takeyama gordon towse developments in the economics of copyright

Richard Leon GoveBook Fund

Developments in the Economics of Copyright

To my mom, Margery Takeyama, with love, and in loving

memory ofmy dad, Wallace Takeyama.LNT

To the memory of Sam Postbrief, and to thefuture of his

grandson, Theo Sam Rubin-Petrovic.

WJG

I Developments in the

j

Economics of i

Copyright ;

Research and Analysis

Edited by

Lisa N. TakeyamaAmherst College, and San Francisco State University, USA

Wendy J. Gordon

Boston University, USA

Ruth Towse

Erasmus University, The Netherlands

Edward ElgarCheltenham, UK Northampton, MA, USA

Lisa N. Takeyama, Wendy J. Gordon, Ruth Towse, 2005

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 byEdward Elgar Publishing Limited

Glensanda House

Montpellier Parade

CheltenhamGlosGLSOlUAUK

Edward Elgar Publishing, Inc.

136 West Street

Suite 202

NorthamptonMassachusetts 01060

USA

A catalogue record for this bookis available from the British Library

ISBN 1 84376 930 1

Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents

List offigures vi

List of contributors vii

Preface Lisa N. Takeyama xiv

General outline of the book xvi

1 . Should economics play a role in copyright law and policy? 1

Pamela Samuelson

2. Risk sharing and the distribution of copyright collective

income 23

Arthur Snow and Richard Watt

3. MP3s and copyright collectives: a cure worse than the disease? 37

Stan Liebowitz

4. Peer-to-peer, piracy and the copyright law: implications for

consumers and artists 60

Anne Duchene and Patrick Waelbroeck

5. 'Fair use' as policy instrument 80

Timothy J. Brennan

6. Towards a differentiated products theory of copyright 103

Christopher S. Yoo

7. Private appropriability and sharing of knowledge: convergenceor contradiction? The opposite tragedy of the creative

commons 120

Giovanni B. Ramello

8. IMS Health or the question whether copyright still deserves a

specific approach in a market economy? 142

Alessandra Narciso and Paul L. C. Torremans

9 The basics matter: at the periphery of intellectual property 1 59

F. Scott Kiejfand Troy A. Paredes

Index 191

Figures

3.1 LP albums sold per capita 38

3.2 Album revenues 49

4. 1 Consumer demand with information-push technology 69

4.2 Consumer demand with information-pull technology 72

4.3 Application with 9 = 0.15 75

4.4 Application with cp= 0.30 75

5. 1 Consumer surplus, profits, and costs for work n: 'equal

diversion' 85

5.2 Consumer surplus, profits, and costs for work n: profit-

maximizing price> fair use price 90

5.3 Consumer surplus, profits, and costs for work n: profit-

maximizing price< fair use price 9 1

vi

Contributors

Timothy J. Brennan is Professor of Public Policy and Economics at the

University of Maryland, Baltimore County and Senior Fellow at Resources

for the Future. Prior to coming to UMBC, he was an associate professor of

telecommunications policy at George Washington University and was a

staff economist with the Antitrust Division of the US Department of

Justice. From 1996-97, he served as the senior economist for industrial

organization and regulation on the staff of the White House Council of

Economic Advisers. His primary research areas are in antitrust, regulated

industries, intellectual property, and communications policy, with over 75

articles and chapters in journals and books covering economics, law, energy

policy, telecommunications, philosophy and politics. He serves on the

editorial boards of the Journal of Regulatory Economics, Information

Economics and Policy, and Communications Law and Policy. He is co-

author of Alternating Currents: Electricity Markets and Public Policy. Hereceived his PhD in economics from the University of Wisconsin, Madison

in 1978.

Anne Duchene completed her PhD in economics at CERAS - Ecole

Nationale des Ponts et Chaussees, after graduate studies at Universite de

Paris 1 (La Sorbonne). Her thesis focuses on intellectual property rights,

and more specifically on Internet piracy, agency relationships between

patent lawyers and innovators, and the (dis)functioning of patent offices.

Wendy J. Gordon is Professor of Law and Paul J. Liacos Scholar in Law at

the Boston University School of Law. With Richard Watt she co-edited The

Economics of Copyright (Edward Elgar 2003), and has published on four

continents scholarship using economics and philosophy to analyze the

merits of the legal institutions that regulate information and culture. She

currently serves as a Vice President for the Society for Economic Research

on Copyright Issues; the Chair Elect for the Intellectual Property Section

of the Association of American Law Schools; and as a Visiting Scholar in

Comparative Media Studies at the Massachusetts Institute of Technology.

Among Professor Gordon's honors are: being named a Fulbright Scholar;

being thrice cited by the US Supreme Court; being elected to the Visiting

Senior Research Fellowship at St. John's College, Oxford University; and

vii

viii Developments in the economics of copyright

receiving a residency at the Rockefeller Foundation retreat in Bellagio,

Italy. Her editorial board service includes The Encyclopedia of Law and

Economics and the Review of Economic Research on Copyright Issues,

and she served as the 2002-03 Area Organizer for Intellectual

Property/Electronic Commerce for the American Law and Economics

Association. Professor Gordon's many articles include An inquiry into the

merits of copyright' (Standard Law Review), 'Fair use as market failure'

(Columbia Law Review), A property right in self-expression' (Yale Law

Journal), 'On owning information' (Virginia Law Review), 'Copyright

norms and private censorship' (forthcoming in the Oxford University Press

volume, Copyright and Free Speech), and the Oxford Handbook on LegalStudies chapter on 'Intellectual Property'. Her writings also address areas

such as rhetoric, restitution and property theory.

F. Scott Kieff is an associate Professor in Technology Law and Business in

Washington University School of Law and 2003-05 W. Glenn Campbelland Rita Ricardo-Campbell National Fellow at the Hoover Institution of

Stanford University. He previously served as Visiting Assistant Professor at

the University of Chicago Law School and the Northwestern University

School of Law and as John M. Olin Senior Research Fellow in Law,

Economics, and Business at Harvard Law School. He is also a member of

the founding faculties of the Munich Intellectual Property Law Center and

the Canadian Centre for Intellectual Property Policy at McGill University

Faculty of Law. Scott graduated with a degree in molecular biology and

applied microeconomics from MIT. He was admitted to the New York Bar

after graduation from the University of Pennsylvania Law School and

served as Law Clerk to the Honorable Giles S. Rich on the US Court of

Appeals for the Federal Circuit. Scott has delivered numerous articles and

speeches about obtaining and enforcing intellectual property rights. Heedited the book Perspectives on the Properties of the Human Genome Project

(Elsevier), and co-authored the popular treatise and casebook Principles ofPatent Law (Foundation Press), now in its third edition. His research inter-

ests generally involve the interface among law, economics, ethics, and crea-

tive endeavors such as science, engineering, medicine, and art, with a focus

on technology law and business, intellectual property, contracts, unfair

competition, antitrust, complex litigation, and the allocation of decision-

making ability and authority in disputes involving technological facts.

Having practiced as an associate with the firm of Pennie & Edmonds in

New York, and as an associate and counsel with the firm of Jenner & Block

in Chicago, he has been called to testify as a legal expert before federal

courts and agencies and maintains his connection to the law firm and busi-

ness communities through his ongoing consulting practice, which focuses

Contributors ix

on the strategy and management of intellectual property, competition, and

litigation.

Stan Liebowitz was trained at Johns Hopkins (BA) and UCLA (PhD). Heis currently a Professor of Economics in the School of Management at the

University of Texas at Dallas after having been on the faculty at the

University of Western Ontario, University of Rochester, and a FacultyFellow at the University of Chicago. In addition to five books, he has pub-lished over 60 articles in journals including the American Economic Review

and the Journal of Political Economy, as well as more popular outlets such

as the Wall Street Journal and CIO Magazine. He is director of the Center

for the Analysis of Property Rights and Innovation (CAPRI) and is Vice

President of the Society for Economic Research on Copyright Issues

(SERCI). He also serves on the Editorial Board of the Review ofEconomic

Research on Copyright Issues (RERCI) and is an adjunct scholar at the Cato

Institute. Professor Liebowitz's research interests include the economic

impact of new technologies on copyright owners (most recently, the impactof filesharing), the economics of networks, pricing issues, and antitrust. His

work (with Steve Margolis) on network effects and lock-in, culminated in

two books, Winners, Losers and Microsoft (Independent Institute, 1999)

and The Economics of Qwerty (NYU Press/Palgrave, 2001, edited by Peter

Lewin). His most recent book titled Rethinking the Networked Economywas published by the American Management Association in the fall of

2002. His research has been the focus of articles in The Economist, the Wall

Street Journal, the New York Times, the Financial Times, and a program on

the BBC. He has consulted and testified in the United States and Canadaon issues related to technology and intellectual property.

Alessandra Narciso is a doctoral researcher in the Department of Law of the

University of Leeds (UK). She holds a first degree in law from La Sapienza

University in Rome and an LLM degree from the University of Leicester

(UK). Her area of specialization is the legal protection of folklore.

Troy A. Paredes is an Associate Professor of Law at Washington University

School of Law in St. Louis. His work primarily focuses on corporate

governance and securities regulation, as well as on property theory and law

and economics. He has written numerous articles on a wide range of topics,

such as corporate governance and venture capital in developing countries,

hostile takeovers, executive compensation, CEO overconfidence, informa-

tion overload and decision theory, asset securitization, intellectual property

and bankruptcy, and hedge funds. Before joining Washington University,

Paredes was a corporate and regulatory attorney.

x Developments in the economics of copyright

Giovanni B. Ramello is Assistant Professor of Economics at the Universita

Carlo Cattaneo - LIUC, Italy. He received an education both in econom-

ics (PhD) and in computer sciences in Italy (Universita degli Studi di

Torino, Universita Bocconi di Milano) and in France (Universite Jean

Moulin - Lyon 3 and CNSM, Lyon). He has been research fellow at the

Fondazione Giovanni Agnelli, Torino (Italy), at the Banff Centre (Alberta,

Canada), and has served as an advisor for the EU Phare-Twinning Project

on the Harmonisation of Antitrust in the European Countries (topic: inter-

play between antitrust and intellectual property), for the Forum for the

Information Society of the Italian Presidency of the Ministers' Council, for

the Federation of the Italian Universities and in some Italian antitrust

cases. He is currently a member of the editorial board of the EuropeanJournal of Comparative Economics. His main research interests concern

industrial economics, antitrust, economic analysis of law and institutions,

intellectual property, information goods and knowledge production. Hehas published in these fields a number of papers in national and interna-

tional journals and books.

Pamela Samuelson is a Chancellor's Professor of Law and Information

Management at the University of California at Berkeley. She is a Director

of the Berkeley Center for Law and Technology, as well as an advisor to the

Samuelson High Technology Law and Public Policy Clinic at Boalt Hall.

She teaches courses on intellectual property, cyberlaw and information

policy. She has written and spoken extensively about the challenges that

new information technologies pose for traditional legal regimes, especially

for intellectual property law. She is co-author of the Software and Internet

Law casebook and of The law and economics of reverse engineering'

(jointly with Suzanne Scotchmer), published in 2002 in the Yale LawJournal and of 'Intellectual property rights in data?' (jointly with J.H.

Reichman) in 1997 in Vanderbilt Law Review, as well as sole author of a

large number of law review and computing journal articles. She is a Fellow

of the Association for Computing Machinery (ACM) and a ContributingEditor of Communications of the ACM. A 1971 graduate of the University

of Hawaii and a 1976 graduate of Yale Law School, Samuelson practiced

law as a litigation associate with the New York law firm Willkie Farr &Gallagher before turning to more academic pursuits. From 1981 throughJune 1996 she was a member of the faculty at the University of Pittsburgh

Law School, from which she visited at Columbia, Cornell, and Emory LawSchools. She has been a member of the Berkeley faculty since 1996.

Since 2002, she has been an Honorary Professor of the University of

Amsterdam.

Contributors xi

Arthur Snow is a Professor in the Department of Economics in the Terry

College of Business at the University of Georgia. He earned a PhD in eco-

nomics from the University of Wisconsin-Madison in 1979, and was

employed as Assistant Professor of economics at the University of Virginia

and at Georgetown University before joining the faculty at the University

of Georgia. His research on the economics of asymmetric information,

decision making under uncertainty, and on issues in public economics has

appeared in the Journal of Political Economy, the Journal of Public

Economics, the Journal ofRisk and Uncertainty, the International Economic

Review, the Economic Inquiry, and Economics Letters. He presided as pres-

ident of the Risk Theory Society at the 2004 Seminar, and currently serves

on the editorial board of the Geneva Papers on Risk and Insurance Theory.

Lisa N. Takeyama received her PhD in economics from Stanford University

under a National Science Foundation Fellowship. She currently holds a

Research Associate position in the department of economics at Amherst

College. Prior to coming to Amherst, she was on the faculty in the depart-

ment of economics at the University of Oregon. Included among her most

well-known publications on piracy and intellectual property are TheWelfare Implications of Unauthorized Reproduction of Intellectual

Property in the Presence of Demand Network Externalities' (Journal

of Industrial Economics) and 'The Intertemporal Consequences of

Unauthorized Reproduction of Intellectual Property' (Journal of Law and

Economics). She has also published other important work on a wide variety

of topics including durable goods monopoly, auctions and economic

growth in the Journal of Industrial Economics, the European Economic

Review, Economics of Innovation and New Technology, and Economics

Letters. Her article (with Eric B. Budish), 'Buy Prices in Online Auctions:

Irrationality on the Internet?' (Economics Letters) was listed by the pub-lisher as being among the top ten most requested articles for 2001 . In 1997,

she was named among the top 15 economics scholars at the Assistant

Professor rank at liberal arts colleges. She currently serves on the editorial

board of the Review ofEconomic Research on Copyright Issues. In 2003, she

was the local organizer for the second annual Congress of the Society for

Economic Research on Copyright Issues, held in Northampton, MA,USA.

Paul L.C. Torremans is the City Solicitors' Educational Trust Professor of

Intellectual Property Law at the School of Law of the University of

Nottingham (UK). He also holds an appointment as Professor of Private

International Law at the Faculty of Law of the University of Gent in

Belgium. He was awarded his PhD and LLM degrees by the University of

xii Developments in the economics of copyright

Leicester (UK) and he also holds law, notary law and law teaching degrees

from the University of Leuven (Belgium). He specializes in intellectual

property law and private international law and more specifically, in the

interaction between intellectual property law and private international law

and between intellectual property law and competition law. He has pub-lished various articles and books, including Holyoak and Torremans

Intellectual Property Law, (3rd edition, Butterworth, 2001; 4th edition,

Oxford University Press, 2005), and Intellectual Property and Private

International Law, with J.J. Fawcett (Clarendon Press, 1998). His latest

book is entitled Copyright and Human Rights, P. Torremans (ed.), (KluwerLaw International, 2004).

Ruth Towse is Reader in Cultural Industries at Erasmus University

Rotterdam. Her main area of expertise is in cultural economics with special

reference to the economics artists' labor markets and copyright in the cul-

tural industries. She has published widely in cultural economics and on the

economics of copyright, editing major collections in both fields, as well as

being the author of a number of books and articles in journals and in other

publications. She was Joint Editor of the Journal of Cultural Economics

from 1993-2002. She is President of the Society for Economic Research on

Copyright Issues (SERCI) (2004-6).

Patrick Waelbroeck is a research fellow from the National Science

Foundation (FNRS) in Belgium. He earned his PhD in economics from

Universite de Paris 1 (La Sorbonne) in December 2000 with highest honors.

He also holds an MA from Yale University and is a Fulbright alumnus. His

current research proposes both empirical and theoretical perspectives on

Internet piracy and technological protection in the music industry.

Richard Watt is Professor of Economic Theory at the Universidad

Autonoma de Madrid. After completing a Bachelor's and a Master's

degree in economics at Canterbury University in New Zealand, he earned

his doctorate in economic theory from the Universidad Autonoma de

Madrid in 1 990. His interest in the economics of copyright is manifested

in his book Copyright and Economic Theory: Friends or Foes! (Edward

Elgar, 2000). He is also actively researching the economics of risk bearingand risk sharing. He has published several papers in international journals

on this topic, as well as on the theory of oligopoly. He is currently the

General Secretary of the Society for Economic Research on CopyrightIssues (SERCI) and the managing editor of the Review of Economics on

Copyright Issues (RERCI).

Contributors xiii

Christopher S. Yoo is Associate Professor of Law at the Vanderbilt

University Law School. A graduate of Harvard College, the Anderson

School at UCLA, and the Northwestern University School of Law, his

research focuses primarily on how technological innovation and cutting-

edge economic theories are transforming the regulation of telecommunica-

tions and the electronic media. He is also pursuing research on the

application of imperfect competition models to copyright. His principal

work in these areas has been published in the New York University Law

Review, the Northwestern University Law Review, the Cornell Law Review,

the Georgetown Law Review, the Vanderbilt Law Review, the Southern

California Law Review, the Emory Law Journal, and the Yale Journal on

Regulation. Before joining the Vanderbilt law faculty, he clerked for Justice

Anthony M. Kennedy of the Supreme Court of the United States and was

associated with the DC law firm of Hogan & Hartson.

Preface

Lisa N. Takeyama

The Society for Economic Research on Copyright Issues (SERCI) was

established in 2001 as a response to the surging interest in copyright and the

expanding body of research applying economic analysis and theory to

copyright issues. As stated on the society's website (www.serci.org),

SERCI's objective is to actively promote and internationally distribute

quality academic research that relates economic theory to all aspects of

copyright and intellectual property of a cultural nature.

Since its inception, SERCI has held annual international congresses in

Spain, the USA, and Italy. From these congresses, two collections of papershave been published, including the present book. Additionally, SERCItakes great pride in the launch of its new journal entitled, Review ofEconomic Research on Copyright Issues (RERCI). Its inaugural issue has

already appeared, and RERCI will now appear twice yearly.

The first book-length collection of papers, The Economics of Copyright:

Developments in Research and Analysis (2003), edited by Wendy J. Gordonand Richard Watt and published by Edward Elgar, contains papers selected

from the inaugural SERCI congress held in Madrid in 2002. The present

book presents further developments in the economics of copyright. It con-

tains selections from the second annual SERCI congress held in

Northampton, Massachusetts, in 2003. Both congresses were richly fueled

by the productive interactions among the many economists, legal scholars,

and other copyright professionals present.

The chapters in this book, each of which has undergone peer-review and

revision, consider a wide variety of topics and issues in the economics of

copyright, including the role of economics in copyright law and policy,

peer-to-peer music file sharing, optimal fair use standards, the benefits of

copyright collectives, copyright and market entry, alternatives to copyright,

the impact of copyright on knowledge production, the proper balance

between copyright and competition law, and the application of systematic

principle to issues that arise at the periphery of intellectual property law.

The chapters also contain a diversity of methodologies employed - from

formal theoretical economic modeling to institutional and case study anal-

yses. Given the diversity of topics and analyses employed, this book should

xiv

Preface xv

be of interest not only to economists, but also to legal scholars and other

professionals interested in copyright issues and management.We would like to thank all of those involved in making this book pos-

sible, including, of course, our contributors, peer-reviewers, and congress

participants, all of whom contributed insightful ideas and suggestions that

helped make this a better book. We would also like to thank Richard Watt,General Secretary of SERCI, without whose vision and efforts SERCIwould not exist. We are also grateful to the Spanish Author's Society

(SGAE) for their continued support of, and belief in, SERCI. Many thanks

are also owed to Edward Elgar Publishing for their assistance in bringingthis volume to press.

Finally, I would like to thank my wonderful co-editors, Wendy Gordonand Ruth Towse, who provided an abundance of invaluable input on this

project and with whom I greatly enjoyed working.

General outline of the book

The book begins with a chapter by Pamela Samuelson, a legal scholar

famous for her pioneering work on the copyright issues raised by new infor-

mation technologies. In Chapter 1,she addresses what may be perhaps the

most fundamental question regarding economics and copyright: just what

role does economics have in copyright law and policy? Samuelson suggeststhat economics has not thus far been as influential in intellectual propertylaw and policymaking as it has been in other areas of economic regulation,

such as antitrust. She offers several possible explanations for this, includ-

ing for example, lack of economic expertise on the part of the relevant pol-

icymakers, and 'the tight nexus between the copyright industry and the

policymaking community'. Samuelson explains how economics does have

an important and useful role to play in copyright law and policy- touch-

ing also on ways economics can be misused - and offers several examplesand case studies.

The chapter by Samuelson, in addressing some of the institutional and

methodological difficulties that contribute to a law/economics divide,

serves to introduce the book as a whole. The primary focus of the book's

succeeding sections is the application of economic tools and analysis to

examine particular issues within copyright. They begin with consideringthe economics of copyright collectives.

In Chapter 2, Arthur Snow and Richard Watt offer a novel view of the

benefits of copyright collectives in their capacity to pool the risks of their

members. While it is well-known that collective administration of copyrightcan reduce transaction costs, the potential risk-sharing benefit of copyrightcollectives has not been previously considered. Using a formal analytical

model, Snow and Watt show that when the final market value of each

creator's work is uncertain, a copyright collective can improve the welfare

of its risk-averse members by alleviating individual risks. Snow and Watt

also point out that despite this potential risk-sharing benefit, the actual

allocation rules employed by copyright collectives in fact preclude this

benefit for their members.

In Chapter 3, Stan Liebowitz carefully examines the benefits and costs of

certain proposals that have been put forth as alternative solutions to the

problem of peer-to-peer music filesharing. Under these proposals, a copy-

right collective would be formed to administer a tax-and-subsidy system

xvi

General outline of the book xvii

whereby the record industry's traditional revenue sources would be

replaced (or supplemented) by revenue generated from taxation of ancil-

lary products (e.g., CDs, ISP services, etc.). Such a system could either

entirely replace the current copyright system or it could serve as a restitu-

tion mechanism to compensate copyright owners for losses induced byunauthorized copying. Some of the benefits of such a system mightinclude, for example, decriminalization of widespread illegal behavior,

enhanced incentives for the production of creative works, and increased

consumption. Liebowitz argues, however, that the costs of such a system,

including for example, those induced by tax distortions and a mispriced

royalty rate, as well as the potential for 'gaming' the system, make such pro-

posals (as his title indicates) 'a cure worse than the disease'.

Peer-to-peer music filesharing is also addressed by Anne Duchene andPatrick Waelbroeck in Chapter 4. However, in contrast to Chapter 3's

emphasis on proposals to resolve the problems induced by peer-to-peermusic filesharing, Chapter 4 is premised on the idea that firms may benefit

from peer-to-peer networks in their capacity to convey information to con-

sumers about new music cheaply, as compared to the more costly tradi-

tional mechanisms (e.g., advertising and promotions). Duchene andWaelbroeck's main thesis is that increasing copyright protection may have

significantly disparate welfare consequences, depending upon whether

information is conveyed to consumers via peer-to-peer networks ('demand-

pull') or via traditional mechanisms ('demand-push'). They develop a

model from which they derive welfare implications of increasing copy-

right's criminal penalties under both demand-pull and demand-push mech-anisms when firms can also employ their own additional technological

protection against piracy (e.g., digital rights management). They find,

among other things, that increasing criminal penalties benefits firms

employing demand-push mechanisms and induces them to increase their

level of technological protection, while the reverse is true for firms employ-

ing peer-to-peer networks. The effect of potential audience size on the

firm's preferred information transmission mechanism is considered as well.

In Chapter 5, Timothy Brennan addresses the issue of fair use, the

flexible doctrine employed by US law that allows behavior to occur - such

as some copying in educational, private, and journalistic contexts - that

would otherwise constitute prima facie copyright infringement. Brennan

examines the fair use doctrine in light of the diversity of policy goals that

have been proposed for its application. To this end, he develops a modelfrom which he derives optimal fair use standards under several different

policy objectives. Specifically, in addition to economic efficiency, he consid-

ers four possible alternate policy objectives: maximization of number of

works supplied, aggregate consumer surplus, gross benefit, and numbers of

xviii Developments in the economics of copyright

uses. By deriving ordered rankings for the optimal levels of fair use under

each of the above standards, Brennan shows how fair use might be adjusted'to address the breadth of policy goals that might be considered'.

In Chapter 6, Christopher Yoo criticizes the standard economic analysis

of copyright that typically assumes copyright turns creators into monopo-lists. Yoo argues that the well-known 'access-incentives' tradeoff is largely

driven by the assumed monopolization of the market. If one instead con-

siders the impact of copyright within a framework that allows for productdifferentiation and entry, the access-incentives tradeoff may be less

significant. By increasing producer appropriability and profit, increased

copyright protection can stimulate entry of competitors producing similar

works, resulting in lower prices, increased product variety, and increased

access. Yoo suggests that this approach also makes a broader set of policyinstruments potentially available for use, but disentangling the effects of

one from another can be quite complicated.In Chapter 7, Giovanni Ramello argues that the 'standard' economic

analysis of copyright is too narrowly focused on individual creation. In par-

ticular, economists have failed to consider the inherently social nature of

knowledge production, which takes place in what Ramello calls the 'social

creative commons'. While economists have recognized that copyrightrestricts access to output (i.e., to the copyrighted good), they have been

insufficiently sensitive to the ways in which copyright also restricts access

to inputs (i.e., to the creative commons and the production of knowledge).Ramello argues that restricting access to the creative commons and hence,

to knowledge sharing, reduces the future productive capacity of the crea-

tive commons, a consequence which Ramello calls an 'opposite tragedy of

the creative commons'. For these reasons, Ramello argues for a 'minimal-

ist' rather than a 'maximalist' approach to copyright.In Chapter 8, Alessandra Narciso and Paul Torremans examine the

interaction and proper balance between copyright and competition (anti-

trust) law. Their analysis largely focuses on the German IMS Health case,

a case that well illustrates the tension that exists between copyright and

competition law. Narciso and Torremans' main thesis is that the applica-tion of competition law to cases in which copyright exists should factor in

and avoid undermining the structural system originally put in place by

copyright. So, for example, in antitrust cases involving copyright, it may be

inappropriate to employ as precedent court decisions in non-copyright

cases, at least without explicit recognition of the special structure intention-

ally put into place by copyright. Narciso and Torremans apply this notion

to several other policy issues raised by the IMS Health case, with a focus

on the essential facilities doctrine.

In Chapter 9, Scott Kieff and Troy Paredes provide a strong argument

General outline of the book xix

for applying, both at the judicial and legislative levels, 'a basics approach'to issues in dispute that arise at 'the periphery of intellectual property law'.

Too often, they argue, such disputed matters are resolved in a piecemealfashion by creating specialized approaches and unique doctrines. This

approach, they argue, 'selectively emphasizes or alternatively ignores par-

ticular features of various legal disciplines in crafting specialized doctrines

for IP'. In contrast, under their competing 'basics approach,' Kieff and

Paredes propose that such matters would be best resolved by applying the

basics of the relevant substantive law, such as intellectual property law,

antitrust law, and property law. Although Kieffand Paredes focus on patent

law, their analysis applies equally to copyright. For example, they arguethat the anti-circumvention provisions of the Digital Millennium

Copyright Act are inconsistent with the 'basics approach' since the trans-

actions it regulates are better governed under existing contract law or the

IP law of indirect infringement.

As this brief review of the contents of this volume shows, both econo-

mists and legal scholars are applying their analytical skills to a broad rangeof theoretical and applied areas in the economics of copyright. We hopethat the volume will stimulate further research and debate in this importantarea.

1 . Should economics play a role in

copyright law and policy?

Pamela Samuelson

1.1 INTRODUCTION

The principal justification for intellectual property (IP) laws in the Anglo-American tradition is economic. 1 Without a grant of exclusive rights,

authors and inventors would have too little incentive to invest in socially

beneficial innovations for this simple reason: developing the first embod-iment of an innovation generally requires very costly investments; subse-

quent copies are generally far less costly to make, and are often trivially

cheap and easy.2 If creators cannot prevent others from selling products

embodying their innovations - which copyists can offer at a lower price if

able to free-ride on the innovator's first copy costs - innovators will not

be able to recoup their research and development (R&D) costs and justify

further investments in innovations. 3Although the non-excludable, non-

rival character of intellectual creations resembles public goods,4

it is

socially desirable for private actors, rather than the government, to makethe investments to overcome this public goods problem.

5 Intellectual

property rights are an ingenious device for addressing this problem: the

government grants rights to qualifying innovators, and the rights provideincentives to individuals to invest in innovation because the rights

granted provide assurance that innovators can recoup their R&D costs if

their intellectual creations prove to be valued by the public.6Original

works of authorship are particularly vulnerable to market-destructive

copying, which is why few question the economic wisdom of a copyright

regime.7

Intellectual property would, for this reason, seem to be 'a natural field

for economic analysis of law.' 8It would, hence, be logical for economic

analysis to have considerable influence in shaping copyright and other intel-

lectual property laws.9 Yet, as compared with other fields of economic regu-

lation, particularly antitrust law, economics has had very little influence

thus far in the intellectual property law and policymaking process. This

chapter considers why this might be so in the field of copyright law and

2 Developments in the economics of copyright

whether economics might become more influential in this field of law and

policy in the future.

The lack of influence is not attributable to inattention from the scholarly

community. Law and economics scholars have produced a substantial and

growing literature that analyzes many aspects of copyright law. 10 Amongthe copyright rules that are widely viewed as economically sound are: the

idea/expression distinction (that is, copyright protects only an author's

expression, not her ideas),11 the independent creation defense,

12 the deriva-

tive work right (that is, the right to control adaptations of the work, such

as a motion picture version of a novel),13 the separation of the ownership

of a copy of a work from ownership of the copyright in it,14 the ability of

authors to assign 'slices' of their exclusive rights,15 limits on the duration of

copyright,16 and the work made for hire rule (which treats employers as the

'author' of works by employees performed within the scope of employ-

ment).17 The economics of copyright collectives and of compulsory licens-

ing has also garnered scholarly attention. 18 Cultural economics, a subfield

of economics, has studied the market for art and cultural products, such as

opera and theatre, and offered insights about policies needed to induce

investment in these cultural forms. 19 One bold work considers whether eco-

nomics is friend or foe of copyright,20 but rare is the work that attempts to

analyze the economics of copyright as a whole.

