivals for the input are not foreclosed, harm from vertical foreclosure would appear to be substantially mitigated or at least difficult to determine.9 One benefit of backward vertical integration is that 5 ibid 64–66. 6 ACCC v Cement Australia Pty Ltd [2013] FCA 909 (hereinafter ‘Cement Australia’). 7 See, eg, M Motta, Competition Policy: Theory and Practice (Cambridge, Cambridge University Press, 2004) 307–09. 8 ibid. 9 See, eg, M Salinger, ‘Vertical Mergers’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics (Oxford, Oxford University Press, 2015) vol 1, 551, 553; G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011) 363–68; D Besanko, D Dranove, M Shanley and S Schaefer, Economics of Strategy, 7th edn (Hoboken, NJ, Wiley, 2016) 97 Motta (n 7) 389–90. The literature on vertical mergers is beyond our scope, however, some aspects relating to vertical foreclosure are relevant to ‘raising rivals’ costs’ theory.

180  Raising Rivals’ Costs the firm obtains security of supply of the input, ie, solves the ‘holdup’ problem and achieves transaction cost efficiencies.10 We might conclude that if the vertically integrated firm secures efficiencies in supply for itself and does not foreclose supply to rivals for the intermediate input, they too will benefit (though they will have to pay for the benefit). It appears that harm from vertical foreclosure is generally assessed by comparing the profit that the dominant firm would forgo in the upstream market (if it refused to supply competitors) with the profit it may gain in the downstream market by increasing rivals’ costs of obtaining the input (which depends on demand response in the downstream market), after taking into account efficiency benefits of vertical integration.11 If the input is a relatively small part of the total costs of producing the downstream product, foreclosure would have little effect on downstream competition.12 If rivals were disadvantaged, they might seek to backward integrate themselves to mitigate the risk of harm.13 What we may take away from the discussion of vertical foreclosure is that, whereas SCP thinking regards vertical integration as an anticompetitive aspect of market structure, whether vertical integration is efficient or contributes to anticompetitive harm is a matter that can only be resolved by empirical enquiry in the circumstances of the case. Again, we learn little from the Cement Australia case about this matter, though, as we will see, we may suspect from subsequent events that it played a part in the ACCC’s reasoning. Let us return to the analysis of bidding wars under the paradigm of zero-sum competition. If we are focused on the efficient operation of the market, we would enquire whether the bidding war, by raising costs and possibly prices, results in reduced output and hence unmet demand. Thus, we need to consider demand response and supply response. The analysis seems to correspond to that we encountered in the Boral case, where a demand shift reduction causes supply to respond by firms exiting the market. In the present context, if a bidding war for inputs raises the costs of both the incumbent and its competitors, we would expect prices to rise and demand to fall. If this effect is of a sufficient magnitude, it may well result in firms leaving the market. Clearly, the firms engaging in a bidding war wish to be among the survivors. We would expect firms that are, or can become, more efficient would survive. In order for this strategy to force less efficient firms out of the market, we would expect that demand would have to reduce significantly in response to a price rise. On the other hand, if demand does not respond much to a price rise, we might expect cost increases to flow through to consumers, unless competition between firms diminishes that effect. In the Cement Australia case, we have an input where available substitutes constrain prices and demand, where the input in question may form only a small part of the price of

10 See

Besanko et al (n 9) 113. Jenkins and Kavanagh (n 9) 364; Motta (n 7) 390. 12 Niels, Jenkins and Kavanagh (n 9) 365. 13 ibid 367. 11 Niels,

Introduction  181 the downstream finished product and where the price of the final product may be constrained by effective competition between firms. As we will see, neither the legal case nor the subsequent related administrative decisions of the ACCC give us the data we need to assess actual effects. Proceedings against companies in the Cement Australia corporate group (which I refer to collectively hereinafter as ‘CA’), arising from events in the period 2001–06 were commenced in 2008 and were decided by Greenwood J in the Federal Court on 10 September 2013. Final declaratory orders were made on 28 February 2014.14 Final orders regarding penalties and costs were made on 16 May 2016.15 The Court ordered CA to pay A$18.6 million in pecuniary penalties. On 6 June 2016, the ACCC appealed, seeking a higher penalty, having proposed penalties of A$97 million to the trial judge.16 During this extended period, the ACCC made two administrative decisions to permit conduct of the same kind that would otherwise contravene the CCA. We can contrast the facts and reasoning in those matters with the facts and reasoning in Cement Australia. The case concerned contracts entered into or given effect to by CA in the period 2001–06 to acquire fly ash in South East (SE) Queensland. Fly ash is an input used by CA and its competitors as a partial substitute for Portland cement in producing cementitious material used to make concrete. Fly ash is said to be a partial substitute for cement, but the actual usage depends on the properties of the fly ash and the structural application of the concrete, so fly ash can vary between 15 per cent and 60 per cent of the cementitious material component of concrete.17 It seems that fly ash and Portland cement may be supplied in blended form or separately. The cementitious component of concrete is generally around 15 per cent by weight, so if there is any fly ash in the concrete at all, it would probably not exceed 10 per cent of the concrete by weight.18 In economic terms, fly ash and cement are complements as well as substitutes because they must be used together. It seems that fly ash is less expensive than Portland cement, though the degree of substitution depends on technical matters. All other things being equal, the price of fly ash appears to be constrained by the price of cement.

14 ACCC v Cement Australia Pty Ltd [2014] FCA 148. 15 ACCC v Cement Australia Pty Ltd [2016] FCA 453 (hereinafter ‘CA Penalties’). 16 ACCC, ‘ACCC Appeals Cement Australia Level of Penalties’, press release 99/16, 6 June 2016. Judgment of the Full Court was reserved on 22 February 2017. On 5 October 2017, the Full Court in ACCC v Cement Australia Pty Ltd [2017] FCAFC 159 set aside the trial judge’s decision on penalties (for reasons which do not affect our discussion of the case) and substituted its own decision on penalties, awarding penalties of A$20.6 million. 17 Portland Cement Association, ‘Optimizing the Use of Fly Ash in Concrete’, 6 December 2007, https://www.cement.org/docs/default-source/fc_concrete_technology/is548-optimizing-the-use-offly-ash-concrete.pdf. 18 ibid 4 cites a survey of concrete producers in North America to suggest 46 per cent of ready-mixed concrete in the survey period in 1998 contained no fly ash.

182  Raising Rivals’ Costs CA was formed in 2003 by merger with Holcim and was at the time of the merger owned 50 per cent by Holcim, 25 per cent by Hanson and 25 per cent by Rinker, all global cement manufacturers.19 In 2007, Heidelberg Cement acquired Hanson.20 Rinker was subsequently acquired by Cemex, which in turn was acquired by Holcim in 2009.21 By 2013, CA was owned 50 per cent each by Holcim and Hanson/Heidelberg.22 In July 2015, Holcim merged with Lafarge, another global building materials manufacturer.23 As we will see, in addition to the significant forces in play at the relevant times concerning global industry concentration, the circumstances concerning the production and availability of fly ash in SE Queensland continued to change following the period in question in the legal proceedings.

II.  Cement Australia: Factual Matrix Unprocessed fly ash is a waste product that arises from burning coal in electricity generation facilities. It can be processed to obtain fine grade fly ash (called concrete grade fly ash or ‘cgfa’) that is suitable for use in the manufacture of cementitious materials. According to Craig Heidrich, in 2002 only about 10 per cent of fly ash produced in Australia and New Zealand was used in cementmaking, and although there are some other commercial uses, about two-thirds of fly ash produced must be disposed of as waste in compliance with environment protection regulations.24 The coal-fired power stations in SE Queensland are distant from power stations in north Queensland (600–700 kilometres away) and power stations in northern New South Wales (700–1,000 kilometres away).25 The following map gives a schematic representation of Annexure D to the judgment, showing sources of supply of fly ash in SE Queensland in 2001.

19 Holcim, ‘Holcim Confirms Merger of its Australian Interests into the Newly Formed Cement Australia Pty Ltd’, press release, 2 June 2003, https://www.lafargeholcim.com/holcim-archives-holcimconfirms-merger-of-its-australian-interests-into-the-newly-formed-cement-australia-pty-ltd. 20 Reuters, ‘Heidelberg Cement Agrees to Buy Hanson’, 15 May 2007, www.reuters.com/article/ us-hanson-heidelberg-idUSL1530198520070515. 21 Holcim, ‘Acquisition of Cemex Australian Successfully Completed’, press release, 1 ­October 2009, https://www.holcim.co.id/en/media/latest-press-releases/latest-release/article/acquisition-ofcemex-australia-successfully-completed. 22 Holcim, ‘Holcim and Heidelberg Cement Intend Joint Control of Cement Australia’, press release, 10 July 2015, https://www.lafargeholcim.com/holcim-and-heidelbergcement-intend-jointcontrol-cement-australia. 23 Holcim, ‘Holcim and Lafarge Merge to Create LafargeHolcim’, press release, 10 July 2015, https:// www.lafargeholcim.com/holcim-and-lafarge-complete-merger-and-create-lafargeholcim-a-newleader-building-materials-industry. 24 C Heidrich, ‘Ash Utilisation: An Australian Perspective’, conference paper, 23 September 2003, www.flyash.info/2003/03heid.pdf. 25 Cement Australia [2013] FCA 909, 310 ALR 165, Annexures A–E.

Cement Australia: Factual Matrix  183 Figure 7.1  Fly ash distribution in SE Queensland

Brigalow

Noosa Heads Tarong Kingaroy Natural market Energy for Tarong Coral 70 kt Caloundra Sea

Total South East Queensland Ash Market (2001) 270 kt

Dalby Millmerran Power

Natural market for Millmerran 3 kt

Shared market with Tarong 167 kt Shared market with Swanbank 10 kt Warwick 100 km (approx)

Brisbane CS Energy (Swanbank)

Natural market for Swanbank 20 kt Gold Coast

Pacific Ocean

It appears that the anticipated lifespan of Swanbank was uncertain at that time.26 Swanbank coal-fired plant A ceased operating in 2005, towards the end of the period in question, and coal-fired plant B ceased operating in 2010, ending production of fly ash at Swanbank.27 Millmerran power station opened in 2002.28 The legal proceedings concern the implications of competitive tenders for contracts to acquire unprocessed fly ash from the Tarong and Millmerran power stations in 2002. It appears that it is economic to export processed fly ash in bulk by sea to other States in Australia and to New Zealand,29 but presumably would be costly for cement-makers in SE Queensland to secure transport by road from more distant power stations. This situation could be expected to continue to change as coalfired power stations convert to gas-fired generation technology. Not shown on the map is Kogan Creek coal-fired power station (150–200 kilometres to the west of Tarong), the construction of which commenced in 2004 and was completed in late 2007.30 The judgment on liability is 665 pages long and addresses evidence about the purpose and effect of contracts entered into by CA in the period 2001–06 to

26 ibid [696]. 27 ‘Shutdown at Swanbank B’, Queensland Times, 25 May 2012, www.qt.com.au/news/shutdown-atswanbank-b/1392591. 28 Cement Australia [2013] FCA 909, 310 ALR 165 [715]. 29 ibid [2036], evidence of CA. 30 ibid [2852].

184  Raising Rivals’ Costs acquire fly ash from the three power stations in SE Queensland. There was held to be market power, but no ‘taking advantage’, so the section 46 claim failed. However, the case succeeded under the section 45 (which prohibits agreements having the purpose, effect or likely effect of SLC). Thus, we may anticipate the possible application of the Harper Review section 46 based on the same SLC test. In summary, the court held as follows:31 (a) CA giving effect in the period 2001–05 to contracts previously entered into concerning exclusive access to Swanbank fly ash had the effect and continuing likely effect of preventing rivals from securing access to unprocessed fly ash and preventing rivals from entering the market for processed fly ash, and thus of SLC in the markets for unprocessed and processed fly ash. In addition, CA making an agreement to extend those contracts to 30 June 2005 had the purpose and effect or likely effect of SLC in those markets (that date is towards the end of the period in question in the proceedings, by which time the ash from Swanbank had become variable and of doubtful quality).32 (b) Although key executives who negotiated the contract on behalf of CA were not called to give evidence, it can be inferred that entry into the original Millmerran contract in 2002 for exclusive access to Millmerran fly ash had the substantial subjective purpose of excluding rivalry as noted above, and thus of SLC in those markets in SE Queensland.33 (c) Although the quality of Millmerran fly ash was not then known, by excluding rivalry, the immediate effect of CA entering into the contract was to SLC in those markets, and there was a real chance (ie, likely effect) of an ongoing effect of SLC; however, that effect was dissipated by the end of 2003 due to the compromised quality of the Millmerran fly ash.34 (d) entry by CA into the amended Millmerran contract in 2004 had a ­substantial subjective purpose of excluding rivalry as noted above, and thus of SLC in the markets for unprocessed and processed fly ash (but not that effect or likely effect).35 (e) CA’s conduct in 2005 in proceeding to invest capital in further processing facilities at Millmerran did not have the purpose or likely effect of SLC.36 (f) Entry by CA into the Tarong contract in 2003 for access to any and all fly ash from the power station, and subsequently giving effect to it, had the purpose

31 ibid interim orders at 310 ALR 165, 817–19. 32 ibid [3215]–[3231]. 33 ibid [3070]–[3078]. 34 ibid [3079]–[3087]. Note that ‘making’ and ‘giving effect’ to the agreements are separate contraventions of s 45, which is made clear in the final orders. 35 ibid [3237]–[3238]. Note that ‘making’ and ‘giving effect’ to the agreements are separate contraventions of s 45, which is made clear in the final orders. 36 ibid [3239].

Cement Australia: Factual Matrix  185 and the effect or likely effect, by excluding rivalry as noted above, of SLC in the markets for unprocessed and processed fly ash.37 (g) Because the balance of the Tarong contract term was by then relatively short, the subsequent conduct by CA in May 2006 of proceeding to invest capital in further processing facilities at Tarong did not have the purpose or effect of SLC.38 Thus, we see that the predictive SLC effects test apparently asks an abstract question which does not require assessment of an actual effect on competition. We saw in Chapter 3 that the CJEU held in the Intel case that effects should not be inferred from the nature of conduct and that an actual competition analysis is required. However, judicial authority in Australia to date on the predictive SLC test does not require this. The court asks if the contract, when made, would have or be likely to have the effect of SLC.39 This effect can then be inferred from the exclusive nature of the rights conferred by the contract. However, it is clear that even the simplest contract for specific goods excludes rivals buying those goods.40 So a firm securing supplies for itself at one and the same time harms rivals, but that is the nature of zero-sum competition and should not be capable of harming the process of competition. By contrast, when we ask whether the market has been manipulated, eg, by cornering, we want to know whether the party so doing has obtained supplies it does not need or use, as a result of which the price to competitors has thus been forced up and supplies have been withheld from the market, ie, there is unmet demand. The problem encountered in a competitive tender, a classic bidding war situation, is that the consequences may well be the same, regardless of which firm wins. Since it appears that CA supplied cgfa to its competitors (the three largest buyers of cgfa were Hanson, Readymix and Boral, accounting for 79 per cent of the market),41 it does not appear that CA foreclosed supply to them. We will want to know whether it restricted supply to them or charged prices which affected their ability to compete in downstream markets for cementitious materials or concrete products. That would be indicative of a firm with market power in an upstream market seeking to use its power to squeeze rivals’ margins in a downstream market (as in Queensland Wire). Greenwood J held that market power can arise from contracts securing access to an essential input, though it must be assessed whether the power is enduring or transitory at the time that the conduct is alleged to take advantage of market power.42 His Honour found that CA had substantial power in the market in which 37 ibid [3240]–[3241]. While the period in question concerns 2001–6, the final declarations regarding the Tarong contract appear to be open-ended. 38 ibid. Not noted in final orders. 39 ibid [2999] and [3007]. 40 B Donald and JD Heydon, Trade Practices Law (Sydney, Lawbook Co, 1978) 277–78. 41 Cement Australia [2013] FCA 909, 310 ALR 165 [1887]–[1888], annual usage figures prior to 2005. 42 ibid [974].

186  Raising Rivals’ Costs it supplied cgfa, at the beginning of the relevant period and through it, based on the contracts it held with Tarong and Swanbank, which gave it a dominant share of unprocessed fly ash inputs.43 This view was not swayed by the opinion from Professor Hay that such power was constrained by the three largest buyers of cgfa (accounting for 79 per cent of the market) or that the power was transitory since the Tarong contracts were open for renewal, or that there were other firms in the region with capacity to process fly ash.44 The conduct alleged by the ACCC to involve the use of market power contrary to section 46 was the entry into the Millmerran contract in 2002. The ACCC alleged that CA’s acquisition cost under that contract was so high that it would place CA at a cost disadvantage to its rivals, which would make no commercial sense unless CA has substantial market power in the market for cgfa and so can recoup that cost.45 Pausing here, we may observe that harm to oneself is a curious characteristic of strategies to raise rivals’ costs, but is presumed to be explicable by power in a downstream market unrelated to the input itself. We may doubt that the market for cgfa is a relevant downstream market, ie, the harm we anticipate would be that rivals of CA would be less able to compete in markets for cementitious materials or concrete products (which was not alleged in the case). Absent such harm, it would appear that the harm is that CA would derive above-normal profits as a result of successfully winning the tender to acquire unprocessed fly ash. Greenwood J accepted that legitimate business reasons may indicate against use of market power, and considered evidence going into minute detail concerning CA’s internal decision-making process,46 which ultimately determines the question of liability for contravention of section 45. However, the key reasoning accepted by Greenwood J is that a dominant firm does not take advantage of its market power if the same conduct would be rational (ie, profitable) for a firm lacking market power.47 This is the critical point of departure for the Harper Review, which accepts that the same conduct by the dominant firm may harm competition, whereas it might not if engaged in by a firm without market power. The Millmerran contract was won by CA under a competitive tender. Greenwood J found that the competing bid was substantially similar to CA’s bid,48 so it could be concluded that the price would have been rational for a firm lacking market power, and thus CA did not take advantage of its market power. We can now return to the trial judge’s consideration of the business rationale for CA’s decision to bid for the Millmerran contract. Here we again encounter the ‘dual purposes’ problem and can see that ‘barriers to exit’ are ignored when we consider what a firm might do in a hypothetical competitive market. The ACCC’s

43 ibid

[1835], [1892]. [1835]–[1891]. 45 ibid [1894], [1897]. 46 ibid [1894]–[2277]. 47 ibid [2279]–[2294]. 48 ibid [2293]–[2294]. 44 ibid

Cement Australia: Factual Matrix  187 economics expert was of the view that entry into the Millmerran contract was not rational profit-maximising conduct.49 CA’s argument was that in 2002, it could not be sure it would retain the Tarong contract, which was open to public tender, and felt that it should bid for both Tarong and Millmerran, and would break even on Millmerran.50 The trial judge considered that there was no rational basis for CA to think it could export fly ash to Victoria, New South Wales or New Zealand, as its witnesses had asserted it could if it won both contracts.51 The key reasoning which supported the finding that CA’s purpose was to SLC was that the consequence of losing either contract would be a loss of volume and earnings worth about A$6 million per annum, which would drag the business below the corporate group’s internal profitability benchmarks.52 Thus, the reasoning suggests that a business rationale of maintaining the firm’s own sales at one and the same time represents a substantial purpose of preventing a rival from taking sales away, which was held sufficient to contravene section 45. We must ask whether there are any implications here for the process of competition, ie, the contest between CA and the rival bidder is zero-sum, one will win and the other will lose, but is the process of competition harmed by one of them winning? We appear to be getting perilously close to a proposition that harm to a rival is synonymous with harm to the process of competition. Greenwood J touches on the question whether CA would rationally stand by and lose the business, but concludes that the operative factor was preservation of sales and profits it had derived from market dominance.53 We may recall the discussion in Chapter 5 of exit costs in relation to Boral and how barriers to exit may cause an incumbent to incur costs to stay in the market that would not make sense under competitive conditions where there are no barriers to exit. If we analyse the decision in the context of zero-sum competition, we would be unconcerned about firms competing to survive and we might even be unconcerned about firms leaving the market. However, we would ask whether there is unmet demand which an equally efficient firm could profitably satisfy, and we would be concerned if a more efficient firm was forced out or prevented from entering. CA’s own assessment supported the inference that, due to its dominant positon in the acquisition of unprocessed fly ash, its selling price for cgfa was about A$10 per tonne above the competitive level.54 However, raising rivals’ costs theory suggests (which experience confirms) that CA’s costs would be increased because it would have had to compensate the power stations to give up rights to other commercial opportunities. Presumably either bidder would have borne this cost 49 ibid [1898]. 50 ibid [1910], [1919]. 51 ibid [2250]. 52 ibid [2266]–[2275]. 53 ibid [2298]–[2301]. 54 ibid [1788]. R Smith, ‘Market Definition and Substitution Options’ (2014) 22 Competition & Consumer Law Journal 105, 111 discusses the standard view that a monopolist charges prices above the assumed competitive level, where elasticity of demand is positive, and the implications of this for market definition (the ‘Cellophane fallacy’).

188  Raising Rivals’ Costs and charged a price accordingly. A more sophisticated analysis is required to assess whether the cost-raising strategy would be profitable.

A.  Subsequent Events While the period to which the proceedings relate ended in 2006, it seems that there were dynamic forces under way well before that date which continued to change the markets after the end of the period. Subsequent events demonstrate how the competition analysis is affected by changes in the factual matrix. As noted above, in 2003 Hanson acquired 25 per cent of CA and by 2007 had increased this interest to 50 per cent, and by 2009 Holcim had become the owner of Readymix (through Rinker and Cemex), as well as holding a 50 per cent interest in CA, thus formalising constraint through backward integration. That is, downstream competitors which might be harmed by the strategy have mitigated harm by acquiring interests in CA. As we will see, this factor is taken into account by the court for the purposes of deciding on the penalty, but not when assessing the competitive effect. Swanbank’s coal-fired power plant A was decommissioned in 2005 and its remaining coal-fired plant B was decommissioned in 2010 as it converted to gasfired power plants,55 thus diminishing the production of fly ash in the region. In 2008, Independent Fly Ash Brokers, a consortium of independent concrete producers, commissioned an A$8 million fly ash processing facility at Millmerran power station.56 In 2011, following a public tender process, Tarong Energy awarded a contract to CA to take at least 70 per cent of the fly ash from Tarong, conditional on the ACCC formally authorising the conduct under section 88 of the CCA. The ACCC initially proposed to deny authorisation, but, after further submissions, authorisation was granted for a period expiring in 2014 on conditions capping CA’s offtake volume. Tarong Energy supported CA’s application for authorisation and criticised the ACCC’s draft determination in a number of respects, including that ‘the Commission makes no finding as to the relevant market or the current state of competition in that market’.57 The ACCC in its final determination considered that: (a) the relevant markets are (i) the supply and acquisition of unprocessed fly ash from power stations, and (ii) the supply and acquisition of cgfa, in both cases in SE Queensland;58 55 ‘Shutdown at Swanbank B’ (n 27). 56 ‘Fly Ash Classifying & Load out Facility at Millmerran PS’, Coal Ash Matters, December 2008, www.adaa.asn.au/uploads/default/files/adaa-coal_ash_matters_dec_08.pdf. 57 Submission on behalf of Tarong Energy, 5 May 2011, [4.4], available on the ACCC Authorisations Register, registers.accc.gov.au/content/index.phtml/itemId/961019/fromItemId/401858. 58 ACCC Determination A91261, 14 July 2011, [4.23], available on the ACCC Authorisations Register, registers.accc.gov.au/content/index.phtml/itemId/961019/fromItemId/401858.

Cement Australia: Factual Matrix  189 (b) the market for cement is not relevant (though the ACCC observed that the high level of vertical integration may impact on the supply of downstream products);59 (c) CA is the largest acquirer and supplier in both markets, and competes with Sunstate (a joint venture between cement-makers Adelaide Brighton and Boral), Independent Fly Ash Brokers (a joint venture between independent cement-makers) and Nucon (a cement-maker that acquires ash from New South Wales, which has an option to acquire ash from Tarong, should it decide to invest in processing facilities there);60 (d) potential competitors include waste disposal companies that acquire fly ash from power stations;61 (e) fly ash is a bulky, low-value good, which it is not commercially feasible to transport over long distances (Nucrush is an exception because it runs its own fleet of road vehicles and has the ability to backload);62 (f) it should reject contentions by CA and Tarong Energy that the proposed supply contract would not SLC because the contract arose from a public tender, that especially since 2007 CA’s competitors supply 35–40 per cent of the fly ash in SE Queensland, and that the agreement provides for third-party access to fly ash of up to 30 per cent of Tarong’s output (though no other tenderer had committed to take any substantial amount); on the contrary, the ACCC considered that the proposed supply contract heightens barriers to entry and forecloses competition for fly ash and thus is likely to SLC with the likely result being ‘prices that are above the competitive level for fly ash concrete products’;63 (g) however, with modifications to cap the quantity of fly ash that CA may take, there are sufficient public benefits to authorise the agreement until 1 March 2014.64 While the ACCC determination does not address the state of competition in downstream markets for concrete products, this seems to be the harm which is considered to flow from the SLC, as discussion of vertical integration would not be relevant unless the ACCC were concerned about effects in the downstream market for concrete products. The theory of harm and analysis of downstream markets and competition effects is thus not as well articulated, nor is the factual matrix as well developed, as we might like.

59 ibid [4.18], [4.22]. 60 ibid [4.7], [4.26]. 61 ibid [4.28]. 62 ibid [4.21]. 63 ibid [4.163] and [4.173]–[4.187] (the reference to concrete products appears to be a deliberate choice of words to distinguish the reference from cgfa). Tarong Energy (n 57) submitted to the contrary that there would be no public detriment: [4.5]–[4.6], [4.14]–[4.16], [4.21], [4.29]–[4.30]. 64 ACCC (n 58) [4.195]–[4.199].

190  Raising Rivals’ Costs A final point to note about the determination is that the ACCC rejected the submission by CA that the exclusive agreement it sought would generate costefficiencies.65 On the contrary, the ACCC considered that the conditions it imposed under the authorisation ‘create the potential for cost efficiency benefits from the greater use of fly ash compared to the position under the Agreement in its original form’ proposed by CA.66 To test this, we would want to know if there is unmet demand that could profitably be supplied, and whether the demand response to increased output would increase or decrease revenue in the market. We would also want to know whether the additional resources required to increase output, including additional fixed capital investment, could profitably be deployed. If the result of increased output would be merely to redistribute surplus from producers of cgfa to consumers of it, at the cost of drawing resources from other productive uses in the economy, we might well conclude that the decision to decrease concentration in the industry producing fly ash could be inefficient. CA’s contract expired in 2014. We might speculate that Tarong Energy did not wish to go through another contested authorisation application. Be that as it may, Tarong Energy awarded the contract to Coal Reuse Pty Ltd for a period of 10 years, as a result of which CA was due to remove its processing equipment from the Tarong power stations so that Coal Reuse could install its own processing equipment.67 That year, Coal Reuse lodged notification of exclusive dealing with the ACCC under section 93 of the CCA, ie, that it was acquiring fly ash from Tarong Energy on condition that Tarong Energy would not supply fly ash to any other person. Such a notification confers immunity unless or until the ACCC determines that the conduct would SLC with no offsetting benefit to the public and formally revokes the notification. In 2015, the ACCC decided not to revoke the notification (so immunity conferred by virtue of the notification continued).68 The ACCC’s statement of reasons provided in November 2015 indicated that: (a) the geographic market may now be broader than SE Queensland, since Kogan Creek power station commenced and it is now thought feasible to transport fly ash from north Queensland and overseas; however, ‘the ACCC considers that variations in the precise definitions of these markets will not alter the outcome of its assessment’;69 65 ibid [4.123]–[4.130]. 66 ibid [4.130]. 67 ‘Contractor Begins 10-Year Fly Ash Deal’, South Burnett News, 1 August 2014, southburnett.com. au/news2/2014/08/contractor-begins-10-year-fly-ash-deal. 68 ACCC Notifications Register, Coal Reuse Pty Limited – Notification – N97609, registers.accc.gov. au/content/index.phtml/itemId/1180364/fromItemId/1133393. Notification is a process available for conduct which could contravene s 47 of the CCA, subject to the competition test. In Cement Australia, the ACCC case was based on s 45, or in the alternative s 47. The Court found that the s 47 case was not established, which made no difference to the outcome (except that s 45 conduct cannot be ‘notified’ to gain immunity under s 93 of the CCA). 69 ACCC Notifications Register, Statement of Reasons, 26 November 2015, [5.5]–[5.7] Coal Reuse Pty Limited – Notification – N97609, registers.accc.gov.au/content/index.phtml/itemId/1180364/ fromItemId/1133393.

Cement Australia: Factual Matrix  191 (b) absent the proposed contract, it is most likely that Tarong would grant a contract to a single operator;70 thus, with or without the Coal Reuse contract, it is likely that there would be a single operator taking fly ash;71 to this extent, there is no change in market power, which pre-contract resides with Tarong and is exercised by it in the terms negotiated with Coal Reuse, or post-contract would reside with Coal Reuse; notwithstanding this, there are supplies of fly ash in SE Queensland ‘and the surrounding area’ that will ‘somewhat constrain’ Coal Reuse;72 (c) submissions by CA seeking revocation of the notification, on the grounds that the contract will give Coal Reuse substantial market power and foreclose access to other users, are rejected; the ACCC considered Coal Reuse’s exclusivity is very different from the exclusivity previously enjoyed by other users, because Coal Reuse is simply a reseller, having no use itself for cgfa, and has no incentive to withhold supplies from downstream users;73 (d) the purpose of Coal Reuse in entering into the contract is to maximise sales of fly ash and return on its investment;74 (e) for the reasons above, the ACCC was ‘not currently satisfied’ that the notified conduct has the purpose, effect or likely effect of SLC in any market.75 We can immediately make some observations on the ACCC’s reasoning, and the factual matrix that underpins it, which may be contrasted with the theory of harm that the ACCC pursued in the proceedings against CA. First, Kogan Creek power station commenced in 2007, which was anticipated in the period with which the proceedings were concerned, since construction began in 2004. Second, Greenwood J treated the geographical extent of the market as significant in the assessment of competition, market power and foreclosure in the relevant markets, rejecting the contentions of CA about the feasibility of transporting fly ash over long distances. Third, since it appears that CA supplied competitors in the downstream markets, there appears to have been no greater risk of foreclosure than in the case of Coal Reuse. Fourth, neither in the case run by the ACCC in the proceedings nor in the subsequent CA authorisation matter did the ACCC expressly argue that harm was likely in downstream markets or that vertical integration was likely to cause vertical foreclosure. Fifth, we may infer that it is permissible for Coal Reuse and CA alike to derive some above-normal profits as a result of their success in winning the tender, but it does not appear that the ACCC considers this alone to give rise to any harm, because all downstream competitors bear this cost. However, we have no analysis of downstream markets to assess whether the consequence



70 ibid

[5.13]–[5.15]. [5.20]. 72 ibid [5.23]–[5.24]. 73 ibid [5.26]–[5.27]. 74 ibid [5.20]. 75 ibid [5.19]. 71 ibid

192  Raising Rivals’ Costs in both cases simply involves a redistribution of surplus between upstream and downstream suppliers, or whether it would cause any transfer of surplus from end consumers of concrete products or would impact on efficiency in the production of concrete products. In 2016, Coal Reuse went into liquidation, due to unpaid debts, without having completed the installation of fly ash processing equipment at Tarong, amid accusations that it had disposed of fly ash contrary to environmental regulations.76

B.  Assessment of Harm and Penalties Earlier in 2016, Greenwood J delivered judgment in the Cement Australia case with respect to penalties.77 The judgment runs to 120 pages, addressing considerable evidence and expert opinion on the question of harm arising from CA entering into the exclusive contracts to take fly ash from Millmerran and Tarong as a result of competitive tenders in 2002. We are reminded that the predictive SLC test requires an inference to be made, at the time of entering into the contracts, about effects which are likely in the sense that there is a ‘real chance’ they may eventuate. This does not require an assessment of actual effects, though (seemingly) ‘real chance’ can be ruled out if market circumstances change, eg, the unforeseen problems with the quality of fly ash from Millmerran. According to the prior case law discussed in Chapter 4, hindsight is not relevant to the assessment of likely effect at the date of entering into the contract, but clearly actual effects can be assessed with knowledge of what happened subsequently. Greenwood J describes the assessment of competition effects as ‘qualitative’.78 This, however, rather suggests a qualitative assessment of relevant matters proven by evidence in the usual way, such as matters of market structure, conduct and performance suggested by the SCP approach. I suggest that the actual process of reasoning is more like abstract reasoning based on a view of the world in which competitors do not engage in bidding wars and do not compete for survival. We can draw some conclusions from the case regarding the permitted behaviour of a firm in CA’s position in 2002, but before doing so, we should consider the theories of harm presented to the court on the question of penalty. It is troubling that proof of harm is not required to demonstrate ‘effects’. The case probably demonstrates that this is simply not possible, which is also troubling. Greenwood J traces the causes of this dilemma to the essential fact that the Tarong and Millmerran contracts were won by competitive tender: if CA had not

76 ‘Fly Ash Firm in Liquidation’, South Burnett News, 6 September 2016, southburnett.com.au/ news2/2016/09/fly-ash-firm-in-liquidation; ‘Coal Reuse, Queensland Government c­orporation contractor, facing court claims over unpaid debts’, ABC News, 6 June 2016, www.abc.net.au/ news/2016-06-06/queensland-government-corporation-contractor-facing-court-claim/7479214. 77 CA Penalties (n 15). 78 ibid [818(3)].

Cement Australia: Factual Matrix  193 won them, the competing tenderer would have won them on similar terms as to ­exclusivity.79 I suggest that, with knowledge of the ACCC reasoning in the subsequent events, the harm comes down to a comparison of the downstream effect of above-normal profits that a non-integrated firm would enjoy having cornered the market, and the effect of CA being a vertically integrated firm deriving the same profits. This, however, is not how the ACCC presented the case. Greenwood J pragmatically observes that ‘the extent of that damage inflicted upon future competition or market harm might be measurable (that is, capable of quantification) or it might not’.80 This leads his Honour to consider detailed opinion of the expert economists about how one might assess the impact on market price, ultimately concluding that it is not possible to quantify it and, accordingly, that the best the court can do is make a qualitative assessment.81 However, without any authority to guide the question, Greenwood J pragmatically decides that sales by CA to downstream members of the corporate group should be excluded when assessing market harm.82 Here we are beginning to grapple with the conundrum that ‘raising rivals’ costs’ harms the dominant firm itself and, I suggest, we must therefore follow the effect downstream to assess harm. That is something the ACCC did not do or urge the court to do.

C.  Institutional Implications The events surrounding the Cement Australia case also illustrate an issue we addressed in Chapter 2: the institutional division of functions between the courts and the ACCC/ACT. The ACCC, at the time of the CA’s authorisation to renew the Tarong contract and the Coal Reuse notification, was a party to legal proceedings, prosecuting a case against CA awaiting adjudication. We may question whether the ACCC’s dual roles as a prosecutor and an administrative decision-maker place it in a difficult position when it seeks to give effect administratively to views that are being contested in legal proceedings (especially if all facts and issues on which it proceeds administratively are not raised in the legal proceedings). It may well be that a separation of those functions would provide better assurance that views on important matters have been or will be independently tested. Administrative review by the ACT may not be a timely way to resolve differences of opinion. The solution may require an independent prosecutor and/or immediate reference of administrative functions to the ACT where the ACCC is conflicted.



