Patents and Industry Standards

Patents and Industry Standards Patents and Industry Standards Jae Hun Park Korean Intellectual Property Office, Kore...

Author: Jae Hun Park

66 downloads 1362 Views 1MB Size Report

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

DOWNLOAD PDF

Patents and Industry Standards

Patents and Industry Standards

Jae Hun Park Korean Intellectual Property Office, Korea and formerly PhD Candidate for Law, University of Nottingham, UK

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Jae Hun Park 2010 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2009941261

ISBN 978 1 84980 011 2

02

Typeset by Cambrian Typesetters, Camberley, Surrey Printed and bound by MPG Books Group, UK

Contents Acknowledgement Foreword Prof. Dr Paul L.C. Torremans Table of statutes Table of cases

vii viii x xi

1

1 1 2 3 5 7 7 8 12 15 20 23 23 24 42 55 55 78 82 96 96 97 107 118 118 119 133 162

2

3

4

5

6

Introduction 1 Background 2 Research questions 3 Methodological approach and source materials 4 Limitations Industry standards 1 Introduction and definition of standards 2 Classification of standards by their economic effects 3 Standard setting process 4 Standardisation and technical change 5 Standards and competition Patents and standards in the US 1 The legal issues 2 Patents and informal standards 3 Patents and formal standards Patents and standards in the EU 1 Patents and informal standards 2 Patents and formal standards 3 Exceptional circumstances and industry standards Patent laws and standards 1 Introduction 2 Reverse doctrine of equivalents 3 Compulsory licences Costs and benefits of patent systems 1 Introduction 2 Costs and benefits of patent systems 3 Costs of patent systems in relation to standards 4 Review and implications

v

vi

7

8

Patents and industry standards

Alternative solutions and a suggested solution 1 Introduction 2 Proposals for patent reform 3 Suggested solution 4 Implementation of the proposed model Conclusion

References Index

165 165 167 200 214 216 219 231

Acknowledgement The views portrayed in this book are those of the author and do not reflect the beliefs of the Korean Government that Jae Hun Park currently works for.

vii

Foreword Patent law is subject to a lot of scrutiny, but most of it concentrates on its internal workings and thresholds. Questions focus on the level of inventive step and the scope of patentable subject matter, to name just these. From an external point of view there is the issue of competition law and of course there is the development agenda. But that is often where matters stop. The issue of standardisation and the international system of industry standards is left completely out of the equation. Standards are often simply considered to be a technical issue for standardisation bodies as part of the administrative regime that is not considered to be part of intellectual property law. In fact there is very little proper legal analysis of the whole system of standardisation. But one should not forget that the newest and most valuable standards and the items that are taken into account as candidates for standardisation are often protected by one or more patents. Whether one likes it or not, the two areas do touch each other. For the purposes of standardisation a standard pre-supposes access to technology for all the actors of a certain industry. As an exclusive right a patent makes that access conditional on the consent of the right holder, who has the right to refuse certain licences for certain potential licensees and who can set the terms of the licence. Exclusivity and standards simply do not go together very well. A decision by a standardisation body cannot invalidate or overrule the patent though. There is no mechanism in patent law to require the right holder to renounce the patent and the exclusive right that goes with it. Somehow a modus vivendi will have to be worked out. Very real issues arise therefore and very little attention has been paid to them in terms of legal scholarship and analysis. Dr Jae Park is to be congratulated for turning our attention to this difficult and underexplored area. His work focuses on standards and patents but goes well beyond an initial analysis. He examines the finer points of both sets of rules in order to find out exactly where the problem lies and he then looks at the existing mechanisms that could provide a solution. Many of these have their roots in the area of competition law, but his thorough analysis shows that competition law in its current form and with its current limitations is not the perfect tool to address the problems that arise when patented technology becomes the object of standardisation. This leads Dr Park to develop his own viii

Foreword

ix

solution for the problem at hand: a solution which he finds in the dynamic liability rules regime. This book really breaks new ground and provides a first and thorough analysis of this rarely addressed but increasingly important area. Prof. Dr Paul L.C. Torremans Professor of Intellectual Property Law School of Law, University of Nottingham

Table of statutes International

Commission Notice on the Definition of the Relevant Market for the Purpose of Community Competition Law, C372 Official Journal 5 (1997) 56, 57 EU, Council Directive 91/250/EEC of 14 May 1991 on the legal protection of computer programs, Official Journal L122 (1991): 42 6

Agreement on Trade-Related Aspects of Intellectual Property Rights, 1994 76, 88–90, 109, 112–14, 214, 215, 218 Paris Convention for the Protection of Industrial Property, 1883 89, 108, 113, 114 Europe Commission Decision No. 2002/165/ EC, L 59 Official Journal 18 (2002) 67, 68

x

Table of cases Europe

Case C-7/97 Oscar Bronner GmbH Co KG v. Mediaprint Zeitungsund Zeitschriftenverlag GmbH & Co. KG, 4 C.M.L.R. 112 (1998) 65–67, 74, 76, 77 Case C-62/86 AKZO Chemie BV v. EC Commission, E.C.R. I-3359 (1991) 58 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., 4 C.M.L.R. 28 (2004) 57, 59, 64, 67–69, 82–84, 86, 88, 90 Case COMP/C-3/37.792 Microsoft, 32 Official Journal 23 (2004) 70 Case T51/89 Tetra Pak Rausing SA v. EC Commission, II European Court Reports 00309 (1990) 77, 78 Case T184/01 R II IMS Health Inc. v. EC Commission, 4 C.M.L.R. 2 (2002) 67–69 Cases 6–7/73 Istituto Chemioterapico Italiano SpA and Commercial Solvents Corp. v. EC Commission, 1 C.M.L.R. 309 (1974) 61 Derry v. Peek, 14 LR 337 (1889) 52 Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission, 4 C.M.L.R. 718 (1995) 56, 60, 63–64, 77, 94, 209

Case 22/78 Hugin Kassaregister AB and Hugin Cash Registers Limited v. EC Commission, 3 C.M.L.R. 345 (1979) 58, 87 Case 24/67 Parke, Davis & Company v. Probel and Others, C.M.L.R. 47 (1968) 60 Case 27/76 United Brands Company and United Brands Continentaal BV v. EC Commission, 1 C.M.L.R. 429 (1978) 56–58 Case 53/87 Consorzio Italiano della Componentistica di Ricambio per Autoveicoli and Maxicar v. Regie Nationale des Usines Renault, 4 C.M.L.R. 265 (1990) 62, 63 Case 85/76 Hoffmann La Roche & Co. AG v. EC Commission, 3 C.M.L.R. 211 (1979) 56, 58 Case 102/77 Hoffmann–La Roche & Co. AG and Hoffmann–La Roche AG v. Centrafarm Vertriebsgesellschaft Pharmazeutischer Erzeugnisse mbH, 3 C.M.L.R. 217 (1978) 60 Case 238/87 Volvo AB v. Erik Veng (UK) Ltd, 4 C.M.L.R. 122 (1989) 60, 62, 96 Case 322/81 Nederlandsche BandenIndustrie Michelin N.V. v. EC Commission, 1 C.M.L.R. 282 (1983) 59 xi

xii


Kirin-Amgen Inc. v. Hoechst Marion Roussel Ltd., R.P.C. 9 (2005) 98 T-201/04 Microsoft Corp. v. EC Commission, 5 C.M.L.R. 11 (2007) 69–74, 76, 77, 81 Tierce Ladbroke SA v. EC Commission ECR-II 923 (1997) 64, 65 USA A.C. Aukerman Company v. R.L. Chaides Construction Co., 960 F.2d 1020 (1992) 47 Alaska Airlines Inc. v. United Airlines Inc., 948 F.2d 536 (1991) 35, 37, 39, 40 Aspen Skiing Co. v. Aspen Highlands Skiing Co., 472 U.S. 585 (1985) 6, 28–33, 151, 155, 160 Atari Game Corp. and Tengen, Inc., v. Nintendo of America Inc. and Nintendo Co., Ltd., 975 F.2d 832 (1992) 83 Berkey Photo Inc. v. Eastman Kodak Co., 603 F.2d 263 (1979) 39, 40 Bonjorno v. Kaiser Aluminum & Chemical Corp., 752 F.2d 802 (1984) 153–154 Boyden Power-Brake Co. et al. v. Westinghouse et al., 170 U.S. 537 (1898) 104 Data General Corp. v. Grumman Systems Support Corp., 36 F.3d 1147 (1994) 29, 30, 36 Data General Corp. v. Grumman Systems Support Corp., 761 F.Supp. 185 (1991) 35 Festo Corp. v. Shoketsu Kinzobu Kogyo Kabuskiki Co., Ltd, et al., 535 U.S. 722 98, 101–103

Graver Tank & MFG. Co., Inc., et al. v. Linde Air Products Co., 339 U.S. 605 (1950) 98, 99, 104 Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195 (1997) 29–32 In re Dell, No. 931-0097, Lexis 466 (1995) 44, 50 In re Independent Service Organizations Antitrust Litigation, 203 F. 3d. 1322 (2000) 27 Intergraph Corp. v. Intel Corp., 195 F.3d 1346 (1999) 36 Intergraph Corp. v. Intel Corp. (Reply Brief of DefendantAppellant Intel Corporation), West Law 1 (1998) 36 Lotus Development Corporation v. Borland International Inc., 49 F3.d 807 (1995) 191, 192 MCI Communications Corp. v. AT&T, 708 F.2d 1081 (1983) 35, 37 Miller Insituform, Inc. v. Insituform of North America, F. 2d 606 (1987) 26, 27 SCM Corp. v. Xerox Corp., 645 F.2d 1195 (1981) 25, 26 Sega Enterprise Ltd. v. Accolade Inc., 977 F.2d 1510 (1992) 183, 184, 185 Sony Computer Entertainment, Inc. v. Connectix Corporation, 203 F3.d 596 (2000) 185 Stambler v. Diebold, 11 U.S.P.Q.2d 1709 (1988) 48 United States v. Microsoft Corp., 253 F.3d 34 (2001) 33 Verizon Communications, Inc. v. Trinko LLP, 540 U.S. 398 (2004) 33, 38, 40

Table of cases

Wang Labs., Inc. v. Mitsubishi Elecs. Am., Inc., 103 F.3d 1571 (1997) 49

xiii

Warner-Jenkinson Company, Inc., v. Hilton Davis Chemical Co., 520 U.S. 17 (1997) 98–101

1. Introduction 1.

BACKGROUND

This book aims to resolve the issues arising from the clash between two conflicting interests. More precisely, this book seeks to reconcile the conflict between compatibility standards and patents and to find the right balance between them. In network markets,1 consumers purchase not a single product but a system which is composed of several compatible products.2 For instance, when buying a computer, a consumer buys not a single product computer but a computer system consisting of hardware, such as a main board, a keyboard, a mouse, a monitor and speakers, and software, such as operating systems and application software. It is easily verified that the various components of a computer system are often produced by different producers. Since all of the elements of a computer system, which are produced by distinct firms, have to be combined to compose a computer system, they have to be compatible. Thus, compatibility between products is essential in these markets, and this need for compatibility inevitably leads the markets to implement technical standards. Standards3 often include technologies that are covered by patents because standards are likely to be based on advanced technologies rather than obvious ones and because patents are granted to novel and inventive technologies.4

1 ‘Network markets’ has a broad meaning, including information technology markets, communication markets, railway system markets and credit card markets. However, network markets in this book are the markets of information and communication technology (ICT), where ICT is a ‘term that includes any communication device or application, encompassing: radio, television, cellular phones, computer and network hardware and software, satellite systems and so on, as well as the various services and applications associated with them, such as videoconferencing and distance learning’, as is defined in SearchSMB.com TechTarget (cited 5 April 2007); available from http://searchsmb.techtarget.com/sDefinition/0,290660,sid44_gci928405,00.html. 2 Oz Shy, The Economics of Network Industries (Cambridge University Press, 2001) 1–5. 3 As is described in Chapter 2, there are several kinds of standards. It has to be noted that the main concern here is the compatibility standards. 4 Janice M. Mueller, ‘Patenting Industry Standards’, John Marshall Law Review 34 (2001).

1

2


Indeed, where a standard does not incorporate advanced technologies, the standard is not easily accepted in the market. Where some parts of standards are protected by patents, conflict between patents and standards arises because the standards have a public characteristic whereas patents are in the private sector. That is, standards need to be widely used just like public goods, whereas patented technologies are private property, normally created by private parties’ inventive efforts. Thus, the need for widespread use of standards clashes with the exclusivity of private property. A patent is a state-created right whose owner is entitled to exclude others from using the inventions covered by the patent.5 Thus, patent owners may lawfully exercise their exclusive rights to prevent others from using the patented technologies. On the other hand, standards need to be open to those who want to provide products compatible with other products in the market. Without access to the technical standards, a company cannot enter the market because it cannot make its products compatible with other products and because consumers in the market are reluctant to buy products that are incompatible with other products. Where the holders of the patents, the subject matters of which are the technologies incorporated into compatibility standards, utilise the patent rights to exclude competitors, the competition and the innovation in the market are likely to be seriously distorted. While patents, which give exclusive rights to inventors, are essential to promote technological innovations, standards, the precondition of which is open access, are also important for technical development because they provide interoperability and network benefits. Likewise, the issues on patent rights arise from the need for compatibility in standard-based industries6 and it is these issues that this book seeks to resolve.

2.

RESEARCH QUESTIONS

It is clear from the previous discussion that the main research problem is how to strike the right balance between the rights of patent owners and the need for industry standards. In order to find the solution to this question, however, some more questions have to be answered before and/or after the first question is answered. First of all, a standard has to be defined and its characteristics need to be studied. Standards are classified in various ways and thus the standards at issue have to be distinguished. In addition, why and how patent rights 5 Lionel Bently and Brad Sherman, Intellectual Property Law, 2nd edn (Oxford: Oxford University Press, 2004). 6 Eric James Iversen, Esten Oversjoen and Haakon Thue Lie, ‘Standardization, Innovation and IPR’, Telektronikk 2 (2004).

Introduction

3

conflict with standards in some industries have to be examined. As becomes clear later in this book, the conflict arises in some industries but not in other industries and thus the mechanism and causes of the conflict have to be studied. Moreover, how the current legal systems resolve the conflict has also to be investigated. It is clear that current legal systems have their own solutions to the problems and thus, in order to find better solutions, the current solutions have to be evaluated. To sum up, the research questions are as follows: • What are the definition and the characteristics of industry standards? • Why and how do patent rights conflict with industry standards? • What solutions are provided by current legal systems and what can be improved? • How are the rights of patent owners and the need for industry standards to be balanced? • What legal measures are necessary to strike the right balance between patent rights and industry standards?

3.

METHODOLOGICAL APPROACH AND SOURCE MATERIALS

The main methodological approach of this book is the law and economics approach.7 The reason why this book takes this approach is clear. It is generally accepted that patent systems are justified for economic reasons,8 and thus whether patent systems work properly and whether they can be improved needs to be evaluated from an economic perspective. Since patent systems seek to promote innovation, where current patent systems can be changed to promote dynamic efficiency they have to be reformed. This is what this book is about. To evaluate current patent systems, a cost and benefit analysis is used here. As is described in Chapters 6 and 7, this book seeks to find a patent reform which reduces the costs of patent systems in relation to standards but does not decrease the benefits of patent systems. The detailed approach and structure are as follows.

7 For the law and economics approach, see Richard A. Posner, Economic Analysis of Law, 7th edn (Wolters Kluwer, 2007). 8 Alan S. Gutterman, Innovation and Competition Policy: A Comparative Study of the Regulation of Patent Licensing and Collaborative Research and Development in the United States and the European Community (Kluwer Law International, 1997) 129–30, Paul L.C. Torremans, Holyoak and Torremans: Intellectual Property Law, 4th edn (Oxford: Oxford University Press, 2005) 11–25.

4


Above all, standards have to be studied. Standards have to be defined and the features and classification of standards need to be examined. Standards may be implemented by market mechanisms or by standard setting organisations (SSOs). The standard created by market process is called an informal standard and that promulgated by an SSO is called a formal standard. The process and legal features of formal standards are different from those of informal standards and thus formal standards and informal standards are studied separately. Standards are closely related to market competition and innovation and thus the economic effects of standards and how standards affect technical innovation need to be studied. All of this is done in Chapter 2. Chapters 3 and 4 examine how the current legal systems of the United States and Europe can be applied to resolve the conflict between standards and patents. As is clear from those chapters, antitrust law in the US and competition law in Europe may be used to resolve the patent issues in standards. Thus, the relevant legal doctrines of competition law and antitrust law are examined with respect to the patent issues in standards to ascertain to what extent they resolve the conflict. In applying the legal principles of competition law and antitrust law, formal standards are studied separately from informal standards since legal issues of formal standards are different from those of informal standards. In Chapter 5, two patent law principles are examined to see whether they can resolve the issues. The reverse doctrine of equivalents and compulsory licences are used to restrict the exclusive right of patent holders and these principles may be applied to the issues of this research. The reverse doctrine of equivalents is used to limit the enforcement of patent rights in those cases where an alleged product is far different from the patented product, even though the product literally infringes the patent at issue.9 A compulsory licence is ‘an involuntary contract between a willing buyer and an unwilling seller imposed and enforced by the state’,10 which clearly restricts the exercise of exclusive rights granted by patent law. In Chapter 5, the two patent law doctrines, the way in which they are used in relation to standards and the limitations of them are examined. In Chapters 6 and 7, a cost and benefit analysis of patent systems in relation to standards is conducted. The approach of these two chapters is to find a

9 William S. Galliani, ‘Patent Infringement Amidst Rapidly Evolving Technologies: New Equivalents, the Doctrine of Equivalents and the Reverse Doctrine of Equivalents’, Santa Clara Computer and High Technology Law Journal 6 (1990): 86. 10 Gianna Julian-Arnold, ‘International Compulsory Licensing: The Rationales and the Reality’, IDEA: The Journal of Law and Technology 33 (1993): 349.

Introduction

5

solution which reduces the costs of patent systems in relation to standards but does not decrease the benefits of patent systems. To do so, the cost structure of patent systems with respect to standards is analysed and a way in which the costs can be reduced is suggested in Chapter 6. In Chapter 7, patent reforms proposed by others are studied and evaluated with respect to the costs and benefits of patent systems in relation to standards. Then, based on the strengths and weaknesses of other patent reforms, the suggestions of this research are described. The solution proposed in this book is examined with respect to the costs and benefits of patent systems and then it is argued that the suggested solution decreases the costs of patent systems but does not reduce the benefits of patent systems. As is clear from the aforementioned description, the research of this book is conducted on the basis of material from Europe and the US. There are several reasons why Europe and the US are the main focus of this book. Firstly, as far as intellectual property law is concerned, Europe and the US are two of the most influential areas in the world. Indeed, Europe and the US along with Japan are the three main players in the world intellectual property system. They are leading the development of the international framework of intellectual property law and thus, in order to suggest patent reform, it is essential to study these two giants. Secondly, both Europe and the US have developed competition law, which is closely related to this study. In many respects, the patent issues of standards are relevant to both patent law and competition law and thus the well established legal tradition of patent law and competition law is essential. Europe and the US are the jurisdictions where patent law and competition law have been developed and where there are cases which show the conflict between patent law and competition law. The third reason is a practical one. It is impossible to study all the material from all over the world. It is inevitable to select some jurisdictions and focus the research on them. Nevertheless, the approach and the outcome of this research may be used to evaluate other jurisdictions and to suggest patent reforms.

4.

LIMITATIONS

This book focuses on the patent issues in standards, which implies several limitations. Firstly, mainly patents are examined and other types of intellectual property like copyright, trademarks and trade secrets are not considered. Intellectual property in general may be related to the compatibility issues and in particular copyright may be relevant to software compatibility since software has

6


been protected by copyright.11 However, copyright and other intellectual property are not considered here since the logic and the legal principles of them are different from those of patents. Admittedly, there are some similarities between patent law and copyright law, but the protection scheme and legal doctrines of copyright law are different from those of patent law. Since this book conducts a detailed analysis of patent law, copyright law is beyond the scope of the study. Secondly, this book is concerned with standards and thus its analysis may not be applicable to those industries where standards are not essential. The analysis of this book applies to network industries, such as the communication industry and the internet industry, whose features inevitably lead to industry standards. However, in non-network industries, such as the chemical industry and mechanical industry, where industry standards are not essential, the analysis of this book may not apply. Thirdly, competition law and antitrust law are studied only to the extent that they are relevant to patent issues, since the main concern of this book is to propose a patent law reform. Pure competition law issues like cartels, price fixing and predatory pricing are not examined. As far as standardisation is concerned, antitrust issues arise in both formal and informal standardisation but many of the issues are beyond the scope of this book and thus antitrust issues are only considered to the extent that they are relevant to the patent law issues.12 Lastly, it is clear that the source material is limited since the research of this book is conducted on the basis of material from Western Europe and the US. However, the limitation of source material does not restrict the application of the approach and analysis of this book to other jurisdictions, because the proposed solution is obtained by analysing the legal principles and economic characteristics of patents in general. Thus, it would not be correct to argue that the suggestions of this book only reflect the perspectives of the developed countries and that they do not apply to the developing countries. Rather, the approach, analysis and outcome of this book may be used in evaluating and designing the patent law regime in any country, with some modifications.

11 For compatibility issues in relation to copyright protection of computer programs, see EU, ‘Council Directive 91/250/EEC of 14 May 1991 on the legal protection of computer programs’, Official Journal L122 (1991). 12 For the antitrust issues of informal standards, see Herbert Hovenkamp, Mark D. Janis and Mark A. Lemley, IP and Antitrust, vol. 1 (New York: Aspen Law & Business, 2002) Chapters 10–13. For the antitrust issues of formal standards, see Herbert Hovenkamp, Mark D. Janis and Mark A. Lemley, IP and Antitrust, vol. 2 (New York: Aspen Law & Business, 2005) Chapter 35.

2. Industry standards 1.

INTRODUCTION AND DEFINITION OF STANDARDS

In resolving the conflict between patents and industry standards, it is essential to understand the characteristics of standards. That is, the general knowledge about standards, such as the definition, the classification and the economics of standards, has to be studied before analysing the relationship between patents and standards. Indeed, the definition and the classification of standards determine the scope of this book. Therefore, in this chapter, the characteristics of standards are to be studied as follows. Firstly, the definition and the classification of standards are studied and then the relevant standards are identified. On the basis of the definition and classification, the process of standardisation and the dynamics of standards are examined. Lastly, how standards affect technical innovation and market competition is studied. First of all, how to define standards? Various definitions of standards are suggested and are being used. ISO (the International Organization for Standardization) and IEC (the International Electrotechnical Commission) define standards as ‘a document, established by consensus and approved by a recognised body, that provides, for common and repeated use, rules, guidelines or characteristics for activities or their results, aimed at the achievement of the optimum degree of order in a given context’.1 However, this definition seems to be inappropriate for the purpose of this book, since it restricts standards to the form of documents and does not include informal standards, which are developed by market process. Some commentators do not accept the definition provided by standard setting organisations (SSOs) and suggest different definitions. For instance, de Vries defines standards as approved specification of a limited set of solutions to actual or potential matching problems, prepared for the benefits of the party or parties involved, balancing their need, intended and expected to be used repeatedly or continuously, during a certain period, by a substantial number of the parties for whom they are meant.2

1 Henk J. de Vries, Standardization: A Business Approach to the Role of National Standardization Organizations (Kluwer Academic Publishers, 1999). 2 Ibid.

7

8


However, this definition is not satisfactory either, because informal standards are not ‘approved’ but are ‘adopted’ by economic units in markets and some standards, such as proprietary standards, do not always balance the needs of the parties involved. David and Greenstein define standards as ‘a set of technical specifications adhered to by a producer, either tacitly or as a result of a formal agreement’.3 This definition seems acceptable because it includes both formal standards, which are implemented by SSOs, and informal standards, which are generated by market process. In addition, it does not restrict standards to any specific form. Thus, for the purpose of this book, this definition is used.

2.

CLASSIFICATION OF STANDARDS BY THEIR ECONOMIC EFFECTS

2.1

Introduction

There are several ways to classify standards. For instance, where standards are categorised on the basis of the standard setting process, they can be divided into de facto standards and de jure standards.4 Standards can also be classified into product standards and non-product standards, depending on the relationship of standards to product.5 Of the various ways to categorise standards, dividing them by their economic effects is a useful way.6 Where standards are divided according to their economic effects, they can be categorised into compatibility and interface standards, minimum quality and safety standards, variety reduction standards, and information standards. However, it is likely that a standard falls into more than one category since a standard often performs multiple functions. 2.2

Compatibility and Interface Standards

Compatibility and interface standards define technical features required to

3 Paul A. David and Shane Greenstein, ‘The Economics of Compatibility Standards: An Introduction to Recent Research’, Economics of Innovation and New Technology 1 (1990): 4. 4 Ibid. 5 Gregory Tassey, ‘Standardization in Technology-based Markets’, Research Policy 29 (2000): 591–4. 6 G.M. Peter Swann, The Economics of Standardization: Final Report for Standards and Technical Regulations Directorate (Department of Trade and Industry, 2000) 4.

Industry standards

9

facilitate interfaces and interaction between different products. As the information and communication technologies have developed rapidly over the last few decades, compatibility and interface standards have become very important, since connecting to and communicating with other products are essential in information and communication industries. These standards are also important for the innovation of systems, which are composed of a number of different components, because they enable component level innovation.7 These standards have two particular economic features, network effects or network externalities and switching costs.8 Network effects, which work on the demand side, occur when the value derived from the consumption of products is affected by the number of other people who use the products or compatible products.9 There are two kinds of network effects: direct network effects and indirect network effects.10 Direct network effects arise in those circumstances where the value created by being connected to a network increases directly as the number of other members grows. For instance, the value of being connected to a telephone network increases in a direct way as the number of participants of the network grows, because each additional participant of the network creates positive network effects for all other participants. On the other hand, indirect network effects can be illustrated by the hardware–software relationship.11 When making a decision to purchase a personal computer, consumers are likely to be concerned with the number of others using similar or compatible hardware because the amount and variety of software supplied for use with a given computer will increase as the amount of hardware that has been sold increases. Consumers may not care about the size of the network itself but they will consider the benefit of compatible networks. After producers and consumers have decided to use one system or standard rather than another, they invest in the system or standard. The longer they use the chosen system or standard, the higher the cost they should pay to switch to another system or standard. The cost is called the switching cost.12 It is likely that network effects and switching costs cause markets to be locked into a particular standard because producers and consumers are reluctant 7 8

Tassey, ‘Standardization in Technology-based Markets’, 590. Kunt Blind, The Economics of Standards: Theory, Evidence, Policy (Cheltenham: Edward Elgar Publishing, 2004) 15. 9 Mark A. Lemley and David McGowan, ‘Legal Implications of Network Economic Effects’, California Law Review 86 (1998): 483–4. 10 Carl Shapiro and Michael L. Katz, ‘Network Externalities, Competition and Compatibility’, The American Economic Review 75, no. 3 (1985): 424. 11 Ibid. 12 Blind, The Economics of Standards: Theory, Evidence, Policy 15.

10


to switch to other standards even though other standards are superior to the standard they are using. It has to be noted that the technology adopted as a standard is not necessarily the optimal technology from the perspective of technological performance.13 The technology owner who has successfully built a wide network of followers and complementary products is often the winner of the standards war. Since the standards formed by network effects in markets are not formal standards but represent proprietary technologies that reach a dominant position in the market, they are called informal standards or de facto standards.14 2.3

Minimum Quality and Safety Standards

Where the product variety increases and consumers are not sufficiently informed about the products, information asymmetries arise.15 The information asymmetries between sellers and buyers could cause severe market failure because if buyers cannot distinguish high quality products from low quality ones, it is hard for a high quality seller to recoup the additional resources necessary to produce high quality products. Eventually, high quality sellers may not be able to survive such that bad quality sellers drive out good quality sellers, which is called Gresham’s law.16 To some extent, minimum quality and safety standards can reduce the market failure resulting from customers’ adverse selection.17 Where standards on quality exist and are well informed, the buyers can confidently distinguish high quality products from low quality ones before purchasing, and thus high quality sellers can sustain a higher price. In addition, minimum quality standards may also reduce transaction costs and search costs by reducing the risk to buyers, making it easier for buyers to evaluate the products. Likewise, minimum quality and safety standards can play an important role in providing product information to consumers, correcting market failure. 2.4

Variety Reduction Standards

Variety reduction standards limit products to a certain range or number of 13 Paul A. David, ‘Clio and the Economics of QWERTY’, The American Economic Review 75, no. 2 (1985). 14 Blind, The Economics of Standards: Theory, Evidence, Policy 17. 15 Swann, The Economics of Standardization: Final Report for Standards and Technical Regulations Directorate 6. 16 For Gresham’s Law, see Steven R. Smith, ‘Gresham’s Law in Legal Education’, Journal of Contemporary Legal Issues 17 (2008). 17 Swann, The Economics of Standardization: Final Report for Standards and Technical Regulations Directorate 6.

Industry standards

11

characteristics and perform two functions.18 Firstly, they lead to the exploitation of economies of scale by reducing the number of variations of products. It is possible to lower costs per unit by focusing on one standardised model since mass sourcing of the input materials, mass production and mass distribution are possible. However, variety reduction may cause adoption costs for a user because a user’s most preferred specification may be different from the supplied specification. Thus, there should be a trade-off between choice and price. Secondly, there is another important role of variety reduction standards. Variety reduction standards can reduce the suppliers’ risks.19 In general, standards determine the trajectories of future technological development, and these help the emergence and growth of new markets. By reducing the technical trajectories, variety reduction standards can help a market to take off, enabling the pioneering companies to focus on a limited number of new technologies in the early stages of markets. 2.5

Information and Measurement Standards

Information standards play a role in providing assurance that a product is what it is supposed to be, reducing the risks faced by producers and consumers and the transaction costs.20 These standards provide technical information and test methods for measuring product attributes.21 Measurement methods are also crucial to research and development (R&D) because standardised scientific and engineering data and standardised equipment calibration techniques are important to conduct efficient R&D.22 However, it is argued that it is sufficient to treat standards of information and product description as a combination of the other three categories of standards.23 2.6

Further Distinctions

According to the categorisation of David and Greenstein, there are four categories of standards, depending on their authorships.24

18 19 20 21 22 23

Ibid., 7. Ibid. Blind, The Economics of Standards: Theory, Evidence, Policy 21–2. Tassey, ‘Standardization in Technology-based Markets’, 589. Ibid. Swann, The Economics of Standardization: Final Report for Standards and Technical Regulations Directorate 7. 24 David and Greenstein, ‘The Economics of Compatibility Standards: An Introduction to Recent Research’, 4.

12


Firstly, there are unsponsored standards, which are a set of specifications in a well documented form for which no originator is identifiable. The appearance of these standards is due to market processes accompanied by increasing returns to scale and positive feedback. The keyboard arrangement QWERTY is an example of such a standard. Secondly, when there are individuals, firms or groups that own proprietary rights to a set of technical specifications incorporated in standards, the standards are called sponsored standards. Intel’s microprocessors and Microsoft’s personal computer operating systems are typical examples. The owners of standard technologies can strategically use the standard technologies to maximise their profits, which may cause suboptimal standards from the economic perspective. Thirdly, while market processes can create de facto standards, standards can also be formulated by voluntary standard setting organisations. Private standardisation organisations like the British Standards Institution (BSI) arbitrate the standard setting process to enable interested groups to arrive at a standard agreement. Fourthly, governmental regulatory bodies can promulgate mandatory standards, which come under state regulation. Where there is a strong need for standardisation but little incentive for private firms to standardise, government needs to develop technical standards. Thus standards can be classified in various ways. Amongst them, this research is concerned with compatibility standards since patents covering compatibility standards are most likely to give rise to policy concerns. That is, the conflict between patents and standards is most considerable in those cases where patents cover compatibility standards. Thus, hereinafter standards refer to compatibility standards unless there is additional explanation.

3.

STANDARD SETTING PROCESS

3.1

Formal Standardisation

3.1.1

The system of formal standardisation in Europe

The institutions There are three European standardisation institutions: the European Committee for Standardization (CEN), the European Committee for Electrotechnical Standardization (CENELEC) and the European Telecommunications Standards Institute (ETSI).25 CEN, CENELEC and ETSI 25

Isabelle Liotard and Rudi Bekkers, ‘European Standards for Mobile

Industry standards

13

implement European standards (EN), European harmonisation documents (HD) and European pre-standards (ENV).26 As far as European standards are concerned, the national member organisations must adopt the exact wording of European standards. As for harmonisation documents, the national member organisations have to implement national standards corresponding to the technical content of HD. European pre-standards are to be applied to areas with a high degree of innovation.27 CEN and CENELEC are the European standardisation institutions comparable to ISO (the International Organization for Standardization) and IEC (the International Electrotechnical Commission), respectively.28 Only the official national standards institutes of the EU and EFTA countries can be members of CEN. Similarly, CENELEC’s membership is limited to electrotechnical committees of the EU and EFTA countries. CEN produces standards for a whole range of industrial and domestic products, whereas CENELEC sets standards for electrical and electronic equipment used in industries and in the home. The membership of ETSI is less restrictive than that of CEN or CENELEC.29 A variety of parties from EU, EFTA and other European countries are members of ETSI. For instance, national administrations, public telecommunications operators, manufacturers, private service providers, research bodies and users are members of ETSI. An individual or a group can be a member of ETSI through a trade association, and participating in ETSI as an observer is also possible. The formal standardisation process30 The application for a new standardisation can be submitted by members, committees of SSOs or the European Commission. Once submitted, the new proposal is examined by members of the technical board, who decide whether or not to adopt the proposal. The technical committee manages the standardisation work, which is undertaken by individual national standardisation institutions, from drafting standards to the final consensus. After a discussion in the technical committee, a vote takes places among members to decide whether or not to accept the standards. Here, the voting is on a weighted national basis.

Communications: The Tense Relationship between Standards and Intellectual Property Rights’, European Intellectual Property Review 21 (3) (1999). 26 Blind, The Economics of Standards: Theory, Evidence, Policy 79–80. 27 Ibid. 28 Diana Good, ‘1992 and Product Standards: A Conflict with Intellectual Property Rights’, European Intellectual Property Review (1991): 400. 29 Ibid. 30 Blind, The Economics of Standards: Theory, Evidence, Policy 79–80.

14


3.1.2

The international system of standardisation

The institutions There are three organisations for technical standardisation on the international level: IEC, ITU (International Telecommunication Union) and ISO.31 IEC is similar to CENELEC and is responsible for standards for electrotechnics and electronics. ITU plays a role similar to ETSI and closely collaborates with IEC, implementing standards for telecommunication. ISO is comparable to CEN and is the institution for all other standardisation. The standardisation process The standardisation process on the international level is fundamentally identical to the European standardisation process.32 International standardisation organisations adopt the principle of national delegation and thus the national interest group cannot participate in standardisation on the international level. However, they may indirectly take part in the process if they become a part of the delegation of the national standardisation organisation. 3.2

Informal Standardisation

3.2.1 Benefits of compatibility standards Compatibility standards make it cheaper to produce complementary products, reduce the switching costs, and make it easier to use products in combinations.33 The main advantages of compatibility standards are as follows. Firstly, standards enlarge the market for complementary products because they may lead to cost reduction and greater variety of complementary products. They increase competition between complementary producers, which is likely to lower the price of the complementary products, and save the educational cost of training people to use the standards. Secondly, compatibility standards also help to move components from one product to another, lowering switching costs and decreasing the possibility of locking-in to a product. Thirdly, standards make it possible for network products to be connected to networks and thus directly increase the value of the use of the networks, the value created by the positive network effects. While the benefits described above are demand-side advantages, there are also supply-side benefits, which are derived from shared development costs and learning effects in manufacture. 31 32 33

Ibid. 80–83. Ibid. Peter Grindley, Standards, Strategy, and Policy: Cases and Stories (Oxford: Oxford University Press, 2002) 25–6.

Industry standards

15

Likewise, compatibility standards provide various benefits to the market and these advantages increase the need for the compatibility standards. 3.2.2 Dynamics of standardisation The mechanism of market-driven standardisation depends on the path taken and the conditions in the early stages are crucial to the outcome of the standardisation process.34 Where a product is better regarded by customers in the early stage of standardisation, more customers will buy the product. This small advantage in the early stage of standardisation significantly affects the whole picture of the standard war. The early advantage of a market leader attracts more customers, increasing the installed base of the leader. The larger the installed base of users is, the more complementary products are developed for the standard, which helps to increase the credibility of the standard and thus makes the standard more attractive. This leads to more adoptions by users, which further enlarges the installed base. Eventually, the market is dominated by the standard. These ‘bandwagon effects’ or ‘tipping effects’ allow the leader of a market to harvest a large advantage. Thus the early stages of the standard setting process by markets are the critical period because a small advantage over competitors may induce the cumulative support process.35 Users’ decisions for early adoption depend heavily on the expectations of success of a standard,36 the implication of which is that the credibility of the standard is crucial in the early stages of the standard setting process in markets. Firms may increase the credibility of their standards by showing assuring evidence to users, such as evidence of commitment, manufacturing capability, and complementary support. Once the prospects for the standards become clearly understood in the market, the installed base effect starts to be predominant.

4.

STANDARDISATION AND TECHNICAL CHANGE

4.1

Overview

Technical standards can influence technical development in various ways.37 Standards reduce the variety of products from which consumers can choose, excluding special individual wishes. Standards may influence technical 34 35 36

Ibid. 27. Ibid. 28. Joseph Farrell, ‘Standardization and Intellectual Property’, Jurimetrics Journal 30 (1989): 40. 37 Blind, The Economics of Standards: Theory, Evidence, Policy 25–8.

16


change negatively through prevention of possible product variations which may provide the basis for the development of new products. In addition, false standardisation can diminish other variation in the interest of standardisation and lower the incentive to change to another standard. On the other hand, it is argued that the negative effects are counterbalanced by the advantages of standards.38 The decrease in the number of system elements due to variety reduction and the increase in the combination possibilities of a single element through compatibility provide the possibilities for mass production, reduce the costs and price, and enlarge the potential number of consumers. There are additional benefits of a concentration of research and development, which decreases market risks and the cost of searching for information. Technical standards contain information about the state of the art of technology and provide a good basis for researchers to generate new ideas. Information flows and cost savings in the innovation process are generated through free know-how transfer between the authors of formal standards and users.39 4.2

Network Externalities and Technical Change

As far as network externalities are concerned, positive externalities appear when a network good is consumed because the benefit from the use of a network good depends on the consumption of the network good by other economic units. Therefore, although network goods, such as video cassettes, telephones, the internet and typewriter keyboards, may be privately owned, they develop their full usefulness through technical or economic connections to similar network goods of other owners.40 Joining the network is more likely to be considered, the more other members it already has. It is clear that a new technology will not be accepted unless a sufficient number of economic entities has already accepted it, namely, critical mass is reached, even though this technology is advantageous from the perspective of technological performance. However, where nobody wants to adopt the new technology unless others also adopt it, the adoption would be difficult since nobody hopes to be the first one to use the technology. Farrell and Saloner called this inefficiency the ‘penguin effect’,41 by analogy with penguins at the

38 39 40 41

Ibid. 26. Ibid. 27. Ibid. 29. Joseph Farrell and Garth Saloner, ‘Competition, Compatibility and Standards: The Economics of Horses, Penguins and Lemmings’, in Product Standardization and Competitive Strategy, ed. H. Landis Gabel, Advanced Series in Management (Oxford: Elsevier Science Publishers, 1987).

Industry standards

17

South Pole wanting to catch fish in the water but unable to jump for fear of possible predators swimming around. Where some penguins jump in the water and show that it is safe, other penguins will follow those penguins and seek to catch fish in the sea. However, no penguin wants to be the first one to take the risk. Another way to reduce the risk is the simultaneous jump by many penguins since the more penguins are in the water, the less the probability of being a prey. The penguin effect clearly demonstrates the difficulty of changing to a new standard. In this example, changing to a new technology is analogous to a simultaneous jump by all the penguins, which does not occur easily in a market. In addition, it may be difficult for users or vendors to coordinate a switch from an old technology to a new one, even if all would like to do so. Farrell explains this coordination problem with the example of horses tied together.42 Where two horses are tied together, they can go anywhere should they coordinate their behaviours. However, the horses cannot go that far since they fail to agree where and when to go. Likewise, because of the coordination problems, switching from an incumbent technology to a new one may not be easily achieved. Farrell and Saloner describe the economic inefficiencies of excess momentum and excess inertia.43 Excess momentum, which means the inefficient adoption of a new technology, entails excessive costs that the users of the old technology must bear. Since not every new technology is better than the old one, too early change from an old one to a new one may harm the technological development, entailing economic efficiency. Too early changes may create a ‘blind giant’,44 which means the further development of an inferior technology. A society where new technologies are preferred is in danger of adopting every new technology without considering the costs and benefits of the adoption. In contrast, excess inertia, which is the inefficient attachment to an old technology, is caused by the costs incurred in switching from an old technology to a new one, where the costs are higher than the benefits achieved from the adoption of the new technology. Just as in the penguin dilemma, all users may resist the change and stick to an old standard due to lock-in effects. Another economic inefficiency in relation to network effects is the parallel existence of two or more incompatible technologies, which occurs when the

42 43

Farrell, ‘Standardization and Intellectual Property’, 37. Joseph Farrell and Garth Saloner, ‘Standardization, Compatibility and Innovation’, The RAND Journal of Economics 16, no. 1 (1985). 44 Paul A. David, ‘Some New Standards for the Economics of Standardization in the Information Age’, in Economic Policy and Technological Performance, ed. Partha Dasgupta and Paul Stoneman (Cambridge: Cambridge University Press, 1987).

18


positive network effects are not fully utilised.45 Where an existing technology is still young and thus is not widely accepted in the market, a change to a new technology may be accepted by those who are not using the older technology. However, it is not likely to be adopted by the users of the old technology because they have to bear switching costs in adopting the new one. The adopters of the old technology will persist in using the technology and will lose not only the network benefits but also the benefits resulting from the new and better technology. In this case, the adopters of the old technology are called ‘angry orphans’, who hamper technical change.46 On the other hand, once the critical mass of a new and advantageous technology is reached, an increasing number of members will accept the new technology because it is beneficial for the economic units to adopt it. This bandwagon effect is likely to be facilitated by compatibility standards since standards applying to a new technology and having compatibility aspects reduce the risk of failing to achieve critical mass. Also, standards can achieve a regulated coordination of the adoption of network goods and thus they can solve the problem of horses bound together. Where network externalities exist, agreements on a common network with respect to a technology are beneficial to societies. Such agreements between economic entities on a common network can be achieved by a dominant firm in the market, by compatibility agreements among suppliers or by governmental regulation, as described in the following sub-sections. 4.3

Coordination Mechanisms and their Impact on Technical Change

There are three possible coordinating mechanisms for standardisation: market coordination, coordination by mutual agreements between manufacturers, and governmental regulation.47 4.3.1 Market coordination If perfect competition and different product or standardisation preference are dominant in the market, an optimal solution to the coordination problem can be achieved by the market mechanism.48 However, as perfect competition is practically impossible, there are dangers of suboptimal standardisation by market process. Where a firm can set the price of its products lower than competitors’, the firm can increase its market share and eventually come to possess a sponsored standard corresponding to its own products. Industry stan45 46 47 48

Blind, The Economics of Standards: Theory, Evidence, Policy 34. Ibid. 35. Ibid. 37–40. Ibid. 37–8.

Industry standards

19

dards by one manufacturer yield concentrated market structures, making the manufacturer a monopolist for the particular good. As economic theories explain, monopolistic market structure would result in resource misallocation, caused by low production and high price.49 That is, the monopolist would produce the monopolised good to an amount less than the market demand, which gives rise to high price and welfare loss. Furthermore, since the monopolist determines any change in the technology concerned, a lack of competition may reduce the incentive for the monopolist to innovate. 4.3.2 Coordination by agreements Where a common interest to agree on a standard exists on the part of the actors involved, an optimal coordination can be reached by way of mutual agreements between manufacturers.50 The agreement may be formal or informal and binding or voluntary. There are two mechanisms of coordination by agreements: consortia and standard setting organisations. Consortia The strategic alliance between vendors who sponsor specific technologies may generate standards.51 Forming consortia is a popular solution to standardisation problems in the information industries. A consortium helps those companies participating in it expedite the rate of developing complementary components and may also lead other firms to produce complementary products because the consortia can provide assurance for economic units that the technologies sponsored by the consortia will be widely adopted in the industries. However, consortia are not always an optimal coordination method because different consortia sponsoring different technologies may stifle the appearance of a standard, as in the case of UNIX standards.52 Furthermore, there may be anticompetitive effects with consortia should the consortia take concerted actions to exclude competing firms and to build high entry barriers. Standard setting organisations Standards can be made by agreements between firms participating in standard setting organisations (SSOs). Many SSOs are in principle open to interested parties in order to promote the discussion, formulation and dissemination of standards. SSOs play an important role

49 F.M. Scherer and David Ross, Industrial Market Structure and Economic Performance, 3rd edn (Boston: Houghton Mifflin Company, 1990) 21–3. 50 For more detail, see Farrell and Saloner, ‘Competition, Compatibility and Standards: The Economics of Horses, Penguins and Lemmings’. 51 Shane Greenstein, ‘Invisible Hands and Visible Advisors: An Economic Interpretation of Standardization’, Journal of the American Society for Information Science 43, no. 8 (1992): 543–4. 52 Ibid.

20


in solving coordination problems in that they provide a mechanism to develop or choose a standard from many technical candidates. Most of the standardisation organisations are voluntary, which means that adopting the standards is optional and that the degree of involvement is within the firms’ discretion.53 However, this mechanism is not a perfect solution to coordination problems, either. Where standardisation committees are evaluated not by the economic and technological efficiency of the published standards but by quantitative output, suboptimal standardisation may emerge since SSOs are likely to create standards too early or to incorporate the wrong technology in standards without considering the costs and benefits of doing so.54 Moreover, SSOs may fail to develop standards when essential technologies are protected by intellectual property rights. Furthermore, where the participants in the standard setting process have different preferences for standards, no or too little standardisation takes place. As with consortia, there may be anticompetitive effects with these standardisation activities. 4.3.3 Governmental regulation Where markets do not provide sufficient incentive for vendors or users to proceed toward a standard which is beneficial to the public, or where coordination problems are not likely to be solved in the private sector, governments can intervene in standardisation and facilitate the standardisation process. A government agency may be involved in the standardisation process when the agency has a mandate to regulate the industries, as is the case of the telecommunication industry and the broadcasting industry. However, it is argued that the governmental solution is the least promising one, as it cannot be assumed that the governmental decision makers are well informed about technological developments.55 Moreover, government agencies are vulnerable to political power and tend to be slow at making a decision.

5.

STANDARDS AND COMPETITION

5.1

Overview

Both pro-competitive and anticompetitive effects can be generated by standards and standardisation activities. David and Steinmueller present an overview of the pro-competitive and anticompetitive impact of compatibility 53 54 55

Ibid. 544. Blind, The Economics of Standards: Theory, Evidence, Policy 38. Mark A. Lemley, ‘Antitrust and the Internet Standardization Problem’, Connecticut Law Review 28 (1996): 1061–3.

Industry standards

21

standards.56 In general, a standard tends to be pro-competitive because the informational attributes of standards are close to public goods.57 That is, where a firm hoping to use compatibility standards can gain access to the standards without obstacles, the standards can contribute to the increase of market competition by providing essential information to market participants, by lowering barriers to entry into a market, by reducing the switching costs of customers and by restricting monopoly power. However, where the benefits of standardisation cannot be appropriated in the same way by all the firms involved, and where the access to such standards by rival firms is restricted, standards may be used strategically to increase market power.58 5.2

Pro-competitive Effects of Standards

There are various pro-competitive effects of standards, which arise from the fact that standards have the aspects of public goods.59 The technical information contained in standards reduces the information asymmetries, helping competing companies gain access to the same information and thus have equal starting positions. Compatibility standards reduce barriers to entry into a market for network products or multi-component products because companies can concentrate on reducing the cost or increasing the quality of those attributes not specified by the standards. Producers do not have to invest in R&D and production of complete systems, which not only lowers entry barriers but also increases price competition. Publication of compatibility standards, or the achievement of interoperability between different products through converters and gateway technologies, diminishes the switching costs that customers have to bear, increasing the competitive pressure in the market. Moreover, non-proprietary compatibility standards, which are implemented by voluntary SSOs, tend to limit the possibility of monopoly leveraging through interface manipulation. In addition, the formulation and utilisation of compatibility standards is a precondition for the effective decentralisation of network industries like the telecommunication industry that previously were regulated natural monopolies.

56

Paul A. David and W. Edward Steinmueller, ‘Economics of Compatibility Standards and Competition in Telecommunication Networks’, Information Economics and Policy 6 (1994). 57 Ibid. 220. 58 Blind, The Economics of Standards: Theory, Evidence, Policy 41. 59 David and Steinmueller, ‘Economics of Compatibility Standards and Competition in Telecommunication Networks’, 221–2.

22


5.3

Anti-competitive Effects of Standards

The anticompetitive effects of standards arise where one or very few companies internalise the benefits of standards, extracting high rents or imposing high costs on their rivals.60 The procurement standards of large buyers such as governmental bodies tend to favour existing major producers that seem to assure supply or meet reliability criteria, causing the standards to dominate the whole market, irrespective of the needs of other customers. In addition, government agencies that are in charge of issuing standards may be influenced by major domestic producers because the dominant firms own the technical expertise needed to formulate the specifications of standards. This phenomenon may be expanded to international level such that international SSOs may be vulnerable to being dominated by representatives of major vendors rather than minor suppliers or users, because the dominant firms have sufficient resources to undertake relevant R&D and the capability to send experts to participate in the meetings of the standards committees. A group of producers participating in the standard setting process can conspire to implement standards incorporating technologies of which they have control through intellectual property rights, imposing additional costs on rival firms. As for the informal, market-driven process of standardisation, installed base effects, which force consumers to follow the choices of other consumers, may strengthen the market power of a dominant firm. In network markets where network externalities are significant, the importance of the predominant installed base position in markets is likely to lead the standard’s owner to predatory price-setting strategies, which may be used to increase entry barriers. It is clear that the more open and less proprietary the standards are, the more beneficial they are for social welfare, at least from the perspective of static efficiency. However, in order to introduce innovation into the market, it is necessary to provide incentives, which may be in the form of proprietary standards, for companies to invest in R&D. This causes tension between the static and dynamic efficiency of standards.

60

Ibid. 222–4.

3. 1.

Patents and standards in the US THE LEGAL ISSUES

Patent rights, which are granted by statutory patent law, are exclusive rights and thus the patent holders are entitled to exclusive control of the patented subject matters. Patent owners have the right to refuse to deal or refuse to grant access to the patented technologies. However, it may be undesirable to allow the patent holders who have the patents covering industry standards to enjoy absolute exclusivity because they may use the rights for anticompetitive purposes or to impede technical innovation. A company cannot compete in a network market without access to standards and thus the absolute exclusivity of patents covering standards may exclude all the competition in the market, increasing social costs arising from patent systems. Likewise, the legal issues concerning the patent rights and standards in network industries arise in those cases where the owners of patents which have become industry standards refuse to license the patents to those who wish to offer interoperable products or services. Where third parties cannot access the patented technology essential to enter the market because the patent holders do not allow them to do so, the third parties may turn to antitrust law to gain access to the proprietary technology because antitrust liability may grant a compulsory licence. It is here that tension exists between patent law and antitrust law concerning the extent to which the antitrust law limits the exclusive rights of patent holders. Indeed, standards and standardisations involve the complex interactions of antitrust law and patent law and are considered to be at a crossroads between them.1 There are two sorts of standards, formal standards and informal standards, depending on the process which produces the standards. Formal standards are promulgated by standard setting organisations, whereas informal standards are generated by market mechanisms in those markets where the network effects are strong. The patent issues of formal standards are different from those of informal standards because constraints placed on the holders of patents essential to standards are greater in formal standards than in 1 Melissa Landau Steinman, ‘Standards, Antitrust and Intellectual Property: Standard-Setting at the Crossroads’, Practising Law Institute 832 (2005).

23

24


informal standards.2 Informal standards arise from market process and thus the patent holders of informal standards do not have to agree to the terms and conditions of licences of their patents in getting their patented technology adopted as standards. As for formal standards, however, standard setting organisations usually have IPR policies which impose some obligations about patent licences on the patent owner. Thus, in this chapter, the patent issues in formal standards and those in informal standards are examined separately.

2.

PATENTS AND INFORMAL STANDARDS

In order to analyse the legal issues concerning patent rights and industry standards, the general antitrust principles on refusal to license and refusal to supply patented goods need to be examined. In addition, there are three antitrust principles which need to be studied to see whether they can be used to resolve the accessibility problem caused by the refusal to license patents essential to standards. They are the essential facilities doctrine, the monopoly leveraging doctrine and the Aspen Skiing principle.3 Therefore, the general antitrust principle on refusal to license and the three antitrust principles which the third parties hoping to gain access to patented technologies may rely on are examined in the following sections. 2.1

Relevant Antitrust Law: Section 2 of the Sherman Act (15 U.S.C. §2)

Section 2 of the Sherman Act proscribes monopolisation and attempts to monopolise, where monopolisation is acquiring or maintaining market power through anticompetitive conduct.4 Every person who shall monopolise, or attempt to monopolise, or combine or conspire with any other person or persons, to monopolise any part of the trade or commerce among the several States, or with foreign nations, shall be deemed guilty of a felony, and, on conviction thereof, shall be punished by fine.5

2 Penelope A. Preovolos, ‘Antitrust, Intellectual Property, Standards and Interoperability’, Practising Law Institute 566 (1999). 3 James B. Kobak, ‘Intellectual Property, Refusals to Deal and the U.S. Antitrust Laws’, Practising Law Institute, Patents, Copyrights, Trademarks, and Literary Property Course Handbook Series 832 (2005). 4 Mark A. Lemley, ‘Antitrust and the Internet Standardization Problem’, Connecticut Law Review 28 (1996): 1066–7. 5 15 U.S.C. §2.

Patents and standards in the US

25

The definitions and elements necessary to establish the offence of this provision are set up by case law. The monopoly power is defined as ‘the power to control prices or exclude competition’.6 The elements of a monopolisation claim are: (1) monopoly power in the relevant market; and (2) the wilful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.7 The elements of an attempted monopolisation claim are: (1) the relevant geographic and product market(s), (2) predatory or anticompetitive conducts, (3) specific intent to monopolise, and (4) a dangerous probability of successful monopolisation.8 2.2

General Antitrust Principles on Refusal to License Patents

It has long been held by the US courts that a mere refusal to license patented technology does not constitute an antitrust violation because the right to exclude competitors from access to patented subject matters is the essential part of the patent rights.9 The majority of cases involving a refusal to license intellectual property have followed this rule. In SCM Corp. v. Xerox Corp.,10 the US Court of Appeals for the Second Circuit had to decide whether the refusal to license by a patent holder enjoying a monopoly in the relevant market violated the antitrust law. Xerox, possessing a monopoly power in the plain-paper copier market, refused to grant to SCM a licence of patents covering the plain-paper copying technology, and thus SCM could not enter the market without infringing Xerox’s patents. SCM filed a complaint alleging, inter alia, that Xerox’s refusal to license its plain-paper xerography technology violated the antitrust law. In reaction to this allegation, Xerox argued that its refusal to license patents for plain-paper copying was a lawful exercise of patent power. After reviewing the relationship between the patent law and the antitrust law, the court held that [i]f the threat of treble damage liability for refusing to license were imbedded in the minds of potential patent holders as a likely prospect incident to every successful commercial exploitation of a patented invention, the efficacy of the economic incentives afforded by our patent system might be severely diminished. … Where a

6 7

Aspen Skiing Co. v. Aspen Highlands Skiing Co., 472 U.S. 585, 596 (1985). S.W. O’Donnell, ‘Unified Theory of Antitrust Counterclaims in Patent Litigation’, Virginia Journal of Law and Technology 9 (2004): 22. 8 Ibid. 9 Preovolos, ‘Antitrust, Intellectual Property, Standards and Interoperability’, 257. 10 SCM Corp. v. Xerox Corp., 645 F.2d 1195 (1981).

26

Patents and industry standards patent in the first instance has been lawfully acquired, a patent holder ordinarily should be allowed to exercise his patent’s exclusionary power even after achieving commercial success; ... [W]here a patent has been lawfully acquired, subsequent conduct permissible under the patent laws cannot trigger any liability under the antitrust laws.11

The court did not endorse the SCM’s argument that ‘a unilateral refusal to license a patent should be treated like any other refusal to deal by a monopolist’,12 but ruled that where patent rights are lawfully obtained, patent holders are entitled to refuse to license patents unilaterally. This case can be interpreted in the context of industry standards. As the SCM court reasoned, the patented product of Xerox was so successful that it constituted a separate market.13 In the same way, industry standards are likely to create a distinct market because they are not substitutable. In addition, like SCM in this case, a firm cannot enter the relevant market without access to the industry standards of the market. The analogy between the SCM case and the compatibility problem in relation to industry standards implies that the holders of patents covering informal industry standards can lawfully refuse to grant a licence for the patents if the patent holders legally acquired the patents. The reasoning of the Second Circuit was endorsed by the Sixth Circuit in Miller Insituform, Inc. v. Insituform of North America.14 Insituform International, N.V. owned a patented process for the rehabilitation of pipelines. In June 1980, Insituform International granted to Instituform of North America (INA) the exclusive licence to the process throughout the United States, except in California, making it possible for INA to grant an exclusive sublicence to Miller Insituform, Inc. (MII). On 9 May 1984, the exclusive licensee, INA, terminated a sublicence which it had granted to MII. MII argued that the termination of the sublicence agreement was not justified and that INA had illegally attempted to monopolise the relevant market by terminating the sublicence. In reaction, INA claimed that, by virtue of the patent laws, it had the right to exclude others from using the patented process. The court, citing SCM, held that the holder of a patent retains the power to exclude others from manufacturing, using, and selling his inventions without running afoul of the antitrust laws. Here, by terminating the sublicense agreement with the appellant (MII), appellee (INA) merely exercised his power to exclude others from using the Insituform process, as

11 12 13 14

Ibid., 1206. Ibid., 1204. Ibid., 1203. Miller Insituform, Inc. v. Insituform of North America, F. 2d 606 (1987).


27

was its right under 35 U.S.C. § 154. In so doing, it did not violate section 2 of the Sherman Act.15

Likewise, the court reasoned that the patent holder or exclusive licensee was not obliged to grant access to the protected technology. This line of reasoning continued in Independent Service Organizations (ISOs) Antitrust Litigation.16 Xerox refused to provide independent service providers with replacement parts for printers and copiers. An ISO, CSU, filed a suit against Xerox, alleging that Xerox violated section 2 of the Sherman Act. In reaction to this antitrust claim, Xerox counterclaimed that patent and copyright had been infringed and argued that CSU could not assert a patent or copyright misuse defence to infringement counterclaims based on Xerox’s refusal to deal. The Court of Appeals for the Federal Circuit (CAFC) created exceptional circumstances under which the exclusive rights of patent holders are limited such that the patent infringement defendant can succeed in antitrust claims. The circumstances are (1) the patentee obtained its patent through ‘knowing and willful fraud’,17 (2) the patent infringement suit is ‘both objectively baseless and subjectively motivated by a desire to impose collateral, anti-competitive injury rather than to obtain a justifiable legal remedy’18 (sham litigation), or (3) the patent was used as an illegal tying strategy to extend market power. The court held that Xerox’s refusal to sell or license its patented spare parts did not go beyond the scope of the patent because, in the absence of exceptional circumstances, a patent may grant the right to exclude competition altogether in more than one market. It is argued that the decision of the Xerox case makes it unlikely that the patentee’s refusal to license patents or to sell patented products will amount to an antitrust violation, even though it is motivated by anticompetitive purposes.19 Proving fraud at the Patent and Trademark Office (PTO) is a difficult task, requiring an infringement defendant to show that the patentee, by making knowing and wilful misrepresentations to the PTO, obtained a patent which would not have been granted without the misrepresentations.20 In addition, the patentee can rebut the claim by showing that there are reasons for the patent grant other than the misrepresentations. As for the sham litigation, the defendant has to show that the patent infringement suit is objectively baseless 15 16

Ibid., 609. In re Independent Service Organizations Antitrust Litigation, 203 F. 3d. 1322

(2000). 17 18 19

Ibid., 1326. Ibid. Nicolas Oettinger, ‘In re Independent Service Organizations Antitrust Litigation’, Berkeley Technology Law Journal 16 (2001): 332–3. 20 Ibid., 333.

28


and subjectively motivated, which imposes a great burden of proof on the defendant.21 As far as illegal tying is concerned, refusal to license essential patents does not have to do with illegal tying. In this way, the holding of this case implies that a patent holder has broad rights to refuse to license or to refuse to sell a patented product and that such refusal is not likely to be condemned even in those cases where the motive of the refusal is anticompetitive. Likewise, the US courts have shown support for patent holders who face antitrust charges by refusing to license patents, even though the patent holders have a monopoly in the relevant markets. It is submitted that the courts have taken this position because they considered that imposing a duty to license or to deal may reduce the incentive to innovate underlying the patent system such as to chill the effort and investment in innovation.22 Thus, although a company is denied access to informal industry standards which are essential to compatibility and without access to which the company cannot enter the relevant markets, it is unlikely that the courts will impose the antitrust liability on the patent holders on the basis of the need of the company. 2.3

Antitrust Principles on Informal Standards for Compatibility

In general, it is difficult to gain access to the patented technologies by resting on antitrust principles on refusal to deal or license. However, there are three antitrust principles which may be applied to the cases of refusal to license patents covering standards. They are the Aspen Skiing principle, the essential facilities doctrine and the monopoly leveraging doctrine, which are described in the following sections. 2.3.1

Aspen Skiing principle

The Aspen Skiing principle and the rebuttable presumption The case of Aspen Skiing Co. v. Aspen Highlands Skiing Co.23 is not an intellectual property case but is considered a landmark case concerning the refusal to deal by a monopolist. The defendant, Aspen Skiing Co., which owned three ski resorts in Aspen, had previously cooperated with the plaintiff in offering an ‘all Aspen ticket program’. After acquiring control of three of the four ski slopes, the defendant virtually refused to deal with the plaintiff, Aspen Highlands Skiing Co., by requiring the plaintiff to receive an unviably low level of revenue from 21 22

Ibid., 333–4. James B. Kobak, ‘Intellectual Property, Refusals to Deal and the U.S. Antitrust Laws’, 402. 23 See Aspen Skiing Co. v. Aspen Highlands Skiing Co.