Among the most studied subjects has been the economics of the fair use

doctrine of US copyright law. 21 Economists and economically minded legal

scholars often view uses of copyrighted works as 'fair' if high transactions

costs or other factors have impeded the effective establishment of a market

for clearing rights.22

Particularly influential has been Wendy Gordon's

analysis of the fair use issues in the Sony Betamax case. Gordon argued that

use of Betamax machines to tape broadcast television programs for later

viewing should be fair because the costs of negotiating rights clearances

would greatly exceed fees collectable for this kind of socially beneficial

use. 23 Because the economic effects of a challenged use are almost alwaysof central importance in judging whether the use will be deemed non-

infringing as a fair use, economics is especially useful in providing guidanceon this copyright issue.

24

Economists and economically minded legal scholars have questioned or

criticized certain copyright rules as economically unsound. For example, it

is unnecessary for personal letters and diaries to be automatically qualified

for copyright protection, given that these works would be created without

the economic incentives of copyright.25

Special exemptions from copyright

liability for certain events organized by agricultural cooperatives and vet-

erans' groups seem to be the product of special interest lobbying, rather

than rigorous economic analysis.26 Extension of the duration of copyrights

Should economics play a role in copyright law andpolicy? 3

in existing works27 and so-called 'restoration' of foreign copyrights that had

long been in the public domain for failure to comply with US formalities28

(e.g., putting copyright notices on published copies of the work) are also

economically questionable.29

As interesting and provocative as the economics of copyright literature is,

even its most ardent fans would have to admit that economics has rarely

played a significant role in the copyright law and policymaking process.

Section 1 .2 will consider various reasons why economics has not been more

influential in copyright up until now. Section 1 .3 will discuss the economics

of extending the terms of existing copyrights and the Supreme Court's deci-

sion in the Eldred vs. Ashcroft case. 30 Although an initial reading of the

Supreme Court's decision might suggest that economics will have little influ-

ence in future court decisions interpreting copyright law, Section 1 .3 explains

why this interpretation of the Eldred decision is incorrect. Section 1 .4 will

discuss particular venues in which economics would be most useful in the

copyright law and policymaking process going forward. Legislators should

(and occasionally do) seek independent economic expertise when consider-

ing copyright and related intellectual property proposals. Courts mayfind economic analysis a useful input to sound decision-making, especially

in cases involving new technology issues unanticipated when legislatures

adopted copyright laws. Section 1.5 will consider some reasons why copy-

right professionals might resist the incursion of economics into the copy-

right policy process, and why economics should nonetheless play a more

important role in copyright policymaking in the future.

1 .2 WHY HAS ECONOMICS HAD SO LITTLEIMPACT ON COPYRIGHT SO FAR?

One salient factor explaining the low impact of economic analysis on copy-

right law and policy is a lack of economic expertise in the relevant policy-

making community. Many copyright professionals have backgrounds in

fields other than economics and are more inclined to embrace a romantic

conception about art and literature, and the people and firms who create and

commercialize copyrighted works that would disincline them to look to eco-

nomics for guidance about how the law should be crafted. 31 Unlike the US

Department of Justice Antitrust Division, the Federal Trade Commission,and the Federal Communications Commission (FCC), for example, all of

whom regularly participate in policymaking as to other economic regula-

tions, the US Copyright Office (as well as its patent and trademark coun-

terpart) has no in-house economic experts who might contribute insights

about economic effects of various policy proposals.32


For path-dependent historical reasons, copyright and other intellectual

property policy matters in the US have generally been considered by sub-

committees of the House and Senate Judiciary Committees, rather than byCommerce Committees, from which economic regulations often emerge.

These subcommittees have relied heavily on industry witnesses when

considering copyright and other intellectual property-related legislation.

Copyright industry groups, in fact, often write the laws that the legislature

enacts. 33 A revolving door enables copyright experts to move from the

legislature or legislative committee staff positions to law firms represent-

ing major copyright firms to industry associations to government agencies

responsible for framing copyright law and policy.34 This contributes to

insularity in copyright policy analysis.

The US is not alone in its systematic, if perhaps unconscious, exclusion

of or inattention to economic analysis in the copyright policymaking

process. Nations whose legal rules are grounded in the natural right of

authors in their works,35

especially those that affirm that authors have

moral rights in their works, have less reason to perceive economics as a

useful input in the policymaking process than the US which at least histor-

ically conceptualized copyright in utilitarian terms.

Although the tight nexus between the copyright industry and the policy-

making community is the main reason economics has thus far had so little

influence on copyright law and policy, it is not the only factor at work.

An impediment to influence is that economists who study copyright issues

often speak in a language inaccessible to policymakers and copyright pro-

fessionals. Few copyright professionals are able, for example, to penetrate

the mathematical expressions that often pervade the academic economic

literature. If economists wish for their work to inform the copyright policy

process, they will have to learn to translate their insights into a more ver-

nacular form. They may also need to publish policy-relevant work in venues

where policymakers would be likely to read it.

Cultural differences are also a factor. Lawyers and policymakers view

some problems that deeply engage economists (e.g., the optimal duration

of a patent or of copyright protection for computer software36) as a waste

of time. Other issues of interest to economists (e.g., network externalities)

may have policy implications that are not readily apparent to copyright

professionals.37 A legal and policy audience may find incomplete an eco-

nomic analysis whose policy implications are not spelled out. If economists

want their work to inform the copyright policymaking process, they will

have to invest the extra time necessary to articulate the policy implications

that flow from their analyses.

The standard economic practice of articulating simplifying assumptionsbefore constructing an economic analysis does not mesh well with the


mindset of lawyers and policymakers either. If the whole edifice of an

analysis depends on assumptions A, B, C, D and E being true and at least

one of them is questionable, lawyers and other policy analysts may be

inclined to discredit the analysis as a house of cards.

Influencing copyright law and policy will also require economists to give

this field somewhat greater attention. Economic research on intellectual

property issues has tended to focus on subjects for which funding from

government or foundation grants is available. Until quite recently, grant-

makers rarely supported research on copyright issues. Far more economic

research has been done on patent issues, mainly because of their significance

for the predominant industrial economy of the nineteenth and twentieth

centuries, but also because more funding has historically been available to

study the economics of patents. The emergence of an information-based

economy in the late twentieth century has attracted the attention of econo-

mists and economically-informed lawyers to exploring the economics of

copyright.38

Also an impediment to influence is the plain fact that economics will not

always yield an unimpeachable policy prescription for copyright issues.

This is partly due to the complex intricacies of the many industries that rely

upon or intersect with copyright law. Since copyright law has many provi-

sions of general applicability that affect a wide range of industries - from

book publishing to architecture to computer software to theatrical

performances, just to name a few - it may be very difficult, if not impossi-

ble, to assess with any precision the economic effects likely to flow from

adoption of particular copyright rules, let alone the economics of the copy-

right regime as a whole. 39

It is to be expected, moreover, that on many issues, economists will differ

in their analyses of the same issue. In A&M Records, Inc. vs. Napster, Inc.,

for example, economic experts offered widely varying analyses of the effects

of peer-to-peer file sharing on the market for digital music.40 Some econo-

mists considered file sharing to be benign or beneficial, while others asserted

it was harmful to the market for sound recordings.41

Policymakers will

understandably be reluctant to put much weight on economic expertise

if equally qualified experts offer irreconcilably conflicting analyses of the

same phenomenon.Another barrier to influence is the understandable reluctance of copy-

right lawyers and policy professionals to acknowledge the public choice

problems presented by the current copyright law and policy-making

process.42 These problems are well-documented in the scholarly literature.43

Major copyright firms are well-organized and well-funded. They typically

have a common interest in getting stronger legal rules from the legislature.

Hence, it may be a sound investment for them to lobby to achieve the


concentrated benefit a legislature can grant them. The public may ultim-

ately have to pay higher costs if copyright lobbyists are successful, but these

costs are diffuse and distributed over a broad base of people and firms.

Collective action problems make it difficult for parties that will be nega-

tively affected by higher protection rules to organize effective resistance to

copyright industry lobbying. This mix of concentrated benefits and dis-

tributed costs is likely to yield the best laws money can buy. The CopyrightTerm Extension Act (CTEA),

44 the Digital Millennium Copyright Act

(DMCA) anti-circumvention provisions,45 and state 'super-DMCA laws46

are three widely recognized examples from the US experience.

1 .3 WHAT EFFECT WILL ELDRED VS. ASHCROFTHAVE ON THE FUTURE OF ECONOMICS INCOPYRIGHT?

A stark example of the battle between economic and non-economic rea-

soning in the copyright law and policy process is evident in the Eldred vs.

Ashcroft case decided in 2003 by the US Supreme Court.47 Eric Eldred, an

online publisher of public domain works, challenged the CTEA, which

extended copyright terms an additional twenty years, based on an eco-

nomic theory of the Intellectual Property Clause of the US Constitution.48

Article I, Sec. 8, Cl. 8 gives Congress power 'to promote the progress of

Science and useful Arts, by securing to Authors and Inventors for limited

times the exclusive Right in their respective Writings and Discoveries.'

Boiled down to its essence, Eldred's theory was that this clause prohibited

Congress from granting exclusive rights to authors without the quid pro

quo of a 'progress-promoting' act (i.e., authorship of a newly created orig-

inal work) in return.49 Stated more colloquially, Congress violated the

Constitution by extending the term of existing copyrights for an additional

twenty years because the public got nothing in return for that grant. The

CTEA was purely a windfall to existing copyright owners. Such a quidpro

nihilo, Eldred argued, failed to satisfy constitutional standards. 50

Seventeen economists -including five Nobel Prize winners -

signed a

friend-of-the-court brief to the US Supreme Court in support of Eldred's

challenge to the CTEA. 51 The brief made the obvious point that the CTEAcould not provide incentives for the creation of works already in existence. 52

It went on to consider the deadweight loss that a longer term of above-cost

pricing would bring, as well as the substantial transactions costs the CTEAwould impose on subsequent users. 53

Tracking down copyright owners and

negotiating with them is costly, even if the owners do not insist on paymentsfor reproducing or distributing copies of their works. CTEA impedes


public access to many works that are no longer commercially exploited, but

that may still be of interest, as well as preventing the creation of many newderivative works. 54 The economists' brief also pointed out that life of the

author plus 70 years was virtually equivalent to the perpetual term of copy-

right that the Constitution forbids. 55

The Supreme Court rejected Eldred's challenge to the CTEA. Congresshad a rational basis, the Court decided, for believing that CTEA would lead

to more investments in distributing and preserving existing works. 56 The

Constitution gave to Congress the responsibility of crafting copyright legis-

lation, and even if it exercised this power unwisely, the Court was reluctant

to second-guess legislative decisions. 57Only Justice Breyer would have

struck down theCTEA on an economic interpretation of the Constitution.58

The Eldred opinion is starkly different from many of the Court's rulings

on economic regulations. Economic analysis is generally quite influential

when the Supreme Court assesses economic regulations, especially amongthe more conservative members of the Court. 59 Given how often the Court

has emphasized the economic incentive rationale for the existence of copy-

right law,60

it is striking how little attention was paid to economics in Eldred.

Justice Ginsburg, who authored the majority opinion, is not, of course, the

most economically minded of the Justices. But the de-emphasis on eco-

nomics in Eldred has other roots.

There is a deep rift within the Court about constitutional powers of

Congress and about how much deference to give to legislation emanatingfrom it.

61 Eldred's lawyers drew upon several Supreme Court precedents

interpreting the constitutional powers of Congress in a restrictive manner

and establishing rigorous standards for judging the constitutionality of

certain kinds of legislation.62 Further limitations on Congress' power under

Article I, sec. 8, cl. 8 were, Eldred's lawyers argued, consistent with these

recent precedents,63 as well as with the Court's prior rulings under the

Intellectual Property clause.64 Eldred's lawyers hoped the analogy to these

other restrictive rulings would woo the more conservative members of the

Court to further restrictions on Congress' constitutional powers in Eldred.

Eldred's lawyers used a different strategy to persuade Justices Breyer and

Stevens (who had dissented in the earlier challenges to Congress' power) to

support Eldred's cause. 65

There was, in short, a lot going on in the Eldred case that had nothingwhatever to do with copyright law, but a good deal to do with the Court's

internal debate on the scope of Congress' powers under the Constitution.

Other factors were also at play. Most significant was the indisputable fact

that Congress had extended the term of existing copyrights numerous times

before: in 1790, 1831, 1909, and 1976, as well as several times leading up to

enactment of the Copyright Act of 1976.66 Although Eldred's lawyers


developed an elaborate chart to distinguish the CTEA from other exten-

sions,67 this chart may have backfired because it underscored the frequency

of these enactments. To comprehend the scope of Congress' power [to

extend terms of existing copyrights] under the Copyright Clause,' said the

Court,'

"a page of history is worth a volume of logic."'68

The Court was also understandably concerned about the implications of

a ruling in Eldred's favor. If the CTEA was invalid, the retroactive term

extensions in the Copyright Act of 1976 would almost certainly be chal-

lenged, and it was not immediately evident how the Court could find a prin-

cipled basis for distinguishing this extension from the CTEA. 69 Had the

Court acceded to the heightened scrutiny for which Eldred argued,70

it

might have opened the floodgates of constitutional challenges to manyforms of intellectual property legislation, not just to copyright rules. 71

Finally, Eldred was willing to accept that life of the author plus 70 years

was a 'limited time' under the Constitution applied prospectively.72 In view

of this, the Court thought that life plus 70 years must also be sufficiently

limited applied retroactively.

Taking into account the range of explanatory factors, the Eldred decision

may be less of a setback for a significant role for economic analysis in copy-

right law and policy than a cursory glance might suggest. The Court was

not willing to interpret the Constitution as mandating economically sound

results, but this does not mean that the Court believes that copyright law

and policy should be uninformed by economic analysis. The next section

will consider several contexts in which economic analysis has been and is

likely to be a useful input into the policymaking process for copyright and

related forms of intellectual property law.

1 .4 WHEN MIGHT ECONOMICS BE USEFUL?

When legislatures consider proposals to add a new subject matter to copy-

right or to adopt new copyright-like forms of legal protection, they should

(and occasionally do) seek input from impartial economic experts about

whether such legislation is necessary to address a market failure that is

deterring socially optimal investments in particular classes of innovations

or information resources. Economic expertise, for example, had some influ-

ence in the US when the National Commission on New Technological Uses

of Copyrighted Works (CONTU) made economic as well as doctrinal

arguments urging that copyright law be used to protect computer pro-

grams.73 The European Union too relied on economic analysis in crafting

appropriate copyright rules for computer programs.74 Owing in substantial

part to the decisions of the US and EU to protect programs by copyright,


this rule became an international standard in 1 994 after member states of

the World Trade Organization (WTO) acceded to the Agreement on Trade-

Related Aspects of Intellectual Property Rights (TRIPS), which required

copyright protection for programs.75

The US Congress has a more mixed record in relying on economic

analysis when considering sui generis (of its own kind) forms of legal pro-

tection for other innovations. It relied heavily on economic argumentswhen adopting in 1984 a new law to protect original layouts of circuits in

semiconductor chips.76However, it adopted a copyright-like form of legal

protection for vessel hulls and other boat parts in 1998 without serious

economic evidence or expertise to support it.77

Yet, economic analysis has

had some cautionary influence on Congress in the debate about whether

to adopt an equivalent sui generis right in the data in databases such as the

EU adopted in 1996. 78

The European Commission asserted a market failure rationale to justify

the issuance of a Directive to member states of the EU requiring them to

adopt a sui generis form of intellectual property protection for controlling

extraction and reuse of data in databases. 79 Without such a law, the theory

went, there would be too few incentives to invest in database develop-

ment. 80 The new law confers on those who invest substantial resources in

database development with 15 years of exclusive rights to control extrac-

tion and reuse of all or substantial parts of the databases (renewable for

additional 15-year terms so long as substantial investments continue).81

Although the Commission invoked an economic rationale for this law, it

neglected to provide any empirical evidence of market failure or rigorous

economic analysis of the regime proposed to cure the perceived failure. Hadthe Commission been more rigorous in its use of economics, the EuropeanParliament might have realized there was less of a need for this sui generis

law than the Commission posited, and might have adopted a somewhat

narrower form of legal protection for data in databases. 82

Economic input should also be sought when copyright laws are

amended. Had the US Congress, for example, sought impartial advice from

economists about the effects of copyright term extension before enacting

the CTEA, it is conceivable that economic arguments against its adoptionwould have persuaded some in Congress to oppose it.

83 Economic analysis

might also have been useful in assessing the anti-circumvention provisions

of the Digital Millennium Copyright Act (DMCA), which have been

heavily criticized as overbroad and anti-competitive.84

Copyright rules are also sometimes promulgated in the context of agency

rule-making, and economic analysis should be a useful input to decision-

making in this venue as well. For example, the Copyright Office is required

to make determinations about claims for allocation of compulsory license


revenues to particular copyright owners. 85 Economic analysis would be

helpful in making such assessments. Proposals for new compulsory licenses

or revisions to existing compulsory licenses should likewise be scrutinized

for economic soundness. 86 The US Copyright Office has been given powerto conduct rule-makings every three years to consider whether to exemptcertain classes of works or uses of works from anti-circumvention rules. 87

This should be a forum in which the Office might be educated about the eco-

nomic effects of broad anti-circumvention rules.

Once rules have been legislated or issued by an appropriate agency, eco-

nomics may sometimes have a useful role to play, and sometimes not, in

how the law is interpreted thereafter. If a legislature has adopted an open-ended rule (e.g., that copyright protects an author's expression, not his or

her ideas or methods), economic analysis may be a useful tool in interpret-

ing the law in an economically sound way. However, if the legislature has

precisely crafted certain rules that prove to be economically harmful, it maybe more appropriate to invoke economic analysis to support proposedamendments to the law than to interpret the rules in litigation.

An instructive example of the pros and cons of economic analysis in

the interpretation of open-ended copyright rules can be found in the USjudicial experience interpreting copyright law as applied to computer

programs.88 Whelan Associates vs. Jaslow Dental Labs., Inc. was the first

American appellate decision to consider whether the 'structure, sequence,

and organization' ('SSO') of computer programs could be protected by

copyright law. 89 Whelan relied upon both doctrinal and economic groundsin ruling that SSO was copyright-protectable.

The doctrinal analysis relied on this syllogism: computer programs are

considered to be literary works under US copyright law, and since the struc-

ture, sequence, and organization of literary works are generally protectable

by copyright law, the structure, sequence, and organization of programsshould be protected by copyright law as well.90

Complementing this doctrinal analysis was an economic argument that

focused on the need for software developers to have sufficient protec-

tion to recoup development costs and on the locus of value in computer

programs:

By far the larger portion of the expense and difficulty in creating computer pro-

grams is attributable to the development of the structure and logic of the

program, and to debugging, documentation and maintenance, rather than to the

coding. See Frank, Critical Issues in Software 22 (1983) (only 20% of the cost of

program development goes into coding); Zelkowitz, Perspective on Software

Engineering, 10 Computing Surveys 197-216 (June, 1978). See also InfoWorld,Nov. 11, 1985 at 13 ('the "look and feel" of a computer software product often

involves much more creativity and often is of greater commercial value than the

Should economics play a role in copyright law andpolicy? 1 1

program code which implements the product . . .'). The evidence in this case is

that Ms. Whelan spent a tremendous amount of time studying Jaslow Labs,

organizing the modules and subroutines for the Dentalab program, and workingout the data arrangements, and a comparatively small amount of time actually

coding the Dentalab program.91

The Third Circuit took seriously a friend-of-the-court brief submitted bya software industry organization which argued that if copyright protection

did not extend to the SSO of a program, the software industry would be

jeopardized because the law would provide too little protection to induce

an optimal level of investment in development of computer programs.92

The Third Circuit directly responded to this plea: 'The rule proposed here,

which allows copyright protection beyond the literal computer code, would

provide the proper incentive for programmers by protecting their most

valuable efforts, while not giving them a stranglehold over the developmentof new computer devices that accomplish the same end.'93

Between 1986 and 1992, the Whelan decision was influential in subse-

quent US cases, both in its doctrinal and economic reasoning.94 The first

decision that challenged Whelarfs hegemony was Computer Associates vs.

Altai, Inc. in 1992.95 On one important point, the Second Circuit Court

of Appeals agreed with Whelan: the structure, sequence and organization

of a program could, in an appropriate case, be protected by copyright

law.96 Unlike Whelan, which regarded program structure as exempt from

infringement only when there was essentially no other way to structure the

program,97 the court in Altai reasoned that similarities in the structure of

two programs might, however, be due to functional constraints such as the

need to develop a program that would interoperate with another program,

efficiency considerations, or use of the same standard programming tech-

niques, none of which was protected by copyright law.98

The court in Altai directed that these and other unprotectable elements

of programs be 'filtered out' before infringement analysis began99 to ensure

compliance with section 102(b) of the US copyright statute which states

that '[i]n no case shall copyright protection for an original work of author-

ship extend to any idea, procedure, process, system, method of operation,

concept, principle, or discovery, regardless of the form in which it is

described, explained, illustrated, or embodied in such work.' 100 The court

ultimately ruled that some structural similarities in Altai were attribut-

able to the fact that both Computer Associates (CA) and Altai were

developing programs to interoperate with the same three IBM operating

system programs, and other structural similarities were to be expected in

programs of that kind. 101Hence, it affirmed the lower court's finding of

noninfringement.


As in Whelan, CA and its amid predicted dire economic consequencesfor the software industry if courts did not provide strong protection to

program structure. However, the Second Circuit took a different view:

[The Supreme Court's decision in] Feist teaches that substantial effort alone

cannot confer copyright status on an otherwise uncopyrightable work. . . .

[Djespite the fact that significant labor and expense often goes into computerprogram flow-charting and debugging, that process does not always result in

inherently protectable expression. Thus, Feist implicitly undercuts the Whelan

rationale, 'which allow[ed] copyright protection beyond the literal computercode ... [in order to] provide the proper incentive for programmers by protect-

ing their most valuable efforts. . . .' [citation omitted]. We note that Whelan wasdecided prior to Feist when the 'sweat of the brow' doctrine still had vitality. In

view of the Supreme Court's recent holding, however, we must reject the legal

basis of CA's disincentive argument.102

The Second Circuit went on to say:

[W]e are unpersuaded that the test we approve today will lead to the dire conse-

quences for the computer program industry that plaintiff and some amici

predict. To the contrary, serious students of the industry have been highly crit-

ical of the sweeping scope of copyright protection engendered by the Whelan

rule, in that it 'enables first comers to "lock up" basic programming techniquesas implemented in programs to perform particular tasks.' 103

The Second Circuit warned that if courts heeded purely economic argu-

ments for broadening the scope of copyright protection for computer pro-

grams, this would impair the integrity of copyright law:

While incentive based arguments in favor of broad copyright protection are

perhaps attractive from a pure policy perspective, . . . ultimately, they have a cor-

rosive effect on certain fundamental tenets of copyright doctrine. If the test wehave outlined results in narrowing the scope of protection, as we expect it will,

that result flows from applying, in accordance with Congressional intent, long-

standing principles of copyright law to computer programs.104

The Second Circuit pointed out that patent protection might be a moresuitable way to protecting some program innovations than copyright.

105 If

copyright proved to be too 'thin' to provide proper incentives to program

developers, this was a matter for Congress, not the courts, to consider. 106

Altai holds that copyright protection is unavailable to elements of pro-

grams necessary for interoperation with other programs (that is, program

interfaces).107

Judged purely in terms of the creativity and judgment

required to design them, program interfaces might initially seem to be 'ori-

ginal' enough to be protectable expression as a matter of copyright law. 108


However, once developed, program interfaces unquestionably constrain the

design choices of subsequent programmers seeking to develop software

capable of successfully interacting with an existing program.109

The competition policy significance of copyright protection for inter-

faces was recognized in Europe in part because the Competition Policy

Directorate of the European Commission had taken action against IBMarising from its practice of changing interfaces in a manner that had exclu-

sionary impacts on European developers of computer peripherals.110 If

European software developers were going to be able to compete with USsoftware in the world market, these developers would need to be able to

use interface information from American programs. The EC Competition

Policy Directorate intervened in negotiations surrounding the drafting of

a directive on the legal protection for computer software that resulted in

these economic considerations being brought to bear on the scope of

copyright protection for software in Europe.111 The US and EU consen-

sus on limiting copyright to enable interoperability influenced other

nations as well. 112

Sega Enterprises, Ltd. vs. Accolade, Inc. 113 addressed a related interoper-

ability issue as to whether programmers could lawfully decompile or disas-

semble other firms' programs in order to get access to information needed

to develop an interoperable program. This was a dicey issue for copyrightlaw because the decompilation and disassembly process inevitably requires

the making of a number of copies of the target program.114 Such copies

arguably run afoul of the exclusive right that copyright law confers onauthors to control the reproduction of their works in copies.

115 Yet unless

copyright law recognized at least a limited right to decompile or disassem-

ble programs, the decision not to protect interface information would be

significantly undermined.116

Accolade relied on the fair use defense of US copyright law in respond-

ing to Sega's charge of infringement.117 Courts consider four factors in

determining whether a use is fair: the purpose of the defendant's activ-

ities, the nature of the copyrighted work, the amount and substantiality

of the use, and the effects of the use on the market for the copyrightedwork. 118 Accolade persuaded the court that decompilation was the only

way Accolade could obtain access to the unprotectable interface informa-

tion in Sega's game programs (apart from entering into an unacceptable

licensing agreement) and that accessing and reusing this information

had made it possible for Accolade to develop new non-infringing works

to compete in the market with Sega's works. The court consequentlydecided that intermediate copying of programs for the purpose of achiev-

ing interoperability was fair use. 119 The court relied on economic reason-

ing in support of the ruling, although its analysis is not as systematically


economic as some of the law review commentary reaching the same legal

conclusion. 120

Both Altai and Sega vs. Accolade have been widely accepted as sound

rules in the US and abroad,121 and the software industry has continued to

flourish in the decade after these decisions were rendered. 122 Two lessons

from the software copyright caselaw are, first, that courts should be waryof unsubstantiated economic claims that without broad protection, the

affected industry will be severely harmed, and second, that considering evi-

dence of economic impacts of the court's ruling in particular contexts can

assist judges in making sound decisions in copyright cases. While software

law has now achieved a stability it lacked in the 1980s and early 1990s, chal-

lenges posed by new technologies for the application of copyright law to

digital works continue to be troublesome. Perhaps the most potent examplehas been the entertainment industry lawsuits against the developers of

peer-to-peer file sharing technologies.123 While economic analysis has been

of some assistance to judges forced to grapple with the thorny legal issues

these cases present,124

it may be necessary for legislatures to address what

standards ought to apply in regulating technologies with infringing and

non-infringing uses. 125

It is worth noting that the Canadian Supreme Court recently considered

another new technology issue that was certainly not anticipated when the

Canadian legislature enacted its copyright law. The Court concluded that

the economic rights of an artist were not infringed when a purchaser of

copies of his work transferred the copyrighted image from paper to canvas

and then resold the copies.126

Although not invoking economic analysis as

such, the care with which the Canadian Court considered the economic

impacts of the challenged use was impressive and bodes well for further

uses of economic arguments in copyright cases in Canada.

1.5 CONCLUSION

Copyright is an important form of economic regulation. This chapter has

considered several reasons why economic analysis has had relatively little

effect on copyright law and policy thus far. It has also argued that eco-

nomics has much to offer in assisting policymakers and judges in crafting

sound copyright law and policy.

There are several reasons why courts and policymakers may resist the

incursion of economics into the copyright law and policymaking process

and in the interpretation of copyright law. As noted in Section 1 .2, manyin the copyright field, including policymakers, have little or no economic

expertise and little inclination to seek it out. Second, copyright industry


professionals have been successful in getting much if not all of what theywant without the aid of economics, so they may perceive little need for

this analytical tool. Third, copyright industry professionals may be waryof embracing economics because although it may sometimes provide

support for stronger protection rules, it may not always cut in their favor.

They may resist embracing a tool that might provide grounds for policy

outcomes they disfavor. Fourth, until some major participant in the copy-

right policy-making process decides that economic analysis would be valu-

able as input to the policy process, the inertia of the current situation is

likely to persist.

Even so, I predict that economic analysis will have greater impact on

copyright in the future. A new generation of scholars is building up a rich

corpus of policy-relevant copyright research and is articulating the policy-

relevant implications of this research. 127 There is now a Society for

Economic Research on Copyright Issues (SERCI), which hosts an annual

international conference and publishes edited volumes of research on the

economics of copyright and a journal, the Review ofEconomic Research on

Copyright Issues (RERCI).128 The increasing importance of copyright

industries to the US and world economy should make policymakers more

receptive to insights that economics can provide.129

Moreover, the inclusion

of copyright and other intellectual property rights within the framework

for promoting world trade by virtue of adoption of the TRIPS Agreementas part of the agreements establishing the World Trade Organizationshould lead to greater use of economic learning to shape global intellectual

property norms.13 This should lead to greater reliance on economic expert-

ise at the national legislative and policymaking process.

Perhaps it is only a matter of time before the US Copyright Office andsister agencies in other countries will hire not just a resident economist, but

a group of them, one of which might be designated a Chief Economist,and give serious consideration to economic effects before endorsing par-ticular legislative initiatives. Economic principles can contribute to evolv-

ing our understanding of the access/incentives tradeoff so that copyrightwill continue - or perhaps return to - its long-standing tradition of pro-

moting the progress of science and the useful arts to the benefit of

humankind.

ACKNOWLEDGMENT

I wish to thank Wendy Gordon and Richard Watt for inviting me to give a keynote address at

the annual conference of the Society for Economic Research on Copyright Issues in

Northampton, MA in June 2003.