79 ibid

[477]–[485]. [485]. 81 ibid [641]–[644], [818]. 82 ibid [818(15)–(18)]. 80 ibid

194  Raising Rivals’ Costs

D.  The Implications of Bidding Wars We can ask what guidance firms may gain from the case and subsequent events regarding permitted behaviour in bidding wars for inputs. It seems that the ACCC raises no issue with a non-integrated firm winning a competitive tender and thereby gaining exclusivity over (foreclosing) a substantial share of the market. It would have been easy for the ACCC in the case of the Coal Reuse notification to admit that the geographical boundaries of the market have broadened and the significance of tying up ash from Tarong was greatly reduced, but the ACCC maintained that this was not influential. Therefore, we conclude that it was the nonintegrated nature of the winning bidder that removed the concern about market harm. On the other hand, what behaviour is expected of a vertically integrated firm? While some experts believe that the same conduct by a vertically integrated firm is not per se harmful, but must be empirically assessed, we learn from the case that this assessment cannot in practice be undertaken by our experts and courts. Thus, a vertically integrated firm risks sanctions if it wins a bidding war, which is a consequence that may well chill competition. We recall that the rival bidder was also vertically integrated, so the same reasoning would apply had it won the bidding.83 We have seen that bidding by contested tender for inputs raises the costs of the inputs. Raising rivals’ costs theory suggests that the strategy raises the dominant firm’s own costs. We may wonder if this situation is analogous, but opposite, to that of a price war such as that considered in Chapter 5. Roberto Burguet and József Sákovics model the effects of competitive bidding for inputs where there is abundant supply, but the supplier cannot sensibly deal with more than one buyer, and argue that higher input prices may mitigate downstream market power.84 I suggest that we can adapt the central intuition of their very specific and complex model to our purposes, ie, to suggest that where the alleged dominant firm does not have market power in the downstream market, the effect of a bidding war for an input is to raise input prices, thereby transferring surplus upstream. While this does not appear to benefit consumers in the downstream market, we may conclude that there would at least be a transfer of surplus away from the alleged dominant firm. Tarong Energy submitted to the ACCC in the authorisation matter mentioned above that the ACCC had not given sufficient weight to the fact that the contracts were awarded following a public tender process.85 The ACCC acknowledged that

83 ibid [189(17)], [262]–[263]. The rival bid occurred prior to the merger creating CA. The rival bid consortium comprised Hanson (Pioneer), Rinker (Readymix) and Boral. Post-merger, all these parties except Boral acquired interests in CA. 84 R Burguet and J Sákovics, ‘Competitive Foreclosure’, research paper, School of Economics, University of Edinburgh, 15 February 2017, www.ed.ac.uk/files/atoms/files/279_-_competitive_ foreclosure_2017.pdf. 85 Tarong Energy (n 57) [6.2].

Gaps in the Factual Matrix  195 ‘a competitive tender can reduce concerns about the competitive effects of vertical supply arrangements’, but considered that ‘a firm with market power may be willing to bid more in a competitive tender to protect its market power’.86 However, by the time of the subsequent Coal Reuse notification matter, the ACCC observed that market power during the bidding resided with Tarong Energy and thereafter with the successful bidder.87 I suggest that we need to address the point that by bidding up the price, surplus is transferred away from the winner of the tender (and that there must be some market power independent of the contracts awarded if harm is to result). Accordingly, we can conclude that competitive bidding may well raise the price of inputs above the level needed to meet demand and that this is a consequence of zero-sum competition, which we may be able to tolerate because it transfers surplus away from the alleged dominant firm.

III.  Gaps in the Factual Matrix From the Cement Australia case and the subsequent events, we again see that choices made by a plaintiff as to how it runs its case may well affect the outcome and, for our purposes, may create gaps in the factual matrix we would need to consider other approaches to the problem of assessing market power, competition and efficiency effects. Despite the vast amount of evidentiary detail considered by the trial judge for the purposes both of determining liability and imposing a penalty, there are a number of gaps in the factual matrix that raise important questions. First, we presume that the ACCC made a strategic decision not to raise the subject of market power in the downstream markets for cementitious products (in which cgfa and cement appear to be both substitutes and complements) or the market for concrete products. We do not have data to assess those markets, but might be tempted to assume that the presence of powerful competitors in those markets indicates against there being market power which CA might need to obtain a competitive advantage from cornering the market in SE Queensland for unprocessed fly ash. Second, absent downstream market power, we may doubt that vertical integration is relevant, though it appears to concern the ACCC in the authorisation case and to distinguish the ACCC’s approach to the notification case. Absent these issues being raised in Cement Australia, we could conclude that the issue solely concerns market power from cornering the market for unprocessed fly ash. In this case, we ask whether the monopolist is meeting demand or whether there is unmet demand, and whether it is efficient to introduce competitors to that market.

86 ACCC 87 ACCC

(n 58) [4.175]. (n 69) [5.23].

196  Raising Rivals’ Costs Third, we may well be concerned about the lack of data concerning the above matters or that significant analysis of competitive harm occurs only after the determination of liability, ie, as a matter relevant to penalty. As we have seen, the predictive SLC test can be based on inferences rather than data. We may recall from Chapter 2 that this is regarded as a strength of the structuralist approach, which acknowledges that actual economic effects are difficult if not impossible to determine. We are now in a position to assess whether we should be troubled by this and whether the market manipulation approach may address our concerns.

IV.  Cement Australia under Harper Section 46 On the face of it, the Harper Review section 46 should reach the same outcome as Greenwood J reached applying section 45 to the Millmerran contract. Clearly, Greenwood J was satisfied that exclusive supply contracts can create market power, even (seemingly) at the stage of a tender, before the contract becomes enforceable, if the party can be confident it will be awarded.88 This is slightly troubling, but we will proceed on the assumption that an exclusive contract over inputs, such as it had with Tarong and Swanbank, would give CA power in the market for cgfa. The issue we then have to determine is whether the effect of entering into an exclusive contract with Millmerran, by public tender at a competitive price to compensate Millmerran for exclusivity, would be to SLC. The Harper Review amendment appears to accept the proposition that we should neutralise the argument that market power is not implicated when a firm lacking market power could have engaged in the same conduct. However, the proposition accepted by the Review is that the same conduct by a firm lacking market power might not have the effect of SLC, whereas that conduct by a firm having market power may well have the effect of SLC. These propositions are not the same: (a) I suggest that the Harper section 46 should still require a causal connection between the effect and the dominant firm’s market power, which remains absent in the Cement Australia case (the cause of foreclosure is entering into the contract, not market power); (b) in the Cement Australia case, I suggest that if a firm lacking market power enters into an exclusive contract with Millmerran, the effect on competition could well amount to SLC and needs to be assessed (certainly we would expect this in the case of Coal Reuse entering into a contract for 10-year exclusivity over Tarong).



88 Cement

Australia [2013] FCA 909, 310 ALR 165 [2271].

Cement Australia under Harper Section 46  197 If the argument for the new entrant lacking market power is simply that exclusivity over Millmerran forecloses a trivial share of the market, this should be good for the dominant firm as well. The Harper Review wants us to take efficiency into account, though the Review does not limit this enquiry by defining what it means by efficiency or how it may be forensically assessed. For the reasons discussed in Chapter 4, I suggest that this cannot be benchmarked against the presumed ‘competitive level’. Thus, we would ask whether the conduct in question is efficient in the sense in which it can be adapted and redefined for our purposes. I suggest that this requires consideration of investment in facilities (given that power stations do not want multiple fly ash processing facilities at a power station,89 which may well be inefficient­ duplication) and whether there is unmet demand that could profitably be supplied by new entry. These questions require forensic determination. It is too simplistic to say that foreclosure necessarily harms the process of competition when market size may well foreclose new entry without any effect on competition. It seems clear that we cannot make reliable inferences about harm to competition on the basis of price and profit margin. The trial judge in Cement Australia considered that conduct directed toward protecting sales volume where prices and profits are above the ‘competitive level’ could be regarded as having the purpose and effect of SLC.90 When the case turned to consider the harm resulting from the conduct the trial judge made a pertinent observation: [S]ince the proposition is that prices for [SE Queensland cgfa] were higher than they would be in a competitive market but for the identified provisions of the contracts … it is important to keep in mind the contribution of the relevant provisions to the contended higher prices because, for example, higher prices due to an exercise of market power per se in the commercial sense (that is, absent a contravention of s 46 of the Act), are not necessarily a consequence of the s 45 contraventions.91

Thus, the problem that remains with the Harper Review section 46 is one of causation, particularly where market power is said to arise from entering into contracts. Cement Australia arguably demonstrates that we do not need section 46 in this situation. However, the case does show that the analysis of ‘competition effects’ for the purposes of determining liability remains a problematic abstract question. As we noted in Chapter 2, it may well be that competition law cannot function if competition effects have to be forensically proven. Be that as it may, I suggest that price and profit alone cannot indicate harm: we expect that price in any model other than perfect competition will be above marginal cost because supply responds to marginal revenue, not price; thus, the profit margin is a function of elasticity of demand; and in order to consider efficiency, we need to know whether there is unmet demand which can profitably be supplied and

89 ibid. 90 ibid 91 ibid

[2266]–[2271]. [752]–[756] Millmerran; [1560] Swanbank.

198  Raising Rivals’ Costs how demand responds to price and output decisions. That is, as we saw in ­Chapter 1, above-normal profits do not necessarily indicate that market power has been misused. It remains to be seen whether the Harper Review section 46 will be interpreted as a ‘no fault’ or strict liability provision, ie, whether causation will be required between market power and effect. Similarly, it remains to be seen whether the Harper recommendation that efficiency should be considered will have any role after Parliament removed that requirement. As we have seen, on the facts of the Cement Australia case, these issues leave many problems that remain to be addressed.

V.  Cement Australia under Market Manipulation This approach solves many problems to do with the traditional concept of market power, eg, whether it may exist before or after contracts are entered into. Price and output decisions may be manipulative. Any firm can engage in manipulative conduct. We may suspect that a firm is withholding output to force prices up, and that consumers would be better off if new entrants expand output. However, ‘abovenormal’ prices are caused, or contributed to, by elasticity of demand. We need to understand how demand responds in order to consider whether the outcome is consistent with the operation of undistorted forces of supply and demand (ie, an efficient market) or whether the market has been manipulated. It may be inefficient to throw resources into the market to expand output. It may be that there is no unmet demand and that the incumbent has successfully managed technical and business risks, eg, the problems encountered by Cement Australia with variable quality unprocessed fly ash, and the investment risks, environmental issues and solvency problems encountered by Coal Reuse. Perhaps profits that appear to be above what might hypothetically be expected under a competitive market standard may be well earned in the real world. Perhaps the market mechanism, which uses price to ration scarce resources, is ill-suited to deal with a waste product which appears to be in abundance, but is an environmental burden rather than a free gift of nature. We see that zero-sum competition may be inefficient when firms engage in a bidding war for inputs, but we see that to intervene in the market to select which firm may succeed risks equating ‘harm to competitors’ with ‘harm to competition’. Thus, we may chill competition by intervening.

VI. Conclusion We saw that the theory of raising rivals’ costs comes with some warnings which would suggest that the case of Cement Australia was not going to be straightforward. The case would have been similar to margin squeeze if CA had substantial

Conclusion  199 power in the concrete products market. I suspect that the ACCC may well have believed this, on account of vertical integration, but the case was not run on that basis, and we could infer that there was likely to be sufficient competitive constraint in the concrete products market. The predictive SLC test is not necessarily an economic effects test, but may be seen as an abstract test requiring only broad structuralist-informed inferences that foreclosure of competition and vertical integration are per se harmful, regardless of efficiency considerations. We can only speculate whether, in the context of the new section 46, Australian courts may tend towards the EU approach, which requires an actual competition analysis rather than inferring effects from the nature of the conduct. Absent a more reliable measure of ‘competition effects’, I suggest we should go straight to considering efficiency, as we can adapt the concept for our purposes, and ask if the market has been manipulated. This would enable us to place big business and small business on an even footing, as all firms to some degree have the power to manipulate the market. We have seen in this chapter that the ACCC’s dual roles as a prosecutor and an administrative decision-maker create a conflict of interest that might be better addressed by the separation of those functions. A decision-maker which has to justify its views only to itself lacks the rigour that external review provides. It is true that judges are not expert economists and that in Australia, there is a lack of depth in terms of precedent and critical literature on law and economics that other jurisdictions enjoy; however, judicial training enables judges to distinguish good arguments from poor arguments. In the next chapter, we examine a case where the section 46 claim succeeded: Baxter bundled supply of products in which it had no effective competitors with products where it did not have a competitive advantage. The court did not appear to consider whether there were efficiencies in Baxter supplying all products (ie, economies of scope) in a bundled deal. The case is interesting because, as we have seen, Queensland Wire could be viewed as a bundling case, where the remedy was to give the competitor access to the dominant firm’s competitive advantage. We will want to see if the Harper Review section 46 deprives a dominant firm of its competitive advantage or, as conventional wisdom would have it, allows the firm to profit from its success.

8 Bundling I. Introduction We may suspect that when a firm with market power over a product bundles it with other products in which the firm does not have market power, the result may be to extend its market power into those other product markets with resulting harm. According to Motta, the outcome may not necessarily be harmful, eg, consumers may benefit from reduced prices for the bundled products, and the effect must be analysed having regard to the particular circumstances including efficiency benefits.1 Niels et al suggest that ‘even if competitors are harmed … this does not necessarily lead to a conclusion that consumers will suffer’.2 The Antitrust Modernisation Commission (AMC) proposes a simple approach, analogous to that for predatory pricing, to assess potential harm from bundling: (a) ascertain the amount by which the bundled price is lower than the aggregate price for the products in the bundle if supplied separately; (b) allocate the price reduction to the product in which the firm lacks market power (since we can conclude that the firm would derive above normal profits on the product over which it has market power); (c) ask if the effective price for the product in which the firm lacks market power is below avoidable cost and, if so, whether the firm is likely to be able to recoup those losses.3 Niels et al would add to this an enquiry, if effective pricing is below avoidable cost, that we should ask whether cost efficiencies generated by bundling the ­products justify the price reduction.4 We saw in Chapter 5 that analysis of a­ voidable costs

1 M Motta, Competition Policy: Theory and Practice (Cambridge, Cambridge University Press, 2004) 461–68. 2 G Niels, H Jenkins and J Kavanagh, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011) 257–58. 3 Antitrust Modernisation Commission, Report and Recommendations (2007) 12. 4 Niels, Jenkins and Kavanagh (n 2) 260. Pfizer [2018] FCAFC 78 [370] involved bundling, but no issue of below-cost pricing was pursued on appeal to the Full Court. The case is of interest to us in this chapter because: the trial judge found that below-cost pricing was a strategy of manufacturers of generic pharmaceutical products during the launch stage of products for a relatively short period, so that similar conduct by the dominant firm would be a normal incident of competition at the time that its patent expires (noted by the Full Court at [241]); the Full Court at [559] analysed what I have called

Introduction  201 may require an assessment of the incumbent’s exit costs, which will become relevant under zero-sum competition when we seek to understand why an ­ ­incumbent might not simply shut down or exit the market when faced with an apparently more efficient new entrant. Besanko at al suggest that the AMC’s three-step approach will indicate whether the purpose of bundling is to deter entry into the product market in which the firm lacks market power.5 However, due to the ‘dual purposes’ problem in zero-sum competition, we know that purpose alone is not a reliable indicator of harm and we must consider economic effects including efficiency. Clearly, careful attention must be given to the economies of scale and scope which might apply to the bundled products. Further, I suggest that it is not obvious why we should effectively compel the firm always to derive above-normal profits in the market in which it has market power, ie, the approach may condemn a strategy which is profitable across the bundle and might otherwise benefit consumers. In Baxter First Instance,6 the ACCC’s bundling case under sections 46 and 47 was dismissed by Allsop J on the basis that Baxter, as a supplier of products to government agencies, was entitled to derivative crown immunity, ie, not subject to the CCA. This aspect of the case was dismissed by the High Court in ACCC v Baxter Healthcare Pty Ltd,7 so we need not concern ourselves with it. The case was then remitted to the Full Court of the Federal Court to resolve appeals regarding findings of the trial judge, in the alternative that immunity did not apply. The Full Court in Baxter Full Court8 by a majority (Dowsett J dissenting) held that Baxter had taken advantage of its market power for an exclusionary purpose in contravention of section 46 and also had made and given effect to an agreement having the purpose, effect or likely effect of SLC in contravention of section 47 (exclusive dealing). Thus, we would expect the same result under the Harper Review section 46 as enacted. Indeed, as we have seen in our discussion of the Cement Australia case in Chapter 7, there is no real need for section 46 where conduct concerns bilateral agreements which can be assessed under the competition test in section 45 or 47. We can, however, examine how considerations of efficiency may affect the outcome of the case under the form of section 46 recommended by the Review which mandates such consideration, or under the market manipulation approach proposed in this study, which focuses on such consideration.

the ‘dual purposes’ problem differently from the Full Court constituted in the Baxter case, which is considered in this chapter. As the Pfizer case did not raise any issue of anticompetitive effect, and the ‘purpose’ case under both ss 46 and 47 failed, we would expect the same result under the Harper Review s 46, ie, the ACCC case would fail. 5 D Besanko, D Dranove, M Shanley and S Schaefer, Economics of Strategy, 7th edn (Hoboken, NJ, Wiley, 2016) 206. 6 ACCC v Baxter Healthcare Pty Ltd [2005] FCA 581 (hereinafter ‘Baxter First Instance’). 7 ACCC v Baxter Healthcare Pty Ltd (2007) 232 CLR 1. 8 ACCC v Baxter Healthcare Pty Ltd [2008] FCAFC 141 (hereinafter ‘Baxter Full Court’).

202  Bundling

II.  Baxter: Factual Matrix Baxter, Gambro and Fresenius are global corporations which manufacture and supply certain sterile fluids for clinical use.9 Some other global corporations ceased manufacturing sterile fluids in Australia in the 1990s due to the small size of the market and the capital costs of manufacturing here.10 The conduct in question concerned contracts in the period from 1998 to 2003 under which Baxter supplied a range of sterile fluids to State government purchasing authorities (called SPAs), mainly by public tenders.11 The ACCC’s case distinguished between peritoneal dialysis (PD) fluids and a range of other sterile fluids (SF) for various clinical uses.12 The latter are bulky water-based products described as ‘high volume low value’ products in respect of which Baxter has market power by virtue of having the only manufacturing facility in Australia.13 At the same facility, Baxter also manufactured PD fluids, described as ‘low volume high value’ products.14 The conduct concerned bundling ‘monopoly’ SF products with PD products at a bundled price significantly below unbundled prices for SF products only, thereby effectively forcing SPAs to accept the bundled offer and so eliminating competition from rival suppliers of PD products.15 Baxter spent millions of dollars over about 20 years up to the mid-1990s developing its Australian manufacturing facility at Toongabbie in New South Wales.16 In the mid-1990s, Baxter launched ‘major PD products in Australia which had been developed in Europe and the US’.17 Different SF and PD products are produced in batches, sharing the same facilities, which are cleaned and flushed between batches.18 From its Australian facility, Baxter manufactures PD products not only for Australia but also New Zealand and the Pacific Islands.19 The ACCC claimed that Baxter’s local manufacturing gave it an advantage (effectively a monopoly) in SF, but did not give at as significant an advantage for PD products, which were open to import competition.20 Gambro could have produced PD products from its manufacturing facility in Australia, but chose not to do so, or to import PD products, while it was unable to gain a relevant market share.21 The impediments to other manufacturers entering 9 Baxter First Instance (n 6) [6]–[7], [79]–[93]. 10 ibid [94]–[105]. 11 ibid [8]–[9]. 12 ibid [36]. 13 ibid [38]. 14 ibid. 15 ibid [9]. 16 ibid [107]–[122]. 17 ibid [110]. 18 ibid [112]–[119]. 19 ibid [394]. 20 ibid [38]; there appeared to be some degree (albeit low) of potential import competition for SF products (at [484]). 21 ibid [636].

Baxter: Factual Matrix  203 the market are the size of the market, the risk of large sunk costs and Baxter’s local production facility.22 The contracts won or lost at tender are a factor, but do not raise barriers to entry.23 Baxter has a very high market share in SF products in Australia and a corresponding market share for PD products.24 Baxter also has a high market share worldwide in PD products, as shown in Figure 8.1 below.25 We may find this surprising in a case in which it is alleged that market power in SF products is being used to leverage its position in PD products, where seemingly it lacks market power. Figure 8.1  Market shares in PD products Region

Baxter

Gambro

Fresenius

77%

2%

16%

Worldwide Asia (excluding Japan) Australia

77%

6%

-

~90%

~2%

5%

There were held to be separate Australia-wide markets for each kind of SF and, ultimately, the constraining power of the SPAs was held to be insufficient to negate Baxter’s having substantial market power in SF products.26 Nevertheless, we may want to consider what influence public tenders called by SPAs have on competition between firms, in particular the extent to which bidding practices respond to buyer demand. There is much discussion in the first instance judgment about the ­difference between peritoneal dialysis and haemodialysis (HD). These are to some extent alternative treatments, though each carries risks and benefits that must be assessed by clinicians in the circumstances of particular patients.27 The market position of the three companies in HD products is set out below.28 Figure 8.2  Market shares in HD products Region Worldwide Australia

Baxter

Gambro

Fresenius

7%

19%

~30%

~9%

33%

58%

How the situation regarding HD products affects the findings in the case is not clear. The data may suggest that Gambro and Fresenius have market power in

22 ibid

[639]. [639]–[640]. 24 ibid [126]–[128], [131]. 25 ibid [130]–[132]; selected figures only reproduced in Figure 8.2. 26 ibid [515], [518]. 27 ibid [23]–[25]. 28 ibid [81], [130]–[132]. 23 ibid

204  Bundling products that compete to some degree with Baxter’s PD products, thus constraining Baxter’s pricing of PD products. However, it seems that in economic terms, the substitutability between PD and HD products was thought to be low.29 The theory advanced by the experts in the case to assess harm appears consistent with the literature cited above.30 However, differences between the experts as to cost-price calculations led the trial judge to remark that ‘unfortunately, accuracy usually necessitates complexity … which may mean that the results are either incalculable or incomprehensible’.31 The ACCC’s expert calculated that effective prices for PD products were below avoidable cost in all States, whereas Baxter’s expert argued that this was the case in only one or two States.32 Accordingly, the trial judge decided to disregard the ACCC’s expert data for the purposes of his finding with regard to ‘taking advantage’ of market power and to use them as a ‘guide or framework’ for a finding with regard to anticompetitive ‘effect’ under section 47.33 I will consider below whether the tests under sections 46 and 47 are at cross-purposes, and whether this may have contributed to this dilemma. Proceeding in the way he indicated, the trial judge concluded that Baxter had not taken advantage of its market power,34 though he concluded that its substantial purpose was to ‘foreclose the likelihood or restrict the possibility of a competitor’s bid having any realistic possibility of success’.35 The exception was in relation to South Australia, where Baxter refused to offer unbundled prices for SF products (so the bundled offer including PD products was the only option open to SPAs). The trial judge held that Baxter’s conduct in South Australia did in fact take advantage of market power.36 Ultimately, therefore, it is hard to assess what weight, if any, was given to the issues of recoupment and efficiency. The ACCC’s experts did not suggest that recoupment of Baxter’s profit sacrifice on PD products was likely, which Baxter’s expert suggested was inconsistent with the possibility of market power being extended from the market for SF products to the market for PD products.37 The profit sacrifice might be justified by cost-saving efficiencies,38 but it does not appear that Baxter was able to articulate any efficiencies. The trial judge considered evidence put forward by Baxter about the cost savings that would have been achieved if it had lost the PD business, but concluded that such savings ‘do not explain nor were they put forward as explaining any significant part of the large difference’ between the unbundled prices for SF and the bundled offer.39

29 ibid

[499]. [500]–[504]. 31 ibid [513]. 32 ibid [537]–[538]. 33 ibid [545]–[546]. 34 ibid [598]. 35 ibid [616]. 36 ibid [582], [596]. 37 ibid [510]. 38 ibid [515]. 39 ibid [392]. The high prices were derived from the low-volume private hospital price list (at [477]). 30 ibid

Baxter: Factual Matrix  205 We can see that Baxter may have been seeking to explain the volume-related benefits it derived from bundling. It might have been better to address this directly through evidence of volume-related efficiencies of scale and scope. Considering the purpose of Baxter in relation to section 46, the trial judge rejected the evidence of Baxter’s key witnesses, who denied that the bundling strategy was intended to defeat Fresenius and Gambro in the PD market and claimed that the purpose was to maintain volume.40 It was found that ‘Baxter’s high market share and profit in the PD market was under threat’.41 It was found that Baxter’s purpose ‘was to bid in such a way as would prevent rival bids in the PD market being “competitive”, that is, likely to succeed over Baxter’s bid’.42 This was held to be a substantial subjective purpose falling within section 46(1)(c), ie, an exclusionary purpose.43 However, the trial judge agreed with Baxter that at a more general level, the purpose was not directed to the ‘competitive process’ by which Fresenius or Gambro bid at tenders or otherwise offer their products for sale in Australia.44 By contrast, the court at first instance and on appeal in Pfizer accepted the evidence of Pfizer’s executives that they expected sales of their original pharmaceutical product to be ‘slaughtered’ once it came off patent, and generics manufacturers entered the market, and that their substantial subjective purpose was to compete effectively for sales after that time.45 While that decision turned on the facts of the case, we can discern a recognition by the court in Pfizer of the dual purposes problem in zero-sum competition. We will discuss the case further below. Considering the purpose of Baxter in relation to the section 47 SLC test, the trial judge held that the purpose of Baxter’s bidding strategy was to substantially hinder the tender process and thus to SLC.46 However, the trial judge indicated that, if that conclusion were wrong, ie, the process of competition was broader than the tender process, he would hold that Baxter’s purpose to win the tender was not a proscribed purpose.47 When considering effect and likely effect, His Honour seems to have adopted the broader view that the competitive process is the broader framework in which the firms remain rivals after the contract period.48 Ultimately it appears that the trial judge would have held Baxter liable for contravening section 47 in four States.49 The ACCC challenged those findings on appeal to the Full Court, arguing for the wider characterisation of the competitive process, which it said had been affected.50 Baxter also appealed, arguing that the tender



40 ibid

[462], [479], [483]–[491]. [471]; see also [476]. 42 ibid [608]. 43 ibid [611]–[617]. 44 ibid [613]. 45 Pfizer (n 4) [443], [447]–[448], [455], [546], [553] and [558]–[559]. 46 Baxter First Instance (n 6) [629]. 47 ibid [630]–[631]. 48 ibid [634(e)], [641]. 49 ibid [643]. 50 Baxter Full Court (n 8) [228]. 41 ibid

206  Bundling process itself is not the ‘competitive process’.51 The final determination of factual matters regarding the ‘process of competition’, purpose and effect therefore fell to the Full Court. The Full Court reversed the trial judge on the question of ‘taking advantage’ of market power, so Baxter was held liable under section 46 in relation to all the impugned conduct. However, I am more interested in the SLC test, which the Harper Review section 46 will adopt. The leading judgment was delivered by Mansfield J, who interpreted the first instance judgment as holding that Baxter’s conduct in bidding the way it did contravened both the purpose and effect limbs of the SLC test, but that entering into the contracts did not.52 Mansfield J agreed with the trial judge’s conclusion that there was no purpose or effect of SLC in the wider process of competition, but also agreed with the trial judge that hindering the tender process did in fact amount to SLC (provided it was legitimate to aggregate the effect of the conduct, as it appears the trial judge did).53 Mansfield J went further to hold that entering into the contracts resulting from the tender process also contravened section 47.54 Dowsett J dissented, concluding that Baxter’s relevant purpose in the bids was to win the tender, and the only thing it prevented was the rivals winning the tender.55 Thus, in His Honour’s view, ‘it is difficult to accept that hindering one transaction in a market necessarily involves hindering competition as a whole in the market’.56 I suggest there is merit in this view, which addresses the dual purposes problem and raises the question, not squarely addressed in the case, as to what point conduct in relation to single transactions becomes significant in the aggregate. As we will see below, the Full Court in Pfizer saw fit to approve the statement of applicable legal principles set out by Dowsett J in his dissenting judgment in Baxter Full Court. We may wonder whether that casts any doubt on the decision of the majority. Subsequent to the Baxter case, we see again that market circumstances continue to change. For example: (a) in 2013, the EC approved Baxter’s acquisition of Gambro, which appears to have combined their HD and PD businesses;57

51 ibid [239]. 52 ibid [203]–[206] and [217]; Gyles J agreed (at [387]). Their Honours refer to the trial judge’s reasons in Baxter First Instance (n 6) [629], though [632] seems to deal differently with the ‘effect’ limb, which may support a different interpretation, as I suggest above. 53 ibid [245]–[246]. 54 ibid [237]; Gyles J agreed (at [387]). 55 ibid [324] in relation to s 46. 56 ibid [353] in relation to s 47. 57 EC, ‘Commission Approves Acquisition of Swedish Medical Technology Company Gambro by US Rival Baxter, Subject to Conditions’, press release, 22 July 2013, europa.eu/rapid/ press-release_IP-13-724_en.htm.

Gaps in the Factual Matrix  207 (b) NxStage is an HD product manufacturer which entered the market in 1998, and is a potential PD product manufacturer, which at the time of writing is being acquired by Fresenius.58

III.  Gaps in the Factual Matrix To apply the Harper Review section 46 and the market manipulation approach, we need to address significant gaps in the factual matrix. First, we have no explanation of Baxter’s success in the PD market following its launch of new products in the mid-1990s. There is an implication that its market share in Australia is attributable to its bidding practices. However, we may doubt that its success in the global market is significantly contributed to by its bidding strategy in Australia. We do not know whether that strategy has been used, or challenged, in other jurisdictions. Baxter’s dominant market share in the global market for PD products suggests this is an unlikely case in which to apply the ‘bundling’ theory of harm. We would like to know whether its global market share results from the clinical success of its PD products. The trial judge recited substantial evidence of clinicians in Australia regarding the competing products, which was apparently directed to the question of market power, but found it to be inconclusive.59 Second, we would like to know whether Baxter was meeting demand in Australia and whether rivals were excluded by the size of the market. The ACCC made the point on appeal that there was no finding by the trial judge that there was a minimum efficient scale which meant that only one firm could supply the Australian PD market.60 Their argument was that, absent such a limitation, Baxter’s conduct in preventing rivals from achieving minimum efficient scale must have an anticompetitive effect in the wider market. The point was not held to be significant for the appeal; however, it emphasises the need to understand the size of the market to assess whether rivals are excluded by the market itself or by the conduct of the incumbent. Third, we lack data concerning avoidable cost and any efficiency benefits of bundling. We may wonder why this was a matter for expert opinion, as opposed to merely a matter for evidence of a factual nature. As we noted in Chapter 1, it seems that avoidable cost (a surrogate for marginal cost) is not directly observable and

58 See, eg, M Thibault ‘NxStage Medical Details New Features, Hints at Peritoneal Dialysis System’, Device Talk, 8 August 2016, www.mddionline.com/nxstage-medical-details-new-features-hints-peritoneal-dialysis-system; NxStage, ‘Fresenius Medical Care to Acquire NxStage Medical, Inc. to Strengthen its Vertically Integrated Dialysis Business’, press release, 7 August 2017, ir.nxstage.com/news-releases/ news-release-details/fresenius-medical-care-acquire-nxstage-medical-inc-strengthen; B P ­ erriello, ‘Fresenius Extends Deadline for $2B NxStage Buyout’, Mass Device, 31 July 2018, https://www. massdevice.com/fresenius-extends-deadline-for-2b-nxstage-buyout-tops-q2-earnings-forecast. 59 Baxter First Instance (n 6) [337]–[339]. 60 Baxter Full Court (n 8) [236].

208  Bundling so can only be derived by manipulating available data. The trial judge’s conclusion that the information we need is incalculable and incomprehensible is troubling, as is the judicial method of proceeding to make qualitative judgements about critical matters that are quantitative. We could infer that if experts cannot undertake the calculation, we should not reasonably expect firms to be able to articulate their price-setting decisions by reference to economic principles that the experts cannot comprehensibly explain. We know that experts have trouble explaining themselves to laypersons and that judges are a tough audience. However, we have seen that there are substantial objective difficulties both with the economic concepts and the attendant forensic issues. I suggest that the focus of the market manipulation approach on demand response, ie, data concerning prices and revenue and the size of the market, may avoid some of the forensic uncertainties arising from costprice comparisons. Fourth, we know that government policy in the area of healthcare reflects the view that markets do not provide an appropriate mechanism to determine access to healthcare because people who do not have the ability to pay would be excluded by the market. Thus, government reimbursement and health insurance schemes both ensure access and provide a constraint on the pricing power of providers of medicines and medical devices.61 The fact that demand for the clinical products in question in the Baxter case responds to patient need rather than price, so that demand is inelastic,62 does not mean that the power to raise the price is unconstrained. We may well want to know how reimbursement schemes might constrain Baxter’s market power regarding SF products and PD products (or the market power of Gambro and Fresenius regarding HD products). We can now consider how we might address some of the unresolved issues that remain after disposition of the Baxter case when we test proposed new laws.

IV.  Baxter under Harper Section 46 It is likely that the same result would occur if the Harper Review section 46 as enacted were applied to the factual matrix of Baxter. Market power has been found to exist, and so has the necessary purpose and effect of SLC. We may wonder what role remains for the theory of harm described in the US and EU literature, ie, do we need to consider all the indicia (pricing below avoidable cost, serious risk of recoupment and absence of efficiency gains) to demonstrate effect and causation? Is it satisfactory for these matters to be assessed qualitatively?

61 See, eg, Independent Hospital Pricing Authority, ‘Home Delivered Dialysis Costing Study to Inform the National Efficient Price 2015’, report prepared by PwC, 11 November, 2014, www.ihpa.gov. au/sites/g/files/net636/f/publications/hdd_report_for_ihpa.pdf. 62 Baxter First Instance (n 6) [572].

Baxter under Harper Section 46  209 Clearly, the Harper Review intends that efficiency effects be considered. However, whether section 46 as enacted permits this and, if so, how it might be done remain to be considered by the courts. We might also pertinently ask what guidance firms with market power gain from the Baxter case about how they should conduct themselves when bidding for public tenders. Mansfield J observed that Baxter could have legitimately bid by reference to ‘price, quality or service differentials in relation to PD fluids’, but not by engaging in exclusive dealing having the quality of failing the SLC test.63 While that observation reflects the legal provisions, I respectfully suggest that the law offers no guidance to elucidate the difference between conduct that is permitted and conduct that is to be condemned. We recall the premise accepted by the Harper Review that market power may give conduct that would be lawful for a firm lacking market power an added impact under the SLC test. We may also recall the analysis of Smith and Round discussed in Chapter 5 that a new entrant lacking scale is vulnerable to predatory price-cutting by the incumbent above avoidable cost. We noted in Chapter 3 that in the US it is thought to be impossible to determine whether such pricing conduct may be characterised as predatory. Is this potentially to be unlawful under the Harper section 46 SLC test? We saw from Cement Australia that a firm can gain market power from winning contracts by bidding in public tenders and may have to consider itself to have power once it can be confident of winning. We might infer from the judgment of Allsop J in Baxter First Instance that a bid must allow rivals a chance to win (since it is not permissible to make a bid that they cannot win). Clearly, in multiple contemporaneous tenders, it is problematic to take into account the probability of winning any one of more bids.64 I respectfully suggest that we can derive no practical guidance from the proposition that a firm with market power must bid in a way that gives its rivals a chance to win. I suggest that when we apply the AMC test for permitted conduct, we can see that sections 46 and 47 are at cross-purposes, which might explain why the experts in Baxter failed to provide comprehensible assessments. Let us test the above issues by asking how the proposed new law would apply if Baxter complied with the AMC test, ie, priced PD products equal to avoidable cost (or only below it to the extent justified by economies of scale and scope). This would not exclude an equally efficient competitor for whom the market is large enough to enter and achieve minimum efficient scale, but would require the new entrant to be able to sustain losses until it achieves that scale. I suggest there is a real chance that conduct permitted by the AMC test could be held to raise barriers to entry and thus be condemned under the SLC test. However, we can see that the real issue is whether the market is large enough. If it is not, we can conclude that the market excludes new entry. 63 Baxter Full Court (n 8) [243]. 64 This caused the economics experts some insurmountable problems in Baxter First Instance (n 6) [523]–[529].