29

the ticket sales of the joint venture. The plaintiff proposed several programs in order to resume the corporative relationship but all of them were rejected by the defendant. The plaintiff filed a suit against the defendant, alleging that the defendant violated the antitrust laws. In response, the defendant argued that even a monopolist was not obliged to engage in joint marketing with competitors and that it did not violate the antitrust law because none of its behaviours could be exclusionary. The Supreme Court reasoned that the absence of a duty to deal with others was the fundamental right of a businessman since it is the counterpart of the independent businessman’s right to select business partners or customers. The court stated that where there was no purpose to create or maintain a monopoly, the antitrust law did not ‘restrict the long recognised right of trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal’.24 The court also held that ‘whether Ski Co.’s conduct may properly be characterized as exclusionary’ should be answered by considering its impact on consumers, its effect on the plaintiff and the business justification of the defendant.25 The court reasoned that the All-Aspen ticket had superior quality because it could provide the skiers with a variety of experiences and convenience. The court also stated that the defendant’s refusal to deal had an adverse impact on the plaintiff and the injury caused by the refusal was substantial. In addition, the court reasoned that the defendant’s conduct was not justified by any normal business purposes since the defendant sacrificed the short-term benefits in an effort to reduce competition in the market in the long term by harming the competitor. On the basis of these analyses, the Supreme Court ruled that a refusal to deal by a monopolist violated section 2 of the Sherman Act when such refusal produced an important change that could not be justified by normal business purposes. The ruling of Aspen Skiing was applied to intellectual property cases by Circuit Courts. Following the reasoning of Aspen Skiing, the First Circuit and the Ninth Circuit adopted the ‘rebuttable presumption’ principle in Data General Corp. v. Grumman Systems Support Corp.26 and Image Technical Services, Inc. v. Eastman Kodak Company, respectively.27 In Data General Corp. v. Grumman Systems Support Corp.,28 Data General, a computer manufacturer, filed a copyright infringement suit against

24 25 26

Ibid., 601–2. Ibid., 605. For more detail, see Data General Corp. v. Grumman Systems Support Corp., 36 F.3d 1147 (1994). 27 Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195 (1997). 28 See Data General Corp. v. Grumman Systems Support Corp.

30


Grumman, an independent service provider, which was a competitor in the market for service to Data General computers. Grumman counterclaimed that Data General illegally maintained its monopoly in the market for service of Data General computers by refusing to license a diagnostic software, ADEX. In response, Data General argued that ‘a unilateral refusal to license a copyright can never constitute exclusionary conduct’.29 The First Circuit held that in order for Grumman to prove that Data General wilfully maintained its monopoly in the service market, Grumman must demonstrate ‘(1) Data General’s possession of monopoly power in the market, and (2) Data General’s maintenance of that power through exclusionary conduct’.30 The court adopted the rebuttable presumption by holding that ‘while exclusionary conduct can include a monopolist’s unilateral refusal to license a copyright, an author’s desire to exclude others from use of its copyrighted work is a presumptively valid business justification for any immediate harm to consumers.’31 The court’s decision suggested two situations where the presumption may be rebutted.32 Firstly, the presumption may be overcome in an Aspen Skiing type case; that is, where a copyright owner initially licensed its copyrights in a competitive market and then changed its behaviour after the competition was curtailed, the presumption might be rebutted. Secondly, where the copyrights were acquired unlawfully, the presumption could be rebutted. The court reasoned that because Data General had always been a monopolist in the relevant market and thus there had never been a competitive market, the Aspen Skiing analogy was not established. This case concerns copyrights but the same line of reasoning is applied to patent cases in Technical Services, Inc. v. Eastman Kodak Company (Kodak II).33 In Kodak II, Kodak produced, sold and serviced photocopiers and micrographic equipment. In the late 1970s, the service market was small compared with the equipment market and thus manufacturers did not consider it an important market. In the early 1980s, independent service organisations (ISOs) began entering the service market. As the market grew bigger and ISOs became more competitive, Kodak, finding that ISOs were strong competitors in the service market, stopped sales of micrographic parts to them. ISOs filed a suit against Kodak, alleging that Kodak violated the antitrust laws by monopolising or attempting to monopolise the sale of service for Kodak machines. Kodak argued, inter alia, that the essential facilities doctrine is the sole legal theory which could force Kodak to sell spare parts but the element 29 30 31 32 33

Ibid., 1184. Ibid., 1181. Ibid., 1187. Ibid., 1187–8. For more detail, see Image Technical Servs., Inc. v. Eastman Kodak Co.


31

of the doctrine was not established34 and that its refusal to sell replacement parts was justified by its patent rights.35 The court rejected Kodak’s argument that the essential facilities doctrine is the only theory concerning the refusal to deal by a monopolist and instead adopted the reasoning of Aspen Skiing. In addition, the court noted that there are two principles concerning the relationship between patent laws and antitrust laws: (1) patent holders are not immune from antitrust liability, and (2) patent holders may refuse to sell or license protected works.36 The court sought to reconcile these conflicting principles by modifying the rebuttable presumption theory established by Data General. The court, citing Data General, held that ‘while exclusionary conduct can include a monopolist’s unilateral refusal to license a [patent or] copyright, or to sell its patented or copyrighted work, a monopolist’s desire to exclude others from its [protected] work is a presumptively valid business justification for any immediate harm to consumers’37 (citations omitted). The court also held that the presumption could be rebutted in three circumstances: (1) the Aspen Skiing analogy: namely, where a monopolist changed its practice after freely licensing its intellectual property in a competitive market; (2) where the monopolist acquired the intellectual property rights unlawfully; or (3) where protection of intellectual property is merely a pretext, and the real reason for the intellectual property owner to refuse to license is anticompetitive. The Ninth Circuit found evidence of pretext based on testimony that Kodak had not considered intellectual property issues when changing its parts policy. Thus the court modified the rebuttable presumption by supplementing evidence of pretext to the conditions where the presumption might be rebutted. However, the Kodak II decision has been widely criticised by commentators who argue that it is incompatible with the Patent Act of 1988 and with the ruling of the Supreme Court.38 Section 271(d) of the Patent Act provides that: No patent owner otherwise entitled to relief for infringement … of a patent shall be denied relief or deemed guilty of misuse or illegal extension of the patent right by reason of … [the patent owners’] (4) refusal to license or use any rights to the patent.

It is argued that the Kodak II court disrespected the intent of Congress.39 34 35 36 37 38

Ibid., 1209. Ibid., 1212. Ibid., 1215. Ibid., 1218. Sergio Baches Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, Fordham Intellectual Property, Media and Entertainment Law Journal 11 (2001): 485. 39 Ibid.

32


It has also been claimed that the intent-based presumption in Kodak II is inconsistent with the Supreme Court’s holding in Professional Real Estate v. Columbia Pictures Industries (PRE).40 In PRE, the Supreme Court held that where an antitrust claim is premised upon an allegation of sham litigation, a test must be satisfied: the lawsuit must be objectively baseless in the sense that no reasonable litigant could expect success on the merits and only if the challenged litigation is objectively meritless may a court examine the litigant’s subjective motivation.41 The Kodak II ruling is inconsistent with the PRE ruling in that, under PRE, a patent owner who files an infringement suit in order to exclude the alleged infringer from some markets may have claims on the merits of patent laws regardless of his intent, whereas the patent holder who merely refuses to license his patent is subject to antitrust scrutiny depending on his intent.42 Indeed, imposing more stringent constraints on refusal to license than on patent infringement suits is not justifiable. Accordingly, it seems that the modified version of the rebuttable presumption, which includes the pretext as a circumstance where the presumption can be rebutted, is not likely to be endorsed in other courts. Implication of the Aspen Skiing principle on informal standards These cases may have implications for patents in relation to industry standards. There are situations where the Aspen Skiing analogy may be established concerning informal standards. In a network market, because of the network effects, the more consumers use a product, the more likely it is that other consumers will decide to purchase the product. While there is no dominant technology in a relevant market and several technologies are competing to be an industry standard, competing firms may agree to grant licences of their patented technologies, which are necessary for compatibility between competing products, to each other, because where their products are compatible with each other they can increase the number of their potential customers and the installed base in the market. As noted, the installed base plays a significant role in the network market. If a firm’s product is compatible with those of competitors, consumers have more choices and do not have to worry about being ‘angry orphans’, that is, being locked into old technology and thus unable to obtain the benefits from the improved quality of new technologies as well as from the network

40 41

Ibid. David T. Pritikin and Bruce M. Zessar, ‘The Sham Litigation Doctrine in Patent Litigation: The Nexus Bad Faith Enforcement of Patent Rights and the Antitrust Laws’, Practising Law Institute 414 (1995). 42 Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 486–7.


33

effects of new networks.43 Thus, the compatibility between competing products is likely to increase customer welfare. After the competition is diminished and a technology becomes an industry standard, the holder of the dominant technology may refuse to license his patented technology. This causes serious damage to the competitors because consumers are not willing to buy a product which is not compatible with the predominant product. In addition, there may not be any compelling business justification of the patent holder other than excluding all competition in the market. It is this situation where the reasoning of Aspen Skiing may be applied and the presumption of valid business justification may be rebutted. Indeed, the reasoning of Aspen Skiing has been relied on by plaintiffs in those cases where a monopolist withdraws the access to patented technology, which was licensed in a competitive market, after gaining market power.44 The application of the Aspen Skiing principle to the case of refusal to license is, however, much limited to exceptional cases and, indeed, the Supreme Court stated that ‘Aspen Skiing is at or near the outer boundary of § 2 liability’.45 This reasoning is only applicable to an established relationship between competitors in a competitive market and, on top of that, the monopolist’s changes of dealing may be justified by legitimate business purposes like the prevention of free riding.46 In addition, the D.C. Circuit refused to uphold the claim that a change in the monopolist’s product design distorted competition in a rapidly changing market.47 Thus, a monopolist having the informal standards can lawfully refuse to grant licences or sell products in various ways and competitors gaining access to the patented informal standards by relying on the Aspen Skiing principle is exceptional. 2.3.2

Essential facilities doctrine

In general The essential facilities doctrine is concerned with the general duty to deal arising from the monopoly condition.48 This doctrine has special characteristics in that it imposes an obligation on the monopolist not because 43 G.M. Peter Swann, The Economics of Standardization: Final Report for Standards and Technical Regulations Directorate (Department of Trade and Industry, 2000) 22. 44 Preovolos, ‘Antitrust, Intellectual Property, Standards and Interoperability’, 264. 45 Verizon Communications, Inc. v. Trinko LLP, 540 U.S. 398, 879 (2004). 46 James B. Kobak, ‘Intellectual Property, Refusals to Deal and the U.S. Antitrust Laws’, 392. 47 United States v. Microsoft Corp., 253 F.3d 34, 65 (2001). 48 Herbert Hovenkamp, Mark D. Janis, and Mark A. Lemley, IP and Antitrust, vol. 1 (New York: Aspen Law & Business, 2002) 13-11–13-12.

34


of the monopolist’s affirmative conducts but because of the monopolised market condition.49 The essential facilities doctrine applies to a situation where a dominant firm which controls a facility essential to its competitors denies competitors access to that facility.50 This principle has been adopted by some lower courts, but the Supreme Court has never endorsed it in relation to intellectual property rights.51 It is generally accepted that this doctrine has its origin in United States v. Terminal Railroad Association, even if the court did not expressly state the ‘essential facilities’.52 The Terminal Railroad Association, composed of six railroads, acquired the St Louis rail bridge. Because the bridge was essential for any rail company to enter St Louis, the Association could exclude competitors. The Supreme Court ordered the Association to either accept any railroads as its members or grant access to the bridge on fair terms. This case concerns the violation of section 1 of the Sherman Act, which proscribes collective conducts such as contract, combination or conspiracy which restrain trade or commerce. This doctrine expanded to the case on unilateral denial by a monopolist in Otter Tail Power Co. v. United States.53 Otter Tail was a monopolist for electric power distribution in the upper Midwest of the United States, supplying both wholesale and retail electrical services in the area. A number of municipalities, after building their own generation facilities, decided to provide local distribution themselves. The municipalities asked Otter Tail either to supply the wholesale power only or to transmit another supplier’s electricity over Otter Tail’s power lines, which Otter Tail refused to do. The Supreme Court held that Otter Tail had violated section 2 of the Sherman Act by refusing to supply or transmit power to the municipalities. The elements of the essential facilities claim were established in MCI, where the court held that

49 50

Ibid. Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 419. 51 Christopher J. Meyers, ‘European Union Competition Law and Intellectual Property Licensing: Trans-Atlantic Convergence and Compulsory Licensing’, Practising Law Institute 842 (2005): 152. 52 Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 416. 53 W. Greg Papciak, ‘Sherman Act Violations: Essential Facilities Doctrine – Intergraph Corp. v. Intel Corp.’, Berkeley Technology Law Journal 14 (1999): 330–31.


35

[t]he case law sets forth four elements necessary to establish liability under the essential facilities doctrine: (1) control of the essential facility by a monopolist; (2) a competitor’s inability practically or reasonably to duplicate the essential facility; (3) the denial of the use of the facility to a competitor; (4) the feasibility of providing the facility.54

Concerning the first element of the doctrine, the plaintiff has to show that ‘control of the facility carries with it the power to eliminate competition in the downstream market’.55 As for the second element, ‘a plaintiff must show more than inconvenience, or even some economic loss; he must show that an alternative to the facility is not feasible’.56 As far as the third element is concerned, the denial includes outright refusal and terms or conditions to deal which are unacceptable to a reasonable person.57 As to the fourth element, not only the physical accessibility to the essential facility should be taken into account, but also the business justification for refusal of access to the facility has to be considered.58 Cases rejecting the essential facilities claim In Data General Corp. v. Grumman Systems Support Corp.,59 the district court of Massachusetts rejected the plaintiff’s essential facilities claim with persuasive logic. As described in the previous sections, in reaction to the copyright infringement suit by Data General, Grumman counterclaimed that Data General violated the antitrust laws. Grumman argued in the district court of Massachusetts that, among other things, ADEX was an essential facility and thus Data General had to license it to competitors. The court held that the plaintiff’s essential facilities claim, which required the defendant to provide the plaintiff with the proprietary information of the defendant, considerably overstepped the boundaries of the essential facilities doctrine. The court also held that superior knowledge in the design of Data General computers [was] insufficient to invoke the essential facilities doctrine; … If manufacturers of complex and innovative systems were required to share with competitors the development of accessories, because they had a possibly absolute advantage through producing the system, the incentives of copyright and patent laws would be severely undermined.

54 55 56 57

MCI Communications Corp. v. AT&T, 708 F.2d 1081, 1132–3 (1983). Alaska Airlines Inc. v. United Airlines Inc., 948 F.2d 536, 544 (1991). Ibid. James B. Kobak, ‘Intellectual Property, Refusals to Deal and the U.S. Antitrust Laws’, 394. 58 Ibid., 394–5. 59 Data General Corp. v. Grumman Systems Support Corp., 761 F.Supp. 185 (1991).

36

Patents and industry standards Not only would the manufacturer, who is in the best position to create these accessories, have less incentive to do so, but also the impetus for competitors to … produce competing solutions would be reduced.60

In Intergraph v. Intel,61 the Federal Court refused to apply the essential facilities doctrine to a circumstance where the plaintiff and the defendant did not compete in the downstream market. In 1993, Intergraph, which had previously manufactured workstations using various microprocessors, began to incorporate Intel ‘open architecture’ microprocessors into its workstations. In 1996, a dispute arose between Intergraph and Intel when Intel requested Intergraph to license, on a royalty-free basis, the patents covering the Clipper technologies. When Intergraph refused to grant licences of the patents covering the Clipper technologies to Intel, Intel cancelled related agreements with Intergraph, ceased to provide it with confidential or pre-release information, and denied it an allocation of microprocessors or pre-release samples. Intergraph sued Intel for, inter alia, antitrust violations, arguing that Intel’s technical assistance and other customer benefits are essential to compete in the workstation market.62 Intel argued that it should not be condemned by the monopolisation claim under section 2 of the Sherman Act because it did not compete with Intergraph in any market where Intel supposedly had or threatened to have a monopoly.63 The Court of Appeals rejected Intergraph’s claim that Intel violated antitrust laws under the essential facilities doctrine of monopolisation and endorsed Intel’s claim, holding that in order for the essential facilities doctrine to be applied there had to be a market where Intergraph and Intel competed. The court held that ‘absent such a relevant market and competitive relationship, the essential facilities theory [did] not support a Sherman Act violation.’64 Implication of the essential facilities doctrine on informal standards The essential facilities doctrine has its origin in a natural monopoly market where a single firm can meet the market demand at the lowest price and thus the market supports only one firm.65 In natural monopoly markets, the duplication

60 61 62 63

Data General Corp. v. Grumman Systems Support Corp., 192. Intergraph Corp. v. Intel Corp., 195 F.3d 1346 (1999). Ibid., 1356. Intergraph Corp. v. Intel Corp. (Reply Brief of Defendant-Appellant Intel Corporation), West Law 1, 1–4 (1998). 64 Intergraph Corp. v. Intel Corp., 1357 (1999). 65 David McGowan, ‘Regulating Competition in the Information Age: Computer Software as an Essential Facility under the Sherman Act’, Hastings Communications and Entertainment Law Journal 18 (1996).


37

of the essential facility requires an unreasonable cost such that the claimant of the essential facility has a high probability of establishing the second element of the essential facilities doctrine, which is ‘(2) a competitor’s inability practically or reasonably to duplicate the essential facility’.66 Thus, some have argued for the application of the essential facilities doctrine to informal standards, such as internet web browsers, by analysing the similarities between a natural monopoly market and the internet web browser market.67 However, there is uncertainty over whether informal industry standards have the characteristics of a natural monopoly. In order to claim that a facility is essential, ‘a plaintiff must show more than inconvenience, or even some economic loss; he must show that an alternative to the facility is not feasible’, as the Alaska Airlines court reasoned.68 However, even considering the network effects, there might be more than one competing technology or product in a network market, as is the case in the microprocessor market.69 In those situations where the market sustains more than one technology or product, it is not likely that the facility in question is essential, because a competitor may find an alternative.70 Another limitation of this doctrine is that in order for a firm to rest on the essential facilities claim, there should be a competitive relationship between the monopolist and the claimant, as the ruling of Intergraph shows. On top of these limitations, there are strong arguments against the essential facilities doctrine, which the US courts have often endorsed. Firstly, it is argued that compelling a monopolist to deal with competitors may benefit the competitors but does not improve customer welfare.71 Even though a monopolist is forced to deal with its competitors, the monopolist will grant access to its monopolised facility to its competitors at a monopolistic price level, and thus consumers have to pay the monopoly price anyhow. In fact, the forced deal may give rise to inefficiency, restraining the fundamental right of a businessman. In addition, even if the monopolist allows the competitors access to the facility at non-monopolistic rates, the competitors can charge

66 67

MCI Communications Corp. v. AT&T, 1132–3. Teague I. Donahey, ‘Terminal Railroad Revisited: Using the Essential Facilities Doctrine to Ensure Accessibility to Internet Software Standards’, AIPLA Quarterly Journal 25 (1997). 68 Alaska Airlines Inc. v. United Airlines Inc., 544. 69 Michael Kanellos, AMD’s Market Share Gains Accelerate (2005) (cited 19 May 2006); available from http://news.com.com/AMDs+market+share+gains+ accelerate/2100-1006_3-5916167.html. 70 McGowan, ‘Regulating Competition in the Information Age: Computer Software as an Essential Facility under the Sherman Act’, 805. 71 Adam Candeub, ‘Trinko and Re-grounding the Refusal to Deal Doctrine’, University of Pittsburgh Law Review 66 (2005): 830–31.

38


monopolistic rates as the monopolist does. In any case, the compulsory dealing does not improve consumer welfare but only results in the splitting of monopolistic profits between competitors. Thus it is argued that the total output and the price are not changed by mandating access to the facility claimed to be essential. Secondly, it is claimed that the courts are not competent to expect the outcome of the forced deal and to administer judicial remedies enhancing social welfare.72 The judicial remedy compelling a monopolist to deal with competitors requires judges to set prices and decide the terms and conditions of the dealing, which is ‘beyond the practical ability of a judicial tribunal to control’.73 Indeed, the reasonable royalties at the moment are not necessarily the reasonable level of royalties in the future since the reasonableness of the licensing fees varies over time, requiring continuous monitoring of the licensing contract.74 The Supreme Court recognised this point in Verizon v. Trinko, where the court held that [e]ffective remediation of violations of regulatory sharing requirements will ordinarily require continuing supervision of a highly detailed decree. … No court should impose a duty to deal that it cannot explain or adequately and reasonably supervise. The problem should be deemed irremedia[ble] by antitrust law when compulsory access requires the court to assume the day-to-day controls characteristic of a regulatory agency. … An antitrust court is unlikely to be an effective dayto-day enforcer of these detailed sharing obligations.75

Thirdly, it is contended that the essential facilities doctrine decreases dynamic efficiency.76 Requiring a firm to share its facilities with its competitors may diminish the incentives to innovate, reducing the investment in innovative activities which are necessary to create alternatives to the facilities. The compulsory dealing may also decrease the incentive of incumbent monopolists to introduce innovation because the mandatory sharing will reduce their return. These arguments have generally been endorsed by the US courts.77 To sum up, the US courts have shown great reluctance to apply the essential facilities doctrine to refusal to license cases and thus it seems difficult for a firm to gain access to the patented technologies covering informal standards by resting on this doctrine. In addition, the firms producing the complemen72 Phillip Areeda, ‘Essential Facilities: An Epithet in Need of Limiting Principles’, Antitrust Law Journal 58 (1989): 853. 73 Verizon Communications, Inc. v. Trinko LLP, 414. 74 Herbert Hovenkamp, Mark D. Janis and Mark A. Lemley, ‘Unilateral Refusals to License’, Journal of Competition Law & Economics 2 (2006): 8. 75 Verizon Communications, Inc. v. Trinko LLP, 414–15. 76 Candeub, ‘Trinko and Re-grounding the Refusal to Deal Doctrine’, 835–6. 77 Verizon Communications, Inc. v. Trinko LLP, 879.


39

tary products of informal industry standards cannot claim the essential facilities doctrine because they are not competitors of the firm owning the industry standards. Thus the essential facilities doctrine, as is used by the US courts, is not likely to be applied in resolving the compatibility problem of informal industry standards. 2.3.3 Monopoly leveraging doctrine The monopoly leveraging doctrine applies to those situations where a company uses or leverages its monopoly power in one market to obtain a competitive advantage in a second market, or to monopolise or attempt to monopolise the leveraged market.78 In the context of patent rights and standards, this doctrine refers to the situations where the patent owner uses his monopoly power in a primary market to leverage the monopoly into the related market for interoperable or compatible products or services.79 The federal circuit courts have made different rulings concerning the monopoly leveraging doctrine. This doctrine was established by the Second Circuit in Berkey Photo Inc. v. Eastman Kodak Co.80 The court held that ‘the use of monopoly power attained in one market to gain a competitive advantage in another is a violation of s 2 [section 2 of the Sherman Act], even if there has not been an attempt to monopolise the second market’.81 Thus the court established the elements of the monopoly leveraging doctrine: (1) there must be a monopoly power in one market, and (2) such power must be exercised to the detriment of competition in the second market. It does not preclude a violation whereby competition in the second market is merely distorted and not destroyed. However, the Ninth Circuit expressly rejected this doctrine. In Alaska Airlines v. United Airlines,82 the Ninth Circuit refused to endorse the reasoning of Berkey Photo, holding that the plaintiff must establish all elements of a monopolisation or attempted monopolisation claim to show a violation. That is, according to Alaska Airlines, in accusing the defendant of a monopolisation offence, the plaintiff has to show that the defendant has monopoly power in the primary and secondary markets, and that the monopoly power in the secondary market is wilfully acquired or maintained in a way distinguished

78 See Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’ 79 Preovolos, ‘Antitrust, Intellectual Property, Standards and Interoperability’, 268. 80 Berkey Photo Inc. v. Eastman Kodak Co., 603 F.2d 263 (1979). 81 Ibid., 276. 82 For more detail, see Alaska Airlines Inc. v. United Airlines Inc.

40


from growth or development as a consequence of a superior product, business acumen, or historic accident.83 As for the attempted monopolisation claim, the plaintiff has to define the relevant geographic and product markets and to show the defendant’s predatory or anticompetitive conducts, the defendant’s specific intent to monopolise, and a dangerous probability of successful monopolisation.84 Merely obtaining a competitive advantage in the second market was not enough to prove that the defendant violated section 2 of the Sherman Act. The different reasoning of the two courts as to the monopoly leveraging doctrine is due to the different approaches toward monopolies.85 The court of Berkey Photo showed hostility to monopolies, considering that monopolies should be allowed to exist only when it was the outcome of efficiency and that even the lawfully achieved monopolies should be subject to careful investigation when they were exercised.86 On the other hand, the court of Alaska Airlines considered that certain monopolies such as efficient monopolies and natural monopolies were tolerated by the antitrust laws, and reasoned that the elements of monopolisation and attempted monopolisation are essential to distinguish lawful monopolies from unlawful ones.87 Most other courts rejected the reasoning of Berkey Photo88 and, especially, the Supreme Court overruled the monopoly leveraging doctrine in Trinko. Here, the Supreme Court ruled that ‘to the extent the Court of Appeals dispensed with a requirement that there be a “dangerous probability of success” in monopolising a second market, it erred’.89 Thus, monopoly leveraging doctrine should be interpreted within the boundary set up by the monopolisation claim or the attempted monopolisation claim. The monopoly leveraging claim may be used by firms producing complementary products in network markets. For instance, if a monopolist in the PC operating system market refuses to license the patents protecting the PC operating system to a firm hoping to produce a compatible word processor, the firm may turn to the monopoly leveraging claim, arguing that the monopolist is attempting to monopolise the word processor market by leveraging the monopoly power in the PC operating system market. 83 84 85

O’Donnell, ‘Unified Theory of Antitrust Counterclaims in Patent Litigation’, 22. Ibid. James P. Puhala, ‘Antitrust Law–Berkey Photo and Alaska Airlines: Independent Approaches to Monopoly Leveraging Claims’, Western New England Law Review 16 (1994). 86 Berkey Photo Inc. v. Eastman Kodak Co., 274–5. 87 Alaska Airlines Inc. v. United Airlines Inc., 548. 88 James B. Kobak, ‘Intellectual Property, Refusals to Deal and the U.S. Antitrust Laws’, 396–7. 89 Verizon Communications, Inc. v. Trinko LLP, 415.


41

It has to be noted that a firm should show that all the elements necessary for the monopolisation or the attempted monopolisation claims are established, because merely obtaining a competitive advantage in the secondary market is not enough to allege the violation of section 2 of the Sherman Act. Therefore, the possible use of this doctrine in relation to informal standards is much limited. In addition, this doctrine cannot be applied to the case where the monopolist in the PC operating system market refuses to license patents covering its PC operating system to a competing firm producing another kind of PC operating system because the monopolist is not leveraging the monopoly power in the PC operating system market to a secondary market but is exercising its monopoly power in the same market where its monopoly power exists. 2.3.4

Review of antitrust principles on informal standards for compatibility The US courts have repeatedly held that the exclusive control by patent owners over the subject matter is the core of patent protection and, with the exception of Kodak II, there has been no reported case where a court imposes antitrust liability on a patent owner who refuses to license patents.90 As noted, the ruling of Kodak II has been highly condemned for its incompatibility with other cases and thus it is not likely to be adopted by other courts. Since the application of the Aspen Skiing principle, the essential facilities doctrine and the monopoly leveraging doctrine is limited to exceptional cases, the monopolist having the informal standards has various options to legitimately deny other firms access to the standards. The courts consider that forcing a firm, even if it is a monopolist, to deal with competitors reduces the social welfare rather than increasing it. To put it in relation to patents and standards, the courts consider that providing the incentive to innovate through the strong protection of proprietary technologies will increase the social welfare, even if the technologies are incorporated in standards. However, whether the strong protection of standards indeed increases the social welfare by increasing dynamic efficiency in the market is questionable. Many innovations have been achieved through gradual and incremental processes combined with the diffusion of the innovations.91 Secondary inventions such as improvement inventions and application of the primary invention to various uses are critical for an invention to be fully utilised in markets.92

90 Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 505. 91 Peter S. Menell, ‘Tailoring Legal Protection for Computer Software’, Stanford Law Review 39 (1987): 1338. 92 Ibid.

42


However, strong protection of an innovation may stifle innovations subsequent to the primary innovation because it may impede the wide use of the primary innovation. Indeed, the theoretical and empirical analyses do not necessarily teach that dynamic efficiency is improved by strong protection of intellectual property.93 Furthermore, the network effects also provide arguments against strong protection of standards. A dominant firm may lack the competence to meet the various demands of consumers and thus may not be able to expand the network enough to maximise the network benefits.94 Other firms hoping to provide goods or services that the dominant firm does not supply need to negotiate with the dominant firm to obtain licences, which raises the transaction cost. In addition, the dominant firm may refuse to license. Thus the strong protection of standards may reduce the positive network effects. Moreover, strong protection may strengthen the ‘excess inertia’,95 the inefficient attachment to the old technology, by preventing the appearance of new technologies and products. As economists explain, even if the new technology is superior to the existing technical standards, it is an uphill battle for the new technology to replace the incumbent standards.96 If the incumbent standards’ owner refuses to grant access to the protected technologies, it will be more difficult to introduce a superior alternative in the market. Developing new technologies or products which are compatible yet non-infringing will require a substantial amount of investment, which may stifle the technical advancement and impair the dynamic efficiency. As such, the imbalance between patent protection and the need for standards may suppress innovation, reduce network benefits and distort competition.

3.

PATENTS AND FORMAL STANDARDS

3.1

Issues of Formal Standards Incorporating Patents for Compatibility

3.1.1 In general Standard setting organisations (SSOs) often adopt standards which are covered by patents or are the subject matters of pending patent applications. The SSOs face a series of issues concerning patents: (a) whether or not to 93 94 95

Candeub, ‘Trinko and Re-grounding the Refusal to Deal Doctrine’, 836–7. Menell, ‘Tailoring Legal Protection for Computer Software’, 1342. Joseph Farrell and Garth Saloner, ‘Standardization, Compatibility and Innovation’, The RAND Journal of Economics 16, no. 1 (1985): 70–72. 96 Ibid.


43

adopt a standard incorporating patented technology and, if patented technologies are incorporated into standards, what the conditions should be to adopt the technologies; (b) what the procedures have to be to determine whether a proposed standard reads on patents; (c) what should be required of a patent owner whose patented technology is incorporated into a standard; and (d) what measures are necessary when relevant patents are discovered after a standard is adopted and the patent owner attempts to enforce the patents.97 Many SSOs implement intellectual property right policies (hereinafter ‘IPR policies’) to handle these issues. 3.1.2 Adoption of standards incorporating patents While some SSOs prohibit the adoption of standards which read on patents, most SSOs allow members to own patent rights in standards, even though they often discourage it or require royalty-free licences to members.98 For instance, the IPR policies of ANSI (American National Standards Institute) do not prohibit, in principle, the incorporation of patented technology in standards if technical reasons justify that.99 Patents are granted to non-obvious technologies and thus patented technologies are likely to be the advanced ones. Where a technology is excluded from the standard setting process only because it is covered by patents, the standards resulting from the standardisation may be inferior and therefore may not be widely accepted in the industry.100 In addition, the exclusion of patented technologies from standardisation may be condemned by antitrust laws.101

97

Preovolos, ‘Antitrust, Intellectual Property, Standards and Interoperability’,

274. 98 Mark A. Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’, California Law Review 90 (2002). 99 Steinman, ‘Standards, Antitrust and Intellectual Property: Standard-Setting at the Crossroads’. Also see ANSI Patent Policy (2007) (cited 30 December 2008); available from http://publicaa.ansi.org/sites/apdl/Documents/Forms/DispForm.aspx?ID= 6376&RootFolder=%2fsites%2fapdl%2fDocuments%2fStandards%20Activities%2fA merican%20National%20Standards%2fProcedures%2c%20Guides%2c%20and%20F orms&Source=http%3a%2f%2fpublicaa%2eansi%2eorg%2fsites%2fapdl%2fDocume nts%2fStandards%20Activities%2fAmerican%20National%20Standards%2fProcedur es%2c%20Guides%2c%20and%20Forms. 100 James C. De Vellis, ‘Patenting Industry Standards: Balancing the Rights of Patent Holders with the Need for Industry-Wide Standards’, AIPLA Quarterly Journal 31 (2003). 101 David M. Schneck, ‘Setting the Standard: Problems Presented to Patent Holders Participating in the Creation of Industry Uniformity Standards’, Hastings Communications and Entertainment Law Journal 20 (1998).

44


3.1.3 Notice and search of patents covering standards In many SSOs, it is required that members disclose the essential or relevant patents in relation to the proposed standards in order that the standard setting bodies may decide whether or not to incorporate particular technologies into the standards.102 Many SSOs have IPR policies which impose on members the duty to disclose patents related to standards, and failure to disclose the patents may preclude the patent owners from enforcing the patent rights.103 It is argued in relation to the Dell case104 that the extensive duty to disclose relevant patents may cause problems.105 Although a search is not required in many SSOs, searching the patent portfolio is inevitable for the members to disclose relevant patents and firms with a considerable patent portfolio may confront difficulties in searching and disclosing all the relevant patents. Thus, it is contended that a company which has important patents that can be relevant to a standard may be discouraged from participating in the standard setting process due to the uncertainty caused by the duty to disclose.106 However, it is not clear whether the duty to disclose the relevant patents would chill the participation of a firm having many patents in SSOs. Firms having a large patent portfolio are likely to have resources to manage their patent database and to train employees who attend meetings for standard setting work.107 Furthermore, the firms are likely to have systems to address the problem of the duty to disclose since, in order to obtain a patent grant, the applicant has to disclose relevant prior arts, including relevant patents, to the Patent Office. Whether they have large patent portfolios or not, firms have already been handling the problems arising from the duty to disclose information relevant to the patentability.108 Thus, it seems unlikely that the duty to disclose the relevant patents would require firms participating in SSOs to make an unreasonable effort. 3.1.4 IPR licensing policies Most SSOs require patent owners of standards to grant royalty-free (RF)

102 For more detail, see Lemley, ‘Intellectual Property Rights and StandardSetting Organizations’. 103 Schneck, ‘Setting the Standard: Problems Presented to Patent Holders Participating in the Creation of Industry Uniformity Standards’, 648. 104 In re Dell, No. 931-0097, Lexis 466 (1995). 105 Michael J. Schallop, ‘The IPR Paradox: Leveraging Intellectual Property Rights to Encourage Interoperability in the Network Computing Age’, AIPLA Quarterly Journal 28 (2000): 222–4. 106 Ibid. 107 Janice M. Mueller, ‘Patenting Industry Standards’, John Marshall Law Review 34 (2001): 929–34. 108 Ibid.


45

licences or reasonable and non-discriminatory (RAND) licences.109 The majority of SSOs adopt RAND licensing policies whereas open source proponents are in favour of RF licensing.110 It is a complicated question which licensing policy is optimal both to promote the wide adoption of standards and to incorporate advanced technology into standards. RF licences provide equal access to any company without any payment for licences and thus they may result in the widespread adoption of standards. In addition, RF licences may reduce the time delay caused by the vested interest of patent owners in the formal standard setting process.111 However, patent owners whose patented technology is the best one for standards may be reluctant to participate in the standard setting process because they are unwilling to grant licences on a royalty-free basis.112 Thus, RF licensing policies may produce inferior standards that will not be widely accepted in industry. As for RAND licences, they also have advantages and disadvantages. Under RAND licensing policies, SSOs have more choice to adopt patented technologies optimal for standards, which may lead to superior standards.113 On the other hand, because the meaning of ‘reasonable’ is not clear and SSOs deliberately leave this meaning equivocal to avoid antitrust liability, disputes over licensing fees may arise.114 3.1.5 Subsequent discovery of patents covering the standard It is possible that patents covering some parts of standards are discovered after the publication of standards and that the patent holder having the patents refuses to follow the rules of SSOs. Some SSOs have provisions for this problem. For instance, where this problem arises in ISO, the technical committee will consider the standards further, and where the patent is discovered after the standard is adopted, ANSI requires a patent owner to give ‘the same assurances to ANSI that are required in situations where patents are known to exist prior to the standard’s approval’, or withdraw the standard.115

109 Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’, 1905–06. 110 Ibid. 111 Joseph Farrell, ‘Standardization and Intellectual Property’, Jurimetrics Journal 30 (1989): 41. 112 De Vellis, ‘Patenting Industry Standards: Balancing the Rights of Patent Holders with the Need for Industry-Wide Standards’, 335. 113 Ibid. 114 P.D. Curran, ‘Standard-Setting Organizations: Patents, Price Fixing, and Per Se Legality’, University of Chicago Law Review 70 (2003): 983. 115 Kevin J. Arquit et al., ‘Antitrust, Intellectual Property, Standards and Interoperability’, Practising Law Institute 524 (1998): 200.

46

3.2


Legal Principles on Formal Standards Incorporating Patents

The IPR policies of SSOs impose on patent owners obligations such as the duty to disclose essential patents and the duty to grant licences to members. However, since the IPR policies are not laws, they are not enforceable by themselves but need to be exercised on the basis of legal principles. In this sub-section, the legal principles on formal standards incorporating patents which may enable the IPR policies to have binding force are examined.116 In addition, how the legal principles may work to resolve the conflict between patents and standards is analysed. 3.2.1 Contracts Because a firm is required to follow the IPR policies of an SSO when it becomes a member of it, the IPR policies may be considered a contract between a member and an SSO.117 The IPR policies may become an enforceable contract in various ways: the members may sign a document agreeing to follow the IPR policies; the members may sign a document which requires the members to read and comply with general bylaws, including IPR policies, of the SSO; or the contract may even be implied from sufficient factual circumstances without any written document and oral agreement.118 If a member company violates IPR policies by failing to disclose essential patents or refusing to grant licences on reasonable terms, the member company can be accused of breach of contract. However, the contract has some limitations. Firstly, it is between a member and an SSO that the contract exists such that not only nonmembers but also members are not contracting parties. Thus, nonmembers of an SSO cannot rely on the IPR policies to require licences for patents covering standards and to argue that the patent owner is contractually obliged to grant licences. Even though the patent owners have contractually agreed to license patents to the public, nonmembers lack standing to file a suit for breach of contract because they are likely to be incidental beneficiaries, who generally may not enforce contracts.119 As for the members of an SSO, even though they are not the contracting parties, they may request the SSO to enforce the contract on behalf of them. In those cases where there are no staff in an SSO who may file a suit instead of members because the SSO is just the combination of members without any independent administrating organisa116 For an excellent analysis of this issue, see Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’. 117 Ibid.: 1909. 118 Ibid.: 1910–11. 119 Ibid.: 1914–16.


47

tion, the members are likely to have standing to sue for breach of contract since they may be considered intended beneficiaries.120 Secondly, the remedy for breach of IPR policies is not likely to fully compensate the cost caused by the breach.121 The remedy for contract damage is usually not an injunctive relief but the compensation of the expected gain of the injured party from the contract. Therefore, in the case of a breach of an agreement to license on a reasonable and non-discriminatory basis, the injured firm will be entitled to compensation of the value, which it could have obtained by using the standards, after subtraction of the royalty rates, which it would have paid, from the value. Even though society suffers from the harm caused by the breach, there is no additional penalty to the patent owner and thus the patent owner has nothing to lose by breaching the contract. Furthermore, when it comes to a breach of the duty to disclose, measuring the damages is extremely difficult. Thus, even though IPR policies can be considered enforceable contracts, applying contract laws to enforce IPR policies has many limitations. Third parties have no right to enforce the contract and the remedy for the breach of contract is not likely to fully compensate the damage. Thus, applying contract laws is not considered an efficient measure to resolve the possible problems arising from the opportunistic behaviours of patent owners. There are patent law principles which may be applied to formal standards incorporating patents: equitable estoppel and implied licence. 3.2.2 Equitable estoppel The equitable estoppel can be used to regulate the duty of a patent holder to disclose patents covering standards. The equitable estoppel applies when the following elements are established: (1) ‘the patentee, through misleading conduct, leads the alleged infringer to reasonably infer that the patentee does not intend to enforce its patent against the alleged infringer’, (2) the alleged infringer relies on the conduct and (3) ‘due to its reliance, the alleged infringer will be materially prejudiced if the patentee is allowed to proceed with its claim.’ 122 The equitable estoppel applies not only to the case where the patent owner affirmatively misleads an SSO but also to the case where the patent owner remains silent despite a clear duty to speak. Thus, where a patent owner takes on the duty to disclose relevant patents which the IPR policies of an SSO 120 121 122

Ibid. 1914–15. Ibid. 1916–17. A.C. Aukerman Company v. R.L. Chaides Construction Co., 960 F.2d 1020, 1028 (1992).

48


impose on members, the violation of the duty may cause the equitable estoppel. It is not necessary to prove the intent of the patent owner to mislead the SSO in establishing the elements of the equitable estoppel and thus the estoppel may be applied to the duty to disclose regardless of the patent owner’s intent.123 In order that the claimant may allege the equitable estoppel, he has to prove his reliance on the patent owner’s conduct. It is not sufficient that the claimant benefits from non-enforcement of a patent; he must reasonably rely on the patent owner’s conduct or silence in determining his conducts. In addition, the enforcement of the patent must prejudice the claimant. Where all the elements of the equitable estoppel are established, the outcome of the equitable estoppel is to prevent the patent owner who violates a duty to disclose the relevant patents from enforcing the patents, as in Stambler v. Diebold.124 Stambler v. Diebold was an early case where the court applied the equitable estoppel and declined to enforce a patent that covered a standard. Leon Stambler, who was the owner of a patent concerning the card validation system activating automatic teller machines, participated in a committee of ANSI. Even though Stambler recognised that the standard was covered by his patent, he failed to inform the committee of this fact while working in the committee. After the standard became widespread in the industry, Stambler sought to enforce his patent right by filing a patent infringement suit against Diebold Inc. Diebold argued that it had invested considerable amounts of money in its production facilities, research and marketing because of the patentee’s misleading conduct and thus the infringement suit was barred by the equitable estoppel. The court endorsed Diebold’s argument and held that the owner had a duty to disclose his patent and that a failure to do so was affirmatively misleading. The court applied the equitable estoppel and ruled that the patent holder ‘could not remain silent while an entire industry implemented the proposed standard and then when the standards were adopted assert that his patent covered what manufacturers believed to be an open and available standard.’125 Thus the equitable estoppel may be applied to the case where the patentee participates in a standard setting process and does not disclose the relevant patents. The equitable estoppel may also be applied to the case where the patent owner affirmatively misleads an SSO by asserting that his patents will not be enforced, and then seeks to enforce them after they are adopted as standards. 123 Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’, 1918–19. 124 Stambler v. Diebold, 11 U.S.P.Q.2d 1709 (1988). 125 Ibid., 6.


49

3.2.3 Implied licence While the equitable estoppel may be used in situations where the patent owners breach the duty to disclose relevant patents, it cannot be applied to the case where patent owners disclose the relevant patents and agree to license on a reasonable and non-discriminatory basis.126 By disclosing the essential patents and agreeing to grant licences on reasonable and non-discriminatory terms, a patent owner demonstrates that his patents will be enforced in the form of licences. That is, the patent owner has never misled others to believe that his patent rights will not be exercised. Thus, the element of equitable estoppel cannot be established in such cases. This is still the case even though the patent owner refuses to grant a licence on reasonable and nondiscriminatory terms after the patented technology becomes the industry standard, because the patentee did not mislead others to reasonably infer that he had no intention to enforce the relevant patent against the alleged infringer. Instead, the breach of agreement to license on reasonable and nondiscriminatory terms may be regulated by the theory of implied licence. Where a patent owner licensed patent rights expressly or implicitly, received consideration for the licence, and then sought to impair the right granted, the implied licence may be established.127 In relation to standards, where the patent owner agrees to grant a licence of the patent covering standards on reasonable and non-discriminatory terms, a licence may be implied from the conduct of the patent owner even though there is no express agreement between the patent owner and users of the standards. Thus, if the patent owner seeks to enforce his patents by filing a patent infringement suit, the alleged infringer may rely on the implied licence theory to defend himself against the infringement charge. Indeed, the Federal Circuit granted an implied licence to a company adopting the standard technology in Wang Labs. v. Mitsubishi.128 Wang Labs. invented single in-line memory modules (SIMMs) and encouraged Mitsubishi to incorporate SIMMs in 256K modem chips. Wang sought to persuade the Joint Electron Device Engineering Council (JEDEC) to adopt SIMMs as a standard without disclosing relevant pending patent applications while JEDEC was considering the standard. After the SIMM standard had become widespread in the industry, Wang attempted to enforce its patent rights against Mitsubishi. Mitsubishi counterclaimed that, among other things, Wang’s conduct created an implied licence. The court held that Mitsubishi was entitled to an irrevocable,

126

Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’,

1924. 127

Wang Labs., Inc. v. Mitsubishi Elecs. Am., Inc., 103 F.3d 1571, 1578–82

(1997). 128

Ibid.

50


royalty-free implied licence under Wang’s patents because of their relationship for six years and Wang’s failure to disclose pending patent applications at the time of Wang’s lobbying for the SIMM standard. The implied licence theory provides legal remedies distinct from the legal measures obtainable from contract theory.129 While the remedy for breach of contract seeks to compensate the loss of the expected value caused by the breach, the implied licence theory allows the user of standards to obtain a licence. As such, the implied licence theory may be used in those cases where the patentee participates in standard setting processes and agrees to grant a licence for the relevant patents. However, contracts, the equitable estoppel and the implied licence theory are used in defensive ways against the patent infringement suit. There are more aggressive legal measures: the antitrust liability and fraud. 3.2.4 Antitrust liability The failure to disclose patents relevant to standards and a misrepresentation of willingness to license the patents on a reasonable and non-discriminatory basis may be condemned by antitrust liability. Indeed, in Dell, the FTC alleged that a firm failing to notify an SSO of the relevant patent violated the antitrust law.130 Dell was the first case where the antitrust authority imposed antitrust liability on a firm that failed to disclose patents essential to standards. Dell Computer Corp., which owned essential patents of the VL bus,131 failed to disclose the patents while participating in the standard setting process of the VL bus of VESA (Video Electronics Standards Association). According to the FTC complaint, a Dell representative allegedly certified that he knew of no intellectual property rights that the bus design would violate. Dell sought to enforce the patents after the standard was adopted, which led the FTC to allege that Dell violated the antitrust laws. This case ended up with a consent decree, calling for Dell not to enforce its patents covering the VL bus standard. The patent owner’s misrepresentation, whether it is from the failure to disclose relevant patents or from a false representation about a licence, may mislead an SSO to adopt a standard which is covered by his patents and which would otherwise be rejected. If the standard confers market power on the patent owner, the patent owner’s conduct may be condemned by antitrust laws.

129


1925. 130 131

In re Dell, No. 931-0097. VL bus is a 32-bit bus that provides a high-speed data path between the CPU and peripherals. See VESA Local Bus (PCMAG.COM, [cited 30.12 2008]); available from http://www.pcmag.com/encyclopedia_term/0,2542,t=VL-bus&i=54017,00.asp.


51

The most likely antitrust claim which may be used in the context of standards is an attempted monopolisation claim under section 2 of the Sherman Act.132 The elements of an attempted monopolisation claim are: (1) the relevant geographic and product market(s), (2) predatory or anticompetitive conduct, (3) specific intent to monopolise, and (4) a dangerous probability of successful monopolisation.133 In order to allege an antitrust violation under an attempted monopolisation claim against a patent owner who did not disclose essential patents during the standardisation process, the antitrust plaintiff must show that there is a causal link between the misrepresentation of the patent owner and the adoption of the standard in question.134 That is, the plaintiff has to prove that the standard in question would not have been adopted had it not been for the misrepresentation of the patent owner. Where the SSO would adopt the standard even without the misrepresentation because of the fact that there is no alternative or for any other reasonable reason, the misrepresentation of the patent owner may not provoke antitrust concerns. In addition, in order to establish the condition of a dangerous probability of successful monopolisation, the standard should be likely to dominate the market.135 Even though there is a high probability of monopolisation in the relevant market, the monopolisation should be caused not by the patent right but by the misrepresentation because the patent right may lawfully confer market power on the patent owner. Moreover, the patent owner can rebut the attempted monopolisation claim by showing that the market power could eventually be obtained through the de facto standard competition in the market: that is, without the adoption of the standard by the SSO.136 On top of that, the plaintiff must establish that the patent owner had a specific intent to monopolise the relevant market, which is difficult. Indeed, the consent decree of Dell was criticised by commentators because there were no allegations that Dell’s failure to disclose relevant patents was intentional.137 Thus it is difficult to establish the elements of an attempted monopolisation claim and therefore it seems that antitrust laws can be used in limited circumstances in relation to standards incorporating patents. 132


1928–9. 133

O’Donnell, ‘Unified Theory of Antitrust Counterclaims in Patent Litigation’,

22. 134


1931–2. 135 136 137

Ibid. Ibid. Donahey, ‘Terminal Railroad Revisited: Using the Essential Facilities Doctrine to Ensure Accessibility to Internet Software Standards’, 322–3.

52


3.2.5 Fraud There are two sorts of fraud: actual fraud and constructive fraud.138 Actual fraud needs an intentional deception whereas constructive fraud does not require intentional deception but can be established by ‘a breach of duty by one in a confidential or fiduciary relationship to another that induces justifiable reliance by the other to his or her prejudice’.139 It is not likely that the patent owner and a user of standards have a confidential or fiduciary relationship and thus constructive fraud is not relevant to standards incorporating patents. However, actual fraud may be constituted when the patent owner violates the duty to disclose the relevant patents. The elements of actual fraud are: (1) a false representation (2) of a material fact, (3) made with knowledge of that material fact (4) with the intent to induce reliance (5) where the other party takes action in justifiable reliance and (6) this results in damage.140 A false representation is constituted not only by an affirmative statement but also by concealment. In the context of formal standards, the failure to disclose patents relevant to standards and a misrepresentation of willingness to license the patents on a reasonable and non-discriminatory basis may constitute a false representation. Where the concealed fact leads the defrauded party to behave differently, it is considered material. Thus, materiality exists where the patent owner’s misrepresentation leads the SSO to adopt standards which would otherwise not be accepted as standards. As for the third element of the fraud claim, the defrauded party must prove that the patent owner actually knew that he had patents covering the standard.141 As far as the fourth element is concerned, the defrauded party has to prove that the patent owner’s misrepresentation was intentional to induce reliance by others.142 In order to establish the fifth element, the defrauded party must prove justifiable reliance on the false representations. The members of an SSO may presumptively have justifiable reliance because the SSO imposes on members the duty to disclose relevant patents.143 As for the sixth element, the

138 Peter David G. Sabido, ‘Defending against Patent Infringement Suits in Standard-Setting Organizations: Rambus Inc. v Infineon Technologies’, Federal Circuit Bar Journal 13 (2003–04): 641–2. 139 Ibid. 140 Ibid. 141 This may be a heavy burden of proof for the defrauded party. In the UK, the House of Lords held that where an alleged fraudster honestly believes that the statement is true, he is not exposed to an action in deceit by the statement. For more detail, see Derry v. Peek, 14 LR 337 (1889). 142 Sabido, ‘Defending against Patent Infringement Suits in Standard-Setting Organizations: Rambus Inc. v Infineon Technologies’, 644. 143 Ibid.


53

defrauded party must show that there was monetary or economic loss caused by the misrepresentation. If the patent owner is successful in his patent infringement suit against the defrauded party, the defrauded party may prove damage by showing the cost necessary to obtain a licence or to design around the patents.144 In some respects, the fraud theory provides effective remedies because the defrauded party may recover actual damage and it is less onerous to establish a fraud claim than an antitrust claim, which requires the definition of relevant markets and the existence of market power.145 However, the fraud theory also has limitations. Like the breach of contract claim, the fraud claim cannot be relied on by nonmembers. The defrauder should have some duty to the defrauded party in order for the defrauded party to claim fraud, but it is not likely that the patent owner will have a duty to nonmembers. 3.2.6 Review There are various legal principles which can be used to limit the enforcement of patent rights when the owners of the patents fail to perform their obligations to comply with the IPR policies of an SSO. These legal principles are fairly effective in preventing the patent owners from seeking undue benefits from their opportunistic behaviour.146 Nevertheless, even though the principles may work effectively in regulating the unfair conducts of patent owners, they do not resolve the problem of blocking patents. The patent owners always have options: they may not participate in SSOs or they may withdraw from SSOs which they once joined. Therefore, if a patent owner considers it disadvantageous to be a member of an SSO, he may choose not to participate in or withdraw from an SSO, and seek to enforce the patents in markets. In this case, the patents block the standardisation process, causing the SSO to find alternative technologies which may be inferior to the patented technologies. Blocking patents are likely to delay the standard setting process, stifle innovations, and entail additional social costs. Especially where there is no alternative to the patented technology, as is the case in the standardisation of GSM in Europe, the patent owners may debilitate an SSO. The owners of essential patents may choose a standard competition in markets rather than the coordinated standard work of an SSO, if they consider it advantageous. They may reject the standard setting of an SSO even in those cases where it is not likely that market processes would produce a de facto standard. This may entail the parallel existence of several 144 145

Ibid., 645. Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’,

1935–6. 146

Ibid., 1936–7.

54


incompatible technologies. Where two or more incompatible technologies exist in a network market, the positive network effects are not fully utilised and thus inefficiency arises. Even though the patent owner refuses to grant licences to an SSO, it is not likely that the patent owner’s refusal to license the essential patents will be condemned by a monopolisation claim or an attempted monopolisation claim, either because the relevant market does not exist at the early stage of standard setting work or because, if the relevant market exists, the patent owner may not have market power in the market. Moreover, the US courts have been reluctant to compel a firm, even if it is a monopolist, to deal with others by imposing antitrust liability. Furthermore, it is unlikely that the legal principles such as contract, equitable estoppel, implied licence and fraud will be used when the patent owner refuses to grant licences to an SSO. Therefore, it is difficult to resolve the problems caused by blocking patents, which may cause serious harm to standard setting efforts. All this implies that it is necessary to analyse the costs and benefits of patent systems in relation to standards and to examine whether it is necessary to reform or fine-tune patent systems.

4. Patents and standards in the EU 1.

PATENTS AND INFORMAL STANDARDS

1.1

Introduction

The legal issues regarding patent rights and standards for compatibility which arise when the patent owner of standards refuses to license the patent to others who need to access it demonstrate the conflict between competition laws and intellectual property rights. As is the case with US antitrust laws, the EU competition laws may be used in these issues. The EU competition laws are embodied in Articles 81 and 82 of the EC Treaty. Article 82 proscribes the abuse of a dominant position in the relevant market, which can be applied to cases of refusal to license by the patent owner. In order to examine the relationship between patent rights and the abuse of a dominant position, it is necessary to study each aspect of Article 82. Thus, in this chapter, the general principles of Article 82 as well as their application to patents covering standards are examined. 1.2

Relevant Competition Law (Article 82 of the EC Treaty)

Article 82 of the EC Treaty states that: Any abuse by one or more undertakings of a dominant position within the common market or in a substantial part of it shall be prohibited as incompatible with the common market in so far as it may affect trade between Member States. Such abuse may, in particular, consist in: (a) directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions; (b) limiting production, markets or technical development to the prejudice of consumers; (c) applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage; (d) making the conclusion of contracts subject to acceptance by the other parties of supplementary obligations which, by their nature or according to commercial usage, have no connection with the subject of such contracts.

55

56


1.3

The Principles of Article 82

1.3.1 Relevant market Article 82 is applied only to cases where an undertaking has a dominant position in the relevant market. Therefore, the first step in ascertaining the dominancy of an undertaking is to determine the relevant market.1 The definition of the relevant market is a critical factor in deciding whether a firm is in a dominant position and thus is one of the main issues in applying Article 82. The key factor in finding the relevant market is the interchangeability. As to this point, the ECJ held in Hoffman–La Roche that [t]he concept of the relevant market in fact implies that there can be effective competition between the products which form part of it and this presupposes that there is a sufficient degree of interchangeability between all the products forming part of the same market in so far as a specific use of such products is concerned.2

In order to find the relevant market, the relevant product market and the relevant geographical market should be defined.3 The relevant product market is determined according to the demand substitutability, supply substitutability and potential competition, but the demand substitutability is considered the most important factor.4 For example, where a consumer buys a product A in response to the rise in price of a product B, the product B is a substitute of the product A and both products are in the same product market.5 Thus, where a product is reasonably interchangeable with its substitute product, both products are included in the same product market. As for the relevant geographic market, the Commission provides guidance on the definition of the geographic market in the Notice on the definition of the relevant market, where it says, ‘[i]n short, the Commission will identify possible obstacles and barriers isolating companies located in a given area from the competitive pressure of companies located outside that area, so as to

1 Richard Whish, Competition Law, 5th edn (London: LexisNexis UK, 2003) Ch. 1; Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission, 4 C.M.L.R. 718, paras 173–84 (1995). 2 Case 85/76 Hoffmann–La Roche & Co. AG v. EC Commission, 3 C.M.L.R. 211, para. 28 (1979). 3 Case 27/76 United Brands Company and United Brands Continentaal BV v. EC Commission, 1 C.M.L.R. 429, paras 10–11 (1978). 4 Alison Jones and Brenda Sufrin, EC Competition Law, 3rd edn (Oxford: Oxford University Press, 2008) 60–84. 5 David A. Balto and Andrew M. Wolman, ‘Intellectual Property and Antitrust: General Principles’, IDEA: The Journal of Law and Technology 43 (2003).

Patents and standards in the EU

57

determine the precise degree of market interpenetration at national, European or global level.’6 Some factors like transportation costs, the location of factories and legal barriers, which hinder the movement of the relevant products, are important in defining the geographic market.7 In general, since the relevant market is found by an analysis as to whether the product at issue is substitutable, intellectual property rights are not considered significant in determining the relevant market.8 Intellectual property rights need to be considered in examining barriers to entry into a market.9 However, according to the ruling in IMS, the relevant upstream market may be hypothetical or potential, and thus there is always the market where (1) the input in an upstream market is essential for undertakings to operate on a downstream market and (2) there is an actual demand for the input on the part of undertakings that hope to enter the downstream market where the input was essential.10 That is, the court reasoned that a separate market for intellectual property rights could be defined even in those cases where the intellectual property holder did not license them and only used them as an input into the development of another product. This point is examined in more detail later in this chapter. 1.3.2 Dominant position in relevant markets Once the relevant product market and the relevant geographic market have been identified, the next step under Article 82 is to determine whether the relevant undertaking has a dominant position in the market. The ECJ defined the dominant position in United Brands, stating that [t]he dominant position referred to in this Article relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately its consumers.11

6 Commission Notice on the Definition of the Relevant Market for the Purpose of Community Competition Law, C372 Official Journal 5, para. 30 (1997). 7 Steven D. Anderman, EC Competition Law and Intellectual Property Rights (Oxford: Oxford University Press, 1998) 165. 8 Guy Tritton, Intellectual Property in Europe, 2nd edn (London: Sweet & Maxwell, 2002) 818. 9 Ibid. 10 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., 4 C.M.L.R. 28, AG55–AG59 and paras 37–47 (2004). 11 Case 27/76 United Brands Company and United Brands Continentaal BV v. EC Commission, para. 65.

58


This definition contains two elements, the ability to prevent effective competition and the ability to behave independently, but the latter is considered to be crucial.12 In order to determine whether an undertaking has a dominant position, many factors should be considered. Amongst them, as was held by the ECJ in Hoffmann–La Roche, very large market shares, which have been held for some time, can be important evidence of dominance.13 However, it is not clear what percentage of market share amounts to a very large market share. There are several cases that show the relationship between market shares and dominance. For instance, the ECJ held in AKZO that a market share of 50 per cent could be a very large market share, in the absence of exceptional circumstances.14 In United Brands, the ECJ held that a market share of 40 to 45 per cent did not permit the conclusion that United Brands automatically controlled the market but that market dominance must be determined having regard to the strength and number of the competitors.15 In this case, after considering other factors indicating dominance, the court found that United Brands had a dominant position.16 Even though market share is an important factor in determining dominance in the relevant market, it is necessary to consider other factors indicating dominance, such as superior technology, capital investment, economies of scale and legal monopoly.17 As for intellectual property rights, the courts have repeatedly confirmed that mere ownership of intellectual property rights does not necessarily confer a dominant position.18 However, intellectual property rights may be relevant to dominance because they can constitute a barrier to entry in the relevant market. Especially, in the field of spare parts, where the relevant market is defined narrowly enough for a single product to constitute a separate market,19 the mere holding of intellectual property rights which cover the spare parts may amount to dominance because other firms cannot produce the product. Furthermore, as to the exceptional circumstances, which are established in Magill and reaffirmed in IMS, the relevant upstream market may be hypothetical or 12 13 14

Whish, Competition Law Ch. 5. Case 85/76 Hoffmann–La Roche & Co. AG v. EC Commission, para. 41. Case C-62/86 AKZO Chemie BV v. EC Commission, E.C.R. I-3359, para. 60

(1991). 15 Case 27/76 United Brands Company and United Brands Continentaal BV v. EC Commission, paras 108–10. 16 Ibid., paras 108–29. 17 Jones and Sufrin, EC Competition Law 407–24. 18 Anderman, EC Competition Law and Intellectual Property Rights Ch. 12. 19 Case 22/78 Hugin Kassaregister AB and Hugin Cash Registers Limited v. EC Commission, 3 C.M.L.R. 345, paras 6–7 (1979).