1 6 Developments in the economics of copyright

NOTES

1 . See, e.g., Paul Goldstein, Copyright's Highway: From Gutenberg to the Celestial Jukebox,168-69 (1994); Mark A. Lemley, The Economics of Improvement in Intellectual PropertyLaw, 75 Tex. L. Rev. 989, 1074-76 (1997). Although 'original' works of authorship are

not necessarily 'novel' in a patent sense, I am using the terms 'innovators' and 'innova-

tion' to encompass the works of both authors and inventors.

2. See, e.g., Wendy J. Gordon, Asymmetric Market Failure and Prisoner's Dilemma in

Intellectual Property, 17 U. Dayton L. Rev. 853 (1992).

3. See, e.g., Pamela Samuelson, Randall Davis, Mitchell D. Kapor, and J.H. Reichman,A Manifesto Concerning the Legal Protection of Computer Programs, 94 Colum. L. Rev.

2308 (1 994) (making this sort of economic argument for legal protection against cloningof computer program innovations).

4. See, e.g., Wendy J. Gordon, Fair Use as Market Failure: A Structural and Economic

Analysis of the Betamax Case and its Predecessors, 82 Colum. L. Rev. 1600, 1610-11

(1982).

5. Id. at 161 1-12.

6. See, e.g., J.H. Reichman, Electronic Information Tools - The Outer Edge of WorldIntellectual Property Law, 24 Int'l Rev. Indus. Prop. & Copyright L. 446 (1993).

7. See, e.g., William M. Landes and Richard A. Posner, An Economic Analysis ofCopyright Law, 18 J. Legal Stud. 325, 333-44 (1989) (presenting a formal modelof copyright). But see Stephen Breyer, The Uneasy Case for Copyright: A Study ofCopyright in Books, Photocopies, and Computer Programs, 84 Harv. L. Rev. 281 (1970)

(questioning the economic need for copyright). See also Barry W. Tyerman, TheEconomic Rationalefor Copyright Protection for Published Books: A Reply to Professor

Breyer, 18 UCLA L. Rev. 1100 (1971); Stephen Breyer, Copyright: A Rejoinder, 20UCLA L. Rev. 75 (1972).

8. Landes and Posner, supra note 7, at 325.

9. See, e.g., Richard A. Posner, Economic Analysis of Law (5th Ed. 1998) (surveying eco-

nomic analysis of many bodies of law).

10. Among the classic works are: Breyer, supra note 7; Gordon, supra note 4; Gillian K.

Hadfield, The Economics of Copyright: A Historical Perspective, 38 Copyright L.

Symposium (1992); Robert M. Hurt and Robert M. Schumann, The Economic Rationale

of Copyright, 56 Am. Econ. Rev. Papers & Proceedings 421 (1966); Landes and Posner,

supra note 7; Arnold Plant, The Economic Aspects of Copyright in Books, 1 Econometrica167 (1934). These and other important articles on copyright can be found in Wendy J.

Gordon and Richard Watt, Economics of Copyright: Developments in Research and

Analysis (2003); Ruth Towse and Rudi Holzhauer, The Economics ofIntellectual Property(2002 Vol. I, Parts II-V). Other important, but more recent, work is cited throughout this

article.

11. 17 U.S.C. sec. 102(b). See, e.g., Landes and Posner, supra note 7, at 347-53.

12. Id. at 345-^7.

13. 17 U.S.C. sec. 106(2). See, e.g., Landes and Posner, supra note 7, at 354-55.

14. 17U.S.C.sec.202.

15. 17 U.S.C. sec. 201(d).16. 17 U.S.C. sec. 302-304. See infra notes 51-55 and accompanying text for economic argu-

ments in favor of limits on the duration of copyrights.17. 17 U.S.C. sec. 201 (b). See, e.g., I.T. Hardy, An Economic Understanding of Copyright

Law's Work-Made-For-Hire Doctrine, 12 Colum.-VLA J.L. & Arts 181 (1988).18. See, e.g., Stanley M. Besen, Willard G. Manning, and Bridger M. Mitchell, Copyright

Liability for Cable Television: Compulsory Licensing and the Coase Theorem, 21 J. Law& Econ. 67 (1978); Stanley M. Besen, Sheila Kirby and Steven Salop, An Economic

Analysis of Copyright Collectives, 78 Va. L. Rev. 383 (1992); Robert P. Merges,

Contracting Into Liability Rules: Intellectual Property Rights and Collective Rights

Organizations, 84 Calif. L. Rev. 1293 (1996).


19. See, e.g., Joni M. Cherbo and Margaret Jane Wyszomirski, eds. The Public Life

of the Arts in America (2000); Ruth Towse, ed., A Handbook on Cultural Economics

(2003).

20. See Richard Watt, Copyright and Economic Theory: Friends or Foesi (2000).

21 . See, e.g., Timothy J. Brennan, Harper & Row vs. The Nation, Inc.: Copyrightability andFair Use, 33 J. Cop. Soc'y 368 (1986); William T. Fisher III, Reconstructing the Fair Use

Doctrine, 101 Harv. L. Rev. 1659 (1988); Gordon, supra note 4; Stanley J. Liebowitz,

Copying and Indirect Appropriability: Photocopying of Journals, 93 J. Pol. Econ. 945

(1985); Landes and Posner, supra note 7, at 357-61; Robert P. Merges, Are You MakingFun of Me? Notes on Market Failure and the Parody Defense of Copyright, 21 AIPLAL.Q. 305 (1993); Alfred C. Yen, When Authors Won't Sell: Parody, Fair Use, and

Efficiency in Copyright Law, 62 U. Colo. L. Rev. 79 (1991). See also Lisa N. Takeyama,

Welfare Implications of Unauthorized Reproduction of Intellectual Property in the

Presence of Demand Network Externalities, 42 J. Ind. Econ. 155 (1994).

22. See, e.g., Gordon, supra note 4, at 1652-57.

23. Id. at 1602-22. Gordon argues that fair use should be awarded in a wide range of other

cases as well. See id. at 1627-36 (discussing, e.g., externalities, antidissemination motives,

and nonmonetizable interests); Wendy J. Gordon, Excuse and Justification in the Law ofFair Use: Transaction Costs Have Always Been Only Part of the Story, 13 J. Copyright

Society 149(2003).24. The effect of a challenged use on the market, or potential market, for a copyrighted work

is one of the principal criteria for judging whether a use is fair. See 17 U.S.C. sec. 107.

25. See, e.g., William M. Landes, Copyright Protection of Letters, Diaries, and Other

Unpublished Works: An Economic Approach, 21 J. Legal Stud. 79 (1992).

26. See 17 U.S.C. sec. 1 10(6), 1 10(10).

27. See Sonny Bono Copyright Term Extension Act, Pub. L. No. 105-298, 112 Stat. 2827

(1998), codified at 17 U.S.C. sees. 302(a), 302(c), discussed infra notes 47-72 and accom-

panying text.

28. 17 U.S.C. sec. 104A. See, e.g., Edward Lee, The Public's Domain: Evolution of LegalRestraints on the Government's Power to Control Public Access Through Secrecy or

Intellectual Property, 55 Hastings L.J. 91 (2003) (discussing economic and constitutional

problems with 'restoration' of these copyrights).

29. See, e.g., Richard A. Epstein, The Dubious Constitutionality of the Copyright Term

Extension Act, 36 Loy. L.A.L. Rev. 123 (2002); Dennis S. Karjala, Judicial Review of

Copyright Term Extension Legislation, 36 Loy. L.A.L. Rev. 199 (2002).

30. Eldred vs. Ashcroft, 123 S.Ct. 769 (2003).

3 1 . See, e.g., Peter A. Jaszi, Toward a Theory of Copyright: The Metamorphoses ofAuthorship,1991 Duke L.J. 455 (discussing the romantic conception of copyright). See also Breyer,

supra note 7, at 284-91 (discussing non-economic justifications for copyright).

32. A decade ago, I recommended that government offices charged with responsibilities for

intellectual property policy should hire staffeconomists. See Pamela Samuelson, Will the

Copyright Office Be Obsolete in the Twenty-First Century?, 13 Cardozo Arts & Ent. L.J.

55 (1994).

33. See, e.g., Jessica Litman, Digital Copyright (2001); William F. Patry, Copyright and the

Legislative Process: A Personal Perspective, 14 Cardozo Arts & Ent. L.J. 139, 141 (1996).

34. The President of the Association of American Publishers, for example, is Pat Schroeder,

formerly a Congresswoman. The recently retired head of the Business Software Alliance,

Emery Simon, was with the US Trade Representative Office where he worked on intel-

lectual property issues before joining BSA. Bruce Lehman who served as head of the

Patent & Trademark Office during the Clinton Administration had been a legislative

staffer in Congress before becoming a lawyer who represented copyright industry groups.Shira Perlmutter, formerly with the US Copyright Office, is now a Vice President and

Associate General Counsel for Intellectual Property at AOL Time Warner.

35. France and Germany are among the countries that protect authors' rights on this basis.

See, e.g., Goldstein, supra note 1, at 168-69; Alain Strowel, Droit d'auteur and


Copyright: Between History and Nature, in Of Authors and Origins: Essays on Copyright

Law, 235 (Brad Sherman and Alain Strowel, eds, 1994).

36. See, e.g., Richard Gilbert and Carl Shapiro, Optimal Patent Length and Breadth, 21

RAND J. Econ. 106 (1990) (arguing for a thin scope and long-term of protection for

patents); Peter S. Menell, Tailoring Legal Protectionfor Computer Software, 39 Stan. L.

Rev. 1045 (1989) (arguing for short duration for legal protection for software);

Samuelson et al., supra note 3, at 2423 (arguing for a short duration of legal protection

against cloning of computer program behavior and other industrial design features of

programs).37. See, e.g., Michael L. Katz and Carl J. Shapiro, Network Externalities, Competition and

Compatibility, 75 Am. Econ. Rev. 424 (1985) (discussing economics of network externali-

ties). An important contribution to the legal literature on network effects is Mark A. Lemleyand David F. McGowan, Legal Implications of Network Economic Effects, 86 Cal. L. Rev.

479 (1998) (drawing out several possible legal implications of network externalities).

38. See, e.g., Joseph Farrell, Standardization and Intellectual Property, 30 Jurim. J. 35 (1 989);

Wendy J. Gordon, On Owning Information, 78 Va. L. Rev. 149 (1992); Lemley &McGowan, supra note 37; Menell, supra note 36; J.H. Reichman, Legal Hybrids Between

the Patent and Copyright Paradigms, 94 Colum. L. Rev. 2432 (1994); Samuelson et al.,

supra note 3. See also sources infra note 127.

39. See, e.g., David McGowan, Copyright Nonconsequentialism 69 Missouri L. Rev. 1 (2004)

(emphasizing indeterminate relationship between copyright rules and desired outcomes).Cf. George Priest, What Economists Can Tell Lawyers About Intellectual Property, in Res.

Law & Econ. 19 (J. Palmer and R. Zerbe, eds. 1986) (questioning the usefulness of eco-

nomic analysis in respect of intellectual property law because of inadequacies in empiri-cal understanding of innovation).

40. 239 F.3d 1004, 1016-17 (9th Cir. 2002).

41 . Plaintiffs' economic expert David Teece focused on the harmful effects on the plaintiffs'

ability to offer competing digital music services. Id. at 1017. Defendants' expert arguedthat use of Napster for sampling purposes enhanced the willingness of users to buyrecorded music. Id. Plaintiffs commissioned two empirical studies of the effects of peer-

to-peer filesharing on sales of recorded music which concluded that among college stu-

dents, purchases had declined because of Napster filesharing. Id. at 1016.

42. On public choice generally, see, e.g., William N. Eskridge, Jr., Implications of Public

Choice Theoryfor Statutory Interpretation, 74 Va. L. Rev. 275 (1988); Daniel A. Farber

and Philip P. Frickey, Law and Public Choice: A Critical Introduction (1991).

43. See, e.g., Glynn S. Lunney, Jr., The Death of Copyright: Digital Technology, Private

Copying and the Digital Millennium Copyright Act, 87 Va. L. Rev. 813 (2001); Neil W.

Netanel, Locating Copyright Within the First Amendment Skein, 54 Stan. L. Rev. 1

(2001); Patry, supra note 33; Stewart E. Sterk, Rhetoric and Reality in Copyright Law, 94

Mich. L. Rev. 1197(1996).44. Pub. L. No. 105-298, sec. 102(b) and (d), 112 Stat. 2827-2828 (amending 17 U.S.C. sees.

302, 304). A Canadian legislative proposal to extend the term of copyrights in unpub-lished works has been controversial. See James Adams, Estates' Rights in Canadian

Copyright Re-examined, Globe and Mail, Sept. 23, 2003, available at http://www.the-

globeandmail.com/servlet/ArticleNews/TPStory/LAC/20030923/COPY23/TPEntertain

ment/TopStories.45. Pub. L. No. 105-304, 112 Stat. 2860 (1998), anti-circumvention provisions codified at 17

U.S.C. sec. 1201.

46. See, e.g., Alex Breeding, What Are the Implications of the New Super-DMCA Laws?,

TechRepublic, 16 July, 2003, available at http://techrepublic.com/5100-6314-5054407.html. A status report on enactment of state super-DMCA laws can be found at

http://www.freedom-to-tinker.com/superdmca.html.47. Eldred vs. Ashcroft, 123 S.Ct. 769 (2003). Justices Stevens and Breyer dissented in sepa-

rate opinions, the former offering an alternative historical and constitutional analysis,

and the latter mainly focusing on the economic effects of the CTEA. See id. 790-801

(Stevens dissent) and 801-15 (Breyer dissent).

Should economics play a role in copyright law and policy? 1 9

48. See Brief for Petitioners in Eldred vs. Ashcroft, at 15-28, available at http://eldred.cc/

legal/supremecourt.html. Eldred also argued that the CTEA violated the First

Amendment. See id. 34-48.

49. Id. at 23.

50. Id.

51. See Brief of George A. Akerloff et al. in Eldred vs. Ashcroft, available at http://eldred.cc/

legal/supremecourt.html. Akerloff, Kenneth Arrow, James Buchanan, Ronald Coase,

and Milton Friedman were the Nobel Prize winning signatories of this brief. See id.,

Appendix A.

52. Id. at 8.

53. Id. at 11-13.

54. Id. at 12- 14.

55. Id. at 8.

56. Eldred, 123 S.Ct. at 781-82.

57. Id. at 783-84.

58. Id. at 801 (Breyer, J., dissenting) ('The economic effect of this 20-year extension - the

longest blanket extension since the Nation's founding- is to make the copyright term

not limited, but virtually perpetual. . . . And more importantly, its practical effect is not

to promote, but to inhibit the progress of "Science" - by which word the Framers meant

learning or knowledge. . . .').

59. See, e.g., Eastman Kodak Co. vs. Image Technical Services, Inc., 504 U.S. 451 (1992);

Jefferson Parish Hosp. Dist. No. 2 vs. Hyde, 466 U.S. 2 (1984).

60. See, e.g., Mazer vs. Stein, 347 U.S. 201, 219 (1954); Sony Corp. of Am. vs. Universal City

Studios, Inc., 464 U.S. 417, 429 (1984).

61 . See, e.g., United States vs. Lopez, 514 U.S. 549 (1995); United States vs. Morrison, 529

U.S. 598 (2000). In both decisions, Justices Breyer, Ginsburg, Souter and Stevens dis-

sented from rulings that Congress lacked power under the Commerce Clause to enact

legislation to regulate guns near schools and gender-based crimes of violence. See, e.g.,

Paul J. Heald and Suzanna Sherry, Implied Limits on the Legislative Power The

Intellectual Property Clause as an Absolute Constraint on Congress, 2000 U. 111. L. Rev.

1119 (2000) (discussing numerous Supreme Court decisions restrictively interpreting

Congress' power and considering their implications for various intellectual property

provisions).

62. See Brief for Petitioners, supra note 48, at 1 1-12 (invoking Lopez and Morrison).

63. Id.

64. Id. at 12-13, 20-21.

65. In view of Justice Breyer's earlier economic skepticism about copyright, see Breyer, supranote 7, Eldred's lawyers expected him to be skeptical of the CTEA on economic groundsand receptive to arguments based on the constitutional history of the Intellectual

Property Clause indicating that the Framers embedded limits in it in order to guard

against rent-seeking. See Brief for Petitioners, supra note 48, at 23-26. Eldred's lawyers

thought Justice Stevens would be receptive to their arguments because his opinion in

Sony Corp. of Am. vs. Universal City Studios, Inc., 464 U.S. 417, 429 (1984) emphasized

copyright as a limited monopoly, the primary purpose of which was to promote public

access to knowledge. (The monopoly privileges that Congress may authorize are neither

unlimited nor primarily designed to provide a special private benefit. Rather, the limited

grant is a means by which an important public purpose may be achieved. It is intended

to motivate the creative activity of authors and inventors by the provision of a special

reward, and to allow the public access to the products of their genius after the limited

period of exclusive control has expired.')

66. Eldred, 1 23 S.Ct. at 775-76.

67. See Addendum to Reply Brief for the Petitioners in Eldred vs. Ashcroft (chart

distinguishing CTEA from other extensions), available at http://eldred.cc/legal/

supremecourt.html.68. Eldred, 123 S.Ct. at 778, quoting New York Trust Co. vs. Eisner, 256 U.S. 345, 349 (1921).

69. Eldred, 123 S.Ct. at 790.


70. Brief for Petitioners, supra note 48, at 39-47; Eldred, 123 S.Ct. at 788-90.

71. See, e.g., Pamela Samuelson, The Constitutional Law of Intellectual Property AfterEldred vs. Ashcroft, 50 J. Cop. Soc'y 547, 556 (2003).

72. Eldred, 123 S.Ct. at 775.

73. See Final Report of the National Commission on New Technological Uses of CopyrightedWorks 14-26 (1979). TheCONTU analysis focused on the economics of developing indi-

vidual programs, but did not consider the appropriate scope of copyright protection for

computer programs or vexing issues such as the interoperability and decompilationissues discussed infra notes 98-120 and accompanying text.

74. See Council Directive 91/250 on the Legal Protection of Computer Programs, 1991 O.J.

(LI 22) 42 (cited hereinafter as 'Software Directive'). Article 6 of this Directive permits

decompilation of computer program code for purposes of achieving interoperability

among programs. See, e.g., Pamela Samuelson, Comparing U.S. and E.C. CopyrightProtection For Computer Programs: Are They More Different Than They Seem?, 13

J. Law & Comm. 279, 287-88 (1994) (discussing the concerns of the EuropeanCommission's competition directorate about the software directive that influenced the

rule in favor of decompilation for interoperability purposes).

75. See Agreement on Trade-Related Aspects of Intellectual Property Rights, in The LegalTexts: The Results of the Uruguay Round of Multilateral Trade Negotiations 321, art.

10(1) at 325 (Cambridge 1994) (cited hereinafter as TRIPS').76. See Semiconductor Chip Protection Act, Pub. L. No. 98-620, 98 Stat. 3347 (1 984), now

codified at 17 U.S.C. sec. 901 et seq. See, e.g., Prepared Testimony of F. Thomas Dunlap,

Jr., Corporate Counsel and Secretary of Intel Corp., Hearings Before the Subcommittee

on Courts, Civil Liberties, and the Administration of Justice of the House Committeeon the Judiciary, on H.R. 1028, 98th Cong., 1st Sess. (8/3/83) explaining the industry's

need for this legislation; Prepared Statement of Sen. Charles McC. Mathias, Jr., House

Hearings, supra, at 3: '[Chip] innovators are being ripped off by onshore and offshore

"chip pirates" who, for a fraction of the developer's cost, can now legally appropriateand use these chip designs as their own.' Of particular concern was the loss to Japanese

industry of a substantial share of the market for random access memory chips to

Japanese competitors whose superior quality control made their chips very competitive.See Stephen P. Kasch, The Semiconductor Chip Protection Act: Past, Present, and Future,

7 High Tech. L.J. 71,79(1993).77. See Vessel Hull Design Protection Act, Pub. L. No. 105-304, tit. V, 1 12 Stat. 2905 (1998),

codified at 17 U.S.C. sec. 1301-1332. Had this law been subjected to impartial economic

analysis, it is difficult to believe Congress would have passed it.

78. See, e.g., J.H. Reichman and Paul F. Uhlir, Database Protection at the Crossroads: Recent

Developments and Their Impact on Science and Technology, 14 Berkeley Tech. L.J. 793

(1999) (discussing the evolution of legislative proposals).79. See European Parliament and Council Directive 96/9/EC of 1 1 March 1 996 on the Legal

Protection of Databases, 1996 O.J. (L. 77) 20, available at http://europa.eu.int/comm/

internal_market/en/intprop/docs/. For a history of the justification for and evolution of

this Directive, see, e.g., J.H. Reichman and Pamela Samuelson, Intellectual Property

Rights in Data?, 50 Vand. L. Rev. 51, 80-97 (1997).

80. EU Database Directive, supra note 79, recitals 1-12.

81. Id., arts. 7, 10.

82. See, e.g., Reichman and Samuelson, supra note 79, at 113-36 (criticizing the EU sui

generis regime because of its potential harmful effects on competition and innovation as

well as for science).

83. See, e.g., Brief of AkerlofT, supra note 51; Karjala, supra note 29.

84. See, e.g., Pamela Samuelson and Suzanne Scotchmer, The Law and Economics ofReverse Engineering, lllYale L.J. 1575, 1630-49 (2002) (critical of anti-circumvention

regulations).

85. See, e.g., 17 U.S.C. sees. 1003-1007.

86. See, e.g., Neil W. Netanel, Impose a Noncommercial Use Levy to Allow Free Peer-to-Peer

File Sharing, 17 Harv. J.L. & Tech.l (2003).


87. 17 U.S.C. sec. 1201(a)(l)(C). See Memorandum to James Billington from Mary Beth

Peters regarding Recommendations of RM 2002-4, Rulemaking on Exemptions from

Prohibition on Circumvention of Copyright Protection Systems For Access Control

Technologies, dated 27 Oct., 2003, available at http://www.copyright.gov/1201/docs/

registers-recommendation.pdf.88. The subsequent paragraphs on Whelan, Altai and Sega vs. Accolade derive from Pamela

Samuelson, Economic and Constitutional Influences on Copyright Law in the United

States, 23 Eur. Intell. Prop. Rev. 409 (Sept. 2001).

89. 797 F.2d 1 222 (3rd Cir. 1 986).

90. Id. at 1239-40.

91. Id. at 1231.

92. See Brief Amicus Curiae of ADAPSO, The Computer Software and Services Industry

Association, Inc., Whelan Associates, Inc. vs. Jaslow Dental Laboratory, Inc. (Civ. A.

No. 85-1358) (1986).

93. Whelan, 797 F.2d at 1237.

94. See, e.g., Johnson Controls, Inc. vs. Phoenix Control Systems, Inc., 886 F.2d 1 173 (9th

Cir. 1989); Lotus Development Corp. vs. Paperback Software Int'l, 740 F. Supp. 37

(D. Mass 1990).

95. 982 F.2d 693 (2nd Cir. 1992).

96. Id. at 702-3.

97. Whelan, 797 F.2d at 1 240.

98. Altai, 982 F.2d at 708-11.

99. Id. at 707.

100. 17 U.S.C. sec. 102(b).

101. Altai, 982 F.2d at 710.

102. Id. at 71 1-12.

103. Id. at 712 (citation omitted).

104. Id.

105. Id.

106. See id.

107. Id. at 707-11.

108. See, e.g., Arthur R. Miller, Copyright Protectionfor Computer Programs, Databases, and

Computer-Generated Works: Is Anything New Since CONTU?, 106 Harv. L. Rev. 977

(1993) (arguing that interfaces should be protectable by copyright).

109. /4/to/,982F.2dat710.110. See, e.g., Alan K. Palmer and Thomas C. Vinje, The EC Directive On The Legal

Protection Of Computer Software: New Law Governing Software Development, 2 Duke

J. Comp. & Int'l L. 65 (1992).

111. Id. at 71-78.

112. See, e.g., Jonathan Band and Masanobu Katoh, Interfaces on Trial: Intellectual

Property and Interoperability in the Global Software Industry 271-97 (1995).

113. 977 F.2d 1510 (9th Cir. 1992). The Ninth Circuit had endorsed Whelan in Johnson

Controls, Inc. vs. Phoenix Control Systems, Inc., 886 F.2d 1173 (9th Cir. 1989), but

repudiated it in Sega, 977 F.2d at 1524-25.

1 14. See, e.g., Andrew Johnson-Laird, Software Reverse Engineering In The Real World, 19

U. Dayton L. Rev. 843 (1994).

115. 17 U.S.C. sec. 106(1).

116. Although some software developers openly publish APIs (application programming

interfaces) for their programs, many do not. The willingness of program developers to

license APIs to other firms varies considerably. Some firms only license APIs on terms

that some developers regard as unacceptable (e.g., giving up the right to make a version

of the same program for other platforms). See Sega, 977 F.2d at 1514.

117. 17 U.S.C. sec. 107.

118. The fair use defense is discussed in Sega, 975 F.2d at 1 520-28.

1 19. See, e.g., Sega, 977 F.2d 1510, 1527-8.

120. For scholarly commentary making economic arguments in favor of fair use for


decompilation for purposes of interoperability, see, e.g., Lawrence Graham andRichard O. Zerbe, Jr., Economically Efficient Treatment of Computer Software: Reverse

Engineering, Protection and Disclosure, 22 Rutg. Comp. & Tech. L.J. 61, 132-34 (1996);

Lemley and McGowan, supra note 37, at 525; Samuelson and Scotchmer, supra note

84, at 1621-26.

121 . See, e.g., Band and Katoh, supra note 1 12, at 271-82, 294-97; Jonathan Band, SoftwareReverse Engineering Amendments in Singapore and Australia, J. Internet L. 17, 20

(Jan. 2000).

1 22. See, e.g. ,PriceWaterhouseCoopers, Contributions of the Packaged Software Industry to

the Global Economy (1999), http://www.bsa.org/usa/globalip/econ/pwcl999.pdf.123. See, e.g., A&M Records, Inc. vs. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001) (prelim-

inary injunction against peer-to-peer software developer as a contributory and vicari-

ous infringer of digital music copyrights because of centralized directory and indexing

functionality); In re Aimster Litigation, 334 F.3d 643 (7th Cir. 2003) (summary judg-ment against peer-to-peer developer as contributory infringer because of direct assist-

ance to infringements). But see Metro-Goldwyn-Mayer Studios, Inc. vs. Grokster, Ltd.,

259 F. Supp.2d 1029 (C.D. Cal. 2003)(summary judgment in favor of peer-to-peer soft-

ware developer because of capability for substantial noninfringing uses).

1 24. See supra note 4 1 .

125. See, e.g., Brief Amicus Curiae of Forty Intellectual Property Professors in MGM vs.

Grokster, available at http://www.sims.berkeley.edu/~pam/papers (arguing that the

legislature, not the courts, should regulate technologies with non-infringing uses).

1 26. See, e.g., Theberge vs. Galierie d'Art du Petit Champlain, Inc., 2002 S.C.C. 24.

127. See, e.g., Yochai Benkler, Intellectual Property and the Organization of Information

Production, 22 Int'l Rev. of L. & EC. 81 (2002); supra note 1 Lemley; Glynn S. Lunney,

Jr., Reexamining Copyright's Incentives-Access Paradigm, 49 Vand. L. Rev. 483 (1996);

Michael Meurer, Copyright Law and Price Discrimination, 23 Cardozo L. Rev. 55

(2001); Christopher S. Yoo, Towards a Differentiated Products Theory of Copyright,

Chapter 6 of this volume (for an expanded version see Christopher S. Yoo, Copyrightand Product Differentiation, 79 N.Y.U.L. Rev. 212 (2004)).

128. See, e.g., Gordon and Watt, supra note 10. For information about SERCI and its con-

ferences and journal, see http://www.serci.org/.

129. See, e.g., International Intellectual Property Alliance, Copyright Industries in the U.S.

Economy: The 1998 Report (Exports of copyright-related products and services are

increasing at an impressive rate - in 1996 their value surpassed that of every other

export sector in the United States.)

130. See, e.g., Pamela Samuelson, Challenges For the World Intellectual Property

Organization and the Trade-Related Aspects of Intellectual Property Rights Council In

Regulating Intellectual Property Rights In The Information Age, 21 Eur. Intell. Prop.Rev. 578 (Nov. 1999).

2. Risk sharing and the distribution of

copyright collective income

Arthur Snow and Richard Watt

2.1 INTRODUCTION

Hollander (1984, p. 200) sums up the basic functions of copyright collect-

ives as follows:

Copyright collectives are associations to whom authors transfer copyrights for

purposes of exploitation. In general, collectives are concerned with the follow-

ing functions: (1) they grant licenses for the use of works in their repertory,

(2) they negotiate and collect royalties, and distribute them to their members,

(3) they take legal action against those who infringe the copyrights to which theyhold title.

Each of these three aspects of collective administration of copyrights (as

well as the monopoly position that such collectives enjoy) has been subject

to the analysis of economic theory. In particular, the literature asserts that

the underlying reason for copyright collectives to form naturally is that by

joining together and marketing a blanket license to the entire repertory,

each individual member can achieve a better utility position than by acting

alone. In this chapter, we shall consider the possible ways in which collect-

ive administration improves the welfare of its members.

In the above quote from Hollander, it is clear that the principal idea is to

share certain transactions costs over collective members (costs of negotiat-

ing, receiving payment for, and protecting license agreements). Clearly,

then, the most obvious reason for collective administration is the welfare

gains that the members can enjoy as transactions costs are diminished.

The Coase theorem implies that final resource allocations may, in general,

be socially inefficient whenever transactions costs exist. When it comes to

intellectual property markets, in which access to intellectual products

between individuals is traded, transactions costs not only exist, but they mayeven be more significant than the final value of the object of the transaction.

The reason for this is the fact that intellectual products are typically of a

public good nature, and so there is a very low cost of free-riding associated

23


with them. Once access has been granted to one user, it becomes very costly

to then exclude access being taken by the entire set of potential users.