210  Bundling The AMC test also requires us to consider whether recoupment of the incumbent’s forgone profits would be likely. I suggest that Boral should be regarded as good authority for the recoupment test in the context of predatory pricing, which in essence is what we confront in the case of bundling. The recoupment test essentially asks whether the market for PD products can in fact be monopolised, ie, whether prices can be raised above the assumed competitive level. While we may doubt on the facts of the Baxter case that this would be caused by bundling (as opposed, eg, to Baxter’s apparent success in global PD markets), judicial views on the recoupment test, and how it can be assessed forensically, remain to be decided. Accordingly, we see that the section 46 test is based on assessing predation and that efficiency is relevant, but may be limited to justifying pricing slightly below avoidable cost. However, the section 47 exclusive dealing test concerns itself with foreclosure of significant parts of the market from competition for significant periods of time, without reference to efficiency, because it is implicitly assumed that the market is large enough to support multiple firms. It would be strange if conduct that might be cleared of liability under section 46 would nevertheless be condemned under section 47. In fact, the Harper Review recommended that section 47 should be repealed, as it has no work to do once the Harper section 46 SLC test is adopted.65 However, this recommendation has not been accepted. If the Harper mandate to consider efficiency is followed, I suggest we would need to ask whether the incumbent is meeting demand or whether there is unmet demand. If it is meeting demand, we should then ask whether the size of the market justifies the resources required to facilitate new entry. Establishing a new multimillion-dollar plant in Australia would not seem to be justified without a positive demand response, but, as we have seen, demand for therapeutic products responds to the incidence of the medical condition rather than the price of products. We would thus want to know whether the incidence is increasing or stable and whether demand can be met with imports which do not need substantial capital investment in Australia. Since the relevance of efficiency to the current SLC test and the Harper Review section 46 as enacted is a matter of some doubt, yet to be judicially determined, we can turn to consider the market manipulation approach, to which considerations of efficiency are central.

V.  Baxter under Market Manipulation When we conceive of market power as the power to manipulate the market, we avoid the issues we saw in Cement Australia and Baxter to do with contractually based market power, estimating probabilities of winning bids and aggregating

65 Commonwealth

of Australia, Competition Policy Review, March 2015, 64.

The Implications of Pfizer  211 the foreclosing effects of contracts. Our concern is with the size of the market, whether there is unmet demand that could profitably be supplied, and whether new entry will result in positive demand response to justify the expenditure of additional resources to produce the output that the market requires. We want to know if the new entrant is more efficient. We want to know if the incumbent is more efficient or faces exit costs that might explain pricing that appears to be below avoidable costs (issues not addressed by the AMC test). That is, we want to know if new entry is excluded by the market or by conduct of the incumbent which cannot be justified on efficiency grounds. We might doubt, when powerful corporations are engaged in a struggle to take sales away from each other, whether competition law should endeavour to protect one from another, as any damage they do to each other may be compensated by the benefit to consumers from lower prices. The more detailed discussion of these issues in Chapters 5 and 6 need not be repeated here. Without data, we can only speculate on the outcome of the Baxter case under the market manipulation approach. However, we can see how the principles could be applied. This approach may well solve some of the forensic problems we are confronted with under the SLC test.

VI.  The Implications of Pfizer The Full Court of the Federal Court of Australia delivered judgment on 25 May 2018 in Pfizer.66 It is not necessary for the present purposes to consider the facts of the case in detail. It is of interest primarily for the implicit recognition that the decision gives to the ‘dual purposes’ problem in zero-sum competition, though the court did not use those terms in its reasons. Pfizer is an ‘originator’ of patented medications. In the few months before the expiration of a patent on one of Pfizer’s ‘blockbuster’ drugs, it sought to launch its own generic version of the product and to bundle it with supply of the patented product in order to gain a market presence ahead of anticipated entry by generics manufacturers.67 It was anticipated that generics manufacturers would flood into the market, seeking, in the initial period after patent expiration, considerable benefits from the subsidy provided by the government under the Pharmaceutical Benefits Scheme.68 The Full Court reversed the trial judge on the question of market power, holding that Pfizer still had market power and took advantage of that power when it established the bundled offer.69 Nevertheless, the Full Court held that Pfizer’s substantial subjective purpose was neither an exclusionary



66 Pfizer

(n 4). [443], [553] and [558]–[559]. 68 ibid [546]. 69 ibid [353]–[356], [417], [422], [523]. 67 ibid

212  Bundling purpose nor a purpose of SLC. Thus, the ACCC’s claim against Pfizer under both sections 46 and 47 failed and the Full Court dismissed the ACCC’s appeal. As the SLC test applies under section 47, the case may offer some insights into how the SLC test might apply to misuse of market power under the new section 46. It is possibly for this reason that the ACCC sought (unsuccessfully) leave to appeal the decision to the High Court to clarify the ‘purpose test’.70 Since the case was run on the basis of ‘purpose without effect’, it would seem to have little to offer us regarding the new ‘effects test’ that has been introduced into section 46 following the Harper Review proposals. Importantly, the Full Court recognises, in circumstances akin to zero-sum (or survival) competition, that a dominant firm may permissibly use its market power to protect its own sales, without it necessarily following that the firm has a substantial subjective purpose of excluding competitors or substantially lessening competition.71 A key element in this reasoning is that upon the expiry of Pfizer’s patent and the entry of generics manufacturers, Pfizer had to decide whether to exit the market for the particular product or to alter its traditional business model as an ‘originator’ of patented medicines to become a generics manufacturer itself.72 The court’s decision in Pfizer appears to have turned on the facts of the case, as the Full Court considered in meticulous detail the findings of the trial judge concerning the ‘substantial subjective purpose’ of Pfizer, based on the evidence of its executives in the case. We noted earlier in the present chapter that in the Baxter case, the trial judge rejected evidence of Baxter’s executives that their substantial subjective purpose was to protect the company’s sales. The Pfizer decision may nevertheless invite us to reconsider the application of relevant principles by the majority of the court in Baxter Full Court. This is because the Full Court in Pfizer takes the slightly unusual course of approving the principles enunciated by Dowsett J in his dissenting judgment in Baxter Full Court.73 Relevantly for our purposes, the Full Court in Pfizer endorses the proposition set out by Dowsett J that one competitor injures another by attracting away sales and that ‘the effect of conduct upon competition is not to be equated with its effect upon competitors, however the latter effect may be relevant to the former’.74 I suggest that this proposition reflects the nature of zero-sum competition and that the Pfizer case more clearly presents us with a case of competition for survival.

70 ACCC, ‘ACCC Seeks Special Leave to Appeal Full Federal Court’s Pfizer Judgment’, press release 115/18, 25 June 2018. On 19 October 2018, the High Court of Australia dismissed the application by the ACCC for special leave to appeal the decision in Pfizer: see ACCC, ‘High Court Refuses the ACCC Special Leave to Appeal Pfizer decision’, press release 212/18, 19 October 2018. 71 Pfizer (n 4) [558]–[559]. 72 ibid [395], noting that a high number of Pfizer’s patented products were coming off patent in the period 2010–14. 73 ibid [470], [475], [477]–[478]. 74 ibid [477].

Conclusion  213

VII. Conclusion The analysis of ‘bundling’, particularly in the market circumstances considered in Pfizer, is highly analogous to that set out in Chapter 5 in relation to predatory pricing, so I will not repeat what is said there. The Full Court in Pfizer, I suggest, demonstrates the acute difficulties in zero-sum competition of grappling with the ‘dual purposes’ problem. We can see in the present chapter that new entry did not apparently respond to unmet demand, but to the opportunity to appropriate, or redistribute, surplus from the incumbent to new entrants. There does not appear to have been any suggestion in Pfizer of any harmful competition or efficiency effects. On the contrary, there may well be inefficiencies caused or contributed to by the paradigm of zero-sum competition itself. We can now proceed in the next chapter to consider the implications of our analysis of competition and efficiency effects in relation to market power for the institutional framework, which divides the functions of the courts from the functions of the ACCC/ACT, and some procedural issues, such as the burden of proof.

9 Institutional and Procedural Implications I. Introduction In this chapter we will first briefly review the decision of the High Court in NT Power.1 This case discloses the High Court’s views concerning its role, and the role (if any) of efficiency in section 46, compared to the role of the ACCC/ACT. Our interest concerns issues of efficiency that, consequently, the High Court did not need to decide, so we can pass over much of the case without comment. After discussing the institutional implications of the study, we will briefly discuss the question of where the burden of proof should lie with regard to the issues of efficiency necessary to determine liability for misuse of market power.

II.  NT Power The Power and Water Authority (PAWA) is an agency of the Northern Territory government.2 It is a vertically integrated electricity generator, distributor and retailer, owning its own transmission lines. In 1996, the owner of the Mt Todd gold mine contracted NT Power to generate electricity for the mine using the mine owner’s generation facilities. NT Power sold surplus electricity to PAWA, or bought electricity from PAWA to satisfy excess demand at the mine. NT Power leased the transmission line used to connect to PAWA’s grid. In late 1997, the gold mine fell dormant for about two years. When the mine suspended operations, NT Power had the option to acquire the generation assets, which it exercised in 1998 and was granted regulatory approval to sell its electricity to the public. However, NT Power needed access to PAWA’s distribution grid in order to make retail sales. In late 1998, PAWA refused to provide access at

1 NT Power Generation Pty Ltd v Power and Water Authority (2004) 219 CLR 90 (hereinafter ‘NT Power’). 2 All facts are taken from the judgment in ibid [30]–[45] (McHugh ACJ, Gummow, Callinan and Heydon JJ).

NT Power  215 that time (seeking to delay access for two years). NT Power commenced proceedings claiming breach of section 46 by reason of PAWA’s refusal to supply access to transmission facilities. The trial judge held that PAWA was entitled to crown immunity and so was not subject to the CCA, but that if it were, PAWA had contravened section 46.3 The Full Court of the Federal Court upheld the trial judge’s decision, agreeing with the trial judge’s conclusion regarding section 46, in the alternative that immunity was held on appeal not to apply.4 The High Court reversed the courts below on the question of immunity, holding PAWA liable for contravening section 46. PAWA raised a number of efficiency arguments that were rejected by the trial judge and again by the High Court.5 NT Power was a more efficient electricity generator than PAWA and would likely undercut the prices of PAWA. ­Section 46 does not permit an inefficient monopolist to give itself time to reorganise, in order to become more efficient, by obstructing the entry of new competitors.6 PAWA’s retail market spans a very large area of northern Australia, servicing many remote communities. PAWA was obliged by its statutory charter to cross-subsidise remote communities and feared that unless it had time to become more efficient and to ameliorate cross-subsidisation, NT Power would target easily reached retail customers, leaving PAWA those that are more distant and less profitable. The High Court pointed out that ‘reasoning of that kind might be relevant in an application to the [ACCC] for authorisation under s 88, at least in so far as the purposes were likely to mature into effects generating public benefits’, but that ‘s 88 does not make authorisation available for s 46 conduct, and in any event authorisation is not a matter for the courts’.7 We can consider how these issues would be dealt with under the Harper Review section 46 if we follow the recommendation of the Review that the courts should consider efficiency effects. Clearly, the ACCC/ACT will be given the power to authorise conduct that would otherwise contravene section 46 if there are overriding efficiency or other public benefits. Parties seeking authorisation can only obtain it in relation to prospective conduct, ie, must be able to await the outcome of their application and any appeals before embarking on the conduct. This would not have assisted PAWA, as there does not appear to be a way to refuse access conditional on authorisation being granted (since the refusal is immediate). As we saw in Chapter 7 in the aftermath of the Cement Australia case, authorisation is a cumbersome process in the case of competitive tenders or other time-critical commercial decisions.

3 NT Power Generation Pty Ltd v Power and Water Authority (2001) 184 ALR 481. 4 NT Power Generation Pty Ltd v Power and Water Authority (2002) 122 FCR 399. 5 The facts set out below are taken from the judgment in NT Power (n 1) [137]–[141] (McHugh ACJ, Gummow, Callinan and Heydon JJ). 6 ibid [138]. 7 ibid [137].

216  Institutional and Procedural Implications We do not know the scale of NT Power’s generating capacity or the extent of its cost-efficiency advantage.8 NT Power may well be more efficient with a small capacity designed to serve a dedicated user, ie, the Mt Todd gold mine (although NT Power clearly needed to draw electricity from PAWA at times to satisfy demand at the mine). However, if NT Power does not have the capacity to serve a substantial portion of the broader market, we would need to understand the capacity of PAWA to meet market demand, the cost-price impact of PAWA suffering a reduction in volume, and the response of market demand to the new entrant in order to assess whether new entry is efficient. We can see that more information is required to address efficiency than merely cost data and would like to know more about market demand. We can see that if PAWA’s capacity will still be required to meet market demand after NT Power’s entry, there is merit in allowing PAWA the opportunity and incentive to become more efficient. New entry may, of course, do this. However, it may be overly simplistic to take the view that PAWA is inefficient and should exit the market, without considering the substantial exit costs that this may well entail. We can also see that NT Power had previously sold excess electricity to PAWA and we might conclude that NT Power had the ability to continue to do so after 1998 (subject to agreement on prices and terms). That would appear to have been a simple resolution (like Queensland Wire buying fence posts from BHP at wholesale prices, as we saw in Chapter 4). However, NT Power presumably sought higher margins by using PAWA’s transmission system to access retail customers, ie, to become a vertically integrated electricity generator and retailer like PAWA. While it is not necessary to discuss the matter of AGL/MacGen,9 mentioned in Chapter 1, in any detail here, we might note the ACCC’s opposition to vertical mergers between generators and retailers in the national electricity market on the basis that they may SLC in the retail market (an argument rejected by the ACT in that case). The structural separation of generating and retailing appears now to be a feature of the Northern Territory market, which might ameliorate the concerns of PAWA raised in the NT Power case.10 Accordingly, without further data and economic analysis, it may not be possible to conclude that entry by NT Power as a vertically integrated generator/retailer would necessarily result in benefits to consumers (rather than merely redistributing surplus from PAWA to itself). We can see that the separation of functions between the courts and the administrative bodies, the ACCC/ACT, can lead to different outcomes depending on

8 The first instance judgment, NT Power Generation Pty Ltd v Power and Water Authority (n 3) [40]–[41] and [48], does give some figures for generating capacity, which suggest that NT Power might have up to 12 per cent of combined capacity, but the figures may not be exhaustive; see also [83]. 9 AGL/MacGen [2014] ACompT 1. 10 PAWA, ‘Interim Northern Territory Electricity Market’, 27 May 2015, www.powerwater.com.au/ networks_and_infrastructure/market_operator.

The Burden of Proof  217 which forum is chosen or reasonably available in the context of time-constrained commercial decisions. If efficiency were clearly a matter within the jurisdiction of the courts, we would be less concerned about the institutional division of functions. The Harper Review attempted to address efficiency in its recommended section 46, but the ACCC’s opposition to the Harper section 46(2) was instrumental in Parliament rejecting that aspect of the recommendation. Clearly, the ACCC considers the administrative agencies the appropriate forum to determine efficiency issues. With respect, given that section 71 of the Commonwealth of Australia Constitution Act effectively reserves to the courts the power to determine liability for penalties and damages, I suggest that the Harper recommendation is to be preferred. Further, I suggest that we should recognise the central role of efficiency considerations in determining whether market power has been misused and that efficiency, as we can adapt and redefine it for our purposes, should be no less capable of judicial determination than the SLC test. We also saw in the preceding chapters in Part II that the outcome of section 46 cases often depends on decisions that parties (including the ACCC) make about the way they run their case. Firms may be expected to use section 46 for commercial purposes to leverage commercial outcomes. Particularly at the appellate level, legal outcomes are notoriously dependent on market definition, about which views may reasonably differ. In Chapter 7 we saw that the dual prosecutorial and administrative decision-making roles of the ACCC may result in cases being run in court in a way that may differ from the way in which the ACCC approaches administrative decision-making. Further research would be required to assess whether that is a systemic concern, in which case it may be desirable to consider separating the prosecutorial and administrative decision-making roles to facilitate independent critical evaluation of the approach to be taken to cases. Nevertheless, I suggest that, for a variety of reasons, judicial and administrative decisions concerning section 46 may well be influenced by the decision-maker’s view of the world, which will rarely (if ever) be explicitly stated. I suggest that a focus on efficiency considerations should enable us to assess market power and its misuse on a more objective basis that would diminish the influence of differing worldviews.

III.  The Burden of Proof Before proceeding to draw some conclusions from our review of the cases in Part II, there is a further procedural aspect of market power cases that could conveniently be mentioned for the sake of completeness. However, a detailed discussion of procedural aspects is beyond the scope of this monograph. I mentioned in Chapter 3 the question of negative and affirmative defences, which raises the question of burden of proof. We may recall that it is unclear in competition law whether efficiency is or should be a defence that negates a claimed anticompetitive effect, and so negates a contravention, or whether it is an affirmative defence that is raised

218  Institutional and Procedural Implications upon admission of a contravention, and so ameliorates harm. The Harper Review recommended section 46(2), had it been accepted by Parliament, would not elucidate that question because it does not elevate efficiency to become an essential element of the contravention, but treats it merely as a relevant matter that the court must consider. Thus, had the Harper section 46(2) been enacted as recommended, the defendant would appear to bear the risk of not being able to adduce any probative evidence on the subject of efficiency.11 We saw in Chapter 3 that efficiency is not a defence in Australia to a complaint alleging an anticompetitive effect, whereas it is in the US, but there the nature of the defence and its implications for the burden of proof are unclear.12 So, for example, the ABA submission to the Harper Review on the review’s issues paper states that the defendant may rebut a claim of monopolisation by showing a legitimate business rationale, eg, to generate efficiencies, in which case the plaintiff must show that the conduct in question was not reasonably necessary to generate those efficiencies.13 Nelson and Smith argue that ‘the party asserting the efficiency rationale should bear the burden of proof ’, since it is typically that party which ‘has the best access to the information needed to identify and factually verify relevant efficiencies’.14 I have argued in Chapter 3 that we should give considerations of efficiency primacy, as it is conceptually distinct from competition effects and is a higherorder objective of competition law. However, contrary to the views of Nelson and Smith, I suggest that under the market manipulation approach, the plaintiff should be required to prove all issues pertaining to efficiency and inefficiency because these matters will likely require evidence of demand response and efficiency effects from sources beyond the defendant. Such sources would include, eg, the rival firm or firms. As the defendant will never know whether a rival is more or less efficient, I suggest that the plaintiff should bear the burden of proving that conduct is inefficient, eg, that a more efficient competitor has been excluded. When we come to the category of conduct which is inefficient, but which is a natural and probable consequence of zero-sum competition, and may be tolerated because, eg, there is a transfer of surplus away from the defendant, the situation is more difficult. However, on balance, I suggest that the plaintiff should still bear the

11 See generally, eg, CR Williams, ‘Burdens and Standards in Civil Litigation’ (2003) 25 Sydney Law Review 165. 12 See, eg, P Nelson and D Smith, ‘Efficiencies in Antitrust Analysis: A View from the Middle of the Road’ (2015) 60 Antitrust Bulletin 128, 149. 13 ABA, ‘Joint Comments of the American Bar Association Section of Antitrust Law and Section of International Law on the Australian Competition Policy Review Issues Paper’, 13 November 2014, 4. See, eg, Aspen Skiing Co v Aspen Highlands Skiing Corporation 472 US 585, 609–11. 14 Nelson and Smith (n 12) 149. Contra: P Akman, The Concept of Abuse in EU Competition Law (Oxford, Hart Publishing, 2012) 282–83 argues that absence of efficiency should be regarded not as a defence, but as an essential element of liability. I would go further to suggest that demonstrated inefficiency should be an essential element of liability.

Conclusion  219 burden of proving harm, having regard to those matters, which should be readily established by objective evidence, none of which is peculiarly within the control of the defendant. I suggest that the court’s ability to apply Jones v Dunkel15 to draw adverse inferences if a party fails to adduce evidence that is within its power to put before the court should be sufficient to address the risk of defendants withholding data adverse to their positon. We saw this applied in the Cement Australia case discussed in Chapter 7, when Cement Australia failed to call evidence from key executives about the commercial rationale for certain conduct.

IV. Conclusion We have seen that the traditional institutional division of functions has supported the view that Australian courts should not get involved in considerations of efficiency. It appears that the administrative jurisdiction of the ACCC and the ACT is truly polycentric, since it involves weighing up conflicting economic welfare objectives (ie, allocative and productive efficiency) as well as non-economic objectives. We might question whether a polycentric law raises serious rule of law concerns, even if administered by an agency of the executive branch of government; nevertheless, that question is beyond the scope of this monograph. The purpose of this monograph is to propose a unicentric legal theory of liability for misuse of market power, the application of which can more clearly fall within the constitutional and institutional competence of the courts. The basis of the theory of harm and theory of legal liability proposed in this monograph is efficiency, as it can be adapted and redefined for such a purpose. I suggest that it follows that the plaintiff ought to bear the burden of proving inefficiency. While procedural and institutional aspects of the enforcement and application of section 46 are important, especially if they have substantive effects, our main concern is the way we assess market power and its use in the context of zero-sum competition, and how we assess competition and efficiency effects. We can now form some conclusions from testing the Harper Review section 46 and the market manipulation approach against the factual matrices of the key court decisions considered in this Part. Part III sets out the conclusions from the study and offers some recommendations regarding market power and the proposed legal theory of liability for misuse of market power.



15 Jones

v Dunkel (1959) 101 CLR 298.

220

part iii Conclusion

222

10 Findings and Recommendations I. Introduction Australian court decisions in section 46 cases have been criticised because they appear to be influenced by unstated policy criteria. Differing worldviews are manifest in the debate about the policy objectives of competition law. Many believe that the objective should be to promote productive efficiency, which would maximise production and income from society’s scarce resources, and in so doing would produce goods and services for the least cost. Others believe that the objective should be to promote consumer welfare in each market by redistributing surplus from producers to consumers, even though this may be inefficient because it draws resources away from other uses of benefit to society. Since resources are limited, such an objective could not be achieved across the board. From our review of key Australian cases, we have seen other worldviews at work. We may suspect that firms having substantial market power (which I have called ‘dominant firms’ for ease of reference) achieve or maintain dominance not through superior products or efficiency, but through anticompetitive practices. We may suspect that harm to competitors necessarily harms the competitive process. We may suspect that consumers are powerless when markets are dominated by large firms. I suggest that this complex web of underlying worldviews may explain why market power and its use or misuse is generally accepted to be the most problematic and controversial area of competition law. It is said that we lack a positive and normative theory of market power and that consequently there is an ongoing difficulty distinguishing: (i) conduct of a dominant firm that should be permitted (which I have called ‘normal competitive conduct’, sometimes called ‘competition on the merits’); and (ii) conduct of a dominant firm that should be condemned (sometimes called ‘exclusionary conduct’ or ‘abusive conduct’). This study began first by examining the origins of our legal concept of market power and the paradigm of competition from which it is derived – the hypothetical competitive market (epitomised by the neoclassical economic model of perfect competition, sometimes called ‘atomistic competition’). The traditional concept

224  Findings and Recommendations of market power has not been questioned in the recent law reform debate in Australia about market power and the regulation of its misuse. However, the relevance of the hypothetical competitive market in which market power is absent has been questioned as a benchmark (ie, positive or normative standard) to assess the conduct of dominant firms. This debate informed the consideration of the matter by the Harper Review and led that review to propose that conduct of a dominant firm should be assessed not by reference to exclusionary purpose, but by reference to anticompetitive purpose, effect or likely effect (the SLC test). The Harper Review also recommended that when assessing anticompetitive purpose or effect, the court should be required to take into account any offsetting efficiency effects. Amending legislation to implement the Harper proposals has been passed by Parliament. However, when the Bill was considered by the relevant Senate committee, the ACCC opposed the requirement that the court should consider efficiency effects, and Parliament removed that subsection from the amending Bill. I suggest that the Review was correct to draw attention to efficiency. I would, however, go further to suggest that our concept of market power can and should be reconsidered, and that economic efficiency can be adapted to become the centrepiece of our positive and normative legal theory of market power. The Harper Review did not explain what it meant by ‘efficiency’. However, the expression is conventionally understood to refer to social efficiency, and I use it in this sense, unless otherwise indicated. However, social efficiency is legally indeterminate (as is the ‘competitive level’ of price and output in a market, which is conventionally presumed to be a surrogate for social efficiency), so we must adapt the concept of ‘efficiency’ for use in our legal theory. While we cannot identify the optimum, we might adopt a similar incremental approach to that of the competition test. I use the terms ‘efficient’ and ‘inefficient’ in this incremental sense of enhancing or diminishing efficiency. I suggest that we can adapt and redefine efficiency in the context of a market or industry. I use the term ‘efficient market’ similarly to its legal usage in securities markets manipulation laws to refer to a market in which forces of demand and supply operate undistorted. This study also has implications beyond the scope of section 46 which could inform further research – eg, regarding the SLC test in sections 45 and 47, consideration of merger efficiencies under section 50, and institutional arrangements – more generally in competition law. The section 46 cases reviewed in the present study suggest that the predictive effects test may well involve qualitative inferences rather than evidence-based findings of harmful economic effects and may be too easily influenced by evidence of hostile intent (facilitated by section 4F, which provides that a ‘substantial’ proscribed purpose is sufficient for a contravention). These may well be matters of general concern under competition law that should be the subject of further research. For the purposes of the present study, we confine our attention to the implications for section 46.

Paradigms of Competition and Norms of Conduct  225

II.  The Concept of Market Power The traditional legal concept of market power, as the power to ‘give less and charge more’, or the ‘power to behave differently than a competitive market would enforce’, implies that market power can be possessed by a dominant firm. I suggest that when we analyse the traditional concept of market power, we can alternatively view it as a function of elasticity of demand, ie, that market power may be caused or contributed to by the response of market demand to firms’ price/output decisions. Accordingly, I suggest that market power can alternatively be legally defined as the power of the market, which cannot be possessed, but can be manipulated. It follows that all firms have this power to some degree. Thus, a new entrant may increase output in the market, causing prices to fall, thus instigating a demand response and possibly triggering a ‘price war’. From our review of leading Australian cases, we have seen that parties on the other side of the market also have power to manipulate the market to some degree. Examples would include: (i) buyers contributing to bidding contests between sellers in a price war or by calling for public tenders; (ii) sellers of raw material inputs creating bidding wars among buyers by calling for public tenders. The traditional concept of market power recognises that a dominant firm may be constrained by countervailing power, which is given statutory recognition in section 46(3). However, assessment of constraints on market power is an inexact process that is reduced to a binary outcome under which the dominant firm is assessed either to have substantial market power or not. Accordingly, I suggest that the traditional legal concept of market power is insufficiently determined and so may not assist us in analysing the harmful effects of market power or determining the standard of conduct which is expected of firms if they are to comply with section 46. Since section 46 is couched in the familiar legal form of a prohibition of specified conduct, it inherently lacks specific guidance as to permitted conduct, which in the case of section 46 is non-obvious. I suggest that the alternative legal concept of market power as power to manipulate the market provides us with the opportunity to identify positive and normative legal standards to assess the use of market power. However, in order to fulfil this promise, we need to consider the implications of our real-world paradigm of competition.

III.  Paradigms of Competition and Norms of Conduct We have seen that the hypothetical competitive market is an idealised paradigm with many assumptions that may not be satisfied under real-world conditions.

226  Findings and Recommendations No firm has market power as it is traditionally conceived, because no firm can influence the market by its price/output decisions. Thus, firms can sell more without taking away sales from competitors. Further, there is free entry and exit, so that when supply responds to, eg, a demand shift reduction, firms freely leave the market without being harmed. Real-world competition is quite different – it has a number of characteristics that lead me to propose the paradigm of ‘zero-sum competition’ as an alternative framework in which to assess the use of market power. It is generally accepted that firms in the real-world compete for the same object (ie, sales) and that a firm gains sales by taking them away from its competitors. Clearly, this is the case when a new firm enters the market. It would also be the case when a firm wishes to grow its sales faster than the market. We anticipate that firms may overestimate their chances of success, so (contrary to conventional economic wisdom) may well not be daunted by the challenge of entering a market and taking sales away from an incumbent in a fight for survival, though many will fail. We know that there are exit costs associated with leaving the market – eg, assets may have to be sold at ‘fire sale’ values – so sunk costs will not be recovered as they otherwise might have been. This means that, under zero-sum competition, incumbent firms may have little choice but to stay in the market and fight for survival, even if they have to sell products at less than avoidable cost. When we consider market power as the power to manipulate the market, we can distinguish firms’ price/output decisions, and other decisions, which may well represent both a privately and socially efficient response to market demand from conduct which manipulates the market, distorting the market’s operation. That is, we can hold conduct to be permissible not because it is rational to a business person, but because it can be considered to be efficient, ie, in the sense that while we cannot identify the point of optimum social efficiency, we can say that increasing output beyond the profit-maximising position may well be both privately and socially inefficient. We cannot measure the impact on social efficiency, but equally we cannot assume it to be neutral. We must therefore adapt economic concepts of efficiency for use in our legal theory. We start with the legal concept of an efficient market from securities markets manipulation laws. This is a market in which undistorted forces of demand and supply operate. Conduct which is inefficient is prima facie manipulative. Our forensic enquiry considers whether there is unmet demand, ie, whether an expansion of output would be profitable and also whether the demand response would justify drawing resources from other socially valuable uses in the economy. We look at the relative efficiency of firms. If a more efficient new entrant has been driven out, this would suggest that the market has been manipulated. However, we know that more efficient firms can miscalculate the risks of business and the chances of an incumbent becoming more efficient in response to competition. If an equally efficient firm has been driven out, we consider whether the firm was needed to satisfy demand.

Forensic Assessment of Misuse of Market Power  227 On the other hand, we have seen that zero-sum competition can produce inefficient outcomes. From our review of leading Australian cases, we have seen that bidding conduct in price wars and bidding wars for inputs can follow from zerosum competition, with consequences that could well be inefficient. I suggest that we can hardly blame firms if they do what we expect them to do under zero-sum competition, ie, if in the circumstances the natural and probable consequence of zero-sum competition would be inefficient. I suggest further that we can tolerate this inefficiency if it is mitigated by a windfall transfer of surplus to consumers or, at least, a transfer of surplus away from the dominant firm. We thus have a basis to assess classes of conduct commonly associated with misuse of market power, eg, refusal to deal and margin squeeze, predatory pricing, bundling and raising rivals’ costs, using the market manipulation approach and efficiency to provide positive and normative standards.

IV.  Forensic Assessment of Misuse of Market Power: Conclusions from the Case Studies I suggest that our review of the case studies demonstrates that the following fundamental and widely accepted propositions about section 46 provide us with no real guidance and on the contrary may mislead us, in the essential forensic task of distinguishing permitted conduct of a dominant firm from conduct which should be condemned: (a) a monopolist is not to be deprived of the fruits of success; (b) competition law protects the process of competition, not competitors; (c) competition is the process by which firms compete for the custom of consumers.

A.  Proposition (a) We saw in Chapter 3 that proposition (a) has been thought to hold good in Australia, ie, section 46 does not make it unlawful for a firm to have market power or to derive profit from its market power. However, the matter has not been pronounced upon conclusively by the Australian courts. We have seen that there is typically no probative factual material in the reviewed section 46 cases going to the question as to whether the dominant firm’s success is well deserved. Instead, there tends to be an assumption, particularly in cases of prolonged dominance, that it is the result of anticompetitive practices rather than success in satisfying the demands of the market. Michal Gal supports proposition (a), arguing that to penalise superior efficiency or performance would distort the incentives for firms to create superior products

228  Findings and Recommendations or become more efficient.1 Gal argues that the Australian cases we considered in Chapter 4 may run counter to proposition (a) as a result of applying the hypothetical competitive market standard to assess ‘taking advantage’ of market power.2 Thus, she argues that the old section 46 may be used to prohibit pricing above the competitive level.3 The Harper Review section 46 of course removes the ‘taking advantage’ requirement. However, I suggest that the hypothetical competitive standard still underlies our concept of market power. I suggest that we lack clarity whether the Harper section 46 may be used to challenge excessive pricing or ‘limit pricing’ (ie, pricing above avoidable cost that may hinder or prevent new entry). These issues were not addressed by the Review, but may be a consequence of the recommended changes as they have been implemented by Parliament. It would clearly be better if this issue had been explicitly addressed in the reform process so that we could discern whether it is Parliament’s intention to change the law in this respect.

B.  Proposition (b) We saw in Chapter 3 that proposition (b) has been used in the EU to criticise formalistic (ie, non-economic) enforcement of abuse of dominance provisions, which is said to place too great an emphasis on protecting competitors, leading to over-enforcement. We saw that in Australia, where the Harper Review criticises the current section 46 for placing too much emphasis on protecting competitors, the law has been criticised for under-enforcement, ie, not sufficiently protecting the process of competition. It is hard to reconcile these conflicting uses of the proposition, which therefore appears to be of no real assistance. Further, in Chapters 7 and 8 we have seen that the SLC test in sections 45 and 47 of the CCA, which the Review recommends be adopted into section 46, has already been applied to conduct in section 46 cases. We find that harm to competitors can be taken to represent harm to the process of competition, ie, that the courts may qualitatively infer the r­ equisite purpose and likely effect of SLC from evidence of harm to competitors. We may accordingly conclude that the Harper section 46 SLC test as enacted could well take us in a direction that has been discredited in the EU.

1 MS Gal, Competition Policy for Small Market Economies (Cambridge, MA, Harvard University Press, 2003) 70. 2 ibid 87–88. Gal relies in particular on ASX Operations Pty Ltd v Pont Data Australia Pty Ltd (1990) 21 FCR 385, where the trial judge held excessive prices to indicate that the defendant took advantage of its market power over stock exchange trading data, and ordered the defendant to supply a certain data stream to the plaintiff for resupply. The Full Court ‘not without hesitation’ upheld the decision on appeal: ASX Operations Pty Ltd v Pont Data Australia Pty Ltd (1990) 97 ALR 513 [87], [102]–[104]. The plaintiff sought access on reasonable terms, but the final form of relief does not appear from the report. We do not know whether the plaintiff was a more efficient competitor. It appears that Pont Data filed for bankruptcy in the US in 1992: ‘A Pont Too Far’, Global Custodian, 1 March 1992. 3 Gal (n 1).

Forensic Assessment of Misuse of Market Power  229 I suggest that a focus on efficiency should help us overcome the ‘dual purposes’ problem we find to arise under the paradigm of zero-sum competition.