59

potential.20 Because the market is hypothetical or potential, the conventional market share approach is not useful in determining whether an undertaking in an upstream market is dominant or not. This issue is examined in more detail later in this chapter. 1.3.3

Abuse

Abuse in general Having a dominant position is not a violation of Article 82 but an abuse of the dominance falls within the prohibition laid down by Article 82. Article 82 provides examples of abuse such as charging an unfair price, limiting production, discrimination and imposing irrelevant conditions in contracts, but these are not considered exhaustive examples. From an economic perspective, an abuse of a dominant position requires that the conduct by a dominant undertaking must prejudice consumers or competitors and has to be feasible only because of the dominance of the undertaking.21 The European Court of Justice provided the legal interpretation of an abuse of a dominant position in Michelin v. Commission, stating that: Article 86 [now Article 82] covers practices which are likely to affect the structure of a market where, as a direct result of the presence of the undertaking in question, competition has already been weakened and which, through recourse to methods different from those governing normal competition in products or services based on trader’s performance, have the effect of hindering the maintenance or development of the level of competition still existing on the market.22

That is, according to the court in Michelin, a dominant firm’s conduct can be found to be abusive when it involves two factors: practices having the effect of weakening competition and recourse to methods different from those governing normal competition based on the trader’s performance. Intellectual Property and Abuse A. Introduction Where dominance in the relevant market is due to intellectual property rights, the question of how the prohibition of abuse can be reconciled with 20 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., AG55–AG59 and paras 37–47. 21 Tritton, Intellectual Property in Europe 828. 22 Case 322/81 Nederlandsche Banden-Industrie Michelin N.V. v. EC Commission, 1 C.M.L.R. 282, para. 70 (1983).

60


the exclusivity of intellectual property rights arises. As to this problem, the ECJ and the CFI take the position that competition laws should not distinguish between the exercise of intellectual property rights and that of other economic power.23 According to this approach, owners of intellectual property rights in a dominant position are not allowed to extract the maximum benefit from the protected goods but must show that the exercise of intellectual property rights does not prevent or distort competition. B. Abuse and the exercise of intellectual property rights In Parke, Davis, the ECJ held that: Under Article 86 [now Article 82] of the Treaty it is prohibited, ‘in so far as trade between member-States is liable to be affected by it, for one or more undertakings to exploit in an improper manner a dominant position within the Common Market or within a substantial part of it’. For an act to be prohibited it is thus necessary to find the existence of three elements: the existence of a dominant position, an improper exploitation of it, and the possibility that trade between member-States may be affected by it. Although a patent confers on its holder a special protection within the framework of a State, it does not follow that the exercise of the rights so conferred implies the existence of the three elements mentioned. It could only do so if the utilisation of the patent could degenerate into an improper exploitation of the protection.24

Thus the ECJ held that the mere exercise of patent rights could not per se be an abuse of a dominant position and this principle has repeatedly been reaffirmed by the court.25 However, the court did not give absolute immunity to the exercise of intellectual property rights but held that, in some cases, the exercise of patents could constitute an abuse of a dominant position. This was confirmed in Hoffmann–La Roche v. Centrafarm where the court held that where the exercise of intellectual property rights was used as an instrument for the abuse of a dominant position, it could constitute an abuse.26 The ECJ has tried to refine in several cases the difference between the mere exercise of intellectual property rights and the abusive use. For the purpose of this book, the refusal to license by a dominant firm having intellectual prop-

23 24

Tritton, Intellectual Property in Europe 836. Case 24/67 Parke, Davis & Company v. Probel and Others, C.M.L.R. 47, para. 4 (1968). 25 Case 238/87 Volvo AB v. Erik Veng (UK) Ltd, 4 C.M.L.R. 122 (1989), Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission. 26 Case 102/77 Hoffmann–La Roche & Co. AG and Hoffmann–La Roche AG v. Centrafarm Vertriebsgesellschaft Pharmazeutischer Erzeugnisse mbH, 3 C.M.L.R. 217, para. 16 (1978).


61

erty rights covering standards is principally considered. Where an undertaking owns intellectual property rights in formal or informal standards, it might be argued that refusal to license such intellectual property rights is an abuse of a dominant position. These cases concern the exceptional circumstances which are discussed in the following sub-sections. C.

Exceptional circumstances

(1) Essential facilities doctrine versus exceptional circumstances Two legal principles may be applied to the refusal to deal. The essential facilities doctrine may be used with respect to tangible assets, whereas the principle of exceptional circumstances may be applied in relation to intangible property.27 The essential facilities doctrine has its origin in the United States and traces back to the Terminal Railroad Association case of 1912.28 In the US, the doctrine has been developed in cases involving the natural monopoly market, where a single firm can meet the market demand at the lowest price so that the market supports only one firm.29 In Europe, the doctrine has not been limited to the natural monopoly market but has been applied to various circumstances. The ECJ has not explicitly endorsed the doctrine, but in a number of cases concerning refusals to supply by a dominant firm the court implicitly utilised it.30 It is submitted that the principle of the doctrine has developed from the ECJ’s ruling of Commercial Solvents,31 where the court held that a refusal to supply by a firm having a dominant position in the production of a raw material could be abusive in those cases where the effect of the refusal would eliminate all competition in the downstream market.32 As to the refusal to deal in relation to intellectual property, European courts have established the exceptional circumstances under which a refusal to license by a dominant firm is condemned by competition laws. The following sub-sections describe the exceptional circumstances in detail.

27 Estelle Derclaye, ‘The IMS Health Decision and the Reconciliation of Copyright and Competition Law’, European Law Review 29 (2004). 28 Mark Furse, ‘The “Essential Facilities” Doctrine in Community Law’, European Competition Law Review 16, no. 8 (1995). 29 David Begg et al., Economics, 8th edn (London: The McGraw-Hill Companies, 2005) Ch. 18. 30 Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 427–8. 31 Whish, Competition Law 664–7. 32 Cases 6–7/73 Istituto Chemioterapico Italiano SpA and Commercial Solvents Corp. v. EC Commission, 1 C.M.L.R. 309 (1974).

62


(2)

Relevant cases

Volvo AB v. Erik Veng Ltd33 Facts – Volvo owned registered design rights in the UK for front wing panels of the Volvo Series 200. Veng imported, from Italy and Denmark, front wing panels which were manufactured without Volvo’s authorisation, and sold them in the UK. Volvo refused to license the design rights although Veng was willing to pay a reasonable royalty. Volvo sought to prevent Veng from importing and selling the front wing panels by instituting proceedings against Veng for infringement of its design rights. Veng argued, among other things, that Volvo’s refusal to license constituted an abuse of a dominant position. The High Court of the UK referred to the ECJ, inter alia, the question as to whether the refusal to license others to supply such body panels, even in return for a reasonable royalty, was prima facie an abuse of a dominant position. The ECJ’s judgment – The court held that the proprietor of a protected design was not obliged to grant a licence to third parties, even in return for a reasonable royalty. It concluded that a refusal to grant a licence was not itself an abuse of a dominant position because the freedom to refuse to license intellectual property constituted the very subject matter of exclusive rights. According to the ECJ in this case, however, the exercise of an exclusive right might have violated Article 82 if it involved additional abusive conduct such as the arbitrary refusal to supply replacement parts to independent repairers, the fixing of prices at an unfair level, or a decision to cease production of spare parts for a particular model when many cars of that model were still in circulation. CICRA v. Renault34 Facts – A trade association comprising a number of Italian undertakings which manufactured and sold spare parts for the bodywork of motor vehicles attempted to invalidate Renault’s design rights for spare parts or alternatively sought a finding that the manufacture and marketing of non-original spare parts did not amount to an offence under the national legislation on unfair competition. The Italian Court referred two questions to the ECJ, the second of which was whether the registration of design rights in respect of car bodywork parts and the exercise of the resultant exclusive rights constituted an abuse of a dominant position.

33 34

Case 238/87 Volvo AB v. Erik Veng (UK) Ltd. Case 53/87 Consorzio Italiano della Componentistica di Ricambio per Autoveicoli and Maxicar v. Regie Nationale des Usines Renault, 4 C.M.L.R. 265 (1990).


63

The ECJ’s judgment – The ECJ reaffirmed the conclusion of Volvo, holding that the mere fact of obtaining protective rights could not be considered an abuse of a dominant position but that the exercise of the exclusive right could be an abuse in those cases where it involved additional abusive conduct set out in Volvo. The court also held that Renault’s higher pricing did not necessarily constitute an abuse, since the owner of design rights might lawfully call for a return on the amounts invested to perfect the protected design. Radio Telefis Eireann v. Commission (Magill)35 Facts – The British Broadcasting Corporation (BBC), Independent Television Publications Ltd (ITP) and Radio Telefis Eireann (RTE), the three television broadcasters in the UK and Ireland, distributed the weekly programme listings, which were protected as literary works and compilations under the UK and Irish copyright laws, to third parties who were permitted to issue them without charge. Unlike the other EC countries, Ireland and Northern Ireland had no comprehensive weekly TV guide which contained all weekly programme listings for channels which most television viewers could watch, because these broadcasters refused to license others to produce a weekly listing in advance. Third parties like newspapers, therefore, could only publish the TV programme listings for a day or for both days of a weekend at a time. In 1986, Magill TV Guide Ltd, an Irish publisher seeking to produce a comprehensive TV guide containing weekly programme listings for all channels, complained to the Commission that such refusal to license constituted an abuse of a dominant position. In its decision of 1988, the Commission held that the refusal to license others constituted an infringement of Article 82 and ordered the broadcasters to make their programme listings available to third parties on a non-discriminatory basis and, if desired, against payment of a reasonable licence fee. ITP, BBC and RTE sought the annulment of the Commission’s decision. The ECJ’s judgment – As for the existence of a dominant position, the ECJ held that ‘the mere ownership of an intellectual property right cannot confer’ a dominant position.36 However, the court emphasised that the broadcasters enjoyed a de facto monopoly over the information such that they were in a position to prevent effective competition on the relevant market. The court, therefore, concluded that the companies occupied a dominant position.37 As regards an abuse, the court held that ‘the exclusive right of reproduction forms part of the author’s rights, so that refusal to grant a licence, even if it is 35 For more detail, see Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission. 36 Ibid., para. 46. 37 Ibid., para. 47.

64


the act of an undertaking holding a dominant position, cannot in itself constitute abuse of a dominant position’.38 However, the court established exceptional circumstances under which a refusal to license by a dominant firm falls within the prohibition laid down by Article 82. The exceptional circumstances are: firstly, there is no actual or potential substitute for the product protected by copyrights and thus the refusal to license copyrights prevents the emergence of a new product for which there is a constant consumer demand; secondly, there is no objective justification for the refusal; and thirdly, the copyright holders, by denying access to the basic information, have reserved to themselves the secondary market by excluding all competition on that market.39 However, the Court did not make it clear whether the conditions of the exceptional circumstances are cumulative or alternative, which gave rise to an issue in IMS.40 Tierce Ladbroke SA v. Commission41 Facts – Tierce Ladbroke was a bookmaker in Belgium on horse races of foreign countries. PMU (Pari Mutuel Urbain) had exclusive rights to take bets abroad on French horse races organised by authorised sociétés de courses. PMI (Pari Mutuel International), whose majority shareholder was PMU, was assigned by PMU the right to market television pictures and sound commentaries on French horse races for Germany and Austria. DSV (Deutscher Sportverlag Kurt Stoof GmbH & Co.) was granted the exclusive right to use televised pictures and sound commentaries of French races in Germany and in Austria. Tierce Ladbroke asked the sociétés de courses, PMU, PMI and DSV to provide it with pictures and sound commentaries of French races, which was refused. In reaction, Tierce Ladbroke lodged a complaint with the Commission, arguing, inter alia, that the refusal to supply the French sound and pictures to Ladbroke was an abuse of a dominant position which could not be justified. 42 The CFI’s judgment – The CFI referred to Magill, holding that the refusal to supply could not be condemned by Article 82 unless the product or service concerned was either essential for the business in question or was a new product which might not appear despite a constant demand by consumers.43 Thus

38 39 40

Ibid., para. 49. Ibid., paras 48–58. See Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH &

Co. KG. 41 42 43

Tierce Ladbroke SA v. EC Commission ECR-II 923 (1997). Ibid., paras 13–14. Ibid., para. 131.


65

the CFI added a new condition (indispensability) for the exceptional circumstances and took the position that the two conditions, prevention of the emergence of a new product and indispensability, were alternative rather than cumulative. On the one hand, the CFI contributed to the establishment of the exceptional circumstances by supplementing a new condition, indispensability. On the other hand, the Court made the application of the exceptional circumstances muddled by holding that the conditions were alternative. As noted, whether the conditions of the exceptional circumstances are cumulative or alternative is an issue in IMS. Oscar Bronner v. Mediaprint44 Facts – Mediaprint, a publisher of two Austrian newspapers with a large market share, refused to grant access to its home-delivery distribution system to its competitor Oscar Bronner, a publisher of a daily newspaper. Bronner complained that Mediaprint was obliged to grant access to its distribution system, because it was required for a dominant firm to grant access to competitors in the downstream market unless the refusal could be objectively justified, and because it was not economically feasible to establish its own distribution system. The Austrian court referred to the ECJ, inter alia, the question whether the refusal by a press undertaking in a dominant position, operating the only nationwide newspaper home-delivery scheme, to allow the rival newspaper company, which by reason of its small circulation is unable either alone or in cooperation with other publishers to set up and operate its own home-delivery scheme in economically reasonable conditions, to have access to that scheme for appropriate remuneration constitutes the abuse of a dominant position. Opinion of Advocate General – In this case, Advocate General Jacobs comprehensively analysed the essential facilities doctrine. He noted a number of general points: First, it is apparent that the right to choose one’s trading partners and freely to dispose of one’s property are generally recognised principles in the laws of the Member States … Secondly, … if access to a production, purchasing or distribution facility were allowed too easily, there would be no incentive for a competitor to develop competing facilities. Thus while competition was increased in the short term, it would be reduced in the long term. Moreover, the incentive for a dominant undertaking to invest in efficient facilities would be reduced if its competitors were, upon request, able to share the benefits. Thus the mere fact that by retaining a facility for its own

44 Case C-7/97 Oscar Bronner GmbH Co KG v. Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG, 4 C.M.L.R. 112 (1998).

66

Patents and industry standards use a dominant undertaking retains an advantage over a competitor cannot justify requiring access to it. Thirdly, in assessing this issue it is important not to lose sight of the fact that the primary purpose of Article 82 is to prevent distortion of competition – and in particular to safeguard the interests of consumers – rather than to protect the position of particular competitors. It may therefore, for example, be unsatisfactory, in a case in which a competitor demands access to a raw material in order to be able to compete with the dominant undertaking on a downstream market in a final product, to focus solely on the latter’s market power on the upstream market and conclude that its conduct in reserving to itself the downstream market is automatically an abuse.45

He added that refusal of access might eliminate or substantially reduce competition in those cases where ‘access to a facility is a precondition for competition on a related market for goods or services for which there is a limited degree of interchangeability’.46 He also argued that intervention by applying the essential facilities doctrine or as a response to the refusal to supply goods or services, ‘can be justified in terms of competition policy only in cases in which the dominant undertaking has a genuine stranglehold on the related market’ and ‘it is not sufficient that the undertaking’s control over a facility should give it a competitive advantage’.47 According to Advocate General Jacobs, whether the cost of duplicating a facility alone constituted an insuperable barrier to entry is an objective test, that is, ‘in order for refusal of access to amount to an abuse, it must be extremely difficult not merely for the undertaking demanding access but for any other undertaking to compete’.48 Against these reasonings, he found that the distribution system of Mediaprint did not qualify as an essential facility. The ECJ’s judgment – The ECJ endorsed Advocate General Jacobs’ opinion. Without citing the essential facilities doctrine, the ECJ held that Mediaprint’s refusal to grant access to its home-delivery system did not amount to an abuse of a dominant position. It held that it was undisputed that other methods of distributing daily newspapers, such as mailing, retail shops, and kiosks, even though these alternatives were less advantageous for the distribution of certain newspapers, existed and were used by daily newspaper companies. In addition, the ECJ noted that there were no technical, legal, or economic obstacles that would make it impossible or even unreasonably difficult for any other publisher, alone or in partnership with other publishers, to establish its own nationwide home-delivery scheme.

45 46 47 48

Ibid., paras 55–58. Ibid., para. 61. Ibid., para. 65. Ibid., para. 66.


67

It is accepted that the ECJ’s ruling in this case restricted the broad application of the essential facilities doctrine to refusal to deal cases.49 IMS Health Inc. v. Commission (IMS)50 Facts – Intercontinental Marketing Services Health (IMS Health), in cooperation with a working group composed of its own employees and representatives from the pharmaceutical industry, developed a brick structure for dividing the entire German market into 1860 pharmacy groups, known as bricks. Each of these contained postcode areas identified by a seven-digit number, where the first five numbers represented a particular metropolitan area and the last two numbers represented the individual ‘brick’ within that area. The brick structure was designed to contain a comparable number of pharmacies in the corresponding geographic area and was protected by the copyright laws of Germany. Pharma Intranet Information AG (PII), established by a former general manager of IMS Health, and AzyX, a smaller Belgian company, entered the market in 1999. In August 2000, some of PII’s assets were acquired by National Data Corporation (NDC). In late 2000, IMS Health obtained interim injunctions from the Frankfurt Regional Court against all these companies, preventing them from using the 1860 brick structure or any structure derived from it. The three companies appealed and, in addition, NDC complained to the Commission, seeking a compulsory licensing. In the course of hearing the main case against NDC, the Frankfurt Regional Court referred to the ECJ, inter alia, the question whether the refusal to license its copyright in the 1860 brick structure would constitute an abuse of a dominant position under Article 82. In 2002, the Frankfurt Higher Regional Court overturned the lower court’s decision in relation to copyright in the main action against PII. The court held that because the copyright was not owned by IMS Health but by the members of the working group, IMS Health could not enforce that copyright. The court also held that refusal to grant a licence to copy the structure could not violate Article 82 and refused to refer this question to the ECJ. The Commission’s decision51 – As for the complaint submitted by NDC, the Commission found that the refusal to license what had become an industry standard amounted to an abuse of a dominant position. The Commission held that the exceptional circumstances had been established because the brick 49

Valentine Korah, ‘The Interface between Intellectual Property and Antitrust: The European Experience’, Antitrust Law Journal 69 (2002): 818–20. 50 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG; Case T184/01 R II IMS Health Inc. v. EC Commission, 4 C.M.L.R. 2 (2002). 51 Commission Decision No. 2002/165/EC, L 59 Official Journal 18 (2002).

68


structure became the de facto industry standard and the refusal by IMS Health, which had no objective justification, had the effect of excluding all competition in the market. The Commission took the position that it did not have to find the prevention of the emergence of a new product in order to establish the exceptional circumstances set up in Magill.52 The Commission imposed the interim measures which required IMS Health to license the 1860 brick structure to NDC and AzyX on a non-discriminatory basis. Order of the President of the CFI53 – IMS Health appealed and asked the CFI to suspend the interim measure. The president of the CFI ruled in favour of IMS Health. In his order, he reviewed Magill and noted that while the exceptional circumstances established in Magill required that the products of the copyright owners and those of the party requesting a licence belonged to distinct markets, the parties in this case wished to offer services ‘on the same market and to the same potential clients and differing only as to detail from the services’ offered by a dominant firm.54 He also found that the Commission’s decision appeared to be a ‘non-cumulative interpretation of the conditions’ concerning exceptional circumstances in Magill.55 He suspended the Commission’s decision, holding that it would have to be determined by the court in its main judgment whether the concept of exceptional circumstances required the prevention of the appearance of a new product or service for which there was potential consumer demand and whether the market of the party should be separate from that on which the copyright owner was dominant. After this ruling, the Commission withdrew its decision. NDC appealed to the president of the ECJ, but in vain. The ECJ’s judgment for the question56 – As noted, the Frankfurt regional court referred to the ECJ, inter alia, the question whether the refusal to license the copyright in the 1860 brick structure would constitute an abuse of a dominant position. The court held that there should be no obligation to license intellectual property rights unless ‘two different stages of production’ could be identified and ‘they are interconnected, the upstream product was indispensable in as much as for supply of the downstream product’.57 The court reasoned that a separate market for intellectual property rights could be defined even in those cases where the intellectual property holder did

52 53 54 55 56

Ibid., para. 67. For more detail, see Case T184/01 R II IMS Health Inc. v. EC Commission. Ibid., para. 101. Ibid., para. 100. For more detail, see Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG. 57 Ibid., para. 45.


69

not license them and only used them as an input into the development of another product. According to the court, it is sufficient that there is a ‘potential market or even hypothetical market’ which can be identified.58 As for the emergence of a new product, the court followed the wording of Magill, holding that for a refusal to license to be abusive ‘the undertaking which requested the licence does not intend to limit itself essentially to duplicating the goods or services already offered on the secondary market by the owner of the copyright, but intends to produce new goods or services not offered by the owner of the right and for which there is a potential consumer demand’.59 Microsoft v. Commission60 The most recent and famous case regarding the exceptional circumstances is the Microsoft case ruled by the CFI. Two issues were argued by both Microsoft and the EC Commission: (i) Microsoft’s refusal to disclose and grant licences of interoperability information of the work group server operating systems and (ii) tying of the Windows client PC operating system and Windows Media Player.61 Of the two issues, it is the refusal by Microsoft to disclose and authorise the use of interoperability information that is concerned with the exceptional circumstances established by Magill and IMS. Thus, only the first issue of the case is studied here. Facts62 – On 15 September 1998, a vice-president of Sun Microsystems sent a letter to Microsoft, requesting Microsoft to supply Sun with the complete interoperability information of Windows operating systems. However, Microsoft refused to provide any more information about the interoperability than is already disclosed in public. In reaction to the refusal by Microsoft, Sun complained to the EC Commission, alleging that Microsoft violated Article 82 of the EC Treaty by refusing to supply the interoperability information on work group server operating systems that is needed to achieve the interoperability with Microsoft’s PC operating systems. The Commission began investigating the case and, after sending three statements of objections, adopted Decision 2007/53/EC, called ‘the contested decision’ in the CFI’s ruling. The Commission’s decision63 – The Commission made a decision that Microsoft violated Article 82 of the EC Treaty by abusing its dominant position

58 59 60 61 62 63

Ibid., para. 44. Ibid., para. 49. T-201/04 Microsoft Corp. v. EC Commission, 5 C.M.L.R. 11 (2007). Ibid., paras 36–50. Ibid., paras 1–20. Ibid., paras 21–50.

70


in the client PC operating systems market. The Commission defined three product markets, the markets for client PC operating systems, work group server operating systems and streaming media players, and found that the geographic market for the three product markets had a worldwide dimension. The Commission also found that Microsoft had a dominant position in two of the three markets, the client PC operating systems market and the work group server operating systems market.64 The Commission found that Microsoft abused a dominant position in the client PC operating systems market by refusing to supply interoperability information, which was defined by the Commission as the complete and accurate specifications for all the protocols [implemented] in Windows work group server operating systems and … used by Windows work group servers to deliver file and print services and group and user administrative services, including the Windows domain controller services, Active Directory services and group Policy services to Windows work group networks (Article 1(1) of the contested decision).65

The Commission applied the exceptional circumstances and concluded that Microsoft’s refusal caused a risk of elimination of competition on the market for work group server operating systems because of the indispensability of the interoperability information, stifled technical advancement to the prejudice of consumers and was not objectively justified.66 In reaction to the Commission’s decision, Microsoft appealed to the CFI, arguing, inter alia, that the exceptional circumstances were not present in this case. 67 The CFI’s judgment68 – The Court of First Instance endorsed the arguments of the Commission, ruling that Microsoft infringed Article 82. The CFI examined the case and decided whether the exceptional circumstances were established in this case. As far as the indispensability is concerned, the CFI examined Microsoft’s arguments that the Commission’s decision was erroneous due to an error of law and an error of fact. In relation to the error of law, Microsoft argued that the Commission falsely assumed that the competing server operating systems had to be able to interoperate with Windows client PC and server operating

64 It is the markets for client PC operating systems and work group server operating systems that are relevant in relation to the exceptional circumstances. 65 T-201/04 Microsoft Corp. v. EC Commission, para. 37. 66 Case COMP/C-3/37.792 Microsoft, 32 Official Journal 23, paras 585–778 (2004). 67 T-201/04 Microsoft Corp. v. EC Commission, paras 291–300. 68 Ibid.


71

systems ‘in exactly same way that Windows server operating systems [did]’.69 Microsoft contended that the degree of interoperability, adopted in the contested decision, required ‘virtual identity’70 between Windows server operating systems and non-Windows server operating systems, which was improper. In addition, Microsoft argued in relation to error of fact that five methods could be used to achieve interoperability between non-Windows server operating systems and Windows client PC and server operating systems, which was evidenced by the fact that non-Windows server operating systems were competing with Microsoft in the market without the interoperability information at issue.71 However, the CFI rejected Microsoft’s arguments. As regards the alleged error of law, the Court upheld the Commission’s arguments that the degree of interoperability required by the contested decision was necessary for the competing work group server operating systems to ‘remain viably on the market’.72 The Court analysed various data such as experts’ opinion, market surveys and marketing documents, and reached this conclusion.73 The Court also rejected Microsoft’s arguments that the Commission intended to have competitors develop work group server operating systems which were virtually identical to Microsoft’s.74 Concerning the alleged error of fact, the CFI also denied Microsoft’s arguments. The Court held that customers considered interoperability important in choosing the work group server operating systems, that the interoperability issue appeared amongst Microsoft’s competitors with the introduction of Windows 2000 operating systems, that the growth of competing work group server operating systems was only marginal, and that the suggested five methods could not achieve ‘the high degree of interoperability’75 which was necessary in this case.76 Thus the CFI concluded that the interoperability information at issue was indispensible. As for elimination of competition, the Court concluded that the refusal at issue risked eliminating competition on the relevant market.77 In reaching this conclusion, the CFI first examined the Commission’s market definition and endorsed it, identifying a work group server operating systems market as a

69 70 71 72 73 74 75 76 77

Ibid., paras 337–9. Ibid., para. 340. Ibid., paras 343–7. Ibid., para. 376. Ibid., paras 371–422. Ibid., para. 375. Ibid., para. 435. Ibid., paras 423–36. Ibid., para. 620.

72


relevant second product market.78 Secondly, the Court considered the method used by the Commission in finding the market shares of Microsoft in the work group server operating systems market and held that the method adopted by the Commission in calculating the market shares was not manifestly incorrect.79 In addition, the Court held that since it was Microsoft’s dominant position on the client PC operating systems market rather than on the work group server operating systems market that was at issue in relation to abusive conducts, even if there were some mistakes in calculating market shares it could not be concluded that the Commission was wrong in condemning the abusive conduct of Microsoft.80 Thirdly, the CFI rejected Microsoft’s arguments that the Commission failed to apply the strict criterion established by case law, namely that there should be a high probability that the conduct at issue would result in the elimination of all competition.81 The Court held that the Commission was right in applying Article 82 of the EC Treaty before all the competition on the relevant market was eliminated since the elimination of competition was likely to be irreversible due to the strong network effects.82 Fourthly, the CFI assessed the market data and the competition on the market, concluding that the refusal in question caused a risk that competition on the work group server operating systems market would be eliminated in this case.83 As to the new product, the CFI also upheld the Commission’s arguments. Most of all, the Court emphasised that the condition of preventing the emergence of a new product had to be considered under Article 82(b) of the EC Treaty, which condemned the abuse of ‘limiting production, markets or technical development to the prejudice of consumers’.84 That is, the Court did not confine the condition to the case of the prevention of a new product, but interpreted it broadly, including the limitation of markets and of technical development as well.85 Holding that the Commission’s finding was not manifestly incorrect, the CFI supported the Commission’s decision. In the first place, the CFI held that without interoperability consumers were locked into Windows work group server operating systems even though consumers considered that non-Windows work group server operating systems were superior to Windows

78 79 80 81 82 83 84 85

Ibid., paras 480–532. Ibid., paras 533–57. Ibid., para. 559. Ibid., paras 560–64. Ibid., para. 562. Ibid., paras. 565–620. Ibid., para. 643. Ibid., para. 647.


73

ones with respect to important factors such as reliability and security.86 In the second place, the Court held that Microsoft’s refusal to supply interoperability information discouraged competitors from innovating work group server operating systems, to the prejudice of consumers.87 In the third place, the CFI rejected Microsoft’s arguments that its refusal did no harm to customers, ruling that consumers considered non-Windows work group server operating systems to be better than Microsoft’s and that Microsoft harmed the effective market structure by being a dominant firm on that market.88 As far as objective justification is concerned, the Court rejected Microsoft’s contention that its refusal was justified by the fact that the technology at issue was protected by intellectual property rights.89 The CFI held that, judging from Magill and IMS, refusal to grant a licence could not be justified by the mere fact that the technology concerned was covered by intellectual property rights and ruled that in those cases where exceptional circumstances were established, the refusal could be abusive.90 The Court also held that it was Microsoft rather than the Commission that had to justify its conduct and that Microsoft failed to prove that the compelled disclosure of the interoperability information would significantly reduce the incentive to innovate.91 Thus the CFI ruled that the exceptional circumstances were established in this case and endorsed the Commission’s decision, which is given as follows. (a) Microsoft … shall, within 120 days of the date of notification of [the contested decision], make the interoperability information available to any undertaking having an interest in developing and distributing work group server operating system products and shall, on reasonable and non-discriminatory terms, allow the use of the interoperability information by such undertakings for the purpose of developing and distributing work group server operating system products.92

(3) Review According to the cases, the refusal to grant a licence, which is the straightforward exercise of intellectual property rights, cannot in itself constitute an abusive conduct. However, under exceptional circumstances, it can amount to an abuse. The exceptional circumstances provoked hot debate and, especially, it was argued that the owner of a basic patent might be prevented from enjoying its exclusivity by the holder of an improvement patent because the latter

86 87 88 89 90 91 92

Ibid., paras 650–52. Ibid., para. 653. Ibid., paras 660–65. Ibid., paras 688–9. Ibid., paras 690–91. Ibid., para. 697. Ibid., para. 48.

74


might be able routinely to require the former to grant a licence.93 There are uncertainties in interpreting the exceptional circumstances because the meaning of new products or services is not clear and what constitutes the objective justification is ambiguous.94 It is considered that in its ruling in Oscar Bronner the ECJ narrowed the scope of the exceptional circumstances of Magill by emphasising the objective indispensability.95 It has to be noted that, as is confirmed in IMS, in order for the exercise of intellectual property rights to be abusive, the exceptional circumstances have to be established in a cumulative way. Especially, the refusal has to block the appearance of a new product for which there is potential customer demand. Conceptually this condition is reasonable because intellectual property holders, who have exclusive rights over the protected subject matters, do not have the obligation to create competition in the market where the protected products are concerned by granting licences to those who merely seek to duplicate the protected product. It has to be only in those cases where the licensee plans to provide products different from the ones provided by the intellectual property holders that the compulsory licences need to be considered. It is argued that this strict application of the exceptional circumstances to intellectual property cases is necessary to realise the proper role of both intellectual property laws and competition laws.96 Nevertheless, it seems that Microsoft expands the concept of ‘a new product’ by including therein not only a product but also markets and technical development, endorsing the Commission’s decision that Microsoft has to ‘make the interoperability information available to any undertaking having an interest in developing and distributing work group server operating system products’.97 This ruling is controversial. In the first place, the ruling seems to adopt the liability rules rather than property rules in those cases where the intellectual property holder is in a monopoly position. This approach is right in the first place but some more conditions have to be followed. It is generally accepted that one of the main disadvantages of liability rules is that they may reduce incentives to inno93

Korah, ‘The Interface between Intellectual Property and Antitrust: The European Experience’, 811. 94 Christopher Stothers, ‘IMS Health and Its Implications for Compulsory Licensing in Europe’, European Intellectual Property Review 26, no. 10 (2004). 95 Mark D. Powell, ‘Competition Law and Innovation: The Interface between Competition Law and Intellectual Property’, Practising Law Institute 708 (2002). 96 Alessandra Narciso and Paul L.C. Torremans, ‘IMS Health or the Question Whether Copyright Still Deserves a Specific Approach in a Market Economy?’, in Developments in the Economics of Copyright, ed. Lisa N. Takeyama, Wendy J. Gordon and Ruth Towse (Cheltenham: Edward Elgar Publishing, 2005). 97 T-201/04 Microsoft Corp. v. EC Commission, para. 48.


75

vate.98 However, the ruling fails to consider this. The ruling seems to be based merely on the unqualified assumption that any undertaking intending to develop a work group server operating system product by using the interoperability information of Microsoft will contribute to technical advancement and that liability rules do not reduce the incentives for the intellectual property owner to innovate in this case. The assumption is not necessarily true. The product made by a third party may create far less value than that created by the interoperability information. That is, the interoperability information itself may be more valuable than the product made by a third party which includes the interoperability information. In this case, the value created by the third party’s product mainly arises from the interoperability information and thus the third party may obtain benefits not from his own efforts but from Microsoft’s R&D efforts. This would cause the free-rider problem, the very problem that intellectual property systems seek to prevent. Instead, as is argued later, legal entitlement has to be allocated according to the value created by each inventor in order to strike the right balance between the original inventor and the follow-on inventors.99 It is reasonable that stronger legal protection has to be provided to the economic entity that created more value. In addition, in order for the intellectual property systems to work properly, whether the holder of intellectual property rights has recouped the R&D expenditure has to be considered in restricting the exclusive rights of the intellectual property. Where the holder has made profits enough to cover the R&D costs, restricting the exclusive right of the right holder may be justified since the intellectual property right may be considered to have completed its mission. However, where the right owner has not recovered the R&D costs, limiting exclusive rights would chill the incentive to innovate. Thus, in order to guarantee that intellectual property systems are not debilitated, care has to be taken in restricting intellectual property rights, which the Microsoft Court failed to do. In short, the approach of the Microsoft Court is too much biased towards the follow-on inventors without considering the negative impact on the incentives which encourage the frontier inventors to innovate.

98

Robert P. Merges, ‘Contracting into Liability Rules: Intellectual Property Rights and Collective Rights Organizations’, California Law Review 84 (1996): 1303–8; Nicos L. Tsilas, The Perils of Imposing Compulsory IP Licensing to Achieve Interoperability (cited 24 August 2006), available from http://www.metrocorpcounsel. com/current.php?artType=view&artMonth=October&artYear=2005&EntryNo=3739. 99 For more detail, see Chapter 5, Section 2.3. Also see Mark A. Lemley, ‘The Economics of Improvement in Intellectual Property Law’, Texas Law Review 75 (1997).

76


On top of that, under the TRIPS Agreement, authorisation of compulsory licences has to ‘be considered on its individual merits’,100 but the Court fails to consider compulsory licences on individual merits. Concerning this issue, the CFI held that, according to the case law, the Court need not consider the TRIPS Agreement in examining the legality of the Community institutions’ measures101 and that even in those cases where Microsoft had to grant compulsory licences of its patents it would be able to negotiate the terms and conditions of the licences on individual merits.102 However, the Court failed to recognise that it would be impossible to decide whether or not to grant a compulsory licence on a case-by-case basis. Admittedly, it may be possible to implement the terms of the contract on individual merits but that the standard owner has to grant a licence to any undertaking is not the case where the compulsory licence is ‘considered on its individual merits’.103 Where the Community Court is not obliged to follow the TRIPS Agreement, the CFI’s ruling in Microsoft may not be problematic, but the decision is not compatible with the international framework of intellectual property law. Thus, in applying the logic of the case to the general cases concerning standards covered by patents, this incompatibility of Microsoft with the international framework of intellectual property law has to be noted. Many commentators criticise the Commission’s decision and the CFI’s ruling in that the decisions are not consistent with precedents like Oscar Bronner.104 It is argued that while Oscar Bronner reduced the scope of the exceptional circumstances by interpreting indispensability in strict ways, Microsoft falsely expanded the scope of the doctrine by replacing indispensability with viability.105 Indeed, the Microsoft Court upheld the Commission’s arguments that the degree of interoperability required by the contested decision was necessary for the competing work group server operating systems to

100

Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 31 (a). 101 T-201/04 Microsoft Corp. v. EC Commission, paras 801–4. 102 Ibid., paras 805–11. 103 Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 31 (a). 104 Ian S. Forrester, ‘Article 82: Remedies in Search of Theories?’, Fordham International Law Journal 28 (2005); Daryl Lim, ‘Copyright under Seige: An Economic Analysis of the Essential Facilities Doctrine and the Compulsory Licensing of Copyrighted Works’, Albany Law Journal of Science and Technology 17 (2007); Kathryn McMahon, ‘Interoperability: “Indispensibility” and “Special Responsibility” in High Technology Markets’, Tulane Journal of Technology and Intellectual Property 9 (2007). 105 McMahon, ‘Interoperability: “Indispensibility” and “Special Responsibility” in High Technology Markets’, 135–9.


77

‘remain viably on the market’. 106 It seems that the viability held in Microsoft is far less strict than the indispensability ruled in Oscar Bronner. To sum up, the ruling of Microsoft has some problems and thus care has to be taken in applying the CFI’s decision to the general cases where the exceptional circumstances may be used.107 The exceptional circumstances, which were established by the ruling of Magill and reaffirmed by IMS and Microsoft, may be applied to cases where a dominant undertaking owns patent rights which cover some parts of standards which third parties require access to in order to enter the downstream markets for products compatible with the standards. For instance, in network industries such as computers and telecommunication, manufacturers of complementary products need to use the compatibility standards of the networks in order to enter the network markets. This issue is analysed in more detail later in this chapter.108 1.3.4 Additional conditions for Article 82 Article 82 has additional requirements that the abuse needs to affect trade between Member States and that the abuse should be within the common market or in a substantial part of it. As regards the effect on trade between Member States, the ECJ held that it was not necessary that the conduct should in fact have substantially affected trade between Member States but that it was sufficient to establish that the conduct was capable of having such an effect.109 Article 82 also suggests that more than one undertaking can abuse a dominant position. For the purpose of this book, the conditions unrelated to intellectual property are not examined in detail. 1.3.5 The relationship between Article 81(3) and Article 82 Article 81(3) provides exemption from the application of Article 81(1) in those cases where an agreement, a decision or a concerted practice contributes to the improvement of the production or distribution of goods or to the advancement of technical or economic progress. In Tetra Pak Rausing,110 the CFI examined 106 107

T-201/04 Microsoft Corp. v. EC Commission, para. 376. As is argued by some commentators, competition laws may be used to accomplish what at any time seems to be fair and right in business practice. See John N. Adams et al., Franchising, 5th edn (Tottel Publishing, 2006) 45. In this sense, it may be argued that the Microsoft case is properly held in a given circumstance. 108 See Section 3. 109 Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission, para. 69. 110 Case T51/89 Tetra Pak Rausing SA v. EC Commission, II European Court Reports 00309 (1990).

78


the relationship between Article 81(3) and Article 82. The Court held that exemption granted under Article 81(3) does not bar the application of Article 82. It is held that there is no exception to the prohibition laid down by Article 82. Rather, the Court held that, ‘where agreements to which undertakings in a dominant position are parties fall within the scope of a block-exemption regulation, the effects of block exemption on the applicability of Article 86 [now Article 82] must be assessed solely in the context of the scheme of Article 86.’111

2.

PATENTS AND FORMAL STANDARDS

In Europe, many of the studies and discussions about the patents and standards originated from the formal standardisation works of ETSI (the European Telecommunications Standards Institute). The standard setting process of GSM and the dispute over the IPR policies of ETSI are two typical examples which demonstrate the tense relationship between patents and standards. In this section, the two cases are examined in order to clarify the issues concerning patents and formal standards. 2.1

Telecommunication Standardisation: the GSM Case112

GSM (Global System for Mobile communications), which is the most well known European telecommunication standard, is a good example of the tense relationship between patent rights and standards. The development of the GSM standard influenced the development of ETSI’s IPR policy, provoking controversy concerning intellectual property rights covering standards.113 2.1.1 From CEPT to ETSI Standardisation efforts for GSM were first made in CEPT (the European Conference of Postal and Telecommunications Administrations). Following the request of CEPT for a technical proposal for GSM, vendors offered eight different technical proposals. Four broadband designs, which were the outcome of R&D activities funded by a Franco-German programme, were based on CDMA (Code Division Multiple Access) technologies and optimised

111 112

Ibid., para. 3. Iversen et al., ‘Standardization, Innovation and IPR’; Liotard and Bekkers, ‘European Standards for Mobile Communications: The Tense Relationship between Standards and Intellectual Property Rights’. 113 Roger Tuckett, ‘Access to Public Standards: Interoperability Revisited’, European Intellectual Property Review 14 (12) (1992).


79

for high traffic communications. In contrast, the four Scandinavian designs were TDMA (Time Division Multiple Access)-based technologies, which showed the best performance for medium and low traffic intensities. Controversy arose over what should be adopted as standards and, as a consequence, the implementation of standards could not progress. Vendors searched for a strategic alliance to break a path for the resumption of the standard setting process. Eventually, Ericsson decided to cooperate with Siemens and a French company, which led CEPT to adopt a narrowband design as the GSM standard. From 1987 onwards, telecom operators (TOs) were organised by means of a Memorandum of Understanding (MOU), the main objective of which was to coordinate the launch of the GSM system in 1991. The synchronised launch of the GSM system and the compatibility of the equipment were crucial to the success of the GSM system. In 1987, TOs from 15 CEPT countries entered into the MOU and thus the MOU was implemented from the TOs’ perspective. In 1988, the standardisation work was transferred from CEPT to ETSI, the newly established standard setting organisation for European telecommunication. This implied that vendors from outside Europe could join the standard setting process, since ETSI was open to companies outside Europe. 2.1.2 Disputes over IPRs A substantial number of essential IPRs existed on basic GSM technology: by the end of the 1990s, about 140 patents were claimed to be essential to the GSM standard and this figure increased over time.114 Licences for building GSM products or operating GSM networks had to be negotiated individually and, in particular, the standard setting process was in danger of being stopped or significantly delayed when IPR owners refused to license their technology. Thus, a number of operators required the suppliers of the network to sign a declaration by which they agreed to serve the whole GSM community on fair, reasonable and non-discriminatory conditions. Most manufacturers agreed to such a statement but some manufacturers, in particular Motorola, refused to sign any arrangement that was not related to individual purchase contracts.115 Motorola had three times more essential patents than its rival had and therefore was unwilling to lose the additional returns afforded by licensing royalties. Motorola only wanted a cross-licence, rather than a licence on a monetary basis, with the intention to use its patent portfolios to gain access to market share in the European market.116 Eventually Siemens, Alcatel, Nokia and 114 115 116

Iversen et al., ‘Standardization, Innovation and IPR’, 71. Ibid., 72–3. Ibid.

80


Ericsson agreed to such cross-licensing agreements with Motorola between 1990 and 1993, and these five companies began dominating the equipment markets of GSM.117 Even though the ‘1993 ETSI IPR Policy and Undertaking’ only allowed monetary consideration of licensing contracts, it was abandoned in 1994 due to strong opposition from large IPR holders.118 2.1.3 Review of the GSM case Two issues arise from the standard setting process of GSM: the problem of blocking patents and the problems caused by a plurality of right holders. Firstly, the problem of blocking patents can result from the situation where the holder of essential patents refuses to license or refuses to grant licences on a fair, reasonable and non-discriminatory basis. The blocking patents can stop or significantly delay the standardisation process. This is the issue which needs to be analysed from the perspective of intellectual property policy. Secondly, crucial problems can arise from situations where the standards include a number of patents. The vested interests of a plurality of right holders may prevent or significantly delay the standard setting process because no one of them wants others’ patents included in standards.119 However, this is a procedural issue rather than an intellectual property policy one. That is, this issue may be treated in proper ways by designing a fair and reasonable procedure of formal standardisation such that each right holder can accept the procedure. Thus, this issue is not examined and only the problem of blocking patents is considered in this book. 2.2

Dispute over ETSI IPR Policy

2.2.1 Establishment of ETSI ETSI is Europe’s third standardisation body, established in 1988 in response to the EC’s 1987 Green Paper calling for the establishment of a third European standard setting body.120 The function of ETSI is to set the standards for telecommunication, information technology and broadcasting. According to the Commission, the telecommunication industry was vulnerable to the intel-

117 Liotard and Bekkers, ‘European Standards for Mobile Communications: The Tense Relationship between Standards and Intellectual Property Rights’, 123. 118 Mark Shurmer and Gary Lea, ‘Telecommunications Standardization and Intellectual Property Rights: A Fundamental Dilemma?’, StandardView 3, no. 2 (1995). 119 Farrell, ‘Standardization and Intellectual Property’, 41–2. 120 Martin Schiessl, ‘The New European Telecommunications Standards Institute Policy: Conflicts Between Standardisation and Intellectual Property Rights’, European Intellectual Property Review 15 (8) (1993).


81

lectual property rights used to erect barriers to trade and competition and that was the main reason why the Commission recommended the establishment of ETSI, which had quite different characteristics from the other European standardisation bodies.121 Hence, ETSI endeavoured to contribute to the creation of a common market for the telecommunication industry in Europe by eliminating obstacles, created by intellectual property rights, to open access to standards. Considering this background, it is understandable that ETSI sought to develop procedures to resolve the potential conflict between intellectual property rights and standardisation. 2.2.2 IPR Policy and Undertaking ETSI announced a proposal of its IPR Policy and Undertaking in 1991 and adopted a modified version of the proposal in 1993. When implementing the IPR Policy and Undertaking, ETSI sought to reconcile the conflict between intellectual property rights (IPRs) and standardisation by imposing certain limits on the IPR holders.122 The main provisions of the IPR Policy and Undertaking are as follows:123 • ETSI members are required to grant a licence of essential IPRs to other members on fair, reasonable and non-discriminatory terms. • The licence must be for monetary consideration and the licensors have a duty to inform ETSI in advance of the maximum royalty rate. • The IPR holders have ‘the right to withhold’ the licences if they notify ETSI of the IPRs and the relevant standards within 180 days after the date on which the Technical Assembly decides to put the draft standard into its work programme. Although the ETSI IPR Policy and Undertaking attempted to resolve the potential conflict between IPRs and standardisation, it provoked strong resistance from major IPR holders in the world including the Computer and Business Equipment Manufacturers Association (CBEMA), a United States trade organisation, which complained to the EC Commission.124 There were two main issues which provoked heated debates in relation to the policies. Firstly, it was argued that the right to withhold could not be exercised

121 122

Ibid. Shurmer and Lea, ‘Telecommunications Standardization and Intellectual Property Rights: A Fundamental Dilemma?’, 56–8. 123 Liotard and Bekkers, ‘European Standards for Mobile Communications: The Tense Relationship between Standards and Intellectual Property Rights’, 121. 124 Ibid.

82


due to the lack of necessary information.125 ETSI was not obliged to search relevant IPRs and thus the information about which IPRs would be involved in standardisation work was not supplied to IPR holders by ETSI. The IPR holders were the ones who should claim the right to withhold the licence. The only information available to IPR holders about standardisation work was the summaries about the Technical Assembly’s work programme, and it was argued that the summaries did not provide sufficient information for IPR holders to monitor all the relevant standardisation activities and to conduct IPR searches.126 Secondly, there was a criticism that ETSI restricted the right to make free contracts by requiring that licences should be on a monetary basis and that the maximum royalty rate should be notified in advance.127 ETSI imposed these limitations in order to clarify the meaning of a fair, reasonable, and nondiscriminatory agreement, but it was argued that the licensing agreement should be left to negotiation between the relevant parties.128 In 1994, after the efforts to resolve the controversies failed, ETSI abandoned the IPR Policy and Undertaking and adopted a new one. According to the new policy, IPR holders would be rewarded in a suitable and fair manner, and members could choose not to license IPRs. The dispute over ETSI IPR policy represents the fundamental conflict between the strong protection of IPRs for technical innovation and the difficulties of implementing standards caused by strong IPRs. It also implies that the SSO’s attempt to resolve this conflict by imposing the obligation on the IPR holders may not be successful.

3.

EXCEPTIONAL CIRCUMSTANCES AND INDUSTRY STANDARDS

3.1

IMS and Standards

In Magill and IMS, the ECJ established the exceptional circumstances where a dominant undertaking’s refusal to license violates Article 82 of the EC Treaty. The exceptional circumstances may be relied on by those who hope to 125 Williams Ellis, ‘Intellectual Property Rights and High Technology Standards’, in Standard Policy for Information Infrastructure, ed. Brian Kahin and Janet Abbate (Cambridge: The MIT Press, 1995). 126 Ibid. 127 Shurmer and Lea, ‘Telecommunications Standardization and Intellectual Property Rights: A Fundamental Dilemma?’, 57–8. 128 Ibid.


83

obtain licences of patents covering standards. Indeed, the IMS case has important implications for standards because the 1860 brick structure has characteristics of network products such as network effects and collective switching costs, which are typical features of information goods.129 As is described in Chapter 2, network effects, which work on the demand side, occur when the value derived from the consumption of products is affected by the number of other people who use the same or compatible products.130 After producers and consumers have decided to use one system or standard rather than another, they invest in the system or standard. The longer they use the chosen system or standard, the higher the cost they should pay to switch to another system or standard, which is called the switching cost. Network effects and switching costs cause markets to be locked into a particular standard because producers and consumers are reluctant to switch to other standards. These are the specific features of the network industries. Here it is argued that the brick structure of IMS has the characteristics of network effects and switching costs. The pharmaceutical companies, which were the customers of IMS, participated in the creation of the brick structure, which led the companies to depend on the structure and to prefer the structure to others.131 The brick structure was distributed free of charge to customers such as pharmacy accounting centres and associations of health insurance schemes,132 which increased the installed base of the brick structure. Once the installed base of the brick structure exceeds the critical mass, which is the minimum number of customers needed to initiate the positive feedback, the network effects begin to work. That is, the more customers utilise the brick structure, the more other customers decide to use the structure because customers can gain more benefits by using the widely used structure. This positive feedback leads the market to tip in favour of one product, which becomes a standard in the market. Indeed, according to the German court, the brick structure became an industry standard in the relevant market.133 The pharmaceutical companies invested their resources in utilising the brick structure; they developed their information systems in accordance with the brick structure of IMS and

129 Carl Shapiro and Hal R. Varian, Information Rules (Boston: Harvard Business School Press, 1999) Ch. 7. 130 Lemley and McGowan, ‘Legal Implications of Network Economic Effects’, 482–3. 131 Stothers, ‘IMS Health and Its Implications for Compulsory Licensing in Europe’, 470. 132 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., AG7. 133 Ibid.

84


managed their marketing data using the information systems.134 Since the companies invested their resources in the standards, they became reluctant to switch to another system unless the other system provided benefits which were big enough to offset the switching costs. Even though the individual switching cost, which an individual company has to pay to switch to another standard, may be trivial, the collective switching costs, which the industry as a whole has to pay to convert from the existing standard to a new standard, can be significant, and the problem of coordination, which is necessary to make the whole industry switch to another standard, is difficult to overcome, as the historical example of the QWERTY keyboard layout demonstrates.135 Thus, the market was locked into the 1860 brick structure, and the brick structure shows the typical features of network products. 3.2 Markets, Dominance and Exceptional Circumstances How can the exceptional circumstances be applied to the patent issues of industry standards? It has to be noted that the exceptional circumstances only apply where the patent owner has a dominant position in relevant markets. Therefore, it is necessary to define the relevant markets and to find that the undertaking has a dominant position in the markets. As for the relevant markets, the ECJ held in IMS that the relevant market was not necessarily the actual one but could be a potential or hypothetical one.136 In IMS, the ECJ endorsed the Advocate General’s opinion that [s]ince it has therefore been established that in order to be able to identify a market for upstream inputs it is not necessary for them to be marketed independently by the undertaking controlling them, it seems plain to me that such a market may by definition be always identified where: (a) the inputs in question are essential (since they cannot be substituted or duplicated) to operating on a given market; (b) there is an actual demand for them on the part of undertakings seeking to operate on the market for which those inputs are essential.137

From an economic perspective, a market is defined as ‘a set of arrangements by which buyers and sellers exchange goods and services’,138 that is, both supply and demand are necessary in a market. However, according to the ruling, the actual demand is crucial in finding the relevant markets, whereas

134 135 136

Ibid. Shapiro and Varian, Information Rules 184–6. Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., para. 44. 137 Ibid., AG59. 138 Begg et al., Economics 30.


85

supply does not matter. Therefore, where the above conditions of (a) and (b) are met, the market can always be identified although it may be a potential or hypothetical one. As for the dominance, whether the undertaking controlling the essential input is dominant or not depends on the scope of the relevant markets. As noted, the market is identified by the above conditions of (a) and (b). According to the condition (a), the inputs in question are essential to enter the downstream market, but there might be other suppliers who supply the inputs or alternatives to the inputs. Thus, the dominance of the undertaking controlling the essential input is not always established and only in those cases where the undertaking is in a dominant position can the exceptional circumstances be used. In the context of standards, however, the undertaking controlling the standards is likely to be dominant. After it has been found that the patent owner has a dominant position in the relevant markets, it has to be ascertained whether the elements of the exceptional circumstances have been established. The elements are as follows: 1. 2. 3. 4.

the product or technology in question is indispensable to carrying on a particular business; the refusal is such as to exclude all competition on a secondary market; the refusal prevents the appearance of a new product for which there is a potential consumer demand; the refusal is not objectively justified.

The issues of informal standards are different from those of formal standards and thus they need to be analysed separately. In the remainder of this section, the issues of informal standards and formal standards are examined and then to what extent the exceptional circumstances provide solutions to these issues is studied. 3.3

Exceptional Circumstances and Formal Standards

The critical issue concerning the formal standards and patents is blocking patents, as is described above. Some of the blocking patent problems can be resolved by applying the exceptional circumstances, but the others are not likely to be reconciled by this doctrine, as shown in the following subsections. 3.3.1 Blocking patents The issue of blocking patents arises from the situation in which the patent owners refuse to participate in an SSO and seek to enforce the patents in markets, or from the situation where the patent holders refuse to comply with

86


the IPR policies requiring them to grant licences on reasonable and nondiscriminatory terms or on a royalty-free basis. If the SSO can use alternative technology, this problem may be resolved, even though the alternative technology may be inferior to the patented technology. Where there is no alternative technology, the blocking patents are likely to hinder the standard setting process. Here it is assumed that there is no alternative technology because where there is alternative technology, whether it is superior or inferior to the blocking patents, the blocking patents are not indispensable and thus it is impossible to apply the exceptional circumstances. The problems of blocking patents can be divided into two categories. The first one is the case where the patent owner’s refusal to license blocks the standardisation process such that the process is halted or severely delayed (hereinafter ‘Type I’). The second one is the case where, because there is no alternative, the patented technology is adopted as the standard even though the patent owner reserves the right to refuse to license, as in the case of the GSM standard (hereinafter ‘Type II’). Application of the exceptional circumstances to Type I In order to allege an abuse of a dominant position under Article 82 of the EC Treaty, it is necessary to define the relevant markets and to establish that the patent owner has a dominant position in the markets. According to the ECJ’s ruling in IMS, the relevant market can always be identified ‘where the products or services are indispensable in order to carry on a particular business and where there is an actual demand for them on the part of undertakings which seek to carry on the business for which they are indispensable.’139 Thus the relevant upstream market does not have to be an actual market but can be a potential or hypothetical market. This concept is useful in finding the relevant market in Type I because, in this case, the actual relevant markets for the technology which the blocking patents cover may or may not exist, because the standard setting work is at an early stage and thus the technology protected by blocking patents may or may not have been introduced into the markets. Furthermore, the owner of the blocking patents may not carry on a licensing business but may merely seek to manufacture the patented products by himself and thus there may not be an actual market for the technology. In those cases where there is no actual market for the technology protected by the blocking patents, the relevant upstream market can be a potential or hypothetical one. The subject matters of the blocking patents are, by assumption, not substitutable with other technolo-

139 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., para. 44.


87

gies and thus the owner of the blocking patents is a monopolist or in a dominant position in the relevant market. In some cases, there can be an actual market for the patented technology. Considering that patents can be traded like tangible products and that essential patents are not interchangeable with other patents, the patents themselves may constitute a separate market just like spare parts.140 Especially when the patent owner actively engages in licensing businesses, the actual relevant technology markets are likely to exist. Because the relevant market is the one for the technology which the blocking patents cover, the patent owner has a dominant position in the market. Therefore, as far as Type I is concerned, the owner of the patented technology is in a dominant position in the relevant market, whether or not there is an actual market for the patented technology. Once it is established that the undertaking having the blocking patents is in a dominant position in the relevant market, it has to be determined whether each condition of the exceptional circumstances is met. The first element of the exceptional circumstances is indispensability, that is, the technology protected by the blocking patents has to be indispensable for operating in a downstream market. This raises the question of how to define the downstream market in this case where the standards are not yet implemented. Indeed, there is no downstream market. One may argue that, just as the upstream market can be hypothetical, the downstream market may be potential. However, this argument fails to consider the balances between the interest of the patent owner and the interest in the protection of free competition. This argument is biased in favour of competition without considering the adverse effect, caused by limiting the patentee’s rights, on technological innovation. It is improper to restrict the economic freedom of the patent owner in order to create competition in a hypothetical market. There is no guarantee that the compulsory licence would create a downstream market and competition therein, whereas it is clear that the compelled dealing would reduce the incentive to invent or innovate.141 Patent rights are statutory rights and any limitation on the rights should have clear grounds, but this is not the case here. Thus, 140 Case 22/78 Hugin Kassaregister AB and Hugin Cash Registers Limited v. EC Commission, paras 6–7. 141 Patents do not directly give an incentive to innovate. For this aspect, see Direct Protection of Innovation, ed. William Kingston (Lancaster: Kluwer Academic Publishers, 1987); Hermann Kronz, ‘Patent Protection for Innovation: A Model’, Part I and Part II, European Intellectual Property Review 7 (1983). However, Kitch emphasised the ‘incentive to innovate’ aspect of patents. See Edmund W. Kitch, ‘The Nature and Future of the Patent System’, Journal of Law and Economics 20 (1977). For more detail, see Chapter 6, Sections 2.2 and 2.3. Suffice here to say that patents provide an incentive to innovate at least in an indirect way.

88


this condition should not be established without a clearly defined downstream market. The second condition is the elimination of all competition in a downstream market. This condition is closely connected to the indispensability condition and, just like the condition of indispensability, it is not likely to be met, because there is no downstream market. The third element is the prevention of the emergence of a new product for which there is a potential consumer demand. As the ECJ held in IMS, a refusal to license patents constitutes an abuse of a dominant position, ‘where the undertaking which requested the licence … intends to produce new goods or services not offered by the owner of the right and for which there is a potential consumer demand’.142 However, the meaning of ‘new goods or services’ is not clear because how novel a new product should be is not clarified.143 Indeed, NDC, the plaintiff in IMS, argued that its products were different from those of IMS but this point was not clearly examined in the Court’s ruling.144 In addition, there is another obstacle which makes the application of the exceptional circumstances to Type I difficult. In Chapter 2, standards are defined as ‘a set of technical specifications adhered to by a producer, either tacitly or as a result of a formal agreement’.145 That is, standards are not new ‘products’ but technical specifications needed to make products. Standards may be used in making a bundle of new products but they are not new products themselves. It may be argued that, since standards will be used to manufacture new products, the prevention of the appearance of the standards with the blocking patents has to be regarded as the prevention of the emergence of new products for which there is a potential consumer demand. However, this argument is unacceptable. Whether a product is new or not has to be decided by comparing the ‘new product’ with the product covered by the patents. It is unreasonable to compare the patented product with a bundle of potential or hypothetical products which may or may not be produced in the future. Moreover, according to the TRIPS Agreement, a compulsory licence should ‘be considered on its individual merits’.146 Granting a compulsory licence to

142 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., para. 49. 143 Christian Ahlborn, David S. Evans, and Jorge Padilla, ‘The Logic and Limits of the “Exceptional Circumstances Test” in Magill and IMS Health’, Fordham International Law Journal 28 (2005): 1125. 144 Stothers, ‘IMS Health and Its Implications for Compulsory Licensing in Europe’, 470. 145 David and Greenstein, ‘The Economics of Compatibility Standards: An Introduction to Recent Research’, 4. 146 Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 31(a).