This problem is often overcome, at least partially, by collective adminis-

tration. The example of copyright is perhaps the most prevalent case. Whena copyrightable product is created, along with it is created an entire set of

access rights, some of which can be transacted with low transactions costs,

but others of which can only be transacted at a very high transactions cost.

For example, consider the case of a musical composition. Once the compos-ition has been fixed onto a physical support, say a CD Rom, it can easily be

listened to by whoever has that CD Rom. When a consumer purchases the

CD Rom from a recognized retailer, the right to a limited set of access priv-

ileges is automatically purchased as well. These include the right to listen to

the music at any time and the right to sell the CD Rom to a third party. In

most countries, the consumer has the right to make a copy of the CD Romfor personal use as well. However, with the purchase price of the CD Rom,the consumer will not acquire the right to make copies of the music for

resale, or the right to play the music to an audience, especially for economic

gain. The problem for the owner of the copyright in the music is how to

monitor the use that the consumer makes of the CD Rom at a reasonable

cost. One partial method lies in collective administration of the copyright.

Whenever the set of users of a particular type of intellectual product is

common to many (say, n) copyright holders, in principle they can, by

joining together and selling a license to the entire repertory, save on the

transactions costs implied by n - 1 individual rounds of negotiation, and

likewise they can also save on both the monitoring costs and the costs of

litigation whenever such recourse is needed. 1 In this way, the option of col-

lective administration can make feasible a set of transactions that is

unprofitable when carried out individually. In economic terms, copyright

collectives enjoy the benefits of a natural monopoly.

However, saving on transactions costs is not the only manner in which

collective administration can be of value to individual copyright holders. If

we assume, as is indeed very natural, that the final market value of each

copyright is a random variable (that is, there is uncertainty as to its final

value), then a copyright collective is able to alleviate the risk-bearing situ-

ation of its members. If the members are risk-averse (again, a very natural

assumption), then an improved risk situation provides for expected utility

gains which are just as important as the gains that the transactions costs

savings provide for. It is curious to note that this aspect of collective admin-

istration seems to have been entirely ignored, both in theoretical contribu-

tions and in real-life cases.

Given the diversity in the market value of each composition in a copy-

right collective's repertory, it is quite common for copyright collectives to

Risk sharing and the distribution of copyright collective income 25

take great pains in analysing the markets in which their repertories are used

in order to determine how much use is made of each particular compos-ition in the repertory. This is puzzling since it goes quite against the idea

that the songs are bundled in the first place, in a blanket license, in order to

save on such transactions costs. 2 The reason why this is done is simply to

be able to base the final distribution of net income on the use made of each

individual copyright. In this way, the copyright holders of the songs that

are most played can be given a proportionately greater share of the collec-

tive's surplus. Abstracting from the fact that monitoring the market use of

individual songs increases the transactions costs of running the collective,

there are important reasons strongly rooted in the economic theory of risk

bearing that suggest that doing this is inconsistent with the best interests of

the pool of members, thereby making current copyright collective activity

inefficient.

In this chapter, we think of a copyright collective as a simple coalition, in

the sense of traditional game theory. Each potential individual member has

the opportunity to join the coalition, or to act alone. In such a setting, it is

very clear that the decision on whether or not to join the coalition is affected

both by how much of the coalition's income is distributed among the

members, and by exactly how that income is distributed. Therefore, the dis-

tribution rule will affect the number of members that the coalition has

(in equilibrium) as well as the welfare of each individual. We shall imposetwo basic requirements of the distribution mechanism of a coalition's net

income. First, we will require that the distribution method be such that indi-

vidual copyright holders prefer to become members of the coalition rather

than acting alone (we refer to this as the 'participation condition'). Second,

we will require that existing members of the coalition prefer new members to

join (we refer to this as the 'incentive compatibility condition'). Given these

two requisites, we first ask if a distribution method that consists of learning

the true market value of each composition and then paying each individual

member accordingly satisfies these requirements. We then propose alterna-

tive distribution mechanisms that satisfy the two requirements and are

preferred by all individuals to the 'learn-then-distribute' option.

2.2 A SIMPLE MODEL OF A COPYRIGHTCOLLECTIVE

Let us assume that there is a set of n individuals, each of whom can create a

composition that has strictly non-negative market value, but where the

market value of the composition is a random variable. Here, we make the

simplest possible assumption, namely that all composition lotteries are


identical. Indeed, the assumption that all compositions are identical, albeit

deterministic, is common in the literature (see, for example, Besen et al.

1992).

Moreover, we will assume that the composition lotteries are not only

identically distributed, but that they are also statistically independent with

a probability of success that is common knowledge.3Concretely, we assume

that each composition can be either a 'success', in which case its market

value is x = 1,or a 'flop', in which case its market value is x = 0. We assume

that the probability of success is/?, which is assumed strictly between and

1,and we denote the true outcome of composition i by xr

Now, assume that the (monetary equivalent) costs of creation of each

composition are constant and equal to c, and that if the composition is

indeed marketed by the copyright holder (acting alone), then there exist

transactions costs (costs of negotiation, monitoring, and litigation),

denoted by t.

All copyright holders are assumed to be strictly wealth-loving and

strictly risk-averse. In order to avoid the need for formally considering

utility, instead we take the approach of mean-variance analysis. Concretely,we shall assume that all copyright holders have a strict preference for

greater expected wealth, and a strict preference for lower variance of

wealth. If copyright holder i goes ahead with the creation and marketingof the composition on an individual basis, then the expected value of his

final wealth (assuming that he has no wealth outside of what the compos-ition generates) is given by

(2.1)

On the other hand, the variance of his wealth is given by

E(E(x -c-t}-(x-c- t)Y = E(p- x)2=p(l -p), (2.2)

which depends only on the probability of success.

Now, consider the situation of the coalition. Assuming that the coali-

tion has m < n members, the final market value of the blanket license

follows a binomial distribution. It can take on any of the discrete values

0, 1, 2, 3, . . ., m, depending on exactly how many of the m member com-

positions are successes. Using the well-known formula for the binomial

distribution, the probability that exactly h of the m compositions will be

successful is given by

m\


Given this, the expected final market value of the coalition's repertory is

True to the public good assumption of the compositions in question, and

to the fact that each composition has the same user set, we assume that the

total cost of marketing the blanket license to all the compositions in its

repertory is the same as for any particular individual; that is, the entire

repertory can be marketed at a (transactions) cost of t.4 Given this, the

coalition (with m members) earns a net income of X(m) t for distribution

to the members.

Now assume that the collective can learn the true outcome of each indi-

vidual composition lottery at some cost, which we fix at A:, and then give

individual member / a payment of

t + k

Under this 'learn-then-distribute' system, a member's expected wealth is

'

x - c-i^LJi- c-iAU(i-Jo-c- ) (2.3)

=p-c m

with a variance of

x-c- \-L-c- \\=E(p-xY=p(\-p\ (2.4)mI \

m//

Clearly, under this distribution rule, all members get a final wealth lottery

with the same variance as acting alone. It is also clear from (2.3) that the

expected wealth of each collective member is strictly increasing in m, so that

all collective members will welcome new members to the coalition, imply-

ing that the incentive compatibility condition is satisfied. However, it is not

clear that the participation condition is satisfied. Comparing the expected

wealth in (2.3) with that in (2.1), participation (that is, all individuals prefer

to become members of the coalition) can be guaranteed if and only if

t + k

m


Clearly, this requirement is more easily satisfied when k is small, when t is

large, or when m is large, each of which is intuitively understandable.

In short, for any given values of / and A:, we have:

Proposition 2.1 A distribution rule under which a collective learns (at a

cost of k) the final result of each independent composition in its reper-

tory, and distributes final income accordingly, is incentive compatible.

There is a minimum membership, given by m* = (t + k)/t, beyond which

the distribution rule also satisfies the participation condition. 5

Assuming that the coalition has sufficient members, the fact that the

learn-then-distribute system satisfies both incentive compatibility and par-

ticipation is due entirely to the transactions costs savings that the coalition

offers. However, the risk-bearing situation of each coalition member has

not changed, since the coalition offers the same variance of the final wealth

lottery as acting alone. Hence, while the coalition increases the expected

value of income (by saving on transactions costs) it does nothing at all to

alleviate the risk situation of the members. Given that the creators are risk-

averse, it would be welfare enhancing if the coalition could offer a better

risk situation as well as transactions costs savings. We now go on to con-

sider if this is indeed possible.

2.3 RISK SHARING DISTRIBUTION MECHANISMS

Consider first the opportunity that the copyright collective has of offering

full and fair insurance to all members. Since no learning is done as far as

the final result of each individual lottery, the only costs of administering

the set of copyrights is /. Now, assume that the collective pays each

member the expected value of his composition less his share of the trans-

actions costs with certainty. In this case, each creator receives an expectedwealth of

p~ c ~~ <2 - 5>

with zero variance. Clearly, this distribution rule satisfies both participation

and incentive compatibility. Moreover, it also Pareto dominates the learn-

then-distribute rule, since all members get a lower variance of wealth and

a higher expected final wealth. Thus, we have:

Proposition 2.2 A sharing rule stipulating that each individual copy-

right holder receives the expected value of his composition with


certainty satisfies both participation by all individuals and incentive

compatibility. It also strongly Pareto dominates the learn-then-distribute

rule.

The fact that the welfare benefits of the full insurance mechanism are so

obvious begs the question of why this system is not typically used. Perhapsthe answer lies in the fact that, if the collective pays each member his

expected value, then there is a strictly positive probability that the final

market value of the repertory will either exceed or fall short of the total

outlay paid to the members as royalties. Unless the collective has access to

a perfect reinsurance market, the members of the collective would still bear

some risk with respect to their final wealth positions, although this risk

becomes smaller as the membership increases. Clearly though, the full and

fair insurance option may not be feasible. However, there are intermediate

options that have full insurance, although not fair insurance, that satisfy

both participation and incentive compatibility, and that still Pareto dom-inate the learn-then-distribute rule.

Consider the following simple option. The collective distributes a fixed

sum to each individual member such that the expected value of each

member's income is the same as if the distribution rule had been learn-then-

distribute. That is, each member is given a fixed payment of

t + k t kp =p .

m mmOnce again, this rule clearly satisfies incentive compatibility, and satisfies

participation under the same proviso as in Proposition 2. 1 .

6Moreover, this

system is strictly preferred by all copyright holders to the learn-then-

distribute rule since it offers the same expected wealth with a lower variance.

Since no learning is done, this rule implies that total royalty payments are

mp t k, while the total net income of the collective is X(m) t. Under

this system, the probability that the collective runs into* a bankruptcy

problem can be significantly reduced (although, naturally, the reduction

depends on the size of k). For example, if we assume that m =100, and if we

assume that k = 10, then the probability of bankruptcy is under 3 per cent. 7

Naturally, the accumulated but not distributed net income of the collective

would, at certain times, be distributed to the members as profit (as opposedto strict royalty payments). Whether or not such a system is indeed feasible

depends on the true values of the parameters m, p and k, and on the access

that the collective has to other financial products (short-term loans, third

party insurance, etc.), but at least in principle it would seem to be feasible in

most modern societies.


Proposition 2.3 A sharing rule stipulating that each member of the col-

lective receives with certainty a payment equal to the expected value of

the payment under the learn-then-distribute rule, satisfies incentive com-

patibility always, and satisfies participation under the same restriction

on total collective membership as in Proposition 2.1. Moreover, this rule

Pareto dominates the learn-then-distribute rule, since it sets the variance

of each copyright owner's final wealth to 0. The probability of bank-

ruptcy under this distribution rule becomes negligible as k increases.

Even if the full but not fair insurance option were not feasible, there are

other options that also Pareto dominate the learn-then-distribute rule and

that never run into bankruptcy problems. To see some of these options,

note that the expected value of the repertory is8

EX(m) = mp (2.6)

and the variance of the market value of the repertory is

E(mp-X(m))

2 = mp(\-p). (2.7)

Now assume that the collective's net income is distributed according to

pre-agreed shares, such that member / receives a proportion \iof the final

net income of the collective, where

in order to avoid all bankruptcy issues (that is, the collective will distribute

exactly all of its net income, whatever that turns out to be). Using this rule,

the expected final wealth of member / is

E\\\X(m)--\-c\ = E\.X(m)-c--mj I

m

= \iEX(m)-c--m

= \mp -c-- (2.8)m

with variance

\mp - c - - -I \

iX(m)- c - -

} }

=E(\.(mp

-X(m)))

2

m \ tn

= \2E(mp-X(m))2.

Risk sharing and the distribution of copyright collective income 3 1

Using (2.7), we find that the variance of final wealth under this distribution

rule is

\}mp(\-p\ (2.9)

Now, since each composition is, ex ante, identical, let us simply assume

that \f= \/m V/. In this case the expected value of final wealth for each

member of the collective is

mwhich is the same as under the full and fair insurance mechanism, and

greater than under the learn-then-distribute rule. On the other hand, the

variance of the final wealth of each of the collective members is

~r-which is greater than any full insurance mechanism (under which variance

is 0), but less than what the learn-then-distribute rule offers (which from

(2.4) is/>(l -/>))

Finally, note that this sharing rule satisfies both incentive compatibility

(expected value is increasing in m and variance is decreasing in m) and par-

ticipation (expected value of being a collective member is strictly greater

than acting alone because of the transactions costs savings, and the vari-

ance that being a member offers is strictly less than that corresponding to

acting alone), and that it Pareto dominates the learn-then-distribute rule

(since it offers both a greater expected value, as the learning costs are saved,

and a lower variance). Thus, we have:

Proposition 2.4 A sharing rule stipulating that each collective memberis paid an equal share of the final net income of the collective is both

incentive compatible and satisfies participation. It also strictly Pareto

dominates the learn-then-distribute rule.

Although we have found a manner in which the net income of a copy-

right collective can be distributed that strictly Pareto dominates the cur-

rently used method, and that will not imply any bankruptcy issues, one

wonders if this method can be improved upon. This type of question has

been analysed extensively in the literature on efficient risk sharing. Twomain results of that literature are worth mentioning here. First, the

so-called 'mutuality principle' of Karl Borch (see Borch 1960) implies that

if a group of risk-averse agents each contribute a random variable to a

common pool (known as a mutual), then a Pareto-efficient risk-sharing


rule must make the final wealth of each individual member independentof the particular contribution that he made. In other words, the final

wealth of each member of a copyright collective should depend only on

the final value of X(m), and not on the individual composition lotteries.

This has been achieved in our proposal, where each individual receives an

equal share of X(m), but is clearly not achieved in the learn-then-distribute

rule.

Second, the question naturally arises: can the equal-share rule also be

improved upon. In particular, since it satisfies the mutuality principle, the

equal-share rule fully diversifies all diversifiable risk, but it leaves the final

wealth of each individual dependent upon the aggregate, or social, risk as

determined by the random variable X(m). The question is exactly how this

aggregate risk should be shared. This question was solved by Wilson

(1968), who proves that full risk sharing efficiency will require that each

member of a mutual should receive a fraction of the total pooled wealth

(here X(mj) that is equal to his individual risk tolerance divided by the sumof the risk tolerances of all members.

The sharing rule based on risk tolerances would, almost certainly, implythat members who supply identically valued compositions to the repertoryreceive different shares of the final pooled wealth. Besen et al. (1992,

pp. 395-7) note that discriminatory practices in the distribution of collect-

ive income are often frowned upon by regulatory authorities. 9 Given this,

it remains to be seen if the fully efficient sharing rule offered by Wilson is

indeed feasible in the regulatory scenario of most copyright collectives.

2.4 HETEROGENEOUS AND CORRELATEDCOMPOSITION LOTTERIES

Our simple model of a copyright collective assumes that the compositionlotteries are identical and independent. In this section we briefly examine

the implications of relaxing these assumptions.When different compositions have different probabilities of success, an

additional condition comes into play that is satisfied vacuously when prob-abilities are identical, namely, the distribution rule must satisfy the require-

ment of 'group rationality.' That is, it must not be the case that a subset of

the members prefers to defect to an alternative collective. As an example,

suppose that half of the compositions in the repertory have a high prob-

ability of success while the others have a low probability. Clearly, the equal

sharing rule introduced in the preceding section will not be viable if the

difference in probabilities is so large that those more likely to succeed prefer

defecting to a collective of their own, even though transactions costs are


then spread over half as many members as in the original collective.

However, group rationality can be restored by appropriately increasing the

share allotted to those with a high probability of success. As long as the col-

lective can categorize copyright holders according to their probabilities, the

benefits of risk pooling described in the previous section can be realized byviable copyright collectives.

Although we have focused on the case of independent composition lot-

teries, correlated lotteries would yield the same qualitative results, except in

the unrealistic case of perfect positive correlation where all compositionseither succeed or fail together and there are no insurance possibilities.

Negative correlation introduces hedging opportunities that improve the

collective's own risk situation. Indeed, with perfect negative correlation, the

collective can attain a riskless position, just as it can whenever it has access

to a perfect reinsurance market. In these cases the collective can offer its

members full and fair insurance for their composition lotteries. As correl-

ation among the lotteries and reinsurance opportunities become imperfect,

the insurance benefits a collective can offer become less valuable, but

nonetheless dominate the learn-then-distribute rule.

2.5 ASYMMETRIC INFORMATION

Thus far we have abstracted from the informational asymmetries associ-

ated with moral hazard and adverse selection. The learn-then-distribute

rule provides the strongest incentive to create a profitable copyrightable

work, since the author is rewarded for the effort to produce a successful

composition precisely in proportion to the success actually realized. In this

manner the rule addresses the problem of moral hazard, whereby a com-

poser's costly effort to produce a successful piece cannot be monitored andrewarded directly. However, the rule is an inefficient response to the

problem, since it leaves the risk-averse composer bearing all the risk of

failure when, as we have seen, insurance opportunities are available. If

moral hazard poses a significant problem for the membership of a collect-

ive, then it would be inefficient to take full advantage of the available insur-

ance opportunities, since the members would then have the least incentive

to be productive. As a general matter, an efficient response to moral hazard

entails balancing the benefits of enhanced incentives to put forth produc-tive effort against the risk-bearing costs that this necessarily entails. Anoptimal balance would call for a distribution rule that ties royalty payments

partially to a composition's realized success, offering some incentive for

productive effort, and partially to the realized success of the entire reper-

tory, offering some insurance against failure. Thus, moral hazard would


mitigate, but not eliminate the advantage we have identified for distribution

rules that incorporate risk pooling.

Adverse selection, whereby individual composers and the collective

have different perceptions of the likelihood of success, poses a different

problem. In a heterogeneous population, when individual composers are

better informed about the probability of success than the collective, risk

classification through categorization and screening, as commonly practiced

in insurance markets, would allow the collective to provide insurance to its

members, although not as effectively as when information is symmetric.10

For those more likely to be successful, screening mechanisms would pay a

higher proportion of royalties based on actual performance and a smaller

proportion through insurance. Thus, while all of the benefits of risk

pooling could not be realized, it would still be possible for members of the

collective to enjoy some of those benefits through a distribution mechanism

that does not rely exclusively on the learn-then-distribute rule.

It is also possible for the collective to be better informed about the like-

lihood of success than overly optimistic copyright holders. 11Indeed, it is

possible for the divergence in beliefs to be so great that the insurance terms

offered by the collective are unacceptable to the members. In this case the

learn-then-distribute rule would be the only viable system available to a

copyright collective.

2.6 CONCLUSIONS

In this chapter we have proposed that copyright collectives may not be

offering their members as great a welfare benefit as they could. By focusing

all of their attention on the transactions costs savings of their services, they

have ignored the opportunity to offer their members important risk-sharing

benefits. We have described several methods under which the net income of

copyright collectives is distributed to members in a manner that takes

advantage of risk-sharing opportunities and that Pareto dominates exclu-

sive reliance on a distribution rule that calls for first learning the outcome

of each composition lottery and then paying out royalties accordingly.

Two explanations may be offered for the failure of copyright collectives

to depart from complete reliance on the learn-then-distribute rule. First,

a partial explication can be found in the existence of asymmetric informa-

tion - adverse selection and/or moral hazard. In these cases, it is well knownthat the agent (the informed party, here, the creator) must retain risk in order

to satisfy incentive compatibility. Secondly, copyright holders may be unre-

alistically optimistic about their chances of success and therefore unwilling

to accept financially viable alternatives to the learn-then-distribute rule. 12


Both of these explanations imply that the full potential of copyright collec-

tives to provide benefits for their members could be realized if their

members were better informed about the value of the risk pooling services

available through the collectives.

NOTES

1 . We note that in some instances, some of the monitoring costs are borne by the purchas-

ing entity which is required to report usage rates to the collective. In any case, the col-

lective still needs to suffer some costs in processing and administering this information.

2. Blanket licensing may also serve to eliminate competition among composers and allow

them to act as a cartel. However, in many jurisdictions, tariffs are regulated, limiting the

opportunity to exercise monopoly power.3. In sections 2.4 and 2.5 we consider the implications of relaxing these assumptions.4. Really we only require that the collective has a transactions cost of t' <mt\ that is, the

collective is more efficient than the sum of its members. Setting t' =

qt, the requirementis that q< m, and in the chapter we have assumed q

=1 . In fact, the more the underlying

product (here compositions) resembles a pure public good, the lower is the value of q,

and q goes to 1 in the extreme case of a pure public good. It seems likely that for the case

at hand, we will typically have a very low value of q, and so our assumption of q= 1 is

only meant to be a rough approximation that facilitates the argument significantly.

5. Clearly the condition can be expressed equivalently as k<mt-t\ that is, the cost of

learning is less than the total transactions cost savings.

6. Indeed, participation can be achieved with a lower membership than under the learn-

then-distribute rule, since each member gets a better risk-bearing situation. Exactly howthe limit value of m is affected is impossible to determine unless we introduce an analy-sis based on utility functions.

7. Concretely, the probability is 1 .76 per cent if/?= 0.5. This compares to the figure of 46.02

per cent when fair insurance is offered.

8 . These results are not so easy to prove. See, for example, Chapter 4 of Canavos ( 1 989) for

proofs.

9. In spite of this, the American Society of Composers, Authors and Publishers (ASCAP)in particular does discriminate. They distribute income in a way that discriminates

according to how long an individual has been a member (see Besen and Kirby, 1 989 p. 2 1 ).

Also, these authors (p. 10), note that, when comparing copyright collectives over the

world 'a very wide range of methods are employed to distribute the proceeds of blanket

licences'.

1 0. See Crocker and Snow (2000) for a theoretical analysis of risk classification in the insur-

ance industry.

1 1 . The possibility that authors may be overly optimistic about the chances of success oftheir work was discussed by Watt (2000, p. 89), and also by Besen (1987).

12. We note that in some instances the creator is paid a fixed fee by a third party who then

holds the copyright. Risk pooling offers no benefits to these third parties when they are

sufficiently diversified to be risk neutral.

REFERENCES

Besen, S. (1987), 'New technology and intellectual property: An economic analy-

sis', RAND Report N N-2601-NSF.


Besen, S. and S. Kirby (1989), 'Compensating creators of intellectual property:Collectives that collect', RAND Report N R-3751-MF.

Besen, S., S. Kirby and S. Salop (1992), 'An economic analysis of copyright collect-

ives', Virginia Law Review, 78, 383-41 1.

Borch, K. (1960), The safety loading of reinsurance premiums', Skandinavisk

Aktuarietskrift, 153-84.

Canavos, G. (1989), Applied Probability and Statistical Methods, New York:

McGraw-Hill.

Crocker, K.J. and A. Snow (2000), The theory of risk classification,' in G. Dionne

(ed.), Handbook of Insurance, Boston, MA: Kluwer Academic Publishers,

pp. 245-76.

Hollander, A. (1984), 'Market structure and performance in intellectual property:The case of copyright collectives', International Journal of Industrial

Organization, 2, 199-216.

Watt, R. (2000), Copyright and Economic Theory: Friends or Foes?, Cheltenham,UK and Northampton MA, USA: Edward Elgar.

Wilson, R. (1968), The theory of syndicates', Econometrica, 36, 1 13-32.

3. MP3s and copyright collectives:

a cure worse than the disease?

Stan Liebowitz

Only one thing is impossible for God: to find any sense in any copyright law onthe planet. Whenever a copyright law is to be made or altered, then the idiots

assemble.

Mark Twain: Mark Twain's Notebook, 23 May, 1903

3.1 INTRODUCTION

The advent of massive unauthorized copying by individuals using

peer-to-peer systems and MP3 files, besides generating an enormousamount of press, has brought the affected copyright industries to the

brink of declaring war against many of their customers. This has taken

the form of a zealous attempt by copyright owners to shut down and limit

peer-to-peer filesharing, with the record industry at the forefront but the

movie industry waiting in the wings. First Napster was shut down. Then

Napster's progeny were brought to court, although the Recording

Industry Association of America (RIAA) has for the moment been frus-

trated in its attempt to have the courts shut down these latter firms. 1

Simultaneously, the record industry has been covertly distributing on

peer-to-peer networks 'spoof files that look like popular MP3 files but

contain no music, in order to hinder downloaders. The industry has also

threatened (obliquely, to be sure) to start engaging in activities that would

prevent computers from being able to download MP3 files, although it is

unclear exactly how this would be done.

Finally, the industry is in the process of the most draconian action of all:

suing users for copyright damages. It already has sued and settled with hun-

dreds of users, some of whom have set up web sites to help them pay for their

settlements. The record industry hopes that fear of prosecution will reduce

the usage of peer-to-peer networks sufficiently to help return the industry to

a better financial state. Indeed, initial statistics from comScore/Media

Metrix, the PEW Internet Project, and Nielsen NetRatings indicate a rather

37


6.0

5.5

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

Figure 3. 1 LP albums soldper capita

sharp decline of about 30 per cent in the number of subscribers to major

filesharing networks after announcement of these law suites. 2

The industry has legitimate grounds for concern. As Figure 3.1 dem-

onstrates, sales since 1999 have been on a steady decline. The decline is

considerably larger than any other decline in the last 30 years, although

the current decline is still smaller than the declines brought about from the

introduction of radio in 1921 and the Great Depression. If real revenues per

capita had been the basis of analysis, the decline from 1978-82 would be of

a similar magnitude to the current decline, although the 1978-82 decline

was, in part, cost-based, due to the transition to less expensive cassettes

from vinyl records.

For many analysts, the behavior of the copyright industries appears mis-

guided. In their view, the corporate entities that are trying to enforce current

copyright laws are seen as antiquated behemoths unable to move with the

times. Critics have suggested that the current set of institutions, including

copyright law and the firms that largely administer the business associated

with copyright, do not effectively address the interests of creators or users. 3

Partly as a backlash to the behavior of the copyright owners, many acade-

mics interested in topics associated with copyright or the Internet have argued

that an alternative to the current copyright regime is in order. The proposals

that have been offered are often subsumed under the rubric of a 'compulsory

license', although most are not technically compulsory licenses; they are in

fact more closely related to performing rights tariffs administered by the

American Society of Composers, Authors and Publishers (ASCAP) and

Broadcast Music, Inc. (BMI). A more appropriate term to describe these pro-

posals would probably be to call them variations of a 'copyright collective'.

Proponents of these proposals generally emphasize several possible

MP3s and copyright collectives 39

positive characteristics of these alternative systems-they appear to have

relatively low administrative costs, they decriminalize behavior that has

become widespread, and they might lead to increased incentive for artistic

creation. Economists writing on the subject would also mention that such

a system would allow consumption to reach a level that is more efficient

than traditional copyright. Reading some of these articles, one can get the

impression that artistic nirvana is right around the corner, if only we have

the courage to scrap the current system.

While I will discuss these proposals in more detail below, the basic idea

is that a pool of money would be generated in a secondary market (pre-

sumably related to MP3s) and transferred to copyright owners. Specifically,

the money would be generated by taxes on ancillary products, such as blank

CDs, CD writers, ISPs, stereo equipment, and so forth. Although somecommentators see a copyright collective as a supplement to the current

copyright system, it is also viewed, particularly by its more passionate

advocates, as a complete replacement of traditional copyright, at least for

recorded music.

While such a system does offer some theoretical advantages, I arguebelow that the defects of such a system have not been sufficiently examined.

Although the current system is obviously imperfect, as any system must be,

it is unlikely that a copyright collective would meet even the modest goalsof a net positive impact, to say nothing of the claims of virtual perfectionthat have been attributed to it.

Just because a system could exist and survive does not mean that it is

efficient from an economic perspective. It is the purpose of this chapter to

examine the probable efficiency characteristics of such a system so that a

better understanding of the alternatives can be gained.

3.2 HOW SUCH SYSTEMS WORK - THE CONCEPT

It is important to note that the compulsion aspect of the 'compulsorylicense' refers to an action compelling the behavior of the copyright owner,not the user of the work. A compulsory licensing scheme is one where the

government requires that copyright owners make their works available to

users, usually at a fixed or capped price. In contrast, a copyright collective

may in principle be voluntary, but often it is a foregone conclusion that

membership is essential to receive copyright payments. In that case, it is a

distinction without much of a difference.

One analogy that is often brought up is the blanket license offered by

performing rights societies (ASCAP and BMI) to radio and television broad-

casters. The blanket license grants broadcasters copyright clearance for the


music they broadcast. Mildly simplified, radio and television stations pay a

percentage of their advertising revenues to performing rights societies,

which then distribute the proceeds among their members after deducting

for operating expense. The rates paid to these organizations are set by

semi-government organizations (the courts or a government appointed

board).

This is to be contrasted with a compulsory license favorably put forward

by Lawrence Lessig which, for example, compels the copyright owners of

distant programs to allow cable operators to retransmit these signals to their

viewers without having to negotiate with the copyright owner. Instead, cable

operators pay a small percentage of their revenues, specified by statute, into

a pool (several pools, actually) and this money is then disbursed among

copyright owners with oversight by the Copyright Arbitration Royalty

Panel (CARP), a government entity.

Another compulsory license is for the use of songs on (phono) records.

Once a song has been included on a record that has been made available to

the public, then anyone else wishing to record that song can do so as long

as they pay the composer of the song the deemed compulsory license fee.