C.  Proposition (c) We discussed the market mechanism in Chapter 1. Markets are clearly a platform in which buyers and sellers transact business for mutual benefit. Thus, markets are agnostic about the relative interests of buyers and sellers. We have seen in the present chapter that the ‘competitive level’ of market price can be both conceptually and forensically indeterminate. We may be misled by focusing on competition as a process solely concerning firms, or by focusing solely on the interests of consumers in their capacity as buyers (rather than income producers), into confusing redistributive effects for efficiency effects. We saw in Chapter 4 that zero-sum competition may result in firms taking sales away from each other, with the effect that surplus may be redistributed between them without any efficiency effect (which would require consideration of demand response). We saw in Chapters 5 and 8 that buyers too have power to manipulate the market by exerting pressure on firms to bid prices down in a price war. I suggest that this outcome, while it may be inefficient, could be tolerated because consumers receive a windfall benefit. We saw in Chapter 7 that suppliers of inputs have power to manipulate the market by exerting pressure on firms which need the input to bid prices up in a bidding war. I suggest we may be able to tolerate this redistribution of surplus because while the effect on downstream consumers is unclear, it redistributes surplus away from the firms we suspect of misusing market power. We can now make some observations, based on our review of the cases, about the inter-relationship between competition and efficiency, which I suggest will help us understand ambivalent attitudes towards the propositions discussed above.

D.  The Inter-relationship between Competition and Efficiency This inter-relationship can be illustrated by Figure 10.1 below. Conduct that is competitive is within the circle labelled ‘Competition’. Everything outside that circle is anticompetitive. Conduct which is efficient is within the circle labelled ‘Efficiency’. Everything outside that circle is inefficient. Only in perfect competition do these circles coincide. Under that paradigm of competition, there is no market power as it is traditionally conceived (because no firm can influence the market), but there is free entry and exit, so competitors cannot harm each other. Clearly, we expect competition and efficiency to overlap in many cases (area ‘b’ in the figure). We would have no problems if that were always the case.

230  Findings and Recommendations Figure 10.1  The inter-relationship between competition and efficiency Competition

z

Efficiency

a

b

Other public benefits

c

However, this is not the case in the real world, as reflected by our paradigm of zero-sum competition, and this is where our problems begin. Firms face a ­downward-sloping demand curve under which each firm has, to some degree, the power, by its price/output decisions, to instigate a demand response – that is, the power to manipulate the market, which I suggest is an alternative way in which we can legally conceptualise market power. We see that all firms have the power to manipulate the market to some degree. We see also that both buyers and sellers in any market likewise have the power to manipulate the market to some degree. It is an oversimplification to ask whether the power of buyers constrains and so ‘cancels out’ the power of sellers, since the balance of power in markets is in constant flux as supply and demand adjust to each other. I suggest that economic efficiency can be adapted to provide us with a means of assessing whether the conduct of firms is manipulative and socially harmful, and so should be condemned. By adopting this approach, we may avoid qualitative judgements based on views of the world (which as we noted in Chapter 1 is a criticism that has been levelled at court decisions under section 46). It is generally accepted that under real-world conditions, competition can be a poor surrogate for efficiency. Thus, the CCA provides an administrative process whereby anticompetitive conduct which would contravene the law can be authorised by the ACCC/ACT if it can be justified on efficiency grounds (area ‘c’ in the figure) or on the grounds of other public benefits (indicated by the area within the dashed line). We may recall from Chapter 2 that Edward Mason in 1959 was forthright enough to acknowledge that we need a surrogate competition test because economic effects can be imponderable. We saw in our discussion of Boral in Chapter 5 that Kirby J acknowledged as much, but nevertheless argues that we need the predictive SLC test to prevent anticipated harm. We have seen, based on our review of the cases in Part II, that the courts may tend to qualitatively infer the purpose and likely effect of SLC rather than make a determination from evidence-based economic analysis. I suggest that a test which is both qualitative and predictive is even more at risk of vindicating a view of the world which cannot be tested to the degree we expect of judicial decisions. Accordingly, I suggest that we should legally redefine misuse of market power as market manipulation, to be

Forensic Assessment of Misuse of Market Power  231 assessed by reference to efficiency, which can be adapted and redefined for use in our legal theory in a way that is forensically determinable. I suggest that areas ‘b’ and ‘c’ in Figure 10.1 should thus be within the jurisdiction of the courts. If it were feasible for a firm to seek authorisation prior to implementing a decision, that course of action should, as the Review recommends, be open to it. However, we have seen in Chapters 5, 7 and 8 that zero-sum competition can produce inefficient outcomes, eg, price wars and bidding wars; that is, we can anticipate that there will be significant conduct under zero-sum competition that falls within area ‘a’ in Figure 10. So all our problems are not yet solved. We have seen that the mechanism by which supply responds to demand shifts, through the entry and exit of firms, under zero-sum competition requires firms to compete for survival. Some will succeed and others will not. Thus, there will be losses and gains that would not occur under perfect competition, where we assume that firms may freely enter and leave the market. We hope in the long run that gains will exceed losses, but this is not guaranteed under zero-sum competition. We can see that supply can be expected to respond to demand shifts under zero-sum competition, and we can examine whether the market is operating efficiently (ie, without distortion), though it does not work in the way we would expect under perfect competition. How do we deal with conduct falling within area ‘a’ in Figure 10.1? We have noted Tony Freyer’s intuition that anticompetitive conduct represents attempts by firms to manage the risks of competition.4 I suggest that this is true, but probably not in the way he means. Freyer is presumably referring to conduct which would fall outside area ‘a’, which I have labelled area ‘z’ in Figure 10.1. What conduct within areas ‘a’ and ‘z’ have in common is that both are inefficient and, as we have seen, the distinction we seek between normal competitive conduct and conduct that misuses market power to harm the process of competition is elusive. An example of conduct within area ‘z’ which appears to cause us no difficulty would be cartel conduct. I suggest that inefficient conduct within area ‘a’ is an inescapable consequence of the paradigm of zero-sum competition, in which firms compete for sales (or inputs) by taking sales (or inputs) away from each other and those that succeed survive at the expense of those that fail. This, I suggest, accounts for what we may call the bipolarity of competition law: we encourage firms to compete under zero-sum competition; however, when they do, and harm each other, we i­ ntuitively seek to condemn it, but we are confined by the language of competition law to characterise this as harm to the ‘process of competition’. This may explain why we discern from the cases reviewed in Part II a tendency for harm to competitors to be taken to represent harm to the process of competition.

4 T Freyer, Antitrust and Global Capitalism, 1930–2004 (Cambridge, Cambridge University Press, 2006) 394–95.

232  Findings and Recommendations I suggest that the better approach to dealing with conduct falling within area ‘a’ in Figure 10.1 would be to recognise that the conduct is inefficient. I suggest further that it is hard to regard the conduct as blameworthy when it is the natural and probable consequence of the chosen paradigm of competition. We have examined the inter-relationship between competition and efficiency specifically in relation to market power. I propose that the concept of market manipulation enables us to assess whether conduct which is not efficient should be condemned. It is not obvious why a firm would conduct itself in area ‘a’ unless it is locked in zero-sum competition in the form, eg, of a price war or bidding war for inputs. We have seen that persons on the other side of the market in these situations also have the power to manipulate the market, which both contributes to and benefits from resulting price effects. I suggest that inefficient conduct of the kind we saw in Chapters 5, 7 and 8 might be tolerated if it redistributes surplus away from the firm we suspect of misusing market power. Accordingly, we can say that conduct which is what we expect of firms in zero-sum competition should not be condemned unless it manipulates the market, ie, prevents supply and demand adjusting to each other. Such a case could be where a dominant firm causes a more efficient new entrant to exit or fail to enter in circumstances where there is unmet demand and the efficiency gains forgone exceed the exit costs of the incumbent firm.

E.  Conclusions Regarding Forensic Assessment of Particular Kinds of Conduct We can now review particular findings from the case studies examined in Part II. We are interested to see how the application of the Harper Review section 46 and the market manipulation approach can be forensically assessed, and how it may change the position applying under current law: (1) Refusal to supply and margin squeeze. We saw that, without resulting efficiency benefits, using section 46 to require a dominant firm to supply a competitor may simply result in a redistribution of surplus from the dominant firm to the competitor without any benefit to society. Evidence-based economic analysis is required. Nothing we have seen in the other cases examined in Part II gives us any confidence that it is possible to assess the price that the dominant firm should charge for access to its goods or services. Hence, it is not possible to determine when there is constructive refusal to supply. Nor is it possible to determine terms of supply that should be approved or to monitor changing market circumstances that might affect the ongoing economic analysis. We may doubt that such determinations are any more possible for administrative agencies than they are for courts. Any terms of supply that sought to protect the competitor’s margins would be likely to contravene the prohibition of price fixing (cartel conduct). The Harper Review section 46 as enacted, by omitting any reference to efficiency, may well be capable of being applied to deprive a dominant firm of the fruits of success, eg, by regulating perceived

Forensic Assessment of Misuse of Market Power  233 excessive prices. Even if the Harper recommended section 46(2) were enacted, the concept of market power on which it is based is not apt to secure efficient outcomes in cases raising refusal to supply and margin squeeze issues. By contrast, the market manipulation approach would ensure outcomes consistent with economic efficiency. This may support supply to a rival where there is unmet demand that could profitably be supplied. Remedies are problematic. However, I suggest that by focusing on efficiency, it is more likely that remedies could be framed to avoid distorting markets. There are a variety of feasible remedies should efficiency require it, eg, declarations as in ­Queensland Wire (which leave it to the parties to negotiate terms of access), supply at wholesale prices (where, as in Queensland Wire, the competitor seeks economies of scope arising from supply) or acquisition at wholesale prices (where, as in NT Power, the competitor seeks an outlet for its product, but may distort retail markets). (2) Predatory pricing. The main issue here is causation. We saw that a demand shift reduction triggers firms to compete for survival. Even under market equilibrium, new entry will trigger firms to compete for survival. Price wars may reduce price and increase output to inefficient levels. Evidence-based economic analysis is required because efficiency dictates which firms will survive. A predictive and qualitative SLC test may well not answer the questions that need to be addressed. The Harper section 46 as enacted, by omitting reference to efficiency, may well be capable of being applied to condemn limit pricing (ie, pricing above avoidable cost may be perceived to hinder or prevent new entry). Nothing we have seen in other cases examined in Part II gives us any confidence that we are able to assess the cost-price issues necessary to determine such cases. Even if the Harper recommended section 46(2) were enacted, the concept of market power on which it is based is not apt to analyse conduct under zero-sum competition. By contrast, the market manipulation approach would ensure outcomes consistent with economic efficiency. Mere harm to competitors would not be a basis to infer economic harm, which would be determined by an analysis of the market size and efficiency of competing firms. Inefficient conduct that is a necessary consequence of zero-sum competition should not be condemned, especially if the consequence is mitigated by a windfall transfer of surplus to consumers. (3) Meeting competition. The Harper section 46 SLC test comes with a presupposition that conduct of a dominant firm could be harmful by virtue of its market power. However, the element of causation under that test, and in particular whether a dominant firm will be able to respond to competitive conduct by a smaller firm, remains unclear. The potential discriminatory application of the new law might have been ameliorated by the recommended section 46(2), had it been enacted. However, for the reasons set out above, I suggest that the market manipulation approach should be preferred to avoid inefficient outcomes arising under the new law and to assess whether we can tolerate inefficient outcomes which are a consequence of zero-sum competition itself.

234  Findings and Recommendations (4) Raising rivals’ costs. Zero-sum competition may cause firms to engage in a bidding war for inputs. This may have the opposite effect to a price war, ie, there may be a redistribution of surplus up the vertical chain. End consumers may not necessarily be harmed; there may merely be a redistribution of surplus to the input supplier from the firm we suspect of misusing market power. Evidence-based economic analysis is required. A predictive and qualitative SLC test may well not answer the questions that need to be addressed. Nor would it appear that qualitative analysis of market power and competition effects based on the hypothetical competitive market standard would provide useful guidance on the distinction between normal conduct and harmful conduct under zero-sum competition. For the reasons set out above, I suggest that the market manipulation approach should be preferred to avoid inefficient outcomes arising under the new law and to assess whether we can tolerate inefficient outcomes which are a consequence of zero-sum competition itself. (5) Bundling. The analysis here is highly analogous to that applicable to predatory pricing, particularly in the case of bidding wars, so I will not repeat what is said above in relation to that issue. Suffice to say that the decision in Pfizer demonstrates the acute difficulties in zero-sum competition of grappling with the dual purposes problem. I suggest that the Pfizer decision more clearly presents us with a case of competition for survival. In each category of conduct, we have seen that the market manipulation approach enables us to adapt legal and economic concepts of efficiency for use in our legal theory. Accordingly, we can distinguish efficient conduct, which should not be condemned, and inefficient conduct, which should be condemned if it manipulates the market, preventing market forces operating to bring demand and supply into equilibrium. Conduct which is a consequence of the zero-sum paradigm of competition, ie, which results from firms doing what we expect them to do to compete (such as bidding to win), should not be regarded as manipulative. I suggest the data required to be forensically proven under the market manipulation approach to determine efficiency (as adapted and redefined in the proposed legal theory) should be no more difficult to obtain, and may well be easier to obtain, than cost-price data and analysis required under the SLC test which has been described as incalculable and incomprehensible.

V.  Implications for Competition Law We may critique the Harper Review recommended section 46 and the issues it leaves to be resolved by the courts. I suggest that the problem of deriving our concept of market power from the hypothetical competitive market leaves us uncertain about the causal nexus (if any) that may be required between market power, conduct and effects. I suggest that the Harper Review section 46 should still

Implications for Competition Law  235 require a causal nexus, but this question of interpretation remains to be decided by the courts. It would clearly be better if this issue had been explicitly addressed in the reform process. We could then discern whether it is Parliament’s intention to change the law to adopt a ‘no fault’, or strict liability, regime to regulate unilateral dominant firm conduct and, if so, we could fully consider the implications. Further, from our review of the leading Australian cases, we see that the determination of competition effects which result from market power remains fundamentally problematic at both the conceptual and forensic levels. If courts continue to assess effects under the predictive SLC test by qualitative inference, often from evidence of purpose, that may well ultimately undermine confidence in the law (as it appears to have done in the EU). We have seen some issues with the SLC test that may become more acute in the context of section 46. It is suggested that the following changes would be desirable to ameliorate these issues in the context of the Harper Review section 46 as enacted: (a) the substantial purpose test under section 4F should be replaced with a ‘sole purpose’ test to properly address the ‘dual purposes’ problem and avoid inferences being unduly influenced by the appearance of hostile intent (which is a consequence of zero-sum competition); (b) as ‘purpose’ and ‘effect’ are inter-related, the SLC test should require both purpose and effect, ie, we would then avoid the problem of ‘purpose without effect’ (which is adequately covered by the prohibition of ‘attempts’) and similarly we would avoid the problem of ‘effect without purpose’ (which arguably takes us into the area of a ‘no-fault’ or strict liability regime for unilateral dominant firm conduct). The express recognition in the Harper Review recommendation that efficiency benefits should be taken into account when considering the SLC test is welcome. The removal by Parliament of that requirement is unfortunate. How the courts may deal with this issue remains to be seen. Nevertheless, the Review’s recommended approach, of weighing up competition and efficiency effects, is regarded as problematic in the US and the EU. I suggest that considerations of efficiency should, by reason of being a higher-order objective of competition law, prevail over adverse competition effects. We have seen that economic concepts require adaptation for use in our legal theories if they are to be legally and forensically determinate. In the absence of clear legislative intervention, it is unlikely that the Harper section 46 can be applied by the courts to avoid inefficient outcomes. It is of particular concern that the Harper Review section 46 may have consequences not addressed or debated in the law reform process undertaken by the Review. Since Parliament omitted any reference to efficiency, the Harper section 46 may well be capable of being applied to condemn: (1) perceived excessive pricing (ie, dominant firms being rewarded for success or efficiency by charging prices and deriving profits above assumed competitive levels);

236  Findings and Recommendations (2) limit pricing (ie, pricing above avoidable cost that may be efficient, but nevertheless may hinder or prevent new entry). It is generally accepted that assessment of conduct of the above kinds is highly qualitative, lacking objective measures and prone to influence by unstated policy criteria. This may well result in the Harper Review section 46 tending to constrain the growth of large efficient firms. Further, the institutional division of functions between the courts and the administrative bodies, the ACCC and the ACT, is likely to mean that dominant firms are exposed to the risk of liability for contravention of the Harper Review section 46 where conduct may well be efficient. I suggest that this would be an undesirable outcome which would be insufficiently mitigated by guidelines issued by the ACCC. The Harper recommendation that conduct under section 46 can be authorised by the ACCC/ACT on the grounds of overriding public benefit is to be welcomed. However, this will be of little use to dominant firms making timesensitive business decisions where it is not feasible to delay conduct until after authorisation decisions and appeals have been concluded. Accordingly, if we are to avoid section 46 exposing firms to the risk of legal liability for conduct which may well be efficient, I suggest we need to go further than merely reinstating the Harper mandate to consider efficiency, and provide an effective affirmative defence for conduct which is efficient, as that concept can be adapted for use in our legal theory. On the whole, however, the traditional concept of market power that underpins the Harper Review section 46 is insufficiently determined, so that even if the changes suggested above were made, it would be a second-best solution. The law would still expose firms to the risk of liability for conduct that is encouraged by the very paradigm of competition that the law embraces, ie, zero-sum competition. We have seen that this paradigm of competition may result in inefficient outcomes, which make it doubly difficult to identify harmful outcomes that are caused by the conduct of firms as opposed to the paradigm itself. Accordingly, I suggest that the market manipulation approach is to be preferred as a basis for determining legal liability for a contravention of section 46 because it more effectually addresses the theory of harm underlying the prohibition of ‘misuse of market power’, ie, the efficient operation of markets. Conduct which is efficient can be considered not to be manipulative. Conduct which is inefficient may be manipulative, distorting the efficient operation of markets, unless it can be explained as a normal incident of zero-sum competition. That is, such conduct is not blameworthy, and the benefits of zero-sum competition (which include transfer of surplus to consumers or at least away from dominant firms) may well enable us to tolerate some inefficiency. I suggest that the paradigm of zero-sum competition enables us to reconceptualise market power in our legal theory as the power to manipulate the market and to develop norms of conduct, by reference to that paradigm and to efficiency, to assess dominant firm conduct. Thus, we now have a positive and normative legal theory of liability for misuse of market power. I suggest that the proposed

Implications for Competition Law  237 new legal theory should provide a better framework to assess market power and guide business conduct to distinguish unilateral dominant firm conduct which may be permitted from conduct which should be condemned because it is socially harmful. Unless fundamental concepts underlying section 46, and objectives of competition law, are clarified by the legislature or the courts, the cycle of enforcement failure and amendment experienced over the life of section 46 in its present and previous incarnations may well continue.

APPENDIX Text of the Old and New Versions of Section 46 of the CCA Old (or Pre-Harper) Section 46 (1) A corporation that has a substantial degree of power in a market shall not take advantage of that power in that or any other market for the purpose of: (a) eliminating or substantially damaging a competitor of the corporation or of a body corporate that is related to the corporation in that or any other market; (b) preventing the entry of a person into that or any other market; or (c) deterring or preventing a person from engaging in competitive conduct in that or any other market. (3) In determining for the purposes of this section the degree of power that a body corporate or bodies corporate has or have in a market, the court shall have regard to the extent to which the conduct of the body corporate or of any of those bodies corporate in that market is constrained by the conduct of: (a) competitors, or potential competitors, of the body corporate or of any of those bodies corporate in that market; or (b) persons to whom or from whom the body corporate or any of those bodies corporate supplies or acquires goods or services in that market. [Sub-sections (1AAA), (1AA), (1AB), (1A), (2), (3A), (3B), (3C), (4), (4A), (5), (6) and (6A) are not reproduced; see legislative history below for a summary.]

Appendix  239

Legislative History of the Old (or Pre-Harper) Section 46 Year

Substantive

Procedural

1974 Section 46 prohibits a corporation Section 46(7) provides that ‘purpose’ may be ‘in a position to substantially control a market’ from ‘taking advantage’ of the power ‘for the purpose’ of eliminating a competitor, preventing entry or deterring or preventing competitive conduct.

Section 4F provides that in the case of multiple purposes, one ‘substantial’ prohibited purpose will be sufficient. 1986 Section 46 is amended to lower the threshold test to require ‘a substantial degree of power in a market’ rather than control. The Attorney-General’s second reading speech states that large enterprises are not necessarily more efficient, and an effective section 46 should give small business ‘a measure’ of protection.

inferred from circumstantial evidence, ie, from the corporation’s conduct or other relevant circumstances. Conduct which contravenes section 46 cannot be authorised by the TPC (now the ACCC) or the Competition Tribunal.

Sub-sections (2), (3) and (4) added to ‘assist’ the courts in applying section 46: • sub-section (2) aggregates the power of related corporations; • sub-section (3) requires the court to ‘have regard to’ the extent to which the corporation’s conduct is constrained by the conduct of competitors or potential competitors or customers; • sub-section (4) clarifies that ‘power’ refers to ‘market power’ either as a supplier or buyer of goods or services.

1995 The Hilmer Report recommends

that, while section 46 may prohibit a corporation with substantial market power refusing competitors access, the courts are not the best mechanism to regulate access, eg, because a rational corporation would recoup lost income as a consequence of providing access, which might make it prohibitively expensive. Part IIIA is added to regulate access to essential facilities of national significance. The Hilmer Report recommends against an ‘effects test’ for section 46. (continued)

240  Appendix Year

Substantive The Dawson Committee considers 2003 whether the purpose test ought to be supplemented by adding an ‘effects test’. It recommends against any change because this would not help distinguish normal competitive conduct from behaviour that should be condemned.

Procedural

The ACCC loses two cases on section 46: Boral and Rural Press. The Senate Economics References Committee enquires into adequacy of section 46 to protect small business.

2004

The Senate Economics References Committee report recommends ‘clarifying’ the operation of section 46. Senator Barnaby Joyce conceives The Senate Economics References Committee 2007 of the following changes in the recommendations are implemented by the Birdsville pub (hence colloquially addition of sub-sections (3A)–(3D): known as the Birdsville • sub-section (3A) permits the court to have amendments). Sub-sections on regard to market power that results from predatory pricing are inserted: actual or proposed contracts, arrangements

• sub-section (1AA) prohibor understandings between the corporation its a corporation having a and other parties (note that QCMA market substantial ‘share of a market’ structure factors already cover this); supplying goods or services • sub-section (3C) permits the court to take for a ‘sustained period’ at less cognisance that a ‘substantial degree of than the relevant cost, for a power’ does not require that the corporation purpose prohibited by s 46; ‘substantially control’ the market or ‘have • sub-section (1B) permits the absolute freedom from constraint’. court to have regard to the Sub-section (3D) provides that more than one number and size of the corpo- corporation(s) may have a substantial degree of ration’s competitors in the power in a market. market. Sub-section (4A) was added to permit the (Repealed by Harper Review court to have regard to the conduct of a amendments.) corporation in supplying goods for a ‘sustained period’ at a price that is less than the ‘relevant cost’ to the corporation. (Apparently intended to assist the court determine whether price cutting is ‘predatory’; however, ‘relevant cost’ is not defined.) (continued)

Appendix  241 Year

Substantive

2008

Procedural Sub-section (6A) is added to clarify that the court may have regard, when considering ‘taking advantage,’ to conduct: • ‘materially facilitated’ by market power; • engaged in ‘in reliance’ on market power; • whether it would likely to be engaged in without market power; • ‘otherwise related’ to market power. Sub-section (1AAA) is added to provide that conduct under sub-section (4A) may contravene section 46, even if the corporation might never be able to recoup the loss. (Repealed by the Harper Review amendments.)

2013 The ACCC loses section 46 case

against Cement Australia (but succeeds in its section 45 case, ie, entering into an agreement for the purpose of SLC). The Harper Review recommends that the 2015 The ACCC loses another section 46 case, against Pfizer legislation should direct the court, when (under appeal). determining whether conduct has the purpose The Competition Policy Review or likely effect of SLC, to have regard to the extent to which the conduct has the purpose, (Harper Review) recommends effect or likely effect of: that section 46 be amended:

• to introduce an ‘effects test’, • increasing competition in the market, including by enhancing efficiency, ie, the section will prohibit a innovation, product quality or price corporation with a substantial competitiveness (subsequently rejected in degree of power in a market the Senate when opposed by the ACCC); from engaging in conduct which has the purpose or • lessening competition in the market, includlikely effect of SLC; ing by preventing, restricting or deterring the potential for competitive conduct in the • consequently, the ‘taking market or new entry into the market. advantage’ element is no longer required; Further recommendations of the Harper • the ‘predatory pricing’ Review to assist clarity: amendments of 2007 and 2008 • the ACCC (and on appeal the Tribunal) are unnecessary and can be should have the power to authorise conduct repealed. that has an overriding public benefit but would otherwise contravene section 46; The Harper Review’s final report was submitted in March 2015.

(continued)

242  Appendix Year

Substantive

Procedural • the ACCC should issue guidelines on its approach to enforcing section 46, prepared in consultation with business stakeholders, legal experts and consumer groups, and issued in advance of the commencement of the revised prohibition. 2017 An amending Bill to implement The Harper Review recommends Harper recommendations is sub-section (2) requiring that the court must amended after Senate committee consider pro-competitive and efficiencyreview in February 2017. enhancing effects removed by the Senate; modification of the Bill is accepted by the The amending Bill is passed by Government in March 2017. Parliament, and receives Royal Assent on 23 August 2017. Effective from 6 November 2017.

Harper Review Recommended Version of Section 46 (1) A corporation that has a substantial degree of power in a market shall not engage in conduct if the conduct has the purpose, or would have or be likely to have the effect, of substantially lessening competition in that or any other market. (2) Without limiting the matters that may be taken into account for the purposes of subsection (1), in determining whether conduct has the purpose, or would have or be likely to have the effect, of substantially lessening competition in a market, the court must have regard to: (a) the extent to which the conduct has the purpose, or would have or be likely to have the effect, of increasing competition in the market including by enhancing efficiency, innovation, product quality or price competitiveness in the market; and (b) the extent to which the conduct has the purpose, or would have or be likely to have the effect, of lessening competition in the market including by preventing, restricting or deterring the potential for competitive conduct in the market or new entry into the market. [Sub-sections (3), (4), (5), (6), and (7) are not reproduced; sub-section (4) corresponds with the old sub-section (3) set out above; sub-section (2) was removed from the amending Bill by Parliament.]

Appendix  243

Version of Section 46 Adopted by Parliament (1) A corporation that has a substantial degree of power in a market must not engage in conduct that has the purpose, or has or is likely to have the effect, of substantially lessening competition in: (a) that market; or (b) any other market in which that corporation, or a body corporate that is related to that corporation: (i) supplies goods or services, or is likely to supply goods or services; or (ii) supplies goods or services, or is likely to supply goods or services, indirectly through one or more other persons; or (c) any other market in which that corporation, or a body corporate that is related to that corporation: (i) acquires goods or services, or is likely to acquire goods or services; or (ii) acquires goods or services, or is likely to acquire goods or services, indirectly through one or more other persons. [Sub-sections (3), (4), (5), (6), (7) and (8) are not reproduced; sub-section (4) corresponds with the old sub-section (3) set out above; sub-section (2) is not used.]

BIBLIOGRAPHY ABC News, ‘Coal Reuse, Queensland Government Corporation Contractor, Facing Court Claims over Unpaid Debts’, 6 June 2016, www.abc.net.au/news/2016-06-06/queensland-governmentcorporation-contractor-facing-court-claim/7479214. Addy, G, Bodrug, J and Tingley, C, ‘Abuse of Dominance in Canada: Reflections on 25 Years of Section 79 Enforcement’ (2012) 25 Canadian Competition Law Review 276. Adelaide Brighton Ltd, ‘Adelaide Brighton Acquires a Number Two Position in the Australian Concrete Masonry Products Market’, press release, 19 June 2003, www.adbri.com.au/-/adbri/lib/pdfsarchived/cmrpmaquisition.pdf. Akman, P, The Concept of Abuse in EU Competition Law (Oxford, Hart Publishing, 2012). Allen, WB, Doherty, N, Weigelt, K and Mansfield, E, Managerial Economics, 6th edn (New York, WW Norton, 2005). American Bar Association, ‘Joint Comments of the American Bar Association Section of Antitrust Law and Section of International Law on the Australian Competition Policy Review Issues Paper’, 27 June 2014. Anti-Dumping Commission (Australia), ‘Analysis of Steel and Aluminium Markets’, August 2016. Antitrust Modernisation Commission (US), Report and Recommendations (2007). Ascher, B, ‘Global Beer: The Road to Monopoly’, American Antitrust Institute, November 2012, www. antitrustinstitute.org. Attorney-General’s National Committee to Study the Antitrust Law (US), Report (1955). Australian Competition and Consumer Commission, ‘Merger Guidelines’ November 2008. ——. Determination A91261, 14 July 2011, ACCC Authorisations Register, registers.accc.gov.au/ content/index.phtml/itemId/961019/fromItemId/401858. ——. Notification N97609 by Coal Reuse Pty Ltd, ACCC Notifications Register, 25 July 2014, registers. accc.gov.au/content/index.phtml/itemId/1180364/fromItemId/1133393. ——. ‘Submission to the Competition Policy Review – Response to the Draft Report’, 26 November 2014. ——. ‘Boral Limited Withdraws Takeover Bid, Provides Court Enforceable Undertakings’, press release, 8 October 2004, www.accc.gov.au/media-release/boral-limited-withdraws-takeover-bid-providescourt-enforceable-undertakings. ——. Statement of Reasons, Notification N97609, ACCC Notifications Register, 26 November 2015, registers.accc.gov.au/content/index.phtml/itemId/1180364/fromItemId/1133393. ——. Inquiry into the East Coast Gas Market, April 2016. ——. ‘ACCC Will Not Oppose AB InBev’s Acquisition of SABMiller’, press release, 5 May 2016, www.accc.gov.au/media-release/accc-will-not-oppose-ab-inbev%E2%80%99s-acquisition-ofsabmiller. ——. ‘ACCC Appeals Cement Australia Level of Penalties’, press release, 99/16, 6 June 2016. ——. ‘Framework for Misuse of Market Power Guidelines’, consultation draft, September 2016. ——. Submission to the Senate Economics Legislation Committee, February 2017. ——. ‘Interim Guidelines on Misuse of Market Power’, October 2017. ——. Digital Platforms Inquiry, 4 December 2017, www.accc.gov.au/focus-areas/inquiries/digitalplatforms-inquiry. ——. ‘ACCC Seeks Special Leave to Appeal Full Federal Court’s Pfizer Judgment’, press release, 115/18, 25 June 2018. Avgouleas, E, The Mechanics and Regulation of Market Abuse: A Legal and Economic Analysis (Oxford, Oxford University Press, 2005). Baker, J, ‘Exclusion as a Core Competition Concern’ (2013) Antitrust Law Journal 527.

Bibliography  245 Bakircioglu, O, ‘The Contours of the Right to Self-Defence’ (2008) 72 Journal of Criminal Law 131. Beaton-Wells, C and Fisse, B, Australian Cartel Regulation: Law, Policy and Practice in an International Context (Cambridge, Cambridge University Press, 2011). Bergþórsson, AF, What is Market Manipulation? An Analysis of the Concept in a European and Nordic Context (Leiden, Brill, 2018). Bernheim, D and Heeb, R, ‘A Framework for the Economic Analysis of Exclusionary Conduct’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 2 (Oxford, Oxford University Press, 2015). Berry, S and Pakes, A, ‘Classical Empirical Models of Static Equilibrium’ (2003), research paper, Yale University, pdfs.semanticscholar.org/67ed/868b9a1e2df4f95f9f9c7e41782ca5bde1cd.pdf. Besanko, D, Dranove, D, Shanley, M and Schaefer, S, Economics of Strategy, 7th edn (Hoboken, NJ, Wiley, 2016). Bishop, S and Walker, M, The Economics of EC Competition Law: Concepts, Application and Measurement, 3rd edn (London, Sweet & Maxwell, 2002). Blair, R and Carruthers, C, ‘The Economics of Monopoly Power in Antitrust’ in K Hylton (ed), Antitrust Law and Economics (Cheltenham, Edward Elgar, 2010). Blair, R and Sokol, D (eds), The Oxford Handbook of International Antitrust Economics, vols 1 and 2 (Oxford, Oxford University Press, 2015). Bork, R, The Antitrust Paradox: A Policy at War with itself (New York, Perseus Books, 1978), extracts from the Introduction and Chapters 4–6 reproduced in D Crane and H Hovenkamp (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013). Bork, R and Sidak, JG, ‘The Misuse of Profit Margins to Infer Market Power’ (2013) 9 Journal of Competition Law and Economics 511. Brunt, M, ‘The Australian Antitrust Law after 20 Years: A Stocktake’ (1994) 9 Review of Industrial Organisation 483. ——. Economic Essays on Australian and New Zealand Competition Law (The Hague, Kluwer Law International, 2003). Bumas, L, Intermediate Microeconomics: Neoclassical and Factually-Oriented Models (New York, Routledge, 2015). Burguet, R and Sákovics, J, ‘Competitive Foreclosure’, research paper, School of Economics, Edinburgh University, 15 February 2017, www.ed.ac.uk/files/atoms/files/279_-_competitive_foreclosure_ 2017.pdf. Caves, R, American Industry: Structure Conduct and Performance (Englewood Cliffs, Prentice Hall, 1967). Caves, R, Ward, I, Williams, P and Wright, C, Australian Industry: Structure Conduct and Performance (Sydney, Prentice Hall Australia, 1987). Church, J and Ware, R, Industrial Organisation (Boston, Irwin McGraw-Hill, 2000). Clark, A, ‘The Giant of the Miner League’, The Bulletin, 29 January 1980. Clark, JM, ‘Toward a Concept of Workable Competition’ (1940) 30 American Economic Review 241. Clarke, J, ‘Section 46: Its Purpose and the Proposed New Effects Test’ (2017) Australian Business Law Review 364. Coal Ash Matters, ‘Fly Ash Classifying & Load out Facility at Millmerran PS’, December 2008, www. adaa.asn.au/uploads/default/files/adaa-coal_ash_matters_dec_08.pdf. Coase, R, ‘Durability and Monopoly’ (1972) 15 Journal of Law and Economics 143. Commonwealth of Australia, National Competition Policy (Australian Government Publishing Service, 1993). ——. Minister for Small Business, ‘Government Names Competition Review Panel’, press release, 27 March 2014, bfb.ministers.treasury.gov.au/media-release/014-2014. ——. Competition Policy Review, Issues Paper, 14 April 2014. ——. Competition Policy Review, Final Report, March 2015. ——. Government Response to the Competition Policy Review, 18 November 2015. ——. Prime Minister, Treasurer and Minister for Small Business, ‘Competition Policy’, joint press release, 16 March 2016.