89

implement a standard is not considering an individual merit but considering a group of merits, which cannot be allowed under the TRIPS Agreement. Thus, whether the exceptional circumstances can be applied should be examined at the time when the new product is appearing or is about to appear in the market. Consequently, it is highly unlikely that the condition of the prevention of the emergence of a new product for which there is a potential consumer demand, will be established. The fourth condition of the test is objective justification. The scope of objective justification is not clear because the courts and the Commission provide little guidance on them. It is suggested that the risk of negative returns and serious congestion may be the possible justifications.147 Indeed, Qualcomm sought to justify its refusal to license by arguing that it would be commercial suicide to grant licences of its patents essential to implement UMTS (Universal Mobile Telecommunication System) standards unless UMTS was made compatible with Qualcomm’s CDMA technology.148 It is also suggested that where the grant of a licence would cause safety and quality concerns, the refusal to license may be justified.149 To sum up, where the patent owner’s refusal to license blocks the standardisation process such that the process is halted or severely delayed, it is unlikely that the exceptional circumstances can be applied to facilitate the standardisation process. On top of that, there is one more difficulty in applying the exceptional circumstances to the problem of blocking patents. Granting a compulsory licence to an SSO does not help to solve the problem of blocking patents. The exceptional circumstances are concerned with the refusal to license the blocking patents and thus, where all the elements of the doctrine are established, the party requesting licences (in the present case, the SSO) can obtain compulsory licences. The reason why the SSO needs the licence of the patents blocking the standardisation process is to implement the standards, because the SSO itself does not manufacture products incorporating the standards. Under the Paris Convention, a compulsory licence of a patent is not transferable, ‘even in the form of the grant of a sub-licence’, if it is granted on account of failure to work or insufficient working, which may be the case of

147

Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 441–2. 148 Jos Dumortier and Mark Hyland, ‘The Ericsson/Qualcomm Dispute and the Essential Facilities Doctrine in the Mobile Telecoms Sector’, Computer and Telecommunications Law Review 5, no. 6 (1999). 149 Opi, ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, 441–2.

90


blocking patents.150 In addition, the TRIPS Agreement prescribes that a compulsory licence of a patent ‘shall be considered on its individual merits’.151 Therefore, even though the SSO obtains the compulsory licences of the essential patents from the patent owner, a firm wishing to supply new products incorporating the standards has to obtain a licence from the patent owner, since the compulsory licences are granted not to the firm but to the SSO. Where the patent owner refuses to license the patent covering the standards to the firm concerned, the firm has to rely on the exceptional circumstances again. Therefore, from the perspective of a firm wishing to use the standards, the compulsory licences which the SSO obtains are of little use. Application of the exceptional circumstances to Type II In this sub-section, it is assumed that the patents are already incorporated into the standards and that the patent owner refuses to license. The patent owner did not agree to any restriction on licensing policies and thus he has the same right as the essential patents’ owner of an informal standard has. In this case, it is likely that there are actual downstream markets for goods or services, the markets in which the standards are the essential inputs to operate, and that patent owners have a dominant position or at least a collectively dominant position152 in the upstream markets of the technology which is covered by the patents. Thus, it is possible to apply the exceptional circumstances to Type II. As for the first condition of the exceptional circumstances, it may or may not be established, depending on the circumstances. In those cases where the standards are mandatory rather than voluntary, like the European Standards, this condition is not difficult to establish, because the blocking patents cover some parts of the standards and, without access to them, it is impossible to enter the markets. The blocking patents are likely to constitute ‘technical, legal or economic obstacles capable of making it impossible or at least unreasonably difficult for any undertaking seeking to operate in the market to create, possibly in co-operation with other operators, the alternative products or services’.153

150 Paris Convention for the Protection of Industrial Property (20 March 1884), Article 5 A(4). 151 Agreement on Trade-related Aspects of Intellectual Property Rights 1994, Article 31(a). 152 ‘[C]ollective dominance implies that a dominant position may be held by two or more economic entities legally independent of each other provided that from an economic point of view they present themselves or act together on a particular market as a collective entity’, Whish, Competition Law, 525. 153 Case C-418/01 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG., para. 28.


91

However, in those cases where the formal standards are not mandatory but voluntary, and some comparable technologies are competing with the formal standards in the market, this condition is not established because there are alternatives. According to the ECJ in IMS, to decide whether an input is indispensable or not, it must be determined whether there are products or services which constitute alternative solutions, even if they are less advantageous, and whether there are technical, legal or economic obstacles capable of making it impossible or at least unreasonably difficult for any undertaking seeking to operate in the market to create, possibly in co-operation with other operators, the alternative products or services.154

Thus, where alternative technology exists, even though it is inferior to the patented technology at issue, the patented technology is not considered indispensable. Even if the standards are voluntary, in some cases where the market is sufficiently mature and the competing technologies are only marginal155 this condition may be met because it may be impossible for competitors to ‘remain viably on the market’156 without the standard technologies, as in the Microsoft case. However, as is argued previously, the decision of Microsoft is problematic.157 To sum up, where the standards are mandatory, this condition is likely to be established and where it is economically unviable to stay on the market without the standards, this condition may or may not be established. Concerning the condition of the elimination of all competition, it is also case-dependent. Where the owner of the blocking patents is the sole producer of the goods incorporating the patented technology, this condition can be met. However, as in the GSM case, where several firms are collectively dominant, where the firms themselves provide the goods or services using the standards, and thus where the refusal to license may exclude competition other than the competition amongst the dominant firms, it is not clear whether it can be considered that the refusal eliminates all competition in the markets. It seems difficult to establish the condition because there is competition anyhow. Only in those cases where a third party wishes to supply new products which incorporate standards and are different from the products provided by the dominant firms, and where the new products cater for a new distinct market, is it likely that this condition will be met. 154 155 156 157

Ibid. This is the case in Microsoft. T-201/04 Microsoft Corp. v. EC Commission, para. 228. See Microsoft v. Commission in Chapter 4.

92


The third element is the prevention of the emergence of a new product for which there is a potential consumer demand. The patent owners may provide many of the possible applications of the standards, as in the GSM case, and, if this is the case, other firms do not have many options. Only those firms which seek to supply goods or services which are different from the dominant firm’s may be able to establish this condition, even though how new the products should be is not clear. Although there are uncertainties over the condition of the objective justification, the patent owner’s refusal to grant licences to firms seeking to manufacture products different from the ones produced by the patent owner is not likely to be justifiable. To sum up, where the patent owner’s exclusive rights are not restricted by the IPR policies of an SSO, it is not likely that third parties would gain access to the standards when the patent owner refuses to license, unless they plan to supply new goods or services which are different from the ones provided by the patent owner. In addition, what constitutes the ‘new’ goods or services is not clear. Furthermore, it is not easy to establish the requirement of the indispensability unless the standards are mandatory or the market is governed by a firm in a super-dominant158 or monopoly position. These limitations may cause the market which the standards are concerned with to have a monopolistic or oligopolistic structure, like the GSM markets.159 3.4

Exceptional Circumstances and Informal Standards

Here it is examined whether it is possible to apply the exceptional circumstances to informal standards incorporating patents when the patent owner refuses to license the essential patents. Informal standards are created by the market mechanism in those markets where the network effects are strong.160 It is clear that there are relevant markets in relation to the informal standards and that the owner of the informal standards has dominant positions in the markets. As for indispensability, the first condition of the exceptional circumstances, it is not easy to establish because in many cases there may be alternatives in the relevant markets. For instance, there are alternative technologies in the PC microprocessor market even though the Intel microprocessor is the market standard.161 These examples strongly suggest that there may be more than one 158 159

Whish, Competition Law 189–90. Liotard and Bekkers, ‘European Standards for Mobile Communications: The Tense Relationship between Standards and Intellectual Property Rights’, 123–4. 160 See Chapter 2, Section 3.2.2.. 161 See Kanellos, ‘AMD’s market share gains accelerate’.


93

competing technology in a network market, even though the network effects are strong. Where there are alternatives, even if the alternatives do not provide as many competitive advantages as the patented technology does, the condition of indispensability cannot be met. Since the alternatives exist, the refusal to license does not eliminate all competition in the relevant markets. Therefore, the second condition of the test cannot be established, either. Only in those exceptional cases where there is no alternative technology at all in the network market or where the alternatives are only marginal, as is the case in the market dominated by a monopolist or a super-dominant firm, may the first and second conditions be met. The third element, the prevention of the emergence of a new product for which there is a potential consumer demand, is also established in limited cases. The patent owners may already provide the products which incorporate the informal standards and thus the third condition is only met when a third party seeks to provide new products. Consequently, the application of the exceptional circumstances to informal industry standards is limited and thus it will be difficult for a third party to gain access to the informal standards if the patent owner refuses to license the patents essential to the standards. 3.5

Review and Limitations of the Exceptional Circumstances

The exceptional circumstances may be applied to some of the critical problems of standards incorporating patents but their use is limited to exceptional cases. Only in those cases where the standards are mandatory or the firm concerned is in a super-dominant position, and where a party hoping to obtain licences plans to supply new goods or services which are different from those provided by the patent owner, may the exceptional circumstances be used to impose compulsory licences on the patent owner. Furthermore, even in those cases where the exceptional circumstances are established in relation to the mandatory industry standards or the super-dominance, whether the patent owner can avoid the condemnation of an abuse of a dominant position by granting a single licence or a limited number of licences is not clear.162 Theoretically, the patent owner can circumvent the accusation by a single licence because a licence can create competition in the market and thus the refusal to licence by the patent owner no longer excludes all competition on a secondary market. If this is the case, there arises the question of how to choose the licensee who is lucky enough to obtain the licence. 162 James Turney, ‘Defining the Limits of the EU Essential Facilities Doctrine on Intellectual Property Rights: The Primacy of Securing Optimal Innovations’, Northwestern Journal of Technology & Intellectual Property 3 (2005): para. 26.

94


Another limitation of the exceptional circumstances doctrine is that it does not properly consider the right balance between the standard owner and the improvement inventors of the standard. Where the exceptional circumstances are established, the patent owner has to grant a compulsory licence to those who hope to obtain the licence. The doctrine does not consider the value created by standard technology and that produced by improvement technology, failing to allocate the proper legal entitlement according to the value from each technology. That is, the principle of the exceptional circumstances does not seek to strike the right balance between the incentive for the standard owner to create the standard technology and the incentive for other inventors to improve the standard technology. The incentive balance has to be found by way of allocating legal entitlement but the exceptional circumstances fail to do so. Intellectual property rights in general and patent rights in particular are granted to provide the incentives for economic entities to invent or innovate,163 and therefore, in imposing any restraints on the exclusivity of intellectual property rights, it is important to consider the adverse impact on innovation. In this sense, it has to be noted that Magill and IMS, which established the exceptional circumstances in which compulsory licences may be granted, are concerned with information, which can arguably be protected by copyrights or which has public characteristics. In Magill, the information to which access was sought was TV programme listings, which were provided without any charge to numerous third parties, such as newspapers. It was debatable whether such information was eligible for copyright protection even though the ECJ held that it was the national rules that determined the conditions and procedures for copyright protection.164 In Magill, it is unlikely that the copyright protection of the TV listings will create more TV listings or better ones. TV listings will be created regardless of the copyright protection and the copyright which covers them is merely a byproduct that is not essential to their production. In IMS, the information at stake was the brick structure which was based on postal codes and was designed with the participation of industry members. Whether the brick structure is copyrightable or not, and whether the structure needs to be copyrighted or not, it has public characteristics. As the former Advocate General Tesauro stated, the brick structure became an industry standard not because of ‘a genuine, independent, creative effort entirely coming from IMS’, but because of ‘the contribution and involvement of the pharmaceutical companies’.165 Although IMS contributed to the creation of the brick 163 164

Torremans, Holyoak and Torremans: Intellectual Property Law 11–25. Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission, para. 25. 165 Forrester, ‘Article 82: Remedies in Search of Theories?’, 944.


95

structure and to making the structure an industry standard, the customers and the market played a far more important role in creating the brick structure and giving IMS a monopoly position in the market. Therefore, it may be reasonable to consider the customers’ welfare in the first place rather than the adverse effect on IMS’s innovative effort. Clearly, introducing a new product for which there is a potential consumer demand is desirable from the customers’ point of view. In addition, it is not clear whether strong protection of the copyright covering the brick structure would increase the dynamic efficiency, introducing more or better products into the market. As some commentators have argued, once the 1860 brick structure was widely used in the market, producing any other data structure might be illegal because cross-referencing of the 1860 brick structure and the other one might constitute a violation of the German data protection law.166 On the other hand, patents are generally the outcome of considerable research and development of a private undertaking and are closely related to the incentives to invent and innovate. Unlike the copyright which subsisted in TV listings, patent rights are essential factors in technical innovation in that they encourage the economic entities to participate in innovative activities by providing exclusive rights to patent owners. Unlike the brick structure in IMS, a patent is created from the risk taking effort of private undertakings and thus it is privately owned property. Therefore, it may be argued that the uniform application of the exceptional circumstances to all sorts of intellectual property rights is not appropriate. Rather, it seems persuasive to argue that it is necessary to take into consideration the nature of the subject matters protected by intellectual property laws, and as for patents, which are generally created by the considerable research and development of a private undertaking, the application of the exceptional circumstances should be restricted and care should be taken not to punish the successful innovator by imposing an improper duty.167 All this reasoning suggests that applying the exceptional circumstances is not necessarily desirable because it may reduce the incentive to innovate and it is not a good means for striking the right balance between the incentive needed for standard owners and that necessary for improvement inventors. In addition, there are many uncertainties in applying the doctrine to cases involving patent rights.

166 Robert Pitofsky et al., ‘The Essential Facilities Doctrine under U.S. Antitrust Law’, Antitrust Law Journal 70 (2002): 444. 167 Turney, ‘Defining the Limits of the EU Essential Facilities Doctrine on Intellectual Property Rights: The Primacy of Securing Optimal Innovations’, 80.

5. Patent laws and standards 1.

INTRODUCTION

In Chapters 3 and 4, the patent issues in industry standards and the question of whether it is possible to resolve the conflict between patents and standards by applying the principles of competition laws or antitrust laws are examined. The issues can be classified into two categories: the issues of informal standards and those of formal standards. The issues of informal standards, which are generated by the market process, arise when the patent owner refuses to license his patents, covering some parts of industry standards, to others who hope to gain access to the patented technology. It is highly unlikely that the legal principles of competition laws such as the exceptional circumstances can be applied to these issues, except for some limited cases. The issues of formal standards, which are promulgated by standard setting organisations (SSOs), include those of informal standards and other issues such as the standard hold-up problem. This is where the patent owners refuse to have their patents included in standards and there is no alternative, so the patents block the standardisation process. Even though the legal principles of competition laws and antitrust laws may resolve some of the issues, they can only be used in limited circumstances. Then, how can the issues be resolved from the perspective of patent laws rather than competition laws? In general, patent laws grant exclusive rights to inventors and innovators in order to promote inventive and innovative activities. Indeed, the very essence of intellectual property in general and patents in particular is their exclusivity.1 Even in patent laws, however, there are some legal principles which restrict the exclusive rights of patent owners. It is these patent law principles which a third party may rely on when a patent owner refuses to license his patent, which is essential for the party to carry on a business in the related market. There are two patent law principles which need to be studied to see whether they can be used to resolve the patent issues in standards. They are the reverse doctrine of equivalents and compulsory licences. The reverse doctrine of 1

Case 238/87 Volvo AB v. Erik Veng (UK) Ltd, 4 C.M.L.R. 122, para. 8 (1989). 96

Patent laws and standards

97

equivalents is used to limit the enforcement of patent rights in those cases where an alleged product is far different from the patented product, even though the product literally infringes the patent at issue.2 A compulsory licence is ‘an involuntary contract between a willing buyer and an unwilling seller imposed and enforced by the state’,3 which clearly restricts the exercise of exclusive rights granted by patent laws. In this chapter, both the reverse doctrine of equivalents and compulsory licences are studied and whether they can be applied to the patent issues in standards is examined. In addition, to what extent they resolve the patent issues in standards is analysed.

2.

REVERSE DOCTRINE OF EQUIVALENTS

Some commentators have suggested that the reverse doctrine of equivalents should be used to overcome the bargaining difficulties between the initial patent owner and subsequent inventors.4 In general, the reverse doctrine of equivalents has been used to resolve the conflict between the frontier patent and its improvements, but some have argued that the doctrine should be applied to relieve the hold-up problems in network markets.5 In this section, the reverse doctrine of equivalents is studied and whether the doctrine can be used for the patent issues in standards is also examined. In order to analyse the reverse doctrine of equivalents, the doctrine of equivalents should be studied first because the two doctrines are closely related. Thus, the doctrine of equivalents, the reverse doctrine of equivalents and its implication are examined in the following sub-sections. 2.1

Doctrine of Equivalents

The doctrine of equivalents is a legal principle which is used to determine the boundary of patent rights in US courts and is employed to broaden the scope 2 William S. Galliani, ‘Patent Infringement Amidst Rapidly Evolving Technologies: New Equivalents, the Doctrine of Equivalents and the Reverse Doctrine of Equivalents’, Santa Clara Computer and High Technology Law Journal 6 (1990): 86. 3 Gianna Julian-Arnold, ‘International Compulsory Licensing: The Rationales and the Reality’, IDEA: The Journal of Law and Technology 33 (1993): 349. 4 Galliani, ‘Patent Infringement Amidst Rapidly Evolving Technologies: New Equivalents, the Doctrine of Equivalents and the Reverse Doctrine of Equivalents’; Mark A. Lemley, ‘The Economics of Improvement in Intellectual Property Law’, Texas Law Review 75 (1997); Robert P. Merges, ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’, Tennessee Law Review 62 (1994a). 5 Lemley, ‘The Economics of Improvement in Intellectual Property Law’, 1066–7.

98


of the patent claims beyond the literal meaning of the claims.6 The doctrine has been developed by the rulings of the US Supreme Court in three patent infringement suits.7 After adopting the doctrine in Graver Tank,8 the Supreme Court reaffirmed and modified it in Warner-Jenkinson9 and Festo.10 2.1.1 Graver Tank & Manufacturing Co. v. Linde Air Products Co.11 The patent in this case was concerned with the electric welding process and welding fluxes which are used in the process. The Linde Air Products Company filed a patent infringement suit against Graver Tank & Manufacturing Company and others. The plaintiff’s patent covered fluxes including alkaline earth metal silicates, whereas the defendant’s flux contained manganese silicate. The issue in this case was whether the defendant’s flux infringed the patent claims covering fluxes including alkaline earth metal silicates, even though manganese was not an alkaline earth metal. Because manganese was not an alkaline earth metal, it was clear that the defendant’s product did not infringe the patent literally. However, the Supreme Court upheld the lower court’s ruling, holding that the patent was infringed by the defendant’s product under the doctrine of equivalents. The court reasoned that, in determining the infringement, it had to be examined in the first instance whether the claims read on the accused matter and, if so, no more analysis was needed since it was clear that the patent at issue was infringed.12 However, the court recognised that ‘outright and forthright duplication [was] a dull and very rare type of infringement’ and if the patent scope was limited to its literal meaning, the patent right would be useless because a dishonest copyist could be free from the infringement charge with only minor and insubstantial changes and substitutions.13 The court held that the doctrine of equivalents would solve the problem arising from the limitation of the

6 7

Ibid., 1043–4. This doctrine was criticised by Lord Hoffmann as ‘a doctrine born of despair’. See Kirin-Amgen Inc. v. Hoechst Marion Roussel Ltd., R.P.C. 9, para. 41 (2005). In Europe, as the new article 2 of the Protocol on the interpretation of Article 69 of the European Patent Convention expressly provides, equivalents have to be taken into account in interpreting claims. 8 Graver Tank & MFG. Co., Inc., et al. v. Linde Air Products Co., 339 U.S. 605 (1950). 9 Warner-Jenkinson Company, Inc., v. Hilton Davis Chemical Co., 520 U.S. 17 (1997). 10 Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Ltd., et al., 535 U.S. 722 (2002). 11 See Graver Tank & MFG. Co., Inc., et al. v. Linde Air Products Co. 12 Ibid., 607. 13 Ibid.


99

literal interpretation of claims and that the doctrine had been used to ‘prevent an infringer from stealing the benefits of the invention’.14 The court went on to hold that ‘if a device perform[ed] substantially the same function in substantially the same way to obtain the same result’ as the patented device, then the device was in the scope of the patent right.15 According to the court, it was an important factor in determining the equivalency whether a person skilled in the art would have known that an element not included in the patent was interchangeable with one that was contained.16 The court noted the experts’ testimony that manganese was equivalent to and could be used instead of magnesium, which is an alkaline earth metal. The court also noted that the patent specification disclosed the use of manganese silicate influxes and that there was no indication that the defendant developed the flux independently. In this way, the court ruled that even though the defendant’s product did not infringe the patent literally, it fell within the scope of the patent claim under the doctrine of equivalents.17 2.1.2 Warner-Jenkinson Co. v. Hilton Davis Chemical Co.18 About fifty years after the Graver Tank case, the Supreme Court had to examine the doctrine of equivalents in Warner-Jenkinson.19 The patent of this case was for a purification process including the ultrafiltration of dye through a membrane at a pH level between 6.0 and 9.0. The inventors of the patent reduced the scope of their claims by limiting the pH level to between 6.0 and 9.0 during patent prosecution in response to the office action of the patent examiner who provided a prior art disclosing an ultrafiltration process at a pH level above 9.0. It was clear that the upper limit of the pH range was set to overcome the prior art, but the reason why the lower limit was set at 6.0 was not clear. The defendant, Warner-Jenkinson Company, developed an ultrafiltration process which was different from the patented process only in that it operated at a pH level of 5.0. The patentee brought an infringement action against the defendant, alleging that the defendant’s process for purification fell within the scope of the patented process under the doctrine of equivalents. The defendant argued that the doctrine of equivalents should not be applied to this case, for the following reasons:

14 15 16 17 18

Ibid., 608. Ibid. Ibid., 609. Ibid., 612. For more detail, see Warner-Jenkinson Company, Inc., v. Hilton Davis Chemical Co. 19 Ibid.

100


(1) The doctrine of equivalents is inconsistent with the statutory requirement that a patentee specifically ‘claims’ the invention covered by a patent, (2) the doctrine circumvents the patent reissue process – designed to correct mistakes in drafting or the like – and avoids the express limitations on that process, (3) the doctrine is inconsistent with the primacy of the Patent and Trademark Office (PTO) in setting the scope of a patent through the patent prosecution process; and (4) the doctrine was implicitly rejected as a general matter by Congress’ specific and limited inclusion of the doctrine in one section regarding ‘means’ claiming, § 112, ¶ 6.20

The court denied the first three of the four arguments, holding that it had already rejected the arguments in Graver Tank and that there was no basis to overrule the previous ruling. The court also rejected the defendant’s fourth argument, saying that § 112, ¶ 621 was enacted to allow the means claims, which were drafted with the language of a means for performing a specific function. The court reasoned that it would be wrong to infer that Congress intended to abolish the doctrine of equivalents with such limited congressional actions. Thus the court reaffirmed the doctrine of equivalents by refusing to overturn it. The defendant also argued that the doctrine of equivalents should be restricted in accordance with (1) the prosecution history estoppel, (2) a requirement of wrongful intent, and (3) a requirement that the patent should disclose the equivalents.22 The court reasoned that the prosecution history estoppel would bar the application of the doctrine of equivalents only in those cases where the patent application was amended to overcome the rejection related to patentability. Since it was not clear whether the reason for limiting the lower limit of the pH level was relevant to patentability and the Federal Court did not consider it, the court remanded the case to the lower court. As for the requirement of wrongful intent, the court ruled that the known interchangeability of substitutes for an element of a patent is one of the express objective factors noted by Graver Tank as bearing upon whether the accused device is substantially the same as the patented invention … The better view, and the one consistent with Graver Tank’s predecessors and the objective approach to infringement, is that intent plays no role in the application of the doctrine of equivalents.23

20 21

Ibid., 25–26. 35 U.S.C. § 112, ¶ 6: An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof. 22 Warner-Jenkinson Company, Inc., v. Hilton Davis Chemical Co., 30–37. 23 Ibid., 36.


101

Concerning the requirement that the patent should disclose the equivalents, the court held that the time for deciding the equivalency was not at the time when the patent was issued but at the time when the patent was infringed. This ruling implies that the patent scope under the doctrine of equivalents can expand as new technologies are developed since newly developed technologies may be considered equivalents of the claimed technologies. Thus the US Supreme Court not only reaffirmed the doctrine of equivalents but also expanded its application by ruling that the newly developed equivalents were included in the scope of equivalents under the doctrine of equivalents. 2.1.3 Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co.24 In this case, the alleged infringer did not argue that the doctrine of equivalents should be abolished. Rather, the doctrine was accepted as an established legal principle but the scope and the applicability of the doctrine were in question. The issues of this case were concerned with the relationship between the prosecution history estoppel and the doctrine of equivalents. Festo Corporation had two patents, the Stoll patent and the Carroll patent, which covered magnetic rodless cylinders. The patent application for the Stoll patent was amended in order to overcome the patent examiner’s objection during the examination. The examiner rejected the application of the Stoll patent not because it was obvious from the prior arts but because the specification was unclear and the claims were written in an impermissible way. The Carroll patent was also amended in the re-examination proceeding. Both patents were amended to have a new element of a pair of sealing rings at opposite ends of the piston, and the Stoll patent came to have a further limitation that ‘the outer shell of the device, the sleeve, be made of a magnetizable material’.25 Festo, the patentee, filed an infringement suit against SMC, the defendant, which sold cylinders similar, but not identical, to the ones covered by Festo’s patents. The defendant’s rodless cylinder used a single two-way seal rather than two one-way seals and the sleeve was made of a non-magnetisable material. The defendant argued that, since the plaintiff narrowed the patent scope to obtain patents and the elements added by the narrowing amendment were the only difference between the patented cylinder and the accused one, the plaintiff was estopped from arguing that the defendant’s device was equivalent to the patented device under the doctrine of equivalents.26 24 For more detail, see Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Ltd., et al. 25 Ibid., 728. 26 Ibid., 729.

102


The Court of Appeals for the Federal Circuit held that any amendment that narrowed the scope of claims to obtain patents gave rise to estoppel. Especially, the Federal Circuit held that the narrowing amendment barred the application of the doctrine of equivalents completely to the elements added in the amendment. The approach of ‘complete bar’ was controversial and was the main issue of this case. The Supreme Court began its reasoning by stating that the language in the patent claims may not capture every nuance of the invention or describe with complete precision the range of its novelty. If patents were always interpreted by their literal terms, their value would be greatly diminished. Unimportant and insubstantial substitutes for certain elements could defeat the patent, and its value to inventors could be destroyed by simple acts of copying. For this reason, the clearest rule of patent interpretation, literalism, may conserve judicial resources but is not necessarily the most efficient rule. The scope of a patent is not limited to its literal terms but instead embraces all equivalents to the claims described.27

Thus the court reaffirmed that the doctrine of equivalents was necessary to decide the proper scope of patents, even though the court acknowledged that the doctrine would make the scope of patents uncertain.28 The court also reasoned that the prosecution history estoppel was necessary for the proper determination of the boundary of the patent scope and held that the prosecution history estoppel … preclud[es] a patentee from regaining, through litigation, coverage of subject matter relinquished during prosecution of the application for the patent. Were it otherwise, the inventor might avoid the PTO’s gatekeeping role and seek to recapture in an infringement action the very subject matter surrendered as a condition of receiving the patent.29 (citation omitted)

After the general discussion of the doctrine of equivalents and prosecution history estoppel, the court began examining the two specific questions of this case. The first question is what kind of amendment gives rise to an estoppel? That is, it is accepted that an estoppel arises when the amendments are made to avoid the prior art, but how about the amendment made to comply with requirements on the form of the patent application? The court held that ‘a narrowing amendment made to satisfy any requirement of the Patent Act may give rise to an estoppel’30 but if an amendment is purely cosmetic, then the

27 28 29 30

Ibid., 731–2. Ibid., 732–3. Ibid., 734. Ibid., 736.


103

amendment would neither reduce the patent scope nor give rise to an estoppel.31 The second question is whether the estoppel is a complete bar or a flexible bar; that is, whether or not the estoppel completely bars the patentee from asserting infringement under the doctrine of equivalents. The court ruled that by amending the application, the inventor is deemed to concede that the patent does not extend as far as the original claim. It does not follow, however, that the amended claim becomes so perfect in its description that no one could devise an equivalent. After amendment, as before, language remains an imperfect fit for invention. The narrowing amendment may demonstrate what the claim is not; but it may still fail to capture precisely what the claim is. … As a result, there is no more reason for holding the patentee to the literal terms of an amended claim than there is for abolishing the doctrine of equivalents altogether and holding every patentee to the literal terms of the patent.32

As such, the court justified with the limitation of language the application of the doctrine of equivalents to the amended claims. To sum up, the doctrine of equivalents, which is used to broaden the scope of the claims of patents beyond the literal meaning of the claims, has been repeatedly affirmed by the US Supreme Court. After Graver Tank, the scope of the doctrine was expanded by the ruling of Warner-Jenkinson, which had the new equivalents, developed at the time of infringement, included in the equivalents, whereas the ruling of Festo narrowed the scope of the doctrine in those cases where patent claims were amended. 2.2

Reverse Doctrine of Equivalents

While the doctrine of equivalents is used to broaden the scope of patent claims, the reverse doctrine of equivalents is employed to reduce it. The reverse doctrine of equivalents is a judicially created principle and is used to limit the enforcement of patent rights in those circumstances where an alleged product is considerably different from the patented product, even though the patent claims read on the alleged product.33 In Graver Tank, the Supreme Court recognised the reverse doctrine of equivalents, holding that [t]he wholesome realism of this doctrine,34 is not always applied in favor of a patentee but is sometimes used against him. Thus, where a device is so far changed in principle from a patented article that it performs the same or a similar function in a

31 32 33

Ibid., 737. Ibid., 738. Galliani, ‘Patent Infringement Amidst Rapidly Evolving Technologies: New Equivalents, the Doctrine of Equivalents and the Reverse Doctrine of Equivalents’, 86. 34 Here ‘this doctrine’ refers to the doctrine of equivalents.

104


substantially different way, but nevertheless falls within the literal words of the claim, the doctrine of equivalents may be used to restrict the claim and defeat the patentee’s action for infringement.35

The concept of this doctrine was established in Boyden Power-Brake v. Westinghouse.36 The patent in question was issued in 1887 to George Westinghouse, Jr, the plaintiff, and was concerned with a train braking system using the power of compressed air from a main air reservoir, located under the locomotive, and a local air reservoir in each brake cylinder. George Boyden, the defendant, improved the braking system and invented a train braking system which had a mechanism for supplying compressed air into a brake piston from both the main air reservoir and the local reservoir, increasing the braking power and reducing the time required to stop the train. The plaintiff filed a patent infringement suit against the defendant, arguing that the defendant’s braking system infringed the plaintiff’s patent. After examining the alleged device and the patent claims, the Supreme Court held that the defendant’s device was covered by the first and the fourth claims of the plaintiff’s patent.37 The court stated, however, that literal infringement did not necessarily lead to the conclusion that the alleged device infringed the patent rights, holding that [t]he patentee may bring the defendant within the letter of his claims, but if the latter has so far changed the principle of the device that the claims of the patent, literally construed, have ceased to represent his actual invention, he is as little subject to be adjudged an infringer as one who has violated the letter of a statute has to be convicted, when he has done nothing in conflict with its spirit and intent.38

The court reasoned that even though the functions of both the alleged device and the patented device were the same, the mechanisms to accomplish the functions were so distinct that they were not equivalents. Thus, where the infringer’s device is much different from the patented device, the reverse doctrine of equivalents sets the infringer free from the patent infringement liability. This has provided the doctrinal ground for those who invented the improvement of a patented invention but were accused of patent infringement.39

35 36 37 38 39

Graver Tank & MFG. Co., Inc., et al. v. Linde Air Products Co., 608–9. Boyden Power-Brake Co. et al. v. Westinghouse et al., 170 U.S. 537 (1898). Ibid., 568. Ibid. Merges, ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’, 93.


2.3

105

Implication of the Reverse Doctrine of Equivalents

It is argued that the reverse doctrine of equivalents should be used to resolve the bargaining problem between the owner of an initial patent and the followon inventors.40 The argument is persuasive and can be summarised as follows.41 Suppose that there are an initial patent, called a frontier patent or the first patent, and three sorts of improvers of the patented invention: minor improvers, significant improvers, and radical improvers. Minor improvers are those who, by inventing improvement technology of the first invention, create additional value to an amount less than the value of the first patent. Minor improvers include those who change the patented technology obviously or trivially from the perspective of a person skilled in the art. In this case, their modification is not protected by patent laws since the value created by the improvement is so small that the improvement invention does not overcome the threshold set up by the patentability requirements, especially the inventive step requirement. Thus, minor improvers cannot prohibit others, including the patent owner, from using their improvement. Accordingly, where the patent owner and the minor improver bargain over the exploitation of the patented invention and the improvement, the patentee has absolute power over the bargaining. This extreme imbalance of bargaining power is reasonable since the value created by the patented invention and the improvement mostly arises from the patented invention. Significant improvers are those who improve the value of the first patent to a level above the value of the first patent but less than some high value. The value created by a significant improvement to an invention is greater than that created by a minor improvement but less than that by a radical improvement. The significant improvers add to the first patent a value greater than the threshold for patentability and thus they can protect their contribution with patents, which are called dependent patents. The significant improver can be in a stronger position than the minor improver in bargaining with the first patent’s owner because the significant improver has the patent protecting the improvement invention. Where the owner of the first patent agrees to give licences to the significant improver, both the owner of the first patent and the significant improver can enjoy the benefit from the exploitation of the improvement. Where they fail to agree, the improver cannot exploit the improvement, while

40 Lemley, ‘The Economics of Improvement in Intellectual Property Law’; Merges, ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’. 41 The argument is described in Lemley, ‘The Economics of Improvement in Intellectual Property Law’, 1008–13.

106


the owner of the first patent cannot obtain benefits from the improved invention. This is called the case of blocking patents between the first patent and the dependent patent. Radical improvers are those who add to the first patent a value far greater than that of the first patent. It is here that the reverse doctrine of equivalents works. In the case of a radical improvement, the radical improver creates far more value than the inventor of the first patent and thus it seems reasonable to argue that the legal entitlement has to be assigned to the radical improver rather than the initial patent owner. The reverse doctrine of equivalents enables the legal entitlement to be allocated in such a way since the radical improver is not condemned as an infringer of the first patent, even though the claims of the first patent read on the radically improved invention. As such, the argument is based on the spectrum of value created by the improvement inventions, from minor improvements to radical improvements, and is plausible in the case of improvement inventions.42 However, the reasoning is not likely to be applicable to patent issues in standards, as the following example shows. Suppose that A is the compatibility standard and is protected by patent rights. The reverse doctrine of equivalents implies that when a following inventor improves A to another invention A1, which is much different in principle from A, the device incorporating A1 is not deemed to infringe A, even though the device literally infringes the patent covering A. In this case, A1 is the invention which is radically improved from A and thus far different from A. Then, is it possible to resolve the compatibility problem arising from the patentee’s refusal to license the essential patent by not condemning A1 with the patent infringement charge? The answer depends on the characteristics of A and A1, but it seems that A1 is not likely to be compatible with A or with other products which are compatible with A. Rather, A1 is likely to be incompatible with A and other network products because it is not A1 but A which makes network products compatible with each other and, by assumption, A1 is radically different from A. The reverse doctrine of equivalents is only applied to those cases where the accused product is radically different from the patented product and thus the two products are likely to be distinct ones or at least to be incompatible with each other. The key for a product to be compatible with other network products is to incorporate the standard, A. Therefore, the structure of products in the relevant network market has to be A+B, A+C, A+D, and so on. Where a subsequent researcher invents A+B, an improvement invention, and obtains patents covering the invention, he can prevent the owner of the patent covering the invention A or others from exploiting the invention A+B, even in those cases where

42

Ibid.


107

he failed to obtain a licence from the owner of the patent A. Thus, it is true that the problem of blocking patents arises in relation to standards, but it is unlikely that the reverse doctrine of equivalents could be applied to the patent issues in standards. To sum up, the reverse doctrine of equivalents is useful in settling the bargaining breakdown between the owner of the first patent and improvers of the first patent. The reverse doctrine of equivalents only applies where the patented technology and the alleged technology are so different that it does not make sense to argue that the alleged technology is covered by the patent at stake. Therefore, the doctrine provides a way to promote the improvement invention subsequent to the first invention. However, it does not follow that the doctrine resolves the patent issues in standards. What matters in relation to standards is not the immunity from the infringement charge, which the reverse doctrine of equivalents may provide, but the availability of the very same technology protected by the patent essential to standards. Compatibility, which is essential to enter the network market, can be obtained not from the radical improvement of the standard technology but only by incorporating the standard technology. In this respect, the reverse doctrine of equivalents is not likely to solve the compatibility issues of standards. Then, is there any other way within patent laws to gain access to the patented technology when the patent owner refuses to license? In other words, is there any element of patent laws which makes it possible for a third party to use the standards covered by patents even though the owner of the patents refuses to grant a licence? There is a patent law principle which may compel the patent owner to deal with a third party even though the patentee does not want to. It is compulsory licences, described in the following section.

3.

COMPULSORY LICENCES

3.1

Introduction

A straightforward way to resolve the problems caused by the refusal to license by a patent owner is to impose a compulsory licence. A compulsory licence is defined as ‘an involuntary contract between a willing buyer and an unwilling seller imposed and enforced by the state’.43 In the context of patents and standards, the unwilling seller is the owner of a patent covering standards and the willing buyer is the party hoping to use the standards.

43 Julian-Arnold, ‘International Compulsory Licensing: The Rationales and the Reality’, 349.

108


Various provisions provide the grounds for compulsory licences of patents. Compulsory licences may be forced by patent laws in those cases where the relevant patents are not worked or the demand for the product covered by the patents is not met, where a dependent patent cannot be worked without the licences of the first patent, and where the patent is related to food and medicine.44 A compulsory licence may also be imposed on a patent owner as a remedy for the violation of competition laws or antitrust laws.45 In the United States, statutory provisions of the Atomic Energy Act and the Clean Air Act provide the grounds for compulsory licences.46 In the following sub-sections, the grounds for compulsory licences are studied and whether compulsory licences can resolve the conflict between patents and standards is examined. 3.2

International Framework

The Paris Convention for the Protection of Industrial Property (hereinafter ‘Paris Convention’) has provisions on compulsory licences of patents. Article 5 section A (2)–(4) of the Paris Convention states that: (2) Each country of the Union shall have the right to take legislative measures providing for the grant of compulsory licenses to prevent the abuses which might result from the exercise of the exclusive rights conferred by the patent, for example, failure to work. (3) Forfeiture of the patent shall not be provided for except in cases where the grant of compulsory licenses would not have been sufficient to prevent the said abuses. No proceedings for the forfeiture or revocation of a patent may be instituted before the expiration of two years from the grant of the first compulsory license. (4) A compulsory license may not be applied for on the ground of failure to work or insufficient working before the expiration of a period of four years from the date of filing of the patent application or three years from the date of the grant of the patent, whichever period expires last; it shall be refused if the patentee justifies his inaction by legitimate reasons. Such a compulsory license shall be non-exclusive and shall not be transferable, even in the form of the grant of a sub-license, except with that part of the enterprise or goodwill which exploits such license.47

44 Joseph A. Yosick, ‘Compulsory Patent Licensing for Efficient Use of Inventions’, University of Illinois Law Review 2001 (2001): 1279–82. 45 Ibid., 1282–4. 46 Cole M. Fauver, ‘Compulsory Patent Licensing in the United States: An Idea whose Time has Come’, Northwestern Journal of International Law and Business 8 (1988): 670–71. 47 Paris Convention for the Protection of Industrial Property, (20 March 1884), Article 5 A(2)–(4).


109

The Agreement on Trade-Related Aspects of Intellectual Property Rights 1994 (hereinafter ‘TRIPS Agreement’) also has provisions on compulsory licences. Article 31 of the TRIPS Agreement makes it more onerous to obtain a compulsory licence, providing that: Where the law of a Member allows for other use48 of the subject matter of a patent without the authorization of the right holder, including use by the government or third parties authorized by the government, the following provisions shall be respected: (a) authorization of such use shall be considered on its individual merits; (b) such use may only be permitted if, prior to such use, the proposed user has made efforts to obtain authorization from the right holder on reasonable commercial terms and conditions and that such efforts have not been successful within a reasonable period of time. This requirement may be waived by a Member in the case of a national emergency or other circumstances of extreme urgency or in cases of public non-commercial use. In situations of national emergency or other circumstances of extreme urgency, the right holder shall, nevertheless, be notified as soon as reasonably practicable. In the case of public non-commercial use, where the government or contractor, without making a patent search, knows or has demonstrable grounds to know that a valid patent is or will be used by or for the government, the right holder shall be informed promptly; (c) the scope and duration of such use shall be limited to the purpose for which it was authorized, and in the case of semi-conductor technology shall only be for public non-commercial use or to remedy a practice determined after judicial or administrative process to be anti-competitive; (d) such use shall be non-exclusive; (e) such use shall be non-assignable, except with that part of the enterprise or goodwill which enjoys such use; (f) any such use shall be authorized predominantly for the supply of the domestic market of the Member authorizing such use; (g) authorization for such use shall be liable, subject to adequate protection of the legitimate interests of the persons so authorized, to be terminated if and when the circumstances which led to it cease to exist and are unlikely to recur. The competent authority shall have the authority to review, upon motivated request, the continued existence of these circumstances; (h) the right holder shall be paid adequate remuneration in the circumstances of each case, taking into account the economic value of the authorization; (i) the legal validity of any decision relating to the authorization of such use shall be subject to judicial review or other independent review by a distinct higher authority in that Member; (j) any decision relating to the remuneration provided in respect of such use shall be subject to judicial review or other independent review by a distinct higher authority in that Member; (k) Members are not obliged to apply the conditions set forth in subparagraphs (b) and (f) where such use is permitted to remedy a practice determined after judicial or administrative process to be anti-competitive. The need to correct anti-competitive

48

‘Other use’ refers to use other than that allowed under Article 30.

110


practices may be taken into account in determining the amount of remuneration in such cases. Competent authorities shall have the authority to refuse termination of authorization if and when the conditions which led to such authorization are likely to recur; (l) where such use is authorized to permit the exploitation of a patent (‘the second patent’) which cannot be exploited without infringing another patent (‘the first patent’), the following additional conditions shall apply: (i) the invention claimed in the second patent shall involve an important technical advance of considerable economic significance in relation to the invention claimed in the first patent; (ii) the owner of the first patent shall be entitled to a cross-licence on reasonable terms to use the invention claimed in the second patent; and (iii) the use authorized in respect of the first patent shall be non-assignable except with the assignment of the second patent.49

Within the boundary granted by the international agreements, the governments of most countries enacted the patent provisions on compulsory licences. There are three situations under which compulsory licences can be imposed on patent owners: (1) where the patent is not worked or the demand for the product covered by the patents is not met, (2) where compulsory licences are essential for the public interest, and (3) where the licences are necessary for a dependent patent, which cannot be exploited without infringing another patent, to be worked. 3.3

Grounds for Compulsory Licences

3.3.1 Insufficient use Where the patent owner fails to exploit the patents or where the patents are not worked enough to meet the demand for the products covered by the patents, compulsory licences may be imposed on the patent owner. The rationale for these sorts of compulsory licences is that the exclusive rights to exploit a new invention are granted to a patentee in exchange for the benefits produced by the exploitation of the new invention. Thus, if the inventor does not provide the benefits from the exploitation of the invention, the exclusive rights conferred by the patents need to be limited to some extent, and a compulsory licence, which makes the invention available to society, is the necessary limitation.50 Third parties hoping to obtain compulsory licences on this ground need to wait four years from the date of patent application or three years from the date of patent grant, whichever period expires last, before they apply for

49 Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 31. 50 Julian-Arnold, ‘International Compulsory Licensing: The Rationales and the Reality’, 351–2.


111

compulsory licences.51 Where a patentee justifies the non-work or insufficient use by legitimate reasons, compulsory licences are not granted.52 There are various reasons why patentees do not fully exploit the patented inventions.53 Firstly, the working of the invention may not be profitable. Where there is no demand for the product incorporating the patented invention in the market, where the cost of production of the invention is so expensive that profits are not expected, or where the patentee underestimates the value expected from the exploitation of the invention, the invention is not likely to be worked. Secondly, although commercialisation of the invention is possible, the patentee would not exploit it in those circumstances where other technologies are more promising than the patented invention. Thirdly, while the patented technology is commercially feasible, it may be impossible for the patent owner to apply the patented technology to the field which he is concerned with. In all of these cases, if the patent owner is rational, he is likely to grant licences to third parties hoping to use the invention because the licences provide the benefit of licensing fees, and thus compulsory licences may not have to be considered. However, there are some cases where the patent owner does not exploit the patented technology, even though the exploitation may increase social welfare. For instance, where the product manufactured by the patented technology would compete with the products which the patentee hopes to keep commercialising, the patented invention is not likely to be exploited, no matter how beneficial it is to society. In those cases, the private benefits are conflicting with the social benefits and the patent is likely to be used to suppress the exploitation of the patented technology by others. It is this circumstance where compulsory licences may be used. 3.3.2 Public interest In many countries, compulsory licences may be forced on the basis of the public interest. In those cases where patents are related to food and medicine, compulsory licences may be granted. In particular, producing the medicine to cure or combat deadly diseases can be the main reason why a country imposes a compulsory licence on patent owners. Indeed, the United States, where there is no provision on compulsory licences in patent laws, considered a compulsory licence of the medicine for anthrax when people were killed and threatened by mail contaminated by anthrax bacteria, and South Africa enacted compulsory licensing legislation to produce HIV/AIDS drugs at a cheap price, 51 52 53

Paris Convention for the Protection of Industrial Property, Article 5 A(4). Ibid. Julie S. Turner, ‘The Nonmanufacturing Patent Owner: Toward a Theory of Efficient Infringement’, California Law Review 86 (1998): 182–6.

112


which evoked a heated debate between the developed countries and the developing countries.54 More recently, some countries have considered compulsory licensing of patents covering an anti-flu drug when swine flu is raging fiercely. As the Clean Air Act and the Atomic Energy Act of the United States show, pollution control, national security and energy supply can be reasons to impose compulsory licences.55 In addition, as described in Chapters 3 and 4, a compulsory licence can also be forced as a remedy for antitrust violation or an abuse of patent rights, which is another case where compulsory licences are imposed in the public interest. 3.3.3 The dependent patent Article 31 (l) of the TRIPS Agreement provides that where a patent (‘the dependent patent’) cannot be utilised without infringing another patent (‘the first patent’), a compulsory licence may be permitted on condition that the dependent patent is technically advanced with considerable economic significance in relation to the first patent, that the owner of the first patent obtains a cross-licence on reasonable terms to use the invention claimed in the dependent patent, and that the compulsory use of the first patent is non-assignable.56 This ground for compulsory licences is related to the hold-up problem between the first patent and the improvement patent. Even though the dependent patent cannot be exploited without the licences of the first patent, it is not always desirable to allow the owner of the dependent patent to obtain compulsory licences automatically, because compulsory licences will restrict the exclusive rights granted to the owner of the first patent, reducing the incentive to invent the frontier technology. The compelled dealing seems to be justified in those cases where the expected value obtained from the exploitation of the improvement patent is far greater than the expected value of the first patent. The condition of ‘an important technical advance of considerable economic significance in relation to the invention claimed in the first patent’ in Article 31 (l) demonstrates this point and requires that the dependent patent should be more advanced than the first patent and thus create large value.

54

Anthony P. Valach, ‘TRIPS: Protecting the Rights of Patent Holders and Addressing Public Health Issues in Developing Countries’, Chicago-Kent Journal of Intellectual Property 4 (2005). 55 Fauver, ‘Compulsory Patent Licensing in the United States: An Idea whose Time has Come’, 670. 56 Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 31 (l).


113

3.3.4 Review and implication There are three major grounds for compulsory licences under the Paris Convention and the TRIPS Agreement. They are the public interest, dependent patents and insufficient use. Then, to what extent can the compulsory licences under patent laws be applied to the conflict between patents and standards? It is improbable that compulsory licences due to the public interest will become an issue in relation to patents covering standards, because, as far as patent laws are concerned, the patents related to the public interest are mostly patents protecting the technologies for pollution control, medicine, food, energy and weapons. Admittedly, standards have public characteristics but it is unlikely that the patented inventions incorporated into standards will be regarded as inventions related to the public interest under patent laws. The patents related to the public interest under patent laws have to affect the general public more extensively than the patents covering standards do. For instance, patents for food and medicine may have considerable influence on human beings as a whole. However, the influence of patents covering standards is mostly limited to the relevant markets and the related innovation trajectories. Therefore, it is unlikely that compulsory licences resulting from the public interest will be imposed on the patents covering standards. As for insufficient use, the patents incorporated in standards are not likely to be unused or insufficiently used because most standards are widely used in the market. Thus, as to the patents which are already incorporated in standards, it is unlikely that compulsory licences would be granted on account of insufficient use. However, in some cases, compulsory licences can be considered in relation to standards for this reason. As described in Section 3.3.1 of this chapter, patent rights can be enforced to prevent others from using the patented technology, although the invention is not exploited by the patent owner. That is, the patent owner may refuse to license the unexploited patents in order to prevent the appearance of a standard for goods or services which will compete with those that the patent owner is providing. This instance can be divided into two categories depending on who is implementing the standard at stake. Firstly, the standard at issue can be implemented by an SSO. In this case, a compulsory licence of the patent on account of insufficient use, which is granted to the SSO, is of limited use for the standard. According to the Paris Convention, a compulsory licence of a patent ‘shall not be transferable, even in the form of the grant of a sub-licence’57 and the TRIPS Agreement also prescribes that a compulsory licence of a patent ‘shall be non-assignable’.58 In addition, according to the TRIPS Agreement, a 57 58

Paris Convention for the Protection of Industrial Property, A(4). Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 31(e).

114


compulsory licence ‘shall be considered on its individual merits’.59 Therefore, although the SSO implements the standard with a compulsory licence, it cannot grant a sub-licence to a firm hoping to produce goods incorporating the standard. The firm has to negotiate with the patent owner for a licence and where the patent owner refuses to license the patent covering the standards, the firm has to seek to obtain a compulsory licence again. This clearly impedes the widespread use of the standard. Secondly, a firm rather than an SSO may seek to create an informal standard in markets by using the unexploited patent. In this case, the firm may obtain a compulsory licence for the unworked patent. However, it is extremely difficult for a firm to get its technology accepted as a standard in the markets without granting licences to other firms, because the critical mass, which is the minimum number of customers needed to initiate positive feedback, will not be gained without cooperation with other firms producing the same or compatible goods. Thus, the firm obtaining a compulsory licence has to be able to grant a sub-licence of the patent to other firms in order to make its technology a standard in the markets. However, this is not allowed under the Paris Convention and the TRIPS Agreement. Therefore, where the patent owner refuses to license the unexploited patents in order to prevent the appearance of a standard, it is not likely that a compulsory licence due to insufficient use would resolve this problem. As far as the dependent patent is concerned, compulsory licences on dependent patents may be applied to the patent issues in standards. The problem of the dependent patents can arise in standards incorporating patents and indeed is closely related to the bargaining breakdown between the owner of patents covering standards and a user of the standards. There is significant imbalance of the bargaining power between the owner of standards and a user, which is likely to result in asymmetric agreements favouring the patent owner. Thus, it is necessary to increase the bargaining power of a standard user for fair bargaining between them, and it is the provision of compulsory licences on the dependent patent which can increase the bargaining power of a standard user. According to the TRIPS Agreement, where the dependent patent cannot be exploited without infringing the first patent and where the first patent owner refuses to license the first patent, a compulsory licence can be granted only in those circumstances where the dependent patent has ‘an important technical advance of considerable economic significance in relation to the invention claimed in the first patent’.60 However, what constitutes the ‘important technical advance of considerable economic significance’ is not clear. It seems that

59 60

Ibid., Article 31(a). Ibid., Article 31 (l).


115

the condition is established in those cases where the value created by the dependent patents is far greater than that of the original patent, and thus the radical improver of the first patent may meet this condition.61 It is clear that a minor improver is not entitled to obtain compulsory licences on the basis of this provision. As far as a significant improver is concerned, it is not clear whether he can obtain a compulsory licence depending on this provision. The condition of ‘an important technical advance of considerable economic significance’ seems to imply that a compulsory licence is only granted in those exceptional cases where the value created by the compelled deal is so great that it offsets the decrease in the incentive to innovate and provides a proper ground for restricting the freedom of contract. Therefore, it seems that a significantly improved invention is not enough for a compulsory licence. Accordingly, as far as the dependent patent is concerned, a compulsory licence may only be granted in those circumstances where the invention covered by the dependent patent is a radical improvement on the invention protected by the first patent. However, another difficulty in applying this sort of compulsory licence is that the invention protected by the dependent patent has to be evaluated ex ante. That is, since the radically improved invention, which is protected by the dependent patent, cannot be exploited without licences of the first patent, whether the dependent patent meets the condition of ‘an important technical advance of considerable economic significance’ has to be decided before it is used, which is clearly difficult. This reasoning on dependent patents can be applied to the patent issues in standards. According to the above reasoning, a third party hoping to obtain a licence from the owner of patents covering standards can be granted a compulsory licence only in those cases where the third party has a dependent patent protecting an invention which is a radical improvement on the patented invention covering standards. Therefore, there can be minor, moderate or even significant improvements of standards which are not eligible to obtain a compulsory licence when the patent owner refuses to license. In addition, it is not clear how to define the radical improvement in relation to standards prior to the exploitation of the improved invention. Accordingly, there are many uncertainties in applying patent laws to the patent issues in standards and thus compulsory licences will be used in highly limited circumstances. It is interesting to note the difference between this outcome and the one derived from the application of competition laws. As noted in Chapter 4, competition laws may grant compulsory licences only in those cases where the standards are mandatory or the firm concerned is in a super-dominant position

61 For a minor improver, a significant improver and a radical improver, see Section 2.3 of this chapter.

116


and where a party hoping to obtain licences plans to supply new goods or services which are different from those provided by the patent owner. It seems that competition laws care more about the market competition than the technological innovation, because competition laws do not provide any ground for compulsory licences in relation to dependent patents. Where there is competition in the market, competition laws do not grant a compulsory licence to the owner of a dependent patent, even though the dependent patent will lead to a great technical advance. In contrast, as far as patent laws are concerned, what matters is whether the dependent patent is a radical improvement or not. Patent laws focus on the technical innovation and thus, in granting compulsory licences, competition in the relevant market is not a main concern. Therefore, on the one hand, competition laws seem to have narrower application than patent laws in relation to patent issues in standards. On the other hand, patent laws have more limitations than competition laws in imposing a compulsory licence on patent owners, because the party hoping to obtain a licence should have its own patent, the dependent patent, and, on top of that, the dependent patent has to be a radical improvement on the patent covering standards. Thus, patent laws and competition laws provide their own solutions to the conflict between patents and standards to some extent. Accordingly, it is wrong to argue that patent laws have a broader application to the patent issues in standards than competition laws or vice versa. As described above, compulsory licences can only be used in limited circumstances and the reverse doctrine of equivalents is not likely to resolve the conflict between patents and standards. It has to be noted that competition law or antitrust law has limitations in relation to the patent issues in standards, as is argued in Chapters 3 and 4. Given that both legal regimes have limitations, there are several approaches to find a way to resolve the conflict. One may seek to resolve the conflict by fine-tuning patent law, by modifying the existing competition law, or by revising both patent law and competition law. This book adopts the first approach, by taking the position that, given that applying the competition law has many problems and limitations, it may be desirable to find the right balance between standard users and standard owners within the scope of intellectual property law itself, as some commentators seek to do.62 Here it is sought to find the solution by answering the question of whether the solutions provided by existing patent laws are appropriate. To answer the question, it needs to be examined whether there are problems in the 62 Michael L. Katz, ‘Intellectual Property Rights and Antitrust Policy: Four Principles for a Complex World’, Journal of Telecommunications & High Technology Law 1 (2002); Burton Ong, ‘Anticompetitive Refusal to Grant Copyright Licences: Reflections on the IMS Saga’, European Intellectual Property Review 26, no. 11 (2004).


117

measures which current patent laws provide to deal with the issues of the conflict between patents and standards. If this is the case, it can be concluded that current patent systems need to be fine-tuned. In answering the question, it is essential to understand how patent systems work to promote technological innovation and what are the costs and benefits of patent systems, which is studied in the next chapter.

6. Costs and benefits of patent systems 1.

INTRODUCTION

The competition laws and the patent laws provide limited solutions to the patent issues in industry standards, as described in previous chapters. That is, the exceptional circumstances doctrine of competition laws may be applied in those cases where the standards at issue are mandatory or the firm concerned is in a super-dominant position and where a party hoping to obtain licences plans to supply new goods or services which are different from those provided by the patent owner, but it is not likely that the doctrine will be used in other cases. As far as patent laws are concerned, compulsory licences can only be used in limited circumstances and the reverse doctrine of equivalents is not likely to resolve the conflict between patents and standards. Then comes the question of whether there are problems in the solutions provided by existing laws and, if so, whether the existing laws can be modified to provide better solutions. There can be various approaches to these questions. The answer may be sought from the perspective of maximising static efficiencies or promoting market competition. In this book, however, ways to resolve the issues of patents and standards are sought from the perspective of patent laws. That is, it will be sought to find ways to resolve the conflict between patents and standards from the perspective of technological innovation. In order to do so, it is essential to study the functions of patent systems and the costs and benefits of patent systems. By analysing the costs and benefits of patent systems in relation to standards, it will be possible to fine-tune patent systems in such a manner as to increase the benefits and/or decrease the costs. As is argued later in this chapter, the costs of patent systems are closely related to the benefits and thus the attempt to reduce the costs may also diminish the benefits or, alternatively, increasing the benefits of patent systems may also enlarge the costs. Therefore, any measure needs to be examined with respect to both the costs and the benefits of patent systems. The approach is as follows. In this chapter, the general costs and benefits of patent systems are examined and then the cost structure of patent systems in relation to standards is analysed. In the next chapter, ways to reduce the costs of patent systems regarding standards are suggested, and lastly the suggested solutions are examined with respect to the benefits of patent 118

Costs and benefits of patent systems

119

systems to see whether the reduction of costs outweighs that of benefits, which is clearly desirable.

2.

COSTS AND BENEFITS OF PATENT SYSTEMS

2.1

Costs

One of the significant costs of patent systems is the resource misallocation generated by monopolistic pricing of the patented invention.1 Suppose that a market is monopolised by a supplier of a product who owns patents covering the product. As the economic theories teach, the monopolist can obtain maximum profits by producing the product to the amount, say Q1, where the marginal cost equals the marginal revenue.2 Thus, the product protected by patents is produced to the amount of Q1. At this point, the customers will obtain some benefits from the product and the patent owner can earn the monopoly profit. However, the monopoly profit is not a pure profit of the patent owner because the patent owner invested resources in R&D to invent the product. Therefore, the profit of the patent owner is the monopoly profit minus the R&D cost. If the monopoly profit is less than the R&D cost, the patent owner cannot recoup the investment cost and will not make an effort to invent or innovate any more. Conversely, where the monopoly profit is large enough to cover the cost incurred during the R&D, the patent owner is likely to invest. On the other hand, in a perfectly competitive market, the price of a product equals the marginal cost of it and the product is produced to an amount, say Q2, which is greater than Q1.3 At this point consumers can buy the product at a price lower than the monopoly price. In addition, the total social welfare in the perfectly competitive market is greater than that of the monopolistic market. Therefore, the difference between the social welfare of a perfectly competitive market and that of a monopolistic market is the loss of the social welfare resulting from resource misallocation by the monopolistic pricing, which is called the ‘deadweight loss’. It has to be noted that, while the consumers can buy the product at a cheaper price in the perfectly competitive market, the patent owner cannot recoup the R&D cost. Therefore, it is likely that there will not be enough incentives for an economic entity to carry on

1 F.M. Scherer, Industrial Market Structure and Economic Performance, 2nd edn (Chicago: Rand McNally College Publishing Company, 1980) 450–51. 2 David Begg et al., Economics, 8th edn (London: The McGraw-Hill Companies, 2005) Ch. 6. 3 Ibid. Ch. 8.

120


inventive activities in a perfectly competitive market. Thus, perfect competition is not conducive to increasing dynamic efficiency and this is one of the reasons why Schumpeter argued that ‘perfect competition is not only impossible but inferior.’4 Thus the comparison between the monopolistic market and the perfectly competitive market shows the conflict between dynamic efficiency and static efficiency caused by patent protection. While patent protection is necessary to encourage economic entities to participate in inventive or innovative activities, it entails a loss in social welfare. In reality, it is rare that a patent right creates market power, providing monopoly profits to its owner.5 Most patents fail to create market power in related markets and, in many cases, competitors are capable of inventing around the patented technology and introducing competing products into the market. Therefore, in some sense, the main function of patent protection is not to provide monopoly profits to patent owners but to prevent others from copying the patented technology. When the patented technology covers standards, however, patent rights are likely to create market power. Competitors cannot invent around the standards because, without incorporating the standards, they cannot make their products compatible with other products and because, where a product is not compatible with other products in the network market, customers do not want to buy it. Thus, incorporating the standards is essential for a firm to enter the network market. This means that the owner of patents which cover the standards can create high barriers to entry into the market which the standards are concerned with. Consequently, the patent owner can increase his market power, prevent competitors or potential competitors from entering the market, and leverage his monopoly power to other markets. In some cases, a would-be standard is not widely accepted by consumers and thus fails to create market power, as Sony’s mini-disk standard does. If this is the case, the patent covering the standard neither builds high barriers to entry nor creates market power in the relevant markets. There is no reason why special attention needs to be paid to these sorts of standards, because the harm caused by them is not very different from that caused by the majority of patents which do not create market power and are merely used to prevent illegal copies of the patented technology. They can be treated well within the existing legal framework. It is the standard which creates market power in the relevant market that this book is concerned with. Thus, it is reasonable to assume in this book that standards are widely accepted in the market and that 4 Joseph A. Schumpeter, Capitalism, Socialism, and Democracy (London: Henderson and Spalding, 1943) 106. 5 Edmund W. Kitch, ‘Elementary and Persistent Errors in the Economic Analysis of Intellectual Property’, Vanderbilt Law Review 53 (2000).