Let me illustrate how this works with an historical example. Suppose

Jimmy Webb writes a song with incomprehensible lyrics called 'MacArthur

Park'. He can sell or give the song to the highest bidder (the song was actu-

ally sung by deceased actor Richard Harris). If a song appears likely to be a

hit ('MacArthur Park' went to number one in the charts), the negotiated price

for this first recording might be very high. Once the Richard Harris recording

of the song is distributed to the public, however, the compulsory license kicks

in. Any other performers who wish to sing about cakes melting in the rain can

now also record 'MacArthur Park', if they pay the compulsory license, which

is set by fiat to approximately one half of one cent per minute, per record dis-

tributed.4 Once the compulsory license kicks in, the composer is compelled to

allow additional users to record his song at the regulated price.

Since the blanket and compulsory licenses seem to 'work', to use a phrase

often found in writings on this topic, they are often held out favorably as a

demonstration that such a system can work to replace the record industry's

traditional revenue sources. However, as I explain below, there are important

differences between these current blanket/compulsory licenses and the sug-

gested new licensing systems. Further, these new systems are more likely to

fail a cost/benefits test than is the case for, say, the traditional blanket license.

Restitution vs Replacement

There are at least two possible types of copyright collective systems that can

be imagined. In one system, the copyright collective would entirely replace

MP3s and copyright collectives 4 1

the current market-based copyright system, at least for digital products. For

those industries for which it would be used, such a system would eliminate

copyright as we currently know it. Adherents of copyright replacement

argue that the current copyright system provides few financial incentives for

artists to create products anyway, and that alternative forms of payment(such as concerts) are the true current drivers of incentives to create. In their

view, the digital world provides vast new opportunities to improve on a

poorly functioning copyright system.

The choice doesn't need to be this stark, however. It is possible to allow

the current market-based copyright regime to operate as best it can while

using a copyright collective as a form of restitution for losses in the market

due to unauthorized copying activity.

Take the case of the compulsory license of distant broadcast signals in

cable retransmission. The television broadcast market is one that generatesmost of its revenues from local viewers in a regular marketplace. By almost

anyone's calculations distant signals carriage on cable systems comprise a

minor component of this market. One can argue that in order to limit the

costs of having thousands of cable operators negotiate with hundreds of

broadcasters for relatively minor rights in distant signals, a compulsorylicense makes sense.

Similarly for the compulsory license in records, the transactions

surrounding the first recording allows the market to determine and gener-ate the value of the primary use; again, it is only the later, presumably less

valuable recordings, that can take advantage of the compulsory license.

If a copyright collective system for copying records were to work as an

add-on to traditional copyright, the goal of the copyright collection systemwould be to compensate copyright owners for the damage brought about byMP3 downloads. The system as usually envisioned would provide immun-

ity from copyright violations to private individuals engaging in MP3 down-loads and it would also provide revenues to the copyright owners through a

tax on ancillary products.

Further, the current copyright system would continue to exist in its

present form, providing copyright protection against the many forms of

copying other than file sharing. Such a proposal has recently been putforward by Neil Netanel. I refer to this as a quid pro quo system, where

copyright owners forgo revenues from peer-to-peer file sharers in return for

payments based upon taxes on ancillary products.A somewhat different alternative, and one that hasn't to my knowledge

been proposed, is to put in place a tax-and-subsidy system to recompense

copyright owners without any change in the rights or obligations of the users

with regard to the legality or illegality of their MP3 downloading activities.

The 1992 Audio Home Recording Act's provision for a tax on digital audio


tape (DAT) can be thought of as this type of system, as can the current

Canadian tax on blank CDs, which is accumulated in a pool and used to pay

copyright owners even though MP3 uploading would still remain a copy-

right violation. 5 This can be thought of as a status quo with restitution.

The Use of MP3s is Not Exogenous to the Copyright Collective

In both the status quo and quid pro quo systems, a tax and distribution

scheme is set up to compensate copyright owners for losses brought about

by MP3 downloads. There is an important difference between them,

however. The status quo system does not provide users with any new rights

in terms of making copies (or put another way, it doesn't abrogate the tra-

ditional rights of the copyright owners). The quid pro quo system, in con-

trast, essentially creates new rights to the traders of MP3 files who are no

longer in violation of copyright. The proponents of copyright collectives

clearly favor the quid pro quo system. Yet, in this latter world, we might

expect a change in behavior on the part of MP3 downloaders that weakens

the case for a quid pro quo system.

Under the current copyright regime, there is a potential cost to the MP3traders if they are discovered violating copyright. The recent crackdown by

the RIAA on downloaders is an example of increasing this cost. Not only

is there the moral conundrum involved with knowingly violating the law,

but MP3 downloaders must bear the risk of being charged with copyright

infringement.

One can predict that if these costs to MP3 users disappear there will be

a larger usage of MP3 files and a greater impact on the sale of CDs. It is

difficult to predict the magnitudes, however, since it is unclear how muchof a restraint these costs are currently imposing on the behavior of users.

But it is important not to predicate the amount of revenues necessary to be

generated by a quid pro quo system based on the amount of MP3 down-

loading that occurs before the system is in place, although that is exactly

what these proposals do.

3.3 THEORETICAL BENEFITS OF A COPYRIGHTCOLLECTIVE SYSTEM

The current evidence supports a claim that file trading is damaging the

record industry.6 Yet most MP3 traders are normally law-abiding citizens.

Many are students. The typical MP3 trader appears to consider MP3downloads to be a perfectly acceptable form of behavior. The number of

downloaders is very large, constituting an important percentage of the


population.7

It is so large that copyright supporters risk labeling a large

portion of their fellow citizens as thieves. Although a bit far-fetched, one

might imagine the downloading issue rending the fabric of the country just

as the drug enforcements efforts have. That, at least, is the more nightmar-ish scenario that has been suggested.

Under these circumstances, society could well decide that the cost of

trying to reduce unauthorized MP3 downloading under the current copy-

right regime is too great- in particular, greater than the costs of switching

to some form of copyright collective. This is particularly true given the well-

known imperfections in the copyright system, which I summarize below.

The Imperfection of the Copyright System

Intellectual products, such as stories and music, are not in and of them-

selves physical commodities. Instead, these products generally must be putinto some physical representation (records, tapes, books, etc.) before theycan be enjoyed. Without some form of intellectual property right, such as

copyright, the creator of an intellectual product would not have control

over the reproduction and sale of his creation. Anyone could reproduce andsell copies of the creator's work, and any producer of reproductions whodid not pay the creator could undercut the price that was charged by pro-ducers who did pay the creator. This state of affairs would be expected to

eliminate from the market any producers who paid the creator, leaving norevenue for the creator of the intellectual property.

Since it is unreasonable to expect creators, or artists, to expend effort

(full-time or professional effort) producing artistic works without the

expectation of concomitant payment (anymore than we would expecthome builders to build homes, or janitors to clean floors), economic analy-sis predicts that without copyright (and patent) laws, the production of

intellectual products would be far below levels that would be considered

socially desirable. 8

Even with the existence of intellectual property protection, the quantityof intellectual products created and consumed will be less than the ideal

level. This is the unavoidable underproduction and underconsumption of

intellectual properties that results from intellectual products being 'non-

rival' goods (sometimes called public goods).9Non-rivalry means that once

a work is produced it does not get used up. A physical manifestation of the

work can be used up, but not the work itself.

To start, there is a problem in the consumption of an intellectual product.There is no cost to society (other than the cost of a reproduction) in letting

an individual consume a unit of a non-rivalrous good, since there is nodiminution in the possible consumption choices of anyone else. Therefore,


an economic requirement for efficient consumption of a non-rivalrous goodis that any consumers who would like to consume it (e.g., a particular song)be allowed to consume the good as long as they are willing to pay the repro-

duction cost. Note that the only single markup over the reproduction cost

that could achieve this requirement would be a markup of zero. 10 A markupof zero, of course, would then provide no revenues to the creators (as

opposed to the producers of reproductions) of intellectual products. With

no revenues, many creators would abandon the creation of non-rivalrous

goods, which almost everyone agrees is not an efficient solution.

This is the economic paradox of non-rivalrous good pricing. There is no

practical mechanism that will produce the ideal quantity of titles and the

ideal number of reproductions for any given title.11

Charging a positive

markup will generate revenues, allowing for the creation of non-rivalrous

works, but the consumption would then be inefficient since some individu-

als with a value for the product greater than the reproduction cost will not

purchase it.

There is a further inefficiency involved with non-rivalrous goods. Non-rivalrous goods would be expected to be produced at less than the ideal

level. In an ideal world, every product that has a value to consumers that

surpasses its cost of production will be produced. For any single intellec-

tual work, even a monopoly seller of that work will not be capable of

appropriating the full potential value of that title since some consumers will

pay less than their reservation price while other consumers will be priced

out of the market. Thus, some intellectual products will not be produced,even where the potential value is greater than the cost of production. Onlya perfect price discriminator would be able to appropriate sufficient rev-

enues to guarantee the efficient production of works. 12 This inefficiency is

not restricted to non-rivalrous goods, however. Every potential product has

the possibility of not being produced due to a lack of appropriation- in

other words, there are some markets that do not exist, but would exist in an

ideal world.

These potential production-based imperfections have greater resonance

in the case of non-rivalrous goods because the markets for any particular

type of non-rivalrous good - books, movies, and so forth - consist of

individual, non-homogeneous titles, each of which is essentially a separate

market. Thus the generic market, for books say, exists, and it is easier to

imagine the harm from not having all the titles written that we might want

than it is to imagine the harm from an undefined market that does not

exist at all.

These market inefficiencies are sometimes taken as the intellectual

stepping-stones on the path to suggesting that some mechanism other than

traditional copyright be considered.


Copyright Collectives can Lead to Efficient Consumption, But Only at

a Zero Price

Since the cost of the creation of the intellectual product, the music in the

case of records, is a fixed cost, and unlimited copies can be made from this

one original, the opportunity cost to society in allowing any individual to

consume the product is merely the cost of making (and delivering) the copy.This means that society gains each time a potential consumer with a

willingness-to-pay for the product that is greater than the cost of copyingis allowed to consume the product.

As explained above, under a copyright regime, successful projects musthave prices that deter some consumers from generating surplus because the

creator must receive some payment. Consumption, therefore, will be below

the ideal level.

This logic also argues for a copyright term, if copyright is the chosen para-

digm, that is only long enough to allow the creator to receive sufficient

payment to voluntarily provide the work so as to minimize the time duringwhich consumption remains below the ideal level.

These consumption inefficiencies in the copyright system can provideammunition to those arguing for changes. If only we were willing to forgo

providing any system of payment to the creators of works, they suggest,

we could then be guaranteed that consumption of these works wouldbe efficient.

Yet few appear willing to argue that copyright owners should not be com-

pensated at all for their efforts. It has sometimes been suggested, therefore,

that alternatives to copyright, such as government grants or prizes, be used.

The current proposals for copyright collectives suggest that a tax be placedon some ancillary product, such as blank CDs, or CD burners, or perhaps

high fidelity sound equipment. The proceeds from this tax would then be

used to compensate copyright owners for their efforts.

This, of course, introduces its own inefficiency by reducing consumptionin the ancillary market that has the tax. Even ignoring the fact that the tax

may fall on the wrong individuals (not all blank CDs, for example, are used

to copy MP3 files; not all CD burners are used to burn MP3 files), by reduc-

ing the uses of blank CDs or burners (through the higher prices induced bythe tax) a new market distortion is created in these markets. If we lower the

price of the copyrighted work to its marginal cost, but make up the rev-

enues by imposing taxes in another market, we are merely shifting the

inefficiency from the market for the copyrighted work to a related market.

Even if the performers making CDs were to be paid through general

taxes, as opposed to taxes on ancillary products, there would still be the

general inefficiencies involved with the tax system. Therefore, even if a


copyright collective system eliminates an inefficiency in its home market,that is hardly grounds for its use since it creates inefficiencies in other

markets.

3.4 REAL-WORLD DIFFICULTIES

It is impossible to determine the proper size and direction of markets in any

way other than by the examination of freely functioning markets. Althoughthe difficulty of emulating markets has historically been grossly underesti-

mated, whether we are talking about centralized economic planners or justthe more prosaic regulatory bodies that abound even in market-based

economies, history has revealed that outside observers cannot divine the

characteristics of a market from alternative sources of information.

Markets arrive at results that are often very difficult to predict in advance.

Often, with enough effort, economists can come to understand why market

outcomes take the particular form that they do. A few simple examples can

be used to illustrate cases where economic theory would even have a

difficult time predicting whether the price is positive or negative, to say

nothing of how much should be paid.

Begin with a case that is often misunderstood -payola. Payola is the

pejorative term used when record companies pay radio stations to play par-

ticular songs. How would an outside analyst determine the proper paymentfor the use of the records? Typically, in the economy, firms pay for their use

of inputs. Radio uses music as its main input. Yet radio stations do not payfor the CDs that they play. Quite the opposite

- there is a history of record

producers paying radio stations in order to play their records. Further,

radio is a substitute for listening to recorded music, and the recorded music

industry would likely benefit if radio could be restricted or eliminated. 13

Although payola is a practice that governments have tried to ban, it is in

fact indicative of competition between record companies.

Similarly, many companies pay television and movie studios for product

placement, as BMW has recently done to have its automobiles and motor-

cycles featured in James Bond movies. M&M Mars is thought to have madeone of the great marketing gaffes of all time when it refused to allow the

makers of ET to use its product; instead, Reese's Peanut Butter Cupsenjoyed the benefits of global publicity from having an immensely popularextraterrestrial enjoying its products.

How would any third-party entity know whether the producers of James

Bond movies should pay for their use of cars, or whether they should get

paid by the automobile companies for having the vehicles appear in the

film? There is no simple economic test other than a market test.


Automobiles are costly to produce. Automobiles are an input in James

Bond movies. The producers pay for actors, cameras, microphones, and

most other inputs. Cars turn out to be different only because automobile

manufacturers believe that the publicity associated with the movie will gen-erate sufficient additional sales that they would benefit from giving the

automobiles to the studio. Should this belief change, then once again movie

studios would need to pay for automobiles. In a movie where the automo-

bile was represented as being unreliable, movie studios will need to pay for

the automobiles they use and are likely to use fictitious vehicles to escapethe wrath of the automobile company so characterized.

The point here is to demonstrate the difficulty of determining the most

rudimentary aspect of pricing- whether a product receives a positive or

negative price. Obviously, if the sign of the price is difficult for an analyst

to determine, the magnitude of the price is even more difficult still.

Since copyright is the subject of this chapter, it is useful to examine the

workings of some copyright markets. For example, should authors be paidfor their work, or should they pay publishers to publish their work?

Academic authors often pay journals to have their work published (the

submission fee is nominally to pay, in part, the outside 'referees' for their

time and effort but authors would pay even if referees worked for free).

Academics are willing to pay to be published because their reputationsand incomes are expected to increase when their articles are published.The author's financial gain from publication is greater than the gain of

the journal.

Similarly, many individuals would like to be published authors. It is

gratifying to have a book to show to friends and neighbors. There has

always been a 'vanity' press for individuals who pay publishers to print

their books.

If one restricted one's attention to only these vanity and academic markets

one might think that it was the norm for authors to pay publishers, that fame

was the main purpose in writing. Yet we know that the bulk of the bookmarket (measured in revenues) consists of works where a small number of

authors are paid some very large sums (both up-front and in royalties) to

produce books even though the authors often become quite famous.

If we did not have the market to point the way, how would an observer

or agency know to charge negative prices to one set of authors and positive

prices to another?

Similarly, it might be (and has been) argued that CD sales serve mainlyas a useful device for artists to increase concert endorsement revenues. It

could be the case that this was the market solution. How would a govern-ment entity know the sign or size of the correct price for payments to

recording artists except for the fact that we have the market to tell us?


Admittedly, the direction of payment is often not as difficult to predict

as in the examples I have chosen. Nevertheless, even in markets where the

direction of the payments is not difficult to determine, the amounts clearly

are difficult. I cannot over-emphasize how difficult it is to emulate the

results that come from market transactions.

3.5 HOW MIGHT A COPYRIGHTCOLLECTIVE WORK?

The overriding determination that must be made in a copyright collective

system is determining the amount of revenue (the royalty) to be raised. Thetwo secondary decisions that must be made are: (1) how the money is goingto be raised - in other words, who is going to be taxed, and (2) how the

money is going to be distributed. For expositional convenience, we can

assume that there is a 'Copyright License Board' (CLB) whose task is to set

rates and determine payouts.

Boards like the CLB usually determine amounts by having competing

experts (often economists) testify about the conditions in the marketplace.

These experts usually work for competing parties (e.g., performing rights

societies and broadcasters) who are the payers and the payees of the various

copyright royalty rates. The competing parties in this instance are likely to

be the artists and record industry versus the retailers and producers of

ancillary equipment who will bear the burden of the tax.

Determining the Amount of Revenue to Raise

The most difficult issue, in my opinion, is determining the amount of

revenue to be raised, which we can refer to as the 'quantum'. Although there

are many arguments that can be made about how one might make such a

determination - arguments based on ideas such as fairness, cultural value,

and national identity, a primary goal is often thought to be to mimic markets

since markets are thought to provide efficient solutions that maximize

economic values. Boards such as a CLB usually state that they try to emulate

markets, although this is not necessarily the exclusive criterion.

If we charge the CLB with the goal of mimicking the market, it will

become increasingly more difficult for the CLB to choose the quantum as

time goes on since the initial market values will recede into the past.

Initially, the market values will be well known since we have been in a periodwhere the market was the mechanism generating values. If, for example, the

current system were to be scrapped today, the CLB would have a fairly goodidea of what the quantum should be if it wanted to match the revenues of

9000-

8000-

g 7000-

1 6000 -

c'

5000-

4000-

3000-

2000


Real album revenue

O> O^ ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON

Figure 3.2 Album revenues

m vo r- ooON ON ON ONON ON ON ON

record companies or performers. Both of these statistics are available from

the current market participants. If it is thought that current numbers are

below what they would have been without the untoward influence of MP3downloading, the numbers could be adjusted to take this into account.

Determining the quantum in this case is relatively easy.

As we move further away from the period when values were generated in

markets, it will become increasingly difficult to divine what the market move-

ments would have been. And markets do move. Figure 3.2, for example, gives

inflation adjusted revenues in the market for full-length audio recordings

since 197 3. Note the 50 per cent increase from 1975 to 1978, followed by a 45

per cent drop from 1 978 to 1 982 and a doubling from 1 982 to 1 994. Based on

a history of CLB-like entities, it is very unlikely that such dramatic changeswould be countenanced, approved, or understood. Demographics were

largely unchanged. Income changes did not follow this pattern. Prices of

substitute products did not suddenly rise or fall. Thus, the usual explanatoryvariables that experts look at would not have predicted these changes well.

I suggested in my 2004 article (see note 6) that this increase in revenues

beginning in the early 1980s was caused by the emergence of portable

devices to play recorded music. Even if a CLB had grasped the importanceof portable devices, how would it have known to double the revenues and

not increase them by, say, 40 per cent, or 250 per cent? These very prosaic

difficulties are at the heart of the mispricing that is rampant in non-market

economies.

It is unreasonable to expect even a highly skilled CLB dominated byunbiased market experts to correctly mimic the workings of a market. The


reality is that the CLB will have to wade through reams of testimony that

is often far from unbiased and is often outside the realm of the expertise of

the members of the CLB (who are often lawyers or industry professionals).

Although I will not discuss in detail the actual performance of such boards

in this chapter, there are numerous instances of decisions indicating that

these boards, for whatever reason, have reached decisions that are incon-

sistent with basic economic principles. For example, I think it would be

agreed by virtually all knowledgeable commentators that it would be an

error in economic logic for the CLB to set the quantum to grow with the

overall inflation rate, holding the real quantum constant. This would be a

particularly egregious error if the inflation-adjusted market for related

copyrighted goods was rapidly growing at the time. Yet there have been

such decisions in the case of blanket licenses. In a similar vein, the com-

pulsory license for mechanical rights remained constant from 1909 until

1976. Is it even remotely possible that the value of such a right would have

remained constant in nominal terms for 77 years if it were market deter-

mined? Of course not.

If the copyright collective system was one of restitution instead of

replacement, one might argue that these problems would not be as severe

as implied in the preceding paragraphs. If part of the market continued to

function as a market, might not the market signals be used to provide the

type of detail I have argued that a CLB would have difficulty inferring?

There is some truth to this claim, but as conditions changed in the CDmarket, for example, it would be difficult to know whether the then current

CD purchasers represented the typical market anymore. For example, some

audiophiles claim that MP3 compression lowers the quality of the sound

relative to CDs, making audiophiles less likely to participate in MP3 down-

loads. If the market for CDs became dominated by audiophiles, who are

well-known for spending large sums on esoteric audio equipment, the

market price for CDs might rise dramatically. This would hardly be indica-

tive of an overall increase in demand, however. Thus, the problems dis-

cussed in this section would still exist, even though they might be mitigated

to some extent.

Using MP3 Downloads as a Proxy for the Market

It might be (and has been) suggested that the number of MP3 downloads

should be used to determine changes in the quantum set by the CLB. After

all, wouldn't MP3 downloads bear the same relationship to revenues as did

the quantity of records sold? Although there might be such a relationship,

there also might not be. Even if there were, we do not know whether we can

get accurate statistics on total MP3 downloads.

MP3s and copyright collectives 5 1

Not every download is actually a download. Many downloads are unsuc-

cessful or aborted for one reason or another. Not all downloaders are

equal. Some downloaders have the capability to purchase CDs and others

do not. If a large number of five-year-olds are convinced to download a

song from a television cartoon while their parents sleep-in, should that get

equal weight to downloads from adults?

Obviously, not every download represents a lost sale.

Finally, and most important, the measured number of downloads can be

manipulated at relatively low cost. What is to keep record companies, or

other organizations that represent large numbers of recording artists, from

using banks of computers to download songs? What is to keep them from

running contests to convince private individuals to download more songsthan they otherwise would? It would be relatively inexpensive to manipu-late MP3 download figures. Additionally, as we shall see below, individual

creators have incentives to inflate their downloads, also skewing measured

downloads.

Who Pays the Tax?

There are several issues with the placement of the tax (tariff). First, one

needs to determine which ancillary market to tax. Second, one needs to

determine the proper rate to charge.

As a general statement, it is correct to say that taxes distort markets. As a

matter of simple fairness, however, it is appropriate that the inevitable losses

from a tax be borne as much as possible by those who benefit from MP3downloading, and not by individuals having no association with MP3download activity. It is difficult to make a serious case that MP3 download-

ing is of sufficiently important national concern that we all should be willing

to chip in to support it. This would rule out using general tax revenues.

With this as a backdrop, the choices available are not terribly appealing.

Blank CDs or CD-writers, although likely to be used by MP3 download-

ers, are also heavily used by individuals interested in storing computer data.

There is no compelling logic to having other computer users subsidize MP3downloaders.

The purchase of audio equipment, on the other hand, is more directly

related to listening to recorded music than is the case for CDs and CDwriters. For this reason, one might argue that a tax on audio equipmentwould more precisely target music listeners, the group that seems to benefit

from MP3 downloads and thus the group that should pay the tax. There

are still variations within this group, however. Some individuals are more

interested in watching DVDs than listening to music, yet both require

similar equipment.


There is a more serious problem in the case where the CD market con-

tinues to exist. A tax on audio equipment would be paid not only by MP3downloaders but also by those who purchase CDs and do not download

MP3s. Although the continued existence of part of the CD market was a

mitigating factor in terms of determining the quantum of the tariff, here it

works to increase the distortion caused by a tax.

Finally, the proposal that is currently enjoying the most support is one

to tax ISPs. The claim is often made that if all American Internet users were

to pay $5/month, they could be allowed to download as much as they want.

Who wouldn't be willing to pay $5 a month for this privilege?

Although this is an interesting rhetorical device, there are two problemswith the logic. First, the $5 per month figure is based on the current esti-

mate of losses suffered by the record industry, and those losses are likely to

increase when everyone gets the green light to stop paying for CDs. A more

complete payment would be in the order of $20 per month.

Second, plenty of individuals would balk at paying these amounts. Manyindividuals using the Internet do not buy CDs and see no reason theyshould subsidize CD consumers.

Who Gets the Revenue?

Once a quantum of money is generated, how would a CLB determine howthe money is to be divided among the copyright owners?

One possibility is that a single copyright collective would be formed, cov-

ering all copyright owners, and that the collective would be charged with

distributing the revenues. Even if this were to be the case, however, the col-

lective would still have to decide how the revenues were distributed amongits members, and it would face the same decisions that I am going to ascribe

to the CLB in the following paragraphs. Further, when there are different

parties involved in the creation of a product (musicians, composers, record-

ing companies), it is likely that there will be multiple groups striving to get

larger shares of the royalty payments, and the Board would have to set or

approve any division that occurs. In the case of retransmission compulsory

license, for example, the Copyright Arbitration Review Panel (CARP)determines the relative amounts received by the sports claimants, the

creators of local programming (news and information), Hollywood pro-

ducers and others. In other words, the Board determines the revenues goingto different genres and it is not impossible that something like that could

happen if the recording industry were put under the auspices of such a

board.

Regardless of how the various parties are aligned, how would revenues

likely be distributed? The suggestion that has most often been made is to


use data on MP3 downloads as a basis for rewarding creators. If method-

ologies could be created to measure downloads, this view holds, and we

could award the dollars based upon these measurements.

The comparison is often made to the performing rights societies

(ASCAP and BMI) who sell blanket licenses to broadcasters and use the

proceeds to pay their members based upon estimates of how frequently a

composer's music is played on radio or television. In fact, the distribution

systems are more complex than this, often taking the length of the music

and the type of music into account when determining their distributions.

Nevertheless, the frequency of play clearly is a factor. There are difficulties

in emulating this system in the case of MP3s, however.

One fly in this ointment is that MP3 downloads are amenable to manipu-

lation, as I have mentioned when discussing the determination of the

quantum. It is possible for an individual artist to pay someone to download

massive numbers of files in order to increase that artist's share of the dis-

tributed revenues.

One might suggest that such manipulation can be prevented by some

form of user authentication that could discover instances of the same user

repeatedly downloading the same file. But how would the CLB discover

cases of massive fraudulent downloading? Perhaps there could be some

system that only allows one downloaded instance of a song for each IP

address, but I presume creative coders could find ways to fool such a system.

Might the Board be forced to engage in the same type of behavior that the

RIAA is now engaged in - trying to prevent 'unauthorized' downloading?This would be ironic, since it is largely that behavior which has created the

pretext for creating a copyright collective system to begin with.

Gaming the system in this manner is not a problem under current com-

pulsory or blanket license systems. The products upon which these licenses

are placed are purchased in the market (for example, in the case of per-

forming rights, broadcasters buy television programs, and in the case of

mechanical rights, individuals buy records) and it would be a money-losing

proposition for copyright owners to try to inflate the measured use of their

creations by purchasing large numbers of records or paying television sta-

tions to play programs containing their music. 14

A second problem in using download statistics is that downloads maynot match up well with purchases. Bands could convince their fans to

download songs for the sake of downloading. It is likely easier to get youngfans with much time but little money, to spend some time downloading

songs to help out their favorite band than it might be to get them to buymore CDs. Such results are not likely to mimic the market very well.

Finally, it is not clear that we can measure downloads very well.

Although I have seen estimates on some web sites, I am not sure how


accurate they are. Those who purport to measure such numbers often keeptheir techniques shrouded in secrecy.

3.6 COMPARISON TO BLANKET LICENSES

When the argument for copyright collectives is put forward, analogies are

frequently made to the blanket license sold by the performing rights asso-

ciations ASCAP and BMI. The purchase of a blanket license allows the

purchaser to use the entire repertoire of the association for a yearly fee that

is usually related to the revenues of the firm purchasing the license. The

prices of the blanket license (royalty rates) are not market determined but

instead are controlled by Rate Courts and the CARP. The thinking appears

to be that the performing rights markets seem to be functioning well, so

there is no reason to believe that a similar regime for the recording indus-

try might not also work well.

There are three problems with the suggestion that the performing rights

markets are an exemplar to be followed by the record industry. First, it isn't

clear how close the royalty rates that have been chosen by the Rate Courts

and CARP are to the 'correct' rate. Second, the performing rights markets

tend to be small offshoots of larger markets, so any inefficiency would tend

to be relatively small relative to the market as a whole. Third, there are often

front-end market arrangements that would serve to limit the damage done

by inefficient performing rights royalties which are not available in the case

of a replacement for the record industry. I will take each of these points

in turn.

Are Performing Rights Efficiently Priced?

Can we take the history of blanket license rates as a harbinger that a copy-

right collective will work well as a replacement for the record industry?

Are performing rights efficiently priced? The answer is that we have no

earthly idea. 15

There are several markets where blanket licenses are used. In the case of

television and radio, the two largest markets, the blanket license purchased

by the broadcaster confers only the right by the composer to broadcast the

music. Ironically, the performing right does not confer any permissions bythe performer of the music.

How would we know if the tariff rates were set too high or too low?

Perhaps we would know that the rate was too high if composers were

willing to pay broadcasters to include their music in a television or radio

broadcast. Yet, the rate could be higher than the optimal rate without


being so high that it entirely eliminated the payment normally going to the

composer of music used in broadcasts. And, of course, the rate could be

too low with no obvious way to make such a determination. If one peruses

the economic arguments that have been put before various copyright tri-

bunals, one finds no clear method to tell whether rates are too high or too

low.

A major saving grace of most performing rights tariffs is that they are

linked to the overall size of the market. The performing rights tariff rate for

radio, for example, is a percentage of advertising revenues. Therefore, the

royalty payments will change as the industry grows or declines. This assures

some modest linkage between them and is likely to keep the royalty pay-

ments from getting too far out of line.16

If the current record industry paradigm is replaced with a copyright col-

lective, however, it will be impossible to tie royalty payments to some indus-

try measure since there will be no market measures to be found. There will

be no restriction on the possibility that the royalty payments might become

much too high or much too low.