246  Bibliography ——. Senate Economics Legislation Committee, Report on the Competition and Consumer Amendment (Competition Policy Review) Bill 2016, February 2017. ‘Contractor Begins 10-Year Fly Ash Deal’, South Burnett News, 1 August 2014, southburnett.com.au/ news2/2014/08/contractor-begins-10-year-fly-ash-deal. Corones, S, Competition Law in Australia, 6th edn (Sydney, Lawbook Co, 2014). Country Press SA, ‘Murray Valley Standard’, www.sacountrypress.com.au/newspaper-lists/murraybridge. ——. ‘River News’, www.sacountrypress.com.au/newspaper-lists/the-river-news. Crampton, P, ‘“Abuse” of “Dominance” in Canada: Building on the International Experience’ (2005) 73 Antitrust Law Journal 803. Crane, D, ‘Market Power without Market Definition’ (2014) 90 Notre Dame Law Review 31. Crane, D and Hovenkamp, H (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013). Crawford, L, ‘Can Parliament Confer Plenary Executive Power? The Limitations Imposed by Sections 51 and 52 of the Australian Constitution’ (2016) 44 Federal Law Review 287. Cross, J, Richards, D, Stucke, M and Waller, S, ‘Use of Dominance, Unlawful Conduct, and Causation under Section 36 of the New Zealand Commerce Act: A US Perspective’ (2013) New Zealand Business Law Quarterly, ssrn.com/abstract=2170538. Cudahy, R and Devlin, A, ‘Anticompetitive Effect’ (2010) Minnesota Law Review 59. Demura, P, ‘Productivity Change in the Australian Steel Industry: BHP Steel 1982–1995’ in Productivity and Growth (Reserve Bank of Australia, 1995), www.rba.gov.au/publications/confs/1995/pdf/ demura.pdf. Department of Justice (US), ‘Competition and Monopoly: Single-Firm Conduct under Section 2 of the Sherman Act’ (2008). ——. ‘Justice Department Withdraws Report on Antitrust Monopoly Law’, press release, 11 May 2009. Department of Justice and Federal Trade Commission (US), ‘Horizontal Merger Guidelines’, 19 August 2010. Devlin, A and Jacobs, M, ‘Antitrust Divergence and the Limits of Economics’ (2010) 104 Northwestern University Law Review 270. Donald, B and Heydon, JD, Trade Practices Law, vol 1 (Sydney, Lawbook Co, 1978). Dooley, P, Elementary Price Theory (New York, Appleton-Century-Crofts, 1967). Drexl, J ‘On the (A)political Character of the Economic Approach’ in J Drexl, W Kerber and R Podszun (eds), Competition Policy and the Economic Approach: Foundations and Limitations (Cheltenham, Edward Elgar, 2011). Dreyfuss, RC, Zimmerman, DL and First, H (eds), Expanding the Boundaries of Intellectual Property (Oxford, Oxford University Press, 2001). Duns J and Duke A, Competition Law Cases and Materials, 4th edn (Chatswood, LexisNexis Butterworths, 2015). Eichner, A and Kregel, JA, ‘An Essay on Post-Keynesian Theory’ (1975) 13 Journal of Economic Literature 1293. European Commission, ‘Guidance on the Commission’s Enforcement Priorities in Applying Article 82 EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings’ [2009] OJ C45/7. ——. Directive 2014/104/EU, ‘Damages for Infringement of Competition Law Provisions’ [2014] OJ L349. ——. ‘Commission Approves Acquisition of Swedish Medical Technology Company Gambro by US Rival Baxter, Subject to Conditions’, press release, 22 July 2013, europa.eu/rapid/press-release_ IP-13-724_en.htm. ——. ‘Antitrust: Commission Opens Formal Investigation into Aspen Pharma’s Pricing Practices for Cancer Medicines’, press release, 15 May 2017, europa.eu/rapid/press-release_IP-17-1323_ en.htm. ——. ‘Google Fined €2.42 Billion for Abusing its Market Dominance as a Search Engine’, press release, 27 June 2017, europa.eu/rapid/press-release_IP-17-1784_en.htm.

Bibliography  247 ——. ‘Google Fined €4.34 billion for Abusing its Market Dominance by Placing Restrictions on Android Device Manufacturers and Network Operators’, press release, 18 July 2018, europa.eu/ rapid/press-release_IP-18-4581_en.htm. Evans, D and Hylton, K, ‘The Lawful Acquisition and Exercise of Monopoly Power and its Implications for the Objectives of Antitrust’ (2008). 4 Competition Policy International 203, ssrn.com/ abstract=1275431. Evans, D and Schmalensee, R, ‘The Antitrust Analysis of Multisided Platform Businesses’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 1 (Oxford, Oxford University Press, 2015). Evans, M, Equity and Trusts (Chatswood, LexisNexis Butterworths, 2012). Ezrachi, A (ed), Research Handbook on International Competition Law (Cheltenham, Edward Elgar, 2012). Facey, B and Assaf, D, ‘Monopolization and Abuse of Dominance in Canada, the United States, and the European Union: A Survey’ (2002) 70 Antitrust Law Journal 513. Farrell, J and Katz, M, ‘The Economics of Welfare Standards in Antitrust’ (2006) Competition Policy International, available at escholarship.org/uc/item/1tw2d426. Feinstein, R, ‘2010 Revisions to the US Horizontal Merger Guidelines’ (2011) 7 Competition Law International 6. Fels, A, ‘The Past and Future of Competition Law’ (2004) 23 Economic Papers 3. ——. ‘The Competition Review’, speech at the University of Melbourne, 19 May 2014. Finlay, L and Kirchengast, T, Criminal Law in Australia (Chatswood, LexisNexis Butterworths, 2015). Finn, C, ‘The Justifiability of Administrative Decisions: A Redundant Concept? (2002) 2 Federal Law Review 239. ‘Fly Ash Firm in Liquidation’, South Burnett News, 6 September 2016, southburnett.com.au/ news2/2016/09/fly-ash-firm-in-liquidation. Foer, A, ‘On the Inefficiencies of Efficiency as the Single-Minded Goal of Antitrust’ (2015) 60 Antitrust Bulletin 103. French, R, ‘The Role of the Courts in the Development of Australian Trade Practices Law’ in F Hanks and P Williams (eds), The Trade Practices Act: A Twenty-Five Year Stocktake (Sydney, Federation Press, 2001). ——. ‘Declarations – Homer Simpson’s remedy – Is There Anything They Cannot Do?’ (2007) Federal Judicial Scholarship 24. ——. ‘Dolores Umbridge and the Concept of Policy as Legal Magic’ (2007) Federal Judicial Scholarship 14. Freyer, T, Antitrust and Global Capitalism, 1930–2004 (Cambridge, Cambridge University Press, 2006). Frontier Economics, ‘Competition Issues’, 26 March 2014, Expert’s Report prepared for AGL in AGL/MacGen [2014] ACompT 1. Gal, MS, Competition Policy for Small Market Economies (Cambridge, MA, Harvard University Press, 2003). Gans, J and King, S, ‘Paying for Loyalty: Product Bundling in Oligopoly’ (2006) 54 Journal of Industrial Economics 43. Gans, J, King, S, Byford, M and Mankiw, NG, Principles of Microeconomics, 7th edn (Melbourne, Cengage Learning, 2018). Gans, J, Sood, R and Williams, P, ‘The Decision of the High Court in Rural Press: How the Literature on Credible Threats May Have Materially Facilitated a Better Decision’ (2004) 32 Australian Business Law Review 337. Gavil, A, ‘Burden of Proof in US Antitrust Law’ (2008) 1 Issues in Competition Law and Policy 125. ——. ‘Imagining a Counterfactual Section 36: Rebalancing New Zealand’s Competition Law Framework’ (2015) 46 Victoria University of Wellington Law Review 1043. Geradin, D, Layne-Farrar, A and Petit, N, EU Competition Law and Economics (Oxford, Oxford University Press, 2012).

248  Bibliography Gerber, D, Global Competition: Law, Markets, and Globalisation (Oxford University Press, 2010). Gibbs, S, ‘Google Maps: A Decade of Transforming the Mapping Landscape’, The Guardian, 8 February 2015, www.theguardian.com/technology/2015/feb/08/google-maps-10-anniversary-iphone-androidstreet-view. Global Custodian, ‘A Pont Too Far’, 1 March 1992, www.globalcustodian.com/Magazine/1992/March-/ A-Pont-too-far. Lovdahl Gormsen, L, A Principled Approach to Abuse of Dominance (Cambridge, Cambridge University Press, 2010). ——. ‘Are Anti-competitive Effects Necessary for an Analysis under Article 102 TFEU?’, (2013) 36 World Competition 223. Griggs, L, ‘Can a Corporation No Longer Maximise Profits – Misuse of Market Power and the Decision in Melway v Hicks’ (2000) 19 University of Tasmania Law Review 145. Griggs, L and Sharp, E, ‘The Homogeneity of Object in Part IV – Consumer Sovereignty as the Answer?’ (2003) 11 Competition and Consumer Law Journal 1. Gummow, WMC, ‘Equity: Too Successful?’ (2003) 77 Australian Law Journal 30. Gundlach, G and Moss, D, ‘The Role of Efficiencies in Antitrust Law: Introduction and Overview’ (2015) 60 Antitrust Bulletin 91. Hanks, F and Williams, P, ‘The Treatment of Vertical Restraints’ (1987) 15 Australian Business Law Review 147. ——. ‘Implications of the Decision of the High Court in Queensland Wire’ (1990) 17 Melbourne University Law Review 437. Hay, G, ‘Practices that Facilitate Cooperation’ in J Kwoka and L White (eds), The Antitrust Revolution (Glenview, Scott, Foresman & Co, 1989). Hay, G and Smith, R, ‘“Why Can’t a Woman Be More Like a Man?” – American and Australian Approaches to Exclusionary Conduct’ (2007) 31 Melbourne University Law Review 1099. Heerey, P, ‘Shootout at the Real Chance Café: Aiming at s 50’ (2011) 39 Australian Business Law Review 457. Heidrich, C, ‘Ash Utilisation – An Australian Perspective’, conference paper, 23 September 2003, www.flyash.info/2003/03heid.pdf. Heydon, JD, Trade Practices Law (Sydney, Lawbook Co, 1989). Holcim, ‘Holcim Confirms Merger of its Australian Interests into the Newly Formed Cement Australia Pty Ltd’, press release, 2 June 2003, www.holcim.com/media-relations/latest-releases/ latest-release/article/holcim-confirms-merger-of-its-australian-interests-into-the-newly-formedcement-australia-pty-ltd.html. ——. ‘Acquisition of Cemex Australian Successfully Completed’, press release, 1 October 2009, www.holcim.com.au/about-us/latest-media-releases/latest-release/article/acquisition-of-cemexaustralia-successfully-completed.html. ——. ‘Holcim and Heidelberg Cement Intend Joint Control of Cement Australia’, press release, 10 July 2015, www.holcim.com/media-relations/latest-releases/latest-release/article/holcim-andheidelbergcement-intend-joint-control-of-cement-australia.html. ——. ‘Holcim and Lafarge Merge to Create LafargeHolcim’, press release, 10 July 2015, www.holcim. com/media-relations/latest-releases.html. Hovenkamp, H, ‘Post-Chicago Antitrust: A Review and Critique’ (2001) Columbia Business Review, reproduced in D Crane and H Hovenkamp (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013). ——. Federal Antitrust Policy (St Paul, West Publishing, 2005). ——. Antitrust Enterprise: Principle and Execution (Cambridge, MA, Harvard University Press, 2005). ——. ‘Schumpeterian Competition and Antitrust’ (2008) University of Iowa College of Law, research paper, ssrn.com/abstract=1275986. ——. ‘Roger Blair and the Goals of Antitrust’ (2016) 61 Antitrust Bulletin 382. ——. ‘Reimagining Antitrust: The Revisionist Work of Richard Markovits’ (2016) 94 Texas Law Review 1221.

Bibliography  249 Hylton, K, Antitrust Law: Economic Theory and Common Law Evolution (Cambridge, Cambridge University Press, 2003). ——. ‘The Law and Economics of Monopolization Standards’ in K Hylton (ed), Antitrust Law and Economics (Cheltenham, Edward Elgar, 2010) 82. ——. (ed), Antitrust Law and Economics (Cheltenham, Edward Elgar, 2010). ——. ‘Antitrust Enforcement Regimes: Fundamental Differences’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 1 (Oxford, Oxford University Press, 2015). Huffman, M, ‘Marrying Neo-Chicago with Behavioural Antitrust’ (2012) 78 Antitrust Law Journal 105. Independent Hospital Pricing Authority (Australia), ‘Home Delivered Dialysis Costing Study to Inform the National Efficient Price 2015’, report prepared by PricewaterhouseCoopers, 11 November 2014, www.ihpa.gov.au/sites/g/files/net636/f/publications/hdd_report_for_ihpa.pdf. Kaplow, L, ‘On the Relevance of Market Power’ (2017) 130 Harvard Law Review 1303. Karmel, PH and Brunt, M, The Structure of the Australian Economy (Melbourne, FW Cheshire, 1966). Kaysen, C and Turner, D, Antitrust Policy: An Economic and Legal Analysis (Cambridge, MA, Harvard University Press, 1959). Kelliher, J, ‘Market Manipulation, Market Power, and the Authority of the Federal Energy Regulatory Commission’ (2005) 26 Energy Law Journal 1. Kemp, K, ‘Uncovering the Roots of Australia’s Misuse of Market Power Provision: Is it Time to Reconsider?’ (2014) 42 Australian Business Law Review 329. ——. ‘Taking Advantage of Substantial Market Power, and Other Profit-Focussed Tests for Unilateral Anticompetitive Conduct’ (2015) 41 Monash University Law Review 655. ——. ‘The Big Chill? A Comparative Analysis of Effects-Based Tests for Misuse of Market Power’ (2017) 40 University of New South Wales Law Journal 493. ——. ‘Misuse of Market Power in Australia and Abuse of Dominance in Canada: Two Legislated Effects Tests for Unilateral Conduct’ (2018) 26 Australian Journal of Competition and Consumer Law 174. ——. Misuse of Market Power: Rationale and Reform (Cambridge, Cambridge University Press, 2018). Kirkwood. J, ‘The Essence of Antitrust: Protecting Consumers and Small Suppliers from Anticompetitive Conduct’ (2013) 81 Fordham Law Review 2425. Kobayashi. B, ‘The Economics of Loyalty Discounts and Antitrust Law in the United States’ (2005) George Mason University, ssrn.com./abstract_id=794944. Kovacic. W, ‘The Intellectual DNA of Modern US Competition Law for Dominant Firm Conduct: The Chicago/Harvard Double Helix’ (2007) Columbia Business Law Review 1. Krämer, A, Jung, M and Burgatz, T, ‘A Small Step from Price Competition to Price War: Understanding Causes, Effects and Possible Countermeasures’ (2016) 9 International Business Research 1, dx.doi. org/10.5539/ibr.v9n3p1. Landes, W and Posner, R, ‘Market Power in Antitrust Cases’ (1981) 94 Harvard Law Review 937. Landolt, P, Modernised EC Competition Law in International Arbitration (The Hague, Kluwer Law International, 2006). Lannin, S, ‘Fortescue Metals Boosts Production by 14pc’, ABC News, 27 July 2016, www.abc.net.au/ news/2016-07-27/fortescue-metals-boosts-production-by-14pc/7664322. Lipman, M, ‘Intel to Get Answer on €1.1B EU Fine Appeal in Sept’, Law 360, 30 June 2017, www. law360.com/articles/940358. Lipsey, RG, An Introduction to Positive Economics, 2nd edn (London, Weidenfeld & Nicolson, 1963). Lipsey, RG and Chrystal, A, An Introduction to Positive Economics, 8th edn (Oxford, Oxford University Press, 1995). Lipsey, RG and Lancaster, K, ‘The General Theory of Second Best’ (1956) 24 Review of Economic Studies 11. Livingston, M, ‘Liquor Licensing: Research and Policy’, Turning Point Drug and Alcohol Centre, presentation to the Drug Policy Modelling Program Symposium, 16 March 2012.

250  Bibliography MacCormick, N, Rhetoric and the Rule of Law (Oxford, Oxford University Press, 2005). McMahon, K, ‘Competition Law, Adjudication and the High Court’ (2006) 30 Melbourne University Law Review 782. Marchionatti, R, ‘On the Methodological Foundations of Modern Microeconomics: Frank Knight and the “Cost Controversy” in the 1920s’ (2003) 35 History of Political Economy 49. Markovits, R, ‘Second-Best Theory in Law and Economics’ (1998) Chicago-Kent Law Review 73. Mason, A, ‘Law and Economics’ (1991) 17 Monash University Law Review 167. ——. ‘The Pursuit of Excellence in Tribunal Decision-Making in Australia, the United Kingdom and Canada’ (2016) 29 Canadian Journal of Administrative Law and Practice 235. Mason, E, ‘The Current Status of the Monopoly Problem in the United States (1949) 62 Harvard Law Review 1280. ——. ‘Market Power and Business Conduct: Some Comments’ (1956) 46 American Economic Review 471. Maughan, C, Maughan, M and Thornhill, A, ‘Confronting Adversarial Attitudes to Negotiation: What the Red/Blue Exercise Can Teach Us and Our Students’ (1998) 32 Law Teacher 79. Meese, A, ‘Reframing Antitrust in Light of Scientific Revolution: Accounting for Transaction Costs in Rule of Reason Analysis’ (2010) 62 Hastings Law Journal 457. Melway, ‘History’, www.melway.com.au/melway-history. ——. ‘Internet Directory’, www.street-directory.com.au/sd_new/home.cgi. Merrett, A, ‘The Assessment and Regulation of Market Power in Australia’, doctoral dissertation (University of Melbourne, 2010). ——. ‘Are Shopper Dockets Anti-competitive? Good Luck Proving it’, The Conversation, 5 September 2013, theconversation.com/are-shopper-dockets-anti-competitive-good-luck-proving-it-17499. ——. ‘Understanding Market Power’ in J Duns, A Duke and B Sweeney (eds), Comparative Competition Law (Cheltenham, Edward Elgar, 2015). Miller, R, ‘What if … The Ongoing Australian Struggle with Misuse of Marker Power’ (2016) 24 Australian Journal of Competition and Consumer Law 242. Ministry of Business, Innovation and Employment (New Zealand), ‘Targeted Review of Commerce Act 1986’, www.mbie.govt.nz/info-services/business/competition-policy/targeted-review-of-thecommerce-act/phase-one. Motta, M, Competition Policy: Theory and Practice (Cambridge, Cambridge University Press, 2004). Muris, T, ‘Improving the Economic Foundations of Competition Policy’ (2003) 12 George Mason Law Review 1. Nagarajan, V, ‘The Paradox of Australian Competition Policy: Contextualising the Coexistence of Economic Efficiency and Public Benefit’ (2013) 36 World Competition 133. Nargundkar, S, Karakaya, F and Stahl, M, ‘Barriers to Market Exit’ (1996) 13 Journal of Managerial Issues 239. Nelson, P and Smith, D, ‘Efficiencies in Antitrust Analysis: A View from the Middle of the Road’ (2015) 60 Antitrust Bulletin 128. Niblett, A, Gans, J and King, S, ‘Structural and Behavioural Market Power under the Trade Practices Act: An Application to Predatory Pricing’ (2004) 32 Australian Business Law Review 83. Niels, G, Jenkins, H and Kavanagh, J, Economics for Competition Lawyers (Oxford, Oxford University Press, 2011). NxStage, ‘Fresenius Medical Care to Acquire NxStage Medical, Inc. to Strengthen its Vertically Integrated Dialysis Business’, press release, 7 August 2017, ir.nxstage.com/releasedetail.cfm? releaseid=1036185. O’Bryan, M, ‘Section 46: Law or Economics? (1993) 1 Competition & Consumer Law Journal 64. ——. ‘Section 46: Legal and Economic Principles and Reasoning in Melway and Boral’ (2001) 8 Competition & Consumer Law Journal 1. Office of Fair Trading (report prepared by National Economic Research Associates), ‘The Role of Market Definition in Monopoly and Dominance Enquiries’, Economic Discussion Paper 2, July 2001.

Bibliography  251 OECD, Reference Guide on Ex-post Evaluation of Competition Agencies’ Enforcement Decisions, April 2016, www.oecd.org/daf/competition/Ref-guide-expost-evaluation-2016web.pdf. Oxford Economic Research Associates, Costs and Benefits of Market Regulators, Parts 1 and 2, report for the Dutch Ministry of Economic Affairs, October 2004. Pace, L (ed), Competition Law: The Impact of the Commissions’ Guidance on Article 102 (Cheltenham, Edward Elgar, 2011). Pengilley, W, ‘Queensland Wire and its Progeny Decisions: How Competent are the Courts to Determine Supply Prices and Trading Conditions?’ (1991) 21 Western Australian Law Review 225. Perriello, B, ‘Fresenius Extends Deadline for $2B NxStage Buyout’, Mass Device, 31 July 2018, www. massdevice.com. Perloff, J, Karp, L and Golan, A, Estimating Market Power and Strategies (Cambridge, Cambridge University Press, 2007). Perloff, J, Smith, R and Round, D, Microeconomics (Frenchs Forest, Pearson Australia, 2014). Pirrong, C, ‘Commodity Market Manipulation Law: A (Very) Critical Analysis and a Proposed Alternative’ (1994) 51 Washington & Lee Law Review 945. Portland Cement Association, ‘Optimizing the Use of Fly Ash in Concrete’, 6 December 2007, www. cement.org/docs/…concrete…/is548-optimizing-the-use-of-fly-ash-concrete.pdf. Posner, R, Antitrust Law, 2nd edn (Chicago, University of Chicago Press, 2001). ——. Economic Analysis of Law, 7th edn (New York, Aspen Publishers, 2007). Power and Water Authority, ‘Interim Northern Territory Electricity Market’, 27 May 2015, www. powerwater.com.au/networks_and_infrastructure/market_operator. Productivity Commission (Australia), Final Report – National Access Regime, 25 October 2013. Quo, S, ‘Interpretation and Application of the Purpose Test in s 46 of the Competition and Consumer Act 2010 – Part 1’ (2011) 19 Competition and Consumer Law Journal 90. Raitt, G, ‘Taking Advantage of Market Power: The Competitive Market Hypothesis’ (1999) 73 Law Institute Journal 70. ——. ‘Insights for Legal Reasoning from Studies of Literary Adaptation and Intertextuality’ (2013) 18 Deakin Law Review 191. ——. ‘Misuse of Market Power: Why Policy Objectives Matter’ (2014) 22 Competition & Consumer Law Journal 1. ——. ‘Changing Views Concerning “Competition Effects” in the US, EU and Australia: the Competition Policy Review’, presentation at the Australasian Law Teachers Association Conference, 16–18 July 2015. ——. ‘The More Things Change …’ (2015) 1 Short Supply 27. ——. ‘Competition and Efficiency Effects in Europe, North America and Australia’ (2016) 24 Competition & Consumer Law Journal 187. ——. ‘Flight Centre v ACCC: Thought-Crimes and Legal Fictions’ (2016) 27 International Company and Commercial Law Review 35. ——. ‘Shifting the Goalposts: Current Issues in Australian Competition Law Affecting the Energy Sector’ (2016) 9 Journal of World Energy Law and Business 424. Redmond, W, ‘Three Modes of Competition in the Marketplace’ (2013) 72 American Journal of Economics and Sociology 423. Reid, W, ‘Aspen Skiing in a Sunburnt Country’ (2005) 73 Antitrust Law Journal 209. Reuters, ‘Heidelberg Cement Agrees to Buy Hanson’, 15 May 2007, www.reuters.com/article/us-hansonheidelberg-idUSL1530198520070515. Richards, B, de Zwart, M and Ludlow, K, Tort Law Principles (Melbourne, Lawbook Co, 2013). Richardson, M and Hadfield, G (eds), The Second Wave of Law and Economics (Sydney, Federation Press, 1999). Robinson, EAG, Monopoly (Cambridge, Cambridge University Press, 1941). Rosario, M, Torre, A, Mendelson, M and Griggs, L, ‘Software Developer Policies: Economics and s 46 of the Competition and Consumer Act 2010’ (2011) 21 Journal of Law, Information and Science 37. Rosen, H, Public Finance (Boston, Irwin McGraw-Hill, 1995).

252  Bibliography Rudaz, C, ‘Did Trinko Really Kill Antitrust Price Squeeze Claims? A Critical Approach to the Linkline Decision through a Comparison of EU and US Case Law’ (2010) 43 Vanderbilt Journal of Transnational Law 107. Salinger, M, ‘Vertical Mergers’ in Blair R and Sokol D (eds), The Oxford Handbook of International Antitrust Economics, vol 1 (Oxford, Oxford University Press, 2015). Salvadori, N and Signorino, R, ‘The Classical Notion of Competition Revisited’ (2013) 45 History of Political Economy 149. Savov, V, ‘BlackBerry’s Success Led to its Failure’, The Verge, 30 September 2016, www.theverge. com/2016/9/30/13119924/blackberry-failure-success. Scheffman, D and Higgins, R, ‘Raising Rivals’ Costs’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 2 (Oxford, Oxford University Press, 2015). Schroeder, D, ‘Normative and Institutional Limitations to a More Economic Approach’ in J Drexl, W Kerber and R Podszun (eds), Competition Policy and the Economic Approach: Foundations and Limitations (Cheltenham, Edward Elgar, 2011). Schinkel, M and Larouche, P, ‘Continental Drift in the Treatment of Dominant Firms: Article 102 TFEU in Contrast to Section 2 Sherman Act’, in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 2 (Oxford, Oxford University Press, 2015). Scherer, FM and Ross, DR, Industrial Market Structure and Economic Performance, 3rd edn (Boston, Houghton Mifflin, 1990). ——. ‘The Innovation Lottery,’ in RC Dreyfuss, DL Zimmerman and H First (eds), Expanding the Boundaries of Intellectual Property (Oxford, Oxford University Press, 2001). ——. ‘Joseph Schumpeter’ (2007) in H de Jong and W Shepherd (eds), Pioneers of Industrial Organisation (Cheltenham, Edward Elgar, 2007). ——. ‘Conservative Economics and Antitrust: A Variety of Influences’ (2008), originally published R in Pitofsky (ed), How the Chicago School Overshot the Mark (Oxford, Oxford University Press, 2008), reprinted in D Crane and H Hovenkamp (eds), The Making of Competition Policy: Legal and Economic Sources (Oxford, Oxford University Press, 2013). Schmidt, H, ‘Market Power – The Root of All Evil? A Comparative Analysis of the Concepts of Market Power, Dominance and Monopolisation’ in A Ezrachi (ed), Research Handbook on International Competition Law (Cheltenham, Edward Elgar, 2012). ——. ‘Taming the Shrew: There’s No Need for a New Market Power Definition for the Digital Economy’, paper presented at the Academic Society for Competition Law conference, June 2017, ssrn.com/ abstract=3048266. Schumpeter, J, Capitalism, Socialism and Democracy (London, George Allen & Unwin, 1976). Shelanski, H, ‘Efficiency Claims and Antitrust Enforcement’, in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 1 (Oxford, Oxford University Press, 2015). ‘Shutdown at Swanbank B’, Queensland Times, 25 May 2012, www.qt.com.au/news/shutdown-atswanbank-b/1392591. Sidak, JG, ‘Abolishing the Price Squeeze as a Theory of Antitrust Liability’ (2008) 4 Journal of Competition Law and Economics 279. Sidak, JG and Teece, D, ‘Dynamic Competition in Antitrust Law’ (2009) Journal of Competition Law and Economics, available at ssrn.com/abstract=1479874. Simon, D, ‘Incumbent Pricing Responses to Entry’ (2005) 26 Strategic Management Journal 1229. Sims, R, ‘Is Australia’s Economy Getting More Concentrated and Does This Matter?’, speech delivered at the RBB Economics Conference, Sydney, 27 October 2016, www.accc.gov.au/speech/keynoteaddress-rbb-economics-conference-0. Smith, R, ‘Market Definition: Going, Going, Gone? Developments in the United States’ (2010) 18 Competition and Consumer Law Journal 119. ——. ‘Market Definition and Substitution Options’ (2014) 22 Competition & Consumer Law Journal 105. Smith, R and Merrett, A, ‘Unwrapping a Fallacy: Market Definition, Market Power & Cellophane’ (2014) 18 Competition 1.

Bibliography  253 Smith, R and Round, D, ‘A Strategic Behaviour Approach to Evaluating Competitive Conduct’ (1998) 5 Agenda 25. ——. ‘Section 46: Oligopoly and Predatory Pricing’ (1998) 6 Competition & Consumer Law Journal 1. Solum, L, ‘Indeterminacy’ in D Patterson (ed), A Companion to Philosophy of Law and Legal Theory (Oxford, Wiley-Blackwell, 2010). Steinwall, R, ‘The Dawson Committee on Competition, Economic Efficiency, Universality and Public Policy’ (2003) 26 University of New South Wales Law Journal 226. Stephens, A, ‘The Melways is Turning 50. But Will Melbourne’s Back-Seat Bible Survive the GPS?’, Sydney Morning Herald, 22 April 2016, www.smh.com.au/entertainment/the-melways-is-turning50-but-will-melbournes-backseat-bible-survive-the-gps-20160413-go5cvn.html. Stroux, S, US and EC Oligopoly Control (The Hague, Kluwer Law International, 2004). Stucke, M, ‘Reconsidering Antitrust’s Goals’ (2012) 53 Boston College Law Review 551. Sullivan, M, ‘A Brief History of GPS’, PC World, 9 August 2012, www.pcworld.com/article/2000276/abrief-history-of-gps.html. Sunstein, C, ‘Interpreting Statutes in the Regulatory State’ (1989) 103 Harvard Law Review 405. Sutton, J, Sunk Costs and Market Structure (Cambridge, MA, MIT Press, 1991). Tarong Energy, Submission to the ACCC, Authorisation A91261, 5 May 2011, ACCC Authorisations Register, registers.accc.gov.au/content/index.phtml/itemId/961019/fromItemId/401858. Thibault, M, ‘NxStage Medical Details New Features, Hints at Peritoneal Dialysis System’, Device Talk, 8 August 2016, www.mddionline.com/blog/devicetalk/nxstage-medical-details-new-features-hintsperitoneal-dialysis-system-08-08-16. Trebilcock, M, ‘The Value and Limits of Law and Economics’ in M Richardson and G Hadfield (eds), The Second Wave of Law and Economics (Sydney, Federation Press, 1999). Trindade, R and Smith, R, ‘Modernising Australian Merger Analysis’ (2007) 35 Australian Business Law Review 358. Vickers, J, ‘Competition Law and Economics: A Mid-Atlantic Viewpoint’ (2007) 3 European Competition Journal 1. Wang, Z, ‘Supermarkets and Gasoline: An Empirical Study of Bundled Discounts’ (2011) 5th Annual Conference on Empirical Legal Studies, ssrn.com/abstract=1628770. Whish, R, ‘National Competition Law Goals and the Commission’s Guidance on Article 82 EC: The UK experience’ in L Pace (ed), Competition Law: The Impact of the Commissions’ Guidance on Article 102 (Cheltenham, Edward Elgar, 2011). White, L ‘Monopoly and Dominant Firms: Antitrust Economics and Policy Approaches’ in R Blair and D Sokol (eds), The Oxford Handbook of International Antitrust Economics, vol 1 (Oxford, Oxford University Press, 2015). Williams, CR, ‘Burdens and Standards in Civil Litigation’ (2003) 25 Sydney Law Review 165. Williams, P, ‘The Counterfactual Test in s 46’ (2013) 41 Australian Business Law Review 93. Williamson, O, ‘Economies as an Antitrust Defense: The Welfare Tradeoffs’ (1968) 58 American Economic Review 18. Witt, AC, The More Economic Approach to EU Antitrust Law (Oxford, Hart Publishing, 2016). Wright, J, ‘The Antitrust/Consumer Protection Paradox: Two Policies at War with Each Other’ (2012) 121 Yale Law Journal 2216. ——. ‘Simple But Wrong or Complex and More Accurate? The Case for an Exclusive Dealing Approach to Evaluating Loyalty Discounts’ (2013) Bates White 10th Antitrust Conference, Washington DC.