121

the patents covering the standards can be used to erect barriers to entry into the market. Thus, the owner of patents covering standards can charge a price higher than that in the competitive market, even though the price may be lower than the monopolistic price. Consequently, the resource misallocation, including the deadweight loss, caused by the market power of the patent owner is the cost of patent protection in relation to standards. Secondly, the technology suppression and the limitation on the widespread use of the patented idea are entailed by patent protection. As is described later in this chapter, the bargaining between a patent owner and the improver is not always successful. The bargaining breakdown between the owner of the first patent and its improver can give rise to the suppression of the improving technology. Especially where the first patent has great possibilities to be developed in various ways, the bargaining breakdown can cause serious harm to technical development. The patents of the Wright brothers are an example of this.6 While the Wright brothers greatly contributed to technical advancement by inventing the powered flying machine, they also delayed further advancement of airplane technology. They obtained a basic patent on the stabilising and steering system of an airplane and sought to develop the system by themselves.7 They refused to grant licences to others who, by improving the Wright brothers’ system, hoped to enter the aircraft market, which entailed litigation between them and the improvers.8 The Wright brothers sought to prevent the emergence of competing technologies by exercising their patent rights, which demonstrates that patents can be used to suppress beneficial technologies. In addition, a patentee may accumulate a considerable number of patents around a technology in order to prevent others from using it.9 It is not common for a single patent or a small number of patents to cover every possible variation of a technology and thus in many cases it is possible to invent around the patented technology. Therefore, it is common that a firm files a number of patent applications to fence in a technology which is considered important. By accumulating a considerable portfolio of patents and by exercising the patent rights, a firm can prevent the appearance of competing technologies. As far as standards are concerned, the patent covering standards has great possibilities to be further developed in various ways because standards are the starting point from which many improvement technologies are developed.

6 Robert P. Merges and Richard R. Nelson, ‘On Limiting or Encouraging Rivalry in Technical Progress: The Effect of Patent Scope Decisions’, Journal of Economic Behavior and Organization 25 (1994): 13–16. 7 Ibid. 8 Ibid. 9 Scherer, Industrial Market Structure and Economic Performance 451.

122


Therefore, a bargaining breakdown between the owner of patents covering standards and the user of standards can cause serious harm to technical development. Furthermore, the owner of patents covering standards may accumulate a considerable number of patents around the standards to have a sizeable patent portfolio and seek to suppress competitors. Therefore, technology suppression by patent rights is the social cost of the patent system in relation to standards. Thirdly, the administration cost is another main cost of patent systems. In order to obtain patents, inventors have to file a patent application in the Patent Office. It is possible for inventors to file an application by themselves, but preparing the necessary documents and setting up the proper patent scope need special skills. Thus, it is common that inventors hire legal specialists such as patent attorneys and have them manage the patent application, which incurs high costs. Once filed, the patent application is handled by the Patent Office. The Patent Office manages the procedures from application to rejection or issuance of patents and the renewal of the registered patents. It is clear that the Patent Office needs a budget to pay salaries to its employees, manage the documents, maintain the information systems and so on, which is not trivial. For instance, the European Patent Office spent 336.9 million euros for fiscal year 200510 and the budget of the United States Patent and Trademark Office for fiscal year 2006 was 1.68 billion dollars.11 On top of that, enforcing the patent rights through litigation costs tremendous amounts of money. It was estimated that in the US the total cost of patent application and patent litigation accounted for a quarter of the cost of industrial basic research.12 However, it is difficult to ascertain the precise relationship between patents covering standards and the administration cost. It is not likely that the patent covering standards increases or reduces administration costs. It can arguably be concluded that the patents covering standards do not significantly affect the amount of administration costs. Nevertheless, administration costs are an important element of social costs resulting from patent systems. Besides the costs described above, some have argued that patent systems may divert productive activities from research in unprotected fields to research in fields where the patent protection can be obtained,13 which generates social

10 11

‘Annual Report 2005’ (European Patent Office, 2005), 66. ‘Performance and Accountability Report Fiscal Year 2006’, (United States Patent and Trademark Office, 2006), 60. 12 William Kingston, ‘Innovation Needs Patents Reform’, Research Policy 30 (2001): 409–10. 13 Fritz Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights of the Committee on the Judiciary, United States Senate (U.S. Government Printing Office, 1958), 24.


123

costs, and that the desire to obtain patent rights may cause patent races between competitors,14 which also entails social costs. Indeed, patent systems give rise to various social costs, including costs not mentioned in this sub-section, but it is impossible to make an exhaustive list of costs resulting from patent systems. In any case, patent systems entail social costs and, especially where patents cover standards, the costs resulting from the patents may be considerable because the patents are more likely to create market power, distorting competition in the market. 2.2

Benefits

The benefits of patent systems are well described in the justification theories for patent systems. The theories for patent systems argue for the necessity of patent systems in various ways. Some argue for patent systems from the perspective of natural rights, while others contend that patents have to be given to inventors because it is an injustice to let others use the inventions without a fair share of the reward in return.15 However, it is generally accepted that the patent systems are justified for economic reasons,16 as described in the following sub-sections. 2.2.1 Economic justification of patent systems The economic rationales for patent systems are basically incentive-based arguments. According to these theories, economic incentives are necessary to drive economic entities to participate in inventive activities which they would not otherwise do.17 Broadly speaking, these arguments can be divided into two groups. The first group is composed of the incentive to invent theory, the incentive to disclose theory and the innovation enhanced competition theory. This group of theories focuses on the need for patent systems before the invention is created and thus they may be called ex ante justification theories of

14 Birgitte Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’ (2003), 21. 15 Birgitte Andersen, ‘If “Intellectual Property Rights” is the Answer, What is the Question? Revisiting the Patent Controversies’, Economics of Innovation and New Technology 13 (5) (2004): 420. 16 Alan S. Gutterman, Innovation and Competition Policy: A Comparative Study of the Regulation of Patent Licensing and Collaborative Research and Development in the United States and the European Community (Kluwer Law International, 1997) 129–30; Paul L.C. Torremans, Holyoak and Torremans: Intellectual Property Law, 4th edn (Oxford: Oxford University Press, 2005) 11–25. 17 Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights, 20–21.

124


patent systems.18 According to these theories, the incentives are essential to make the inventions come into being and thus, after the creation of the inventions, the incentives have completed their mission and the cost associated with the enforcement of the patent rights is an undesirable price which society has to pay.19 Conversely, the other group of theories, which may be called the ex post justification theories of patent systems, emphasise the role of patent rights after inventions are created and argue that exclusive rights are necessary in order for the right holders to manage and exploit the inventions efficiently.20 These theories are also incentive-based arguments but the incentives which these theories are concerned with are those necessary after the inventions have been created. The prospect theory belongs to this.21 However, while it may be helpful in understanding the function of patent systems to classify the justification theories into these two groups, the distinction is not correct in the strict sense because both the ex ante and ex post aspects of patent systems are mixed in each group of theories. In any case, the benefits of patent systems provided by the justification theories of patent systems are as follows. Firstly, patent systems encourage economic entities to participate in inventive activities by providing the incentive to invent, as the incentive to invent theory argues. The argument of this theory is that inventions are necessary for industrial progress, but that inventions will not be created in the absence of the exclusive rights because inventions are likely to be appropriated by competitors who free ride on the inventor’s effort without sharing in the costs of inventions.22 The profits which may be gained from the competitive exploitation of inventions are not sufficient for the inventor to recoup the R&D cost, which causes underinvestment in inventions.23 This underinvestment problem can be resolved by patent systems which grant temporary monopolies in the form of exclusive rights to inventors.24 By using the exclusive rights, the inventors may extract a price which is higher than that in the competitive market and thus can be in a better position to recoup the R&D expenditure. Thus patent 18 Mark A. Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’, University of Chicago Law Review 71 (2004). 19 Rebecca S. Eisenberg, ‘Patents and the Progress of Science: Exclusive Rights and Experimental Use’, University of Chicago Law Review 56 (1989): 1037–8. 20 Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’. 21 Ibid. 22 Eisenberg, ‘Patents and the Progress of Science: Exclusive Rights and Experimental Use’, 1024–8. 23 Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights, 20–21. 24 Ibid.


125

systems increase the profit expectation by granting temporary monopolies in the form of exclusive rights to inventors, which leads individuals or firms to make more inventions.25 Secondly, the incentive to disclose theory argues that patent systems provide the incentives necessary to encourage the owners of useful technologies to disclose them. This theory assumes that without patent protection inventors will not disclose inventions but will instead keep them secret in order to prevent competitors from exploiting them.26 If the inventions are not disclosed, useful inventions will not become available or the exploitation of the inventions will be significantly delayed, which may cause wasteful duplicative research and deprive society of the opportunity to use the inventions.27 Hence, society needs to bargain with the inventors to make them disclose the inventions and the patent systems are the outcome of the bargain. Thirdly, it is argued that innovation enhances competition in the long term. Since knowledge can be consumed by anyone who knows it without preventing others from using it simultaneously, it is considered that knowledge has the qualities of public goods, the non-rival and non-excludable aspects.28 It is easily copied and spread, making it easy for imitators to free ride on the inventor’s effort and making it hard for the inventor to control it. Hence, in the absence of intellectual property protection, one is not willing to endeavour to create knowledge but is likely to wait for the emergence of new knowledge and copy it, which will prevent the appearance of new ideas. In order to resolve the inefficiency created by the non-rival and non-excludable aspects of knowledge, rival and exclusive aspects have to be introduced into knowledge. The exclusive right to use knowledge conferred by intellectual property rights provides knowledge with these aspects, which is claimed to enhance competition in the long term by providing an incentive to invent new knowledge.29 It is contended that, just as property rights promote competition in the production of goods by restricting the consumption of products, intellectual property rights increase competition at the innovation level by limiting the production of goods.30 That is, the limitation of the production of goods may distort market competition in the short term, but the limitation increases dynamic efficiency at the innovation level, promoting market competition in the long term, 25 Richard A. Posner, Economic Analysis of Law, 7th edn (Wolters Kluwer, 2007) 38–40, 279–83. 26 Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’, 34–5. 27 Torremans, Holyoak and Torremans: Intellectual Property Law 20–23. 28 Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’, 129–30. 29 Torremans, Holyoak and Torremans: Intellectual Property Law 13–20. 30 Ibid.

126


since, without the limitation, innovative activities will be significantly diminished due to the free-rider problem. Thus, this theory takes the position that the restrictions on competition created by patent rights are only justified in so far as the restrictions promote innovative activities.31 Lastly, it is argued that patent rights are necessary in order for the right holders to manage and exploit their inventions efficiently after the inventions are created.32 The fundamental basis of such arguments is the tragedy of the commons theory.33 This theory argues that common property, such as pastures and lakes, open to the public is likely to be overused because the private profits expected by the overuse in the first place outweigh the private cost anticipated by the overuse.34 The theory argues that the best way to prevent the tragedy is to internalise the costs and benefits of using valuable resources by granting property rights of the commons to individuals. Property rights enable their owners to exclude others from using the resources, protecting the resources from being overused by others. In addition, the owners of property rights will not overuse the resources but rather use them effectively because the property owners are in a position to realise the full costs and benefits of the exploitation of the property. If the transactions between the right holders are costless, the owners of the property rights will trade with each other to produce the best possible use.35 The idea of the tragedy of the commons theory is used to justify intellectual property systems by Edmund Kitch. The argument of the tragedy of the commons in relation to intellectual property is that, just as property rights need to be granted to tangible resources in order to facilitate the efficient use of scarce resources, property rights have to be provided to intangible resources for the same reason. However, whether there is a tragedy of the commons in inventions is not clear. While the tragedy of the commons arises in those cases where the resources are scarce or finite, inventions can be used to an unlimited extent without depletion.36 Unlike the overuse of tangible property, the overuse of inventions is not likely to cause a tragedy but is likely to lead to a comedy, increasing social welfare, because wide use of an invention leads to the introduction of a variety of products into markets and makes it easier for

31 32

Ibid. Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’,

129–33. 33 34 35

Ibid. Garrett Hardin, ‘The Tragedy of the Commons’, Science 162 (1968). Mark A. Lemley, ‘The Economics of Improvement in Intellectual Property Law’, Texas Law Review 75 (1997): 1045. 36 Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’, 23.


127

subsequent inventions to emerge. Edmund Kitch notes this point and argues that, while information may be used to an unlimited extent, the resources which can be used to exploit information are limited, and property rights given to inventions make it possible to effectively manage and exploit the inventions.37 Edmund Kitch has suggested the prospect theory by analysing the function of the patent system.38 This theory is along the same lines as the argument of the tragedy of the commons39 and assumes that there is no transaction cost, which can only be realised in the hypothetical framework called the Coasean theorem.40 While the tragedy of the commons theory claims that inventions need to be protected for the prevention of overuse,41 the prospect theory argues that patent protection is essential to encourage the patent owner to improve, maintain, or exploit the invention. The term ‘prospect theory’ comes from the analogy between the patents and awards of exclusive mineral claims in the American West. Kitch argues that just as the mineral claims, which provide property rights to the newly discovered valuable mineral for those who find it, encourage the landowner to exploit the land efficiently, patent rights promote further commercialisation and exploitation of inventions. Kitch maintains that [a] patent ‘prospect’ increases the efficiency with which investment in innovation can be managed. … Technological information is a resource which will not be efficiently used absent exclusive ownership. … The patent owner has an incentive to make investments to maximise the value of the patent without fear that the fruits of the investment will produce unpatentable information appropriable by competitors.42

Hence Kitch argues that exclusive rights over further exploitation of inventions such as improvements and marketing are essential. 2.2.2 Critiques of economic theories for patent systems The economic justification theories for patent systems are criticised in various ways.

37 See Edmund W. Kitch, ‘The Nature and Future of the Patent System’, Journal of Law and Economics 20 (1977). 38 Ibid. 39 See Hardin, ‘The Tragedy of the Commons’. 40 Ronald H. Coase, ‘The Problem of Social Cost’, Journal of Law and Economics 3 (1960). 41 Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’, 129–32. 42 Kitch, ‘The Nature and Future of the Patent System’, 276.

128


As for the incentive to invent theory, some argue that inventions will be made even without patents because a firm has to keep up with the technological progress of competitors or because the head start provides sufficient incentive to promote investment in R&D.43 It is also argued that the patent system may cause underinvestment in inventive activity because, even though inventions are socially created from the cumulative ideas of various inventors, a patent is granted to the person who happens to complete the invention, while others who participate in the social activity of inventing are excluded.44 This theory is also criticised for the heavy cost associated with patent systems. Patent rights limit the widespread use of the patented idea during the patent terms, resulting in social costs. Patent systems may divert productive activity from research in unprotected fields to research in fields where patent protection can be obtained.45 The desire to obtain patent rights may cause patent races between competitors,46 which entail social costs, and the patents owned by an inventor may reduce the incentives for inventive activities of those who could improve the patented invention.47 In addition, marketing, licensing of ideas and the bureaucracy necessary to administrate and enforce patent rights incur social costs.48 There are also many objections to the incentive to disclose theory. Some argue that inventions are social or collective and thus, even if an inventor seeks to keep the invention secret, other inventors will develop and exploit the same or a similar idea.49 It is also contended that secrecy cannot be successful because competitors will eventually find out the idea through reverse engineering or in other ways. In those cases where a technical idea can be exploited in secrecy by its inventors, the infringers will also use the technology in secret, which means that finding the infringement and enforcing the patent rights would be difficult.50 One other strong argument against this 43 Eisenberg, ‘Patents and the Progress of Science: Exclusive Rights and Experimental Use’, 1026–7. 44 Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’, 15–16. 45 Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights, 24. 46 Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’, 21. 47 Eisenberg, ‘Patents and the Progress of Science: Exclusive Rights and Experimental Use’, 1027–8. 48 Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’, 16–21. 49 Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights, 24–5. 50 Eisenberg, ‘Patents and the Progress of Science: Exclusive Rights and Experimental Use’, 1028–9.


129

theory is that patent rights will only be granted to protect an invention which will be disclosed anyhow and that inventors will not disclose their invention if they believe that keeping it secret will be successful.51 This argument is supported by empirical studies showing that secrecy is preferred to patent protection in many industries.52 The challengers to innovation enhanced competition theory argue that patent rights do not necessarily enhance competition or that patents may reduce the incentive to invent. Some commentators contend that rivalry is good for inventive progress and broad patent protection hinders the creation of inventions subsequent to a patent, blocking the appearance of improvement inventions.53 In many industries, technological advances are cumulative such that new technologies are developed on the basis of the existing technologies. Therefore, a patent protection that is too strong impedes the inventive process by blocking the appearance of secondary inventions, which destroys competition.54 Another argument against this theory is related to network externalities, which work on the demand side and occur when the value derived from the consumption of products is affected by the number of other people who use the products or compatible products.55 As the historical evidence of the QWERTY keyboard layout shows, where network externalities are strong in a market, the market is likely to be locked into dominant technologies.56 Where the dominant technologies are protected by patent rights in the market, the patent owner may use the patent rights to strengthen the lock-in effect.57 Patent rights can be exploited for anticompetitive purposes such as excessive rent creation or the elimination of competition. The prospect theory is also challenged. There are debatable assumptions in the prospect theory.58 Kitch assumes that

51 Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights, 24–25. 52 Richard C. Levin et al., ‘Appropriating the Returns from Industrial Research and Development’, Brookings Papers on Economic Activity 1987, no. 3 (1987). 53 Robert P. Merges and Richard R. Nelson, ‘On the Complex Economics of Patent Scope’, Columbia Law Review 90 (1990). 54 Ibid. 55 Mark A. Lemley and David McGowan, ‘Legal Implications of Network Economic Effects’, California Law Review 86 (1998). 56 See Paul A. David, ‘Clio and the Economics of QWERTY’, The American Economic Review 75, no. 2 (1985). 57 Andersen, ‘The Rationales for Intellectual Property Rights in the Electronic Age’, 28–9. 58 Mark A. Lemley, ‘Policy Levers in Patent Law’, Virginia Law Review 89 (2003): 1600-04.

130


[n]o one is likely to make significant investments searching for ways to increase the commercial value of a patent unless he has made previous arrangements with the owner of the patent. This puts the patent owner in a position to coordinate the search for technological and market enhancement of the patent’s value so that duplicative investments are not made and so that information is exchanged among the searchers. 59

Thus Kitch emphasises that patents function as a way to promote efficiency in research after the grant of patents, enabling the patent holder to monitor and control the research relevant to the patents. However, the efficiency in research can only be achieved with efficient licensing between patent owners and third parties hoping to obtain licences, and the efficient licensing is possible only on condition that the information on the patent rights is perfect, that there is no transaction cost and that the parties concerned are rational.60 The reality is, however, that the information on the patent rights is not perfect since the information on the identity and the scope of the relevant patents is often uncertain. Moreover, it is rare that transaction costs are zero since finding the patent owner, negotiating licensing conditions, and exchanging payment or services entail costs. Another fundamental assumption of this theory which is criticised is that the invention and its improvement would be available to the public at reasonable prices. Economic knowledge teaches that the patent owner has no reason to reduce the price without competition. Rather the patent owner is likely to set the price of his product higher than the marginal cost. Concerning this point, Kitch argues that patents in most cases do not confer market power and the patent owner is likely to face competition, which forces him to lower the price.61 However, it is not uncommon that the patented product is priced in excess of full costs required to produce and distribute the product.62 2.2.3 Summary of the benefits of patent systems As the theories for patent systems describe, patent systems provide the ex ante and/or ex post economic incentives necessary to encourage economic actors to conduct inventive activities which they would not otherwise do.63 Patent protection is necessary to promote inventions, to facilitate disclosure of inventions, to enhance market competition in the long term by improving dynamic

59 60 61

Kitch, ‘The Nature and Future of the Patent System’, 276. Lemley, ‘Policy Levers in Patent Law’, 1602. Kitch, ‘Elementary and Persistent Errors in the Economic Analysis of Intellectual Property’, 1729–38. 62 Scherer, Industrial Market Structure and Economic Performance 451. 63 Machlup, ‘An Economic Review of the Patent Systems’, Study of the Subcommittee on Patents, Trademarks, and Copyrights, 20–21.


131

efficiency, and to encourage the exploitation of inventions. However, as the critiques of the theories for patent systems argue, inventions may be made and exploited even without patents because a firm has to keep up with the technological progress of competitors or because the head start provides sufficient incentive to promote investment in R&D. In addition, inventions may eventually be disclosed without patent systems or patent rights will only be granted to protect technology which will be disclosed anyhow. Strictly speaking, to find the pure benefits of patent systems, it woould be necessary to measure the inventions which would not be made, disclosed or put into use without patent protection. However, that is impossible. All in all, even though none of the theories for patent systems are without criticisms and objections, they show how patents provide the economic incentive to encourage economic entities to participate in inventive activities. None of the objections to the theories justifying patent systems are successful in arguing that patent systems have to be abolished. While the objections to these theories show that some aspects of patent systems cannot be justified economically, what they imply is not that patent systems should be abolished but that patent systems need to be fine-tuned.64 2.3

Review of the Costs and Benefits of Patent Systems

The costs of patent systems are closely related to the benefits of patent systems. For instance, on the one hand, the resource misallocation caused by monopoly pricing and the technology suppression entailed by the exclusivity of patent rights are clearly costs of patent systems, but, on the other hand, they constitute the incentives of patent rights, which lure economic entities to invent or innovate. That is, the expectation of monopoly profits and the rights to exclude competitors are attractive reasons to take the risk of inventions and innovations. Similarly, the administration costs of patent systems are essential to establish the legal rights and to enforce the rights. Therefore, without the administration costs, it is impossible to achieve the benefits of patent systems, gained from the exclusivity of patent rights. Thus the costs of patent systems are inter-related with the benefits of patent systems. Accordingly, it has to be noted that a measure to reduce costs may also decrease benefits and, conversely, increasing benefits may also raise costs. Therefore, any measure to change the existing patent systems needs to be examined with respect to both the costs and the benefits of patent systems. The best way to fine-tune patent systems in relation to standards is to find the measure which maximises the ratio of benefits to costs with respect to stan-

64

Torremans, Holyoak and Torremans: Intellectual Property Law 19.

132


dards.65 However, finding the optimal solution, which maximises the ratio of benefits to costs of patent systems, is extremely difficult, because the precondition of finding the optimal solution is that the information about the relationship between the costs and benefits of patent systems is perfect but it is impossible to obtain perfect information about them and, in reality, the available information is fragmented. In other words, if the relationship between the costs and benefits of patent systems is given, for instance, in the form of a mathematical formula, and thus the perfect information on the relationship between them is provided, it may be possible to find the optimal solution by solving the formula, as many economists do. However, the current issues are not mathematical problems and the available information is limited. How then to find the measures to improve patent systems? Given the limitation, another reasonable way is adopted in this book. As noted, there are various factors which increase the social costs of patent systems and it is impossible to make exhaustive lists of them. However, it is possible to find some factors of the costs in relation to standards, as is described in the following sections. Thus, if it is possible to find measures which reduce the known costs of patent systems with respect to standards without decreasing the benefits of patent systems, the measures will be one of the ways to resolve the conflict between patents and standards. Of course, there is a precondition in this approach. The condition is that, even though the costs which are required to decrease here are reduced, other unconsidered or unknown costs of patent systems will not increase, or, if they increase, the amount of the increase is negligible. It is impossible to find all the elements of the social costs of patent systems and thus, admittedly, there may be other factors of the costs which are not considered here. If other unconsidered costs of patents are increased significantly by reducing the costs examined, the solution derived from the analysis of the known costs may not work. Thus, it is essential that the costs considered in this approach do not affect or raise significantly other costs not examined here. However, it is not difficult to infer that where the solutions obtained by analysing some elements of the social costs of patent systems are to limit the patent rights, the costs arising from patent rights will be diminished. Since the costs of patent systems mainly arise from the exclusivity of patent rights, it is reasonable to assume that limiting patent rights would reduce the costs of patent systems as a whole. Thus, what is important in deciding whether the solutions obtained from the analysis are feasible is the effect of the solutions on the benefits of patent systems rather than on costs which are not considered here.

65 Kaplow suggested this method in optimising patent terms. See Louis Kaplow, ‘The Patent–Antitrust Intersection: A Reappraisal’, Harvard Law Review 97 (1984).


133

The approach of the analysis is as follows. Firstly, the question of what are the costs of patent systems in relation to standards is studied. As noted, since it is impossible to find every cost of patent systems in relation to standards, only some cost factors, which are significant and manageable, are considered. Secondly, ways are sought to reduce the ‘known’ costs of patent systems in relation to standards and then the suggested solutions are examined with respect to the benefits of patent systems. If the proposed measures decrease costs and increase benefits, they are acceptable. If the suggested measures reduce benefits as well as costs but costs are reduced more than benefits, they are also acceptable. However, if they decrease benefits more than costs, they are not acceptable. Thus, whether the suggested solutions are acceptable or not is examined ex post. It has also to be noted that the suggested solutions may or may not be the optimal ones and that it is impossible to prove or disprove that they are optimal solutions since the information is not perfect. All that can be argued here is that the suggested measures are better than those provided by the existing patent laws and in that respect they are modest ones.

3.

COSTS OF PATENT SYSTEMS IN RELATION TO STANDARDS

3.1

Uniform Patent Systems

Patent systems seek to promote technological advancement by providing incentives to invent, disclose and innovate in the form of exclusive rights and thus in designing patent systems it is essential to consider the innovation structures. Innovation structures vary depending on the characteristics of industries and innovations; that is, innovation arises in a different way in each industry and in each innovation. There are many factors which affect the innovation process. First of all, the R&D cost needed to create an invention and to exploit it differs widely.66 In the case of the pharmaceutical industry, inventing and testing a new drug generally require considerable time and money. Similarly, substantial investments in R&D and production facilities are needed in designing and producing a highly sophisticated semiconductor integrated circuit. In contrast, many of the business method inventions and software inventions are not the outcome of large-scale investments. Similarly, inventions created by individual inventors are not likely to be from considerable investment since individual inventors usually cannot afford that. Thus the resources necessary for inventions and innovations differ by industry and by innovation.

66

Lemley, ‘Policy Levers in Patent Law’, 1581–3.

134


Moreover, the incentives to invent or innovate other than patent protection also vary widely in different industries.67 It is generally accepted that the cost required to imitate an invention is lower than that of creating the invention and that one of the important roles of patent protection is to create artificial lead time during which the inventor may recoup the costs invested to the invention. However, in some industries, a head start is the most crucial factor for high profits such that the lead time arising from the time required for competitors to imitate the original inventions provides sufficient incentive for inventors. In other words, in some industries the lead time rather than patent protection is the main reason why firms invent and innovate. Indeed, according to empirical researches, patents are an important way to protect inventions in the pharmaceutical industry but in many other industries patents are but one of many incentives to innovate and other incentives such as lead time and learning effects are more crucial than patent protection.68 Thus innovation structures vary by industry and by innovation and therefore the need for patent protection differs depending on industries and innovations. However, the current patent system is a ‘one-size-fits-all system’, treating every invention in the same way without considering the special features of it.69 Once patented, every invention is protected for the same time period, usually 20 years from the date of the application, with the same exclusive rights over the patented subject matter. Of course, in the case of pharmaceutical patents, patent terms can be extended under special circumstances, but that is the exception rather than the norm. Since the current patent system does not consider the industry specific features of innovations, it is natural that patent systems work well in some industries but not in some others. Indeed, considering the fact that homogeneous patent systems are applied to heterogeneous industries and innovations, patent systems are the ‘Procrustean bed’70 of technological innovation. Due to this aspect, the costs of patent systems arguably outweigh the benefits of them in some industries. This discrepancy between diverse needs for patent protection and the application of uniform patent systems decreases the effectiveness of patent systems in facilitating technical advancement, most

67 68

Ibid.: 1584–5. For more detail, see Levin et al., ‘Appropriating the Returns from Industrial Research and Development’. 69 Lester C. Thurow, ‘Needed: A New System of Intellectual Property Rights’, Harvard Business Review September–October (1997): 95. 70 ‘Procrustean bed’ has its origin in Greek mythology and means ‘something into conformity with which or subservience to which someone or something is arbitrarily and often ruthlessly or violently forced’. See Webster’s Third New International Dictionary, vol. II (London: Merriam-Webster Inc., 1986) 1809.


135

notably in cumulative technologies and system technologies.71 A cumulative technology is a technology which is built upon other technologies invented by earlier inventors,72 and a system technology means a technology which is composed of many distinct component technologies, which can be developed separately.73 With the exclusive right granted by patents, the holder of the patent covering the first invention can prevent, by refusing to grant licences to others, the appearance of improvement inventions or inventions which use the patented technology. Where the technological trajectory for potential improvements and uses of the first invention is clear and bounded, as is the case in ‘discrete’ technologies74 like pharmaceutical inventions, the technology suppression by the patent owner may not be serious because the original inventor is likely to develop the patented technology more efficiently than others. However, where the technology has broad prospect to be developed along various trajectories, as is the case in cumulative technologies and system technologies,75 it is not necessarily true that the holder of the first patent is in a better position to develop the prospect. Because of the uncertainty associated with research and development, a firm tends to focus on a limited number of applications and thus to develop its technology and product along a limited number of technical trajectories.76 Thus, the patent owner of the first patent may overlook or not recognise useful applications which can be built upon the patented invention. Furthermore, there is no guarantee that the patent owner will be willing to give licences to others hoping to exploit the applications which the patent owner does not develop. There is, of course, the argument that monopolists in general and patent owners in particular do not have any incentive to refuse to license if the bargaining between the patent owners and the users is beneficial to both parties.77 Most notably, Chicago school scholars argue this point based on economic analysis and Kitch, who suggested the prospect theory, contends similarly from patent’s perspective. However, as is argued in the following 71 Merges and Nelson, ‘On Limiting or Encouraging Rivalry in Technical Progress: The Effect of Patent Scope Decisions’, 6–9. 72 Suzanne Scotchmer, ‘Standing on the Shoulders of Giants: Cumulative Research and the Patent Law’, The Journal of Economic Perspectives 5, no. 1 (1991). 73 Merges and Nelson, ‘On Limiting or Encouraging Rivalry in Technical Progress: The Effect of Patent Scope Decisions’, 7. 74 Ibid. 75 Ibid. 76 Merges and Nelson, ‘On the Complex Economics of Patent Scope’, 873–4. 77 David J. Gerber, ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’, Virginia Law Review 74 (1988); Kitch, ‘The Nature and Future of the Patent System’.

136


sub-sections, the bargaining between the patent owners and the users is not always successful and, on top of that, monopolists’ refusal to deal is not necessarily based on efficiency reasons. Suffice it to say here that efficient bargaining is not necessarily achieved. In the absence of efficient bargaining between a patentee and third parties, the technology with broad prospect is likely to be underdeveloped. This reasoning leads to the conclusion that a variety of subsequent researchers can provide a more dynamic and increasing range of inventions and products. This reasoning is supported by empirical evidences. Merges and Nelson describe empirical cases in which patent protection deters technical advancement in cumulative and system technologies.78 The facts that Edison’s basic patent on incandescent lamps significantly hindered the technical advancement of the incandescent lamp industry, that the Selden patent which covered the use of a light gasoline combustion engine in an automobile was mainly used to distort competition in the markets, and that the Wright brothers’ patent was used to suppress potential competitors79 are shown in these cases, demonstrating the high costs caused by applying the uniform patent systems to cumulative and system technologies. Then how can the above reasoning be applied to patents covering standards in network industries? This question can be answered by noting that most of the products and technologies in network markets are cumulative and have the features of system technologies. The cumulative characteristics of them can easily be verified by the fact that the designs and functions of mobile phones are similar to each other and that items of software such as web browsers and word processors, produced by different companies, have functions and user interfaces comparable to each other. Considering that any product in markets where compatibility standards prevail has to incorporate the standards, the inventions which incorporate the standards can be considered inventions subsequent to standards, the initial inventions. Thus, with the patent rights, the holder of the patent covering the standards can preclude, by refusing to grant licences of the patent, the appearance of improvement inventions or inventions which use technical standards. Many of the products and technologies in network markets also have the characteristics of system technologies. It is common that a product in network markets is composed of several compatible products.80 For instance, a computer comprises a group of compatible hardware modules and a set of compatible software, and telecommunication systems are composed of various 78 79 80

Merges and Nelson, ‘On the Complex Economics of Patent Scope’, 884–97. Ibid. Oz Shy, The Economics of Network Industries (Cambridge University Press, 2001) 1–5.


137

compatible components. These characteristics of cumulative technologies and system technologies lead to the reasoning that if the owner of standards needs incentives to invent or innovate, the subsequent inventors also need incentives to participate in inventive activities to improve and expand the technologies incorporated in standards. Thus, the issue of how to strike the right balance between the two different incentives arises81 and this issue is not likely to be resolved within the existing patent systems, since the current patent systems are designed for discrete inventions such as pharmaceutical inventions.82 Accordingly, the one-size-fits-all patent systems do not strike the right balance between the costs and benefits of patent systems in relation to standards. 3.2

Bargaining Theories and Standards

Another factor increasing the costs and reducing the benefits of current patent systems can be found in the bargaining theories. This sub-section is devoted to showing how the bargaining problem between a standard owner and its user can create social costs. One of the most influential theories on bargaining was presented by Ronald Coase, a Nobel Prize winner, who suggested the Coase theorem. The Coase theorem contends that, in the absence of transaction costs, an efficient use of resources is obtained by way of private bargaining regardless of the initial assignment of legal entitlements.83 That is, without transaction costs, the parties at issue will trade until they reach a mutually satisfactory result, which leads to an efficient use of resources. The two assumptions of this theorem are that the parties concerned are rational and that the transaction costs are zero. Indeed, the parties will seek to maximise the utilities of resources since they are rational, by assumption. Given that the transaction cost is zero, as is assumed, the parties will trade until they can maximise the utilities of resources and thus there is no need to make a deal any more, without worrying about costs arising from transactions. Thus, it is clear that the theorem is correct under the assumptions of the bargaining parties’ rationality and zero transaction costs. However, in reality, the assumptions of the theorem are not necessarily met and, especially, the transaction costs are not zero or negligible in many cases. Transaction costs arise in each step of the exchange.84 In order to make a deal, one has to find a bargaining partner who hopes to bargain with one. Then one

81 John H. Barton, ‘Patents and Antitrust: A Rethinking in Light of Patent Breadth and Sequential Innovation’, Antitrust Law Journal 65 (1997). 82 Kingston, ‘Innovation Needs Patents Reform’, 405–6. 83 For more detail, see Coase, ‘The Problem of Social Cost’. 84 Robert Cooter and Thomas Ulen, Law and Economics, 3rd edn (Addison Wesley Longman, Inc., 2000) 87–8.

138


has to negotiate with the partner over the bargaining conditions. Once a bargain is reached, it has to be enforced and monitored. Since search costs, negotiation costs and enforcement costs arise in actual cases, the transaction costs are not zero. Where transaction costs are not zero, bargaining may be successful or unsuccessful, depending on the transaction costs and the value created by the trade. Where the transaction costs are greater than the value generated by the bargaining, the transaction costs will block the bargaining, on condition that the buyer and the seller of the bargain are rational. Conversely, where the value expected from an exchange is larger than the transaction costs, the bargaining parties will make a deal. Considering this effect of transaction costs on bargaining, it is important to allocate legal entitlement properly so as to achieve an efficient use of resources. There are two ways to maximise efficiency by designing law with transaction costs taken into account.85 The first method is to make law in such a way as to lower transaction costs. As noted, the Coase theorem argues that, in the absence of transaction costs, an efficient use of resources is achieved from private bargaining regardless of the initial assignment of legal entitlements. Thus, even though transaction costs are not zero, by lowering transaction costs it is possible to make the bargaining conditions similar to the ideal conditions which the Coase theorem assumes. This method seeks to apply the Coase theorem to real cases by making the practical circumstances similar to ideal conditions. That is, this method seeks to minimise the discrepancy between the ideal conditions and the real circumstances and to find solutions to the problems caused by the transaction costs within the boundary of the Coase theorem. Thus, this method is called the normative Coase theorem and can be expressed like this: ‘Structure the law so as to remove the impediments to private agreements.’86 One way to structure the law to lower transaction costs is to design the law in such a manner as to make the legal rights simple and clear. The other method has a perspective totally different from that of the normative Coase theorem and is called the normative Hobbes theorem.87 While the normative Coase theorem seeks to achieve successful bargaining and thus argues that the obstacles to efficient bargaining have to be removed, the normative Hobbes theorem focuses on the harm caused by bargaining failure. That is, the normative Hobbes theorem presupposes the bargaining breakdown and seeks to minimise the costs arising from the bargaining failure. The normative Hobbes theorem is as follows: ‘Structure the law so as to minimise the harm caused by failures in private agreements.’88 85 86 87 88

Ibid. 93. Ibid. Ibid. 94. Ibid.


139

In order to structure the law in accordance with the normative Hobbes theorem, the legal entitlement should be allocated to those who value it most, the reason for which is easily verified by the following example. Suppose a person A has a widget and values it at £100. There is another person B, who evaluates the widget at £400. Where the transaction cost of the bargaining between A and B is less than £300, they will bargain and both of them obtain benefits from the trade.89 However, if A considers the value of the widget £400 and B evaluates it at £100, that is, if A values the widget more than B does, there will be no bargaining between them, whatever the transaction costs would be. Indeed, there is no need to bargain and the efficient resource allocation is achieved at the initial condition. This is what the normative Hobbes theorem seeks to achieve. In short, since each transaction would create costs, the normative Hobbes theorem seeks to design the law so as to make the transaction unnecessary and thus reduce the transaction costs. The Coase theorem, the normative Coase theorem, and the normative Hobbes theorem clearly show that in real bargaining, where transaction costs are not zero, an efficient use of resources is not necessarily achieved and thus law needs to be designed to resolve the inefficiency. Then what implications do the bargaining theories have in relation to patents covering standards? The bargaining theories are important in examining the costs and benefits of patent systems because, without efficient bargaining, the costs of patent systems grow, whereas the benefits decrease. As noted before, patent rights are provided to resolve the public goods problem of inventions. That is, since inventions can be consumed simultaneously by everyone who knows the technology, which is called non-rivalrous consumption, and since it is costly to exclude others from using the inventions, which is called non-excludability, private markets often undersupply inventions without government interventions.90 In order to solve this public goods problem of inventions, government intervention is necessary. The government can supply inventions by itself, subsidise the supply of inventions, or protect the inventions with property rights.91 Clearly, granting patents to inventions is to protect the inventions with property rights. Since patent rights are exclusive rights, efficient bargaining is essential to achieve an efficient use of resources. However, bargaining of inventions protected by patents is not necessarily efficient because transaction costs related to the bargaining are not zero. In the context of patents and standards, various factors increase transaction costs and block efficient bargaining. 89 90

For simplicity, strategic behaviours are not considered in this example. Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’,

129–30. 91

Cooter and Ulen, Law and Economics 126.

140


Firstly, in order to achieve efficient bargaining, the information on patents has to be perfect, bargaining must be costless and the parties concerned have to be rational. However, transactions require time and money, and costless deals are only possible in hypothetical situations. The costs which are necessary to find and contact the patent owner, negotiate licensing conditions, and exchange payment or services are not trivial. This is also true in the case of patents covering standards. In some respects, the standard setting organisations contribute to lowering the transaction costs between patent owners and users, but there may still be considerable transaction costs arising from the bargaining within SSOs. Many SSOs devise IPR policies to facilitate bargaining between patent owners and users, requiring patent owners to license the essential patent on a reasonable and non-discriminatory basis.92 However, since the members of SSOs have to negotiate with each other over the licensing terms and the meaning of ‘reasonable’, disputes over licensing fees may arise,93 increasing transaction costs. Indeed, some have even argued that the price fixing in standard setting organisations should not be condemned because it could increase the efficiency of bargaining by reducing transaction costs and uncertainty,94 which implies that transaction costs are important factors that need to be considered in the licences of technology incorporated in standards. Given that the formal standards have problems caused by transaction costs, it easily follows that the informal standards also have these problems. Furthermore, even though it is assumed that the relevant parties are rational, which is not always true,95 the rationality assumption does not necessarily lead to efficient bargaining because each party has imperfect information on the value and prospect of the inventions.96 Uncertainty caused by imperfect information can lead to bargaining breakdown in at least two ways.97 Firstly, because of uncertainty, each party may not identify the gains from the bargaining, even though the bargaining is beneficial to each party. Secondly, each party may evaluate the bargaining differently, which may prevent the parties from reaching an agreement on the licensing fees. Thus uncertainty may give rise to the failure of beneficial bargaining. 92 For more detail, see Mark A. Lemley, ‘Intellectual Property Rights and Standard-Setting Organizations’, California Law Review 90 (2002). 93 P.D. Curran, ‘Standard-Setting Organizations: Patents, Price Fixing, and Per Se Legality’, University of Chicago Law Review 70 (2003): 983. 94 Ibid. 95 Lemley, ‘The Economics of Improvement in Intellectual Property Law’, 1059–61. 96 Robert P. Merges, ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’, Tennessee Law Review 62 (1994): 84–9. 97 Lemley, ‘The Economics of Improvement in Intellectual Property Law’, 1055–6.


141

Secondly, strategic behaviours may increase transaction costs, causing a bargain to fail.98 Each party of a bargain has an incentive to act strategically to increase the benefit from the bargain and these strategic behaviours may work because the information which each party has is asymmetric; that is, each one has perfect information on itself but can only obtain limited information on its counterpart. Each party is likely to seek to maximise the profit from the bargaining by using rational and seemingly irrational strategies. The vulnerability of the standard users, combined with the strategic conduct of patent owners, can hinder efficient bargaining. ‘Arrow’s paradox of information’, which arises when one hopes to sell information, explains this point.99 In order to sell information, the owner of the information has to disclose it, but once it is disclosed the counterpart of the bargain does not need to buy the information because he already has it. However, without disclosure of the information, potential buyers do not know what the information is and thus the bargain is not possible. One way to overcome this paradox is to give the seller of the information the power to prevent the buyers from exploiting the information without his consent. Thus, the seller needs to obtain intellectual property rights over his information in order to prevent others from exploiting it freely. This reasoning can be applied to standards protected by patents. The potential user of standards needs to obtain licences from the owner of patents covering standards. In order to obtain licences, he has to explain how he will use the licences, exposing his technical idea. Here Arrow’s Paradox comes in. Where the user does not have any patents over his technical idea and merely is a potential infringer, the bargain is totally up to the patent owner. Thus, the patent owner may steal the idea of the user and exploit it if doing so seems profitable. However, where the user of standards has patents for his invention which incorporates some of the patented standards, the user will have more bargaining power than he would if the improvement invention were not protected by patents. The user of standards who has patents over his technical idea can prohibit the patent owner from using the idea. Thus, in those cases where the bargain between the patent owner and the user of standards fails, the patent owner of standards is not able to enjoy the benefit otherwise gained from the exploitation of the idea of the potential user, whereas the improver of standards cannot exploit his invention at all. This is called the blocking patent problem.100

98 99

Ibid.: 1058–60. Kenneth J. Arrow, ‘Economic Welfare and the Allocation of Resources for Invention’, in The Rate and Direction of Inventive Activity: Economic and Social Factors (National Bureau of Economic Research, 1962). 100 For more detail, see Merges, ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’.

142


In this situation, the patents covering standards block the exploitation of the improved inventions while the patents on improvement inventions also prevent the owner of the patents covering standards from utilising the improved inventions. Thus patent laws seek to encourage the bargaining parties to make a deal by increasing the loss from bargaining failure.101 However, considering the outcome of this bargaining breakdown, the owner of standards is still in a superior position to the user of standards, even though the user has some bargaining power. On top of that, even in those cases where the standard user’s inventions are protected by patent rights, it may still be possible to invent around the inventions because it is usually impossible to perfectly protect an invention with a patent. As far as technical standards are concerned, it is of no use to invent around them because what is necessary is not to invent around the technical standards but to incorporate them for compatibility. As for patented technical improvements, however, it is possible to invent around them in many cases. Should the owner of the patent covering standards believe that inventing around the improvement invention were possible and profitable, he would use the idea of the improvement invention and make a new version of the improvement. All this implies that the user of standards who has an idea to improve or expand the patented technology incorporated in standards has far less bargaining power than the standard owner. This imbalance of bargaining power is likely to result in asymmetric agreements favouring the patent owner, which strongly suggests that an adjustment of bargaining power is needed. The point argued here is not that bargaining power has to be fine-tuned for fairness or economic justice but that bargaining power has to be adjusted for the economic efficiency which patent systems need to achieve. The imbalance of bargaining power would result in asymmetric agreements in favour of the patent owner, which would reduce the incentives to improve the technical standards, hindering technical innovation, and prevent efficient bargaining between the relevant parties. Thus, there should be readjustment of the bargaining power between the owner of standards and the users in order to increase the efficiency of patent systems. Empirical evidences show that mutually beneficial deals can fail such that efficient bargaining is not necessarily achieved. The dispute between Marconi Company and De Forest is a typical example of this.102 De Forest had a patent on triode, which was based on Marconi’s diode patent, and thus the triode patent of De Forest could not be exploited without a licence of Marconi’s 101

Lemley, ‘The Economics of Improvement in Intellectual Property Law’,

1062–3. 102 Merges, ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’, 84–7.


143

diode patent. Should De Forest and Marconi have reached a voluntary agreement on a patent licence, both of them would have gained considerable benefits from the exploitation of patented technologies.103 However, private exchanges between them failed and the dispute was resolved by a court and the Navy, which took more than ten years. Because of this bargaining deadlock, an efficient use of resources was precluded and the advancement of related technologies was significantly delayed. There are other cases which show that beneficial deals could not be achieved104 and there may be other unknown stories about this problem. The above discussion suggests that current patent systems do not necessarily promote the efficient use of resources. Especially when the patent covers standards, the patent owner has strong bargaining power, which may be exercised to result in inefficient bargaining. It seems unlikely that the bottleneck like standards will always be open to the public automatically.105 Rather, standards can only be available through negotiation under the shadow of asymmetric bargaining power between the owners and the users. This implies that there should be some limitations on the rights of a patent owner who has the patents essential to standards, and that adjustment of bargaining power is essential to achieve efficient bargaining. 3.3

Network Externalities and Monopoly

This book is concerned with standards, especially the compatibility standards in network industries, where the network effects or network externalities prevail. Network externalities arising from the compatibility standards should be considered in resolving the patent issues in standards because they can affect the innovation process and market structure. As noted in Chapter 2, network externalities work on the demand side and occur when the value derived from the consumption of products increases as the number of other people who use the products or compatible products increases.106 Network externalities arise from virtual networks, such as computer hardware and software, as well as from real networks, such as telephony and fax machines.107 In a typical economic analysis, demand is represented as a function of price.

103 104 105

Ibid. Ibid. Joseph Farrell, ‘Creating Local Competition’, Federal Communications Law Journal 49 (1996): 211–12. 106 Lemley and McGowan, ‘Legal Implications of Network Economic Effects’, 483–4. 107 Carl Shapiro and Michael L. Katz, ‘Network Externalities, Competition and Compatibility’, The American Economic Review 75, no. 3 (1985): 424.

144


It is well known that the demand curve is down-sloping with respect to price, and thus where the price of a good rises demand for it falls and where the price of a good falls demand for it rises. In network markets, however, customer demand for a product is affected not only by the price of the product but also by the size of the network to which the product belongs. Beside the value created by using a product, the network provides additional value to its members and these value creation effects of a network are called network effects or network externalities. Network externalities have both positive and negative aspects. On the one hand, network externalities enhance the value of network products. Take computer systems. Clearly, computer systems have stand-alone value since they enable users to use software such as word-processors and spreadsheets. However, it is also clear that the value of computer systems is enhanced in those cases where the computer systems are connected to the internet. The value enhancement effect resulting from network externalities is often described by Metcalf’s law.108 Suppose the number of people in a network is n, and the value which is created by being connected to the network to each person in the network is proportional to the number of other people in the same network. Then the total value of the network is approximately proportional to the square of the number of people in the network, since the total value is n (n – 1) = n2 – n. Accordingly, where the size of a network increases ten times, the value of the network increases approximately a hundred times. Of course, it is not necessarily true that the value of the network to each user is proportional to the number of other users, but what Metcalf’s law demonstrates is that the value of a network may grow rapidly as the size of the network increases. Thus network externalities provide additional value to network members and increase the utility of a network product. On the other hand, network externalities also have negative aspects. Network externalities, along with switching costs, which the consumer has to pay to switch from one system to another, can cause markets to be locked into a particular standard because market participants are reluctant to switch from incumbent standards to other standards regardless of whether other standards are inferior or superior to the standard which they are currently using. That is, in network markets, a consumer who is enjoying the value created by network externalities does not want to lose the network benefits and thus is reluctant to join a new network which substitutes the incumbent network unless other consumers join the new network. This phenomenon is called the ‘penguin effect’, because consumers’ behaviours are similar to those of penguins that

108 Carl Shapiro and Hal R. Varian, Information Rules (Boston: Harvard Business School Press, 1999) 183–4.


145

want to catch fish in the water but are unable to jump for fear of possible predators swimming around.109 The longer consumers remain in a network, the more resources they invest in the network and thus, where they hope to switch from an old network to a new one, they have to pay high switching costs. Where the benefits which are expected from transition from an old network to a new network do not outweigh the switching costs, consumers are not likely to join the new network. Due to these network externalities and switching costs, consumers in network markets are likely to be locked into a technology. Lock-in effects may cause the network market to stick to an old technology, even though the old technology is inferior to a new one. Farrell and Saloner call this effect excess inertia, which refers to the inefficient attachment to the old technology.110 Because of the lock-in effect, the dominant firm having standards in network industries can enjoy the protection of high entry barriers. The high barriers to entry cause the market to have a monopolistic or highly concentrated oligopolistic structure. The harm caused by these market structures includes not only the resource misallocation resulting from monopoly pricing but also fewer innovations.111 Even though a dominant firm in a concentrated market is in a better position to engage in innovative activities, the firm may have less incentive to innovate because the pressure for innovation in concentrated markets is not so intense as in competitive markets.112 Indeed, many empirical researches support this argument, as follows. Baker studies the automobile industries of the US and finds that fringe firms can be more innovative than dominant firms.113 He argues that the innovation by dominant firms can be more costly than that by fringe firms because, in order to develop new products, dominant firms have to give up the investment made to manufacture old products and such costs may discourage dominant firms from innovating.114 He also argues that the risk arising from R&D

109

For more detail, see Joseph Farrell and Garth Saloner, ‘Competition, Compatibility and Standards: The Economics of Horses, Penguins and Lemmings’, in Product Standardization and Competitive Strategy, ed. H. Landis Gabel, Advanced Series in Management (Oxford: Elsevier Science Publishers, 1987). 110 Joseph Farrell and Garth Saloner, ‘Standardization, Compatibility and Innovation’, The RAND Journal of Economics 16, no. 1 (1985): 70–72. 111 Willow A. Sheremata, ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’, The Antitrust Bulletin 42 (1997): 971. 112 Thomas A. Piraino, ‘Identifying Monopolists’ Illegal Conduct Under the Sherman Act’, New York University Law Review 75 (2000): 815. 113 Jonathan B. Baker, ‘Fringe Firms and Incentives to Innovate’, Antitrust Law Journal 63 (1995). 114 Ibid.: 634–9.

146


may cause dominant firms not to innovate.115 Dominant firms can obtain considerable profits from the status quo and thus they may have less incentive to participate in risky innovation. Tripsas examines the technological and competitive history of the typesetter industry from 1886 to 1990.116 She finds that the incumbent firms’ innovations are handicapped by their previous business and that new firms in the market introduce innovations which are superior to those of established firms.117 Her finding implies that the dominant firms may be reluctant to introduce new innovations which can cannibalise their previous investments and that, due to the limitation, they may be less innovative than new entrants in the market. Using empirical data, Scherer argues that there is an ‘inverted U’ shaped relationship between innovative activities and market concentration.118 That is, innovative activities are not so intense in both highly concentrated markets and competitive markets as in modestly concentrated markets. He finds that R&D/sales ratios increase as market concentration rises and peaks at the point where four firms have market share of 50 to 60 per cent.119 After the peak point, the ratios decrease as market concentration increases. His finding demonstrates that a highly concentrated market structure does not increase dominant firms’ innovative activities. Shelanski studies innovations in the US telecommunication industry and finds that innovations in the industry are deployed more quickly in competitive market conditions than in non-competitive ones.120 He analyses four innovations deployed in monopoly markets, three technologies introduced under concentrated oligopolistic market conditions, and three innovations deployed in competitive oligopoly markets. He finds that deployment of innovations positively correlates with the competitiveness of markets and argues that competition should be presumed to be beneficial to innovation in the telecommunication industry. These empirical studies support the argument that a dominant firm in a concentrated market may have less incentive to innovate. Of course, there is no generally agreed theory on the relationship between market power and 115 116

Ibid. Mary Tripsas, ‘Unraveling the Process of Creative Destruction: Complementary Assets and Incumbent Survival in the Typesetter Industry’, Strategic Management Journal 18 (1997). 117 Ibid.: 139. 118 F.M. Scherer, ‘Antitrust, Efficiency, and Progress’, New York University Law Review 62 (1987): 1012–13. 119 Ibid. 120 Howard A. Shelanski, ‘Competition and Deployment of New Technology in U.S. Telecommunications’, University of Chicago Legal Forum (2000).


147

innovation. Indeed, Joseph Schumpeter famously argues that monopoly and market power are more effective in promoting economic growth and technical advancement than competition, since market power will provide economic rents for the dominant firm to take the risk of innovations.121 Many others present similar arguments, supported by empirical evidences, and their arguments are quite persuasive. However, what is argued here is not that monopoly and market power are always bad for technical innovation but that they can be detrimental to technical innovation. That is, it is not clear whether monopoly and market power will always stifle technical innovation but it is important to note that monopoly and market power have impeded innovation in many cases, as the empirical studies demonstrate. The facts that market power may hinder innovations and that standards create market power by lifting barriers to entry imply that standards may impede technical advancement. Some argue that barriers to entry resulting from network externalities and the advantage of a dominant firm arising from a large installed base are not invincible.122 Their argument is that even though a technology or a product is protected by high barriers to entry, it will eventually be substituted by competing ones through market process. Indeed, compact disk technologies overcame the high barriers to entry and the large installed base of LP technologies in the recording device industry, and MP3 players successfully replaced portable cassette player technologies in the portable music player industry, and these examples support the arguments. Admittedly, any dominant product will surely surrender to new products and a dominant firm is vulnerable to its competitors. However, as one commentator argued, the timing of such market mechanism is important.123 That is, even though a dominant technology may be leap-frogged some time in the future, the delay in innovation caused by the high entry barriers may entail considerable social costs. On top of the entry barriers arising from the nature of the markets, the dominant firm can erect additional barriers by protecting the standards with patents and by denying others access to the standards.124 Without access to the incumbent compatibility standards, technology or a product cannot be made compatible with the standards and thus the product or technology is not likely to survive in markets. Of course, there may be pioneering consumers who seek to buy new network products incompatible with the incumbent standards. Considering that they have to lose the benefits provided by the established

121 122

Schumpeter, Capitalism, Socialism, and Democracy Chapter VIII. Sheremata, ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’, 962–3. 123 Ibid. 124 Carl Shapiro, ‘Exclusivity in Network Industries’, George Mason Law Review 7 (1999): 675–7.

148


network, the number of the pioneering consumers is likely to be small. Furthermore, a network product which is not compatible with the incumbent network is generally adopted slowly because consumers should pay large costs to switch from an existing technology to a new one and because the network benefits created by the new product are not so big at an early stage of the adoption of the new product due to the small number of users.125 In addition, the owner of incumbent standards may stifle the appearance of new standards by making exclusive contracts with potential pioneering users.126 Consequently, the exclusionary conduct of the standard owner is likely to stop or significantly delay the emergence of competing standards, preventing them from gaining the critical mass which is necessary to form another network. In some sense, it is natural that a dominant firm in a market seeks to protect its dominant position with its market power. Where new entrants or potential entrants are likely to be competitors in the market in which the dominant firm carries on a business, the dominant firm will seek to prevent competitors from entering the market and to make the competitors exit from the market.127 However, the exercise of market power by the dominant firm may conflict with the interest of society and impede technical advancement. Allowing a single firm or several firms to decide the trajectories of technical advancement is not necessarily desirable from the perspective of technical innovation. 3.4

Abuse of Market Power

Compatibility standards give rise to concentrated market structures, conferring market power on the owners of the standards. It can be expected that the market power created by standards will be abused to strengthen the market power by excluding competitors in the relevant market, and will be leveraged from the relevant market to another market. There are various ways in which the market power is abused. For instance, exclusive dealing agreements, tieins, refusal to deal, predatory pricing and price discrimination can be used to maintain and enhance market power,128 which a dominant firm may otherwise lose as competitors enter the market and competition increases. The abuse of market power by dominant firms is condemned by competition laws and antitrust laws, but it is also one of the important elements comprising the

125 126 127

Ibid.: 678–80. Ibid. Shane Greenstein, ‘The Three Coms of the Microsoft Antitrust Suit: Competition Policy, Commercial Experimentation, and Computing Platforms’, UWLA Law Review 32 (2001): 98–9. 128 For more detail, see Richard Whish, Competition Law, 5th edn (London: LexisNexis UK, 2003).


149

social costs of patent systems. In order to find a way to reduce the social costs entailed by patent systems in relation to standards, it is essential to understand how the market power of the owner of patents covering standards can be abused to eliminate competition in the related markets, increasing social costs and hampering technical innovation. Economic theories seek to analyse the economic effects of market power and abuse thereof, but most notably the Chicago school scholars established neo-classical economic theories and applied the theories to analyse market power and anticompetitive behaviours. The Chicago school’s analysis has greatly contributed to the understanding of economic efficiency and monopolists’ behaviours, and many courts in the US adopted the neo-classical economic theories to analyse antitrust cases. As can be seen in the following sub-sections, their analysis is based on price theory and economic efficiency. However, the neo-classical economic theories of the Chicago school are not free from challenges and are criticised by so called ‘post-Chicago’ critics.129 The main criticism of the Chicago school’s approach is that the analysis of Chicagoans is heavily dependent on price theory and that their perspectives are too static. In the following sub-sections, the patent issues in standards are analysed by applying economic theories. As is described in Chapters 3 and 4, the main issue of standards covered by patents is the refusal by patent owners to grant licences to those who hope to enter the relevant markets. Refusal to deal in general and refusal to license in particular are the typical methods which can arguably be used to extend market power. Therefore, how the economic theories analyse the exclusionary conducts like refusal to license and how the economic analysis can be applied to standards covered by patents are examined in the following sub-sections. 3.4.1 Chicago school’s approach to refusal to deal The Chicago school’s analysis of refusal to deal can be described with the following example. Suppose that there are widgets A and B, where the widget A is an indispensible input to produce the widget B. Suppose also that the market of the widget A is monopolised by a monopolist, named M, and that the market of the widget B is competitive. The monopolist M is also carrying on a business in the market of the widget B and thus, in the market of the widget B, the monopolist M is competing with the buyers of the widget A, called U as a whole. In short, the market of the widget A, which is an essential input to carry 129 Richard N. Langlois, ‘Technological Standards, Innovation, and Essential Facilities: Toward a Schumpeterian Post-Chicago Approach’ (paper presented at the George Mason University conference on Dynamic Competition and Antitrust, Washington DC, 1999).

150


on businesses in the market of the widget B, is monopolised by the monopolist M, and the monopolist M and the widget A’s buyers U are competing in the market of the widget B. Here, the market of the widget A is the upstream market and the market of the widget B is the downstream market. M is a vertically integrated monopolist since it does its business in the downstream market as well as the upstream market. In this situation, what would the monopolist M do to maximise its profits from the markets of the widgets A and B? Intuitively, the monopolist M may seek to eliminate competition in the market of the widget B by refusing to sell the widget A to the competitors U. It seems that, by excluding the competitors in the market of the widget B, the monopolist M can obtain monopoly profits from the market of the widget B as well as from the market of the widget A. However, the Chicagoans’ arguments, which are based on chain-link theory, are quite different.130 They argue that there is no anticompetitive reason for the monopolist M to refuse to deal with the competitors U in this circumstance. Since the monopolist M can extract monopoly profits by selling the widget A to competitors U at a price providing monopoly profits to the monopolist, the monopolist M does not have to exclude the competitors U from the downstream market and to pay extra costs to monopolise the market of the widget B. Furthermore, the monopolist M cannot obtain monopoly profits from both markets, even though it successfully monopolises the market of the widget B. If the monopolist raises the prices of both widget A and widget B to monopolistic levels, the demand for the widget B as well as for the widget A is decreased. This is easily verified. Since the market of the widget B is monopolised by assumption, the monopolist M is the sole seller of the widget B and thus the monopoly pricing of the widget A will inevitably increase the price of the widget B. Where the monopolist seeks to obtain monopoly profits not only from the widget A but also from the widget B, the price of the widget B will be raised, which reduces the demand for the widget B. Accordingly, the monopolist cannot obtain monopoly profits in both markets. Therefore the chain-link theory, which the Chicagoans adopt, argues that since the monopoly profits are bounded and cannot be increased by monopolising the vertically related markets, the monopolist has no incentive to refuse to deal for anticompetitive purposes.131 Rather, given that there is no anticompetitive reason for a monopolist to refuse to deal with competitors in the downstream market, the monopolist’s refusal to deal should be presumed to

130 Gerber, ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’, 1083–5. 131 Richard A. Posner, ‘The Chicago School of Antitrust Analysis’, University of Pennsylvania Law Review 127 (1979): 925–33.