Performing Rights are an Ancillary Market

With compulsory and blanket licenses, the tariffs make up a very a small

component of the total market. In the case of television broadcasting, the

performing rights tariff is only in the range of 1-2 per cent of the total

revenue of broadcasters. In the case of radio, it is in the vicinity of

3 per cent. For sound recordings, the compulsory license only covers addi-

tional performances of the song after the initial performance, and the

original performance is likely to be the far more important market.

Similarly, for the case of a compulsory license for cable retransmission of

distant signals, the major market is the broadcast television market and

the compulsory license is a very small component. If the tariff rates were

set incorrectly in any of these markets, it would not have too large an

impact on the overall market due to the very small impact that the tariff

has on the overall market.

Why is this important? In these cases, the potential harm from replacing

a market price with a regulated price is small because these tariffs are

small. In fact, it is informative to examine the reasons why the royalty rates

were not allowed to be set in markets in the first place. In many of these

cases, it was thought that transactions costs would be so large relative to

the value of the rights as to tend to cause market breakdown. Imagine the

difficulties that individual composers would encounter trying to track

down each use of their song in a television or radio broadcast relative to

the value of the right in the broadcast of a single program. Imagine the


costs of trying to negotiate rates in each instance, or of broadcasters trying

to track down all the composers to secure the required rights. In this case,

it appears to be far better for everyone involved to have something like a

blanket license.

In principle, you could have market negotiations between ASCAP/BMIand broadcasters to set the blanket tariff rate, but a concern that the per-

forming rights societies would wield monopoly power over broadcasters

(perhaps somewhat unfounded since the broadcasters have been able to

organize into a single bargaining unit of their own) led to the requirementthat either party could appeal to the appropriate regulatory body to have

the rates set.

It is also the case that the rate is set between the seller and the buyers of

the license, with some regulatory authority as an intermediary. Comparethis to the suggested copyright collective for MP3s. The parties involved in

the setting of the rate will be the sellers of the right (record companies,

artists, composers), but the other parties are likely to be the sellers of the

products bearing the tax (ISPs, sellers of blank CDs, and so forth). Their

negotiations are unlikely to be able to lead to anything approaching a

market value since ISPs, for example, are unlikely to care how much music

is created since they are not the consumer of the music. 17

Further, MP3s substitute for the purchase of a record, and the sugges-

tion has been for a compulsory licensing system to replace either a very

large portion of the market or perhaps the entire market, not 1 per cent or

3 per cent. In this case, the copyright collective handles the primary product

being sold. In this case, any mistake will be far more serious.

Unlike Performing Rights, There is No Safety Valve

There is one other important difference that mitigates the damage from

mispricing the royalty rate in a blanket or compulsory license. These

markets often have a separate but related market that can compensate, to

some extent, for errors in the royalty rate.

The television blanket license, for example, is a back-end payment that is

made in addition to front-end payments that are most often paid to com-

posers when their compositions are commissioned for a television program.If the royalty rate is too high, the front-end payment might very well be

reduced since higher royalty rates would make composers more willing to

accept lower front-end payments. Similarly, if the royalty rate were too low,

broadcasters would need to pay higher front-end payments if they wished

to achieve the same quality of music embedded in their programs. The

front-end allows gross mistakes to be at least somewhat ameliorated. There

are other complexities having to do with risk sharing that would require a


particular front-end/back-end balance for true efficiency, but I will largely

ignore that issue at this time.

A copyright collective system that provided the revenues for the entire

recording industry enterprise has no 'front-end' that might be used to com-

pensate for an error in the royalty rate. If a mistake is made in this market,

which as we have seen is very likely, there is no safety valve to ameliorate

the problem.

3.7 CONCLUSIONS

The current difficulty with enforcing copyright in the face of MP3 down-

loads is roiling the economic covenant between music creators and music

users, threatening to overturn current institutional arrangements. Academic

writers have proposed various arrangements to assuage these difficulties.

The most popular of these suggestions has been to replace the current system

with a form of copyright collective.

From a distance, a copyright collective has some very attractive features.

One needs to examine such a system in detail, however, to truly gauge the

likelihood that its adoption might lead to an improvement.A copyright collective throws out the markers, the lighthouses if

you will, that can help guide the prices in these markets. It requires that

prices and revenues be set in some arbitrary manner. Setting prices and

revenues are the very questions that any economic system answers by its

choice of rules. The evidence of the last century has led almost all com-

mentators to agree that markets are superior at allowing consumers to

determine which goods producers produce, how much is produced, and at

providing incentives for quality improvements, compared to commandand control methods.

There are good reasons to believe that mispricing by the copyright col-

lective will be a more serious problem for the record industry than it mighthave been for other markets that have adopted some form of copyright col-

lective. And there would be no safety value in the record industry copyright

collective to limit damage.Before we throw out the baby with the bathwater, we need to investigate

more carefully the arguments that are used to support a movement awayfrom the unfettered market and toward some alternative, such as a copy-

right collective. We also need to consider other proposals, such as enhanced

copy protection, known as digital rights management. Finally, we should

not be so quick to abandon the current market. It is not yet clear howonerous enhanced enforcement of current copyright laws will turn out to

be, or whether such enforcement can feasibly protect the industry. Only as


a very last resort should we replace the current market system with a

copyright collective.

NOTES

An earlier version of this article was first published by the Progress and Freedom Foundationon IPcentral http://www.ipcentral.org/review/ vln2intro.html.

1 . After its victory in the Napster case, the industry hit a roadblock in its attempt to shut

down Napster progeny Grokster, Streamcast and Kazaa. A judge has ruled that those

non-centralized filesharing systems were little different than VCRs and thus were not

liable for the infringing behavior of their customers.

2. See for example, http://www.pewinternet.org/PPF/r/124/report_display.asp.3. There are numerous academic authors who believe the current system should be

changed. See Fisher (2003), Ku (2002), Lessig (2001), Nadel (2002), and Netanel (2002).An overview of the parties and the ideas involved in this debate can be found at the web

siteGrammy.com: http://www.grammy.com/features/2003/0725_complicenses.html.4. There are some other notification requirements that I ignore here. Also, the price is actu-

ally the higher of eight cents per song of five minutes or one-and-a-half cents per minute.

5. The 1992 Act also required that manufacturers of DAT machines restrict the machinefrom being able to make copies of copies. The DAT technology never was very success-

ful in the US whereas CD burners have become extremely common.6. For evidence on the impact of MP3 downloads on the sales revenues in the recording

industry through 2002 see Liebowitz (2004a).

7. Estimates suggest that 40 million Americans have downloaded at least one copyrightedwork. See 'Downloads to save music biz' by Jane Weaver, MSNBC, 12 August, 2002.

8. It has sometimes been argued by economists and others that copyright is not needed

at all. This argument often boils down to nothing more than a claim that the optimalduration of the copyright is less than the time advantage that one gets from being first.

See Plant (1934).

9. I used to use the term 'public goods' but there are at least two definitions in the profession,one that contains only the non-rivalrous consumption assumption and one that also

includes 'non-excludability' (the inability to prevent users from consuming the good). The

difficulty with the latter definition is that non-excludability is a function of the law and the

amount of resources devoted to excluding nonpayers, the subject of this chapter. The twoattributes are not necessarily related and I have always thought it was counterproductiveto treat non-excludability as an inherent attribute of intellectual products, since it is not.

10. I am assuming that only one price can prevail in the market and that consumers have

different valuations for the product, with these valuations arrayed down to zero. Anypositive price then excludes some potential users. If multiple prices are allowed, however,then charging each consumer just slightly below what he is willing to pay can achieve

optimal results. This is known as perfect price discrimination.

1 1 . The only way to imagine these products being produced at this ideal level is with the

imaginary cases of 'perfect price discrimination' or a perfectly omniscient governmentintervening in the market.

12. Do not confuse the claim that markets do not produce the ideal output with the claim

that markets do not produce the efficient output. The efficient output is not the ideal

output if the ideal output cannot be achieved, as Harold Demsetz reminded us several

decades ago in his classic article (1969).

13. In a recent paper, I examined two natural experiments regarding the impact of radio onrecord sales - the introduction of radio in the US and the introduction of commercialradio in England. In neither instance is there any evidence of a positive effect of radio

play on record sales. If anything, the results tend to go the opposite way. Imagine, for


example, that there was no radio. The only way to listen to music in automobiles would

be to listen to prerecorded music, which would certainly increase record sales. See

Liebowitz (2004b). For a detailed history of payola that discusses attempts by the record

industry to outlaw it, see Coase (1979).

14. Payola exists, on the other hand, to try and shape which songs are 'hits' and to increase

record sales, not to increase airtime per se to generate additional performing rights pay-ments.

1 5. For a fuller treatment of this issue see Liebowitz (2002).

16. Although the same linkage had existed in the television performing right tariff, in most

countries and most time periods, that linkage was broken in the US when the Southern

District Court of New York instituted a rule that tied royalty payments to a combin-

ation of changes in the inflation rate and the number of broadcasters, a decision that

threatens to break any linkage between economic value and royalty payments.17. Some percentage of bandwidth will be devoted to music downloads, and ISPs wouldn't

want to lose customers due to having too little music. But the losses to ISPs from lost

customers are likely to be small relative to the payments made in the tariff since most cus-

tomers have multiple uses of the Internet and the number likely to base decisions on the

availability of music is likely to be small.

REFERENCES

Coase, Ronald (1979), 'Payola in radio and television broadcasting', Journal ofLawand Economics, 22, 269-328.

Demsetz, Harold (1969), 'Information and efficiency: Another viewpoint', Journal

ofLaw and Economics, 12, 1-22.

Fisher, W.W. (2003), 'Promises to Keep: Technology, Law, and the Future of

Entertainment', draft.

Ku, Raymond (2002), 'The creative destruction of copyright: Napster and the neweconomics of digital technology', University of Chicago Law Review, 69,

263-324.

Lessig, Lawrence (2001), The Future of Ideas, New York: Vintage Books.

Liebowitz, Stan J. (2002), 'Mission impossible: Determining the value of copyright'in Ysolde Gendreau (ed.), Copyright: Administrative Institutions, CopyrightAdministrative Institutions: Conference Organized by the Centre de recherche

en droit public (CROP) of the Faculty of Law of the Universite de Montreal,11 and 12 October 2001, (Cowansville, Quebec: Les Editions Yvon Blais) 2002,

77-100.

Liebowitz, Stan J. (2004a), 'Will MP3 downloads annihilate the record industry?The evidence so far', Advances in the Study of Entrepreneurship, Innovation, andEconomic Growth, 15, 229-60.

Liebowitz, Stan J. (2004b), 'The elusive symbiosis: The impact of radio on the

record industry', The Review of Economic Research on Copyright Issues, 1,

93-118.

Nadel, Mark S. (2002), 'Questioning the economic justification for (and thus

constitutionality of) copyright law's prohibition against unauthorized copying:

106,'mimeo.

Netanel, Neil W. (2002), 'Impose a noncommercial use levy to allow free P2P file-

swapping and remixing', working paper.

Plant, Arnold (1934), 'The economic aspects of copyright in books', Economica,

May, 167-95.

4. Peer-to-peer, piracy and the

copyright law: implications for

consumers and artists

Anne Duchene and Patrick Waelbroeck

4.1 INTRODUCTION

Many voices in the music industry claim that Internet piracy has reduced

sales of original CDs and that illegal MP3 files have become a substitute to

legal CD purchases. Different from for-profit piracy, end-user piracy seems

to be much more difficult to control. The industry and policymakers have

addressed the issue by reinforcing the copyright law, by implementing tech-

nological protection and by actively enforcing legal protection through

lawsuits targeted at developers and users of peer-to-peer (P2P) networks.

On the other side of the field, advocates of online distribution technologies

have argued that MP3 downloads offer a new way for consumers to try out

new music, a typical experience good, in order to make more informed pur-

chases, which could eventually boost CD sales. The goal of this chapter is

to analyse the effect of increasing copyright protection on the pricing and

protection strategies of the firms and its consequence on consumer surplus.

We consider two distribution technologies that differ as to how consumers

acquire information on new music. Before the formal presentation of the

model, we start by an overview of legal and technological aspects of digital

music distribution and a non-technical summary of our main findings.

4.2 TECHNOLOGICAL AND LEGAL ASPECTS OFINTERNET PIRACY

Digital Music Distribution

Two technologies are currently used to download music files and to listen

to music online without the authorization of copyright owners: file-sharing

technologies and non-authorized audio-streaming technologies.

60

Peer-to-peer, piracy and the copyright law 6 1

The principle of filesharing technologies is very simple. Users run the

search engine of the software, looking for specific files. Typically, a user

types the name of an artist or the title of a particular album or song. In the

second step, the software returns 'file results' found on computers con-

nected to the file-swapping network at the time of the search. In the last

step, the user proceeds to download files directly from other users sharing

the relevant files. Note that filesharing consists not only in downloading but

also in uploading files. While downloading is always active, uploading can

be passive. Indeed, downloaded files are by default on the sharing list and

can thus be automatically uploaded (unless specified otherwise by the

user).1 In addition, users can transfer songs from an original CD into

digital format and upload them in a similar way.

Most P2P technologies include a priority rating system that provides

information about the material shared by users. For instance, the Kazaa

priority rating is a measure of how many megabytes have been uploaded

compared to how many megabytes have been downloaded over a given

period of time. It is clear that such a system benefits users who share large

popular files such as recent theatrical and pornographic movies.

Along with P2P, the Internet also gave birth to audio streaming. Audio

streaming can take two forms. On the one hand, Internet radio stations

owned by sites independent from major technological distribution com-

panies broadcast licensed music. On the other hand, specific streaming

technologies owned by large software and content providers (such as

Microsoft, Apple, AOL) have also obtained license to broadcast music

from copyright owners. Indeed, the Digital Millennium Copyright Act

requires web casters and commercial broadcasters to pay licensee fees:

these fees are set to 0.07 cents per performance with a minimum of $500 a

year; fees are collected by the Royalty Panel (CARP).2

Music as a Digital Product

Music files can be compressed without losing much information or quality,

so digital copies have a technical quality similar to the original. Digital

music files can be assimilated to public goods as they are non-rival and

non-excludable. However, the original digital product is often bundled

with other non-digital components such as a printed booklet (with lyrics,

pictures, song and artist information, etc.) and a CD case for music CDs.

Recent digital technological protection tools that we will discuss later on

can also make digital music files close to private consumption goods.

Finally, music is a typical experience good. When facing new music, con-

sumers like to spend time getting recommendations from music magazines,

listening to music on the radio or in a record store before making their


purchase decision. The traditional way for record companies to provideinformation to consumers on the existence and style of new CDs and

artists is to spend large costs on advertising and promotion. However,after the Napster experience, it has become clear that there is a cheaper

way for consumers to obtain this information: by searching, downloadingand testing digital music files made available through P2P or other file-

swapping technologies. Acquiring information is costly and time consum-

ing: downloaded files could differ from the one users expected, mainlybecause the filename has changed; the downloaded file can be badly

encrypted or there can be a download error, and this can be found out

only once the file is downloaded, which makes users waste time. As a

result, copying and use of P2P is limited to consumers with a low oppor-

tunity cost of spending time online, i.e. mainly teenagers and college stu-

dents. To sum up, we consider traditional marketing and promotions as

an information-push technology, and new filesharing networks (and to a

lesser extent audio-streaming), as information-pull technologies, where

consumers, not firms, spend time and resources. 3

The Law

The Copyright Act aims at protecting authors of 'original works of

authorship', including literary, dramatic, musical and artistic works. The

protection is available for published and unpublished works.

The copyright law includes several exemptions to copyright infringe-

ment. The most ambiguous is 'fair use' in the digital era. Four elements

have to be balanced to determine whether an activity lies within fair use:

the purpose of the use; the nature of the work being used; the amountof the work used; the effect of the use on the market for or value of the

original work.

Unlike traditional means of copying, file-sharing technologies provide a

large-scale diffusion channel that is virtually impossible to monitor, as a

single copy can be downloaded by any user across the world. So far, public

policies have addressed this new form of end-user piracy by strengthening

existing copyright laws, as illustrated by the 1 998 US Digital Millennium

Copyright Act (DMCA).4

Along with the strengthening of copyright laws has come a series of

legal actions aimed at shutting down filesharing technologies. None of

those technologies has come under more scrutiny than Napster, an online

operator which allowed registered users to exchange indexed music files in a

compressed format (MP3), and to do so freely and anonymously. Followinga lawsuit filed by the Record Industry Association of America (RIAA),

Napster was found guilty of copyright infringement. The court found that

Peer-to-peer, piracy and the copyright law 63

the use of Napster harms the music industry on two grounds: reduced sales

of CDs, and heightened barriers to entry by the music industry in the online

distribution. Napster's argument according to which sampling leads to

higher CD purchases was supported by an empirical study, but ruled out

as flawed and non-objective by the court. After monitoring activities over

new filesharing networks, the Recording Industry Association of AmericaRIAA filed 261 lawsuits charging music uploaders with copyright infringe-

ment in September 2003. These legal actions are also based on the argu-ment that music downloads are causing substantial damage to the music

industry.5 Under the current legislation, any user of P2P risks legal attacks.

The current reinforcement of legal protection against Internet piracyincreases the perception of that risk. Hence, P2P users have to take into

account the cost of being caught uploading music files. This legal cost can

be represented by the expected penalty of being caught copying.

Technological Protection

The DMCA also enabled 'digital rights management' (DRM), a small

piece of software that can detect, monitor and block (unauthorized) use of

copyrighted material. There are two types of existing DRM technologies:

watermarking (digital identification inserted in digital files) and finger-

printing (converts the files content into a unique identification number).

Despite a promising future, companies that create such technological pro-tection tools struggle to survive. 6

There are several reasons behind this failure. First, some legally pur-chased copy-protected CDs failed to play on some computers and audio

devices. Second, it is almost impossible to find protection devices for such

mass-consumption products that cannot be hacked. Third, there is the issue

of DRM standards. Finally, DRM software threatens the privacy of users.

Samuelson (2003) argues the DRM goes beyond the Copyright Act.

Indeed, DRM can protect any digital content even if it is not protected bythe copyright law such as documents in the public domain. It reduces the

value of fair use and can force consumers to view content that they do not

wish to (such as ads and FBI warnings). Because of such restrictions, DRMsometimes stands for 'digital restrictions management'. Moreover, it can

potentially protect over an infinite amount of time, which is contrary to

the spirit of the Copyright Act. In a way, DRM creates the basis for a

perpetual payment system.

Another controversial aspect of DMCA is the anti-circumvention rule

that prohibits users bypassing technological measures of protection. Manypeople involved in the copyright law have argued that DMCA and its

corollary DRM fail to strike a balance between the benefits of developers


and users and would, in addition, impede artistic and scientific progress.

Liebowitz (2002) challenges that view. Because DRM allows some form of

first-price discrimination, he argues, the use will not be reduced (however,

his analysis fails to notice that transformative and creative uses that are

authorized under fair use could be greatly reduced by DRM). Moreover,

first-price discrimination leaves no surplus to consumers and is question-

able from a social point of view if the 'intensive' music listener is a low-

income individual who would be charged more than, say, a wealthy

individual who only listens to music occasionally. Liebowitz (2002) also

challenges the view that DRM will slow 'sequential innovations' that rely

on previous work (as argued by Landes and Posner 1989). Here he argues

that DRM will not hinder such innovation because copyright only protects

the expression of ideas and not the ideas themselves.

In addition to these arguments, we would like to discuss three other

important economic dimensions of DRM. First, because DRM can be

implemented in the hardware, at the operating software level and also at

the player level, each of which are all provided by different firms, the issue

of setting standards and making sure that all platforms are compatiblecannot be neglected. For that reason, some industry observers say that

DRM stands for 'down-right messy'. Second, technological protection

software and its inventors create an additional layer between artists and

consumers, introducing a new source of vertical inefficiency. Third, if

DRM is legally mandated in players and software, this could raise the

barrier of entry in the distribution and the software protection markets

and increase market power.

4.3 A SUMMARY OF THE LITERATURE

Despite the technological breakthrough brought by filesharing technolo-

gies, the debate on the welfare implications of piracy goes a long way back

in the economic literature on unauthorized copies of copyrighted material,

reviewed by Peitz and Waelbroeck (2003 a). Instead of duplicating the

review here, we would like to discuss which arguments fit or do not fit the

characteristics of the music industry. When a copyright owner can monitor

the quantity of copies likely to result from the purchase of original mater-

ial, he or she can indirectly appropriate revenues by charging a higher price

for the original (see Liebowitz 1985; Besen and Kirby 1989; and Bakos et al.

1999). The first argument is related to the pricing of a club membershipand the nature of the cost to copy. The second argument is related to the

literature on bundling and how club formations can reduce the variance of

the demand of the club as a whole compared to individual demands.


Potentially, indirect appropriation could arise if users of filesharing tech-

nologies were be ready to pay a premium to purchase the original version

of a popular hit song in order to improve their priority rating when they

upload the files on a P2P network (for a discussion, see Liebowitz 2002).

However, both arguments are unlikely to play a key role in the case of P2P

technologies since it is extremely difficult to monitor file exchanges and

only a minority of P2P users share files. Fair use policy can involve a

tradeoff between the number of works supplied and the number of uses of

each work. Specifically, expanding fair use could reduce the number of

markets served by copyrighted works. In principle, expansion could be

more costly on both efficiency-based and non-economic policy standards.

For example, increasing the fraction of uses regarded as fair could reduce

the number of works produced with little effect on number of uses per

work, reducing total uses of all works. Such a change would move the

outcomes farther away from both overall economic efficiency and non-

economic objectives relating to volume of use of copyrighted works.

Most of the time, the copy is of lower quality than the original and

product differentiation in many models imply that the increases in con-

sumers' surplus more than compensate the static losses of producers. This

argument is easily understood since the ex-post welfare-maximizing price

is equal to the marginal cost, which is assumed to be zero. However, in

a long-term perspective, such profit loss will result in less incentive to

provide quality on the market (an important contribution to this idea

should be credited to Novos and Waldman, 1984). For the reasons that wehave already mentioned, MP3 files arguably have a lower expected value

than an original CD so that some elements of product differentiation

should be part of the debate on Internet music piracy.

In some cases, positive network externalities generated by copiescan benefit copyright owners as shown by Conner and Rumelt (1991),

Takeyama (1994) and Shy and Thisse (1999). There is a case for weaknetwork effects in music consumption if users value the number of people

listening to the same music. These social network effects can result from the

fact that consumers want to belong to a community or be able to talk about

music in social gatherings. In principle, network effects could depend onboth the number of originals and copies. There is another rationale for the

existence of network effects among copiers using filesharing technologies:

the speed of downloading music files grows along with the size of the

network. Such network effects are endogenous, since users of filesharing

technologies can decide whether or not to share their files. This decision

typically depends on the expected gain in the user's rating (making it easier

for him to download new songs in the future) and the cost of sharing related

to the probability of being caught by legal authorities or receiving a virus


and the loss of computer resources. However, if there is only a small

number of users sharing a large number of files, the extent of network

effects will be limited (for a discussion see Peitz and Waelbroeck, 2004b).

Finally, digital copies can provide information on the value of original

products that can be assimilated to an experience good. Takeyama (2003)

shows how copies with information on the characteristics of a durable goodcan solve adverse selection problems. Peitz and Waelbroeck (2003b) show

how a multi-product firm can benefit from better matching consumers to

their ideal products due to better sampling on P2P networks, despite the

negative competition effect due to the availability of digital copies. Webelieve that arguments based on the informational role of copies make a lot

of sense for music, which is a typical experience good, and we will use this

argument in our model.

4.4 NON-TECHNICAL OVERVIEW OF THE MODEL

We propose a simple model of the music industry to analyse welfare impli-

cations of an increase in copyright protection against end-user piracy when

firms can implement technological protection. The model includes three

original and relevant features.

First, we consider two means of intellectual property protection: legal

protection (set by public authorities) and technological protection (imple-

mented by copyright owners). Copyright owners can implement costly

technological protection (DRM) on original material in order to prevent

their duplication (as MP3 files).7However, technological protection alone

cannot eliminate online copying since it is always possible for professional

hackers to crack software/hardware protection, plus anti-protection devices

would be quickly available on P2P networks. Therefore, active enforcement

of copyright laws is also required. Legal protection consists of legal pro-

ceedings against copiers who must internalize the risk of being caught and

the resulting penalty.8

Second, music is an experience good. In addition to the traditional wayof targeting consumers with promotions and ads, we assume that P2P

technologies make it possible for some firms to enter the market at a low

distribution cost, as consumers bear the cost of acquiring information in

lieu of record companies.9 In short, we view traditional distribution as

an information-push technology in which the firm pays to provide infor-

mation to consumers and P2P as an information-pull technology where

consumers spend resources to acquire information on products they have a

potential interest in. Although this is a very simplistic representation of

music distribution, such a conceptual distinction allows us to analyze the


welfare implications of an increase in copyright protection according to

the information transmission technology.

Finally, we assume that a digital copy is less valuable than equivalent

original material, but at a cost that depends on the users' disutility of using

computers and the Internet to download and burn files. On the one hand,

we assume that a digital music file available on a P2P network has a lower

expected value than original material. Some users view digital copies as

poor substitutes to original CDs, mainly because they greatly value lyrics,

pictures and other information such as songwriters and album title. On the

other hand, because the amount of entertainment time is inelastic, we also

assume that consumers have different opportunity costs of spending time

looking for and downloading music files over P2P networks.

Our model contributes to the literature on welfare implications of

illegal copies on the following points. First, we endogenously assess the

level of technological protection chosen by firms, and we emphasize its

connection with legal protection. We find that increasing copyright pro-

tection gives record companies using the information-push technologyincentives to raise the level of technological protection, which increases

their profits. It is detrimental to firms using information-pull technologies.

Second, we show that stronger legal actions against copiers lead to a

smaller consumers' surplus due to the direct negative effect on copiers

and the indirect negative effect on buyers of original products through the

increase in technological protection and prices. Finally, we illustrate the

fact that an increase in copyright protection may decrease profits at

the industry level when there is a large proportion of firms producingsmall-audience music.

4.5 A SIMPLE MODEL OF INTERNET PIRACYWITH AN INCREASE IN COPYRIGHTPROTECTION

We analyze the role of an increase in copyright protection when onlytwo different and extreme forms of distribution technologies are available.

These technologies differ in the ways consumers acquire information.

Information transmission technologies. On the one hand, the information-

push technology allows firms to sell original versions of their records (with

their full set of features: CD box, lyrics, pictures . . .) at a large fixed cost

of marketing and promotions noted by K, which, for simplicity, we assume

to give information about the existence and the characteristics of the

product to all consumers. On the other hand, the information-pull (P2P)

technology allows consumers to search and test digital copies that we


assimilate to downgraded versions of originals (with a limited set of fea-

tures). Thus the P2P technology allows firms to freely distribute their prod-

ucts online but only to informed consumers, i.e. to those who have

downloaded digital copies. In other words, only consumers with a low

opportunity cost of spending time online searching, downloading and

testing digital files can become potential buyers. With both distribution

technologies, the fixed cost of production and the marginal distribution

cost of the original are normalized to zero. 10

Technological and legal protection. Regardless of the distribution tech-

nology, there is online end-user piracy, the amount of which depends on

the extent of technological protection and legal enforcement of copyright

protection. The producer can spend resources on technological protection

of originals in order to make digital copies of his or her product more

difficult to create and to find online. 11Technological protection, a, can be

implemented by the firm at an increasing and convex cost function c(a).

In addition to technological protection, original products can be legally

protected. This protection is modeled by legal fines, cp, that are determined

by the enforcement of laws such as the DMCA. Thus, 9 represents the

additional disutility of copying. In other words, <p is the expected cost of

being caught downloading unauthorized digital copies.

Music experience. Consumers incur an opportunity cost s of spendingtime searching and downloading digital music files using the P2P technol-

ogy. We assume that s is uniformly distributed on the segment [0, s\ of

mass 1 . That opportunity cost not only reflects a direct cost of spendingtime on the Internet (related to the user's level of Internet sophistication),

but it also reflects an indirect cost related to the financial value of what

the user gives up to spend time online (his or her wage or other leisure

activities, for instance). When listening to music (through promotions or

through copying), consumers can either enjoy it or not. These states of

nature occur with probability p (the consumer enjoys the music) and 1-

p

(the consumer does not like the music) respectively (with O^p^l).Copiers who like what they have listened to obtain a utility represented bythe function v(l a), whereas buyers of the original product obtain a

utility v>v(l a) for all 0<a<l. The differential between copies and

originals, av, represents the value of the original over the copy, such as

lyrics, booklet, pictures, song information, and so on. 12

Consumers who do not like the music get utility, regardless of whether

they learned the existence and the characteristics of the music through

advertising or through a digital copy, and regardless of the quality differ-

ential between copies and originals. The implicit assumption behind the

latter requirement is that consumers do not care about additional features

of music they do not like.


To sum up, the total expected cost of copying is s + 9. It is worth notingthat this cost could have taken the form of 5(1 + <p) (as in Novos and

Waldman, 1984). However, we do not consider that copiers with a higher

downloading cost s should suffer more from legal protection than others.

Thus, we assume that legal protection creates an additional burden that is

identical for all copiers.

Consumers' expected utility is defined by u(x, y), where xE {0, 1} is the

decision to purchase the original (x=

1) or not (x=

0), and yE. {0, 1 } is the

decision to download the digital copy (y=

1) or not (y=

0). We assume that

11(0,0) =0.

Information-push Technology

In this section, the firm uses the offline distribution technology with the

fixed marketing and advertising cost K. 13 We assume that advertisement

reaches all consumers who are thus aware of the existence and the char-

acteristics of the product. Hence, consumers take their purchasing deci-

sions in a perfect information environment. Consumers who like the

music (in proportion p) can purchase the original or copy the down-

graded version.