254

INDEX Note: Alphabetical arrangement is word-by-word, where a group of letters followed by a space is filed before the same group of letters followed by a letter, eg ‘economic theory’ will appear before ‘economics’. In determining alphabetical arrangement, initial articles and prepositions are ignored. ABA, see American Bar Association above-normal profits, see profits abuse of dominance, 59 EU, 107 abusive dominant firm conduct, 223 ACCC, see Australian Competition and Consumer Commission acceptable dominant firm conduct, 93 ‘access regime’: Harper Review, 135–36 ACT, see Australian Competition Tribunal actual competition effects, 97 actual effects, see effects Adelaide Brighton: Boral case, 160 Cement Australia case, 189 administrative agencies, see Australian Competition and Consumer Commission; Australian Competition Tribunal administrative authorisation: CCA 2010 role, 87 administrative decisions: ACCC role, 217 judicial reviews of, 73 administrative process: CCA 2010, contemplated by, 86 advertising revenue, 173 affirmative competition defences, see competition: defences agency-centred institutional arrangements, EU, 87 AGL/MacGen, 39, 47, 216 Alcoa case, USA, 178–79 allocative efficiency, 2, 17 congruent optimisation, 124 equilibrium achieving, 49 freedom of choice and, 58 as goal, 16

imperfect competition and, 83 as inconsistent objective, 49 market equilibrium, consumers, 46 objectives, ACCC and ACT weighing up, 219 perfect competition, 55 production possibilities frontier, 118–19 productive efficiency and, 55–57 objectives inconsistency, 120 pursuit not achievable in real-world markets, 58 relative, 119, 121, 123–24 resources allocation satisfying consumer preferences, 48 valuing as synonymous with freedom of choice, 58 Allsop J, 201, 209 AMC, see Antitrust Modernization Commission American Bar Association (ABA), 113 submission to Harper Review, 218 analysis: business decisions, different perspectives, 22 cost-based, prices, 164, 165 economic, see economic analysis efficiency effects, 96 evidence-based, see evidence-based economic analysis harm to competition, 196 structural, see structural analysis supply and demand, 46 analytical constructs, markets as, 44 anti-dumping, 138 Anti-Dumping Authority, 139 anticompetitive aspect of market structure: vertical integration as, 180 anticompetitive conduct: substantially lessening competition, Canada, 105–106

256  Index anticompetitive effects, 6 efficiency: assessment, 13 counteracting, competition defences, 103–104 as defence, 102, 218 effects and, courts weighing against, 11 Harper Review, 153 market power: causal link between, 164 contributing to, Melway case, 150–51 realised efficiency more substantial than, 101 USA, 117 anticompetitive practices: dominant firms, 223 antitrust: Alcoa case, USA, 178–79 policy objectives, 43–44 see also competition Antitrust Modernization Commission (AMC), United States, 16, 41, 200–201, 209–11 Apple iPhone, 62 arbitrariness, market definition, 114 ASIC, see Australian Securities and Investments Commission ASIC v Westpac, 60, 117 assessing competition effects, 93, 96 assessing predation, 210 assessing dominant firm conduct, 224 assets: financing measures, Queensland Wire, 149 loss of value, 157 return on, Queensland Wire, 149 atomistic competition, see competition attempts, prohibition of, 235 Australia and New Zealand Banking Group, 20 Australian Competition and Consumer Commission (ACCC), 1 administrative functions: decision-making role, 217 efficiency effects, 89 appeals to ACT, 98 authorisation, 215 of conduct otherwise contravening CCA, 83 decisions by, 85–86 Baxter, 201–11 Boral, 155–68 Cement Australia, 179–99 competition as imprecise proxy for efficiency, 100

competition law roles, 84–86 conduct authorisation, 99, 130 conduct failing SLC test, power to authorise, 80 courts and Australian Competition Tribunal and, institutional division of functions, 6, 78–86, 193, 216–17, 236 dual prosecutorial and administrative decision-making roles, 217 economic welfare, conflicting objectives, weighing up, 219 efficiency effects: administrative functions, 89 opposition, 11, 13 equity, merger efficiencies and, 121 Harper section 46(2) opposition, 217 market power misuse problem recommendations, 5–7 merger efficiencies, equity and, 121 modified SCP approach, 79 non-economic objectives, weighing up, 219 NT Power, 215 Pfizer, 211–12 polycentric decision-making, 83–84 polycentric jurisdiction, 219 predatory pricing, 155–68 prosecutorial role, 217 public benefit authorisations: administrative powers, jurisdiction of courts and, dividing line becoming blurred, 103 Rural Press, 169–77 vertical mergers opposition, 216 Australian Competition Tribunal (ACT), 6, 13 ACCC: appeals to, 98 and courts and, institutional division of functions, 6, 78–86, 193, 216–17, 236 administrative functions: efficiency effects, 89 appeals from ACCC, 98 authorisation, 215 of conduct otherwise contravening CCA, 83 decisions by, 85–86 competition: law roles, 84–86 process, 79 conduct: authorisation, 99, 130 failing SLC test, power to authorise, 80

Index  257 courts and ACCC and, institutional division of functions, 6, 78–86, 193, 216–17, 236 economic welfare, conflicting objectives, weighing up, 219 efficiency effects, administrative functions, 89 modified SCP approach, 79 non-economic objectives, weighing up, 219 NT Power, 215 polycentric decision-making, 83–84 polycentric jurisdiction, 219 public benefit authorisations: administrative powers, jurisdiction of courts and, dividing line becoming blurred, 103 Australian Safeway Stores, 15 Australian Securities and Investments Commission (ASIC): litigation against banks, 59–60 authorisation: ACCC: decisions by, 85–86 role, 215 ACT: decisions by, 85–86 role, 215 administrative, CCA 2010 role, 87 competitive tenders and, 215 conduct otherwise contravening CCA 2010, 83 non-economic grounds, 85 court decisions, 83 autonomy, private, 71 average costs: European Union, 108 marginal costs equalling, 54 Queensland Wire, 149 average revenue, 60 background norms, 47 backward vertical integration, 179–80, 188 balance of power: in markets, 230 Bank Bill Swap Reference Rate, 59–60 barriers to entry, see entry barriers to exit, see exit Baxter, 35 Baxter First Instance, 201, 209 Baxter Full Court, 201, 206, 212 factual matrix, 202–207 gaps in, 207–208 under Harper section 46: 201, 208–10 under market manipulation, 210–11

behaviour: commercial, see commercial behaviour competition law addressing, 128 firms, see firms strategic, 79 see also conduct below-cost pricing, 111, 165 benchmarks: actual market as, New Zealand, 91 actual position in real-world market, 178 assessment: dominant firm conduct, 90 atomistic competition as, 90 competitive level problematic as, 178 dominant firms: conduct assessment, 90 output, 90 own interests in real-world market, pursuit as, 91 prices, 90 against hypothetical competitive level profits, 60 hypothetical competitive markets, 104, 105, 124 hypothetical market, courts adoption as, 133–34 legal: economic efficiency adaptation to provide, 43 efficiency, 124 New Zealand actual market as, 91 perfect competition as, 22 profits: perfect competition, 60 workable competition, 60 to transcend different worldviews: efficiency adaptation and refinement, 130 benefits, volume-related, 205 Bernheim, D, 108, 115 Bertrand Paradox, 66 Besanko et al, 201 BHP (Broken Hill Pty Co Ltd), see Queensland Wire bidding: for inputs, 194 in price wars, 227 public tenders, 209 wars, see bidding wars bidding wars, 11, 231 bundling, 234 Cement Australia, 179, 194–95 harm to competition process, 175

258  Index harm to competitors, 175 for inputs, 126, 130–31, 175, 179, 194, 227, 232 public tender processes, 194–95 sellers of raw materials creating by calling for public tenders, market manipulation, 225 unmet demand, 180, 185 vertical integration, 194 zero-sum competition, 180–81 bipolarity in competition law, 71, 167, 231 BlackBerry, 31, 62 Blair, Roger, 40, 45, 52–54, 57, 62 Boral, 121–22, 155–58, 166–68, 179, 180, 187, 230 Boral First Instance, 155–56 factual matrix, 158–60 gaps in, 160–63 under Harper section 46, 163–66 under market manipulation, 166 recoupment tests, 210 Boral, Cement Australia case, 185, 189 borderline prices, 134 breach of contract: damages determination, foreseeability requirements, 98 break-even points: producers, 45 bright-line legal rules, 41 Broken Hill Pty Co Ltd (BHP), see Queensland Wire Brunt, Maureen, 81–82, 84 Bryan v Maloney, 71 Budget Bricks, Boral case, 159 bundling, 200–201, 213, 227 AMC, 200–201 Baxter, see Baxter bidding wars, 234 dual purposes problem, 201, 234 economic effects, 201 economies of scale, 201, 205 economies of scope, 201, 205 efficiency, 201 harm, 200–201, 204, 207, 208 purpose, 201 recoupment tests, 210 volume-related benefits, 205 burden of proof, 217–19 Burguet, Roberto, 194 business: Canada, 106 conduct: economic consequences, 82

decisions: analysis, different perspectives, 22 consequences playing out in market, 149 financial accounting, 21 management accounting, 21 rational norms, purpose tests, 128 efficiency, 152 judgement, 152 rule, 71–72, 148 legitimate justification as production efficiency, 81 legitimate rationale, 81, 218 legitimate reasons, 152, 186 buyers: competing to bid up prices, 51 contributing to bidding contests, market manipulation, 225 dominant firms as, market power, 15 market manipulation powers, 230 buying: consumers: reducing when market prices rise, 61 overbuying, 178–79 C&M, Boral case, 159–62, 164–65, 166 Callinan J, 157, 159 Canada: anticompetitive conduct substantially lessening competition, 105–106 business justification, 106 competition: anticompetitive conduct substantially lessening, 105–106 harm to, 106 protection, 129 competitors: harm to, 106 protection, 129 conduct: prohibited, injunctions to restrain, 105, 113 dominant firms: anticompetitive conduct substantially lessening competition, 105–106 effects: prohibition based on purpose and, 102, 105 efficiency: credible, 106 purpose as surrogate for, 127 harm to competition and competitors, 106 injunctions to restrain prohibited conduct, 105, 113

Index  259 market power defined, 106 pro-competitive rationale, 106 prohibited conduct: injunctions to restrain, 105, 113 protection, competition and competitors, 129 purpose: prohibition based on effects and, 102, 105 as surrogate for efficiency, 127 capital: expenditure, 58 fixed, efficient use measures, Queensland Wire, 149 capital markets, 48 capitalism: free market ideology, 44 Carruthers, Celeste, 40, 45, 52–54, 57, 62 cartel conduct, 65, 146, 231, 232 causation: chain, Structure-Conduct-Performance, 79 effects and, 105 Harper Review, 196, 197, 234–35 meeting competition, 233 predatory pricing, 233 CCA, see Competition and Consumer Act 2010 Cement Australia, 179–82, 198–99, 201, 215, 219 factual matrix, 182–88 bidding wars, implications, 194–95 gaps in, 195–96 harm assessment, 192–93 institutional implications, 193 penalties assessment, 192–93 subsequent events, 188–92 under Harper section 46: 196–98 under market manipulation, 198 cementitious materials, 181–82, 185–86, 195 Cemex, Cement Australia case, 182, 188 certainty: policy choices tested against, 17–18 cgfa (concrete grade fly ash), 182, 185–91, 195–97 choices: freedom of, 58 perfect competition assumptions, 119 Chrystal, A, 54 Church, J, 178 CJEU (Court of Justice of the European Union), see European Union: CJEU coal-fired power stations, 182–84, 185–91 Coal Reuse, Cement Australia case, 190–96 collusion, 64–65

commercial behaviour: social engineering of, 42, 83 commercial intuition, courts, 4 commercial transactions, supervision: European Commission, 135 reluctance, US courts, 135 common law: US method of judicial law-making, 87 common market, EU, 111 Commonwealth of Australia Constitution Act, 217 community of interest, 46 community welfare, 119, 121 companies, see firms competition: absence of in perfect competition equilibrium, 51 atomistic, 8, 23 as benchmark, 90 CCA 2010, 79 promotion at expense of efficiency, 78–79 US courts, 78–79 Canada, see Canada between consumers, 51, 53 defences: affirmative, 103, 217–218 acknowledging harm to competition, 101 dominant firms justifying conduct, 128 efficiencies with independent values, 104 harm to competition acknowledging, 101 producer surpluses, 104 SLC tests and, 104 efficiency: counteracting anticompetitive effects, 103–104 EU, 111 meeting competition, 148 EU, 111 negating elements of contravention, 101 negative, 102, 103, 217–18 disequilibrium, occurring in, 51 economic models, 19–22 economic principles adopted, 8 effects, 89–90 actual, 97 adverse, efficiency prevailing over, 235 assessing, 93, 96 competition analysis, EU, 97 court enquiries pushing into, 82

260  Index court jurisdiction, 89 determining, 15 direct measures, 80, 90 downstream markets and, 189 economic enquiries, revealed by, 82 efficiency effects and: hierarchical relationship between, 101 interaction between, 13 relationship between, 175 efficiency gains relevance to assessing, 99 harm to efficiency indicator, 102 inferring from nature of conduct, EU, 97 law focus on, 13 market manipulation assessing, 93 market performance, 93 market power resulting from, 235 negative, positive efficiency effects tests and, EU and USA, 128 probable, 97 proof of harm not required to demonstrate, 192–93 qualitative analysis, predatory pricing, 234 qualitative assessment, 192 raising rivals’ costs, 179 resulting from market power, 235 Structure-Conduct-Performance forensic determination implications, 80 tests, see effects tests efficiency: interrelationship between, 229–32 poor surrogate for, 230 surrogate, 100, 101, 102 EU, see European Union harm to: affirmative defences acknowledging, 101 analysis, 196 Canada, 106 competition-meeting grounds, EU, 110 efficiency grounds, EU, 110 EU, 107, 108, 110 justification by offsetting public benefits, 99 necessity grounds, EU, 110 price alone not indicating, 197 profit alone not indicating, 197 Queensland Wire, 143 imperfect, 60 allocative efficiency and, 83 EU, 110 marginal revenue, 69 market mechanism of perfect competition not operating under, 121 non-zero-sum games, 119

peaceful coexistence, 119 productive efficiency and, 83 supply, 69, 122–23 as imprecise proxy for efficiency, 100 law: ACCC roles, 84–86 ACT roles, 84–86 behaviour, addressing, 128 bipolarity, 71, 167, 231 competition effects, focus on, 13 conduct giving rise to harm, prevention, 162–63 consumer welfare, objectives, 223 courts, see courts dominant firm conduct, application to, 19 economics and, unanimity lacking, 120 effects, focus on, 13 equity and, 121 equity jurisprudence doctrines and, 12 findings and recommendations implications, 234–37 legislative purpose, 72 market power abuse, 73 neoclassical economic models use, 42 objectives: differing opinions, 13 dynamic efficiency, 48 equitable distribution of surplus between producers and consumers, 46–47 role of courts and, 82–84 unanimity lacking, 120 policy objectives debate, differing worldviews, 223 predictive tests requiring foreseeability, 99 productive efficiency, objectives, 223 prohibitions focussing on conduct, 99 proscribed conduct, 71 protection of process of competition, 17 protects process of competition, not competitors proposition, 227, 228–29 publicly owned enterprises exposure to, 53–54 redistributive objectives counter to, 47–48, 121 social engineering of commercial behaviour, 83 success, 81 worldviews, differing, 18–19 legislation, v–vi interpretation, economics and, 41 see also law above

Index  261 lessening, 10 substantially, see SLC tests; substantially lessening competition meeting: causation, 233 conduct harming competition grounds, EU, 110 defence of, 148 EU, 111 dominant firms justifying conduct by, 128 Harper Review, 174–75, 233 market manipulation approach, 233 Rural Press, 169–77 mergers: anticompetitive, prohibition, 14 on the merits, 8, 223 EU, 107–108 non-price, 53 normal incidents of: harmful conduct of dominant firms, distinguishing from, 48 norms of conduct: findings and recommendations, 225–27 oligopoly, see oligopoly paradigm, 102 economic models and, 49–59 findings and recommendations, 225–27 Harper Review, 23 products, 11 perfect, 8, 19, 102 allocative efficiency, 55 assumptions, 119–20 as benchmark, 22 community of interest, 46 economic model of, 62 efficiency and, 118–19 entry without costs, 69 equilibrium: absence of competition in, 51 consumer satisfaction maximisation, 56 price, 45 output, 54 substitution of goods, 56 EU, 110 exit without costs, 69 firms: leaving market, 156, 157 efficiency or innovation incentives under, 123 harm from market power derived from, 118 hypothetical market, 23

inter-relationship between competition and efficiency, 229 market mechanism of, 121 free exit of firms, 122 market power derived from, 118 model, 20–21, 45–46, 62, 118 monopoly comparisons, 21 neoclassical model of, 110, 223 output maximisation under, 119 productive efficiency, 49, 55 profits: benchmarking, 60 maximisation, 56 real-world competition and, distinguishing between, 63 sales increases, 70 shut down principle, 165–66 social welfare, 56 suppliers lack market power, 50–51 supply curves, 45 permitted conduct norms, 71 policy and law, independent review panel, 6 as poor surrogate for efficiency, 74 process, 95 ACT, 79 Baxter case, 205, 206 conduct impact on, 99 efficiency: losses harm to, 100 separate from, 101 harm to, 231 bidding wars, 175 efficiency losses and, 100 equally efficient competitors excluded, 102 EU, 109, 112 less efficient competitors excluded, 102 Queensland Wire, 144 Rural Press, 175 USA, 115 market conduct equated with, 80 market structure equated with, 80 protection of, 17 purpose tests and, 103 separate from efficiency, 101 SLC tests focus, 146 by which firms compete for the custom of consumers proposition, 227, 229 between producers, 51 product differentiation, 53 production efficiency role, 81

262  Index protection: Canada, 129 not competitors, EU, 127, 228 USA, 129 pursuit of: efficiency as higher order objective than, 85 rational business decisions, 8 real world, 63, 226, 230 risks management: firms anticompetitive conduct representing, 231 Schumpeterian dynamic, 31 substantially lessening (SLC), see substantially lessening competition between suppliers, 53 as surrogate for efficiency, EU, 109 surrogate test, 230 survival, see zero-sum below tests, 10, 74 USA, see United States workable, 8, 19 profits, benchmarking, 60 worldviews of, 102 firms locked in increasing output, 130 norms in, 70–75 price to pay for inefficiencies, 166 zero-sum, v–vi, 9, 10–11, 21, 25, 75–76, 226–27 Bertrand Paradox, 66 bidding wars, 180–81 competition for the market relationship, 31 competitors, taking sales from, 123 court acceptance, 102 dominant firm conduct under, 152 dual purposes problem, see dual purposes problem economies of scale, 69 efficiency: implications, 43, 65, 69–70 market manipulation and, 124–26 equally efficient competitors and, 62–63 equally efficient new entrants, USA, 115–16 excluding competitors, 156 exit barriers, 158, 165 firms’ efficiency or innovation incentives under, 123 firms leaving market, 156 firms locked in, 125–26 incumbent firms: EU, 110 seeking to survive, 123 USA, 115–16

inefficient outcomes, 80, 123, 124, 152, 179, 227, 231, 236 inter-relationship between competition and efficiency, 230–32 Kaldor–Hicks efficiency, EU, 109 market efficiency addressing, 49 market manipulation and efficiency and, 124–26 market power reconceptualisation, 236 market structure as primary source of market power, reconceptualised, 79 new entrants seeking to destroy incumbents, 123 new entry implications, 69 as normal market mechanism working under real world conditions, 123 norms in, 70–75 not achievable, 124 output increases, 125 overbuying inputs, harm from, 179 peaceful coexistence, 73 price: increases, 126 reductions, 125 proportionality, 73 Queensland Wire, 144, 145 real-world assumptions, 42 revenue maintenance or increase, 123 Rural Press, 174, 175 sales: growth, 69, 70 taken from competitors, 152 self-preservation, 156 self-selection for entry or exit, 119 steel industry, 139 supply adapting to demand shifts, 157 two firms locked in, 148 see also antitrust Competition and Consumer Act 2010 (CCA): administrative authorisation role, 87 administrative process contemplated by, 86 atomistic view of competition, 79 authorisation of conduct otherwise contravening, 83 efficiency grounds, 230 non-economic grounds, 85 efficiency priority over public interest, 85–86 Part IIIA: access regime, 135–36, 154 Part IV, 8, 14, 83 predictive SLC test, 131 purpose and effect, 117 Queensland Wire, 145

Index  263 Part IVA, 84, 92 prosecutorial role, 87 SCP approach embedded in, 82 section 45: bilateral agreements, 106 Cement Australia, 196 competition test, 117 contravention, 186–87 Melway, 151 SLE tests, 97, 176, 184 section 46: 4–8, 11–14 conduct failing SLC test, power to authorise, 80 conduct having purpose of excluding competition prohibited, 95 contravention, market manipulation as basis for determining legal liability, 236 current effects tests, outside, 95–101 economic efficiency, 91 efficiency, 81 exclusion of rivals, 48 ‘exclusionary purpose’ requirement, 81 Harper Review recommended version, 242 Harper Review with sub-section (2), 24–29 see also Harper Review legal theory of liability, 42 liability for contravention, 14–15 new, 24–29 old, 24–29 pre-Harper legislative history, 238–42 pre-Harper version text, 238 Parliament-adopted version, 243 private economic power and, 82 purpose tests, 91 subject to SLC tests, proposals, 95 taking advantage requirement, 81 section 47: anticompetitive effect, 204 Baxter, 205, 206 bilateral agreements, 106 competition test, 117 contravention, 23 exclusive dealing, 33, 201, 210 Pfizer, 96 SLC tests, 151, 153, 201, 205, 212 terminated distributors, 151 section 50: merger efficiencies, 224 section 51A: future representations, 147–48 section 76: pecuniary liabilities, 14, 96

section 82: private actions, 15 section 83: private actions, 15 section 88: Cement Australia, 188 section 93: Cement Australia, 190 standards requiring normative and evaluative judgments in application, 83 Competition and Consumer Amendment (Misuse of Market Power) Bill, 6 Competition Policy Review, see Harper Review competitive bidding for inputs, 194 competitive conduct: by incumbent firms, Rural Press, 174 normal, 223 competitive constraints: market power, 58 competitive disadvantage: Queensland Wire, 136 competitive dynamics, 1 competitive effects: of vertical supply arrangements, competitive tenders reducing concerns about, 195 competitive level: benchmark, problematic as, 178 dominant firm prices raising above, 178 efficiency and, 143 forensic difficulties of ascertaining, 120 hypothetical, 121 identification, price and output, 21 indeterminate, 178 as indicator of private welfare, 120 as indicator of social welfare, 120 output, dominant firms conduct benchmark, 90 prices, dominant firms conduct benchmark, 90 private welfare indicator, 120 prohibition of pricing above, 228 social welfare indicator, 120 competitive markets: hypothetical, 93, 223–24, 225 benchmark, 104, 105, 124, 133–34 Harper Review, 127 market power concept deriving from, 79 of neoclassical economics, 8 competitive output, 57 competitive paradigms: conduct norms underlying, 94 competitive prices, see prices

264  Index competitive steps: Rural Press, 173 competitive strategies: industry concentration affected by, EU, 109 competitive tenders: authorisation and, 215 competitive effects of vertical supply arrangements, reducing concerns about, 195 competitors: Canada, see Canada customers and suppliers as, 146 dominant firms supplying finished products to: Queensland Wire, 140 economies of scale denied, EU, 108 equally efficient: competitor test, EU, 112 excluded, harm to competition process, 102 EU, see European Union excluding, 156–57 zero-sum competition, 156 harm to, 231 bidding wars, 175 Canada, 106 EU, 107, 108, 109, 112 Rural Press, 175 USA, 115 less efficient: excluded, harm to competition process, 102 lower prices, dominant firms meeting, EU, 111 market power as freedom from constraint by, EU, 112 protection: Canada, 129 EU, 107, 127, 228 purpose tests, 103 USA, 129 relative efficiency of, 102 suppliers and customers as, 146 taking sales from: zero-sum competition, 123 USA, see United States weakened, EU, 108 zero-sum competition taking sales from, 123 concentrated market structures, 1 concentration: industry, see industries market, 1–2

conceptualised market power, see market power concrete grade fly ash (cgfa), 182, 185–91, 195–97 concrete masonry products, 158–66 concrete products, 185–86, 189, 192, 195, 199 condemned conduct, 209 conduct: authorisation: ACCC and ACT, 99, 130 bidding in price wars, 227 business, 82 cartels, see cartel conduct competition law prohibitions focussing on, 99 competition process, impact on, 99 competitive, see competitive conduct condemned: permitted and, difference between, 209 dominant firms, see dominant firms effects proscribing by reference to purpose and, 128 efficient, 9, 130, 229–32, 236 elimination of competition as sole motivation, USA, 113 EU, see European Union exclusionary, 103 USA, 113, 114, 156 forensic assessment conclusions: findings and recommendations, 232–34 form, inferring effects from, EU, 112 giving rise to harm: competition law prevention, 162–63 harmful, see harm harming competition, justification, EU, 110 harming competitors, EU, 109 impact on process of competition, 99 inefficient, 130, 218–19, 231, 232, 236 intent inferred from, USA, 113, 114 manipulative, 152 efficiency and, relationship between, 125 market demand, responding to: power of the market, 130 private efficiency, 130 social efficiency, 130 market manipulation, 60, 226 inefficient, condemnation assessments, 232 market power and: causal connection, 174 Harper Review, 127

Index  265 market structure not determining, 79 normal competitive, 223 norms: competitive paradigms: findings and recommendations, 225–27 underlying, 94 economic models, underlying, 94 market power misuse, 71 permitted conduct, competition, 71 worldviews difference reflecting, 64 permitted: condemned and, difference between, 209 dominant firms, 223 norms, competition, 71 prohibited and, distinguishing between, 96 test, AMC, 209–11 profit-maximising: rational, 133 prohibited: injunctions to restrain, Canada, 105, 113 permitted and, distinguishing between, 96 proscribed: competition law, 71 by reference to effects and purpose, 128 rational profit-maximising, 133 rational self-interested, 102 SLC test, failing, 80 tests based on, USA, 89 unconscionable, 84 USA, see United States weakening process of competition, EU, 109 see also behaviour congruent optimisation: allocative efficiency, 124 conscious parallelism, 65 consequences: economic, 82 of law, 80 monopoly price and output decisions, responsiveness to, 51–52 policy choices informed by, 17–18 constitutional norms, 71 constitutionality of legislation limiting personal freedoms, 73 constraints: absence of, market power as, 147 competitive, market power, 58 dominant firms, 225 see dominant firms market power as freedom from, by competitors, EU, 112 Consumer Law, 84, 148

consumers: buying: reducing on increasing prices, 61 choice: changes in, 48 maximisation, 16 community of interest, 46 competition between, 51, 53 detriment to, 46 domestic, preferences, 57 equilibrium output response, 62 equilibrium prices response, 62 EU, see European Union harm, 48 holding down prices for, 53 market exclusion, 121 perfect competition assumptions, 119 prices: change responsiveness, 51 holding down for, 53 producer welfare redistribution to, 57 producers and, equitable distribution of surplus between: competition law objectives, 46–47 as producers, interest as, 118 responsiveness: market power as function of, 52 price changes, 51 satisfaction: maximisation, 16 perfect competition equilibrium, 56 surpluses, 120, 121 equitable distribution between producers and, 46–47 as indicator of efficiency effects, EU and USA, 128 under market equilibrium, 45 monopolies and, 54 transfers from, 179 transfers to, 166, 227 USA, see United States welfare, 17, 56, 102 ambiguous term, 118 competition law objectives, 223 Harper Review, 17 redistribution from producers, 57 standard, EU, 109–10 USA, 17, 102 contested tenders: input costs raising by bidding by, 194 contracts: exclusive, see exclusive contracts

266  Index law, foreseeability requirements, 98 winning by bidding in public contracts, 209 contractually-based market power, 210–11 contravention: negating elements of, competition defences, 101 contribution margins, 161 cornering, 59, 185 Corones, Stephen, 13, 79, 99 Corporations Act 2001, 59, 117 corporations law: business judgement rule, 148 directors’ duties, 71–72 correct prices, 44 cost-based analysis, prices, 164, 165 cost-based predatory pricing approach, 157, 163 cost benchmarks, EU, 108 cost–efficiency advantages, 216 cost of capital, EU, 108 cost–price calculations, Baxter case, 204 cost–price predatory pricing, issues, 233 costs: average, see average costs data, NT Power, 216 dominant firms, see dominant firms exit, see exit firms, see firms holding, 157 increases: bidding for inputs by contested tenders, 194 industry, 54 inputs, raising by bidding by contested tenders, 194 marginal, see marginal costs opportunity, 21 own, increasing on raising rivals’ costs, 187–88, 194 producer functions, 49 raising rivals, see raising rivals’ costs reopening, 157–58 revenue and, relationship between, 54 monopoly, 65–66 saving, efficiencies, Baxter, 204 social, see social costs sunk, 109, 157, 226 transaction cost efficiencies, 2, 180 countervailing power: dominant firm constraints by, 225 court-centred competition law, 82

court-centred institutional arrangements, USA, 87 Court of Justice of the European Union (CJEU), see European Union: CJEU courts: acceptance of zero-sum competition, 102 administrative agencies and, division of functions between, 80 ACCC and ACT and, institutional division of functions, 6, 78–86, 193, 216–17, 236 authorisation matters decisions, 83 commercial intuition, 4 competition effects, enquiries pushing into, 82 competition law: economic theory distinguished from, 8 objectives and, 82–84 reasoning, 4 roles, 82–84 constitutionality of legislation limiting personal freedoms, 73 discretionary decision-making, 83 dominant firm conduct, commercial intuition, 58 economic analysis, 82 economic theory: competition law distinguished from, 8 efficiency, 217 issues decisions, 83 judicially defined, 11 efficiency effects: anticompetitive effects, weighing against, 11 assessing, 103 taking into account in judicial proceedings, 5, 6 efficient markets, judicial use of term, 117 European Union, see European Union: CJEU hindsight avoidance in competition law cases, 131 hostile intent evidence influencing, 116 hypothetical market, adoption as benchmark, 133–34 injunctions to prevent threatened harm, 168 institutional division of functions, 6, 78–86, 193, 216–17, 236 interlocutory injunctions, 99 judges: normative role in applying standards, 83 judicial attitudes towards economics, 42 judicial determination of efficiency, 217 judicial interpretation of purpose, 96

Index  267 judicial proceedings taking efficiency effects into account, 5, 6 judicial reviews of administrative decisions, 73 judicial use of term ‘efficient markets’, 117 jurisdiction: administrative powers of ACCC and ACT and, dividing line becoming blurred, 103 competition effects, 89 efficiency excluding from, 81 legitimate business rationale, credence to, 81 market conduct tests administration, 80 market power misuse recommendations, 5–7 market structure tests administration, 80 merger efficiencies, exclusion from assessing, 110 polycentric decision-making, 83 public benefit issues decisions, 83 purpose interpretation, 96 purpose tests subjective purpose interpretation, 92 statute analysis, 23 United States, see United States zero-sum competition acceptance, 102 creative destruction, 43 criminal law: self-defence, 72–73 CS Energy (Swanbank), Cement Australia case, 183–84 curves, see demand; supply customers: suppliers and, as competitors, 146 cutting prices, see prices damages, 15 Daniel Shak v JP Morgan Chase & Co, 93 Dawson J, 137 deadweight losses, 57 dealing, see exclusive dealing; refusal to deal Deane J, 71, 137 debt/equity leveraging: Queensland Wire, 149 decision-making: ACCC dual prosecutorial and administrative roles, 217 discretionary, 83 elasticity of demand subjective, 123 firms, subjective process, 123 polycentric, 83–84 subjective, elasticity of demand, 123

decisions: administrative, see administrative decisions authorisation before implementing, 231 dominant firms, see dominant firms implementing, authorisation before, 231 making, see decision-making monopoly output, 51–52 defences: competition, see competition efficiency, see efficiency defensive price cutting, 158 delivery delays: unmet demand indication, 125 demand: curves: downward sloping, 51–52, 55–56, 66–70, 230 in flux, 68 horizontal, 45, 50, 54–55 elasticity of, see elasticity of demand incumbents meeting, 210 market, see market demand monopoly meeting, 195 price changes, responsiveness to, 40 reduced prices as competitive response to changed market conditions of, 158 reduction, effect of, 122 reduction shifts: Boral, 164 supply response, 164 response, 53, 152 Cement Australia, 180, 198 evidence of, 218 market manipulation, 208, 230 market power, 225 to output decisions, 198 to price decisions, 198 Rural Press, 173 shifts: Boral, 165 reductions, 152 predatory pricing, 233 supply adapting to, zero-sum competition, 157 supply responding to, 231 supply and, see supply and demand supply exceeding: price reductions, 51 unmet, see unmet demand Demura, Peter, 138–39 Department of Justice, United States, see United States

268  Index destruction: creative, 43 mutually assured, 65 determinacy, legal, see legal determinacy deterrence theories: entry, 70 detriment to consumers and producers, 46 Devlin, Alan, 18 digital economy: market power application, 30–31 digital platforms, 30–31 see also platform markets diminishing returns, law of, 108 diminishing social efficiency, 224 direct measures: competition effects, 80, 90 directors: duties, 71–72 good faith, 71 discretionary decision-making: courts, 83 discretionary practical justice: inconsistent with rule of law, 12 disequilibrium: competition occurring in, 51 dissipation: monopoly profits, 66 distributional equity: economics of public finance, 47 economics of taxation, 47 market power, 48 monopoly, 48 distributive equity, 85–86 distributive issues: between interest-groups, 47 distributors: terminated, 151 domestic consumers’ preferences, 57 dominance: abuse of, 59 EU, 107 laws, abuse, 127 dominant firms, 1, 9 anticompetitive conduct substantially lessening competition, Canada, 105–106 anticompetitive practices, 223 benchmarks, see benchmarks as buyers, market power, 15 commercial conduct, EU, 107 conduct, 14, 16, 18–19 abusive, 223 acceptable, 93

assessment, 90, 224 competition law application to, 19 courts commercial intuition, 58 creative destruction application, 43 effects assessment, 120 exclusionary, 223 harmful, normal incidents of competition, distinguishing from, 48 justifying, 128 permitted, 223 unilateral, see unilateral conduct below US courts intervention scepticism, 58 under zero-sum competition, 152 constraints: by countervailing power, 225 costs: raising by overbuying, 178–79 decisions: on basis of economic efficiency, 91 process, EU, 108 rational business decisions, 93 harmful effects, 91–92 downstream markets, see downstream markets dual purposes problem, court resolution against, 143 economic efficiency concepts application to, 3 equally efficient competitors, EU, 108 EU, see European Union finished products supplying to competitors: Queensland Wire, 140 harm to self on raising rivals’ costs, 186, 193 harmful conduct: normal incidents of competition, distinguishing from, 48 incumbents: inefficient new entrants, 125 lower prices of competitors, meeting, EU, 111 market power: in downstream markets, 178–79, 194 taking advantage of, 23 output: artificial restrictions, 50, 92 increasing, 125 overbuying, 178–79 own interests in real-world market: pursuit as benchmark, 91 power: prices above competitive level, ability to charge, 134 prices: excessive, 121

Index  269 above normal, 92 raising above competitive level, 178 profits: forgone in upstream markets, 180 gained in downstream markets, 180 above normal, 134 maximising conduct, EU, 108 maximising levels, 143 protecting own interests, EU, 111 raising rivals’ costs: harm to self on, 186, 193 surpluses transfer away from, 227 taking advantage of market power, 134 unilateral conduct, 3, 118 international debate on, 16 overseas experience, 22–24 regulation, 235 see also market power: misuse US Department of Justice conduct report, 18 double marginalisation problem, 179 downstream markets, 178–79 competition effects and, 189 dominant firms: market power, 178–79 without market power, 194 profits gained in, 180 efficiency losses in, 100 higher input prices, market power mitigating, 194 identification, 179 market power in, 179, 195 profits gained in, dominant firm, 180 surpluses: redistribution between upstream and downstream suppliers, 191–92 downward sloping demand curves, see demand: curves Dowsett J, 201, 206, 212 Dranove, D (Besanko et al), 201 dual purposes problem, 124 Baxter case, 206 bundling, 201, 234 Cement Australia, 186–87 courts: resolution against, dominant firms, 143 solving problem, 95–96 different approaches to, 102 EU, 108 malicious conduct, 75 market manipulation theory, 27 Pfizer case, 205, 211 self-interested conduct, 75

sole purpose tests, 235 USA, 113, 115–16 duties, directors, 71–72 dynamic conditions: supply and demand, 173 dynamic efficiency, 2, 48, 56, 118–19 dynamic forces underlying effects tests, 98 dynamic gaps in neoclassical economic models, 48 dynamic potential for innovation and increasing efficiency, 124 dynamics, competitive, 1 EC (European Commission), see European Union economic activity: in firms, 118 in industries, 118 legal precedent role, 4 in markets, 118 public policy role, 4 in whole economy, 118 economic analysis: courts, 82 evidence-based, see evidence-based economic analysis monopoly, 44 positive: policy choices elucidation by, 17–18 Economic Analysis of Law (Posner 2007), 40 economic assessment, 86 economic concepts of efficiency, 130 economic consequences, 82 economic downturns, 157 economic effects: bundling, 201 economic efficiency, vi, 2, 10 dominant firms: concepts application to, 3 decisions on basis of, 91 harm to, 69 Harper Review on, 11, 13–14 legal benchmark, adaptation to provide, 43 legal indeterminacy, 11 manipulative conduct of firms, assessing, 230 market power, positive and normative legal theory centrepiece, 224 Melway First Instance, 145 Queensland Wire, 146–49 SCP approach embedded in CCA impeding, 82 socially harmful conduct of firms, assessing, 230

270  Index economic enquiries: competition and efficiency effects revealed by, 82 economic experts, 41–42 economic models: competition, 19–22, 49–59 conduct norms underlying, 94 neoclassical, see neoclassical economic models perfect competition, 62 economic power: private, 82 economic principles: adopted, competition, 8 economic theory: distinguished from competition law, by courts, 8 predictive power, 20 role in economic activity regulation, deeper understanding required, 4 economic welfare: conflicting objectives, ACCC and ACT weighing up, 219 economics: competition law and, 41, 120 judicial attitudes towards, 42 market power in, 39–43, 76–77 economic models and competition paradigm, 49–59 efficiency concepts, 43–49, 75–76 incumbent response to new entry, 61–70, 73–74 market manipulation, 75–76 laws in securities markets, 59–61 norms in zero-sum competition, 70–75 policy objectives, 43–49 zero-sum competition, 75–76 neoclassical, see neoclassical economic models of public finance, distributional equity, 47 of taxation, distributional equity, 47 US thinking since 1970s, 40 economies of scale, 2, 58 bundling, 201, 205 competitors, denied to, EU, 108 equally efficient firms competing for, 174 EU, see European Union exclusion, causing, EU, 108–109 Rural Press, 174 USA, 115 volume-related efficiencies, 205 zero-sum competition, 69

economies of scope, 2, 58 bundling, 201, 205 volume-related efficiencies, 205 economists: lawyers and, opposed perspectives, 42 non-zero-sum games, 63–64, 119 worldviews, 64–65, 119, 130 effective prices, 200 effects: actual: establishing, EU, 109 international focus on, 97 anticompetitive, see anticompetitive effects assessment: dominant firm conduct, 120 Canada, see Canada causation and, 105 competition, see competition conduct, proscribing by reference to purpose and, 128 dominant firm conduct assessment, 120 efficiency, see efficiency effects EU, see European Union form of conduct, inferring from, EU, 112 harmful, purpose as screen isolating, 127 intent inferred from, USA, 113 no prohibition of purpose without, EU, 102 no prohibition of without purpose, EU, 102 prices, 93 probable: international focus on, 97 problems with concept of, 117 prohibition based on purpose and, Canada, 102, 105 purpose: and, integrally linked, 127 and, inter-linked and inextricable meaning, EU, 112 as staging point towards determining, EU, 109 without, 96, 117 Pfizer case, 212 USA, 102, 114, 116 SLC tests, 143, 235 tests, see effects tests USA, see United States effects-based approach, 86 effects tests, 16, 91–92 ABA recommendation against adopting, USA, 113 current, outside section 46: 95–101