151

increase efficiency.132 Indeed, vertical restraints may increase efficiency by preventing free-rider problems.133 To illustrate this, consider the case of Aspen Skiing Co. v. Aspen Highlands Skiing Co.134 The interpretation of this case shows the typical approach of the Chicago school to market power. Since this case is described in Chapter 3, it is not necessary to repeat it again in detail, but for convenience the facts of the case are briefly repeated here. The defendant (Aspen Skiing Co.), who owned three ski resorts in Aspen, had previously cooperated with the plaintiff in offering an ‘all Aspen ticket program’. After acquiring control of three of the four ski slopes, the defendant refused to make a deal with the plaintiff (Aspen Highlands Skiing Co.). The plaintiff offered several suggestions in order to resume the cooperative relationship but all of them were rejected by the defendant. The plaintiff filed a suit against the defendant, alleging that the defendant violated the antitrust law. In response, the defendant argued that even a monopolist was not obliged to engage in joint marketing with competitors and that it did not violate the antitrust law because none of its behaviour could be exclusionary. This case was concluded in the Supreme Court. The Supreme Court held that a refusal to deal by a monopolist violated section 2 of the Sherman Act when such refusal produced an important change that could not be justified by normal business purposes. The Chicago school’s approach to this case is quite different from the ruling of the Supreme Court.135 It is argued that the defendant may refuse to deal on the basis of efficiency. The argument is that the efficiency may be maximised with the three slopes and that adding one more slope to the optimal size of the defendant’s business may reduce the optimal efficiency. Where the defendant’s business reaches the optimal size, expanding the business by cooperating with competitors may entail inefficiency. Since market process will eventually find the optimum where the efficiency is maximised, the compelled dealing is not necessary and would cause resource misallocation. Thus the Chicagoans argue that the criterion for the anticompetitive refusal to deal has to be efficiency.136 That is, even though a vertical constraint may seem to distort market competition, the conduct should not be condemned if it increases economic efficiency, since the vertical constraint will benefit consumers after all. Only in those cases where a refusal to deal does not

132 Robert H. Bork, The Antitrust Paradox: A Policy at War with Itself (Oxford: The Free Press, 1993) 290–91. 133 Ibid. 134 See Aspen Skiing Co. v. Aspen Highlands Skiing Co., 472 U.S. 585 (1985). 135 See Gerber, ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’, 1107–12. 136 Bork, The Antitrust Paradox: A Policy at War with Itself Chapter 17.

152


increase efficiency can it be considered anticompetitive, but anticompetitive refusal to deal is not common.137 Another notable argument of the Chicago school is that natural barriers to entry are not policy concerns and legal intervention is necessary only in those instances where entry barriers are built artificially through exclusionary practices.138 The product differentiation, advertising and promotion, and capital requirement are often considered natural barriers to entry. However, it is argued that since these difficulties are inherent in entering the market and the incumbent firms have also paid these costs, these barriers are natural and there is no policy concern about them. Rather, these natural barriers to entry are representations of efficiency and thus should be distinguished from the artificial entry barriers, which need to be condemned.139 Vertical constraints such as refusal to deal are explained with the argument of natural and artificial barriers to entry. Vertical constraints can be used by a monopolist to exclude competitors from a downstream market and to monopolise the market. One may argue that vertical integration by way of vertical constraints can create a barrier to entry because any entrant into the monopolised markets has to enter two markets, the upstream market and the downstream market. However, the Chicagoans deny this argument as follows.140 Where the vertical integration is efficient, there is no reason to object to it. Therefore, just consider the case where the vertical integration is not efficient. In this case, the monopolist is losing efficiency and thus new entrants that can be more efficient than the monopolist are encouraged to enter the market even though they have to enter the two markets at the same time. The new entrants may not need to enter the two markets simultaneously since, if the market is attractive enough, someone else will surely enter it at another level. In any case, the natural barriers to entry become lower than they would have been if the monopolist had not monopolised the two markets. Thus the Chicagoans argue that, since there is no reason for a monopolist to restrict vertical relationships with its competitors in downstream markets, it has to be assumed that vertical constraints increase efficiency.141

137 For the elaboration of this argument, see Gerber, ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’. 138 Bork, The Antitrust Paradox: A Policy at War with Itself 310–29. 139 Ibid. 140 Ibid. 321–4. 141 Gerber, ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’, 1084–7.


153

3.4.2 Post-Chicago approach to refusal to deal The economic theories of the Chicago school are challenged by ‘postChicago’ critics142 and the attacks mainly focus on the fact that the analyses of the Chicago school are too static. It is argued that the Chicagoans’ perspectives are so static that they fail to consider the dynamic effect of vertical constraints.143 As can be seen in the chain-link model, the Chicagoans’ interpretation of monopoly mainly focuses on profit maximisation in a given market structure. However, the post-Chicagoans argue that the monopolist may seek to increase the long-term profits by changing the market structure and that in order to change the market structure the monopolist may even sacrifice the short-term profits.144 In that respect, the post-Chicagoans’ perspectives are dynamic ones rather than static ones. To contrast the postChicago theorists’ approach with the Chicago school’s, the post-Chicagoans’ interpretations of vertical constraints, Aspen Skiing,145 and barriers to entry are examined. While the Chicago school denies the possibility of extending monopoly power from one market to another, as the chain-link argument shows, the postChicagoans argue that by way of vertical constraints a monopolist may leverage its monopoly power in one market to another in order to maintain and enhance the monopoly power. Bonjorno146 is an example.147 The defendant (Kaiser) had a dominant position in the market of aluminum coil and sheet, and the plaintiffs (Bonjorno et al.) were the sole stockholders of Columbia Metal Culvert Co., which produced aluminum drainage pipe using the aluminum coil and sheet. Kaiser was a competitor of Columbia in the market of aluminum drainage pipe. In 1974, Kaiser began the price squeeze strategy. That is, Kaiser along with other suppliers of aluminum raw materials supplied aluminum coil and sheet to Columbia at a price similar to the price at which Kaiser sold aluminum drainage pipe. Consequently, the margin between the buying price of raw materials and the selling price of the final products was so small that Columbia could not obtain profits enough to cover the total costs. Because of these financial difficulties, Columbia stopped producing pipe and exited the market.

142 Langlois, ‘Technological Standards, Innovation, and Essential Facilities: Toward a Schumpeterian Post-Chicago Approach’, 12–17. 143 Louis Kaplow, ‘Extension of Monopoly Power through Leverage’, Columbia Law Review 85 (1985): 523–30. 144 Ibid. 145 For more detail, see Aspen Skiing Co. v. Aspen Highlands Skiing Co. 146 Bonjorno v. Kaiser Aluminum & Chemical Corp., 752 F.2d 802 (1984). 147 This example is described in Herbert Hovenkamp, ‘Antitrust Policy After Chicago’, Michigan Law Review 84 (1985): 270–74.

154


Then why did Kaiser, the monopolist, adopt this strategy? According to the Chicago school’s analysis, the monopolist does not have to use a price squeeze strategy. From the Chicago school’s perspective, the monopolist can refuse to deal and can monopolise the downstream market should the vertical integration increase efficiency.148 Alternatively, where the vertical integration does not increase efficiency, the monopolist does not have to monopolise the downstream market, as the chain-link theory explains.149 Therefore, the vertical constraints have to be presumed to enhance efficiency from the Chicago school’s perspective. However, it is argued that the vertical constraints can be used as a way to force the firms in the downstream market to deal exclusively with the monopolist, strengthening the market power of the monopolist.150 Suppose the monopolist uses the price squeeze strategy by selling the raw material at a price with which a firm in the downstream market can make profits only enough to cover the average variable costs but not enough to cover the fixed costs. In this case, the firm in the downstream market has no choice but to keep producing its goods unless it can exit the market with trivial costs. That is, the sunken costs, which are the difference between the value of resources invested by the firm in the downstream market and the salvage value which the firm can obtain by selling its assets to a third party hoping to enter the market, can be utilised strategically by the monopolist. If the monopolist continues the price squeeze for the useful life of plants of the firm in the downstream market, the firm cannot recoup the whole fixed costs and only obtains the salvage costs when it exits the market. In contrast, the monopolist can obtain monopoly profits by selling the raw materials at the price at which it can make monopoly profits. That is, with the price squeeze strategy, the monopolist can successfully threaten vertically related firms without sacrificing monopoly profits. Refusal to deal can be utilised in similar ways. The amount of the sunken costs varies according to industry and in many industries it is not trivial. That is, a firm has to pay considerable sunken costs when it sells its assets to exit the relevant market. The monopolist can exercise its monopoly power strategically, taking advantage of the sunken costs of a firm in the downstream market. Where the sunken costs are considerable, the monopolist can intimidate the firms in the downstream market into dealing exclusively with the monopolist or following the demands of the monopolist. The monopolist may have no intention to refuse to deal because it decreases monopoly profits, but the mere possibility of the refusal to deal can be a real threat to the firms in the 148 149 150

Ibid.: 269. Posner, ‘The Chicago School of Antitrust Analysis’, 925–30. Hovenkamp, ‘Antitrust Policy After Chicago’, 272–3.


155

downstream market, which would have to pay substantial sunken costs should the monopolist refuse to deal.151 Thus, using vertical constraints like refusal to deal, the monopolist can transmit a clear message that exclusive dealing with the monopolist or accepting the requests of the monopolist is the best option that the firms in the downstream market can take. Thus, according to the post-Chicago commentators, vertical constraints can be used to enhance the market power of the monopolist, excluding competitors and potential competitors in the market. In the previous section, it is explained that the Chicago school’s interpretation of Aspen Skiing is based on economic efficiency. How then do the postChicago theorists analyse the case? Aspen Skiing is described by a post-Chicago theorist as an example of how refusal to deal raises competitor’s costs.152 In Aspen Skiing, since most skiers preferred the all-Aspen ticket, a ticket for four slopes,153 the demand for a joint venture between Aspen Skiing and Aspen Highlands would be higher than the demand was when they did their business separately. Thus, if the competitors cooperated, the total profit would increase and the dominant firm, Aspen Skiing, could make more profit than it could when they did not cooperate. From the Chicago school’s perspective, there is no reason for the dominant firm to refuse to deal with the competitor. Then why did the dominant firm, Aspen Skiing, refuse to deal with its competitor, Aspen Highlands? The possible answer is that Aspen Skiing may have predicted that even though total market demand decreased, its market share would increase, providing more profit to it, since its facilities are more attractive than the competitor’s.154 By refusing to deal, the attractiveness of the dominant firm’s business as well as the competitor’s business would decrease but the competitor would suffer more damage than the dominant firm, which would eventually lead to the increase of the dominant firm’s market shares. Thus it is argued that refusal to deal can be used as a way to do harm to competitors, causing economic inefficiency. Lastly, the post-Chicago theorists interpret the barriers to entry in quite different ways from the Chicago school theorists’ perspective. As described before, the Chicagoans argue that natural barriers to entry are not policy concerns and only the artificial barriers to entry through exclusionary practices have to be condemned.155 The Chicago school also argues that since vertical integration does not create barriers to entry, it has to be assumed that the verti-

151 152 153 154 155

Ibid.: 270–74. Ibid.: 280–83. Aspen Skiing Co. v. Aspen Highlands Skiing Co., 605–7. Hovenkamp, ‘Antitrust Policy After Chicago’, 281–2. Bork, The Antitrust Paradox: A Policy at War with Itself 310–29.

156


cal integration increases efficiency.156 However, the post-Chicago scholars argue that vertical integration can create entry barriers.157 Entering a vertically integrated market is more risky and costly than starting a business in one level of market since a wider range of business skills and experiences are required. Furthermore, the risk is not avoidable even though two independent firms enter different levels of market simultaneously since coordination between the new entrants in different levels may not be successful and a barrier to entry at any level of market prevents coordinated entry. The barriers to entry resulting from vertical integration may not preclude new entry entirely but may significantly delay the appearance of new entrants.158 Thus the post-Chicago critics argue that the Chicago school’s analysis fails to consider the dynamic aspects of a market and that a dominant firm may behave strategically to increase its long term-profits, seeking to change the market structure. Their argument is that vertical constraints such as refusal to deal should not be considered as per se legal but rather need to be examined under the rule of reasons.159 3.4.3 Application to standards In the previous sub-sections, the economic theories of the Chicago school and the post-Chicago school, which are used in analysing markets and anticompetitive behaviours, have been studied. Then, how can the theories be applied to analyse the costs of patent systems in relation to standards? In analysing the behaviour of the owner of proprietary standards in network markets, Microsoft’s business strategies provide good examples. Since the time when Microsoft introduced MS-DOS in the personal computer operating system market, it has increased its market power and become a monopolist in the market. However, it is generally accepted that MS-DOS, the operating system software prior to Microsoft Windows, was inferior to competing operating systems like DR-DOS 5.0.160 In addition, Microsoft is not considered to be innovative in the computer industry.161 That is, it is not the superior qualities of MS-DOS or the innovative activities of Microsoft that make Microsoft a monopolist in the market. Then how could Microsoft become a monopolist?

156 Gerber, ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’, 1084–7. 157 Kaplow, ‘Extension of Monopoly Power through Leverage’, 536–9. 158 Ibid. 159 Langlois, ‘Technological Standards, Innovation, and Essential Facilities: Toward a Schumpeterian Post-Chicago Approach’, 12. 160 Sheremata, ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’, 964. 161 Ibid.: 947.


157

There are three possible answers to this question: network externalities, exclusionary practices and technical incompatibility. Firstly, network externalities are one of the main reasons why Microsoft has become a monopolist in the market. As described in Chapter 2,162 consumers’ decisions for early adoption depend heavily on the expectations of success of a standard163 and thus the conditions in the early stages of market standardisation are crucial to the outcome of the standardisation process. A small advantage over competitors may induce the cumulative support process such as to lead to ‘bandwagon effects’ or ‘tipping effects’, which allow the leader of a market to harvest a large advantage and to dominate the market.164 It is likely that consumers in the operating system markets expected that MS-DOS would be successful in the market, because it was adopted as the operating system of IBM computers. The fact that MS-DOS was used in IBM computers provided consumers with credible evidence that MS-DOS would be widely used, leading them to adopt MS-DOS and enlarging the installed base of consumers. The larger the installed base of users was, the more complementary products were developed for MS-DOS, which helped to increase the credibility of the standard and made the standard more attractive. Thus these bandwagon effects are one of the important factors which have enabled Microsoft to dominate the relevant market. Secondly, the licensing practice of Microsoft is also a crucial factor in the enhancement of the market power of Microsoft. Microsoft required original equipment manufacturers (hereinafter ‘OEMs’) of personal computers to accept ‘the CPU license’.165 Under the CPU licence, the price per copy of MSDOS was decided through negotiation between an OEM and Microsoft. An OEM had to agree on a minimum number of copies of MS-DOS, called a minimum requirement, and to pay a lump-sum licence fee for the minimum requirement. Where an OEM needed to buy MS-DOS above the minimum requirement, the licence fee for each additional copy had to be paid on the basis of the negotiated price. Where an OEM refused the CPU licence, Microsoft sold MS-DOS licences at considerably higher prices than it did where an OEM accepted the CPU licence. Furthermore, where an OEM sold computers with competing products like DR-DOS, the OEM could be given

162 163

See Chapter 2, Section 3.2.2. See Joseph Farrell, ‘Standardization and Intellectual Property’, Jurimetrics Journal 30 (1989). 164 Peter Grindley, Standards, Strategy, and Policy: Cases and Stories (Oxford: Oxford University Press, 2002) 27–8. 165 Kenneth C. Baseman et al., ‘Microsoft Plays Hardball: The Use of Exclusionary Pricing and Technical Incompatibility to Maintain Monopoly Power in Markets for Operating System Software’, Antitrust Bulletin Summer (1995).

158


no reduction in licence fees. In addition to the higher price, Microsoft showed its hostility in various ways to the OEMs that did not follow its request. For instance, it stopped providing technical support and sold Microsoft Windows at a higher price.166 Due to these consistent ‘efforts’ of Microsoft, the market share of MS-DOS grew from 70 per cent in 1990 to 82 per cent in 1992.167 Thirdly, keeping MS-DOS incompatible with competing products is another reason for Microsoft’s market power.168 Microsoft did not disclose some of the interface information of MS-DOS and if the unpublished interface information was discovered by others it changed or removed the disclosed information. In addition, Microsoft excluded its competitors from beta testing of MS-Windows and refused to solve possible compatibility problems with its competitors. On top of that, if competing products were detected in a computer system, Windows 3.1 showed a warning message to computer users, stating that there could be problems in running Windows 3.1 with non-Microsoft operating systems. Thus Microsoft sought to be incompatible with its competitors, which made it difficult for competitors to produce software that could be accepted in the market. How then are Microsoft’s business practices to be analysed in the light of economic theories? Firstly, it is clear that network externalities worked as a barrier to entry. The Chicagoans consider barriers to entry ‘anything that makes the entry of new firms into an industry more difficult’ and deny that natural barriers to entry are policy concerns, since whether a firm is an incumbent firm or a new entrant it has to overcome the same barriers in order to enter the market.169 However, it is clear that network externalities are not the same barriers to entry to every firm but make it more difficult for a new firm to enter the relevant market than an incumbent firm. The entry of a new firm into a network market is only possible in those circumstances where the firm pays more costs than an incumbent firm. That is, the barriers which a new firm has to overcome are higher than those that an incumbent firm has already overcome, because of network externalities and consumers’ switching costs. Therefore, it is more reasonable to argue that the barriers to entry erected by network externalities can deter the entry of a new firm even in those cases where the new firm is more efficient than an incumbent firm. Accordingly, 166 The first version of the Microsoft Windows operating system was launched in 1983. See Windows History (cited 30 December 2008), available from http://www. microsoft.com/windows/winhistorydesktop.mspx. In 1991, Microsoft Windows 3.1 was shipped together with MS-DOS in 18.5% of new PCs. See Baseman et al., ‘Microsoft Plays Hardball: The Use of Exclusionary Pricing and Technical Incompatibility to Maintain Monopoly Power in Markets for Operating System Software’, 7. 167 Ibid.: 8. 168 Ibid.: 12. 169 Bork, The Antitrust Paradox: A Policy at War with Itself 310–11.


159

network externalities may cause the market to be locked into a suboptimal network, increasing social costs. Thus, the barriers to entry resulting from network externalities have to be policy concerns. Secondly, the CPU licensing of Microsoft is a prototypical example of vertical constraints which are used to force the firms in the downstream market to accept exclusive dealing with the monopolist.170 Microsoft imposed various penalties on those OEMs that refused to follow its request. The Chicagoans argue that consumers, the OEMs in this case, do not prefer to deal exclusively with monopolists and thus they will not cooperate with the enhancement of the market power of a monopolist, which is quite plausible.171 Thus, it is reasonable to assume that some OEMs hoped to trade with Microsoft’s competitors because they did not want to be solely dependent upon a monopolist, as the Chicagoans argue. This assumption is further supported by the fact that the quality of DR-DOS, the competing product of MS-DOS, was superior to that of MS-DOS.172 However, the reality is that Microsoft’s market power increased, while the competing products of MSDOS were losing market shares. Then why could OEMs not refuse to accept the CPU licensing? Had they collectively refused to accept the CPU licensing, Microsoft might not have gained such a strong market power. The OEMs’ behaviour can be analysed by game theory and, in particular, the prisoner’s dilemma,173 which is one of the most famous games, may explain the OEMs’ behaviour. The prisoner’s dilemma game leads to a dominant strategy which is independent of the other player’s choice and results in suboptimal outcomes.174 Suppose that two criminal suspects, A and B, are arrested for several criminal charges and each of them is questioned in a separate room. Assume that, unless the suspects confess and provide related evidences, there is no way to prove the main charge. Suppose also that the jail term of each suspect depends on whether they confess or not. Where both of them confess their crimes, each of them will be sentenced to three years in prison. Where both of them do not confess, they will be convicted of only minor crimes and will be sentenced to two years in prison, since it is impossible to prove the main charge. Where one of them, say A, confesses and the other one, B, does not confess, A will be sentenced to one year in prison and B to five years, and vice versa. Under these circumstances, what will A and B do if they are rational? Where both of them remain silent, they can be released after two years in prison, which is the

170 171 172

Hovenkamp, ‘Antitrust Policy After Chicago’, 272–3. Kaplow, ‘Extension of Monopoly Power through Leverage’, 531–6. Sheremata, ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’, 964. 173 Begg et al., Economics 150–51. 174 Ibid.

160


optimal outcome of this game because the total jail term of the two criminals is four years. However, this optimal outcome is not achieved and both of them come to confess. How can this happen? Consider A’s decision. In making his decision, the suspect A has to consider what the suspect B will do and how the suspect B’s decision affects him. The suspect A firstly considers the case where the suspect B confesses. In this case, if the suspect A also confesses, each of them will be in prison for three years. However, if the suspect A does not confess, the suspect A has to be in prison for five years, while the jail term of the suspect B will only be one year. Thus, the suspect A makes the rational decision and chooses to confess. Now the suspect A considers the case where the suspect B does not confess. In this case, if the suspect A also remains silent, each of them will be in prison for two years. However, if the suspect A confesses, the jail term of the suspect A will be only one year. Thus, the suspect A chooses to confess. In any case, the suspect A chooses to confess, which is the dominant strategy of this game. The suspect B also makes a decision in the same way. Consequently, the optimal outcome cannot be achieved and the suboptimal solution is obtained. What then has the prisoner’s dilemma to do with the OEMs’ behaviour? Given that other OEMs can behave strategically, each OEM’s decision to accept Microsoft’s licensing offers is quite rational. As in the prisoner’s dilemma, if all OEMs had refused to follow Microsoft’s request, the optimal outcome might have been achieved since they could have freely made their own business decisions without any restrictions imposed by a monopolist. However, since only some of them rejected the CPU licensing and others accepted it, the rebellious OEMs had to suffer from various difficulties resulting from Microsoft’s punishment. Given that competing OEMs may behave strategically, the rational choice of an OEM is to accept the CPU licensing. This may be the reason why a monopolist’s behaviour can increase the market power even though the behaviours are not efficient. Thirdly, making it difficult for competitors to make their products compatible with MS-DOS is a way to raise rivals’ costs. Were MS-DOS compatible with competing products, consumers would have more choice between competing operating systems and would obtain more benefits by selecting superior products. However, Microsoft refused to make its products compatible with its competitors’ products. This incompatibility might reduce the attractiveness of MS-DOS itself, but it eventually did more harm to Microsoft’s competitors. This situation is similar to the circumstance of Aspen Skiing,175 where a dominant firm refuses to cooperate with its competitor even though a joint venture is likely to increase customers’ demand.

175

See Aspen Skiing Co. v. Aspen Highlands Skiing Co.


161

As the Microsoft case shows, a dominant firm may exercise its market power to exclude competitors. Like Microsoft, a dominant firm may use the strategies of vertical constraints and horizontal constraints. The CPU licensing is an example of vertical constraints used by a dominant firm to compel vertically related firms to deal exclusively with the firm. Making it difficult for competing firms to design products to be compatible with the product of a dominant firm is an example of horizontal constraints used to handicap rivals. In addition, as the post-Chicago theorists argue, vertical integration may also be used to deter the entry of potential competitors into the market. Considering the characteristics of network markets, this argument becomes more persuasive. A dominant firm may monopolise vertically related downstream markets and make products in the downstream markets incompatible with competing products in the upstream market.176 By doing so, the dominant firm can increase the installed base of its own network, decreasing rivals’ networks. For instance, it may be possible for Microsoft to monopolise the markets of application software such as word-processors, spreadsheets and media players and then design the application software to be incompatible with that of competing operating systems. By doing so, Microsoft can take advantage of the installed base of the application software, depriving competitors of their customers. The installed base tends to be self-expanding; that is, the larger the installed base is, the more complementary products are developed for the standard, which leads to more adoptions by users, which further enlarges the installed base. Therefore, vertical integration may also build barriers to entry into the market. As the above reasoning shows, the economic analyses of the post-Chicago theorists are more suitable to explain the network markets. From the purely static perspective, there is no reason for a dominant firm to impose vertical or horizontal constraints since such constraints do not increase the profits of the dominant firms. However, such constraints can be used strategically so as to change the structure of the relevant market and to maintain and expand the market power of the dominant firm.177 Therefore, it has to be accepted that vertical and horizontal constraints, including refusal to deal, may be used to increase the market power of a dominant firm, decreasing efficiency and causing social costs. Of course, it would be wrong to say that every vertical or horizontal constraint has to be condemned. The point is that the constraints imposed by a dominant firm may be abused in such a way as to exclude 176

A commentator calls this strategy defensive leveraging. For more detail, see Robin Cooper Feldman, ‘Defensive Leveraging in Antitrust’, Georgetown Law Journal 87 (1999). 177 Langlois, ‘Technological Standards, Innovation, and Essential Facilities: Toward a Schumpeterian Post-Chicago Approach’, 13.

162


competitors, suppressing beneficial technology. Accordingly, it can be concluded that the patents covering standards may also be abused to eliminate competition in the relevant markets and to hamper technological innovation, which is clearly detrimental to technical development.

4.

REVIEW AND IMPLICATIONS

In the previous sections, the costs of patent systems related to standards have been examined. It has been ascertained that the costs of patent systems arise from the facts that uniform patent rights are granted to heterogeneous innovations, that bargaining between patent owners and users can fail, that network externalities drive the network markets to have a concentrated market structure, and that the market power of the standard owner can be abused to deter beneficial innovations. It is necessary to consider each cost factor to find the ways to reduce cost. Firstly, it is clear that the cost arising from the application of the homogeneous patent system to heterogeneous innovations can be reduced by granting different patent rights to different innovations. Granting industry specific patent rights may be one option. One may try to grant different patent terms to each patent, depending on the characteristics of the industry to which an innovation belongs. Innovations in some industries may need more than 20 years’ protection, while innovations in other industries may require less than 20 years. For instance, pharmaceutical innovations may belong to the former category and business model patents may fit in the latter. Thus, it may seem desirable to design industry specific patent systems according to the special need of each industry or, more ideally, to design innovation specific patent systems which grant different rights to each innovation. Designing industry specific patent systems or innovation specific patent systems is ideal theoretically but, in practice, it is extremely difficult to achieve, if not impossible. Implementing innovation specific patent systems or industry specific ones requires insurmountable administration costs. It is impossible to find the optimal patent terms and protection scope in each industry. Given that it is even unclear whether the 20 years’ patent term is optimal or not in any one industry, it is almost impossible to find the proper patent term in a number of industries. Furthermore, it is common that an innovation extends over several industries. The patent classification clearly shows this aspect of an innovation. A patent is classified according to its technical features and, in classifying patents, the international patent classification (IPC) is used.178 A patent may

178

For IPC, see the WIPO web site at http://www.wipo.int/classifications/ipc/en/.


163

be given only one IPC code where it is so distinct that it belongs only to a specific IPC group. However, a large number of patents are given more than one IPC code because they cover several technical areas, which implies that the boundaries between industries are unclear. To illustrate, consider electronic displays such as an LCD (liquid crystal display) and an OLED (organic light emitting diode). Electronic displays cover various technical areas: the control units are related to electronics, the display unit is developed by chemical technology, and the frame of the product is designed by mechanical engineers. Thus, it is common that a patent application on electronic displays extends over electronic engineering, chemical engineering and/or mechanical engineering. These sorts of examples are numerous. How then to define the relevant industry and find the proper patent term which is given by industry specific patent systems? Thus, even if it is possible to design industry specific patent systems, the patent systems are unworkable since an innovation is not limited to a specific industry. There are other arguments against industry specific patent systems179 but suffice it here to say that implementing such patent systems is not feasible. Given that industry specific patent systems are infeasible, it is clear that implementing innovation specific patent systems is almost impossible. How then to reduce the costs arising from the uniform patent system? One way to do so may be to apply patent law doctrines differently to different innovations, as some commentators argue.180 Another way is to facilitate efficient bargaining. A great deal of the costs arising from the uniform patent system can be reduced should efficient bargaining between the patent owner and the user be achieved. This argument is clearly verified by noting that the costs of patent systems resulting from cumulative technologies and system technologies can be diminished by means of efficient bargaining. Thus, the way to reduce the costs of patent systems generated by the uniform patent system becomes the same as the measures to decrease the cost from bargaining breakdown. As for the costs resulting from the bargaining failure between standard owners and standard users, the previous discussion shows that it is necessary to limit the bargaining power of standard owners and to increase that of standard users in order to achieve efficient bargaining.181 Thus, the patent systems have to be reformed in such a manner as to limit the scope of patent rights covering standards. However, it has to be noted that any measures reducing the bargaining power of standard owners has to be balanced against the need to provide sufficient incentives to invent or innovate standards. Consequently, 179 180 181

Lemley, ‘Policy Levers in Patent Law’, 1630–38. Ibid.: 1638–95. For more detail, see Section 3.2 of this chapter.

164


the patent systems have to be reformed to increase the possibility of efficient bargaining between patent owners and users by limiting the scope of patents covering standards and at the same time to provide sufficient incentives to invent. Secondly, the cost arising from the network externalities of network markets can be decreased by facilitating compatibility between products. The network externalities increase the social costs of patent systems because of excess inertia generated by the lock-in effect and switching costs.182 These consumer lockin and considerable switching costs can be mitigated by making products compatible. Where the competing products are compatible, consumers can choose the most appropriate product mainly on the basis of the product’s quality and price. Where competing networks are compatible, consumers can easily switch from one network to another, having no need to worry about switching costs. Thus, compatibility between competing products and competing networks enhances consumer welfare. However, compelling compatibility between competing networks may reduce the incentives to innovate the networks in the first place and thus special care has to be paid to strike the right balance between the need for compatibility and the incentive to innovate. Lastly, the cost from abuse of the market power of patent owners can be diminished by limiting market power and the exercise thereof. In network markets, owning a compatibility standard is one of the main sources of market power and thus limiting the proprietary control of the compatibility standard can reduce the market power of the patent owner. By ensuring horizontal and vertical access of third parties to the compatibility standard at some stages, it may be possible to reduce the market power of the standard owner and to decrease the probability of abuse of market power. By making the compatibility standard available to competitors, it may be possible to have the standard owner feel competitive pressure and make an effort to innovate to survive in the market. Consequently, the measures to reduce the costs arising from abuse of the market power of patent owners become identical with the measures to decrease the costs from network externalities since they are ultimately compatibility problems. To sum up, the ways to reduce the social costs of patent systems in relation to standards have to be able to facilitate efficient bargaining and to restrict the market power of standard owners by ensuring access to the standards at some stages. With this analysis in mind, how to reduce the costs of patent systems arising from the factors discussed in this chapter is studied in the next chapter.

182 For more detail, see Farrell and Saloner, ‘Standardization, Compatibility and Innovation’.

7. Alternative solutions and a suggested solution 1.

INTRODUCTION

In Chapter 6, the costs of patent systems in relation to standards have been discussed. The costs resulting from uniform patent systems, bargaining breakdown, the special characteristics of a network market such as switching costs and network externalities, and the market power of dominant firms are discussed. It is argued that the way to reduce the social costs of patent systems in relation to standards is to find measures to facilitate efficient bargaining and to restrict the market power of standard owners by ensuring access to the standards at some stages. How then to decrease the costs of patent systems arising from the factors discussed in the previous chapter? That is what is studied in this chapter. Studies on the social costs resulting from patent systems are not new. Numerous studies have been conducted on this subject and a number of patent reforms have been suggested to reduce the social costs of patent systems.1 Amongst those studies, some researches argue for the radical change of patent systems,2 some studies suggest the modification of existing patent laws and

1 There are voluminous researches on patent reform. For instance, Ian Ayres and Paul Klemperer, ‘Limiting Patentees’ Market Power without Reducing Innovation Incentives: The Perverse Benefits of Uncertainty and Non-Injunctive Remedies’, Michigan Law Review 97 (1999); John H. Barton, ‘Reforming the Patent System’, Science 2000; Julie E. Cohen, ‘Patent Scope and Innovation in the Software Industry’, California Law Review 89 (2001); Kingston, ‘Innovation Needs Patents Reform’; Kingston (ed.), Direct Protection of Innovation; Kronz, ‘Patent Protection for Innovation: A Model’, Parts I and II; Amy L. Landers, ‘Let the Game Begin: Incentives to Innovation in the New Economy of Intellectual Property Law’, Santa Clara Law Review 46 (2006); Merges and Nelson, ‘On Limiting or Encouraging Rivalry in Technical Progress: The Effect of Patent Scope Decisions’; Merges and Nelson, ‘On the Complex Economics of Patent Scope’; Pamela Samuelson, ‘Patent Reform through the Courts’, Communications of the ACM 2007; Pamela Samuelson and Suzanne Scotchmer, ‘The Law and Economics of Reverse Engineering’, Yale Law Journal 111 (2002); Thurow, ‘Needed: A New System of Intellectual Property Rights’. 2 For instance, William Kingston, ‘Compulsory Licensing with Capital

165

166


legal doctrines,3 and some argue for sui generis law for the protection of specific subject matters.4 Some of the studies on patent reform have important implications for the circumstances where the openness of standards and the exclusivity of patents clash. Thus, some proposals for patent reform, which can be applied to the patent issues in standards, are studied and evaluated with respect to the costs, which are analysed in Chapter 6, and the benefits of patent systems. Since some aspects of the proposals for patent reform are not satisfactory in relation to the patent issues in standards, as is clear later in this chapter, ways to finetune patent systems with respect to standards are suggested. The proposals for patent reform which are analysed in this chapter are as follows: (1) application of liability rules,5 (2) modified compulsory licences – compulsory licences with lump sum payments,6 (3) adoption of the patent fair use doctrine,7 and (4) the competitive platform model.8 Some of the proposals for patent reform are suggested to cope with the problem arising from the patent protection of software-related inventions, but they can also be applied to the patents covering compatibility standards, which is clear later in this chapter. Payments as an Alternative to Grants of Monopoly in Intellectual Property’, Research Policy 23 (1994), Kingston (ed.)., Direct Protection of Innovation, Kronz, ‘Patent Protection for Innovation: A Model’, Parts I and II; Gerard Llobet, Hugo Hopenhayn and Matthew Mitchell, ‘Federal Reserve Bank of Minneapolis Research Department Staff Report 273, ‘Rewarding Sequential Innovators: Prizes, Patents and Buyouts’, (Federal Reserve Bank of Minneapolis, 2000). 3 For instance, Landers, ‘Let the Game Begin: Incentives to Innovation in the New Economy of Intellectual Property Law’; Robert P. Merges, ‘Who Owns the Charles River Bridge? Intellectual Property and Competition in the Software Industry’, (1999); Merges and Nelson, ‘On the Complex Economics of Patent Scope’; Samuelson, ‘Patent Reform through the Courts’. 4 For instance, Jerome H. Reichman, ‘Of Green Tulips and Legal Kudzu: Repackaging Rights in Subpatentable Innovation’, in Expanding the Boundaries of Intellectual Property, ed. Rochelle Cooper Dreyfuss and Diane Leenheer Zimmerman (Oxford: Oxford University Press, 2001); Jerome. H. Reichman, ‘Charting the Collapse of the Patent–Copyright Dichotomy: Premises for a Restructured International Intellectual Property System’, Cardozo Arts and Entertainment Law Journal 13 (1994–1995). 5 Dana R. Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’, Hastings Law Journal 51 (2000). 6 See Kingston, ‘Compulsory Licensing with Capital Payments as an Alternative to Grants of Monopoly in Intellectual Property’. 7 Maureen A. O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, Columbia Law Review 100 (2000). 8 Philip J. Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, Columbia Law Review 103 (2003).

Alternative solutions and a suggested solution

2.

PROPOSALS FOR PATENT REFORM

2.1

Application of Liability Rules

167

Patent rights are property rights which are created to solve the public goods problem of inventions,9 and thus the very essence of patent rights is their exclusivity. Without exclusivity, inventions would be undersupplied because of the free-rider problem. However, it is also true that this exclusivity is the main source of the social costs of patent systems. The deadweight loss and technology suppression arise from the exclusivity of patents and the reason why patent rights can be used in anticompetitive ways is that patents are exclusive rights. Where the costs resulting from the fact that patents follow property rules are considered to outweigh the benefits from it, an attempt can be made to correct this problem by applying liability rules to patent systems.10 Especially given that the life cycle of a product is short these days and that many inventions are created in cumulative ways, the exclusivity of a patent right may stifle technical innovation rather than promote it. In this respect, some may argue that liability rules have to be applied to resolve the problem caused by the property rights – the patent rights – and this argument is discussed in this sub-section. 2.1.1 Arguments for liability rules in systems technology Wagner argues that liability rules have to be applied to intellectual property covering ‘systems technology’ instead of property rules.11 The main concern

9

Lemley, ‘Ex Ante versus Ex Post Justifications for Intellectual Property’,

129–30. 10

For a general discussion of property rules and liability rules, see Guido Calabresi and A. Douglas Melamed, ‘Property Rules, Liability Rules, and Inalienability: One View of the Cathedral’, Harvard Law Review 85 (1972). For a discussion of these rules in relation to bargaining, see Ian Ayres and Eric Talley, ‘Distinguishing between Consensual and Nonconsensual Advantages of Liability Rules’, Yale Law Journal 105 (1995a); Ian Ayres and Eric Talley, ‘Solomonic Bargaining: Dividing a Legal Entitlement to Facilitate Coasean Trade’, Yale Law Journal 104 (1995b); Louis Kaplow and Steven Shavell, ‘Do Liability Rules Facilitate Bargaining? A Reply to Ayres and Talley’, Yale Law Journal 105 (1995); Louis Kaplow and Steven Shavell, ‘Property Rules Versus Liability Rules: An Economic Analysis’, Harvard Law Review 109 (1996). For a discussion of these rules in the context of intellectual property, see Merges, ‘Contracting into Liability Rules: Intellectual Property Rights and Collective Rights Organizations’; Robert P. Merges, ‘Of Property Rules, Coase, and Intellectual Property’, Columbia Law Review 94 (1994b); Yosick, ‘Compulsory Patent Licensing for Efficient Use of Inventions’. 11 For more detail, see Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’.

168


of his study is ‘systems technology’, which is defined as ‘technology that defines and governs the computing environments within which people operate’.12 Systems technology includes computer hardware, such as microprocessors, the bus design and I/O interfaces, and software such as the operating system, internet browsers and application software.13 It is argued that the special features of systems technology require the application of liability rules.14 Firstly, it is claimed that systems technology exhibits network externalities, which arise from the circumstances where the value derived from the consumption of products is affected by the number of other people who use the products or compatible products.15 It is also argued that the systems technology market has the characteristics of both direct network externalities and indirect network externalities, which drives the market to standardisation.16 Since the markets for systems technology have these features, the standard technology is essential for a firm to enter the relevant markets and thus the need for the open standard clashes with the exclusivity of patent rights where the standard is protected by patents. Secondly, it is contended that systems technology has the characteristic of ‘interconnectivity’, which refers to the situation where products have to be interconnected to compose a complete product.17 For instance, in order to make a computer system, systems technology has to be connected to other components such as memory, I/O systems and various peripherals, and all of these components have to be compatible. It is clear that this interconnectivity is closely related to the network externalities. The interconnectivity, along with network externalities, inevitably causes the market to be locked into a standard, increasing the possibility of a conflict between standards and patents. Thirdly, it is also argued that, in systems technology markets, innovation is rapid and the life cycle of a product is short.18 Because of these features, the term of patent protection, usually 20 years from the date of the patent application, is too long for the effective life of a technology. The basic logic of patent systems is that an inventor enjoys the exclusive rights for 20 years, recouping the investment made to create the technology, and after the expiration of the patent term the patented technology is donated to the public. 12 13 14 15

Ibid.: 1081. Ibid. Ibid.: 1096–1109. Lemley and McGowan, ‘Legal Implications of Network Economic Effects’,

483–4. 16 For the discussion of direct and indirect network externalities, see Chapter 2 or Shapiro and Katz, ‘Network Externalities, Competition and Compatibility’. 17 Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’, 1098–9. 18 Ibid.: 1099–1100.


169

According to this logic, patent systems seek to balance social benefits against private benefits by granting the exclusive right for a fixed term. However, where the effective life of a technology is far shorter than the patent term, patents protect innovations more than necessary, suppressing beneficial technology subsequent to the patented technology. Furthermore, since the pace of innovation is fast, not only the outright denial of access to a technology but also the delay of it may result in the exclusion of a firm from the systems technology market. Fourthly, it is argued that the exclusion of competitors by the owner of systems technology by means of keeping the internal information secret or by implementing exclusionary engineering is easy even without legal protection.19 It is contended that, since systems technology is complex, it would be impossible to produce compatible products without the internal information of systems technology. In addition, the owner of systems technology may design the structure of the technology to be incompatible with competing ones, which results in the exclusion of competitors from the relevant markets. Lastly, it is argued that the exclusivity of patent rights can be used in anticompetitive ways.20 The owner of property rights can prevent others from using the subject matters protected by the rights and thus the owner of systems technology may exercise the property rights to exclude competitors by refusing to grant licences or to expand market power from one market to another market by means of tie-in. Wagner argues that the costs arising from these features of the systems technology market can be reduced by adopting liability rules. In relation to network externalities, it is argued that since the optimal level of compatibility is not achieved from the market and since the firms controlling standards have a large incentive to abuse the power gained from standards, the legal system has to be designed to provide compatibility and open access to standards.21 As for the interconnectivity, it is argued that open access to technology for interconnection is crucial for innovation and competition in the market and the need for access to technical information for interconnection increases the likelihood of standard owners’ strategic behaviour.22 In addition, it is claimed that since the term of intellectual property protection is much longer than the effective life of systems technology, the intellectual property rights following property rules increase the social costs, which can be reduced by liability rules.23

19 20 21 22 23

Ibid.: 1100–01. Ibid.: 1107–09. Ibid.: 1102–04. Ibid.: 1105–06. Ibid.: 1106.

170


It is also argued that intellectual property protection following property rules can be abused anticompetitively in various ways. Namely, it is contended that the firm controlling systems technology may refuse to license to the producers of complementary products to increase economic benefits, that the firm may exercise its exclusive rights to prevent competitors from entering the market, and that the firm may grant access to systems technology on condition that other firms buy auxiliary products which they do not need to purchase. Wagner argues that such anticompetitive behaviour can be reduced by adopting liability rules since, under liability rule regimes, the intellectual property owner of systems technology cannot exclude others from exploiting the protected technology but instead is entitled to receive financial compensation for the use of the technology.24 Thus, analysing the effect of liability rules on the characteristics of systems technology such as network externalities, interconnectivity, rapid innovation, excludability and anticompetitive behaviour, he concludes that liability rules are better for systems technology than property rules. Wagner also analyses the possible criticism of the liability rule regimes.25 The main criticisms of liability rules are that they may reduce incentives to innovate in the first place and that deciding the reasonable compensation is difficult.26 However, it is argued that the incentive to innovate is not greatly reduced because the first mover advantages may provide sufficient incentives to innovate and because innovators can extract compensation from the users of systems technology, and that innovation generated by the spillover effects of liability rules would outweigh the reduction of incentives caused by liability rules.27 As far as the valuation problem is concerned, Wagner argues that it can be resolved by establishing processes for valuations.28 Moreover, it is highly likely that other market transactions which may be used to decide the proper compensation already exist since systems technology is widely used in the market.29 All in all, Wagner argues that liability rules are superior to property rules in relation to systems technology in that they facilitate the wide use of the technology, they do not reduce incentives to innovate significantly, and they do not cause additional costs. 24 25 26

Ibid.: 1107–09. Ibid.: 1109–18. See Merges, ‘Contracting into Liability Rules: Intellectual Property Rights and Collective Rights Organizations’, 1303–08; Tsilas, The Perils of Imposing Compulsory IP Licensing to Achieve Interoperability. 27 Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’, 1112. 28 Ibid.: 1115–17. 29 Ibid.


171

2.1.2 Evaluation and implication The subject matter of Wagner’s study is ‘systems technology’30, whereas that of this study is standards in network markets. As noted, systems technology is defined as ‘technology that defines and governs the computing environments within which people operate’31 and includes computer hardware, such as microprocessors, the bus design and I/O interfaces, and software such as the operating system, internet browsers and application software.32 In addition, systems technology has the characteristics of network externalities, interconnectivity, and rapid innovation and is easily abused to exclude competitors.33 Considering the definition, the scope and the features of systems technology, it seems that there is no difference between systems technology in Wagner’s argument and the compatibility standards of this book. Thus, if his argument is correct, applying liability rules to patents covering standards would resolve the conflict between patents and standards. Then do liability rules really solve the problems? In order to answer this question it is necessary to examine liability rules with respect to both the costs and the benefits of patent systems. Consider the costs first. It has previously been argued that facilitating efficient bargaining and restricting the market power of standard owners by ensuring access to the standards at some stages would reduce the costs of patent systems in relation to standards. Thus, it is necessary to examine whether liability rules promote efficient bargaining34 and restrict the market power of standard owners. Firstly, consider bargaining between patent owners and users. It is generally accepted that liability rules are superior to property rules in those cases ‘where there are obstacles to cooperation (i.e., high transaction costs)’ and the reverse is true ‘where there are few obstacles to cooperation (i.e., low transaction costs).’35 As far as patents are concerned, in most cases only two parties, a patentee and a licensee, participate in bargaining and the transaction costs are low and thus property rules are appropriate.36 However, there are some cases where the transaction costs are high in relation to patents covering standards. That is, it is unlikely that the owner of standards will bargain with firms having competing technology which may replace the standard technology. 30 31 32 33 34

Ibid.: 1081. Ibid. Ibid. Ibid.: 1096–1109. It is reasonable to assume that voluntary bargaining between bargaining parties is most efficient bargaining since bargaining parties themselves have more information about the bargaining than others. Thus, efficient bargaining here refers to private and voluntary bargaining. 35 Cooter and Ulen, Law and Economics 103. 36 Merges, ‘Of Property Rules, Coase, and Intellectual Property’, 2664–5.

172


Rather, the owner of standards is likely to seek to exclude the competitors from the markets, however beneficial the competing technology may be. The relationship between competitors is generally ‘hostile’ and thus the transaction costs become high.37 In this case, liability rules are superior to property rules. Consequently, it can be concluded that liability rules do not generally facilitate efficient bargaining between standard owners and standard users but in some cases, where bargaining is infeasible, liability rules are desirable. As for the effects of liability rules on market power, liability rules effectively restrict the market power of standard owners. Under liability rule regimes, anyone hoping for access to patented technology can obtain it as long as he pays for it and thus the owner of patents covering standards is not able to exclude competitors from the market. Consequently, liability rules would lower the barriers to entry into the relevant market, making the market more competitive than property rule regimes would do. Probably this is the main reason why Wagner argues for liability rules. To sum up, as far as the effects of liability rules on the costs of patent systems are concerned, liability rules reduce the costs arising from the market power of patent owners. However, liability rules do not facilitate efficient bargaining between a standard owner and a user in general and they decrease the costs from bargaining failure only in some cases. Now, consider the benefits of patent systems in relation to standards. Briefly speaking, the benefits of patent systems can be summarised as providing incentives to invent, innovate and disclose, as is described in Chapter 6. Outright application of liability rules to patents covering standards would significantly decrease the benefits of patent systems, the reasons being as follows. Firstly, contrary to Wagner’s arguments, liability rules may significantly reduce the incentive to invent or innovate. Wagner argues that the first mover advantage and the compensation from the users of systems technology ‘may’ provide sufficient incentives to innovate.38 However, they ‘may not’ do so. There is no guarantee that first mover advantages would provide a sufficient incentive and that the licensing fees from the users would be sufficient. Wagner’s arguments assume that systems technology would always provide sufficient rents to the owners by way of the advantage gained from the head start or licensing fees, which is not necessarily true. There are historical examples which contradict his arguments. For instance, Sony’s Beta-Max video player was the first video player in the related market but it was defeated by 37 For various elements comprising transaction costs, see Cooter and Ulen, Law and Economics 87–90. 38 Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’, 1109–12.


173

JVC’s VHS video player, and the personal computer was first introduced by Apple but the personal computer market is dominated by IBM compatible computers.39 In both cases, it is not clear whether the first movers have recouped sufficient profits. Thus, even in a network market, first mover advantages may not provide sufficient incentives to innovate in some cases. In addition, licensing fees may not be sufficient to recoup the investment in innovation. Especially, liability rules may be taken advantage of by large firms, which have considerable resources. Suppose that a small or medium size firm is commercialising a systems technology which is protected by patent rights, that the technology is promising, and that the market for the technology is in an early stage. Under property rules, the firm does not have to worry about other firms’ exploitation of the technology without its consent and is likely to obtain sufficient rents as the relevant market expands. However, under liability rules, large firms having sizeable resources for marketing may use the technology with the payment of ‘reasonable’ licensing fees, expelling the small firm that invents the technology. In this respect, Wagner does not recognise the difference between horizontal relationships and vertical relationships or between competitors and complementors in the market. Open access to complementors, who are vertically related to the owner of a systems technology, may not do harm to the owner of the systems technology, but should the owner always have to grant open access to competitors, the incentives to invent and innovate would be significantly diminished. Secondly, Wagner fails to take into consideration the lead time, which may be essential for an inventor to recoup the R&D costs. As the aforementioned example shows, should the owner of the systems technology have to grant competitors open access to the technology in an early stage of the relevant market, the owner may not have enough time to recover the costs of the innovation. One of the main functions of patent protection as an incentive to invent or innovate is that it takes competitors a certain amount of time to invent around the patented invention, which creates the lead time during which the patent owner may obtain rents from his invention. However, where liability rules apply, such lead time cannot be obtained. Competitors may exploit the systems technology at any time they wish without having to negotiate with the owner or to wait until the patent right expires. This may significantly reduce the incentive to invent or innovate in the first place, causing the free-rider problem. Thirdly, the outright application of liability rules to systems technology may be detrimental to the emergence of beneficial technologies. The competitors of a systems technology do not have to make an effort to invent around or

39

Grindley, Standards, Strategy, and Policy: Cases and Stories 75–98, 131–55.

174


to make better products than the systems technology because all they have to do about the systems technology is to pay licensing fees and to exploit it. Thus, the competing technology which would otherwise come into being may not appear. Historical evidence supports this argument. Had Sony’s Beta-Max video technology been open to competitors including JVC, VHS technology might not have been introduced in the VCR market. Similarly, IBM’s PC might not have been invented, had Apple’s Macintosh computer been available to IBM. Thus, applying liability rules to systems technology may impede technical innovation. Consequently, liability rules are likely to significantly reduce the benefits of patent systems. Arguably, the reduction in the benefits of patent systems outweighs the decrease in the costs of patent systems. Indeed, Wagner’s argument seems to be that patent protection of systems technology has to be abolished altogether, which is too radical and unqualified. Admittedly, applying liability rules in some cases may be desirable, but outright application of liability rules to systems technology may destroy the incentive function of patent systems. Thus, it is essential to distinguish between the cases where liability rules are desirable and the cases where property rules are superior. Despite these weaknesses, Wagner’s argument on valuation difficulties is notable. Wagner argues that the valuation of compensation, which inevitably arises from liability rules, is not a problem since a well designed process can resolve the possible problem and since there are many other similar transactions which may provide useful guidance on the proper compensation.40 This argument seems persuasive and can be applied to compatibility standards in the same way. Since standards are, by definition, widely used in the relevant market, it is highly likely that there are similar licensing agreements which can provide useful information on the proper level of licensing fees, and thus, even if the dealing between a standard owner and a user is not voluntary but compelled by courts or by third parties, setting the appropriate licensing fee may not be problematic. In addition, various valuation processes such as finaloffer arbitration,41 which is based on the information provided not by courts but by bargaining parties themselves, may resolve the possible valuation problem. Furthermore, the analysis of liability rules’ effects on the costs and benefits of patent systems implies that liability rules may be used in fine-tuning patent systems in relation to standards. As previously described, liability rules are superior to property rules in reducing some costs of patent systems but are inferior to property rules in increasing the benefits of patent systems. Thus, 40 Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’, 1115–17. 41 Ibid.


175

where patent law is designed to magnify the advantages of liability rules and to minimise the disadvantages, it would be possible to use liability rules in fine-tuning patent systems. 2.2

Compulsory Licences with Capital Payments

Professor Kingston suggests several notable patent reforms in reaction to the costs arising from the current patent systems, including compulsory arbitration of patent disputes, direct protection of innovation and compulsory licensing with capital payments.42 Amongst the patent reforms which he proposes, the suggestion for compulsory licensing with capital payments is related to the bargaining problem between patent owners and the potential users, which may be applied to the patent issues in standards. Thus, the proposed regime of compulsory licensing with lump sum payments is studied and its implications to patents covering standards are examined in this sub-section. 2.2.1 Arguments for compulsory licences with capital payments Compulsory licensing with capital payments is to grant compulsory licences to anyone hoping to obtain licences of patents, on condition that he pays a multiple of the investment made in the patented innovation, where the multiple is determined to reflect the risk of the investment.43 Kingston begins the analysis by examining how the current patent systems work to promote inventions and innovations. He finds that patent systems operate well in some industries like the chemical and pharmaceutical industries but work poorly in other industries.44 He argues that in many industries patents are not particularly important ways to protect inventions since there are other ways to do so but that patents are crucial in protecting inventions in the chemical and pharmaceutical industries. Since the chemical formulas of chemical and pharmaceutical inventions are easily identified and copied by competitors, the industries would suffer from free-riding without patent protection.45 By reviewing the history of the development of American patent systems, Kingston argues that patent systems are designed to be suitable for the chemical and pharmaceutical industries and thus work well in those industries, as follows.46 In the early 20th century, many pharmaceutical inventions were not eligible for patent protection since they could not meet the patentability

42 43

See Kingston, ‘Innovation Needs Patents Reform’. Kingston, ‘Compulsory Licensing with Capital Payments as an Alternative to Grants of Monopoly in Intellectual Property’, 665–6. 44 Kingston, ‘Innovation Needs Patents Reform’, 403–10. 45 Ibid. 46 Ibid.: 403–7.

176


requirement. At that time, patent law was designed to protect innovation by individuals and in order to get patented an innovation had to be ‘something more than the work of a skilled’ person or to be the outcome of a ‘flash of creative genius’.47 Yet, many chemical and pharmaceutical innovations were the outcome of massive investment rather than of a ‘flash of creative genius’ and thus they could not meet the requirement of a ‘flash of creative genius’. There was a big gap between innovation structure in practice and patent protection at that time and thus patent law needed to be changed to fill the gap. Consequently, in 1952 American patent law was changed and the patentability requirement of a ‘flash of creative genius’ was replaced by the ‘inventive step’ requirement to make it possible to grant patents for inventions from large investments. Thus it is argued that patent law works well in the chemical and pharmaceutical industries since it is designed to cope with the problems faced by those industries. However, it is claimed that patents are not well fitted to many other industries and are often used as bargaining chips in patent pools and cross-licensing or to exclude competitors, thus failing to work as an incentive system.48 Kingston contends that patent rights suppress technology diffusion, which is essential in the innovation process, and that current patent systems do not solve this problem efficiently.49 Especially, he notes the importance of competitive rivalry and incremental innovations in promoting technical advances and argues that early access to patented inventions is crucial in promoting innovations.50 One way to reduce the costs resulting from technology suppression and to allow early access to patented inventions is to impose compulsory licences. However, compulsory licences are likely to significantly reduce the incentive to innovate, which patent systems seek to provide. Thus the basic problem which Kingston’s proposal seeks to resolve is that while compulsory licences are beneficial to the diffusion of innovation, they are likely to erode the incentive to innovate, failing to compensate the risk which the innovator has to face in making a decision to innovate.51 Thus, a ‘new kind of compulsory licensing’ is required to balance the need for fast diffusion of innovation against the need for incentives to innovate and it is argued that the compulsory licensing with capital payments meets this requirement.52 Kingston argues that the negative effect of compulsory licences on the incentive to invent or innovate

47 48 49 50 51

Ibid.: 405. Ibid.: 407–9. Ibid.: 418. Ibid. Kingston, ‘Compulsory Licensing with Capital Payments as an Alternative to Grants of Monopoly in Intellectual Property’, 663–5. 52 Kingston, ‘Innovation Needs Patents Reform’, 418.


177

can be offset by paying to patent owners the lump sum money as licensing fees.53 The lump sum money is decided by R&D costs and the risks of investment in conducting R&D and, more precisely, the capital payment is the result of the multiplication of R&D costs and risk-compensated multiples. It is clear that the precondition of this proposal is to find the R&D costs and the multiples reflecting the risk of investments. As for the R&D costs, Kingston argues that it is a well-established practice for a patentee to keep the record of research expenditure and thus it is not difficult to find the R&D cost of an innovation. As far as the multiples are concerned, he calculates the risks of investment at several stages of the innovation process with the data from a US research program in order to demonstrate how to find the proper multiples. In finding the appropriate multiples of investments, he uses over 20 000 data from the research projects supported by the US government under the Small Business Innovation Research Act of 1982 (hereinafter ‘SBIR’).54 He divides the innovation process into four stages – the idea formulation stage, the stage after the feasibility study, the reduction to practice stage, and the commercialisation stage – and calculates that the probabilities of reaching each stage are 0.01, 0.09, 0.18, and 0.33, respectively.55 Thus, the probability-weighted multiples, which are the inverses of the probabilities, are 100, 11, 5.6, and 3, respectively. Given the R&D costs and the multiples reflecting the risk, it is possible to calculate the lump sum which the licensee has to pay to the licensor. For instance, where the investment made to reach the stage of reduction to practice is £1M, the licensee has to pay to the innovator the lump sum of £5.6M, which is obtained by multiplying the investment (£1M) by the probability-weighted multiple (5.6). Thus Kingston argues that it is possible to find the proper capital payment with the given R&D costs and sophisticated accounting techniques. He also contends that compulsory licences with lump sum payments would facilitate the improvement of an innovation, provide a proper protection for software, reduce the patent litigation, and resolve the anticompetitive problem arising from the blocking effects of patents, while providing sufficient incentives to innovate.56 He admits that more researches are required to implement such a radical reform of patent systems and finding the appropriate multiples would be especially difficult.57 53 54

Ibid.: 419. Kingston, ‘Compulsory Licensing with Capital Payments as an Alternative to Grants of Monopoly in Intellectual Property’, 666–71. 55 Ibid.: 666–7. 56 William Kingston, Meeting Nelson’s Concerns about Intellectual Property (accessed 8 April 2007); available from http://www.druid.dk/conferences/nw/paper1/ kingston.pdf. 57 Ibid.

178


2.2.2 Evaluation and implication First of all, the feasibility of the proposed model has to be examined. Kingston’s suggestion for compulsory licences with lump sum payments is difficult to implement, since it is hard to find the proper multiples for each patent. Admittedly, it would be possible with reasonable costs to find the R&D cost related to patented technology since it is well established that a firm keeps a record of R&D expenditure. However, as Kingston admits, the main difficulty of applying this model is ‘to develop multiples of these costs (R&D costs) which would offer incentives for risky investment in R&D that are at least as good as from the present system of intellectual property’.58 Finding the proper multiples is crucial in this model since ‘with a lower multiple, the risk would be regarded as too high, and the investment would not be made’, whereas ‘with a higher one, the protection would be more than it needs to be’.59 Since the risk of investments is different by industries, different multiples have to be used for each industry, as Kingston notes.60 However, it is difficult to find each multiple for a number of industries. Indeed, finding appropriate multiples even in one industry is not easy. Kingston himself shows how to compute the proper multiples with the SBIR data, dividing the innovation process into four stages. However, even though the division looks plausible, it is an arbitrary one. The innovation process is continuous, not discrete, and thus there are a number of stages inbetween the four stages. Therefore, it would be difficult to find the multiples of a number of stages in a number of industries. Given that measuring the optimal multiples for each innovation is not feasible, finding the correct lump sum payment is ‘by sheer accident’, just as Kingston criticises the uniform patent terms for every invention,61 and the lump sum payment would be lower or higher than optimal. Accordingly, the costs of finding proper multiples for every patent in a number of stages far outweigh the benefits expected from the model of compulsory licensing with capital payments and thus replacing the current patent systems with the regime is not feasible. However, even though it is not feasible to abolish current patent systems and adopt the model of compulsory licensing with capital payments, limited use of the model in relation to standards may be possible and desirable. The proposed patent reform may be useful in resolving disputes between standard owners and users. The court or the Patent Office can intervene only in those cases where standard owners and users fail to bargain voluntarily, and can use the model of compulsory licensing with capital payments in dispute resolu58 59 60 61

Ibid. Ibid. Ibid. Kingston, ‘Innovation Needs Patents Reform’, 418.