The utility of purchasing the original version is w(l,0)=

v-/?, and

the utility of downloading a digital copy is w(0,l)=

v(l -a) -s- 9. Let

s = p-q>-av denote the opportunity cost of the indifferent consumer

between the original and the copy. Consumers purchase the original if it

provides a higher utility than copying or not purchasing at all:

If the first inequality holds, consumers who like the product (in propor-tion p) purchase it if their opportunity cost of using the P2P technology is

such that s > s. This is illustrated in Figure 4. 1 .

When the price is very low, all consumers who like the product (in pro-

portion p) purchase it. This occurs when s < 0. Similarly, when the price

s s

Figure 4. 1 Consumer demand with information-push technology


is very high, no consumer purchases (s> s or/?

>v). Therefore, the firm

faces the following demand:

d(p, a)= p^^

S

ifp> v (and d(p,oi)=

otherwise), and maximizes its profit function to set

the optimal level of copyright protection a and the optimal price/?:

s sMax IT = pp ---

c(a) K<*./> s

s.t.p< v.

The optimal price/?* and technological protection a* are such that

l>

. andz* z,s

As a result, a* is increasing in <p:

da*- > O.14

A reinforcement of legal protection (as a consequence of the DMCA for

instance) thus makes firms increase technological protection. The intuition

is the following. Increasing copyright protection decreases the number of

copiers because buying becomes more advantageous. Firms can increase

prices somewhat to capture more surplus from buyers, but they may want

to also increase technological protection to prevent consumers at the

margin switching to the copy because of a higher price.

Moreover, the demand for originals and the optimal price are increasing

functions of copyright protection cp.15Proposition 4.1 describes the impact

of a higher legal protection on a firm's profit and consumers' surplus:

Proposition 4.1 Increasing legal protection cp:

(1) Increases the firm's optimal profit IT*;

(2) Decreases consumers' surplus.

Proof

(1) Applying the envelope theorem shows that


(2) First, consumers who still copy the product after an increase

in 9 have a lower utility. Indeed, (0, 1)=

v(l -a) -5 -9. Since

da*/d(p>

0, v(l -a) decreases as 9 increases, so that w(0, 1) also

decreases. Second, the utility of consumers who still purchase the

product after an increase in 9 decreases because of the higher price.

Third, consumers who switch from the copy to the original have a

lower utility. Previously they had a higher utility of copying. After

9 increases, the utility of copying and buying decreases, and theyswitch to buying. Thus, they are worse off.

Proposition 4. 1 states that a reinforcement of legal protection (as a con-

sequence of the Digital Millennium Copyright Act, for instance) makesfirms increase technological protection. Firms who distribute their prod-ucts using the information-push technology benefit from a stronger legal

protection. However, the welfare gains are lowered by the reduction in

consumers' surplus.

Information-pull Technology

In this section, the firm distributes its product using a file-sharing tech-

nology. We assume that consumers are not initially informed about the

characteristics of the product, but they can acquire information by down-

loading digital copies. Consumers who have copied the product and

enjoyed the music have the possibility to directly purchase the original

from the firm. The expected utility of downloading the digital copy is

w(0, l)=

pv(l -a) -s- 9 (the probability of liking the music being p).16

The consumer who is indifferent between downloading a copy or not is

such that H(0,l)=

w(0,0)<=>,s=

pv(l -a) -9. Hence, only consumers

with a low search cost and who like the music can become potential

buyers. On the one hand, copiers who did not like the music do not pur-chase the original version, as their utility of purchasing the original

product would be-/?-s-<p<0. On the other hand, consumers whohave copied the product (of mass pv(l

-a)-9 /s) and who liked it (in pro-

portion p) have an expected utility w(l, 1)=

p(v(l-a) + va -p)

- s - 9.17

It

is clear that copiers who like the music will purchase the original providedthat /?< av. Indeed, copiers who like what they listened to have an extra

utility av-p of purchasing the original instead of just copying. This is

illustrated in Figure 4.2.

When legal protection is very strong, there are no copiers, and therefore

no potential demand for originals. Even with a strictly positive potential

demand, no consumer purchases when the price is very high. Therefore, the

firm faces the following demand:


s s

Figure 4.2 Consumer demand with information-pull technology

if /?<av (and d(p,a)=

otherwise). This leads to the following optimiza-

tion program:

f .Max TT = --c(a)

a,P 5

s.t.p<av and <p<

pv(l a).

Here the level of technological protection, a, is a decreasing function of

copyright protection cp.18

Increasing copyright protection reduces the

potential demand for original products distributed online, as potential

buyers need to download a digital copy first and this activity is more costly

when copyright protection increases. In order to compensate this loss, it is

optimal to reduce the level of technological protection. Overall, profits also

decrease and so does buyers' surplus.

Proposition 4.2 An increase in copyright protection:

( 1 ) Decreases the profits of firms distributing their products using P2P;

(2) Decreases the social loss due to the downloading cost;

(3) Decreases buyers' surplus.

Proof

(1) Applying the envelope theorem shows that

dix* ova

~T~= ~--'

fl(p 5

(2) We first show that

S^ *a

acp


Straightforward calculations yield:

ds da v/5

Substituting the expressions found in note 19 and rearranging

shows that ds/dy < 0. A reinforcement of the copyright law from

cpjto cp2 (with <p2 >(pj) makes a proportion sfo^ s(<p2) stop copy-

ing. There is a reduction of the number of copiers, which means that

there is a smaller proportion of copiers who waste resources on

copying without enjoying the product (1-

pX-stoj)-

.?(<P2)].

(3) The surplus of copiers who become purchasers is given by

/6= <;(pv[l-a'

o

After rearranging, we obtain that. ;-

-<x*(p)]

-J -pv

-

The expression in between brackets is negative. Thus Ubis decreas-

ing in <p.

The intuition behind Proposition 4.2 is that an increase in legal pro-

tection reduces the demand for copies and thus the potential demand for

originals. Since the firm extracts all buyers' surplus, consumers only obtain

the surplus from the copy.

The main difference with the previous distribution technology is that

there is always both a benefit and a cost of copyright protection on con-

sumers' surplus. The reason is that increasing 9 reduces the number of

copiers and thus both the number of copiers who waste resources on down-

loading music they do not like and the number of copiers who become

buyers of the original product.


Application: Effect of an Increase in Copyright Protection on the

Profits of the Industry

In this section we assume that C(OL)= ca2

/2.

When the firm distributes its product with the information-push technol-

ogy, it sets a price

+ _ cs(y + s)

2cs v2p

and a level of technological protection

2cs v2p

(the denominator is positive by the second order condition). This yields a

profit of

cp(9 + s)2

1 4cs 2v2p

using the information-push technology.

When the firm uses the information-pull technology, it sets a price

p* = a*v and a level of technological protection

^ = vp(vp-

9)

It is easy to check that the optimal profit is

~<P)

2

22s(cs + 2v2p2)<

In Figures 4.3 and 4.4, we compare the profits using both distribution tech-

nologies for two different legal environments (<p= 0.15 and <p

=0.30). The

other parameters are set to s = 1,v = 1

,K= 0.45, c = 1 .

19 The flattest curve

corresponds to the profit obtained by using information-pull technology,

and the steepest curve to the profits of a firm using the information-push

technology. The trade-off between the two distribution technologies lies in

the potential audience of the music, p.

For any given values of <p, music with a low potential audience p is

not distributed, and firms with an average potential audience choose the

information-pull technology, although they could have used the other.

0.2-

0.1-


Information-push

Information-pull

0.2 0.4S 0.6 0.8 1

-0.1 -

-0.2-

Potential audience

Figure 4. 3 Application with cp= 0. 75

0.6i

0.5 -

0.4-

0.2-

0.1 -

Information-push

Information-pull

-0.10.6 0.8 1

Potential audience

Figure 4.4 Application with q>= 0.30

For instance, in Figure 4.3, firms with 0.8<p<0.9 use the information-

pull technology although the traditional distribution technology was prof-

itable. Finally, firms producing music with a large audience prefer the

information-push technology.

Increasing <p from 0.15 to 0.3 has several noticeable effects (see Figure

4.4). First, the range of p that supports the information-push distribution

technology increases. Second, more firms prefer the information-push tech-

nology. Third, the minimum audience which makes it profitable to enter

the market with the information-pull technology increases too. Thus there

are fewer firms in the industry and a more important proportion chooses


the information-push technology. The aggregate effect of an increase in

copyright protection at the industry level depends on the distribution of

audience p. When there is a high proportion of superstars, an increase in

copyright protection leads to an increase of industry profits. However,when there are proportionally more small-audience artists, industry profits

might fall.

4.6 CONTRIBUTION, AGENDA AND CONCLUSION

We have argued that information-pull technologies offer a possibility of

entry through cheap distribution channels, and open the market to newartists and small distribution companies. We have shown that a stronger

legal protection can shift too much the balance between consumers and

producers in favor of producers of large audience music. In a way, these pro-

ducers could use legal protection as a foreclosure tool against new products.

However it is foreseeable that as a growing number of new productsbecomes available, consumers will ignore the existence of one particular

product in which they could have a potential interest. We believe that newdistribution technologies could open the market for informational inter-

mediaries. Indeed, in our model, both distribution technologies are

inefficient at transmitting relevant information. On the one hand, produc-ers who use information-push technologies waste the fixed cost of adver-

tisement and promotion on reaching consumers with no interest in their

music. On the other hand, some P2P users waste resources on download-

ing music files they finally do not like. Informational intermediaries could

reduce these inefficiencies, by promoting and recommending new products

(a system already available on Amazon.com and similar Internet sites), tar-

geting only consumers who have the highest likelihood of purchasing the

original. Some of the new emerging business models are based on cross-

platforms, like Kazaa for instance, who work together with record compa-nies as a new vehicle to reach mass markets. As P2P plays the part of an

advertiser, record companies are becoming simple distributors. MP3.comhas been using their data on consumers to provide record companies infor-

mation and advice on their marketing strategies.

Other business models improve the quality of the original by offering a

number of features that could not appear on the copy: a booklet contain-

ing lyrics, pictures, or an access code to online chat rooms, forums, making-

offs, additional music and a discount on live performances, etc. Indeed, the

possibility of offering an Internet connection with login and password to a

site with additional features for purchasers of original products reduces

the value of a digital copy. In fact, the emerging new business models are


beginning to take into account other assets than just the music. The mar-

ketplace is thus changing from a commodity market to a service one, as

illustrated by MP3.com, who offer complementary services to the music,

which represents additional value for users.

ACKNOWLEDGMENT

We wish to thank B. Caillaud, G. Llobet, P. Pereira, J. Pouyet, M. Bourreau, and session

participants of the 2nd ZEW Conference on the Economics of ITs in Mannheim and of the

2nd Annual Conference of the Society for Economic Research on Copyright Issues,

Northampton, MA, of the 29th Annual Conference of the European Association for

Research in Industrial Economics (EARIE), Madrid and of 18th Annual Congress of the

European Economic Association (EEA), Stockholm.

NOTES

1 . For instance, eMule automatically shares files that are being downloaded.2. According to an Arbitron/Edison Media Research study, over 35 per cent of American

Internet users aged 12 and older were 'streamers' in July 2002. However, the trend is flat

mainly because of the shutting down of small webcasters who were not able to pay roy-alties to copyright owners.

3. Recent survey data show that a large share of Internet users download files for sampling(69 per cent of downloaders listen to new music and 31 per cent listen to music by artists

they never heard before according to PEW Internet tracking and IPSOS-INSIGHTreports). Moreover, a survey of IPSOS of 2002 finds that 73 per cent of music down-loaders 'like being able to sample music online before making a purchase decision'.

Another survey of IPSOS of 2002 finds that 30 per cent of Internet users have changedthe genre of music they typically listened to, mainly because they were able to experi-ment with different forms of music. See Peitz and Waelbroeck (2004b). We believe that

these numbers understate the potential of file-sharing systems for sampling purposesbecause cross-recommendations and profiles of downloaders could be greatly improved.

4. The DMCA makes it a crime to circumvent anti-piracy measures built into most com-mercial software (except for research purposes, non-profit libraries, etc.). While the

DMCA limits liability of copyright infringement of Internet service providers and insti-

tutions of higher education, it requires webcasters and commercial broadcasters to paylicensee fees. Finally, the DMCA does not affect conditions of copyright infringements,

including fair use. The DMCA lays the legal foundation of pay per use even for mater-

ial that is no longer protected by copyright and so on a perpetual basis.

5. This argument is only partly validated by cross-country data analyzed by Peitz andWaelbroeck (2004a) who find that while music downloads may have reduced music sales

by as much 20 per cent in the early days of P2P networks, other factors are also playingan important role.

6. MIT Technology Review (March 2002) comments on the struggles of ContentGuardand TrustTechnologies, two important producers of DRM solutions.

7. DRM alone is not sufficient per se to secure digital content, and has to be coupled with soft-

ware or hardware solutions. Eckersley (2003), citing sources from IBM's Thomas J. WatsonResearch Center and extrapolating from the cost of a 4758 cryptographic co-processor,estimates the current cost of a complete security device around US$4000 per computer.Further R&D could reduce this figure to US$1 00- 150 on a large-scale production basis


and possibly to US$1 0-20 if it is possible to develop a single-chip device. In any case, the

cost of developing secure technological protection is far from being negligible.

8. Our model also applies to audio streaming where we analyze the effect of increasing roy-

alties paid by webcasters. Fewer webcasters means less music variety and less consumer's

surplus.

9. When Napster created its 'New Artist program', which enabled starting bands to set upa hyperlink to a merchant site selling their original product, more than 10000 profiles

were submitted within the first two weeks (see The Standard, 19 June, 2000).

10. Even with a strictly positive cost of creation, our results would not be affected, since we

compare profits between technologies and between different legal environments, which

do not affect this cost. Moreover, the marginal cost of distributing information goods is

negligible compared to the fixed marketing cost.

11. In another more elaborated version, we allow technological protection to decrease the

value of the original as well, through a reduction of its fair use value. The results pre-

sented in this simpler version are qualitatively similar.

12. This parametric example corresponds to the specification used by Besen and Kirby

(1989). Note however that in their model a is exogenous, while this parameter will be a

choice variable in our model.

1 3. For simplicity, we do not make a distinction between the artist and his label. The implicit

assumption is that it is possible to write a contract that eliminates vertical inefficiencies.

14. Applying the implicit functions theorem to

we get

da _ pv/2s~

Since the denominator is negative by the concavity of the profit function, da./d(f>Q.

d<f 2\ d(f> I

'**

dcp 2\*

d<p

16. The process of purchasing the original requires the availability of a connection to an

Internet site or file-sharing software, which then proposes a connection to a merchant

site. We thus assume that consumers cannot buy random products for which the fixed

costs of marketing and promotion K have not been spent. In other words, productswhich have not been distributed traditionally are not available in brick-and-mortar

stores and can only be discovered through a time-consuming online search process.

17. It is worth noting that in this section a consumer might have both versions (copy and

original) of the same product, contrary to the last section. This is due to the fact that the

informational structures are different. Since consumers are initially uninformed about

the product, they first have to test the product before knowing whether they like it or not.

Thus their actions are sequential. Moreover, since they are initially uninformed, they do

not know if the music that they download is copyrighted or not. For that reason, they

still anticipate an expected fee, although as we will show, artists using information-pull

technologies have no incentives to sue them.

1 8. At the optimum, the effect of increasing a on profit is given by

F(a,<p)= =

-_ [pv(pv(l-a)

-<p)]

-P2av2 + c'(a) .

Applying the implicit function theorem to F(a, cp)=

0, we obtain that

da. _ pv/s

d(f dF/dot


Since the numerator is positive, it remains to show that the denominator is negative. Wehave:

= _ 2v2 -c"(a) < 0. Thus < 0.

da s dq>

19. We could not find any permissible configurations of the parameters that yielded other

qualitative results.

REFERENCES

Bakos, Y, E. Brynjolfsson and D. Lichtman (1999), 'Shared information goods,'Journal ofLaw and Economics, 42 (1), 1 17-55.

Besen, S. and S.N. Kirby (1989), 'Private copying, appropriability, and optimal

copying royalties,' Journal ofLaw and Economics, 32, 255-80.

Conner K.R. and R.P. Rumelt (1991), 'Software piracy- an analysis of protection

strategies,' Management Science, 37 (2), 125-39.

Eckersley, P. (2003), 'The economic evaluation of alternatives to digital copyright,'

working paper.

Landes, W. and R. Posner (1989), 'An economic analysis of copyright law,' Journal

of Legal Studies, 18, 325-63.

Liebowitz, S. (1985), 'Copying and indirect appropriability: Photocopying of jour-nals,' Journal of Political Economy, 93, 945-57.

Liebowitz, S. (2002), Rethinking the NetworkedEconomy, New York: Amacom Press.

Novos, I. and M. Waldman (1984), 'The effects of increased copyright protection:An analytic approach,' Journal of Political Economy, 92, 236-46.

Peitz, M. and P. Waelbroeck (2003a), 'Piracy of digital products: A critical review

of the economics literature,' CESIfo working paper #1071.

Peitz, M. and P. Waelbroeck (2003b), 'Making use of file-sharing in music distribu-

tion,' mimeo.

Peitz, M. and P. Waelbroeck (2004a), 'The effect of Internet piracy on music sales:

Cross-section evidence,' Review of Economic Research on Copyright Issues, 1 (2),

71-9.

Peitz, M. and P. Waelbroeck (2004b), 'An economist's guide to digital music,'

CESIfo working paper #1333.

Shy, O. and J.F. Thisse (1999), 'A strategic approach to software protection,' Journal

of Economics and Management Strategy, 8 (2), 63-190.

Samuelson, P. (2003), 'Digital rights management and, or, vs. the law,' Communi-cations of the ACM, 46 (4), 41-5.

Takeyama, L.N. (1994), The welfare implications of unauthorized reproduction of

intellectual property in the presence of demand network externalities,' Journal

of Industrial Economics, 42 (2), 155-66.

Takeyama, L.N. (2003), 'Piracy, asymmetric information and product quality,' in

Wendy Gordon and Richard Watt (eds), The Economics of Copyright:

Developments in Research and Analysis, Cheltenham, UK and Northampton,MA, USA: Edward Elgar, pp. 55-65.

5. 'Fair use5

as policy instrument

Timothy J. Brennan

5.1 INTRODUCTION

Battles over copyright are typically framed as between providing moreaccess for the public to created works and giving profits to those whocreated the works. To this observer of copyright policy discussions, grant-

ing any weight to the latter side of this tradeoff often seems to come grudg-

ingly. It is as if only a minimal degree of fairness justifies consideration of

creators in the balance at all.

These battles between access rights and exclusion rights take place on

many intellectual property fronts. Seemingly innocuous or obvious busi-

ness practices have been patented, leading to concerns that they will be

unnecessarily monopolized and promote collusion (Ciminello 2000).l

Cases involving MP3 music filesharing and DVD decoding cases suggest a

conflict between listeners and producers regarding access to music andfilms and on broadband deployment.

2 The DVD cases also reflect contro-

versies regarding the Digital Millennium Copyright Act's restrictions onaccess to decryption technology.

3 The Supreme Court ruled against con-

tentions that extending copyright an additional twenty years violates the

Constitutional mandate '[t]o promote the Progress of Science and useful

Arts by securing for limited Times to Authors and Inventors the exclusive

Right to their respective Writings and Discoveries.'4

One facet of these controversies is whether viewing copyrighted worksas special, e.g., as 'cultural' goods, should change copyright doctrines, par-

ticularly to expand 'fair use.' Fair use is defined statutorily as the right to

copy or use copyrighted works in ways that would otherwise lead to liabil-

ity for infringement, based on four factors: purpose and character of

the use, nature of the work, amount of it used, and effect on its marketor value. 5 The operational content of these factors continues to evolve. 6

Policy goals other than economic efficiency, e.g., maximizing access (Brown1985),

7'public knowledge, political debate, or human health' and First

Amendment considerations (Gordon 1982), or relational goals threatened

by commodification (Gordon 2002) might warrant more generous fair use

80

'Fair use'

as policy ins trumen t 8 1

standards or more radical changes in copyright. The National Research

Council's Computer Science and Telecommunications Board (2000,

pp. 137-8) has observed that fair use might be justified by 'fundamental

human rights' of freedom of expression and the press, and 'public interest

grounds' associated with schools, libraries, and the courts. 8

Models below illustrate modifying fair use to promote objectives other

than aggregate economic efficiency. Alternatives include disregarding profits

to copyright holders, recognizing only gross consumer surplus, or counting

uses of copyrighted works without regard to their different economic valu-

ations. We look at fair use when there may be 'diversion' in that some uses,

for which users could pay and could be identified, may nonetheless be

obtained for free. One might be willing to pay for a copy of a CD to carry in

one's car, but one does not need to do so. In the spirit of justifications based

on avoiding transaction costs for low value uses, we also model fair use as

allowing uses for which there is low willingness to pay.9

The motivating question is whether different policy objectives should

significantly affect copyright. The intuition as to why they might not is

as follows. Suppose, to be simplistic, that a valid proxy for looking at the

effects of copyright is the product of the number of works times the

number of uses per work. By any credible policy standard, one would want

to increase both the number of works and uses per work. One would be at

an optimum only if increasing the number of works leads to a decrease in

uses per work or vice versa. If that tradeoff is not much different under

different goals, e.g., economic efficiency vs. maximizing uses, argumentsabout the basis of copyright policy are excessively heated.

Markets for copyright involve too much detail to make precise recom-

mendations regarding fair use, compulsory licensing (e.g., as under the

Audio Home Recording Act of 1992), term limits and breadth (including

reach over derivative works). Consequently, this research is only a step

toward understanding how much policy objectives matter. Still, narrowly

structured, inherently unrealistic models may yet offer insight into how a

change in objective may lead to a change in policy.10 Such limitations may

be appropriate. Lacking more specific empirical justifications, copyright

policy rules are likely to be relatively and appropriately simple, e.g., in not

applying different fair use standards to a highly popular film compared to

one with a small audience.

5.2 PRIOR ECONOMIC ANALYSES

Formal models of the design of copyright have focused primarily on the

extent to which copying should be made more or less costly (Novos


and Waldman 1984; Johnson 1985; Besen and Kirby 1989; Yoon 2002).

Analyses of fair use typically attempt to justify it when market failures

impede otherwise desirable uses rather than to assess its effects. The fun-

damental insight, from Gordon (1982, pp. 1614-22), is that some con-

sumers may be willing to pay the prevailing price to obtain or use a

copyrighted work, but not be willing to pay that price plus the cost of the

transactions associated with those uses. 1 1 Brennan (1986) takes a more nar-

rowly economic approach, treating fair use as a compulsory license where

the fee is zero, building on the transactions cost-based analysis in Besen

etal. (1978).

From this standpoint, Klein et al. (2002) claim a copyright holder would

not object to fair use, since its profits would be unaffected. An importantlimitation of Klein, et al.'s insight, however, is that it ignores the other three

statutory factors, beside effect on market value, associated with fair use. 12

Copyright holders and user groups could dispute whether the nature and

use of the work in news or educational settings justify unauthorized

copying, even if there is a non-negligible effect on the work's market value.

Some analyses have taken a more strategic approach. Depoorter and

Parisi (2002) suggest that fair use can limit incentives for strategic with-

holding when copyrighted works are complements.13 In essence, each

copyright holder has an incentive to hold out for the full value of a

package of copyrighted works absent some compulsion to make the work

available. 14 In an article not specifically on fair use, Takeyama (1997) offers

a model in which fair use available to those with low reservation prices

could allow a copyright holder to credibly commit to charge high prices

in a multiperiod setting. Absent fair use, the copyright holder could not

commit not to offer a low price in later periods to low reservation price

users, reducing the amount it could extract from users with high reserva-

tion prices.

A noteworthy implication of arguments based on transaction costs is

that if those costs fall, the rationale for fair use defenses against accus-

ations of infringement also falls. As the movement of creative works

becomes increasingly digitized through the Internet, it becomes increas-

ingly feasible for copyright holders to find and charge for uses that mightnot have been worth identifying in the analog environment (NationalResearch Council 2000, p. 136). Advocates of fair use view such develop-ments as reducing free access to works below what users have come to

expect, e.g., in educational and research settings (National Council for

Higher Education, 2003). These discussions only highlight that there mightbe more than economic efficiency at stake in the design of fair use. For

example, greater fair use of general-purpose works in educational settings

might be justified as redistribution of wealth away from copyright holders

'Fair use' as policy instrument 83

of those works to help equalize educational opportunities for all. The

models below do not incorporate such distributional factors directly,

although some fair use settings we do examine, in particular those that

maximize only consumer welfare, or volume of uses of works, could speak

to these considerations.

5.3 THE ECONOMIC MODELS

We look at copyrighted works as a series of monopolies. Each has separ-

ate, independent demands, created at a fixed cost independent of actual

use. As more works are supplied, the 'marginal' work is valued less (has a

lower demand curve) then the previous one. We rule out substitutability

among works from a consumer perspective, e.g., that the availability of

work X reduces willingness to pay for work Y. 15 We also rule out comple-

mentarity in consumption or production, e.g., that availability of work Xincreases demand for Y, or that access to work X reduces the cost of pro-

ducing work Y 16 These assumptions abstract away from complications

and arbitrariness inherent in differentiated product models. As the point

of the models is to suggest the relative position of optimal amounts of

allowed (fair) use under different policy goals rather than to estimate spe-

cific quantitative fair use standards or to look at changes in external con-

ditions (e.g., new distribution technologies), these simplifications may not

be inappropriate.

Fair use policy can involve a tradeoff between the number of works

supplied and the number of uses of each work. Specifically, expanding fair

use could reduce the number of markets served by copyrighted works. In

principle, expansion could be more costly on both efficiency-based and

non-economic policy standards. For example, increasing the fraction of

uses regarded as fair could reduce the number of works produced with little

effect on number of uses per work, reducing total uses of all works. Such a

change would move the outcomes farther away from both overall economic

efficiency and non-economic objectives relating to volume of use of copy-

righted works.

We treat 'fair use' parametrically in two ways: varying the fraction k of

uses available for free at any value, and varying a reservation price ceiling r

below which uses can be obtained for free. We refer to the first as the 'equal

diversion' case, and the second as the 'reservation price' case. In both, we

show how this parameter should be adjusted to address the breadth of

policy goals that might be considered, to see which direction fair use might

be adjusted and the conditions under which substantial adjustments are

more likely to be warranted.


5.4 THE 'EQUAL DIVERSION' SETTING

We index the set of actual or potential copyrighted works by the param-eter n. For tractability, we assume that as n increases, demand D(p,n) for

that work at price/? declines, i.e., Z)/J<0. 17 For convenience, we assume that

each of these works can be created at a fixed cost c, constant across all,

with no marginal cost for individual uses. 18 With demand for each work

independent from the demand for any other, an owner of a copyrighted

work n would set price/? to maximize profit !!(/?,)

Tl(p,n)=pD(p,n), (5.1)

where

pDp(p,n)+D(p,n)=Q.19

(5.2)

Let /?*() be the profit-maximizing price for work n as defined by (5.2), and

let n*Cz)=II(/?*(),) be the maximum profit received by the copyright

holder for work n. 20 From (5.1), (5.2) and the envelope theorem,

11*' CO =P*(n)Dn(p*(n), n) < 0. (5.3)

Profits fall with n because Dn<Q. The number of works n* with no fair

use is determined where the profits just cover the cost of production, i.e.,

where n*(*) = c.

The Fair Use/Number of Works Tradeoff

We first look at fair use that by intent or effect allows some fraction of uses

of copyrighted works to take place without having to pay the copyright

owner. For simplicity, assume that this diversion occurs uniformly over

the demand curve. Let A;E[0, 1] be the fair use parameter, such that the

amount of uses that the copyright owner of work n can sell at price /? will

be (1-k)D(p,n). The quantity of fair uses that takes place will be kD(Q,n).

An advantage of the assumption that k is uniform over the demand curve

is that the price that maximizes profits will be determined by (5.2), inde-

pendent of k. Thus, the extent of fair use has no effect on prices of copy-

righted works that are sold. The amount of uses of work n with fair use k,

USE(n,k),is

USE(n,k) = (1 -k)D(p*(n\n) + fc/>(0,/i). (5.4)


At fair use intensity k, works will be created up to the point where profits

of the marginal work N just cover creation costs, i.e.,

This implicitly defines the number of works as a function of the fair use

regime, N(k). From (5.5), (5.3), and the definition of profit, the derivative

N'(k) is given by

N'(k) =(l-k)H*'(N)

(\-k?p*(N)Dn(p*(N),N)

D(p*(N),N)

(1 -)/)(/>*(#),A^' (5.6)

The fraction is negative because demand is positive and the denominator

is negative by assumption. Equation (5.6) not only defines N'(k) but pro-

vides the basis for the assumption in the derivations below that N'(k)<Q,

i.e., that fair use reduces the number of works created. The magnitudes of

N' and Dnare inversely related. If reducing the number of works just a little

bit would show a large increase in demand for the marginal work, N will

only fall a small amount for a given increase in the degree of fair use.

Illustrating Policy Objectives

Figure 5.1 illustrates potential policy objectives for fair use for a particular

work n. The key for any case will be to recognize tradeoffs. Increasing k may

C cl(\-k)D(p*(n),n)

(\-k)D(p*(n\n) (1 -)D(0, n)

Uses of work n

, n)

Figure 5. 1 Consumer surplus, profits, and costsfor work n: 'equal diversion


increase the contribution to the policy objective from work n against the

reduction in the total number of works N(k) that would be supplied.

In Figure 5.1, Area 1 corresponds to the consumer surplus of those who

pay p*(n) to use work n. Area 2 is the consumer surplus from the kD(0,n)

fair uses. Area 3 is the profit to the creator of work n. Area 4 is the cost c

of producing work . Finally Area 5 is the forgone potential surplus from

those unable to obtain a fair use of work n and unwilling to pay p*(n) for

it. Were the supply of work n guaranteed, we would want to set k 1,and

set Area 2 to the entire consumer surplus, while Area 5 disappears. We can

now look at setting k to maximize different social objectives.