Index  271 dynamic forces underlying, 98 efficiency effects role and, 102–106 focus internationally on actual or probable effects, 97 foreseeability, 99 Harper Review, 156 predictive nature, 96–97, 98 principles developed overseas application to, 109 probability-based, 97 Pfizer case, 212 structural factors, 98 efficiency: adaptation, 11 as legal concept, 118–24 as relevant benchmark to transcend different worldviews, 130 adverse competition effects, prevailing over, 235 allocative, see allocative efficiency anticompetitive effects: assessment and, 13 as defence, 102 assessments and relevance to purpose tests, 99 atomistic competition promotion at expense of, 78–79 average costs, minimising, EU, 108 Baxter, 204 benefits: Harper Review, 235–37 vertical integration, 180 Boral, 163, 164 bundling, 201 burden of proof, 218 business, 152 Canada, see Canada CCA 2010, section 46, 81 competition as imprecise proxy for, 100 competition process, separate from, 101 competition surrogate for, 74, 100, 101, 102, 230, 109 competitive level and, 143 concepts, 43–49, 75–76 conclusions regarding, 68–69 conduct harming competition grounds, EU, 110 costs saving, Baxter, 204 counteracting anticompetitive effects, 103–104 courts and, 217 credible, Canada, 106

defence: EU, 111 rational self-interested conduct as surrogate, 102 determining, 15 dominant firms justifying conduct by reference to, 128 dynamic, see dynamic efficiency economic concepts, 130 economic, see economic efficiency effects, see efficiency effects enhancing, 10 equity and, distinction between, 128 EU, see European Union evaluative issues, 84 excluding from court jurisdiction, 81 financial, EU, 108 firms, assessing, 123 forensic determination, 80 forensic tests, 11, 27–28 gains: as defence to competitive harm claims, USA, 101 offsetting, EU, 107 relevance to assessing competition effects, 99 surrogate tests, USA, 117 harm to: principal mischief SCP seeking to address, 79 Harper Review, 197–98 as higher order objective than pursuit of competition, 85 importance, 82 incentives and assessment, firms, 123 increasing: dynamic potential for, 124 socially inefficient production as price to pay, 123 with independent values: affirmative competition defences, 104 indicator: harm to, competition effects, 102 indirect approach focussing on, EU, 107 issues: court decisions, 83 judicial determination, 217 judicially defined, 11 Kaldor–Hicks, EU, 109, 124, 148 legal benchmark, 124 legal concept, adapting as, 118–24 as legal theory of liability basis, 219

272  Index losses: harm to process of competition and, 100 in upstream and downstream markets, 100 manipulative conduct and, relationship between, 125 market manipulation: considerations, 167 zero-sum competition and, 124–26 market power misuse: considerations, 217 redefinition as market manipulation assessed by reference to, 230–31 market size and, 168 meaning, 224 mergers, 224 monopolisation: American Bar Association submission to Harper Review, 218 normative issues, 81, 84 NT Power, 215–16 operational, EU, 108 optimisation, 123–24 output and: Queensland Wire, 141 Pareto, 123 perfect competition and, 118–19 plaintiffs proving issues pertaining to, 218 preferred approach: Harper Review on, 128 prevailing over market power, 79 prices and: Queensland Wire, 141 priority over public interest, CCA 2010, 85–86 private, see private efficiency problematic role: Harper Review on, 15–16 productive, see productive efficiency profits and: Queensland Wire, 141 purpose as surrogate for, EU, Canada and USA, 127 Queensland Wire, 141, 142 rationale, parties asserting bearing burden of proof, 218 realised, anticompetitive effects more substantial than, 101 recognition as defence, USA, 80 redefining as relevant benchmark to transcend different worldviews, 130 relative: assessing as surrogate test, 124

of competitors, 102 firms, 168 separate from process of competition, 101 social, see social efficiency of survival strategies, 158 as test of social desirability, 158 testing, 129–219 as theory of harm basis, 219 transaction cost, 2, 180 USA, see United States zero-sum competition: implications, 43, 65, 69–70 market manipulation and, 124–26 efficiency effects, 5–7, 13, 89, 90, 117 ACCC and ACT administrative functions, 89 analysis, 96 competition effects and: hierarchical relationship between, 101 interaction between, 13 relationship between, 175 consumer surplus as indicator of, EU and USA, 128 courts, see courts economic enquiries, revealed by, 82 effects tests, role and, 102–106 Harper Review on, 10, 143–46, 209–10, 215, 224 opposition, ACCC, 11, 13 positive, negative competition effects tests and, EU and USA, 128 Queensland Wire, 142 role, effects tests and, 102–106 SLC tests and, 130 Structure-Conduct-Performance forensic determination implications, 80 efficiency incentives and assessment: firms, 123 efficient conduct, 9, 130, 229–32, 236 efficient markets, 2–3, 49, 60 judicial use of term, 117 legal concept, 125, 126, 130, 226 meaning, 224 standard, 25 supply and demand, 117 efficient use: of fixed capital and resources, Queensland Wire, 149 elasticity of demand, 40 decision-making, subjective, 123 equilibrium price and output, 62 Lerner Index correlating with, 52

Index  273 market power: caused or contributed, 52 correlation between, v dependence of, 51 derived from, 93 market share and, 53 profit maximising output and, 50, 92 revenue, 61 electricity market: NT Power, 215–16 end users: market segments and, 153 endogenous sunk costs: industry concentration affected by, EU, 109 energy markets: market manipulation, 59 energy prices: United States, 59 engineering: social, see social engineering enhancing efficiency, 10 enhancing social efficiency, 224 entry: barriers, 58, 62 in imperfectly competitive markets, EU, 110 Melway, 150 predatory pricing, 158 recoupment, 166 Rural Press, 174–75 consumer welfare increased by equally efficient firms, 70 without costs under perfect competition, 69 deterrence theories, 70 incumbent response theories, 70 self-selection for, 119, 168 equality before the law, 47 equally efficient competitors: tests, EU, 110–11 zero-sum competition and, 62–63 equally efficient firms, see firms equilibrium: allocative efficiency, achieving, 49 market, consumer surpluses under, 45 markets at, 46 monopoly, see monopoly output, see output perfect competition, see competition; perfect prices, see prices productive efficiency achieving, 49

equitable distribution of surplus between producers and consumers: competition law objectives, 46–47 equitable doctrines, 12 equity: competition law and, 121 debt leveraging, Queensland Wire, 149 distributional, see distributional equity distributive, 85–86 efficiency and, distinction between, 128 jurisprudence doctrines, competition law and, 12 merger efficiencies and, ACCC, 121 redistributive, 85–86 returns on, Queensland Wire, 149 as social objective, 121 EU, see European Union European Commission (EC), see European Union European Union: abuse of dominance, 107 average costs, 108 barriers to entry, 110 barriers to exit, 110 business decision tests, rational, 111 CJEU: competition on the merits, 107–108 competitor, not competition process, protection, 107 dominant firms protecting own interests, 111 effects, actual, establishing, 109 Intel case, 35, 107, 109, 112, 185 market excluding equally efficient firms, 109 common market, 111 competition: defences: efficiency, 111 meeting competition, 111 effects: competition analysis, 97 inferring from nature of conduct, 97 negative, positive efficiency effects tests and, 128 efficiency surrogate, 109 harm to, 107, 108 competition-meeting grounds, 110 efficiency grounds, 110 necessity grounds, 110 see also process in this sub-entry below

274  Index imperfect, 110 meeting: conduct harming competition grounds, 110 defence, 111 on the merits, 107–108 perfect, 110 process: conduct weakening, 109 harm to, 109, 112 protection, not competitors, 127, 228 as surrogate for efficiency, 109 zero-sum: incumbent firms, 110 Kaldor–Hicks efficiency, 109 competitive strategies: industry concentration affected by, 109 competitors: economies of scale denied, 108 equally efficient, 108, 112 harm to, 107, 108, 109, 112 lower prices, dominant firms meeting, 111 market power as freedom from constraint by, 112 protection, 107, 127, 228 weakened, 108 conduct: commercial, dominant firms, 107 form, inferring effects from, 112 harm and, causal nexus between, tests to find, 112 harming competition, justification, 110 harming competitors, 109 weakening process of competition, 109 consumers: surplus as indicator of efficiency effects, 128 welfare, 17, 109–10 cost of capital, 108 costs: average, minimising, 108 benchmarks, 108 courts, see CJEU above decision-making: dominant firms, 108 defences: competition, see competition above efficiency, 111 diminishing returns, law of, 108 dominance, abuse of, 107 dominant firms: commercial conduct, 107 decision-making process, 108

equally efficient competitors, 108 lower prices of competitors, meeting, 111 profit-maximising conduct, 108 protecting own interests, 111 dual purpose problem, 108 economic approach, 23 economies of scale: competitors, denied to, 108 exclusion, causing, 108–109 effects: actual, establishing, 109 competition, see competition above efficiency, see efficiency below form of conduct, inferring from, 112 purpose and, inter-linked and inextricable meaning, 112 purpose as staging point towards determining, 109 efficiency: average costs, minimising, 108 competition surrogate for, 109 conduct harming competition grounds, 110 defence, 111 effects: consumer surplus as indicator of, 128 positive, negative competition effects tests and, 128 financial, 108 gains offsetting, 107 indirect approach focussing on, 107 Kaldor–Hicks, 109 operational, 108 purpose as surrogate for, 127 endogenous sunk costs: industry concentration affected by, 109 enforcement guidance, 23 entry barriers, 110 in imperfectly competitive markets, 110 equally efficient competitors: dominant firms, 108 tests, 110–11, 112 European Commission: Baxter’s acquisition of Gambro approval, 206 commercial transactions supervision, 135 competitor, not competition process, protection, 107 guidance, 107 effects test, advocating, 89–90 equity, concern for, 121 market power as freedom from constraint by competitors, 112 tests use, 112

Index  275 excessive pricing, prohibition, 111 exit barriers, 110 financial efficiency, 108 firms: strategising, 110 see also dominant firms above gains: offsetting efficiency, 107 harm: competition, see competition above competitors to, 107, 108, 109, 112 conduct and, causal nexus between, tests to find, 112 effects, inference of, 112 theory of, 208 imperfect competition, 110 industry concentration: market size and, relationship between, 109 institutional arrangements agency-centred, 87 Kaldor–Hicks efficiency, 109 law of diminishing returns, 108 marginal costs, 108 market position, 110 market power: as freedom from constraint by competitors, 112 market size: exclusion, causing, 108–109 industry concentration and, relationship between, 109 market structure, 110 merger control: consumer welfare standard, 109 on the merits, competition, 107–108 necessity: conduct harming competition grounds, 110 neoclassical model of perfect competition, 110 operational efficiency, 108 perfect competition, 110 predatory pricing, 110 prices: excessive pricing, prohibition, 111 lower, of competitors, dominant firms meeting, 111 predatory pricing, 110 profits: maximising, dominant firm conduct, 108 purpose: effects and, inter-linked and inextricable meaning, 112 no prohibition of without effect, 102

as staging point towards determining ultimate effects, 109 as surrogate for efficiency, 127 tests, 109 rational business decision tests, 111 SLC tests: Australian version difference, 112 dominant firm conduct, not preferred for, 110 as guiding principle, 109–10 structural analysis, 110 TFEU, Article 102: 107, 111 zero-sum competition, see competition above evaluative issues: efficiency, 84 evidence: of demand response, 218 probative, 149 evidence-based economic analysis: margin squeeze, 232–33 predatory pricing, 233 raising rivals’ costs, 234 refusal to supply, 232–33 excess profits, see profits excessive prices, see prices excessive pricing, see prices excessive profits, 48 excessively high prices, 48 excluding rivalry, 184, 185 exclusionary conduct, 103 USA, 156 exclusionary dominant firm conduct, 223 exclusionary purpose, see purpose exclusive contracts: Cement Australia, 184–85, 190, 192, 196–97 firms lacking market power, 196–97 exclusive dealing, 23, 33, 151, 153, 190, 209, 210 exit: barriers, 62–63, 65, 69 Cement Australia, 186–87 EU, 110 exit costs as, 157 predatory pricing, 158 real world, 123 zero-sum competition, 158, 165 costs, 9, 226 Boral, 165 as exit barriers, 157 incumbent firms, 126 perfect competition without, 69 self-selection for, 168 zero-sum competition, 119 social costs of causing, 69

276  Index factual matrices, 131 Baxter, 202–208 Boral, 158–63 Cement Australia, see Cement Australia Melway, 149–51 Queensland Wire, 136–42 Rural Press, 170–74 failure of markets, 19 falling prices, see prices: reductions fight for survival, 226 financial accounting: opportunity costs and, 21 financial efficiency: firms, 168 assessing, 126 findings and recommendations, 223–24 competition law implications, 234–37 competition paradigms and norms of conduct, 225–27 conduct: norms and competition paradigms, 225–27 particular kinds forensic assessment conclusions, 232–34 market power concept, 225 misuse of market power forensic assessment, 227 competition is the process by which firms compete for the custom of consumers proposition, 227, 229 competition law protects process of competition, not competitors proposition, 227, 228–29 inter-relationship between competition and efficiency, 229–32 monopolists not to be deprived of fruits of success proposition, 227–28 finished products: dominant firms supplying to competitors: Queensland Wire, 140 firms: behaviour: vertical integration, 194 competition risks management, anticompetitive conduct representing, 231 conduct: manipulative, assessing, 230 socially harmful, assessing, 230 costs, objective indicators of, 123 decision-making, subjective process, 123 dominant, see dominant firms

economic activity in, 118 efficiency assessing, 123 equally efficient: competing for economies of scale, 174 entry, consumer welfare increased by, 70 exclusion, 10, 126 forced out of market, 69–70 market exclusion, 62 new entrants, 125 exclusion, 126 not required to satisfy market demand, 62 self-selection for entry or exit under zero-sum competition, 10 social costs of causing exits, 69 financial efficiency, 168 assessing, 126 free exit, 122 perfect competition market mechanism, 122 incumbent: exit costs, 126 innovation incentives, 123 lacking market power: exclusive contracts, 196–97 leaving market: perfect competition, 156, 157 zero-sum competition, 156 less efficient: forced out of market, 69–70 market exclusion, 62 social costs of causing exits, 69 locked in zero-sum competition, 125–26 manipulative conduct of, assessing, 230 market manipulation, 123 with market power, 104–105, 126 without market power, 104–105 market power in upstream markets: squeezing rivals’ margins in downstream markets, 185 market size of firms, market exclusion, 120–21 more efficient: exclusion, 126 number in market, market size determinant, 164 price-takers, 51 profits: maximising, 123 objective indicators of, 123 relative efficiency, 168 revenue, objective indicators of, 123 socially harmful conduct of, assessing, 230

Index  277 strategising, EU, 110 survival competition, 231, 233 technical efficiency, 168 assessing, 126 two locked in zero-sum competition, 148 fixed capital: efficient use measures, Queensland Wire, 149 fixing prices, see prices flux, demand curves in, 68 fly ash, 181–98 Foer, Albert, 167 forensic determination of efficiency, 80 forensic difficulties of ascertaining competitive level, 120 forensic problems: market power, 12 forensic tests: efficiency, 11, 27–28 foreseeability: competition law predictive tests requiring, 99 effects tests, 99 form-based approach, 86 forms: market size, market exclusion, 120–21 France Télécom, 111 free entry, 226, 229 free exit, 122, 226, 229 free market ideology: capitalism, 44 freedom of choice: allocative efficiency and, 58 French J, 83–84 French, Robert, 17, 42, 82–83 Fresenius, 202–204, 205, 207–208 Freyer, Tony, 231 Frontier Economics, 39, 40 gains, efficiency, see efficiency Gal, Michal, 227–28 Gambro, 202–204, 205–206, 208 game theory, 9, 63 Gaudron J, 71, 158 Gavil, Andrew, 1, 13, 16, 91, 97, 101, 103–105, 133 geographical extent of markets, 191, 194 Gleeson CJ, 157, 159 global economy changes, 48 global positioning system (GPS), 150 goals: allocative efficiency as, 16 productive efficiency as, 16 see also objectives

Golan, A, 66 good faith: directors, 71 goods: perfect competition assumptions, 119 scarce, 46 substitution: perfect competition equilibrium, 56 Gormsen, Liza, 112 government: executive branch: polycentric decision-making, 83 public interest functions, 83 GPS (global positioning system), 150 Greenwood J, 181, 185–87, 191–93, 196 Gummow J, 158 haemodialysis (HD) products, 203–204, 206, 208 Hanks, F, 90–91, 133–34, 142, 152 Hanson, Cement Australia case, 182, 185, 188 harm: assessment: Cement Australia, 192–93 Baxter, 204 bundling, 200–201, 204, 207, 208 burden of proof, 218–19 to competition, see competition to competition process, see competition: process conduct: and, causal nexus between, tests to find, EU, 112 dominant firms, see dominant firms consumers, 48 to economic efficiency, 69 effects: dominant firm rational business decisions, 91–92 inference of, EU, 112 purpose as screen isolating, 127 to efficiency: principal mischief SCP seeking to address, 79 to efficiency indicator: competition effects, 102 EU, see European Union to itself on dominant firm raising rivals’ costs, 186, 193 from market power derived from perfect competition, 118 measure of, 28

278  Index to new entrants, Queensland Wire, 143 plaintiffs proving issues pertaining to, 218–19 prevention: predictive SLC tests, 230 to process of competition, see competition: process proof not required to demonstrate competition effects, 192–93 purpose indicator of, 201 to rivals, see rivals theory of, 24–25, 93 efficiency as basis, 219 EU, 208 market power misuse, underlying, 236 USA, 208 threatened, injunctions to prevent, 168 from vertical foreclosure, 179–80 Harper Review, vi access regime, 135–36 American Bar Association submission to, 218 anti-competitive effects, 153 Baxter, 201, 208–10 Boral, 163–66 causation, 196, 197, 234–35 Cement Australia, 196–98 competition: effects: negative, positive efficiency effects tests and, 128 paradigm, 23 protection, not competitor protection, 127 conduct, 127 failing SLC test, power to authorise, 80 consumer welfare, 17, 102 effects test, 89, 90, 95, 98, 102–106, 156 efficiency, 197–98, 217 benefits, 235–37 economic, 11, 13–14 effects, 209–10, 215 enhancing, 10 offsetting, 224 positive, negative competition effects tests and, 128 SLC effect tests, 143–46 ‘exclusionary purpose’ requirement removed, 81 exit barriers, 62 focus on isolated issues, 127 hypothetical competitive markets, 127 limit pricing condemnation, 236 margin squeeze, 232–33

market power, 153 causal connection with SLC, 82 misuse recommendations, 5–7 recommendations, 5–7, 9 meeting competition, 174–75, 233 Melway, 151–53 monopolisation: American Bar Association submission, 218 monopoly, 135 NT Power, 215 perceived excessive pricing condemnation, 235 predatory pricing, 116, 155–56, 233 Queensland Wire, 142–46 reasoning to advocate change, 127 refusal to supply, 232–33 Rural Press, 174–75 section 46(2), ACCC opposition, 6–7, 217 SLC tests, 103, 151, 224 subjective purpose test, 113 taking advantage requirement, 105 removed, 81, 228 Hayne J, 158 HD products, see haemodialysis products health insurance, 208 healthcare, government policy, 208 Heeb, R, 108, 115 Heerey J, 91, 150, 151–52, 155–56, 159–60 Heidelberg Cement, Cement Australia case, 182 Heidrich, 182 Hicks, John, see Kaldor–Hicks efficiency Hicks–Kaldor efficiency, see Kaldor–Hicks efficiency Higgins, Richard, 178–79 high industry concentration, 78, 79 high volume low value products, 202 Hilmer Report, 2, 41, 47, 53–54, 61, 135 Holcim, Cement Australia case, 182, 188 holding costs, 157 horizontal demand curves, 45, 50, 54–55 hostile intent, 224, 235 evidence: influencing, courts, 116 USA, 114, 156 Hovenkamp, Herbert, 4, 11–12, 17, 18, 43–44, 49, 50–51, 57–58, 73, 110, 114, 116 Hylton, Keith, 4, 87, 114–15 hypothetical competitive level, 121 hypothetical competitive markets, see competitive markets

Index  279 hypothetical markets: courts adoption as benchmark, 133–34 perfect competition, 23 idealised foundations: concept of market power built on, 60 identification: downstream markets, 179 imperfect competition, see competition imports: Queensland Wire, 138–39, 141, 144 income maximisation, 16 inconsistent objective: allocative efficiency as, 49 productive efficiency as, 49 incumbents: dominant firms, 125 firms, 126 market manipulation, 166, 175–76 market power misuse, 70 meeting demand, 210 new entrants seeking to destroy, zero-sum competition, 123 profits foregone by: recoupment, 210 response theories, entry, 70 response to new entrants, 61–70, 73–74 Rural Press, 174–75 seeking to survive, zero-sum competition, 123 Independent Fly Ash Brokers, Cement Australia case, 188, 189 indeterminacy, legal, 11 indeterminate competitive level, 178 industries: concentration, 66 Boral, 164 high, 78, 79 market size and, relationship between, EU, 109 costs: supply curves, 54 economic activity in, 118 inefficiency, 10 conduct, see conduct monopoly, 54, 215 new entrants, 125 outcomes: zero-sum competition, 80, 123, 124, 152, 179, 227, 231, 236 plaintiffs’ burden of proof, 219 social, 130

surpluses transferred to consumers, 166 toleration, 131 zero-sum competition price to pay, 166 injunctions: interlocutory, 99 to prevent threatened harm, 168 to restrain prohibited conduct, Canada, 105, 113 innovation: Boral, 163 dynamic potential for, 124 incentives, firms, 123 inputs: bidding, 194 bidding wars for, see bidding wars competitive bidding for, 194 costs: raising by bidding by contested tenders, 194 overbuying, harm from, zero-sum competition, 179 prices: higher, downstream market power mitigating, 194 raising by bidding wars, 194 surpluses: transferring upstream, 194 institutional arrangements: agency-centred, EU, 87 court-centred, USA, 87 merger efficiencies, courts exclusion from assessing, 110 see also Australian Competition and Consumer Commission; Australian Competition Tribunal institutional division of court functions, 6, 78–86, 193, 216–17, 236 institutional implications, 214–19 Cement Australia, 193 insurance, health, 208 Intel case, 35, 107, 109, 112, 185 intent: condemnation, USA, 114 hostile, see hostile intent inferred from conduct, USA, 113, 114 inferred from effect, USA, 113 objective, USA, 113 specific, USA, 113 court apparent rejection, USA, 115 inferring, USA, 116 to monopolise, USA, 116 subjective, USA, 113 substantial subjective, 113, 143

280  Index interest: community, 46 groups: distributive issues between, 47 naked transfers, 47 see also consumers; producers public, see public interest interlocutory injunctions, 99 international controversy: market power misuse, 16 international debate: on unilateral dominant firm conduct, 16 intervention: in markets, 57 scepticism, United States courts, 5 investment: resources, 48 iPhone, 62 Jacobs, Michael, 18 Jenkins, H, see Niels et al Jones v Dunkel, 219 judges, see courts jurisdiction, see courts justice: discretionary practical, 12 Justice Department, United States, see United States: Department of Justice Kaldor, Nicholas, see Kaldor–Hicks efficiency Kaldor–Hicks efficiency, 109, 124, 148 Karp, L, 66 Kavanagh, J, see Niels et al Kaysen, Karl, 79–80, 82 Kelliher, J, 59 Kemp, Katherine, 91–92 Kirby J, 46, 121–23, 153, 157, 163, 164, 230 Kogan Creek, Cement Australia case, 183, 190–91 Kovacic, William, 5 Lafarge, Cement Australia case, 182 Lancaster, K, 21 Landes, William, 21, 40, 53, 54 law: competition, see competition consequences of, 80 of diminishing returns: EU, 108 dominance laws, 127 equality before, 47 manipulation, see manipulation

market power in, 39–43 conclusion, 76–77 economic models and competition paradigm, 49–59 efficiency concepts, 43–49, 75–76 incumbent response to new entry, 61–70, 73–74 market manipulation, 75–76 laws in securities markets, 59–61 norms in zero-sum competition, 70–75 policy objectives, 43–49 zero-sum competition, 75–76 as normative rules system, 71 polycentric, 219 reform initiatives, consequences, 80 securities industry: effacement market in, 8 USA: juries’ role, 114–15 thinking since 1970s, 40 see also corporations law; criminal law; legislation; policy and law; rule of law; tort law lawyers: economists and, opposed perspectives, 42 worldviews, 64–65, 119, 130 Leader, Rural Press case, 173, 174 legal benchmarks, see benchmarks legal concepts, 3 market power, see market power legal determinacy, 3 social efficiency, 124 legal indeterminacy: economic efficiency, 11 legal legitimacy, 4 legal liability: positive standard, 81 legal precedent: role in economic activity regulation, deeper understanding required, 4 legal theoretical problems: market power, 12 legal theory of liability, see liability legislation: competition, see competition limiting personal freedoms, constitutionality of, 73 see also law legitimate business justification, 81 legitimate business rationale, 81, 218 legitimate business reasons, 152, 186 Lerner Index, v, 52

Index  281 less efficient firms, see firms lessening competition, 10 see also substantially lessening competition liability: legal theory of, 42 efficiency as basis, 219 for market power misuse, 236–37 polycentricity and, 84 proxy tests, 74 surrogate tests, 74 unicentric, 84 market power misuse, positive and normative legal theory, 8–11, 126 no fault, see no fault liability positive and normative legal theory for market power misuse, 8–11, 126 strict, see strict liability theory of, 25–28 limit pricing, 165, 228 Harper Review condemnation, 236 USA, 115–16 Linkline case, USA, 135 Lipsey, R, 21, 54, 57–58 lose-lose outcomes, 65 losses: assets values, 157 deadweight, 57 efficiency, 100 predatory pricing: recoupment from price cutting, 155 low volume high value products, 202 MacCormick, Neil, 71 management accounting: opportunity costs and, 21 manipulation: laws: securities markets, 117, 125, 126 market power, v, vi, 9 power of the market, 51, 61, 93 see also market manipulation; securities market manipulation manipulative conduct: efficiency and, relationship between, 125 Mansfield J, 169–70, 206, 209 margin squeeze, 133–54, 227 cartel conduct, 232 evidence-based economic analysis, 232–33 firms with market power in upstream markets on rivals’ margins in downstream markets, 185 Harper Review, 232–33

market manipulation 232–33 overbuying aspects in common with, 179 prices: excessive, 233 fixing, 232 rivals, 185 marginal costs: average costs equalling, 54 EU, 108 marginal revenue equalling, 45 monopoly, 58 curves, 54 producers, 58 profit-maximising output and, 92 marginal rate of transformation (MRT), 56 marginal rate of substitution (MRS), 55 marginal revenue, 60, 66–69 imperfect competition, 69 marginal costs equalling, 45 oligopoly: supply response, 148 profit-maximising output and, 92 supply and, 122–23 marginalisation: double marginalisation problem, 179 margins, contribution, 161 market concentration, 1–2 market conditions: changed, reduced prices as competitive response to, 158 market definition: arbitrariness, 114 market demand, 9 conduct responding to, see conduct equally efficient firms not required to satisfy, 62 monopoly expectations to meet, 125 NT Power, 216 price, responding to, 49 privately efficient responses to, 226 response to price/output decisions relationship, v–vi socially efficient responses to, 226 market efficiency: Queensland Wire, 148 zero-sum competition addressing, 49 market equilibrium: consumer surpluses under, 45 producer surpluses under, 45 market exclusion: consumers, 121 equally efficient firms, 62

282  Index less efficient firms, 62 market size of firms, 120–21 market failure, 19 market intervention, 57 market manipulation, vi approach, 53, 74–75, 128 Baxter under, 210–11 Boral, 166 buyers: contributing to bidding contests, 225 powers, 230 Cement Australia under, 198 competition effects, assessing, 93 conceptualised market power as, 16 condemnation by misuse legal theory, 9 conduct, 60, 226 inefficient, condemnation assessments, 232 cornering, 59, 185 demand response, 208, 230 efficiency and zero-sum competition and, 124–26 efficiency considerations, 167 energy markets, 59 firms, 123 incumbents, 166, 175–76 laws in securities markets, 59–61 legal liability for section 46 contravention, 236 margin squeeze, 232–33 market power: conceptualised as, 14, 16, 19 misuse redefinition as, 230–31 meeting competition, approach, 233 Melway under, 154 new entrants, 166, 175–76 predatory pricing, approach, 233 price, prohibited: normal market price volatility, distinguishing between, 59 proposed approach, 7–12 Queensland Wire under, 146–49 raising rivals’ costs, approach, 234 raw material sellers creating bidding wars, by calling for public tenders, 225 refusal to supply, 232–33 Rural Press under, 175–76 securities, laws, 2–3 securities markets, laws in, 59–61 sellers’ powers, 230 strategic market power, reconciliation, 51 testing, 129–219 theory, 24–30

zero-sum competition and efficiency and, 124–26 see also power to manipulate the market market mechanism, 62 normal, zero-sum competition as, working under real world conditions, 123 of perfect competition, 121 free exit of firms, 122 market performance: competition effects, 93 inferences, SCP approach providing chain of causation, 79 market position, EU, 110 market power: as absence of constraint, 147 abuse, competition law, 73 advantage of, taking, 61 alternative legal conception, 61 anticompetitive effects: and, causal link between, 164 contributing to, Melway, 150–51 Baxter, 204, 208–209, 210 Boral, 159, 163–65 caused or contributed by elasticity of demand, 52 Cement Australia, 184, 186, 209 competition effects, resulting from, 235 competitive constraints, 58 conceived as power of the market, see power of the market conceived as power to manipulate the market, see power to manipulate the market concept: deriving from hypothetical competitive markets, 79 findings and recommendations, 225 reconsideration, 93 conceptualised, 7–8 as market manipulation, 16 conduct and: causal connection, 174 Harper Review, 127 consumer responsiveness and, 52 contractually-based, 210–11 defined, 39, 50, 60 Canada, 106 demand response, 225 dependence on elasticity of demand, 51 derived from perfect competition, 118 digital economy application, 30–31 distributional equity, 48 dominant firms in downstream markets, 178–79

Index  283 dominant firms taking advantage of, 23 dominant firms without in downstream markets, 194 in downstream markets, 179, 195 in economics, see economics efficiency prevailing over, 79 elasticity of demand: and, correlation between, v derived from, 93 firms with, 104–105, 126 firms without, 104–105 exclusive contracts, 196–97 forensic problems, 12 as freedom from constraint by competitors, EU, 112 harm from, 118 Harper Review, 153 recommendations, 5–7, 9 idealised foundations, concept of built on, 60 in law, see law legal concept, 8, 9, 15–16, 19, 25–26 legal theoretical problems, 12 legitimate business reasons indicating against use of, 186 manipulation, v, vi, 9 market definition determinative of, USA, 114 market manipulation, conceptualised as, 14, 16, 19 market share and, 53 market structure as primary source of, reconceptualised under zero-sum competition, 79 mergers creating or enhancing, 14 misuse, 3 borderline prices and, 134 conduct norms, 71 court recommendations, 5–7 efficiency considerations, 217 forensic assessment findings and recommendations, 227 competition is the process by which firms compete for the custom of consumers proposition, 227, 229 competition law protects process of competition, not competitors proposition, 227, 228–29 inter-relationship between competition and efficiency, 229–32 monopolists not to be deprived of fruits of success proposition, 227–28 Harper Review recommendations, 5–7 incumbents, 70 international controversy, 16

legal approach inconsistency, 61 legal theory of liability for, 236–37 liability for, positive and normative legal theory of, 8–11, 126 market manipulation condemnation, 9 overseas experience, 22–24 problem, 3–4 recommendations, 5–7 problem recommendations, ACCC, 5–7 redefinition as market manipulation assessed by reference to efficiency, 230–31 reform process, 4, 5–7 rule of law concerns with jurisprudence on, 12 theory of harm underlying, 236 nature, elucidation of, 19 normative legal theory centrepiece, 224 opinions differing, 22 perfect competition model, derivation from, 118 Pfizer, 211–12 positive legal theory centrepiece, 224 as power of the market, see power of the market as power to manipulate the market, see power to manipulate the market predatory pricing use restraints, 167 price reductions redressing, 121 prohibited use, 26 qualitative analysis, predatory pricing, 234 reconceptualisation, zero-sum competition, 236 redressing by price reductions, 121 regulation, 18, 19 Rural Press, 169–77 second-best way ro assess, USA, 114 self-correction, 18 SLC test: application, irrelevance to, 23 causal connection, 82 strategic, see strategic market power structuralism conception, 9 substantial, 51 Melway, 151 Queensland Wire, 143 taking advantage of: Melway, 150 threshold for, 26 in upstream markets, 179 firms squeezing rivals’ margins in downstream markets, 185

284  Index USA, see United States see also power of the market; power to manipulate the market market relationships: zero-sum competition for, 31 market response: Melway First Instance, 145 market risks: ongoing supply relationships and, 134 market segments: end users and, 153 market share: Boral, 159 elasticity of demand and, 53 market power and, 53 Queensland Wire, 139 role of, 53 market sharing, see cartel conduct market size, 66, 67 efficiency and, 168 exclusion, causing, EU, 108–109 industry concentration and, relationship between, EU, 109 number of firms in market determinant, 164 unmet demand and, 125 market structure: competition process, equated with, 80 concentrated, 1 conduct, not determining, 79 EU, 110 performance, not determining, 79 as primary source of market power, reconceptualised under zero-sum competition, 79 roles, 2 as surrogate test, 80 tests, court administration of, 80 vertical integration as anticompetitive aspect of, 180 market supply curves, 45 markets: as analytical constructs, 44 balance of power in, 230 definition: arbitrariness, 114 opinions differing, 22 USA, 114 downstream, see downstream markets economic activity in, 118 efficient, see efficient markets at equilibrium, 46 excluding producers, 48 geographical extent, 191, 194

hypothetical, see hypothetical market profits sufficiency, 48 relevant, see relevant markets resources investment, 48 self-selection for entry or exclusion, 48 upstream, see upstream markets Mason, Anthony, 40–41 Mason CJ, 71 Mason, Edward, 82, 163, 230 matrices, factual, see factual matrices maximisation: income, 16 monopoly profits, 56 production, 16 profits, perfect competition, 56 McHugh J, 163 McMahon, Kathryn, 4, 42 measures of competition effects, 80 medications, patented, 211–12 meeting competition, see competition Melway, 41 factual matrix, 149–51 under Harper section 46: 151–53 under market manipulation, 154 Melway FFC, 150, 151–52 Melway First Instance, 145 mergers: anticompetitive, prohibition, 14 control: consumer welfare standard, EU, 109 creating or enhancing market power, 14 efficiencies, 224 courts exclusion from assessing, 110 equity and, ACCC, 121 vertical, see vertical mergers vertical foreclosure, 179 vertical integration, 179 merits: competition on the, 8, 223 merits-based strategies, 91 Merkel J, 145, 150, 153 Merrett, Alexandra, 13 Millmerran, Cement Australia case, 183–84, 186–88, 192, 196–97 mischief and institutional response: Australia, 78, 87–88 institutional division of functions between courts and agencies, 81–82 administrative agencies’ roles, 84–86 court role, 82–84 SCP approach in Australia, 78–81 European Union, 86–87 United States, 86–87