179

tion. As is argued, the main problem in applying the suggested proposal is to find the proper multiples for every invention since it is difficult to develop the proper multiples for each invention. However, it may be possible to calculate the proper multiple of some specific patents and, in particular, as far as standards are concerned, it may be possible to find the capital payment in relation to patents covering standards. Patents covering standards are at the commercialisation stage and thus only the probability of reaching the commercialisation stage has to be calculated. In finding the proper multiples for the commercialisation stage, the court or the Patent Office can use the data submitted by each party of the dispute. Furthermore, the court or the Patent Office may conduct empirical research to find the proper multiples in several industries where the dispute is likely to arise and use the data in actual cases. That is, the court or the Patent Office does not have to calculate the proper multiples of every patent but only has to find the multiples for some specific patents in the limited circumstances where bargaining fails. Thus, the cost to find the proper multiple is reduced significantly such that the model may be used in limited ways. Given that the proposed model as such is not feasible, it is of no use examining the model with respect to the costs and benefits of patent systems. Nevertheless, the model has an important implication for the fine-tuning of patent systems. The model shows that in providing incentives to invent or innovate by means of patent systems, it is essential to consider the risk which innovators have to face in making investments in innovation. The monetary investment itself is not the real cost which the innovators have paid in innovation since they had to face the risk of failing. Thus, taking the risk into consideration in calculating licensing fees is necessary and this seems to be the most remarkable feature of this model. Another significant characteristic of this model is to design the incentive structure of patent systems on the basis of the input side of innovation rather than the output side. That is, the lump sum payment is calculated on the basis of the input of innovation such as R&D costs and risks, rather than the output such as the fair market value of the patent. The fair market value of a patent may vary depending on the patent’s prospect and the users’ ability to use the patent. Some patents may have less value than R&D costs, while some patents may have far greater value than R&D costs. On the other hand, R&D costs and risks can be calculated independently of the market value of the patented invention. Where the innovator can recover the real investment, the riskcompensated R&D costs, the incentive structure of patent systems can operate effectively. In this respect, Kingston’s argument is correct. This model can be useful in deciding whether to limit patent rights. Where the patentee has recovered the risk-considered R&D costs, he has arguably recouped sufficient reward from the patented innovation. The patent right has

180


completed its mission in providing sufficient incentive to invent or innovate and thus the exclusivity of the patent right may be limited for other reasons such as compatibility, subsequent innovation and so on. To sum up, the model of compulsory licences with capital payments suggested by Kingston is so radical and difficult to implement that it may not be used to replace the current patent systems. Nevertheless, it can be used to find the minimum incentive which patent systems have to provide. That is, patent systems have to ensure that the patentee recovers his real costs, the risk-compensated R&D costs, and once when the patentee has earned enough money to cover the risk-weighted R&D costs the patent rights should not be restricted. 2.3

Patent Fair Use Doctrine

There can be various approaches to the fine-tuning of patent systems. Some may seek to introduce liability rules into patent law in order to address the problems arising from the fact that patent law follows property rules, as Wagner and Kingston do.62 Also, some may make an attempt to use the legal principles of other intellectual property laws such as trademark law and copyright law to fine-tune patent law. Indeed, Professor Maureen A. O’Rourke proposes the introduction of the fair use of copyright law into patent law to reduce the cost of patent systems.63 She argues that the well designed fair use doctrine of patent law would be useful in overcoming the challenges that current patent systems face in accommodating new technology. Since her arguments and analyses are relevant to the issues of this book, they are examined and evaluated here. In order to do so, the fair use doctrine of copyright law has to be studied first and then the application of the doctrine to patent issues in standards has to be analysed. 2.3.1 The fair use doctrine in copyright law64 The fair use doctrine is prescribed in the US copyright law and is used by

62 See Kingston, ‘Compulsory Licensing with Capital Payments as an Alternative to Grants of Monopoly in Intellectual Property’; Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’. 63 See O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’. 64 The fair use doctrine in the US differs from the fair dealing doctrine in the UK in that the fair dealing only applies for the purposes approved by UK copyright law, namely research, private study, criticism, review or news reporting. In contrast, the US copyright law has a general fair use defence such that where the court considers an alleged infringement fair, there is no infringement. For more detail, see Bently and Sherman, Intellectual Property Law 193–206. To focus on the related arguments, only the fair use doctrine of the US is being studied.


181

alleged infringers to defend themselves against the copyright infringement charges.65 The fair use doctrine is an equitable doctrine used to adjust the copyright and, if successful, the alleged infringers can obtain the right to infringe the copyright with no licensing fees.66 The fair use is justified by market failure caused by market barriers such as insurmountable transaction costs which prevent voluntary bargaining in the market, externalities that make the infringers unable to pay the proper licensing fees, and the unwillingness of copyright owners to disseminate the copyrighted work.67 Firstly, where a bargain of a copyrighted work in a market is unavailable, the fair use of the work may be justified.68 There can be various reasons why the bargain of a copyrighted work in a market is not attainable. Where a copyrighted book is out of print, the bargaining to obtain the book is not likely to happen. In addition, the transaction cost of a bargain which is higher than the benefits expected from the bargain would prevent voluntary bargaining. For instance, should an individual have to obtain permission from copyright owners in using a photocopying machine or a video cassette player at home, the transaction costs arising from the attempt to get the permission would far outweigh the benefits anticipated from the use of the copyrighted works, which would discourage the use of the copyrighted works. Moreover, it would be difficult, if not impossible, and unprofitable for a copyright owner to

65 17 USC § 107. Limitations on exclusive rights: Fair use Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include— (1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work. The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors. (Here, §106 refers to the section dictating ‘Exclusive rights in copyrighted works’ and §106A stipulates ‘Rights of certain authors to attribution and integrity’. For more detail, see 15 USC §106 and §106A.) 66 O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1188. 67 Wendy J. Gordon, ‘Fair Use as Market Failure: A Structural and Economic Analysis of the Betamax Case and Its Predecessors’, Columbia Law Review 82 (1982): 1627–35. 68 Ibid.: 1627–30.

182


enforce the copyright of a book or a movie against individuals infringing the copyright at home since the cost of the enforcement is greater than the profits expected from the enforcement and the enforcement of copyright will stop the use of the copyrighted works, creating no income to the copyright owner and reducing the social benefits from the copyrighted work. In such cases, the fair use doctrine serves the interests of copyright owners, infringers and society. Secondly, the market failure caused by externalities may justify the fair use of copyrighted works.69 The exploitation of copyrighted works by third parties may create a large social benefit but the third parties may not be able to pay the fees for the use of the copyrighted works, since the benefit is not internalised but externalised so as to prevent the parties from earning money to pay the licensing fees. For instance, suppose a critic produced a commentary using a copyrighted work which gave rise to a heated debate in society. Since the commentary became a big issue, it was discussed in TV, newspapers and radios, giving information to people who did not buy the commentary. Because the critic used the copyrighted work, he had to pay a licensing fee to the copyright owner. However, the critic might not gain profits from the commentary because the external benefits, which the general public obtained without buying the commentary, were not internalised to be his benefits. These externalities prevent the appearance of a beneficial work which utilises copyrighted works, causing market failure. Thus, where market failure caused by the existence of externalities prevents the emergence of socially beneficial works, the fair use doctrine might condone the infringement. Thirdly, the unwillingness of copyright owners may hinder the voluntary bargaining of a copyrighted work in a market, causing market failure.70 The typical examples of this market failure are criticism, commentaries and satires. It is to be expected that a copyright owner would be reluctant to grant licences of his copyright to those who criticise his work and to his competitors, however large are the social benefits created from the exploitation of the copyrighted work. Clearly, criticism, commentaries and satires produce social benefits since they may provide valuable insight into the copyrighted work which they deal with and on top of that they themselves are other copyrightable works. Therefore, where the ‘anti-dissemination motives’ of copyright owners are likely to impede the beneficial exploitation of the work, the fair use doctrine may be used to facilitate the dissemination of the work.71 Thus, the fair use doctrine of copyright law can be used in various circumstances but most notably the doctrine has been used in relation to the compatibility issues of network industries. The video game cases such as Atari v. 69 70 71

Ibid.: 1630–32. Ibid.: 1632–5. Ibid.: 1632.


183

Nintendo, Sega v. Accolade, and Sony v. Connectix are the cases where the fair use doctrine is applied in courts’ rulings. In Atari v. Nintendo,72 Atari, a computer game developing company, used reverse engineering to decipher the locking program (10NES) of Nintendo, which was designed to prevent the use of unauthorised game cartridges in Nintendo’s game console, in order to produce games compatible with the console.73 Atari obtained the source code of the 10NES program from the Copyright Office by falsely alleging that the source code was necessary in the litigation where Atari was a defendant, and used it to debug errors in the reverse engineering. With the outcome of the reverse engineering, Atari made the Rabbit program, which unlocked 10NES. Nintendo filed a suit against Atari, alleging that Atari infringed its 10NES copyright, and in response Atari alleged that Nintendo violated antitrust law and misused copyrights. The Federal Court recognised that Atari made intermediate copies of the copyrighted 10NES program in undertaking the reverse engineering of the program.74 The court held that ‘reverse engineering object code to discern the unprotected ideas in a computer program [was] a fair use’75 but that since Atari obtained the copyrighted work in an unauthorised way the fair use defence could not be established. Nevertheless, the court held that should Atari’s reverse engineering be conducted with an authorised copy of the 10NES program, the reverse engineering could be considered a fair use.76 The court’s ruling implies that where the copyrighted work is obtained in legitimate ways, the reverse engineering of the copyrighted work to achieve the compatibility can be deemed to be a fair use. This line of reasoning is confirmed in Sega v. Accolade. In Sega v. Accolade,77 Accolade reverse engineered Sega’s video game cartridges in order to make its video games compatible with Sega’s game console, Genesis.78 Unlike Atari in Atari v. Nintendo, Accolade obtained Sega’s copyrighted works in a proper way by purchasing them, reverse engineered them and produced video games compatible with Sega’s Genesis. Sega responded to Accolade’s sale of Genesis-compatible games by filing a suit against Accolade, alleging, among other things, that Accolade infringed the copyrights covering its video game programs.79 In response, Accolade 72 Atari Game Corp. and Tengen, Inc., v. Nintendo of America Inc. and Nintendo Co., Ltd., 975 F.2d 832 (1992). 73 Ibid., 835–7. 74 Ibid., 842. 75 Ibid., 843. 76 Ibid., 843–4. 77 Sega Enterprise Ltd. v. Accolade Inc., 977 F.2d 1510 (1992). 78 Ibid., 1514–17. 79 Ibid., 1516.

184


argued, inter alia, that ‘disassembly of object code in order to gain an understanding of the ideas and functional concepts embodied in the code [was] a fair use’.80 In determining whether Accolade’s reverse engineering of Sega’s game programs is condoned under the fair use doctrine, the Ninth Circuit Court analysed the facts with respect to the four statutory factors of a fair use defence prescribed in section 107 of the copyright law.81 The four factors are as follows: (1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work.

As for the first element, the court reasoned that even though Accolade’s copyright infringement was for commercial use, the direct purpose of the infringement was to discover the functional elements for compatibility and that the public could obtain benefits from this use, holding that the first statutory element was in favour of Accolade.82 As to the second factor, the court held that the reverse engineering of the object code of Sega’s game cartridges was needed to obtain the program’s functional information for compatibility, which copyright law did not protect, and that in order to protect the idea the creator of a work had to obtain patent rights over the idea rather than copyrights.83 As far as the third factor is concerned, the court held that, even though Accolade copied entire programs of Sega’s, the fact that Accolade did so did not bar the application of the fair use doctrine.84 Concerning the fourth statutory element, the court held that even though Accolade’s product surely affected the market for Sega’s video games, the effect was not significant since a consumer in the market might buy both products.85 Thus the Ninth Circuit endorsed Accolade’s arguments, concluding that ‘where disassembly is the only way to gain access to the ideas and functional elements embodied in a copyrighted computer program and where there is a legitimate reason for seeking such access, disassembly is a fair use of the copyrighted work, as a matter of law’.86 80 81 82 83 84 85 86

Ibid., 1517–18. For § 107 of the copyright law, see supra note 65. Sega Enterprise Ltd. v. Accolade Inc., 1522–3. Ibid., 1524–6. Ibid., 1526–7. Ibid., 1523–4. Ibid., 1527–8.


185

The reasoning of Atari v. Nintendo and Sega v. Accolade was reaffirmed in Sony v. Connectix, where the Ninth Circuit held that ‘Connectix’s intermediate copying of the Sony BIOS during the course of its reverse engineering of that product was a fair use under 17 USC §107, as a matter of law.’87 Consequently, the fair use defence can be used by a party hoping to develop a product which is compatible with a dominant product. The party may reverse engineer the dominant product to acquire the functional elements necessary for compatibility and design its product to be compatible with the dominant product. However, it has to be noted that the final product of the party should not infringe the copyright of the dominant product since the fair use doctrine only allows the intermediate copy.88 2.3.2 The arguments for fair use doctrine in patent law Professor O’Rourke notes the fair use doctrine of copyright law and argues that the doctrine has to be introduced into patent law.89 She compares the fair use doctrine of copyright law with the patent law doctrines used to limit patent rights and finds that, unlike copyright law, patent law does not condone infringement conducted for a commercial purpose.90 Furthermore, she notes that there may be bargaining breakdown between the patentee of the initial patents and the significant improvers even though the blocking patent doctrine increases the bargaining power of the significant improvers.91 She argues that there may be some cases where commercial infringement needs to be allowed and that the fair use defence has to be used in such cases.92 She contends that just as the commercial infringement may be excused in copyright law, which is clear from the video game console cases, there may be some cases where the infringement conducted for commercial purposes needs to be condoned in patent law. In this respect, the main difference between the existing patent law doctrines, such as the reverse doctrine of equivalents, the blocking patent doctrine and the experimental exemption, and the fair use doctrine lies in whether the doctrine excuses commercial infringement for significant improvements. As is described in Chapter 5, the reverse doctrine of equivalents excuses the commercial infringement but only applies to radical improvements.93 In addition, the blocking patent doctrine and the experimental exemption cannot

87 Sony Computer Entertainment, Inc. v. Connectix Corporation, 203 F3.d 596, 608 (2000). 88 Sega Enterprise Ltd. v. Accolade Inc., 1528. 89 See O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’. 90 Ibid.: 1187–1205. 91 Ibid.: 1204. 92 Ibid.: 1205. 93 See Chapter 5, Section 2.2–2.3.

186


be used as a defence for commercial infringement. Thus the fair use doctrine is different from the reverse doctrine of equivalents in that it may be applied to a significant improvement, whereas the fair use doctrine is distinct from the blocking patent doctrine and the experimental exemption in that it condones infringement for commercial purposes. O’Rourke proposes five factors which have to be considered in finding a fair use in patent law as follows: (1) the nature of the advance represented by the infringement; (2) the purpose of the infringing use; (3) the nature and strength of the market failure that prevents a license from being concluded; (4) the impact of the use on the patentee’s incentives and overall social welfare; (5) the nature of the patented work.94

As for the first factor, O’Rourke argues that where the improvement made by the infringement is considerable and thus largely serves public welfare, the infringement has to be granted as a fair use, and that even though the advance is not significant, the fair use doctrine has to be applied in those cases where there is no way to make a product compatible with dominant products other than infringing the patent rights of the dominant products.95 As to the second factor, she contends that where the infringement is conducted for noncommercial use, the infringing use needs to be allowed as a fair use and that where the alleged infringer infringes patent rights in an intermediate step to produce a non-infringing final product, the infringement has to be considered a fair use, as in copyright law.96 As far as the third element is concerned, she argues that the court has to find the relevant market failure and to consider the characteristics of the market failure in deciding a fair use.97 She contends that the fourth factor is the most important factor of the five fair use factors and that the court should seek to strike the right balance between the social benefits obtained from the infringement and the harmful impact on the incentive to invent caused by allowing the infringement.98 As for the fifth factor, she argues that, in order to prevent ‘disproportionate leverage’99, it is necessary to

94 95 96 97 98 99

O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1205. Ibid.: 1206. Ibid. Ibid.: 1206–7. Ibid.: 1207–8. The disproportionate leverage refers to the circumstances where ‘small property rights are being employed to leverage very large market’; for more detail, see Merges, ‘Who Owns the Charles River Bridge? Intellectual Property and Competition in the Software Industry’, 4.


187

apply the fair use doctrine more widely in cases where the advance made by the patented invention is comparatively smaller than in cases where the advance achieved by the patented invention is significant.100 In order to demonstrate how the fair use doctrine of patent law can be put into practice, O’Rourke applies the suggested patent fair use to the compatibility issues of software application programming interfaces (APIs).101 She notes that the API market is a network market, which has the characteristic of network externalities, and finds that the value of the API is determined not only by its inherent value but also by the number of consumers, as other network products are.102 Due to this feature of the network market, the owner of standard APIs comes to dominate the innovation market as well as the product market. O’Rourke reasons that while it is important to provide the incentive to invent APIs by ensuring exclusive rights over them, it is also necessary to allow open access to the standards to stimulate subsequent innovations. Based on this reasoning she argues that the protection regime of APIs has to be dynamic, changing from the proprietary regime to a pro-compatibility regime, which is called a ‘dynamic compatibility regime’.103 It is argued that the existing patent law doctrines, such as the blocking patents doctrine, the reverse doctrine of equivalents, the experimental use exemption and the patent misuse doctrine, are not suitable to implement the ‘dynamic compatibility regime’ but that the fair use doctrine of patent law can achieve that.104 O’Rourke applies the patent fair use to APIs as follows. As far as the first factor of the fair use doctrine, the nature of the advance represented by the infringement, is concerned, O’Rourke argues that this factor is likely to allow the indirect infringement committed in developing a compatible and vertically related product but unlikely to do so in relation to the directly infringing product which is a horizontally related product.105 In the context of APIs, the infringement committed in developing application software is more likely to be considered a fair use than the infringement for competing operating systems. As for the second factor of the fair use doctrine, which is the purpose of the infringing use, it is argued that, just as in the case of the first factor, the court should treat the infringement resulting from a competing product and that from an application in different ways.106 It is clear that a competing operating

100 101 102 103 104 105 106

O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1208–9. Ibid.: 1211–35. Ibid.: 1227. Ibid.: 1218. Ibid.: 1227–30. Ibid.: 1230. Ibid.: 1230–31.

188


system is likely to harm the patentee, whereas vertically related products are not necessarily so. Thus, O’Rourke contends that the second factor may be neither in favour of nor against a vertically related product but would be against a competing product.107 As to the nature and strength of the market failure, which is the third element, she argues that the court has to note the possible anti-dissemination motives of patent owners and to impose a compulsory licence where it is necessary to guard or enhance social welfare.108 In particular, it is argued that where the patentee is competing with firms in the downstream market, the court needs to suspect the patentee’s licensing behaviour of refusing to license to the firms. The argument on the fourth factor, the impact of the use on the patentee’s incentives and overall social welfare, is similar to the other arguments, being that the harm caused by excusing infringement is relatively small in the case of a vertically related product but that the harm resulting from allowing the infringement of a directly competing product may be considerable.109 Thus, it is argued that the fair use has to be used to allow infringement committed in developing vertically compatible products. Regarding horizontally related products, it is argued that where network externalities are strong, producing excess inertia,110 and where the patent owner has already been rewarded substantially from the exploitation of the patented innovation, even infringement committed in developing competing products may need to be considered a fair use.111 As far as the fifth factor, the nature of the patented work, is concerned, it is argued that the court has to consider whether the patented API has significant intrinsic value or whether its value is largely derived from external factors such as consumer expectations and first mover advantages, and that where the API has relatively small value compared with the value created by external factors, the fifth factor of the fair use doctrine has to be in favour of both vertical and horizontal products.112 O’Rourke also contends that, unlike the copyright fair use, the patent fair use needs to be granted for a licensing fee in some cases.113 It is argued that where the patentee’s refusal to license is due to an anti-patent motive, it is

107 108 109 110

Ibid.: 1231. Ibid.: 1231–3. Ibid.: 1233–4. ‘Excess inertia’ refers to the inefficient attachment to the old standard. For more detail, see Farrell and Saloner, ‘Standardization, Compatibility and Innovation’. 111 O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1233–4. 112 Ibid.: 1234. 113 Ibid.: 1234–5.


189

proper to grant a patent fair use free of charge but that where the fair use concerns externalities, compensation to the patentee is needed to strike the balance between competition and the incentive to invent.114 In setting the proper licensing fees of the patent fair use, it is argued that the fees need to be set at an amount less than the fair market value in order to encourage both parties of a bargain to make a voluntary deal. The argument is that immunity from the infringement charge under a fair use and the royalty rates set at an amount less than the fair market value would be a real threat to the patent owner, encouraging the patent owner to voluntarily bargain with the potential infringer. From the perspective of the potential infringer, it would be better to deal with the patent owner since reverse engineering is costly and the potential infringer can be offered know-how and other technical information which are not available from the reverse engineering. Thus it is argued that the patent fair use doctrine can facilitate efficient private bargaining and that it can increase social welfare by excusing a socially beneficial infringement even in those cases where the bargain fails. 2.3.3 Evaluation and implication Professor O’Rourke’s arguments are quite logical and persuasive. She seeks to apply the legal principles of another law, modifying them to be suitable to patent law. She argues for the introduction of the copyright law principle, the fair use doctrine, to patent law by redesigning it. She notes that the fair use doctrine is used to correct market failure in copyright law and then seeks to apply the reasoning of the doctrine to the realm of patent law. It is true that well-established legal practices in other intellectual property law may be useful in reforming patent law since they may be used to reduce the costs caused by trial and error which inevitably follow in introducing new concepts into patent law. The fair use defence is well set up in copyright law and thus the introduction of the principle into patent law may not be prohibitively difficult, since the practices in copyright law provide good precedents. O’Rourke also utilises antitrust law concepts such as the vertical relationship and the horizontal relationship and argues that the products in a vertical relationship have to be treated more favourably than the products in a horizontal relationship under the patent fair use doctrine. Like the fair use of copyright law, such antitrust concepts are well set up and thus they can be used in applying the patent fair use defence. In addition, she argues that since the harm caused by excusing infringement committed in developing complementary products is considerably different from that resulting from infringement by competing products, it is reasonable to treat the two sets of infringement in

114

Ibid.

190


different ways. As such, her attempt to utilise the principles of other relevant law such as copyright law and antitrust law is quite remarkable. The other contribution of her arguments is the suggestion of the five factors of the patent fair use which the court has to consider in deciding whether to grant the fair use defence. From the four factors of the copyright fair use, she drops one factor, ‘the amount and substantiality of the portion used in relation to the copyrighted work as a whole’, and adopts the other three factors in modified forms. This approach is proper since, as is argued by O’Rourke, the dropped factor can be considered in examining the other factors of the patent fair use.115 She also supplements a factor, ‘the nature and strength of the market failure that prevents a license from being concluded’, in designing the patent fair use. This factor means that the limitation on patent rights by applying the fair use defence has to be allowed only in cases of market failure, as is the case in copyright law. O’Rourke’s arguments can quite nicely be applied to the patent issues in compatibility standards. Indeed, her suggestion can be considered the basic framework of the patent reform to reduce the costs of patents in relation to compatibility standards. Her studies show the whole set of relevant factors which the court or a patent reformer has to consider in limiting patent rights. Thus, her arguments have to be adopted in fine-tuning patent systems in relation to patent issues in standards. How then does the proposed patent fair use affect the costs and benefits of patent systems in relation to standards? Firstly, consider the costs. In order to reduce the costs of patent systems with respect to standards, it is necessary to facilitate efficient bargaining and restrict the market power of standard owners by ensuring access to the standards at some stages. Thus, it is necessary to examine whether the patent fair use promotes efficient bargaining and restricts the market power of the standard owner. As for bargaining, the fair use doctrine is likely to facilitate efficient bargaining between patent owners and users since, as is argued by O’Rourke, the patent fair use would increase the bargaining parties’ need for voluntary bargaining.116 As far as the effect of the patent fair use on the market power of standard owners is concerned, it is likely to effectively restrict the market power of standard owners by ensuring open access to standard technology, should the doctrine be applied properly. In particular, O’Rourke’s suggestion for the dynamic compatibility regime, which refers to the protection regime of APIs, changed from the proprietary regime to a pro-compatibility regime, may nicely be applied to compatibility issues in standards, reducing the costs of patent systems in relation to stan-

115 116

Ibid.: 1205. See Section 2.3.2 of this chapter or ibid.: 1234–5.


191

dards.117 Thus, if used properly, the patent fair use is likely to decrease the costs of patent systems in relation to standards. Secondly, regarding the effect of the proposed patent fair use on the benefits of patent systems, it is not likely that the doctrine would seriously harm the incentive structure of patent systems, should it be properly applied. O’Rourke suggests the five factors which the court has to consider in deciding whether or not to apply the fair use doctrine and should the court appropriately apply the five factors to the actual cases, the benefits of patent systems would not be reduced. Rather, the fair use doctrine may promote the diffusion of standard technology and thus increase the benefits of patent systems. Accordingly, the patent fair use is a good candidate for patent reform in relation to standards since it reduces the costs of patent systems and does not decrease the benefits. However, it has to be noted that the precondition of this reasoning is that the patent fair use has to be used in proper ways, which is the weakness of O’Rourke’s arguments. She only suggests the principles and thus there arise uncertainties in applying the patent fair use to actual cases and the cost resulting from the application of the patent fair use may be significant due to these uncertainties. For instance, she argues that where the patentee has already recouped a ‘substantial reward’ from the patented innovation, the fair use doctrine has to be applied more generously.118 This argument seems to be in concurrence with the opinion of Judge Boudin in Lotus v. Borland.119 In Lotus v. Borland, Lotus had copyright over the spreadsheet program Lotus 1-2-3 and the competing firm, Borland, designed its spreadsheet program Quattro Pro to have the same menu trees as those of Lotus 1-2-3.120 In designing the menu structure, Borland did not copy the source code of Lotus 1-2-3 but only copied ‘the words and structure of Lotus’s menu command hierarchy’.121 Lotus filed an infringement suit against Borland, alleging that Borland infringed the copyright of Lotus in implementing Quattro Pro’s menu structure, and, in response, Borland sought a declaration of non-infringement. The majority opinion of the First Circuit was that the menu structure of Lotus1-2-3 was uncopyrightable since it was a method of operation and thus Borland did not infringe Lotus’s copyright.122 However, Judge Boudin opined

117 118 119

Ibid., 1218. Ibid., 1233–4. Lotus Development Corporation v. Borland International Inc., 49 F3.d 807

(1995). 120 121 122

Ibid. Ibid., 809–10. Ibid., 819.

192


differently and argued for ‘the privileged use’.123 He argued that it was accepted that Lotus invested in developing Lotus 1-2-3 menu trees but that users also made an investment ‘in learning the menu and building their own mini-programs … in reliance upon the menu’, being locked into Lotus.124 He also argued that Lotus had been rewarded substantially as a head starter and that Borland had no intention to appropriate Lotus’s investment but merely sought to make the users have the option to use their own investment in Lotus 1-2-3.125 In line with Judge Boudin’s opinion in Lotus v. Borland, O’Rourke contends that the court has to consider in applying the fair use doctrine whether the patentee has recouped a ‘substantial reward’ from the patented innovation. However, what is the meaning of ‘substantial reward’? How to decide whether the patentee is rewarded enough? There may be some cases where this decision is relatively easy, but there may also be some cases where it is difficult. O’Rourke fails to provide clear criteria which the court can use in examining the patent fair use defence. Another example of uncertainty arising from the patent fair use is found in her argument about royalties. She argues that in applying the patent fair use, compensation for a patentee by way of royalties can be necessary and that the royalties have to be set at an amount less than the fair market value to facilitate the private bargaining.126 This argument is feasible in some sense but the problem is to decide how much ‘less’ the royalty rates should be. In the case of a standard, it may be relatively easy to find the fair market royalty rates since other similar licensing agreements are likely to exist. However, it may be difficult for the court to find the proper level of the royalty rate that is less than the fair market value since, whereas there is one fair market value, there are a number of fees which are less than the fair market value. Because of these uncertainties and the fact specific characteristics of the fair use doctrine, the application of the fair use doctrine to patent cases is likely to be quite costly, as O’Rourke admits.127 It is not easy to predict the outcome of the application of the fair use doctrine to real cases since the court has too much discretion. Admittedly, uncertainty may facilitate private bargaining, as some commentators argue,128 but the problem is the level of

123 124 125 126 127 128

Ibid., 819–22. Ibid., 819. Ibid., 821. O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1234–5. Ibid., 1236. For the benefits of uncertainty, see Ayres and Klemperer, ‘Limiting Patentees’ Market Power without Reducing Innovation Incentives: The Perverse Benefits of Uncertainty and Non-Injunctive Remedies’; Ayres and Talley, ‘Solomonic Bargaining: Dividing a Legal Entitlement to Facilitate Coasean Trade’.


193

uncertainty. Contrary to O’Rourke’s arguments, it is not clear whether the patent fair use can achieve the optimal level of uncertainty to facilitate socially beneficial infringement. Moreover, since the fair use doctrine has its origin in copyright law, it is likely that the doctrine only condones the intermediate copy. O’Rourke argues that ‘patent courts should make the same distinction as their copyright counterparts. Copyright courts are more hospitable to infringement that is conducted as an intermediate step in producing a non-infringing end product than towards a directly infringing product.’ 129 Even though she does not exclude the possibility of the application of the doctrine to the case where the end product infringes the patented product, it is likely that the fair use doctrine plays little role in that sort of case. As is argued in Chapter 5,130 however, since what matters in relation to standards is the availability of the very same technology protected by the patent essential to standards, condoning only the intermediate infringement may not be sufficient. To sum up, it is notable that Professor O’Rourke proposes the broad principles of the patent fair use defence but, in putting those principles into practice, some more details are required. Despite the fact that the patent fair use is a new concept in patent law, too many decisions are left to the court having no experience of applying the new doctrine, creating uncertainties. In addition, it is necessary to make the patented standards available to the end product since they are essential for compatibility. Consequently, should the patent fair use be supplemented with some detailed implementation schemes and be modified in such a way as to condone infringement in relation to the end product, it may provide a way to resolve the patent issues in standards. 2.4

Competitive Platform Model

Professor Philip Weiser proposes ‘the competitive platform model’ in relation to the information platform.131 The definition of the information platform is not clearly presented but it is clear from his explanation that the information platform contains the platform standards, which components and complementary products have to be compatible with, and includes operating systems, game consoles, web browers and telecommunication standards like the GSM and CDMA technologies.132 It seems that the information platform is similar to ‘systems technology’ as defined by Wagner earlier in this section.133 Weiser 129 130 131 132 133

O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1206. See Chapter 5, Section 2.3. See Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’. Ibid.: 545–6. For systems technology, see Section 2.1.1 of this chapter.

194


suggests the competitive platform model to balance the conflicting interests of the incentive to innovate and open access to standard technologies, and thus his suggestion is closely related to the topic of this book. 2.4.1 Arguments for the competitive platform model Weiser explains two typical models of intellectual property policies in relation to the information platform: the common model and the proprietary control model.134 As the name implies, the advocates for the common model argue that free access to information is essential for technical innovation in information technology and thus the proprietary control of information has to be avoided. This argument is supported by the evidence that a number of innovations have taken place without intellectual property protection and, especially, the explosive growth of the internet is largely due to the absence of intellectual property protection.135 This model uses the theory of network industries and argues that a dominant standard appears inevitably in information technology markets since network effects prevail in those markets. It is argued that the keyboard layout QWERTY and Microsoft PC operating systems dominate the relevant markets because the markets tend to tip toward a specific standard. As the historical evidence shows, these dominant standards are not necessarily superior to competing technologies.136 That is, it is the features of the network markets such as network effects and switching costs rather than the superiority of the products that make those products be adopted as standards. It is argued that because of these characteristics of the network market, the proprietary control of such standards would cause a long-lasting monopoly, preventing other beneficial technologies from entering the market, and that access to standard technologies is necessary for the furtherance of innovations.137 Thus the common model contends that open access to standards is essential to spur innovations. On the other hand, the proprietary control model’s arguments are totally different from those of the common model. These arguments are along the same lines as those of Schumpeter and Edmund Kitch. Joseph Schumpeter argues that any market power is temporary since a new product will replace the incumbent product through ‘creative destruction’ and that market power is essential to promote technological innovations.138 This Schumpeterian

134 135 136

Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 568–83. Ibid.: 570–71. David, ‘Clio and the Economics of QWERTY’; Sheremata, ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’. 137 Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 572. 138 Schumpeter, Capitalism, Socialism, and Democracy 81–90.


195

perspective is applied to intellectual property by Edmund Kitch, who argues that exclusive control over inventions is essential for the efficient use of the inventions.139 That is, according to this perspective, intellectual property protection is necessary not only to provide the incentive to invent but also to ensure the efficient exploitation of the invented technologies. In the same way, the proponents of this model argue that the proprietary control of platform standards is necessary for the supply and efficient use of the standards. Weiser criticises both the common model and the proprietary control model and seeks to adopt the benefits of each model. He condemns the common model because of the facts that non-proprietary standards like open source standards are exceptional, that refusing to adopt patented technologies as standards in formal standard bodies like W3C (World Wide Web Consortium) has many limitations in creating information platforms, and that network markets are not necessarily tipped toward a single standard since other factors such as the optimal network size, competing networks and diverse consumer demand may create multiple networks.140 He is also critical of the proprietary model for several reasons. Firstly, it is argued that according to empirical evidences, large firms are not necessarily more innovative than small firms, and this is especially the case in information industries where intellectual capital is more important in R&D than physical capital.141 Secondly, even though a new standard would replace the incumbent standard some time in the future, the monopoly power of an incumbent standard may be detrimental to innovations, preventing beneficial technologies from emerging in the markets.142 That is, not only the fact that ‘creative destruction’ will replace the incumbent technologies is important, but also the timing of the creative destruction is important.143 The proprietary control of standards would delay the market mechanisms, impeding the creative destruction. Lastly, the dominant firms may be reluctant to introduce new and innovative technologies but seek to maintain the status quo since the current market provides large profits to the dominant firms.144 The incumbent firms may feel less pressure for innovation than other firms. While criticising the weaknesses of each model, Weiser recognises the advantages of each model and seeks to design a model by combining the advantages. The common model ensures open access to the information platform,

139 140 141 142 143

For more detail, see Kitch, ‘The Nature and Future of the Patent System’. Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 573–6. Ibid.: 579–81. Ibid.: 581. Sheremata, ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’, 962–3. 144 Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 581–2.

196


encouraging multiple entries and stimulating innovation in diverse ways. On the other hand, the proprietary model provides the incentives to supply the information platform, facilitating the appearance of the platform standards. Weiser seeks to combine these advantages of the two models in his competitive platform model. The special feature of the competitive platform model, as proposed by Weiser, is that the model suggests different treatments of the information platform, depending on whether there are rival platforms or not. It is argued that where there are competing platforms, the proprietary control model has to prevail, and that where a platform standard becomes dominant, the common model has to be applied.145 Weiser’s argument is that while two or more standards are competing to dominate a market, the competing firms are likely to be innovative and to ensure open access to the platforms since, in order to win the standards war, a firm has to make its platform more attractive than its competitors’ and to build a large installed base of users.146 Thus, the market condition itself promotes innovation and therefore there is no need to change the existing intellectual property policies at this stage. Conversely, where access to platform standards is allowed by means of reverse engineering at this stage, the incentive to invent a competing platform will be reduced and the benefits which would otherwise be obtained from the competition will not be achieved. Thus, it is argued that competition between rival platform technologies has to be encouraged by adopting the proprietary control model at the stage of ‘competition for’ a standard and that reverse engineering should not be allowed at this stage. On the other hand, once a dominant standard appears or is likely to appear, it is argued that the common model has to be used and that reverse engineering has to be allowed.147 The argument is that once the competition becomes ‘competition within’ a standard, the owner of the dominant platform standard has made sufficient or more than sufficient profits from the dominant standard and thus intellectual property protection is not necessary any more.148 It is argued that the intellectual property protection of the platform standards at this stage would not create more innovations and that allowing open access to the standards would not significantly reduce the incentives to innovate. Rather, open access to the standards would facilitate the appearance of new platform standards and prevent anticompetitive behaviour by the standard owner. That is, the argument is that the benefits gained from the application of the common model outweigh those obtained from the application of the proprietary control 145 146 147 148

Ibid.: 583–94. Ibid.: 585–91. Ibid.: 591–4. Ibid.: 591–2.


197

model. As such, it is contended that the competition within a standard has to be encouraged by adopting the common model in those cases where a standard begins to dominate the relevant markets, and that reverse engineering has to be allowed as a way to allow open access. It has to be noted that Weiser’s analysis mainly focuses on the issues of horizontal access to the platform standards since, he argues, vertical access is generally allowed by the fair use doctrine of the current law in those cases where a standard becomes dominant.149 It has also to be noted that his argument deals exclusively with the issue of reverse engineering of the platform standards as a way to allow open access to them and that other issues like compulsory licences are not discussed. He argues that the legality of reverse engineering of a dominant standard would facilitate voluntary bargaining between the standard owner and the user since both parties of the bargaining have an incentive to make a deal. The standard owner has the incentive to trade with the user since where the bargain fails the user may reverse engineer the standards and obtain the necessary information. Thus, the mere possibility of reverse engineering increases the bargaining power of the potential users and, on top of that, the cost of the reverse engineering regulates the maximum licensing fees.150 The standard user also has reason to seek voluntary bargaining. The facts that the reverse engineering is costly, that the technical support will not be provided by the standard owner, and that other legal issues like trademark issues may arise encourage the user to strike a voluntary deal.151 As for the possible problem of how to decide whether a platform technology becomes dominant, Weiser argues that the antitrust analysis as to market power is well established and that the antitrust analysis would make it possible with reasonable costs to find the market dominance.152 2.4.2 Evaluation and implication Weiser’s analysis is quite remarkable in several respects. Most notably, he seeks to accommodate the competition law policy in the intellectual property policy by introducing the antitrust law principles into intellectual property law. His model utilises in analysing the market conditions the concepts of the vertical relationship, the horizontal relationship and market dominance, which are used in antitrust analysis, and then proposes the application of the intellectual property law principles, such as the fair use doctrine and the misuse doctrine, to the result obtained from the analysis.153 In this way, he successfully seeks 149 150 151 152 153

Ibid.: 600–608. Ibid.: 548–9. Ibid.: 609. Ibid.: 610–12. Ibid.: 552–7, 600–608.

198


to reconcile the seemingly conflicting legal regimes, the intellectual property law and the antitrust law. Secondly, he differentiates ‘the competition for standards’ from ‘the competition within standards’ and argues that horizontal access to a platform technology has to be granted at the stage of competition within standards.154 Considering that a dominant standard is likely to provide considerable profits to the owner by way of the licensing fees or the sale of the standard, his argument that horizontal access to the dominant standards has to be granted is persuasive. In this manner, he seeks to resolve the conflict between the incentives to innovate and the need for open access to a dominant standard. Thirdly, his argument that allowing the reverse engineering of the dominant platform standards would encourage voluntary bargaining between the owner of the platform standards and the user is feasible.155 As is argued in Chapter 6 of this book, the asymmetric bargaining power between the owner of standards and the user may result in inefficient bargaining and thus it is necessary to adjust the bargaining power between them.156 The legality of the reverse engineering is one way to fine-tune the bargaining power, increasing the bargaining power of the user of standards. Thus, as Weiser argues, allowing the reverse engineering of the dominant standard is likely to increase the voluntary bargaining, enhancing the probability of achieving efficient bargaining between the relevant parties. Lastly, he suggests the use of the antitrust analysis in judging whether a standard becomes dominant or not,157 which is also notable. Determining the relevant markets and judging the dominance of a firm in the markets are well established practices in the antitrust analysis, even though they are not easy tasks. Since there are a number of precedents for them, applying the competitive platform model to the actual cases is not impractically onerous. Weiser’s arguments are similar to O’Rourke’s arguments in many respects. Both Weiser and O’Rourke use the concepts of antitrust law such as the vertical relationship, the horizontal relationship and market dominance. Furthermore, O’Rourke’s dynamic compatibility regime is similar to Weiser’s distinction between the stage of competition for standards and that of competition within standards. The main difference between O’Rourke’s model and Weiser’s is that O’Rourke’s main concern is patent law whereas Weiser’s main focus is copyright law. Even though Weiser apparently presents arguments about intellectual property policies in general, his main concern seems to remain in the realm of copyright law. This is verified by his statement that 154 155 156 157

Ibid.: 583–94. Ibid.: 548–9, 609. For more detail, see Chapter 6, Section 3.2. Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 610–12.


199

‘vertical access - which involves products that are complementary to an information platform – should not be restricted. … intellectual property law … generally allows access for the creation of complementary goods.’158 As the legal principle which is used to allow vertical access, he refers to the fair use principle. However, the fair use doctrine is a copyright law principle and there is no fair use doctrine in patent law. In patent law, reproduction for commercial purposes is not allowed without permission from the patentee, even though it is done to achieve compatibility or vertical access. Weiser merely hints that patent misuse may be used to make patent law harmonious with copyright law as to reverse engineering, but he does not describe it in more detail.159 In some sense, the fact that he focuses on copyright law is understandable, since his main concern is the information platform mainly comprising software innovations, which have generally been protected by copyright law. Possibly due to his focus on software and copyright law, Weiser’s argument is limited within the boundary of the legality of reverse engineering, which is the weakness of his model. Allowing the reverse engineering of the dominant platform standards may be insufficient to guarantee open access to the standards. He argues that ‘in many reverse engineering cases, the rival product itself … does not infringe on the original product’s intellectual property’.160 That is, his argument is that in many cases the competing product created by the reverse engineering of a first product does not infringe the intellectual property of the first product. This argument may apply to some software cases but not to the non-software cases nor to some software cases. To illustrate, consider the telecommunication standards which are covered by patents. Without adopting the standard technologies, there is no way to make a product compatible with the telecommunication network and thus a firm has to incorporate the standard technologies in order to communicate with the network. Also consider APIs, which are a software product. As is argued by O’Rourke, incorporating the same APIs is essential for compatibility.161 However much effort a firm makes to invent around the standard by reverse engineering or any other means, it is unlikely that a firm will invent a product which does not infringe the patents and is compatible with the network as well. In many non-software product cases and some software product cases, compatibility is not achieved without the incorporation of the standards, which implies that open access to the standards is not obtained by reverse engineering. The only way to achieve compatibility is to incorporate 158 159 160 161

Ibid.: 564–5. Ibid.: 559. Ibid.: 567. O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1228.

200


the compatibility standards and thus licences of the patented standards are essential for compatibility. Accordingly, Weiser has to consider the way to open access to the protected standards by means of licences and, especially, compulsory licences have to be examined. However, he does not include compulsory licences in his analysis. Thus, in applying his analysis to the compatibility standards, it is essential to consider the issue of compulsory licences. Since Weiser’s competitive platform model is similar to O’Rourke’s dynamic compatibility regime, it follows that the competitive platform model reduces the costs of patent systems in relation to standards, facilitating efficient bargaining and ensuring open access to standard technology, and does not decrease the benefits of patent systems, providing sufficient incentives to invent and innovate, just as O’Rourke’s dynamic compatibility regime does. To sum up, Weiser’s model seems to be a detailed form of O’Rourke’s patent fair use model and is logical and correct in many respects. Especially, the distinction between competition for standards and competition within standards is notable. In addition, he distinguishes between the horizontal relationship and the vertical relationship and treats them in different ways, which is quite correct. However, his analysis does not include the compulsory licence for compatibility, which is the weakness of the proposed model.

3.

SUGGESTED SOLUTION

The model of liability rules, the model of compulsory licences with capital payments, the patent fair use model and the competitive platform model have been examined in this chapter. Each model has advantages and disadvantages. Based on the insight gained from the analysis of each model, the model that this book suggests is described below. As is clear in the following, the basic structure of this model is similar to O’Rourke’s and Weiser’s. Namely, the competition for standards and the competition within standards are differentiated and the vertical relationship and the horizontal relationship are treated in different ways. However, the focus of the analysis is different from that of O’Rourke’s or Weiser’s in that what is analysed here is the incentive structure. That is, the incentive to bargain and to abuse market power is analysed since, as is argued in Chapter 6, in reducing the costs of patent systems in relation to standards, it is crucial to facilitate efficient bargaining and to restrict the market power of standard owners by ensuring other firms’ access to the standards at some stages. By analysing the incentive structure of bargaining and the abuse of market power, it may be possible to find the proper patent reform. In addition, Wagner’s liability rules model is used to reduce the costs arising from the exclusivity of patent rights and Kingston’s model of compulsory


201

licences with capital payments is adopted in limited circumstances to preserve the incentive structure of patent systems. The suggested model is examined with respect to the costs and benefits of patent systems in relation to standards and is applied to the patent hold-up problem in formal standardisation. It is argued that the proposed model would mitigate the patent hold-up problem in a standard setting organisation and thus facilitate the appearance of formal standards. Lastly, how to implement the proposed model is described. 3.1

Analysis of the Incentive Structure

As noted, in reducing the costs of patent systems in relation to standards, it is crucial to facilitate efficient bargaining and to restrict the market power of standard owners by ensuring other firms’ access to the standards at some stages. Thus, in order to design a proper patent system with respect to standards, it is essential to analyse the incentive structure of bargaining and the abuse of market power. Table 7.1 shows the incentive structure in relation to standards. It has to be noted that the incentive varies depending on various factors. Whether a technology is in the stage of competition for standards or competition within standards and whether the relationship between a standard owner and a user is horizontal or vertical are crucial factors in determining the incentive to bargain and to abuse market power. Thus, it is necessary to examine the incentive according to the factors. Firstly, consider the incentive to bargain at the stage of competition for standards. The stage of competition for standards refers to the circumstances where two or more technologies compete to dominate the relevant market. Where a technology has not yet become a standard, the owner of the technology has to obtain an installed base to make the technology a standard. In Table 7.1

Incentive structure Incentive to bargain

Incentive to abuse market power

Intervention

Competition for standards

Vertical Horizontal

Large Uncertain

Small Small

Unnecessary Unnecessary

Competition within standards

Vertical Horizontal

Large Small

Large Large

Necessary Necessary

202


network markets, it is essential for a technology to reach the critical mass to initiate the positive feedback which makes the technology a standard.162 Thus, the owner of the technology is likely to collaborate with vertically related firms to expand the installed base of the technology. Accordingly, the owner of a technology has large incentives to bargain with vertically related firms like the firms producing complementary products. As for the horizontally related firms, however, it is not clear whether the owner of a technology will be willing to bargain with the competing firms or not. The owner of a technology may bargain with the competitors in order to expand his network. By making his network compatible with those of competitors, the owner of a technology may obtain a large installed base and thus increase the probability of being a winner in the standard war. However, the owner of a technology has to share the pie with competitors even though he wins the war. Where the owner of a technology does not wish to share the profits with competitors, he may seek to exclude competitors by refusing to deal with them at the stage of competition for standards. It is not clear which one would prevail. All that can be said is that an owner of a technology may or may not bargain with competitors and whether to bargain or not is not important since no one has market power to abuse. Secondly, consider the incentive to abuse market power at the stage of competition for standards. In this stage, there is no dominant technology and the market power of a firm is not very strong and thus the incentive to abuse market power is small. A firm in this stage has to be favourable to other firms, especially the firms producing complementary products, and to customers since competitors always seek to expand their own installed base by collaborating with complementary firms and by attracting more customers. Furthermore, the firms in this stage have to be innovative to survive in the market. Accordingly, government interventions and patent reform are not needed at this stage. Thirdly, consider the incentive to bargain at the stage of competition within standards. The stage of competition within standards refers to the situations where a technology becomes a standard and dominates the market. The firm with standards has a large incentive to bargain with vertically related firms since it is essential to keep or expand the installed base to sustain the dominance. However, the firm controlling standards will not be willing to bargain with competitors. Indeed, the firm is likely to seek to expel the existing competitors from the market and to prevent potential competitors from entering the market. One of the best ways to achieve this is to refuse to bargain with

162 For the dynamics of standardisation, see Grindley, Standards, Strategy, and Policy: Cases and Stories 27–8.


203

competitors since competitors cannot enter the relevant market without making their products compatible with standards. Thus, it is unlikely that a standard owner will be willing to make a voluntary deal with competitors and thus a standard owner has a large incentive to refuse to bargain with horizontally related firms. Fourthly, as far as the incentive to abuse market power is concerned, the owner of a standard has a large incentive to abuse market power to strengthen his dominance. Market power can be exercised to exclude competitors. As noted, a standard owner may refuse to make his product compatible with those of competitors, just as Microsoft did to DR-DOS.163 In addition, the dominating firm may compel vertically related firms to deal exclusively with it, as is argued by the post-Chicago theorists164 and is demonstrated by Microsoft’s licensing practices.165 Without support from complementary products, a product in a network market is not accepted by consumers and thus fails to create its own network. Where a dominant firm controls the complementary products either by producing the complementary products itself or by forcing the complementary firms not to deal with competitors, it would be difficult for a competitor to enter the market even though the competitor has a superior and beneficial technology. Consequently, the market power of a dominant firm has to be limited at this stage. To sum up, where a firm is at the stage of competition for standards, no measure is necessary to limit the patent rights of the firm. Where a firm becomes dominant in the relevant market, however, the proprietary control of the patents covering standards has to be limited. Where the dominant firm can effectively strengthen and extend its market power by exercising it, innovation in the market is likely to be harmed since the innovation trajectory and the incentive to innovate in the market are wholly up to the dominant firm. Indeed, the result of this analysis is similar to the arguments of Weiser and O’Rourke. O’Rourke argues for the ‘dynamic compatibility regime’, which refers to the protection regime of APIs, changed from a proprietary regime to a pro-compatibility regime.166 However, O’Rourke only suggests the principles and does not suggest the detailed implementation schemes. Weiser proposes a detailed form of O’Rourke’s argument. He differentiates the

163 Baseman et al., ‘Microsoft Plays Hardball: The Use of Exclusionary Pricing and Technical Incompatibility to Maintain Monopoly Power in Markets for Operating System Software’, 12–13. 164 Hovenkamp, ‘Antitrust Policy After Chicago’, 272–3. 165 Baseman et al., ‘Microsoft Plays Hardball: The Use of Exclusionary Pricing and Technical Incompatibility to Maintain Monopoly Power in Markets for Operating System Software’, 9–12. 166 O’Rourke, ‘Toward a Doctrine of Fair Use in Patent Law’, 1218.

204


competition for standards stage from the competition within standards stage and argues that intellectual property reform is necessary only for the competition within standards stage.167 In this respect, his model is similar to the dynamic compatibility regime of O’Rourke. Based on empirical studies, Weiser implements the competitive platform model, which is a detailed form of the dynamic compatibility regime. However, the weakness of his argument is that his suggestion for intellectual property reform is limited to the legality of reverse engineering.168 Admittedly, Weiser’s and O’Rourke’s proposals provide crucial insights into the relationship between patent rights and standards and indeed the argument of this book is in line with their arguments. However, as is clear in the following sub-sections, there are significant differences between the model proposed by this book and their models. On the basis of the incentive structure and the analysis thereof, this book’s proposal for patent reform is described in the following sub-sections. 3.2

Dynamic Liability Rules Regime

3.2.1 Basic principles of patent reform The examination of patent reform suggested by other commentators and the analysis of the incentive structure in relation to standards lead to the following basic principles of this book’s proposal: 1.

2. 3. 4. 5.

The patent reform has to be able to facilitate voluntary bargaining between the owner of patents covering standards and standard users at the stage of competition within standards. The market power of a standard owner has to be limited at the stage of competition within standards. The reform has to be based on well established legal practices to reduce the costs arising from the reform. The fine-tuned patent system has to provide sufficient incentives to invent, innovate and disclose. The limitation on patent rights has to be imposed in such a way as to promote dynamic efficiency.

There is no need to describe numbers 1 to 4 of the basic principles any more since they are sufficiently analysed above. However, the fifth element of the basic principles needs more explanation. To illustrate, consider Figures 7.1 and 7.2. In those figures, A and A′ represent the standards protected by patents,

167 168

Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 583–94. Ibid.: 600–612.


205

A B C

Notes: A: Standards protected by patents. B: Improvement invention based on standards. C: Standards necessary for invention B.

Figure 7.1

A significant/radical improvement

B and B′ are improvement inventions which incorporate some parts of the patented standards, and C and C′ are the parts of patented standards which are included in B and B′ respectively. Thus, Figures 7.1 and 7.2 demonstrate the cases where an improvement invention B or B′ is created on the basis of patented standards A or A′ respectively. The difference between Figures 7.1

A′

B′ C′

Notes: A′: Standards protected by patents. B′: Improvement invention based on standards. C′: Standards necessary for invention B.

Figure 7.2

A minor improvement

206


and Figure 7.2 lies in the value created by improvement inventions. That is, the value created by improvement inventions is far greater in Figure 7.1 than in Figure 7.2. In other words, B is a significant improvement or a radical improvement, whereas B′ is a minor improvement.169 Minor improvements refer to those inventions which are modified obviously or trivially from the patented inventions from the perspective of a person skilled in the art. Minor improvements are not eligible for patent protection since they create so little value that they do not overcome the threshold set up by the patentability requirements, especially the inventive step requirement. Significant or radical improvements refer to inventions which create such large value that they are eligible for patent protection. As described in Chapter 5, the blocking patent doctrine applies to significant improvements and the reverse doctrine of equivalents may apply to radical improvements.170 From the perspective of technical innovation, minor improvements do not contribute to technical advancement very much and thus the incentive to invent minor improvements is not very important. Indeed, any skilled person can invent minor improvements since minor improvement is obvious to a person skilled in the art. On the other hand, significant and radical improvements are not obvious and thus are eligible for patent protection. They are directly related to dynamic efficiency and thus providing the incentive to invent significant and radical improvements by designing patent policies properly is important. It is argued here that the patent rights of standard owners have to be limited to provide incentives for significant and radical improvements but should not be restricted for minor improvements. It is not proper to limit the patent rights of standard owners to facilitate minor improvements since most of the value created by minor improvements results from the patented standards. Should the patent rights be restricted for minor improvements, the incentive structure of patent systems would be seriously harmed. Minor improvements are merely copies of the patented inventions and thus limiting patent rights to encourage minor improvements would cause the free-rider problem. However, limiting the patent rights of standard owners needs to be considered in the case of significant and radical improvements. Since significant and radical improvements are eligible for patent protection, the incentive to invent a first invention needs to be balanced against the incentive to invent the improvement inventions. Designing the proper incentive structure for the first invention and the subsequent inventions thereof is essential to increase dynamic efficiency. Therefore,

169 For the discussion of minor, significant and radical improvements, see Chapter 5, Section 2.3. 170 See Chapter 5, Section 2.3.


207

the patent rights of standard owners may need to be limited in those cases where the improvement inventions are significant and radical improvements. In some sense, the argument here is similar to the exceptional circumstances of Magill and IMS. As is described in Chapter 4, the ECJ established the exceptional circumstances where the exclusive rights of a dominant undertaking are limited.171 The exceptional circumstances are: firstly, there is no actual or potential substitute for the product protected by copyrights and thus the refusal to license copyrights prevents the emergence of a new product for which there is a constant consumer demand; secondly, there is no objective justification for the refusal; and thirdly, the copyright holders, by denying access to the basic information, have reserved to themselves the secondary market by excluding all competition on that market.172 One of the uncertainties in applying the exceptional circumstances is the meaning of ‘a new product’.173 How to define a new product is not clear. The model proposed by this book may provide a valuable insight into the meaning of a new product. A new product may be defined as a product which is eligible for patent protection in patent cases; that is, significant and radical improvements can be considered new products, whereas minor improvements are not considered new products. To sum up, the model suggested by this book takes into consideration significant and radical improvements but does not consider minor improvements. In this model, the patent rights of standard owners are restricted with respect to significant and radical improvements, which are eligible for patent protection, but not to minor improvements, which are not patentable. 3.2.2 Suggested model As a proposal for patent reform which meets the five basic principles, this book suggests the model summarised in Table 7.2. As for the competition for standards stage, no measure is needed since competition in the market would facilitate bargaining and innovations. However, as far as the stage of competition within standards is concerned, proprietary control over the patented standards has to be limited in such a way as to promote compatibility and the market entry of competitors. What is argued for here is to apply liability rules, including compulsory licences, as is suggested by Wagner,174 at this stage.

171 172

See Chapter 4, pp. 61ff. Joined Cases C 241–242/91 P, Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission, paras 48–58. 173 Stothers, ‘IMS Health and Its Implications for Compulsory Licensing in Europe’. 174 For more detail, see Wagner, ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’.

208


Table 7.2

Suggested model Incentive to bargain


Measures


Vertical Horizontal

Large Uncertain

Small Small



Vertical Horizontal

Large Small

Large Large

Liability rules Liability rules

Under liability rules regimes, the owner of patents covering standards cannot exclude others from exploiting the patented technology but instead is entitled to receive financial compensation for the use of the technology.175 More specifically, it is proposed here that in applying liability rules at the stage of competition within standards, the least favourable licensing terms to the standard owner amongst a number of previous licensing agreements have to be applied. As is clear later in this section, applying the least favourable terms will increase the incentives for voluntary bargaining. What effects then do liability rules including compulsory licences at this stage have on the costs and benefits of patent systems in relation to standards? Firstly, consider the bargaining between patent owners and patent users. The standard users are entitled to compulsory licences as long as they provide the financial compensation. This will increase the bargaining power of standard users and thus they can be in a better position in bargaining with standard owners. Nevertheless, a standard user has a large incentive to bargain with standard owners since it may not be possible to obtain the essential technical support from standard owners even though the user can obtain the licence compulsorily. Thus, the standard user is likely to seek to bargain voluntarily.176 Moreover, a standard owner has an incentive to bargain with users since the mere possibility of compulsory licences is a real threat to the standard owner. On top of that, where voluntary bargaining fails, the compulsory licence will be imposed on terms and conditions which are least favourable to the standard owner. Thus, the standard owner is likely to seek to make a deal

175 176

Ibid.: 1079–80. Weiser, ‘The Internet, Innovation, and Intellectual Property Policy’, 608–9.


209

voluntarily with the users on terms and conditions favourable to him. Thus, applying liability rules at this stage would increase the possibility of efficient bargaining between the bargaining parties. Secondly, consider the cost arising from abuse of the market power of a standard owner. Since the standard owner cannot refuse to license, competitors can make their products compatible with the dominant standard, which reduces the possibility of abuse of the market power of the standard owner. On top of that, at the stage of competition within standards, the standard owner is not likely to bargain with competitors and thus liability rules are better than property rules according to conventional wisdom.177 Thus, liability rules have to be applied to the bargain between a standard owner and a competitor. As for the vertical relationship, the standard owner has little reason not to bargain with vertically related firms and thus applying liability rules is not likely to harm the standard owner. By applying liability rules to bargaining between a standard owner and a vertically related firm, it is possible to restrict market power arising from the standard in such a way as to protect abuse. Thus, it is clear that the costs of patent systems arising from bargaining failure and abuse of market power are reduced by adopting the suggested model. Then, how about the benefits of patent systems? Does the proposed model provide sufficient incentives to invent, innovate and disclose? The answer is affirmative with one more measure. The basic assumption of this approach is that the standard owner would recoup sufficient profits at the stage of competition for standards and that he would obtain considerable rents from the standard at the stage of competition within standards even if he did not recoup enough profits at the stage of competition for standards. This assumption is not always established, although it is likely that it is true in most cases. There may be some cases where the standard owner has not recovered the real R&D cost, which is calculated by taking the risk of investment into consideration, as suggested by Kingston. If this is the case, the patent rights of the standard owner should not be restricted. Where a patent system does not ensure sufficient profits even to the successful innovations, the incentive structure of the patent system will not work. Thus, the owner of the patents covering standards has to be able to refuse to deal with others in those cases where the real R&D cost has not been recovered or where the real R&D cost would not be recouped should compulsory licences be imposed. Since it is likely that the patentee has already earned sufficient profits in most cases, the burden of proof has to be on the patentee. Thus the risk-compensated R&D costs, suggested by Professor Kingston, can be used in this model as a safety measure to ensure sufficient incentives to invent, innovate and disclose in patent systems.

177

See Cooter and Ulen, Law and Economics 103.

210


It has to be noted that the proposed scheme adopts the well established legal practices of competition laws. In applying the model, it is necessary to distinguish between the competition for standards stage and the competition within standards stage. Finding the dominance of a firm in the relevant market is well established in competition laws, and thus by applying the practices of competition laws it is possible to differentiate the two stages. It has also to be noted that only the patented technologies using standards are eligible for the compulsory licences in this model. Minor improvements, which are not patentable, cannot obtain compulsory licences. Thus it is argued here that the suggested dynamic liability rules regime reduces the costs of patent systems in relation to standards and does not decrease the benefits of patent systems. Therefore the proposed model is acceptable as a patent reform in relation to standards. 3.3

Possible Criticisms

There may be some criticisms of this model. Firstly, some may criticise the application of liability rules rather than property rules at the stage of competition within standards. The criticisms may be that the application of liability rules is likely to destroy the incentive structure of patent systems,178 that it is difficult to calculate the proper licensing fees reflecting the values of patented inventions,179 and that the court is not able to monitor and enforce the compulsory licences properly.180 However, these criticisms have already been examined when the benefits of patent systems were considered with respect to the proposed model. As for the model’s effect on the incentive structure of patent systems, the harmful effects of this model on the incentives which may otherwise be created by patent systems are minimised by adopting Kingston’s proposal. That is, the suggested patent systems ensure the risk-compensated R&D costs and thus do not harm the incentive structure of patent systems. On top of that, contrary to the critics’ arguments, at least one empirical research demonstrates that compulsory licences have little effect on firms’ incentive to innovate. Scherer investigated 44 companies which were subjected to compulsory licences under antitrust decrees in the US and concluded that there was ‘no significant indication that [the] 44 companies … sustained less intense R&D efforts than other firms of comparable size and industry origin’.181 Therefore,

178 See Tsilas, The Perils of Imposing Compulsory IP Licensing to Achieve Interoperability. 179 Merges, ‘Of Property Rules, Coase, and Intellectual Property’, 2664–7. 180 Hovenkamp et al., ‘Unilateral Refusals to License’, 7–8. 181 F.M. Scherer, The Economic Effects of Compulsory Patent Licensing, The Monograph Series in Finance and Economics (New York: New York University, 1977) 75.


211

it is unlikely that this model would damage the incentive structure of patent systems. The argument that it is difficult to calculate the proper licensing fees is not persuasive in the proposed model. As is argued by Wagner, there are a number of similar licensing agreements in the market and thus finding the proper level of licensing fees is not a problem. Furthermore, since this model adopts the terms and conditions which are least favourable to patent owners and since there can only be one set of least favourable licensing terms, calculating the licensing fees is not a problem at all. As far as the monitoring problem is concerned, it can be overcome by implementing the proper systems, as is argued by Wagner. The court does not need to monitor the licensing agreement day by day. The court only needs to establish dispute resolution systems with respect to the licensing fees. Furthermore, since the suggested model arguably facilitates voluntary bargaining between the relevant parties, the administration costs would not be considerable. Secondly, the other criticism may be that there are uncertainties in determining the relevant market and the dominance of a product. Indeed, one of the main issues in litigation concerning competition law is defining the relevant market and finding the dominance of a product or a firm.182 However, there are a number of precedents for these tasks and, on top of that, it is relatively easy to find the related market and the dominance in network markets due to network externalities. Network externalities, along with the switching costs, lead the network market to have a concentrated market structure, making it easy to find the relevant market and the dominance of a firm. For instance, it is well known that Microsoft is the dominant firm in the PC operating system market and that Qualcomm dominates the CDMA mobile communication technology market. Thus, even though there may be some uncertainties in defining the relevant market and finding the dominance of a firm, these uncertainties do not hinder the application of the suggested model. Thirdly, there may be criticisms about uncertainties arising from the application of the model of compulsory licences with capital payments. As Kingston argues, finding the R&D costs is not difficult but calculating the proper multiples which reflect the risk resulting from the investment may be difficult.183 Therefore, it may be unclear whether a patent owner has recouped the risk-compensated R&D costs, making it difficult to decide whether the patent owner is entitled to refuse to license or not. Admittedly, there are uncertainties in finding the proper multiples but this can be overcome in various

182 183

Whish, Competition Law 178–92. Kingston, ‘Meeting Nelson’s Concerns about Intellectual Property’.