Index  285 misuse of market power, see market power monopolisation: American Bar Association submission to Harper Review, 218 attempted, USA, 116 law, USA, 4–5, 87 legitimate business rationale rebutting claims of, USA 218 manipulation of markets, USA, 93 no fault regime rejected, USA, 92 USA, see United States monopoly, 19 competitive prices, 54, 57, 66 competitive process regulation, 62 consumer surpluses and, 54 cost and revenue relationship, 65–66 demand, meeting, 195 distributional equity, 48 economic analysis, 44 equilibrium, 52 inefficiency, 54 profit maximisation, 56 Harper Review, 135 inefficient, 215 marginal costs, 58 curves, 54 market demand, expectations to meet, 125 market power source in, 52 models, 20, 57 neoclassical economic model of, 48 new entrants, 65–66 not to be deprived of fruits of success proposition, 227–28 output, 55 decisions, 51–52 increases, 124 perfect competition comparisons, 21 power: definition, 52 prices, 47, 50, 55 decisions 51–52 new entry incentives, 66 regulation, 54, 57, 61–62 private efficiency, 125 producer surpluses and, 54 profits, 133 dissipation, 66 maximising, 56 behaviour, 125 output, 133 above normal, Queensland Wire, 141 revenue and cost relationship, 65–66 social efficiency, 125

supply curves under, 45 virtual, Queensland Wire, 143 more efficient firms: new entrants exclusion, 126 Motta, M, 200 MRS (marginal rate of substitution), 55 MRT (marginal rate of transformation), 56 multiple contemporaneous tenders, 209 Murray Valley Standard, 170 mutually assured destruction, 65 Nagarajan, Vijaya, 85–86 naked transfers, 74 interest-groups, 47 Nash equilibrium, 64–65, 119 necessity: conduct harming competition grounds, EU, 110 negating elements of contravention: competition defences, 101 negative competition defences, 102, 103 negligence: duty of care, foreseeability requirements, 98 Nelson, P, 218 neoclassical economic models, 19–20 competition law use, 42 competitive market of, 8 dynamic gaps in, 48 of monopoly, 48 neoclassical focus on marginal costs: EU, 108 neoclassical model of perfect competition, 110, 223 ‘net back’ prices, 145–46 new entrants: equally efficient firms, 125 exclusion, 126 harm to, Queensland Wire, 143 incentives, monopoly prices, 66 incumbents’ response to, 61–70, 73–74 inefficient, 125 incumbent dominant firms, 125 monopoly obstructing, 215 lack, Queensland Wire, 141–42 lacking scale, 209 market manipulation, 166, 175–76 monopoly, 65–66 more efficient firms, exclusion, 126 predatory price cutting by incumbents, 209 seeking to destroy incumbents, zero-sum competition, 123 seeking to enter, 165

286  Index unmet demand and, 125 zero-sum competition implications, 69 New Zealand: actual market as benchmark, 91 newspapers: Rural Press, 169–77 Niels et al, 108–110, 111, 200 no fault liability, 92–93, 105, 127, 130, 165, 175, 198, 235 no fault prohibitions, 130 Boral, 164 Cement Australia, 198 Rural Press, 175 non-dominant firms, 9 non-economic considerations, 2 non-economic objectives: ACCC and ACT weighing up, 219 non-market values, 71 non-price competition, 53 non-zero-sum games, 63–64, 119 normal competitive conduct, 223 normal market mechanism: working under real world conditions, zero-sum competition as, 123 normative issues: efficiency, 81, 84 normative legal theory of liability for misuse of market power, 8–11, 126 normative role of judges in applying standards, 83 normative rules of law, 71 normative standard, 25 norms: background, 47 conduct, see conduct constitutional, 71 in zero-sum competition, 70–75 NT Power, 99, 124, 214–16, 233 Nucon, Cement Australia case, 189 Nucrush, Cement Australia case, 189 NxStage, 207 O’Bryan, Michael, 4 objective reasonableness test, USA, 114–15 objectives: antitrust policy, 43–44 competition law, see competition: law non-economic, see non-economic objectives redistributive, see redistributive objectives social, equity as, 121 see also goals

oligopoly: competition: strategy analysis, 63–64 marginal revenue: supply response, 148 peaceful coexistence, 63 supply: curves under, 45 marginal revenue, response, 148 ongoing relationships, supply, 134 open economy, 57 opportunity costs: financial accounting and, 21 management accounting and, 21 optimal output, 44 optimal prices, 44 optimisation: congruent, 124 efficiency, 123–24 optimum social efficiency, 92 outcomes: inefficient, zero-sum competition, 123, 124 output: artificial restrictions: dominant firms, 92 by increasing prices, 50 competitive: in monopoly model, 57 competitive level: dominant firms conduct benchmark, 90 decisions: demand response to, 198 dominant firms, see dominant firms efficiency and: Queensland Wire, 141 equilibrium: consumer response, 62 elasticity of demand, 62 perfect competition, 54 increases: dominant firms, 125 firms locked in, zero-sum competition, 130 monopoly, 124 beyond profit-maximising position, 226 zero-sum competition, 125 maximisation under perfect competition, 119 maximising profits, 50 monopoly, see monopoly optimal, 44 perfect competition equilibrium, 54

Index  287 profit-maximising, 50, 56 elasticity of demand and, 50, 92 marginal costs and, 92 marginal revenue and, 92 monopoly, 133 overbuying, 178–79 overseas experience: unilateral dominant firm conduct, misuse of market power, 22–24 overseas principles application to effects and purpose tests, 109 pandemics, 49 parallelism, conscious, 65 Pareto efficiency, 123 Part IV, see Competition and Consumer Act 2010 Part IVA, see Competition and Consumer Act 2010 patented medications, 211–12 PAWA (Power and Water Authority), 214–16 payoff matrix, 64 PD (peritoneal dialysis) fluids, 202–10 peaceful coexistence: imperfect competition, 119 oligopoly, 63 zero-sum competition, 73 penalties: assessment, Cement Australia, 192–93 Pengilley, W, 134 perceived excessive pricing, 135 perfect competition, see competition performance: market structure not determining, 79 markets, see market performance peritoneal dialysis (PD) fluids, 202–10 Perloff, J, 66, 157 permitted conduct, see conduct personal freedoms: legislation limiting, constitutionality of, 73 Pfizer, 35, 96, 111, 211–12, 205, 206, 211–13, 234 Pharmaceutical Benefits Scheme, 211 Pincus J, 136–37, 140–42, 146, 149 Pioneer, Boral case, 159, 162, 166 plaintiffs: proving issues pertaining to efficiency, 218 platform markets, 170 see also digital platforms policy: antitrust objectives, 43–44 choices: informed by consequences, 17–18

objectives, 43–49 see also policy and law policy and law: competition independent review panel, 6 polycentric decision-making, 83–84 polycentric jurisdiction: ACCC and ACT, 219 polycentric law, 219 polycentricity: legal theory of liability and, 84 Portland cement, 181 positive and normative legal theory of liability for misuse of market power, 8–11, 126 positive economic analysis: policy choices elucidation by, 17–18 positive standard, 25 Posner, Richard, 20, 21, 40, 51, 53, 54, 57, 62 power, see balance of power; buyers; countervailing power; dominant firms; economic power; monopoly; power of the market; power to manipulate the market; predictive power Power and Water Authority (PAWA), 214–16 power of the market, 9 conduct responding to market demand, 130 manipulation, 51, 61, 93 market power as, 9, 53, 225 power stations, coal-fired, 182–84 power to control markets: Queensland Wire, 137 power to manipulate the market: market power as, 60, 105, 125, 226, 236 Queensland Wire, 147 predation: assessing, 210 claims, objective reasonableness test, USA, 114–15 predatory pricing, 227 ACCC, 155–68 Boral, see Boral causation, 233 competition effects qualitative analysis, 234 cost-based approach, 157, 163 cost–price issues, 233 cutting by incumbents, 209 demand shift reductions, 233 entry barriers, 158 EU, 110 evidence-based economic analysis, 233 exit barriers, 158

288  Index Harper Review, 155–56, 233 losses recoupment from price cutting, 155 market manipulation approach, 233 market power: qualitative analysis, 234 use restraints, 167 predictive and qualitative SLC tests, 233 purpose tests, 156 recoupment: US approach, 155–56 tests, 210 USA, 116, 155–56, 209 predictability, 86 policy choices tested against, 17–18 predictive effects tests, 224 predictive nature of effects tests, 96–97, 98 predictive power: economic theory, 20 predictive recoupment tests, 162 predictive SLC tests, 97, 99 Cement Australia, 185, 199 harm prevention, 230 margin squeeze, 144 predatory pricing, 162, 164, 233 qualitative inference, 235 raising rivals’ costs, 234 refusal to deal, 144 usefulness implications, 131 predictive tests: competition law requiring foreseeability, 99 preferences: domestic consumers, 57 price and output: competitive level identification, 21 decisions, market demand response, v–vi price-takers: firms, 51 price wars, 10, 65, 126, 152, 227, 231, 232 Boral, 162, 164–67 prices: alone not indicating harm to competition, 197 below-cost pricing, 111, 165 borderline, 134 buyers competing to bid up, 51 changes: consumer responsiveness, 51 demand responsiveness to, 40 competitive: in context of monopoly, 54 level, 58 above, dominant firm ability to charge, 134

dominant firms conduct benchmark, 90 indeterminate, 229 in monopoly model, 57, 66 constructive refusal to supply, 134 consumers, see consumers correct, 44 cost-based analysis, 164, 165 cutting: defensive, 158 predatory, 209 predatory pricing losses recoupment from, 155 Rural Press, 175 decisions: demand response to, 198 dominant firms, see dominant firms effective, 200 effects, 93 efficiency and: Queensland Wire, 141 equilibrium, 57 consumer response, 62 elasticity of demand, 62 perfect competition, 45 tending towards producer break-even points, 45 excessive: dominant firms, 121 excessively high, 48 margin squeeze, 233 refusal to supply, 233 excessive pricing, 228 perceived, 135 prohibition, EU, 111 falling, see reductions below fixing: contravention, 146 margin squeeze, 232 refusal to supply, 232 holding down for consumers, 53 increases: above-normal profits, 50 by artificial restriction of output, 50 consumer buying reducing, 61 revenue effects, 61 zero-sum competition, 126 inputs, see inputs limit pricing, 165, 228 USA, 115–16 market demand, responding to, 49 market price volatility: prohibited manipulation, distinguishing between, 59

Index  289 monopoly, see monopoly net back: Queensland Wire, 145–46 above normal: dominant firms, 92 optimal, 44 perceived excessive pricing: Harper Review condemnation, 235 perfect competition: assumptions, 119 equilibrium, 45 predatory pricing, see predatory pricing profitability, Queensland Wire, 141 prohibited manipulation: normal market price volatility, distinguishing between, 59 prohibition of pricing above competitive level, 228 prohibitive: Queensland Wire, 140 reductions, 121 as competitive response to changed market conditions, 158 market power redressing by, 121 predatory pricing, 158 redressing market power, 121 revenue effects, 61 supply exceeding demand, 51 zero-sum competition, 125 rising, see increases above uncompetitive: Queensland Wire, 137 wars, see price wars Prisoner’s Dilemma, 64, 130 private autonomy, 71 private economic power: CCA 2010, section 46 and, 82 growth limitations, 82 private efficiency: conduct responding to market demand, 130 market demand responses, 226 Melway, 152 monopoly, 125 private inefficiency: output increases beyond profit-maximising position, 226 Queensland Wire, 141 private welfare: competitive level indicator, 120 standard, 25 pro-competitive rationale, Canada, 106 probability-based effects tests, 97

probable competition effects, 97 probable effects, 97 probative evidence, 149 procedural implications, 214–19 process of competition, see competition producers: break-even points: equilibrium prices tending towards, 45 community of interest, 46 competition between, 51 consumers and: equitable distribution of surplus between: competition law objectives, 46–47 interest as, 118 cost functions, 49 detriment to, 46 marginal costs, 58 markets excluding, 48 perfect competition assumptions, 119 profits, excessive, 48 surpluses, 120, 121 affirmative competition defences, 104 equitable distribution between consumers and, 46–47 under market equilibrium, 45 monopolies and, 54 welfare: redistribution to consumers, 57 see also suppliers production: maximisation, 16 from given resources, productive efficiency, 48–49 possibilities frontier: allocative efficiency, 118–19 dynamic efficiency, 118–19 productive efficiency, 118–19 socially inefficient as price to pay for increasing efficiency, 123 production efficiency: competition role, 81 legitimate business justification as, 81 productive efficiency, 2, 17 allocative efficiency and, 55–57 objectives inconsistency, 120 competition law objectives, 223 congruent optimisation, 124 equilibrium achieving, 49 as goal, 16 imperfect competition and, 83 as inconsistent objective, 49 increasing, 56 market equilibrium, consumers, 46

290  Index maximisation of production from given resources, 48–49 objectives, ACCC and ACT weighing up, 219 perfect competition, 49, 55 production possibilities frontier, 118–19 relative, 123–24 technological innovation, 56 United States, 16 see also dynamic efficiency Productivity Commission, 135 competition as imprecise proxy for efficiency, 100 products: differentiation: competition, 53 perfect competition assumptions, 119 profit-maximising: conduct, 9 monopoly, 125 rational, 133 output, see output profitability: prices and, Queensland Wire, 141 profits: above-normal, 53, 133 monopoly, Queensland Wire, 141 alone not indicating harm to competition, 197 dominant firms, see dominant firms economic: above-normal, 53 efficiency and: Queensland Wire, 141 excess, 57 Queensland Wire, 141 excessive, 48 firms, see firms foregone by incumbents: recoupment, 210 hypothetical competitive level, benchmarking against, 60 maximisation: behaviour, monopoly, 125 firms, 123 monopoly equilibrium, 56 output, 50, 56 increases beyond, 226 monopoly, see monopoly objective indicators of, firms, 123 perfect competition, see competition: perfect producers, excessive, 48 sacrifice: Baxter, 204

sufficiency: capital markets, 48 markets, 48 workable competition, 60 prohibited conduct, see conduct prohibited use of market power, 26 prohibitions: of attempts, 235 based on purpose and effects, Canada, 102 competition law, focussing on conduct, 99 no fault, 130 pricing above competitive level, 228 strict liability, 130 prohibitive prices, Queensland Wire, 140 proof, burden of, 217–19 proportionality: zero-sum competition, 73 proscribed conduct, see conduct prosecutorial roles: ACCC, 217 CCA 2010, 87 protection: competition process, 17 competitors, see competitors revenue, Rural Press, 171–72 proxy tests, 74 public benefits: ACCC and ACT authorisations: administrative powers, jurisdiction of courts and, dividing line becoming blurred, 103 harm to competition justification by offsetting, 99 issues, court decisions, 83 NT Power, 215 small business protection as, 85 public finance: economics of, distributional equity, 47 public interest, 81 government executive branch functions, 83 tests, 84–86 public policy: role in economic activity regulation, deeper understanding required, 4 public tenders: Baxter, 202–204, 209 Cement Australia, 209 bidding, 209 processes, bidding wars, 194–95 publicly owned enterprises: exposure to competition law, 53–54

Index  291 purpose: bundling, 201 Canada, see Canada conduct, proscribing by reference to effects and, 128 without effects, 96, 117 Pfizer case, 212 SLC tests, 235 USA, 102, 114, 116 effects and: integrally linked, 127 inter-linked and inextricable meaning, EU, 112 effects without: SLC tests, 235 EU, see European Union exclusionary, 205 CCA 2010, section 46 requirement, 81 Melway, 150 indicator of harm, 201 judicial interpretation, 96 no prohibition of without effects, EU, 102 problems with concept of, 117 prohibitions based on effects and, Canada, 102, 105 rational self-interested conduct and, 102 as screen isolating harmful effects, 127 SLC tests, 235 sole purpose tests, 235 as staging point towards determining ultimate effects, EU, 109 subjective, 92, 95, 102, 116 Baxter case, 212 Pfizer case, 212 Queensland Wire, 140 substantial, 96 purpose tests, 96 subjective: Baxter case, 212 Pfizer case, 212 as surrogate for efficiency, Canada, EU and USA, 127 tests, see purpose tests USA, see United States purpose tests, 90, 91 Boral, 163 competition process and, 103 competitor protection, 103 efficiency assessments and relevance to, 99 EU, 109 Pfizer case, 212 predatory pricing, 156

principles developed overseas application to, 109 process of competition and, 103 rational business decision-making norms, 128 subjective purpose, court interpretation, 92 substantial purposes, 96, 235 Qantas, 16, 41, 47 qualitative assessment: competition effects, 192 qualitative SLC tests: predatory pricing, 233 raising rivals’ costs, 234 Queensland Co-operative Milling Association, 39 Queensland Wire, 39, 44, 97–8, 185, 216, 233 factual matrix, 136–39 gaps in, 140–42 Harper section 46: 142–46 under market manipulation, 146–49 raising rivals’ costs, 178–79, 227 competition, effects on, 179 evidence-based economic analysis, 234 harm to dominant firm itself, 186, 193 market manipulation approach, 234 market power use restraints, 167 own costs increasing, 187–88, 194 predictive and qualitative SLC tests, 234 see also Cement Australia rational business decisions: competition, 8 dominant firms, 93 harmful effects, 91–92 making, norms, purpose tests, 128 tests, EU, 111 rational profit-maximising conduct, 133 rational self-interested conduct, 102 Readymix, Cement Australia case, 185, 188 real chance, 34, 97, 105, 164 Baxter, 209 Cement Australia, 184–85, 192 real world: competition, 63, 226, 230 exit barriers, 123 assumptions: zero-sum competition, 42 markets: actual position in benchmarks, 178 pursuit of allocative efficiency not achievable in, 58 see also worldviews

292  Index realised efficiency: anticompetitive effects more substantial than, 101 recommendations and findings, see findings and recommendations recoupment: Baxter, 204 Boral, 163, 165, 166, 210 bundling, tests, 210 entry barriers, 166 predatory pricing: losses from price cutting, 155 tests, 210 predictive tests, 162 profits foregone by incumbents, 210 SLC effects tests, application under, 167 redistributive equity, 85–86 redistributive objectives: competition law counter to, 47–48, 121 taxation, 47–48 Redmond, William, 51, 53 reduction of prices, see prices reform process: misuse of market power, 4, 5–7 refusal to deal, 133–54, 179, 227 refusal to supply, 83, 134, 136, 144–45, 147, 149, 153, 215, 232–33 regulation: market power, 18, 19 monopoly competitive process, 62 monopoly prices, 54, 57, 61–62 private autonomy limitation purpose, 71 regulators: worldviews, 64–65, 130 relationships: ongoing, supply, 134 relative allocative efficiency, 119, 121, 123–24 relative efficiency, see efficiency relative productive efficiency, 123–24 relevant markets: Queensland Wire, 137–38 reopening costs, 157–58 resources: allocation satisfying consumer preferences: allocative efficiency, 48 efficient use measures, Queensland Wire, 149 investment: capital markets, 48 markets, 48 of society, 54 restrictive trade practices, 81

returns: on assets: Queensland Wire, 149 diminishing, 108 on equity: financing assets measures, Queensland Wire, 149 revenue: average, 60 changes: supply responding to, 60 costs and: relationship between, 54 relationship, monopoly, 65–66 elasticity of demand, 61 firms, objective indicators of, 123 increase: zero-sum competition, 123 maintenance: zero-sum competition, 123 marginal, see marginal revenue objective indicators of, firms, 123 price increases, effects, 61 price reductions, effects, 61 protection: Rural Press, 171–72 Review, see Harper Review Rinker, Cement Australia case, 182, 188 rising prices, see prices: increases risks: voluntary assumption of, 72 rivals: conduct excluding, 58 costs, raising, see raising rivals’ costs excluding rivalry, 184, 185 exclusion: Competition and Consumer Act 2010, 48 giving chance to win, 209 harm to, 185 process of competition and, 187 margin squeeze, 185 scale efficiencies, prevention from achieving, 207 Robert Hicks Pty Ltd, see Melway Rocla, Boral case, 159–60, 162 Round, D, 63, 157–58, 165, 209 rule of law, 4 discretionary practical justice inconsistent with, 12 misuse of market power, jurisprudence on, concerns with, 12 polycentric law raising concerns, 219

Index  293 Rural Press, 169–70, 176–77 factual matrix, 170–72 gaps in, 172–74 under Harper section 46: 174–75, 176–77 under market manipulation, 175–76 Rural Press First Instance, 169, 173 sacrifice of profits, Baxter, 204 Sákovics, József, 194 sales: growth: zero-sum competition, 69, 70 increases, perfect competition, 70 taken from competitors: zero-sum competition, 123, 152 Salvadori, Neri, 51 saving costs, Baxter, 204 scale: new entrants lacking, 209 scale economies, see economies of scale scale efficiencies: rivals, prevention from achieving, 207 scarce goods, 46 Schaefer, S (Besanko et al), 201 Scheffman, David, 178–79 Schere, FM, 42 Schmidt, Hedvig, 19 Schroeder, Dirk, 86 Schumpeter, Joseph, 20–21, 43 Schumpeterian dynamic, competition, 31, 43 scope economies, see economies of scope SCP, see Structure-Conduct-Performance securities industry laws: efficient market in, 8 securities markets: manipulation laws, v–vi, 2–3, 59–61, 117, 125, 126, 226 self-correction: market power, 18 self-defence, 72–73 self-preservation: zero-sum competition, 156 self-selection: for entry, 48, 119 equally efficient firms, for entry or exit under zero-sum competition, 10 for exit, 48, 119 sellers: market manipulation powers, 230 of raw materials creating bidding wars, by calling for public tenders, market manipulation, 225

Senate Economics Legislation Committee, 6–7, 17, 127 services: perfect competition assumptions, 119 SF (sterile fluids), 202–208 Shanley, M (Besanko et al), 201 Shelanski, Howard, 80 Sherman Act, see United States shortages: unmet demand indication, 125 shut down: principle, perfect competition, 165–66 rule, 157–58 Signorino, Rodolfo, 51 Simon, Daniel, 70 Sims, Rod, 1 size of markets, see market size SLC, see substantially lessening competition SLC tests, 224 affirmative competition defences and, 104 application, market power irrelevance to, 23 Baxter case, 205–206, 209 Boral, 163, 164, 166–67 changes to, 103 conduct failing, 80 effects, 143, 235 Harper Review on, 143–46 without purpose, 235 recoupment application, 167 efficiency effects and, 130 EU, see European Union Harper Review, 151, 164, 224 market power causal connection, 82 Pfizer case, 212 predictive, see predictive SLC tests process of competition focus, 146 purpose without effect, 235 qualitative, see qualitative SLC tests Queensland Wire, 142, 143 Rural Press, 174–75 section 46 subject to, proposals, 95 structural analysis, 110 structuralist, see structuralist SLC tests weakening competitive process, 95 small business: protection as public benefit, 85 Smith, D, 218 Smith, R, 63, 157–58, 165, 209 Smorgon, 139

294  Index social benefits: perfect competition assumptions, 119 social costs: of causing exits, 69 perfect competition assumptions, 119 social desirability: efficiency as test of, 158 social efficiency, 2, 9–10, 11 conduct responding to market demand, 130 diminishing, 224 ‘efficiency’ understood to refer to, 224 enhancing, 224 legal determinacy, 124 market demand responses, 226 Melway, 152 monopoly, 125 optimum, 92 Queensland Wire, 148 social engineering: of commercial behaviour, 42 social inefficiency, 130 output increases beyond profit-maximising position, 226 Queensland Wire, 141 social objective, equity as, 121 social welfare, 24 competitive level indicator, 120 perfect competition, 56 socially inefficient production, 123 society, resources of, 54 sole purpose tests, 235 squeezing, 59 standards: judges’ normative role in applying, 83 legal liability, positive, 81 State government purchasing authorities (SPA), 202 statutes, court analysis, 23 steel industry, 136–49 Steel Industry Plan, 138 sterile fluids (SF), 202–208 strategic behaviour, 79 strategic market power, 39–40 market manipulation reconciliation, 51 strategy analysis: oligopoly competition, 63–64 strict liability, 92, 105, 235 Boral, 164 Cement Australia, 198 prohibitions, 130 Rural Press, 175 see also no fault liability

structural analysis, 90, 106, 110, 163 structural factors: effects tests, 98 structuralism: antitrust, USA, 114 approach, 1 economic, USA, 114 market power conception, 9 USA, 114 structuralist SLC tests, 124 Rural Press, 175, 176 structuralist thinking, 79 Structure-Conduct-Performance (SCP), 1–2, 20, 21–22, 78 causation chain, 79 CCA embedded, impeding economic efficiency, 82 competition effects, forensic determination implications, 80 discredited overseas, 90 efficiency effects, forensic determination implications, 80 embedded in CCA 2010, 82 market performance inferences, chain of causation providing, 79 market structure as primary source of market power, 79 mischief addressed by, 79 modified approach, ACCC and ACT, 79 USA, out of favour in, 79 vertical integration, 180 subjective purpose, 92, 95, 102, 116 substantial market power, see market power substantial purpose tests, 96, 235 substantial subjective intent, 113, 143 substantial subjective purpose, see purpose substantially lessening competition (SLC), 5, 14, 23, 24, 26, 29 Baxter case, 206, 208 Cement Australia, 184–85, 196–97 electricity market, 216 tests, see SLC tests substitution, marginal rate of (MRS), 55 sunk costs, 109, 157, 226 Sunstate, Cement Australia case, 189 Sunstein, Cass, 47, 71–72, 74 suppliers: competing: market power constrained by, 52 competition between, 53 customers and, as competitors, 146

Index  295 lacking market power, perfect competition, 50–51 see also producers supply: adapting to demand shifts, zero-sum competition, 157 constructive refusal, at prices, 134 curves: industry costs, 54 market, 45 under oligopoly, 45 perfect competition, 45 demand and, see supply and demand demand shifts, responding to, 231 exceeding demand: price reductions, 51 imperfect competition, 69, 122–23 marginal revenue and, 122–23 monopoly, curves under, 45 oligopoly: marginal revenue, response, 148 ongoing relationships, 134 reduced prices as competitive response to changed market conditions of, 158 refusal, see refusal to supply response: Cement Australia, 180 revenue changes, responding to, 60 supply and demand: adjusting to each other, 125 analysis, 46 demand exceeding supply, 51 dynamic conditions, 173 efficient markets, 117 interplay of forces of, 117 Rural Press, 173 supply exceeding demand, 51 surpluses: consumers, see consumers downstream markets: redistribution between upstream and downstream suppliers, 191–92 producers, see producers transfers: consumers, 227 inefficiencies, 166 from consumers, 179 away from dominant firms, 227 upstream, 194 surrogate competition test, 230 surrogate structuralist competition test, 163

surrogate tests, 74 assessing as relative efficiency, 124 market structure as, 80 relative efficiency as, 124 surrogate welfare test, 24 survival: competition, see competition: zero-sum fight for, 226 games, 63 strategies, efficiency of, 158 Swanbank, Cement Australia case, 183–84, 186, 188, 196 taking advantage: dominant firms of market power, 134 lacking, Cement Australia, 184 market power, 228 Melway, 150 Queensland Wire, 142 requirement, 23, 81, 104–105 Tarong Energy, Cement Australia case, 183, 184–96 taxation: economics of, distributional equity, 47 redistributive objectives, 47–48 technical efficiency: firms, 168 assessing, 126 technological innovation: productive efficiency, 56 technology changes, 48 tenders: Baxter, 202–204 competitive, see competitive tenders contested, see contested tenders multiple contemporaneous, 209 public, see public tenders terminated distributors, 151, 153 testing: of efficiency, 129–219 tests: competition, 10, 74 conduct, tests based on, USA, 89 dual purposes problem sole purpose tests, 235 dynamic forces underlying effects, 98 efficiency, see efficiency equally efficient competitors, EU, 110–11, 112 forensic, 11, 27–28 legal theory of liability proxy tests, 74 market conduct, court administration of, 80

296  Index market structure tests, courts, 80 objective reasonableness, USA, 114–15 permitted conduct test, AMC, 209–11 policy choices tested against certainty, 17–18 predictive, 99 predictive effects, 224 predictive nature of effects, 96–97, 98 proxy, 74 public interest, 84–86 rational business decision tests, 128 EU, 111 subjective intent test, USA, 114 subjective purpose, 113 surrogate tests: competition, 230 structuralist competition, 163 welfare, 24 see also effects tests; purpose tests; SLC tests TFEU (Treaty on the Functioning of the European Union), see European Union: TFEU theory of harm, see harm theory of liability, see liability thinking, structuralist, 79 threshold for market power, 26 toleration of inefficiency, 131 Toohey J, 137 Tooth and Tooheys, 97 tort law: breach of contract inducing, 74 foreseeability requirements, 98 voluntary assumption of risk, 72 total welfare, 16, 17, 102, 120 trade practices: restrictive, 81 Trade Practices Act 1965, 81, 82 Trade Practices Act 1974, 40 transaction cost efficiencies, 2, 180 transfers, naked, see naked transfers transformation, marginal rate of (MRT), 56 Treaty on the Functioning of the European Union (TFEU), see European Union: TFEU Trinko case, USA, 135 Turner, Donald, 79–80, 82 uncompetitive prices, Queensland Wire, 137 unconscionable conduct: statutory prohibition, 84 unicentric legal theory of liability, 84 unilateral conduct of dominant firms, see dominant firms

United States: Alcoa antitrust case, 178–79 American Bar Association (ABA), 113 anticompetitive effects, 115, 117, 218 antitrust: Alcoa case, 178–79 structuralism, 114 Antitrust Modernization Commission (AMC), 16, 41, 200–201, 209–11 business decisions, courts reluctance to second-guess, 115 common law method of judicial law-making, 87 competition: atomistic, 114 effects: negative, positive efficiency effects tests and, 128 elimination as sole motivation, 113 process: harm to, 115 protection, 129 survival, 115 zero-sum: dual purposes problem, 115–16 equally efficient new entrants, 115–16 incumbent firms, 115–16 competitors: harming, 115 protection, 129 conduct: elimination of competition as sole motivation, 113 exclusionary, 113, 114, 156 intent inferred from, 113, 114 tests based on, 89 consumers: surplus as indicator of efficiency effects, 128 welfare, 17, 102 courts: anticompetitive effects, weighing up against efficiency gains not contemplated, 115 atomistic competition, 78–79 business decisions, reluctance to second-guess, 115 commercial transactions, reluctance to supervise, 135 credible efficiency justifications negate antitrust harm view, 114 dominant firm conduct intervention scepticism, 58

Index  297 efficiency gains, weighing up anticompetitive effects against not contemplated, 115 intervention scepticism, 5 judicial decisions, 23 judicial law-making, common law method, 87 juries’ role, 114–15 Sherman Act promoting atomistic competition view, 114 specific intent, apparent rejection, 115 Department of Justice: dominant firm conduct report, 18 economic structuralism, 114 economics thinking since 1970s, 40 economies of scale, 115 effects: anticompetitive, see anticompetitive effects above efficiency, see efficiency below intent inferred from, 113 efficiency: effects, 117 consumer surplus as indicator of, 128 positive, negative competition effects tests and, 128 gains: as defence to competitive harm claims, 101 surrogate tests, 117 purpose as surrogate for, 127 recognition as defence, 80 energy prices, 59 equally efficient new entrants, 115–16 exclusionary conduct, 113, 114, 156 harm: competitors, 115 theory of, 208 hostile intent, evidence of, 114, 156 incumbent firms, 115–16 institutional arrangements court-centred, 87 intent: condemnation, 114 hostile, evidence of, 114, 156 inferred from conduct, 113, 114 inferred from effect, 113 objective, 113 specific, 113 court apparent rejection, 115 inferring, 116 to monopolise, 116 subjective, 113

juries’ role, 114–15 law: juries’ role, 114–15 thinking since 1970s, 40 legitimate business rationale: rebutting claims of monopolisation, 218 limit pricing, 115–16 Linkline case, 135 market definition, 114 arbitrariness, 114 market manipulation: monopolisation, 93 market power: market definition determinative of, 114 second-best way to assess, 114 monopolisation: American Bar Association submission to Harper Review, 218 attempted, 116 law, 4–5 development, 87 legitimate business rationale rebutting claims of, 218 manipulation of markets, 93 no fault regime rejected, 92 new entrants, equally efficient, 115–16 no fault regime rejected, 92 objective reasonableness test, 114–15 predation claims: objective reasonableness test, 114–15 predatory pricing, 116, 155–56, 209 prices: limit pricing, 115–16 predatory pricing, see predatory pricing above productive efficiency, 16 protection, competition, 129 purpose: Australian position regarding and, 113 without effect, 102, 114, 116 as surrogate for efficiency, 127 Sherman Act, 47, 80 atomistic competition promoting, 114 section 2: 50, 53, 61, 82, 93, 113–14 subjective intent test, 114 structuralism: antitrust, 114 economic, 114 Structure-Conduct-Performance out of favour in, 79 subjective intent, 113 survival competition, 115

298  Index total welfare, 102 Trinko case, 135 anticompetitive, 117 efficiency, 117 intent inferred from, 113 purpose without, 102, 114, 116 tests, ABA recommendation against adopting, 113 zero-sum competition, see competition above Universal Music, 13, 151 unmet demand, 69, 117, 125, 133, 226 Baxter, 210 bidding wars, 180, 185 Cement Australia, 180, 185, 195, 197–98 market size and, 125 Queensland Wire, 141, 143–44 Rural Press, 173 upstream markets: dominant firm profits forgone in, 180 downstream market surpluses: redistribution between upstream and downstream suppliers, 191–92 efficiency losses in, 100 market power in, 179 firms squeezing rivals’ margins in downstream markets, 185 profits forgone in, dominant firms, 180 upstream transfers of surpluses, 194 USA, see United States value judgements, 57 Varney, Christine, 18 vertical foreclosure: mergers, 179 vertical foreclosure, 179 harm from, 179–80 vertical integration, 179 as anticompetitive aspect of market structure, 180 backward, 179–80, 188 behaviour of firms, 194 bidding wars, 194 efficiency benefits, 180 mergers, 179

NT Power, 216 relevance, 195 vertical mergers, 179 ACCC opposition, 216 vertical supply arrangements: competitive effects of, competitive tenders reducing concerns about, 195 virtual monopoly, Queensland Wire, 143 volume-related benefits: bundling, 205 voluntary assumption of risk, 72 Waikerie Printing, Rural Press case, 171–76 Ware, R, 178 weakening competitive process: SLC tests, 95 welfare: community, 119, 121 consumers, see consumers private, see private welfare private standard, 25 producers, 57 social, see social welfare surrogate test, 24 total, see total welfare Whish, Richard, 111, 112–13 whole economy: economic activity in, 118 Williams, Philip, 90–91, 103, 134–35, 142, 152 win-lose strategies, 64, 65 win-win strategies, 64 workable competition, see competition worldviews: of competition, 102 differing: competition law, 18–19, 223 norms of conduct reflecting, 64 transcending, 130 economists, 64–65, 119, 130 lawyers, 64–65, 119, 130 regulators, 64–65, 130 see also real world zero-sum competition, see competition