212


ways. Empirical researches on the average probability of a project in the information and communication industries can be conducted to find the risk of a project in those industries. In addition, as similar disputes arise, the court or the Patent Office is likely to become experienced in finding the relevant multiples. More importantly, since the model is used in limited circumstances where the patent owner seeks to deny the compulsory licences by showing that he has not recovered the risk-compensated R&D costs, the costs arising from the model are not prohibitive. There is no legal system which does not have some level of uncertainties and the model suggested by this study also has uncertainties. However, as some commentators argue, uncertainties may facilitate private and voluntary bargaining.184 It is clear that the model proposed here has more certainty than O’Rourke’s fair use model, arguably reducing the costs arising from uncertainties. This model also takes compulsory licences into consideration and thus makes up for the weakness of Weiser’s competitive platform model. In addition, this model applies liability rules only in those cases where the patent owner has recouped or is likely to recover the real R&D costs, thus adopting the advantages of Wagner’s liability model and removing the disadvantages of it. On top of that, this model uses the main idea of Kingston’s model and applies it to protect the incentive structure of patent systems. This model is created by combining the previous models and redesigning them to be fitted to the conflicts between patents and standards. As is argued here, this model reduces the costs of patent systems in relation to standards, while it does not decrease the benefits of patent systems. This model, which is designed on the basis of informal or market standards, can also be applied to formal standards, mitigating the patent hold-up problem in formal standardisation, which is examined in the following sub-sections. 3.4

Application to the Patent Hold-up Problem in Formal Standards

While informal standards arise from market process, formal standards are implemented by standard setting organisations (SSOs). Not all formal standards become dominant in the relevant markets. However, formally promulgated standards are in a good position to be standards in the relevant markets. Where formal standards become standards in the market, the proposed model can be applied. That is, while a formal standard is not dominant in the market, no measures are necessary, as is the case at the stage of competition for stan184 For more detail, see Ayres and Klemperer, ‘Limiting Patentees’ Market Power without Reducing Innovation Incentives: The Perverse Benefits of Uncertainty and Non-Injunctive Remedies’.


213

dards in informal standards. Once the formal standard becomes dominant in the market, however, the measures for the stage of competition within standards of informal standards have to be applied to the formal standard. Thus the criterion for limiting the patent rights of formal standards has to be the market power of formal standards. By treating formal standards in the same ways as informal standards, it is possible to decrease the costs of patent systems in relation to standards and to provide economic entities with sufficient incentives to carry on inventive activities. As is described in GSM standardisation, one of the crucial issues of formal standardisation is the patent hold-up problem. Namely, the patent owner may refuse to let his patents be included in a formal standard. Indeed, where a patent owner considers it disadvantageous to participate in an SSO and to let his patents be included in a formal standard, he may choose not to participate in or to withdraw from an SSO, and seek to enforce the patents in markets, blocking the standardisation process. This is called the patent hold-up problem and the patents at issue are called blocking patents. The patent hold-up problem becomes more serious where there is no technology which can be an alternative to the patented technology, as is the case in the standardisation of GSM in Europe, since the patent owners may debilitate an SSO. Under the current intellectual property regime, patent owners who control informal standards have fewer obligations for the patents covering standards than patent holders who have their patents incorporated into formal standards. Since informal standards arise from market process, the patent holders of informal standards do not have to agree to the terms and conditions of licences of their patents in getting their patented technology adopted as standards. As for formal standards, however, SSOs usually have IPR policies which require reasonable and non-discriminatory licences of the patent.185 Accordingly, the patent holders may choose not to participate in the formal standardisation process if they do not wish to grant licences of their patents, which causes the patent hold-up problem. The dynamic liability rules regime proposed by this book may mitigate the patent hold-up problem of an SSO. Under the dynamic liability rules regime, the patent owner of informal standards has obligations similar to those of formal standards. Namely, the patent owner has to grant licences of the relevant patents compulsorily and thus the incentive not to participate in the formal standardisation is reduced. Rather, the formal standardisation may be a safer way to make the patents be included in a standard and thus the patent owner is encouraged to be a member of an SSO. Thus the proposed model may

185

1095–6.


214


mitigate the patent hold-up problem in formal standardisation. In this sense, the way to resolve the patent hold-up problem in an SSO which is proposed by this model is an indirect one. It has to be noted that this model does not seek to take private property forcibly to promulgate a standard. Rather, this model seeks to solve the problem by balancing the incentive to have informal standards against that for having formal standards. Under this regime, rational patent owners would find that blocking the formal standardisation process is not a good option.

4.

IMPLEMENTATION OF THE PROPOSED MODEL

The best way to implement the proposed model is to embody it in the TRIPS Agreement. The TRIPS Agreement prescribes the international framework of intellectual property law and each member state has to comply with the provisions of the Agreement. Thus, in order to be implemented nationally or internationally, the model has to be realised in the TRIPS Agreement or to be allowed by the TRIPS Agreement. As is described in Chapter 5, it is unlikely that the proposed model would be accommodated by the TRIPS Agreement as it is, since compulsory licences can only be granted in limited circumstances under the TRIPS Agreement.186 Namely, under the TRIPS Agreement, compulsory licences can be imposed on patent owners in those cases (1) where the patent is not worked or the demand for the product covered by the patents is not met, (2) where compulsory licences are essential for the public interest, and (3) where the licences are necessary for a dependent patent, which cannot be exploited without infringing another patent, to be worked. It is unlikely that the compulsory licences of the proposed model would be imposed due to insufficient use since patents covering standards are likely to be used sufficiently. It is also doubtful that the proposed dynamic liability rules regime would be realised by means of the compulsory licences which are imposed on account of the public interest since the patents in the public interest are those protecting inventions essential to human beings. Patents protecting the technologies for pollution control, medicine, food, energy and weapons are examples. It is also unlikely that compulsory licences for dependent patents would fully realise the proposed model since only radical improvements are entitled to compulsory licences and, as far as significant improvements are concerned, compulsory licences are not likely to be granted. Accordingly, it is unlikely that the proposed model would be realised in the current TRIPS Agreement.

186

For more detail, see Chapter 5, Section 3.


215

Since the proposed model cannot be accommodated by the current TRIPS Agreement, the Agreement needs to be modified to implement the model. Considering that the objectives of the TRIPS Agreement are to design international intellectual property regimes which ‘contribute to the promotion of technological innovation and to the transfer and dissemination of technology, to the mutual advantage of producers and users of technological knowledge and in a manner conducive to social and economic welfare, and to a balance of rights and obligations’187, there is no reason why the Agreement cannot be changed to embody the proposed model. Since the proposed dynamic liability rules regime is likely to meet the objectives of the TRIPS Agreement, the model can be realised in the Agreement. Article 31 prescribes compulsory licences and thus it is the best candidate to be changed to accommodate the model.

187 Agreement on Trade-Related Aspects of Intellectual Property Rights 1994, Article 7.

8. Conclusion In network markets, the need for compatibility leads the market to use standards. Where standards are protected by patents, the need for open access to standards clashes with the exclusivity of patent rights. More seriously, where the owner of patents covering standards refuses to grant licences to others hoping to enter the relevant market by making products compatible with standards, the market competition is seriously distorted. In addition, where the patent owner refuses to let his patent be incorporated into standards, the standardisation process can be halted or seriously delayed. This book seeks to resolve these problems by finding ways to strike the right balance between the rights of patent owners and the need for industry standards. Especially, this book seeks to solve the problem arising from the refusal to license by patent owners who have the essential patents for standards. One of the main findings of this book is that current legal systems do not resolve the issues in a satisfactory way. The antitrust law of the US, the competition law of the EU and the legal doctrines of patent law are examined in relation to standards in Chapters 3 to 5. In Chapter 3, the patent issues in standards are examined from the perspective of the US antitrust law. The US antitrust law does not condemn the patent owner who refuses to license patents, even though the patentee is in a monopoly position. There are legal principles which may restrict the exercise of patents covering informal standards, but these principles may be used only in limited circumstances. Various legal principles may be used in relation to formal standards. While the legal principles are effective in preventing the patent owners from seeking undue benefits from their opportunistic behaviours, they do not resolve some issues such as blocking patents. In Chapter 4, the patent issues in formal and informal standards are analysed from the perspective of EU competition law. The provision studied is Article 82 of the EC Treaty, which condemns the abuse of a dominant position. In particular, the principle of the exceptional circumstances is examined with respect to the patent issues of formal and informal standards. Where the patent owner’s refusal to license blocks the standardisation process so that the standardisation process is halted or severely delayed, it is unlikely that the 216

Conclusion

217

principle of the exceptional circumstances can be applied to facilitate the standardisation process. On top of that, there are uncertainties in applying the exceptional circumstances since the interpretation of each element of the exceptional circumstances is not clear. As for informal standards, gaining access to them by relying on the exceptional circumstances is a difficult task. Moreover, applying the exceptional circumstances test is not a good way to obtain a licence from the owner of patents covering standards because of the facts that the patent owner may circumvent the accusation by a single licence and that the test does not properly consider the right balance between the standard owner and the improvement inventors of the standard. Chapter 5 examines how the issues may be resolved from the perspective of patent law. The reverse doctrine of equivalents and compulsory licences are studied with respect to the patent issues in standards. The reverse doctrine of equivalents may be useful in the case of improvement inventions, but it is not so helpful in resolving the patent issues in standards since what matters in relation to standards is not immunity from the infringement charge, which the reverse doctrine of equivalents may provide, but the availability of the very same technology protected by the patent essential to standards. A compulsory licence is another patent law principle which may be applied to the patent issues in standards. According to the TRIPS Agreement, where the dependent patent cannot be exploited without infringing the first patent and where the first patent owner refuses to license the first patent, a compulsory licence can be granted only in those circumstances where the dependent patent has ‘an important technical advance of considerable economic significance in relation to the invention claimed in the first patent’.1 It may be possible to impose compulsory licences on this ground in those circumstances where the invention covered by the dependent patent is a radical improvement of the invention protected by the first patent. However, it is difficult to apply this ground for a compulsory licence since the invention protected by the dependent patent has to be evaluated ex ante. Given that both legal regimes have limitations, this book seeks to find the right balance between standard users and standard owners within the scope of intellectual property law itself. The cost and benefit analysis of patent systems in relation to standards is conducted as follows. Firstly, the costs of patent systems in relation to standards are studied; secondly, ways are sought to reduce the ‘known’ costs of patent systems in relation to standards; and thirdly, the suggested solutions are examined with respect to the benefits of patent systems.

1

Ibid., Article 31 (l).

218


The cost and benefit analysis of patent systems in relation to standards leads to the second main finding of this book, namely that one possible way to reduce the costs of patent systems without harming the incentive structure of patent systems is to facilitate efficient bargaining and to restrict the market power of standard owners by ensuring access to the standards at some stages. By examining how to decrease the costs of patent systems in relation to standards, this book suggests a patent reform in Chapter 7. It attempts to find the proper patent reform by analysing the incentive structure of voluntary bargaining and the abuse of market power. In addition, this book seeks to adopt the advantages of the models described in Chapter 7 and to reduce the disadvantages of them. As a proposal for patent reform, the dynamic liability rules regime is suggested in Table 8.1. Table 8.1

Suggested model Incentive to bargain


Measures


Vertical Horizontal

Large Uncertain

Small Small



Vertical Horizontal

Large Small

Large Large

Liability rules Liability rules

The best way to implement the proposed model is to embody it in the TRIPS Agreement. Since the model is compatible with the objectives of the TRIPS Agreement, it can be realised within the Agreement, possibly by changing Article 31. Thus this book finds a way to resolve the conflict between patents and industry standards. Patent systems, as an incentive system for technical development, face a great challenge in this fast-changing market. In order to function properly, patent systems have to continuously react to the needs of markets. In this context, fine-tuning of patent systems in response to market needs is essential, and that is what this book seeks to do.

References EUROPE Books Adams, John N., Julian J.B. Hickey and K.V. Prichard Jones (2006), Franchising, 5th edn, Haywards Heath: Tottel Publishing. Anderman, Steven D. (1998), EC Competition Law and Intellectual Property Rights, Oxford: Oxford University Press. Begg, David, Stanley Fischer and Rudiger Dornbusch (2005), Economics, 8th edn, London: The McGraw-Hill Companies. Bently, Lionel and Brad Sherman (2004), Intellectual Property Law, 2nd edn, Oxford: Oxford University Press. Blind, Kunt (2004), The Economics of Standards: Theory, Evidence, Policy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. De Vries, Henk J. (1999), Standardization: A Business Approach to the Role of National Standardization Organizations, Boston, Dordrecht, London: Kluwer Academic Publishers. Grindley, Peter (2002), Standards, Strategy, and Policy: Cases and Stories, Oxford: Oxford University Press. Gutterman, Alan S. (1997), Innovation and Competition Policy: A Comparative Study of the Regulation of Patent Licensing and Collaborative Research and Development in the United States and the European Community, London: Kluwer Law International. Jones, Alison and Brenda Sufrin (2008), EC Competition Law, 3rd edn, Oxford: Oxford University Press. Kingston, William (ed.) (1987), Direct Protection of Innovation, Dordrecht, Boston, Lancaster: Kluwer Academic Publishers. Schumpeter, Joseph A. (1943), Capitalism, Socialism, and Democracy, London: Henderson and Spalding. Shy, Oz (2001), The Economics of Network Industries, Cambridge: Cambridge University Press. Swann, G. and M. Peter (2000), The Economics of Standardization: Final Report for Standards and Technical Regulations Directorate, London: Department of Trade and Industry. Torremans, Paul L.C. (2005), Holyoak and Torremans: Intellectual Property Law, 4th edn, Oxford: Oxford University Press. 219

220


Tritton, Guy (2002), Intellectual Property in Europe, 2nd edn, London: Sweet & Maxwell. Whish, Richard (2003), Competition Law, 5th edn, London: LexisNexis UK. Articles Andersen, Birgitte (2003), ‘The Rationales for Intellectual Property Rights in the Electronic Age’. Andersen, Birgitte (2004), ‘If “Intellectual Property Rights” is the Answer, What is the Question? Revisiting the Patent Controversies’, Economics of Innovation and New Technology, 13 (5), 417–42. Derclaye, Estelle (2004), ‘The IMS Health Decision and the Reconciliation of Copyright and Competition Law’, European Law Review, 29, 687–97. Furse, Mark (1995), ‘The “Essential Facilities” Doctrine in Community Law’, European Competition Law Review, 16 (8), 469–72. Good, Diana (1991), ‘1992 and Product Standards: A Conflict with Intellectual Property Rights’, European Intellectual Property Review, 398–403. Iversen, Eric James, Esten Oversjoen and Haakon Thue Lie (2004), ‘Standardization, Innovation and IPR’, Telektronikk, 2, 65–79. Kingston, William, ‘Meeting Nelson’s Concerns about Intellectual Property’, available from http://www.druid.dk/conferences/nw/paper1/kingston.pdf. Kronz, Hermann (1983), ‘Patent Protection for Innovation: A Model’, Parts I and II, European Intellectual Property Review 7, 178–83, 206–10. Liotard, Isabelle and Rudi Bekkers (1999), ‘European Standards for Mobile Communications: The Tense Relationship between Standards and Intellectual Property Rights’, European Intellectual Property Review, 21 (3), 110–26. Narciso, Alessandra and Paul L.C. Torremans (2005), ‘IMS Health or the Question Whether Copyright still Deserves a Specific Approach in a Market Economy?’, in Developments in the Economics of Copyright, edited by Lisa N. Takeyama, Wendy J. Gordon and Ruth Towse, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 142–58. Ong, Burton (2004), ‘Anticompetitive Refusal to Grant Copyright Licences: Reflections on the IMS Saga’, European Intellectual Property Review, 26, (11), 505. Reichman, Jerome H. (2001), ‘Of Green Tulips and Legal Kudzu: Repackaging Rights in Subpatentable Innovation’, in Expanding the Boundaries of Intellectual Property, edited by Rochelle Cooper Dreyfuss and Diane Leenheer Zimmerman, Oxford: Oxford University Press, pp. 23–53. Schiessl, Martin (1993), ‘The New European Telecommunications Standards

References

221

Institute Policy: Conflicts Between Standardisation and Intellectual Property Rights’, European Intellectual Property Review, 15 (8), 263–6. Shurmer, Mark and Gary Lea (1995), ‘Telecommunications Standardization and Intellectual Property Rights: A Fundamental Dilemma?’, StandardView, 3 (2), 50–59. Stothers, Christopher (2004), ‘IMS Health and Its Implications for Compulsory Licensing in Europe’, European Intellectual Property Review, 26 (10), 467–72. Tuckett, Roger (1992), ‘Access to Public Standards: Interoperability Revisited’, European Intellectual Property Review 14 (12), 423–7. Others ‘Annual Report 2005’, Munich: European Patent Office, 2005.

USA Books Bork, Robert H. (1993), The Antitrust Paradox: A Policy at War with Itself, Oxford: The Free Press. Cooter, Robert and Thomas Ulen (2000), Law and Economics, 3rd edn, New York: Addison Wesley Longman. Hovenkamp, Herbert, Mark D. Janis and Mark A. Lemley (2002), IP and Antitrust, Vol. 1, New York: Aspen Law & Business. Hovenkamp, Herbert, Mark D. Janis and Mark A. Lemley (2005), IP and Antitrust, Vol. 2, New York: Aspen Law & Business. Posner, Richard A. (2007), Economic Analysis of Law, 7th edn, New York: Wolters Kluwer. Scherer, F.M. (1977), The Economic Effects of Compulsory Patent Licensing, The Monograph Series in Finance and Economics, New York: New York University. Scherer, F.M. (1980), Industrial Market Structure and Economic Performance, 2nd edn, Chicago: Rand McNally College Publishing Company. Scherer, F.M. and David Ross (1990), Industrial Market Structure and Economic Performance, 3rd edn, Boston: Houghton Mifflin Company. Shapiro, Carl and Hal R. Varian (1999), Information Rules, Boston: Harvard Business School Press.

222


Articles Ahlborn, Christian, David S. Evans and Jorge Padilla (2005), ‘The Logic and Limits of the “Exceptional Circumstances Test” in Magill and IMS Health’, Fordham International Law Journal, 28, 1109. Areeda, Phillip (1989), ‘Essential Facilities: An Epithet in Need of Limiting Principles’, Antitrust Law Journal, 58, 841. Arquit, Kevin J., Arman Y. Oruc and Richard Wolfram (1998), ‘Antitrust, Intellectual Property, Standards and Interoperability’, Practising Law Institute 524, 200. Arrow, Kenneth J. (1962), ‘Economic Welfare and the Allocation of Resources for Invention’, in The Rate and Direction of Inventive Activity: Economic and Social Factors, Cambridge, MA: National Bureau of Economic Research, p. 609. Ayres, Ian and Paul Klemperer (1999), ‘Limiting Patentees’ Market Power without Reducing Innovation Incentives: The Perverse Benefits of Uncertainty and Non-Injunctive Remedies’, Michigan Law Review, 97, 985. Ayres, Ian and Eric Talley (1995a), ‘Distinguishing between Consensual and Nonconsensual Advantages of Liability Rules’, Yale Law Journal, 105, 235. Ayres, Ian, and Eric Talley (1995b), ‘Solomonic Bargaining: Dividing a Legal Entitlement to Facilitate Coasean Trade’, Yale Law Journal, 104, 1027. Baker, Jonathan B. (1995), ‘Fringe Firms and Incentives to Innovate’, Antitrust Law Journal, 63, 621–41. Balto, David A. and Andrew M. Wolman (2003), ‘Intellectual Property and Antitrust: General Principles’, IDEA: The Journal of Law and Technology, 43, 395. Barton, John H. (1997), ‘Patents and Antitrust: A Rethinking in Light of Patent Breadth and Sequential Innovation’, Antitrust Law Journal, 65, 449. Barton, John H. (2000), ‘Reforming the Patent System’, Science, 1933–4. Baseman, Kenneth C., Frederick R. Warren-Boulton and Glenn A. Woroch (1995), ‘Microsoft Plays Hardball: The Use of Exclusionary Pricing and Technical Incompatibility to Maintain Monopoly Power in Markets for Operating System Software’, Antitrust Bulletin, Summer, 265–315. Calabresi, Guido and A. Douglas Melamed (1972), ‘Property Rules, Liability Rules, and Inalienability: One View of the Cathedral’, Harvard Law Review, 85, 1089. Candeub, Adam (2005), ‘Trinko and Re-grounding the Refusal to Deal Doctrine’, University of Pittsburgh Law Review, 66, 821. Coase, Ronald H. (1960), ‘The Problem of Social Cost’, Journal of Law and Economics, 3, 1.

References

223

Cohen, Julie E. (2001), ‘Patent Scope and Innovation in the Software Industry’, California Law Review, 89, 1. Curran, P.D. (2003), ‘Standard-Setting Organizations: Patents, Price Fixing, and Per Se Legality’, University of Chicago Law Review, 70, 983. David, Paul A. (1985), ‘Clio and the Economics of QWERTY’, The American Economic Review, 75 (2), 332–7. David, Paul A. (1987), ‘Some New Standards for the Economics of Standardization in the Information Age’, in Economic Policy and Technological Performance, edited by Partha Dasgupta and Paul Stoneman, Cambridge: Cambridge University Press, pp. 206–39. David, Paul A. and Shane Greenstein (1990), ‘The Economics of Compatibility Standards: An Introduction to Recent Research’, Economics of Innovation and New Technology, 1, 3. David, Paul A. and W. Edward Steinmueller (1994), ‘Economics of Compatibility Standards and Competition in Telecommunication Networks’, Information Economics and Policy, 6, 217–41. De Vellis, James C. (2003), ‘Patenting Industry Standards: Balancing the Rights of Patent Holders with the Need for Industry-Wide Standards’, AIPLA Quarterly Journal, 31, 301. Donahey, Teague I. (1997), ‘Termainal Railroad Revisited: Using the Essential Facilities Doctrine to Ensure Accessibility to Internet Software Standards’, AIPLA Quarterly Journal, 25, 277. Dumortier, Jos and Mark Hyland (1999), ‘The Ericsson/Qualcomm Dispute and the Essential Facilities Doctrine in the Mobile Telecoms Sector’, Computer and Telecommunications Law Review, 5 (6), 150–57. Eisenberg, Rebecca S. (1989), ‘Patents and the Progress of Science: Exclusive Rights and Experimental Use’, University of Chicago Law Review, 56, 1017. Ellis, Williams (1995), ‘Intellectual Property Rights and High Technology Standards’, in Standard Policy for Information Infrastructure, edited by Brian Kahin and Janet Abbate, Cambridge: The MIT Press. Farrell, Joseph (1989), ‘Standardization and Intellectual Property’, Jurimetrics Journal, 30, 35. Farrell, Joseph (1996), ‘Creating Local Competition’, Federal Communications Law Journal, 49, 201. Farrell, Joseph and Garth Saloner (1985), ‘Standardization, Compatibility and Innovation’, The RAND Journal of Economics, 16 (1), 70–83. Farrell, Joseph and Garth Saloner (1987), ‘Competition, Compatibility and Standards: The Economics of Horses, Penguins and Lemmings’, in Product Standardization and Competitive Strategy, edited by H. Landis Gabel, Oxford: Elsevier Science Publishers, pp. 1–21. Fauver, Cole M. (1988), ‘Compulsory Patent Licensing in the United States:

224


An Idea whose Time has Come’, Northwestern Journal of International Law and Business, 8, 666. Feldman, Robin Cooper (1999), ‘Defensive Leveraging in Antitrust’, Georgetown Law Journal, 87, 2079. Forrester, Ian S. (2005), ‘Article 82: Remedies in Search of Theories?’, Fordham International Law Journal, 28, 919. Galliani, William S. (1990), ‘Patent Infringement Amidst Rapidly Evolving Technologies: New Equivalents, the Doctrine of Equivalents and the Reverse Doctrine of Equivalents’, Santa Clara Computer and High Technology Law Journal, 6, 75. Gerber, David J. (1988), ‘Rethinking the Monopolist’s Duty to Deal: A Legal and Economic Critique of the Doctrine of “Essential Facilities”’, Virginia Law Review, 74, 1069. Gordon, Wendy J. (1982), ‘Fair Use as Market Failure: A Structural and Economic Analysis of the Betamax Case and Its Predecessors’, Columbia Law Review, 82, 1600. Greenstein, Shane (1992), ‘Invisible Hands and Visible Advisors: An Economic Interpretation of Standardization’, Journal of the American Society for Information Science, 43 (8), 538–49. Greenstein, Shane (2001), ‘The Three Coms of the Microsoft Antitrust Suit: Competition Policy, Commercial Experimentation, and Computing Platforms’, UWLA Law Review, 32, 95–102. Hardin, Garrett (1968), ‘The Tragedy of the Commons’, Science, 162, 1243–8. Hovenkamp, Herbert (1985), ‘Antitrust Policy After Chicago’, Michigan Law Review, 84, 213. Hovenkamp, Herbert, Mark D. Janis and Mark A. Lemley (2006), ‘Unilateral Refusals to License’, Journal of Competition Law & Economics, 2, 1. Julian-Arnold, Gianna (1993), ‘International Compulsory Licensing: The Rationales and the Reality’, IDEA: The Journal of Law and Technology, 33, 349. Kaplow, Louis (1984), ‘The Patent–Antitrust Intersection: A Reappraisal’, Harvard Law Review, 97, 1813. Kaplow, Louis (1985), ‘Extension of Monopoly Power through Leverage’, Columbia Law Review, 85, 515. Kaplow, Louis and Steven Shavell (1995), ‘Do Liability Rules Facilitate Bargaining? A Reply to Ayres and Talley’, Yale Law Journal, 105, 221. Kaplow, Louis and Steven Shavell (1996), ‘Property Rules Versus Liability Rules: An Economic Analysis’, Harvard Law Review, 109, 713. Katz, Michael L. (2002), ‘Intellectual Property Rights and Antitrust Policy: Four Principles for a Complex World’, Journal of Telecommunications & High Technology Law, 1, 325. Kingston, William (1994), ‘Compulsory Licensing with Capital Payments as

References

225

an Alternative to Grants of Monopoly in Intellectual Property’, Research Policy, 23, 661–72. Kingston, William (2001), ‘Innovation Needs Patents Reform’, Research Policy, 30, 403–23. Kitch, Edmund W. (1977), ‘The Nature and Future of the Patent System’, Journal of Law and Economics, 20, 265. Kitch, Edmund W. (2000), ‘Elementary and Persistent Errors in the Economic Analysis of Intellectual Property’, Vanderbilt Law Review, 53, 1727. Kobak, James B. (2005), ‘Intellectual Property, Refusals to Deal and the U.S. Antitrust Laws’, Practising Law Institute, Patents, Copyrights, Trademarks, and Literary Property Course Handbook Series 832, p. 385. Korah, Valentine (2002), ‘The Interface between Intellectual Property and Antitrust: The European Experience’, Antitrust Law Journal, 69, 801. Landers, Amy L. (2006), ‘Let the Game Begin: Incentives to Innovation in the New Economy of Intellectual Property Law’, Santa Clara Law Review, 46, 307. Langlois, Richard N. (1999), ‘Technological Standards, Innovation, and Essential Facilities: Toward a Schumpeterian Post-Chicago Approach’, paper presented at the George Mason University conference on Dynamic Competition and Antitrust, Washington, DC. Lemley, Mark A. (1996), ‘Antitrust and the Internet Standardization Problem’, Connecticut Law Review, 28, 1041. Lemley, Mark A. (1997), ‘The Economics of Improvement in Intellectual Property Law’, Texas Law Review, 75, 989. Lemley, Mark A. (2002), ‘Intellectual Property Rights and Standard-Setting Organizations’, California Law Review, 90, 1889. Lemley, Mark A. (2003), ‘Policy Levers in Patent Law’, Virginia Law Review, 89, 1575. Lemley, Mark A. (2004), ‘Ex Ante versus Ex Post Justifications for Intellectual Property’, University of Chicago Law Review, 71, 129. Lemley, Mark A. and David McGowan (1998), ‘Legal Implications of Network Economic Effects’, California Law Review, 86, 479–611. Levin, Richard C. et al. (1987), ‘Appropriating the Returns from Industrial Research and Development’, Brookings Papers on Economic Activity, no. 3, 783–831. Lim, Daryl (2007), ‘Copyright under Seige: An Economic Analysis of the Essential Facilities Doctrine and the Compulsory Licensing of Copyrighted Works’, Albany Law Journal of Science and Technology, 17, 481. Llobet, Gerard, Hugo Hopenhayn and Matthew Mitchell (2000), ‘Rewarding Sequential Innovators: Prizes, Patents and Buyouts’, Federal Reserve Bank of Minneapolis Research Department Staff Report 273, Federal Reserve Bank of Minneapolis. Machlup, Fritz (1958), ‘An Economic Review of the Patent System’, Study of

226


the Subcommittee on Patents, Trademarks, and Copyrights of the Committee on the Judiciary, United States Senate, Washington, DC: Government Printing Office, pp. 1–89. McGowan, David (1996), ‘Regulating Competition in the Information Age: Computer Software as an Essential Facility under the Sherman Act’, Hastings Communications and Entertainment Law Journal, 18, 771. McMahon, Kathryn (2007), ‘Interoperability: “Indispensibility” and “Special Responsibility” in High Technology Markets’, Tulane Journal of Technology and Intellectual Property, 9, 123. Menell, Peter S. (1987), ‘Tailoring Legal Protection for Computer Software’, Stanford Law Review, 39, 1329. Merges, Robert P. (1994a), ‘Intellectual Property Rights and Bargaining Breakdown: The Case of Blocking Patents’, Tennessee Law Review, 62, 75. Merges, Robert P. (1994b), ‘Of Property Rules, Coase, and Intellectual Property’, Columbia Law Review, 94, 2655. Merges, Robert P. (1996), ‘Contracting into Liability Rules: Intellectual Property Rights and Collective Rights Organizations’, California Law Review, 84, 1293. Merges, Robert P. (1999), ‘Who Owns the Charles River Bridge? Intellectual Property and Competition in the Software Industry’, available at http://www.escholarship.org/uc/item/3704w3hn. Merges, Robert P. and Richard R. Nelson (1990), ‘On the Complex Economics of Patent Scope’, Columbia Law Review, 90, 839. Merges, Robert P. and Richard R. Nelson (1994), ‘On Limiting or Encouraging Rivalry in Technical Progress: The Effect of Patent Scope Decisions’, Journal of Economic Behavior and Organization, 25, 1–24. Meyers, Christopher J. (2005), ‘European Union Competition Law and Intellectual Property Licensing: Trans-Atlantic Convergence and Compulsory Licensing’, Practising Law Institute 842, 135. Mueller, Janice M. (2001), ‘Patenting Industry Standards’, John Marshall Law Review, 34, 897. O’Donnell, S.W. (2004), ‘Unified Theory of Antitrust Counterclaims in Patent Litigation’, Virginia Journal of Law and Technology, 9, 8. O’Rourke, Maureen A. (2000), ‘Toward a Doctrine of Fair Use in Patent Law’, Columbia Law Review, 100, 1177. Oettinger, Nicolas (2001), ‘In Re Independent Service Organizations Antitrust Litigation’, Berkeley Technology Law Journal, 16, 323. Opi, Sergio Baches (2001), ‘The Application of the Essential Facilities Doctrine to Intellectual Property Licensing in the European Union and the United States: Are Intellectual Property Rights Still Sacrosanct?’, Fordham Intellectual Property, Media and Entertainment Law Journal, 11, 409. Papciak, W. Greg (1999), ‘Sherman Act Violations: Essential Facilities

References

227

Doctrine – Intergraph Corp. v. Intel Corp.’, Berkeley Technology Law Journal, 14, 323. Piraino, Thomas A. (2000), ‘Identifying Monopolists’ Illegal Conduct Under the Sherman Act’, New York University Law Review, 75, 809–92. Pitofsky, Robert, Donna Patterson and Jonathan Hooks (2002), ‘The Essential Facilities Doctrine under U.S. Antitrust Law’, Antitrust Law Journal, 70, 443. Posner, Richard A. (1979), ‘The Chicago School of Antitrust Analysis’, University of Pennsylvania Law Review, 127, 925. Powell, Mark D. (2002), ‘Competition Law and Innovation: The Interface between Competition Law and Intellectual Property’, Practising Law Institute, 708, 57. Preovolos, Penelope A. (1999), ‘Antitrust, Intellectual Property, Standards and Interoperability’, Practising Law Institute, 566 , 251–79. Pritikin, David T. and Bruce M. Zessar (1995), ‘The Sham Litigation Doctrine in Patent Litigation: The Nexus Bad Faith Enforcement of Patent Rights and the Antitrust Laws’, Practising Law Institute, 414, 661. Puhala, James P. (1994), ‘Antitrust Law–Berkey Photo and Alaska Airlines: Independent Approaches to Monopoly Leveraging Claims’, Western New England Law Review, 16, 111. Reichman, Jerome. H. (1994–95), ‘Charting the Collapse of the Patent–Copyright Dichotomy: Premises for a Restructured International Intellectual Property System’, Cardozo Arts and Entertainment Law Journal, 13, 475. Sabido, Peter David G. (2003–04), ‘Defending against Patent Infringement Suits in Standard-Setting Organizations: Rambus Inc. v Infineon Technologies’, Federal Circuit Bar Journal, 13, 635. Samuelson, Pamela (2007), ‘Patent Reform through the Courts’, Communications of the ACM, 19. Samuelson, Pamela and Suzanne Scotchmer (2002), ‘The Law and Economics of Reverse Engineering’, Yale Law Journal, 111, 1575. Schallop, Michael J. (2000), ‘The IPR Paradox: Leveraging Intellectual Property Rights to Encourage Interoperability in the Network Computing Age’, AIPLA Quarterly Journal, 28, 195. Scherer, F.M. (1987), ‘Antitrust, Efficiency, and Progress’, New York University Law Review, 62, 998–1019. Schneck, David M. (1998), ‘Setting the Standard: Problems Presented to Patent Holders Participating in the Creation of Industry Uniformity Standards’, Hastings Communications and Entertainment Law Journal, 20, 641–62. Scotchmer, Suzanne (1991), ‘Standing on the Shoulders of Giants: Cumulative Research and the Patent Law’, The Journal of Economic Perspectives, 5 (1), 29–41.

228


Shapiro, Carl (1999), ‘Exclusivity in Network Industries’, George Mason Law Review, 7, 673. Shapiro, Carl and Michael L. Katz (1985), ‘Network Externalities, Competition and Compatibility’, The American Economic Review, 75 (3), 424–40. Shelanski, Howard A. (2000), ‘Competition and Deployment of New Technology in U.S. Telecommunications’, University of Chicago Legal Forum, 85–118. Sheremata, Willow A. (1997), ‘Barriers to Innovation: a Monopoly, Network Externalities, and the Speed of Innovation’, The Antitrust Bulletin, 42, 937–72. Smith, Steven R. (2008), ‘Gresham’s Law in Legal Education’, Journal of Contemporary Legal Issues, 17, 171. Steinman, Melissa Landau (2005), ‘Standards, Antitrust and Intellectual Property: Standard-Setting at the Crossroads’, Practising Law Institute, 832, 241. Tassey, Gregory (2000), ‘Standardization in Technology-based Markets’, Research Policy, 29, 587. Thurow, Lester C. (1997), ‘Needed: A New System of Intellectual Property Rights’, Harvard Business Review, September–October, 95–103. Tripsas, Mary (1997), ‘Unraveling the Process of Creative Destruction: Complementary Assets and Incumbent Survival in the Typesetter Industry’, Strategic Management Journal, 18, 119–42. Turner, Julie S. (1998), ‘The Nonmanufacturing Patent Owner: Toward a Theory of Efficient Infringement’, California Law Review, 86, 179. Turney, James (2005), ‘Defining the Limits of the EU Essential Facilities Doctrine on Intellectual Property Rights: The Primacy of Securing Optimal Innovations’, Northwestern Journal of Technology & Intellectual Property, 3, 179. Valach, Anthony P. (2005), ‘TRIPS: Protecting the Rights of Patent Holders and Addressing Public Health Issues in Developing Countries’, ChicagoKent Journal of Intellectual Property, 4, 156. Wagner, Dana R. (2000), ‘The Keepers of the Gates: Intellectual Property, Antitrust, and the Regulatory Implications of Systems Technology’, Hastings Law Journal, 51, 1073. Weiser, Philip J. (2003), ‘The Internet, Innovation, and Intellectual Property Policy’, Columbia Law Review, 103, 534. Yosick, Joseph A. (2001), ‘Compulsory Patent Licensing for Efficient Use of Inventions’, University of Illinois Law Review, 1275.

References

229

Others ANSI Patent Policy 2007 (cited 30 December 2008), available from http://publicaa.ansi.org/sites/apdl/Documents/Forms/DispForm.aspx?ID=6 376&RootFolder=%2fsites%2fapdl%2fDocuments%2fStandards%20Acti vities%2fAmerican%20National%20Standards%2fProcedures%2c%20Gu ides%2c%20and%20Forms&Source=http%3a%2f%2fpublicaa%2eansi%2 eorg%2fsites%2fapdl%2fDocuments%2fStandards%20Activities%2fAme rican%20National%20Standards%2fProcedures%2c%20Guides%2c%20a nd%20Forms Kanellos, Michael (2005), ‘AMD’s Market Share Gains Accelerate’ (cited 19 May 2006), available from http://news.com.com/AMDs+market+ share+gains+accelerate/2100-1006_3-5916167.html ‘Performance and Accountability Report Fiscal Year 2006’, United States Patent and Trademark Office. SearchSMB.com TechTarget (cited 05 April 2007), available from http://searchsmb.techtarget.com/sDefinition/0,290660,sid44_gci928405,00 .html Tsilas, Nicos L., The Perils of Imposing Compulsory IP Licensing to Achieve Interoperability (cited 24 August 2006), available from http://www. metrocorpcounsel.com/current.php?artType=view&artMonth=October&ar tYear=2005&EntryNo=3739 VESA Local Bus PCMAG.COM (cited 30 December 2008), available from http://www.pcmag.com/encyclopedia_term/0,2542,t=VL-bus&i= 54017,00.asp Webster’s Third New International Dictionary (1986), Vol. II, London: Merriam-Webster Inc. Windows History (cited 30 December 2008), available from http://www. microsoft.com/windows/winhistorydesktop.mspx

Index actual fraud 52 Adams, John N. 77 administration costs, patent systems 122 Ahlborn, Christian 88 Alaska Airlines Inc. v. United Airlines Inc. 37, 40 Anderman, Steven D. 57, 58 Andersen, Birgitte 123, 125, 126, 128, 129 angry orphans 18, 32 ANSI, subsequent discovery of patents 45 antitrust law, USA, Section 2 of the Sherman Act 24–5 Areeda, Phillip 38 Arquit, Kevin J. 45 Arrow, Kenneth J. 141 Arrow’s Paradox of information 141 Aspen Skiing Co. v. Aspen Highlands Skiing Co. 28–9 Chicago school approach 151–2 Post-Chicago approach 154 Aspen Skiing principle 28–33 Atari Game Corp. and Tengen, Inc., v. Nintendo of America Inc. and Nintendo Co., Ltd., 975 F.2d 832 (1992) 183 Atomic Energy Act (USA), compulsory licences 108 Ayres, Ian 165, 167, 192, 212 Baker, Jonathan B. 145 Balto, David A. 56 bandwagon effects 15 Barton, John H. 137, 165 Baseman, Kenneth C. 157, 158, 203 Begg, David 61, 84, 119, 159 Bekkers, Rudi 12, 78, 80, 81, 92 Bently, Lionel 2 Berkey Photo Inc. v. Eastman Kodak Co. 39 Betamax video format 172–3, 174

blind giants 17 Blind, Kunt 9, 10, 11, 13, 15, 18, 20, 21 blocking patent problem 141–2 blocking patents 85–6 Type I 86 application of exceptional circumstances 86–90 Type II 86 application of exceptional circumstances 90–92 Bonjorno v. Kaiser Aluminum & Chemical Corp. 153–4 Bork, Robert H. 151, 152, 155, 158 Boyden Power-Brake Co. et al. v. Westinghouse et al. 104 brick structure 83–4, 94–5 BSI (British Standards Institution) 12 Calabresi, Guido 167 Candeub, Adam 37, 38, 42 CBEMA (Computer and Business Equipment Manufacturers Association) 81 CDMA (Code Division Multiple Access) technologies 78 CEN (European Committee for Standardization) 12–13 CENELEC (European Committee for Electrotechnical Standardization) 12–13 CEPT (European Conference of Postal and Telecommunications Administrations), GSM standard development 78 chain-link theory 150 Clean Air Act (USA), compulsory licences 108 Coase, Ronald H. 127, 137, 167, 171, 210 Coase theorem 137 see also normative Coase theorem Cohen, Julie E. 165 231

232


Columbia Pictures Industries, Professional Real Estate v. 32 compatibility standards (standards) 8–10, 12 benefits 14–15 competition 20–21 anticompetitive effects of standards 22 pro-competitive effects of standards 21 competitive platform model 193–4 arguments for 194–7 evaluation and implication 197–200 compulsory licences 4, 97, 107–8 grounds for dependent patents 112, 115 insufficient use 110–11, 113 public interest 111–12, 113 international framework 108–10 non-transferability 89, 113 proposed system with capital payments arguments for 175–7 evaluation and implication 178–80 review and implication 113–17 consortia, and technical change 19 Consorzio Italiano della Componentistica di Ricambio per Autoveicoli and Maxicar v. Regie Nationale des Usines Renault, Case 53/87 62–3 constructive fraud 52 Cooter, Robert 137, 139, 171, 172, 209 copyright 5–6 fair use doctrine in US copyright law 180–85 reverse engineering of a copyrighted work 183–4 creative destruction 195 cumulative technologies 135–7 Curran, P. D. 45, 140 Data General Corp. v. Grumman Systems Support Corp. 29–30, 31, 35–6 David, Paul A. 8, 10, 11, 17, 21, 32, 36, 88, 129, 194 De Vellis, James C. 43, 45 De Vries, Henk J. 7

deadweight loss 119 Dell Computer Corp., In re, No. 9310097 50–51 dependent patents, grounds for compulsory licences 112, 115 Derclaye, Estelle 61 Diebold, Stambler v. 48 direct network effects 9 doctrine of equivalents 97–103 Donahey, Teague I. 37, 51 DR-DOS 5.0 156, 159 Dumortier, Jos 89 dynamic compatibility regime 203 dynamic liability rules regime 204–10 Eastman Kodak Co. Berkey Photo Inc. v. 39 Image Technical Services Inc. v. 29, 30–32, 41 EC Treaty, Article 82 55 additional requirements 77 dominant position in market 57–9 exceptional circumstances 61 intellectual property and abuse 59–61 relationship with Article 81(3) 77 relevant cases 62–77 relevant market definition 56–7 Eisenberg, Rebecca S. 124, 128 Ellis, William 82 EN (European standards) 13 formal standardisation process 13 institutions involved 12–13 ENV (European pre-standards) 13 equitable estoppel, USA 47–8 Ericsson, GSM standard development 79–80 Erik Veng Ltd, Volvo AB v. 62 essential facilities doctrine 33–9 versus exceptional circumstances 61 ETSI (European Telecommunications Standards Institute) 12–13 establishment 80–81 GSM standardisation work 79 IPR Policy and Undertaking 81–2 EU dominant position in market Article 82 of the EC Treaty 57–9 IMS case 84–5

Index exceptional circumstances Article 82 of the EC Treaty 61 dynamic liability rules regime 207 IMS case 82–4, 88 informal standards 92–3 review and limitations of 93–5 Type I blocking patents 86–90 patents and formal standards 78 patents and informal standards 55 abuse 59 see also EC Treaty, Article 82; ETSI; GSM standard development European Patent Office, administration costs 122 Evans, David S. 88 exceptional circumstances Article 82 of the EC Treaty 61 dynamic liability rules regime 207 IMS case 82–4, 88 informal standards 92–3 review and limitations of 93–5 Type I blocking patents 86–90 excess inertia 17, 42, 145 excess momentum 17 fair use doctrine proposal for patent law arguments for 185–9 evaluation and implication 189–93 in US copyright law 180–85 Farrell, Joseph 15, 16, 17, 19, 42, 45, 80, 143, 145, 157, 164, 188 Fauver, Cole M 108, 112 Feldman, Robin Cooper 161 Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co. 101–3 first mover advantage 170, 172–3 formal/de jure standards 4, 23–4 Forrester, Ian S. 76, 94 Furse, Mark 61 Galliani, William S. 4, 97 Gerber, David J. 135, 150, 151, 152, 156 Good, Diana 13 Gordon, Wendy J. 74, 181 governmental regulation, technical change 20

233

Graver Tank & Manufacturing Co. v. Linde Air Products Co. 98–9, 103–4 Greenstein, Shane 8, 11, 19, 88, 148 Gresham’s law 10 Grindley, Peter 14, 157, 173, 202 Grumman Systems Support Corp., Data General Corp. v. 29–30, 31, 35–6 GSM standard development 78–80 Gutterman, Alan S. 3, 123 Hardin, Garrett 126, 127 HD (European harmonisation documents) 13 Hoffmann–La Roche & Co. AG v Centrafarm Vertriebsgesellschaft Pharmazeutischer Erzeugnisse mbH, Case 102/77 60 Hoffmann–La Roche & Co. AG v. EC Commission, Case 85/76 56, 58 Hopenhayn, Hugo 166 Hovenkamp, Herbert 6, 33, 38, 153, 154, 155, 159, 203, 210 Hyland, Mark 89 ICT (information and communication technology), definitions 1 IEC (International Electrotechnical Commission) 13 Image Technical Services Inc. v. Eastman Kodak Co. (Kodak II) 29, 30–32, 41 implied licence, USA 49–50 IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG, Case C418/01 57 IMS Health Inc. v. EC Commission, Case T184/01 R II (IMS) 67–9, 94 exceptional circumstances 82–4, 88 markets and dominance 84–5 incentive to disclose, patent systems 125, 128–9 incentive to innovate 87 incentive to invent, patent systems 124–5, 128 indirect network effects 9 industry specific patent systems 162–3 informal/de facto standards 4, 10, 23–4 exceptional circumstances 92–3 information platforms 193

234


information standards 11 Insituform of North America, Miller Insituform Inc. v. 26 insufficient use, grounds for compulsory licences 110–11, 113 Intel, Intergraph v. 36 interconnectivity 168 Intergraph v. Intel 36 IPC (international patent classification) 162–3 IPR policies 43 ANSI 43 ISO (International Organization for Standardization) 13 subsequent discovery of patents 45 Istituto Chemioterapico Italiano SpA and Commercial Solvents Corp. v. EC Commission, Cases 6–7/73 61 Iversen, Eric James 2, 78, 79 Jacobs, Advocate General 65–6 Janis, Mark D. 6, 33, 38 Jones, Alison 56, 58 Julian-Arnold, Gianna 4, 97 Kanellos, Michael 37, 92 Kaplow, Louis 132, 153, 156, 159, 167 Katz, Michael L. 9, 116, 143, 168 Kingston, William 87, 122, 137, 165, 166, 175–80, 200, 209, 210, 211, 212 Kitch, Edmund 126–7, 129–30 Kitch, Edmund W. 87, 120, 127, 130, 135, 194, 195 Klemperer, Paul 165, 192, 212 Kobak, James B. 24, 28, 33, 35, 40 Kodak II see Image Technical Services Inc. v. Eastman Kodak Co. Korah, Valentine 67, 74 Kronz, Hermann 87, 165, 166 Landers, Amy L. 165, 166 Langlois, Richard N. 149, 153, 156, 161 law and economics approach 3 Lea, Gary 80, 81, 82 lead time 173 Lemley, Mark A. 6, 9, 20, 24, 33, 38, 43, 44, 45, 46, 48, 49, 50, 51, 53, 75, 83, 97, 105, 124, 125, 126, 127, 129, 130, 133, 139, 140, 142, 143, 163, 167, 168, 213

Levin, Richard C. 129, 134 liability rules proposed application to patents arguments for 167–70 evaluation and implication 171–5 Lie, Haakon Thue 2 Lim, Daryl 76 Liotard, Isabelle 12, 78, 80, 81, 92 Llobet, Gerard 166 Lotus Development Corporation v. Borland International Inc., 49 F3.d 807 (1995) 191–2 Machlup, Fritz 122, 123, 124, 128, 129, 130 Magill see Radio Telefis Eireann & Independent Television Publications Ltd. v. EC Commission mandatory standards 12 Marconi Company, dispute with De Forest 142–3 market for patents 87 McGowan, David 9, 36, 37, 83, 129, 143, 168 McMahon, Kathryn 76 measurement standards 11 Melamed, A. Douglas 167 Menell, Peter S. 41, 42 Merges, Robert P. 75, 97, 104, 105, 121, 129, 135, 136, 140, 141, 142, 165, 166, 167, 170, 171, 186, 210 Metcalf’s law 144 Meyers, Christopher J. 34 Michelin v. Commission see Nederlandsche Banden-Industrie Michelin N.V. v. EC Commission, Case 322/81 Microsoft Corp. computer operating system monopoly 156–61 exclusionary practices 157–8 maintenance of technical incompatibility 158 network externalities 157 v. EC Commission, T-201/04 69–77, 91 Miller Insituform Inc. v. Insituform of North America 26 minimum quality standards 10

Index minor improvements 206 Mitchell, Matthew 166 Mitsubishi Elecs. Am. Inc., Wang Labs. Inc. v. 49–50 monopoly leveraging doctrine 39–41 monopoly power 25 monopoly profits 119–20 Motorola, GSM standard development 79–80 Mueller, Janice M. 1, 44 music recording technologies 147 Narciso, Alessandra 74 natural barriers to entry Chicago school approach 152 Post-Chicago approach 155–6 Nederlandsche Banden-Industrie Michelin N.V. v. EC Commission, Case 322/81 59 Nelson, Richard R. 121, 129, 135, 136, 165, 166, 177, 211 network effects, compatibility and interface standards 9, 83 network externalities and Microsoft monopoly 157 and monopoly 143–8 reduction of costs 164 technical change 16–18 network industries 6 network markets 1 non-excludability 139 non-rivalrous consumption 139 normative Coase theorem 138 normative Hobbes theorem 138–9 objective justification 89 O’Donnell, S.W. 25, 40, 51 Oettinger, Nicolas 27 Ong, Burton 116 Opi, Sergio Baches 31, 32, 34, 39, 41, 61, 89 O’Rourke, Maureen A. 166, 180, 181, 185–93, 198, 199, 200, 203, 204, 212 Oscar Bronner GmbH Co KG v. Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG, Case C-7/97 65–7 Otter Tail Power Co. v. United States 34–5

235

Oversjoen, Esten 2 Padilla, Jorge 88 Papciak, W. Greg 34 Paris Convention for the Protection of Industrial Property, compulsory licences 89, 108, 113–14 Parke, Davis & Company v. Probel and Others, Case 24/67 60 patent fraud, USA 27, 52–3 patent laws 96–7 patent systems administration costs 122 benefits 123, 130–33 competition enhanced by innovation 125–6, 129 costs 131–3, 162–4 abuse of market power 148, 164, 172, 202–3 bargaining theories and standards 137–43, 171–2, 202–3 Chicago school approach to refusal to deal 149–52 homogeneous patent systems 133–7 network externalities and monopoly 143–8 Post-Chicago approach to refusal to deal 153–6 economic rationales 123–7 critiques 127–30 incentive to disclose 125, 128–9 incentive to invent 124–5, 128 prospect theory 127, 129–30 resource misallocation 119–21 social costs 122–3 technology suppression 121–2 tragedy of the commons 126–7 patent systems proposals 166, 216–18 application of liability rules arguments for 167–70 evaluation and implication 171–5 competitive platform model 193–4 arguments for 194–7 evaluation and implication 197–200 compulsory licences with capital payments arguments for 175–7

236


evaluation and implication 178–80 fair use doctrine of patent law arguments for 185–9 evaluation and implication 189–93 suggested model 200–1 application to patent hold-up problem in formal standards 212–14 dynamic liability rules regime 204–10 implementation 214–15 incentive structure 201–4 possible criticisms 210–12 patents 20 year life 168–9 definitions 2 and standards overview 2 research limitations 4–6 research method 3–5 research questions 2–3 penguin effect 16–17, 144–5 perfect competition 120 personal computers, invention by Apple 173, 174 Pharma Intranet Information AG 67 Piraino, Thomas A. 145 Pitofsky, Robert 95 Posner, Richard A. 3, 125, 150 Powell, Mark D. 74 PRE see Professional Real Estate v. Columbia Pictures Industries Preovolos, Penelope A. 24, 25, 33, 39, 43 prisoners’ dilemma 159–60 Pritikin, T. 32 Professional Real Estate v. Columbia Pictures Industries (PRE) 32 prospect theory, patent systems 127, 129–30 public interest, grounds for compulsory licences 111–12, 113 Puhala, James D. 40 Qualcomm, refusal to license 89 QWERTY keyboard arrangement 12 Radio Telefis Eireann & Independent

Television Publications Ltd. v. EC Commission (Magill) 63–4, 68, 94 RAND (reasonable and nondiscriminatory) licences 45 refusal to deal Chicago school approach 149–52 Post-Chicago approach 153–6 refusal to license antitrust principles on 25–8 Qualcomm 89 Reichman, Jerome H. 166 Renault, CICRA v. 62–3 resource misallocation, patent systems 119–21 reverse doctrine of equivalents 4, 96–7, 103–4 implications 105–7 reverse engineering of a copyrighted work 183–4 RF (royalty-free) licences 44–5 Ross, David 19 Sabido, Peter David G. 52 safety standards 10 Saloner, Garth 16, 17, 19, 42, 145, 164, 188 Samuelson, Pamela 165, 166 SBIR (Small Business Innovation Research Act of 1982) 178 Schallop, Michael J. 44 Scherer, F.M. 19, 119, 121, 130, 146, 210 Schiessl, Martin 80 Schneck, David M. 43, 44 Schumpeter, Joseph A. 120, 147, 194 SCM Corp. v. Xerox Corp. 25–7 Scotchmer, Suzanne 135, 165 Sega Enterprise Ltd. v. Accolade Inc., 977 F.2d 1510 (1992) 183–4 Selden patent 136 Shapiro, Carl 9, 83, 143, 144, 147, 168 Shavell, Steven 167 Shelanski, Howard A. 146 Sheremata, Willow A. 145, 147, 156, 159, 194, 195 Sherman, Brad 2 Shurmer, Mark 80, 81, 82 Shy, Oz 1, 136 Siemens, GSM standard development 79–80

Index Smith, Steven R. 10 Sony Computer Entertainment, Inc. v. Connectix Corporation, 203 F3.d 596, 608 (2000) 185 source material 6 sponsored standards 12 SSOs (standard setting organisations), and technical change 19–20 Stambler v. Diebold 48 standards classification according to economic effect 8–12 definitions 7–8 not widely accepted 120 research method 4 see also compatibility standards; information standards; measurement standards; minimum quality standards; safety standards; sponsored standards; unsponsored standards; variety reduction standards standards setting formal standardisation in Europe 12–13 international system 14 informal standardisation benefits of compatibility standards 14–15 dynamics 15 Steinman, Melissa Landau 23, 43 Steinmueller, W. Edward 21 Stothers, Christopher 74, 83, 88, 207 subsequent discovery of patents 45 Sufrin, Brenda 56, 58 Swann, G.M. Peter 8, 10, 11, 33 switching costs 9–10, 83 system technologies 135–7 systems technology 167–8, 171 Talley, Eric 167, 192 Tassey, Gregory 8, 9, 11 TDMA (Time Division Multiple Access)-based technologies 79 technical change 15–16 coordination by agreements 19–20 governmental regulation 20 market coordination 18–19 network externalities 16–18

237

technology suppression, patent systems 121–2 Terminal Railroad Association, United States v. 34 Tesauro, Advocate General 94 Tetra Pak Rausing SA v. EC Commission, Case T51/89 77–8 Thurow, Lester C. 134, 165 Tierce Ladbroke SA v. EC Commission 64–5 tipping effects 15 Torremans, Paul L.C. 3, 74, 94, 123, 125, 131 tragedy of the commons, patent systems 126–7 Trinko LLP, Verizon Communications Inc. v 38, 40 TRIPS Agreement (Agreement on TradeRelated Aspects of Intellectual Property Rights 1994) compulsory licences 88–9, 90, 109–10, 114 for implementation of dynamic liability rules regime 214–15 Microsoft Corp. v. EC Commission 76 Tripsas, Marv 146 Tritton, Guy 57, 59, 60 Tsilas, Nicos L. 75, 170, 210 Tuckett, Roger 78 Turner, Julie S. 111 Turney, James 93, 95 Ulen, Thomas 137, 139, 171, 172, 209 United Airlines Inc., Alaska Airlines Inc. v. 37, 40 United Brands Company and United Brands Continentaal BV v. EC Commission, Case 27/76 57–8 United States, Otter Tail Power Co. v. 34–5 United States v. Terminal Railroad Association 34 unsponsored standards 12 U.S. Patent and Trademark Office, administration costs 122 USA fair use doctrine in copyright law 180–85 legal issues 23–4

238


patentability requirements, flash of creative genius 176 patents and formal standards adoption of standards incorporating patents 43 antitrust liability 50–51 contracts 46–7 equitable estoppel 47–8 fraud 27, 52–3 implied licence 49–50 IPR licensing policies 44–5 issues 42–3, 53–4 legal principles 46, 53 notice and search of patents 44 subsequent discovery of patents 45 patents and informal standards 24 antitrust principles 41–2 antitrust principles on informal standards for compatibility 28–42 antitrust principles on refusal to license 25–8 Aspen Skiing principle 28–33 essential facilities doctrine 33–9 monopoly leveraging doctrine 39–41 Section 2 of the Sherman Act 24–5

see also doctrine of equivalents Valach, Anthony P. 112 Varian, Hal R. 83, 84, 144 variety reduction standards 10–11 Verizon Communications Inc. v. Trinko LLP 38, 40 Volvo AB v. Erik Veng (UK) Ltd, Case 238/87 62 Wagner, Diana R. 166, 167–74, 180, 193, 200, 207, 211, 212 Wang Labs. Inc. v. Mitsubishi Elecs. Am. Inc. 49–50 Warner-Jenkinson Co. v. Hilton Davis Chemical Co. 99–101 Weiser, Philip J. 166, 193–200, 203, 204, 208, 212 Whish, Richard 56, 58, 61, 90, 92, 148, 211 Wolman, Andrew M. 56 Wright brothers 121, 136 wrongful intent 100 Xerox Corp., SCM Corp. v. 25–7 Yosick, Joseph A. 108, 167 Zessar, Bruce M. 32

Patents and Industry Standards

Multimedia Systems: Algorithms, Standards, and Industry Practices

Patents for Chemical Inventions

Classification of Patents

The Law of Patents

Essentials of Patents

Patents, Profits & Power

Polymers, Patents, Profits

Standards And Practicies

Standards and Public Policy

The Economics of Patents and Copyrights

ITV Handbook: Technologies and Standards

Patents, Copyrights & Trademarks For Dummies

Tobacco Industry and Smoking

Frequency Standards: Basics and Applications

Tourism and Travel Industry

Double Standards

Monetary Standards and Exchange Rates

ADSL: Standards, Implementation, and Architecture

Issues In Setting Standards: Establishing Standards

Global Biopiracy. Patents, Plants, And Indigenous Knowledge (Law and Society)

Competition Law and Patents: A Follow-on Innovation Perspective in the Biopharmaceutical Industry (New Horizons in Competition Law and Economics)

Corrosion Tests And Standards Application And Interpretation

Math You Can't Use: Patents, Copyright, and Software

Ethics and the Pharmaceutical Industry

Ethics and the Pharmaceutical Industry

One Night Standards

Standards: Recipes for Reality

Intelligent Transport Systems Standards

Industry Studies

Patents and Industry Standards

Multimedia Systems: Algorithms, Standards, and Industry Practices

Patents for Chemical Inventions

Classification of Patents

The Law of Patents

Essentials of Patents

Patents, Profits & Power

Polymers, Patents, Profits

Standards And Practicies

Standards and Public Policy

The Economics of Patents and Copyrights

ITV Handbook: Technologies and Standards

Patents, Copyrights & Trademarks For Dummies

Tobacco Industry and Smoking

Frequency Standards: Basics and Applications

Tourism and Travel Industry

Double Standards

Monetary Standards and Exchange Rates

ADSL: Standards, Implementation, and Architecture

Issues In Setting Standards: Establishing Standards

Global Biopiracy. Patents, Plants, And Indigenous Knowledge (Law and Society)

Competition Law and Patents: A Follow-on Innovation Perspective in the Biopharmaceutical Industry (New Horizons in Competition Law and Economics)

Corrosion Tests And Standards Application And Interpretation

Math You Can't Use: Patents, Copyright, and Software

Ethics and the Pharmaceutical Industry

Ethics and the Pharmaceutical Industry

One Night Standards

Standards: Recipes for Reality

Intelligent Transport Systems Standards

Industry Studies

Recommend Documents