Number of Works

Let kN be the level of k set to maximize the number of works TV, e.g., to

encourage creativity without regard to other objectives. With this objective,

kN =Q, i.e., have no fair use, since N' <0.21

Net Economic Welfare

From Figure 5.1, the contribution to net economic welfare from work n

with fair use k is the sum of Areas 1, 2, and 3. To reduce some of the clutter

in the notation, define the consumer surplus function CS(p, n) of work n

at price /? as

CS(p,n)="JD(p,n)dp.p

Area 1 is (1-k)CS(p*(n\n)\ Area 2 is &CS(0,); Area 3 is p*(n)(\ -k)

D(p*(ri),n) c. Aggregated over all N(k) works supplied, net economic

welfare NEW(k) is

N(k)

NEW(k) =I

[(1 -k)CS(p*(n\n) + kCS(0,n)

o

+ /?*()(! -k)D(p*(n\ri)]dn-cN(k).

Recalling from (5.5) that the profits of the marginal work just cover its cre-

ation costs, the degree of fair use, kNEW,that maximizes NEW is that which

satisfies the first-order condition

J/ -\r ri TT/- \ ^(K); = f [CS(0,/i)

-CS(p*(n),n)

-P*(n)D(p*(n\n)}dn

dko

0. (5.7)

'Fair use'

as policy instrumen t 87

The integral is the increase in net economic welfare from each work supplied

with more fair use; the net gains to consumers from getting more free uses

less the lost profits to creators. The second term is the lost economic welfare

from the works not produced as fair use is expanded. Because profits from

the marginal works produced are zero, losses are borne entirely by con-

sumers of those works.

Aggregate Consumer Surplus

A third potential standard for fair use policy would be to consider costs but

to leave profits to creators out of the social benefit calculation. In Figure

5.1, this would be to focus only on Areas 1 and 2, leaving out Area 3, profits

to creators. If we let ACS(k) refer to this measure,

IN(k)

I

f p*(n)(l-k)D(p*(n),n)dn

-cN(k) .

Since profits from the marginal work just cover creation cost,

^ = /}WW^^. (5-8)

The term on the right in (5.8) is the effect of increased fair use on profits

to inframarginal works. This was subtracted out in calculating net eco-

nomic welfare. At kNElv, where NEW is maximized, ACS will thus be

increasing in k.

Accordingly and not surprisingly, the level of fair use that maximizes

consumer surplus, kACS,will be greater than kNEW. The lost profits to copy-

right work holders would not matter. The difference between kACS and

kNEW, i.e., the degree to which fair use should be increased if we focus solely

on benefits to users, depends on the magnitude of the integral in (5.8). This

is larger the greater are prices at which copyrighted works are sold. If aggre-

gate consumer surplus is the objective, fair use should be stronger the

higher are prices for copyrighted works.

Gross Benefit

A fourth potential objective might be the value of the uses to consumers of

the works, neglecting both profits and costs except to the degree that they

influence the number of works available. This would comprise Areas 1 , 2,

3, and 4 in Figure 5.1, not subtracting production costs. This gross benefit


GB(k) is defined by adding costs back into the expression for net economic

welfare, i.e.,

GB(k) = NEW(k) + cN(k).

Hence,

d(GB) d(NEW)dk dk

cN'(k\ (5.9)

The first-order condition for gross benefit is given by (5.7) for NEW, addingback cN', which is negative. At kNEW, the level of fair use that maximizes

net economic welfare, d(GB)/dk<0, implying that the level of fair use that

maximizes gross benefits, kGB, is less than kNEW.

This is less surprising than it may initially appear. Ignoring production

costs, the optimal number of works should increase relative to that under

a net welfare standard. The difference depends on N' . If increasing the

degree of fair use has little effect on output, the standard that maximizes

net economic welfare will be close to the standard that maximizes gross

benefit.

Number of Uses

Fifth, suppose that the object of fair use policy is to maximize the number

of uses of copyrighted works, without regard to the value of different uses.

This could follow from a belief that willingness to pay depends on wealth

and does not reflect utility (in the utilitarian, not economic sense) or anyother intrinsic social value in access to works. Letting USE reflect the

volume of access to copyrighted works under fair use regime k, we have

N(k)

USE(k) =I

[(1-k)D(p*(n),n) + kD(Q,n)]dn

o

cKfJSE)m

and

dko

+ N'[(l-k)D(p*(N),N) + kD(0,n)]. (5.10)

The integral term in (5.10) is the net increase in uses of existing works from

extending fair use. The second term is the lost uses that follow when extend-

ing fair use reduces the number of works.

No general comparison between kUSE,the degree of fair use that maxi-

mizes access, and other values of fair use found above is apparent.


Equation (5.10) suggests one possibility. It is similar to the expression for

the first-order condition for maximizing aggregate consumer surplus. From

(5.7) and (5.8), we have

N(k)

I[CS(0,/i)

-CS(p*(n),n)]dn

o

+ N'[(\- k)CS(p*(N\N) + kCS(Q,N)].

The integral here equals the integral in (5.10) times the average consumer

surplus over the supplied works. The second term in the above expression

equals the second term in (5.10) times the average consumer surplus of the

marginal work. If the former average consumer surplus exceeds the latter,

the positive contribution to the condition for CS outweighs the negative

contribution at kUSE, suggesting that

and hence that kACS>kUSE, i.e., that looking at uses leads to less generous

fair use.

This result is plausible. Increasing the amount of fair use will tend to

increase uses of inframarginal works that are relatively highly valuable and

discourage creation of works that are valued less. If these relative values are

ignored and all uses counted equally, this difference will not matter and this

justification for making fair use stronger will not apply.

Summary

Recalling that the optimal fair use regimes are kN for number of works,

kNEW for net economic welfare, kcs for consumer surplus, kGB for gross

value of uses, and kUSE for number of uses,

kACS > kmw>kGB >kN = and ' probably, kACS > kUSE.

The degree of the first inequality depends on the prices at which copy-

righted works are sold. The degree of the second and third inequalities

depends on the sensitivity of the supply of copyrighted works to price. In

addition, one may be able to come up with specifications for this model in

which kNEW>kUSE, i.e., if one were concerned only with uses, one might

want a weaker fair use policy than that motivated by economic efficiency. If

increasing fair use has a sufficiently large negative effect on the number of

works produced without much of an effect on the number of uses per work,

one might get more overall uses of works with less fair use rather than more.


5.5 LOW RESERVATION PRICE FAIR USES

We now model fair use more in line with the transaction costs rationales

Gordon and others have described. We characterize fair use by the param-eter r, where all uses of a work valued at a reservation price below r are

free to the user. We assume throughout this section that this restriction is

enforced, i.e., that there is no diversion of higher valued uses into the fair

use regime.

As in the earlier model, n* is the copyrighted work for which profits

just cover cost, i.e., where p*(n*)=c/D(p*(n*),n*). Obtaining results in

this setting requires the further assumption that as n increases, /?*() falls, i.e.,

p*' < O.22

If/?*' < 0, fair use will not discourage any creation of works for r <

/?*(*). For all w<*,/?*(w) >/?*(*) >r, implying that the copyright holder

will lose no sales. Under these conditions, fair use will enhance all measures

of benefit based on use, without reducing the supply of works, if fair use

allows all uses where the reservation price is less than /?*(*). For notational

convenience, let r* equal p*(n*). For the rest of the analysis, we restrict our

attention to r>r*, the range where policy tradeoffs become relevant,

although with respect to actual copyright rules, r may well be less than r*.

In this setting, there are two different effects of fair use on the market for

a work, depending on whether p*(ri) > r or p*(n) ^ r. The first is displayed

in Figure 5.2.

As with Figure 5.1, Area 1 represents the consumer surplus for those who

buy work n at the copyright holder's price /?*(), Area 3 is the net profits

/*()

clD(p*(n\ n)

D(p*(n), n) D(r, n) >(0, n)

Uses of work n

Figure 5.2 Consumer surplus, profits, and costsfor work n: profit-

maximizing price >fair use price


D(r, n)

Uses of work n

Z>(0, n)

Figure 5.3 Consumer surplus, profits, and costsfor work n: profit-

maximizing price <fair use price

reaped by the copyright holder of work n, and Area 4 is the cost c of creat-

ing work n. These are independent of the fair use regime as long asp*(n) > r.

Area 2 represents the consumer benefits from fair use, to those with reser-

vation prices below r obtaining the work for free. Area 5 represents con-

sumers unwilling to payp*(n) but willing to paymore than r; these consumers

do not get the work and reap no benefits. In this setting, those who value

the work a lot or a little get it; those in the middle do not. Note that r need

not be below c/D(p*(n), n) as drawn in the graph, as long as it is belowp*(ri).

The second case is where p*(n) ^ r. There, the copyright holder of work

n is unable to charge the nominal profit maximizing price, since users whovalue the work more than p*(n) but less than r will obtain it for free. The

copyright holder will charge r for the D(r,n) uses valued more than r.

Relevant effects are displayed in Figure 5.3.

Consumers make all positively valued uses of works. Area 1 represents

the surplus of those consumers who have to pay r for the work. Area 2 is

the surplus captured by consumers with reservation price less than r who

get the work freely under the fair use policy. Area 3 is the profit to the copy-

right holder; Area 4 is the cost of creating this work.

Output-maximizing Fair Use

Let us consider works for which rD(r,n) < c will not be created. Let N(r) be

the value of n for which rD(r,n)=

c. From our assumption that the equi-

librium price increases as n falls, we have that all works for which n< N(r)


will be created. The fair use level that maximizes output, indicated againwith the subscriptN9 is rN<r*. To avoid discouraging uses that would leave

output and prices unaffected, rN =r*. By any policy criterion, the optimalr will be no less than r*.

Net Economic Welfare

For any r, copyrighted works n>N(r) fit two categories. First, for n > N(r)such that p*(n) <r, copyright holders will set the price at r and sell D(r,),as in Figure 5.3. Define N*(r) by p*(N*) = r. Because /?*'<0, 7V'*<0;as r increases, the set of works for which owners can charge the profit-

maximizing price shrinks. Second, for n> N*(r), owners can charge /?*(),with results as seen in Figure 5.2.

Net economic welfareNEM^with fair use r is consumer surplus plus profits

aggregated over works in each of these two categories, less creation cost.

NEW(r) = [CS(p*(n),n) + CSF(r,n) +p*(n)D(p*(n),n)]dno

N(r)

i)dn-cN(r}. (5.11)+J

CS(0,)

N*(r)

As before, CS(p, n) is consumer surplus for those who pay price p to

use work n. CSF(r,), the consumer surplus from fair use for uses with

reservation prices below r obtained for free, is Area 2 in Figure 5.2. For

n G [0, N(r)], the terms in the integral are respectively Areas 1, 2, andAreas 3 and 4 together; Area 4 is subtracted off in the cost term on the left.

For n G [N(r), 7V*(r)], all uses are made of the work, so the total benefit

net of costs is just total available surplus CS(0, n). The last term above is the

subtracted cost of creating N(r) works. If r =/?*(0), N*(r)= 0. Increasing r

above p*(Q) would eliminate consumer surplus from the fewer works that

are created, and those losses would be less than the saved costs. 23 Thevalue of r that maximizes NEW, rNEW, must then be at most/?*(0), makingN*(rNEw) anc* all terms of (5.11) relevant. Consequently, rNEW is the

value of r satisfying

d(NEW) _A

+ CFS(r,N*(r)) + p*N*(r)D(p*(N*(r)\N*(r))-

CS(0, JV*(r))]

'Fair use'

as policy ins trumen t 93

Because CSFr(r,n)

= - rDr(r,n)

24and, by definition, at N*(r\p* = r, the

sum of Areas 1, 2, and 3 constitutes the entire consumer surplus CS(0,

N*(r)) available at TV*. The long second term on the right is thus zero, leaving

(r))-cNf

(r). (5.12)

The first term is the increase in surplus from allowing more uses of works

for which the selling price exceeds the fair use standard. The second term

is the lost surplus from the marginal works no longer profitable when all

uses valued below r are provided free. The last term is the reduced cost from

creating N' fewer works.

Because Drand N' are negative, it will be useful to use absolute values

and change signs as appropriate. Let

\[D(^n)-D(r,n)]dn

Nbe the average amount of fair use of works through TV at maximum reser-

vation price r. We can rewrite (5.12) as

d(NEW}^ = N*(r)rDr(r9 N*(r))-

\N'(r)\[CS(0 9 N(r))-

c]. (5.13)

The first term is, again, the gain in surplus from increasing fair use, which

equals the average increase in consumer surplus times the number of works

for which that surplus is increasing. The second term is the lost net surplusas the supply of works falls by \N'(r)\.

Net economic welfare is maximized when (5.13) equals zero. Setting it

equal to zero and multiplying through by r/N*(r)[CS(Q, N(r))-

c] gives the

equivalent first-order condition

rD(r,N*(r)) _ N(r).6 JVOF

[CS(0, N(r)) -c]~USE

N*(r)

where

^M-W^*'-"* , ....'..,..

is the elasticity of the quantity of fair uses for works where the price is

above the fair use reservation price, and


is the absolute value of the elasticity of supply of works with respect to the

degree of fair use.

The numerator in the fraction on the left-hand side of (5.14) is the ratio

of the product of the marginal value of fair uses times the average fair use

of works where the market exceeds the reservation price of fair uses. Thedenominator is the total net surplus of works that would be excluded if fair

use were marginally enhanced. The fraction on the left-hand side is the ratio

of total works created to works created where price exceeds the marginalfair use.

At r*, the ratio on the left-hand side is 1;the numerator and denominator

both equal *, the number of works where profits cover creation costs.

There are no created works where the profit-maximizing price is less than

the maximum fair use reservation price. The D term in the numerator will

be the average number of fair uses of all works. From (5.14), we would not

want to make fair use any more generous than r* if at r*

r*D(r,n*)LXI i . n t

<e,[CS(0,*)-c]

USE ^W

If the elasticity of the number of works created sufficiently exceeds the elas-

ticity of fair uses at r*, a more generous fair use regime above the minimumwill reduce net economic welfare.

Aggregate Consumer Surplus

As before, the difference between net economic welfareNEWand aggregateconsumer surplus ACS is that the latter does not include profits for works

created, represented by Area 3 in Figures 5.2 and 5.3. As consumer surplus

includes only Areas 1 and 2 in the cases portrayed by those figures, aggre-

gate consumer surplus over all works supplied is

N(r)

ACS(r)= [CS(p*(n\n) + CSF(r,n)]dn+ f [CS(r,n) + CSF(r,n)]dn.

Nr)

f [CS(r,n) +

N*(r)

Because CSr(r,n)

=-D(r,n), maximizing ACS with respect to r gives

25

N(r)

-D(r,n)dn

N*(r)

+ N'(r)(CS(r,N(r)) + CSF(r,N(r))}.

At N(r), the combined consumer surplus for customers with reservation

prices above r who pay for the work and for those who with reservation


prices below r who use the work for free is just the total available surplus

CS(0, N(r}) less cost, as profits just cover cost at N(r) by definition. Thebracketed term in the above is CS(Q,N(r))

-c, implying from (5.12) that

d(ACS) = d(NEW)dr dr

W3[-rDr(r,n)-D(r,n)]dn. (5.16)

The expression in the integral in (5. 16) is the negative of the derivative of

the profit function at r for works where the fair use reservation price exceeds

the profit-maximizing price. For prices above the profit-maximizing price,

that derivative is negative, i.e., rDr(r,n) + D(r,n)<0. This makes the inte-

gral in (5.16) positive. At rNEW, increasing r further would increase ACS.The degree of fair use that maximizes aggregate consumer surplus, rACS,

exceeds rNEW.

The result is qualitatively the same as in the 'equal diversion' case; profits

lost when fair use is increased do not count against aggregate surplus as

they do against net economic welfare. The difference here depends on the

convexity of the profit function, which determines the degree to which the

integral in (5.16) is positive. If rNEW=r* because (5.14) holds, rACS mayequal r* as well. Changing the policy objective from economic welfare to

consumer surplus, ignoring profits to creators, need not lead to a more gen-erous fair use regime.

Gross Benefit

Define again gross benefit GB as the gross consumer surplus, neglectingcosts. For works where the fair use reservation price is less than the profit-

maximizing price, as portrayed in Figure 5.2, the gross benefit is repre-

sented by Areas 1,2,3, and 4. For works where the reservation price exceeds

the profit-maximizing price, the gross benefit will be the consumer surplusobtained if uses were free to all. In both cases, gross benefit equals net eco-

nomic welfare with creation costs added back in. Accordingly, as with

'equal diversion', (5.9) holds, adapted for the low reservation price imple-

mentation, i.e.,

c^ (r) . (5, 7)

At rNEW, where the first term on the right-hand side is zero, the derivative

of gross benefit will be negative (Nr < 0). The fair use regime that maximizes

gross benefit, rGB, will be less than rNEW. If creation costs are disregarded,the social calculus will tilt toward increasing the number of works and awayfrom uses per work relative to where net economic welfare is maximized.


The degree to which rGB falls below rNEW depends on the sensitivity of the

supply of works to the level of fair use. If N' is small, these will differ little.

As rNElvmay equal r*, rGB will be even more likely to equal r*, the value of

fair use that maximizes the supply of works.

Number of Uses

For works where the profit-maximizing price exceeds the fair use reserva-

tion price r, the number of uses will be those purchased at the profit

maximizing price, D(p*(n\ ), plus free uses under the fair use regime,

Z>(0,) D(r,n). For works where the fair use reservation price exceeds

the profit-maximizing price, all D(0, n) uses with a positive value will occur.

The total number of uses, USE(r\ is

7r)

N(r)

[D(p*(n),n) + 0(0, /i)-0(r,)] dn + f D(0,n)dn.

N*(r)

USE(r) is maximized at rUSE, defined by the first-order condition

MT7V1?\ N*W^^=f -D(r,n)dn + N'(r)D(Q,n)

= Q. (5.18)dr J

o

The first term is the increase in uses from making more of them 'fair;' the

second is the lost uses from the reduced output when more generous fair

use reduces the supply of works.

To compare rUSE to other fair use standards, multiply the derivative in

(5.18) through by r. Comparing the result with (5.12) gives

. N'(r)[rD(Q,n)-

(CS(0,#(r))-

c)]. (5.19)dr dr

As N' < 0, this expression is positive at rNEW> i.e., rUSE>rNEW " an(* om<y ^

r0(0,/i)<CS(0,JV(r))-c atr = rNEW.

Dividing both sides by 0(0, n) implies that rUSE> rNEW if and only if rNEWis greater than the average net surplus per use of the marginal work at rNEW.

Even if rUSE>rNEW,rUSE may be below rACS . This ambiguity is greater

than that in the equal diversion case, where we have good reason to believe

that fair use would be stronger under a consumer surplus standard than

under a total use standard. The difference arises because the added uses in

this case from increasing fair use are valued at r. In the 'equal diversion'

setting, the added uses are valued at the average willingness to pay over all

works, which is likely to be greater than the average surplus associated with

forgone uses of the marginal work.

'Fair use'


To compare rVSE and rGB , use (5.17) and (5.19) to find

d(USE) d(GB)

dr drN'(r)[rD(0,n)-CS(0 9 N(r))].

This implies that rUSE>rGB if and only if r/)(0, n)<CS(Q, N(r)) at rGB,

i.e., that r is less than the average willingness to pay for the marginal work

over all uses.

Summary

Recasting fair use to allow only uses below a target reservation price pre-

serves many of the results under equal diversion. Fair use is strongest if

aggregate consumer surplus is the objective and weakest if numbers of

works is the objective. In between those two, the level of fair use that maxi-

mizes net economic welfare exceeds the level that maximizes gross value of

uses, i.e.,

> r > r > r r*r r r ~ r

where r* is the maximum reservation price for free uses that would not

discourage creation of any works.

How a gross benefit standard compares to the net economic welfare

standard is largely the same in both cases. The degree to which rACS exceeds

rNEW depends on the extent to which increasing the level of fair use reduces

profits for works for which the fair use price exceeds the price the copy-

right owner would prefer to charge. The difference between rNEW and rNdepends on how the ratio of the supply elasticity of works compares to the

elasticity of fair uses. At rN=

r*, this ratio may be small enough so that

rNEW= rN

= r*

:>deviations above the minimum reduce net economic welfare.

If the elasticity of the number of works with respect to r is sufficiently large,

we could have strict equality among all the terms in the above expression

and thus not want to increase fair use beyond the minimum regardless of the

policy objective. The degree of fair use that maximizes uses of copyrightedworks may or may not exceed the level that maximizes net economic welfare

or gross benefit. It depends on whether the average net surplus or gross

surplus is less than the maximum value of a fair use.

5.6 CONCLUSION

Our goal is to better understand optimal fair use standards, particularly

how they depend upon criteria by which copyright policy should be judged.


We examined a setting in which copyrighted works are created at constant

cost and serve markets with independent demands. Fair use took two

forms, one in which some fraction of uses were freely available ('equal

diversion'), and a second in which uses below a maximum willingness to

pay were available for free ('reservation price'). Standards were evaluated

according to numbers of works supplied, net economic welfare, aggregate

consumer surplus, gross benefit, and numbers of uses.

Many results held in both contexts. From a net economic welfare per-

spective, the marginal benefit of more fair use is the increase in surplus from

added uses of works produced. This should equal the marginal cost of

more fair use, which from an efficiency perspective is lost net surplus (con-

sumer benefit less production costs) from marginal works that would no

longer be profitable to create if more uses were free. Interest in numbers of

works created would lead to weaker fair use standards; interest in consumer

surplus would lead to a more generous regime relative to net economic

welfare. Focusing on gross benefits leads to less generous fair use relative to

these other standards. Ignoring creation costs under the former standard

tilts the 'uses vs. works' tradeoff toward producing more works.

Why those differences occur and whether they are significant are influ-

enced by different factors in the two settings. Convexity of profit functions

matters more under a reservation price criteria, while price levels of the

copyrighted works play a stronger role with equal diversion. Under the

reservation price setting, if the elasticity of supply of works with respect to

the scope of fair use is sufficiently large, the choice of objective may not

matter. The optimal fair use standard under any criterion may be the

maximum consistent with no reduction in the supply of works at all.

Focusing on uses regardless of surplus led to ambiguous and different

answers in the two settings. Under equal diversion, focusing on uses alone,

regardless of differences in willingness to pay by different users, would

likely lead to a less generous fair use standard than that maximizing aggre-

gate consumer surplus. Neglecting willingness to pay uses would tend to

rebalance the relevant tradeoffs more in favor of additional, less econom-

ically valuable works and away from giving users with high willingness to

pay greater free access to works. Under the reservation price regime, the fair

use regime that maximizes uses depends on how the maximum reservation

price for fair use compares to the average surplus consumers obtain from

the marginal work. In either case, a fair use standard based upon maxi-

mizing uses of copyrighted works could be less generous than one designed

to maximize net economic welfare.

A corollary is that if the copyright policy balance is seen solely as one

between numbers of works and numbers of uses, neglecting values based

on willingness to pay (of wealthy rather than poor users, say) and profits,

'Fair use'

as policy instrument 99

fair use may be less generous than if efficiency is the standard. The optimal

level of fair use would be between that maximizing uses and that maximiz-

ing supply of works, and both standards may be less generous than that

maximizing net economic welfare.

One extension would be to consider that copyright policies today affect

the availability of copyrighted works in the future, e.g., by stimulating

creativity and fostering a sense that creators are enriching an ongoingstream of cultural works from which they have drawn (Gordon, 2003).

These effects could be represented as an external benefit based on reducing

the cost of producing future works. But how this cultural complementarityaffects fair use policy depends very much on how that external benefit is

generated. If future benefit comes from having more works produced or

gross benefit in the present, it argues for reducing fair use relative to a net

economic welfare standard, to stimulate the production of more works. If

that benefit comes from today's consumer surplus, it is likely to suggest

increasing the scope of fair use relative to that standard. Other factors,

including number of uses, are more ambiguous, as the tradeoffbetween lost

surplus from fewer uses of works (increasing fair use) and lost surplus from

having fewer works (decreasing fair use) is more relevant.

We can also investigate different settings to see whether these results are

robust. One could employ a Cournot oligopoly (Watt, 2000, pp. 37-54),

modeling works as perfect substitutes, where each copyright holder chooses

simultaneously how much of the work to make available. A second class of

models could involve competition among suppliers of differentiated prod-

ucts (Carlton and Perloff, 2000, ch. 7). As these models remain arbitrary, we

are unlikely to get precise estimates of how fair use should be designed. The

complexity and lack of generality of differentiated product models mayinhibit their ability to inform judgments on the design of copyright policy.

ACKNOWLEDGMENT

Thanks to Shubha Ghosh, Wendy Gordon, Mark Nadel, Christopher Yoo, the editors of this

volume, and other participants at the 2003 Congress of the Society for Economic Research on

Copyright Issues for comments. Errors remain the author's responsibility.

NOTES

1. State Street Bank & Trust v. Signature Fin. Group, 149 F.3d 1368 (Fed. Cir. 1998);

Amazon.com v. Barnes & Noble, 73 F. Supp. 2d 1228 (1999).

2. A&M Records, Inc., et al. v. Napster, Inc. (9th Circ., Feb. 12, 2001); Universal City

Studios, et al. v. Corley (2nd. Circ., Feb. 7, 2002).


3. Digital Millennium Copyright Act, Pub. L. No. 105-304, 1 12 Stat. 2860 (Oct. 28, 1998).

4. Eldred et al. v. Ashcroft (U.S. Sup. Ct. 01-618, Jan. 15, 2003), interpreting U.S.

Constitution, Article I, Sec. 8.

5. 17US.C.107.6. See Napster, n. 2 supra, Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417 (1984);

Harper & Row v. The Nation Enterprises, 471 U.S. 539 (1985); Campbell v. Acuff-Rose

Music, Inc. 510 U.S. 569 (1994); American Geophysical Union v. Texaco, Inc., 60 F.3d

913(1994).7. I thank Wendy Gordon for this reference.

8. Brennan (1994) reviews the limits of markets in providing social goals.

9. This model does not exactly fit Gordon's ( 1 982) justification for fair use, that some users

value the work but are not willing to pay the transactions costs associated with the work.

In that case, the demand curve that the creator of a work sees will be net of these trans-

actions costs. We show that the fair use policy maximizing net economic welfare is typ-

ically greater than zero. A fair use policy focusing on economic considerations may then

be more generous than that which minimizes transactions costs.

10. To examine changes in one feature of copyright, it is useful to assume that the rest of

copyright law is designed efficiently, to avoid using an inferior policy instrument to

address a problem other than the one at hand (Brennan, 1988).

11. See also National Research Council (2000, p. 138). For an exclusively economic

approach, see Landes and Posner (1989).

12. From 17 U.S.C. 107, the four statutory factors courts are to consider in ascertaining

whether a use is a fair use are:

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; and4. the effect of the use upon the potential market for or value of the copyrighted work.

Klein et al. (2002) deal only with the last of these, the potential market, and not the first

three.

13. Ramello (chapter 7 of this volume) makes a similar point from a qualitative, socio-

anthropological perspective.

14. They do not suggest that the optimal compulsory license fee need be zero; they offer only

a possible rationale for fair use.

15. We also neglect income effects, i.e., that if less money is spent on some copyrighted works,

because their price falls or because they go off the market, that demand will increase for

other works because consumers have more money to spend.1 6. We discuss allowing a work as an 'input' into the production of subsequent works in the

conclusion. Such considerations underlie concerns with parody and derivative works in

the copyright area, and innovations that build on earlier ones in the patent area.

17. We assume Dp<Q and other second-order conditions necessary to ensure unique

internal profit maximizing prices for each work. This assumption implies that the

'marginal' work has a lower demand curve than that of 'inframarginal' works.

1 8. One could regard the demand curve for work n as net of the marginal cost of producing

copies.

19. In positing a single price for the copyright work across all uses, we assume copyright

holders cannot price discriminate beyond what fair use institutes. A more complex model

could allow for some price discrimination, e.g., as in sales of hardcover and paperback

books, or theatrical release compared to video rental for films.

20. 'Profit' here refers to variable profit or producer surplus, ignoring the ex ante costs of

producing the work. The number of works is determined endogenously with entry up to

the point that this profit equals the cost of producing a work.

'Fair use'


21 . A policy to have no fair use might also maximize producer surplus, which undoubtedlywill affect the development of copyright policy in a political environment.

22. This assumption does not follow from Dn <Q. For any n, the value of p* is defined by

ry/7, n)= 0. Consequently, p*'

= - HpnIH

pp. Since the denominator is negative,/;*' <0 if

and only ifO^< 0. From (2), 11^=pD

pfl

+ Dn

. The second term is assumed negative, but

the first could be positive and outweigh it. We have/?*' < when Dpn< 0, i.e., the demand

curve becomes steeper at any price as n increases.

23. The consumer surplus of works created exceeds the creation cost c for r>r*. At r*, the

profit of work *just equals costs. As consumer surplus exceeds profits, consumer

surplus exceeds costs. It thus will be true for all <*, since profits increase as n falls.

24. From Figure 5.2, CSF(r,n) = CS(0,)- CS(r,n)-rD(r,ri). The derivative with respecttor is

CSFr(r, n)

= - CSr(r, )

-D(r, n)

- rDr(r, n).

Since CSr(r, n) D(r, n), the first two terms on the right-hand side cancel, leaving the

expression in the text. The subscript r refers to the partial derivative of demand with

respect to price. For clarity, we use r instead of p throughout this section.

25. Because p*(N*(r = r at N*, CS(p*(N*(r)),ri)=

CS(r,n). Thus, terms with TV*' cancel, as

with the derivative of NEW.

REFERENCES

Besen, S. and S.N. Kirby (1989), 'Private copying, appropriability, and optimal

copying royalties', Journal ofLaw and Economics, 32, 255-80.

Besen, S., W. Manning and W. Mitchell (1978), 'Copyright liability for cable televi-

sion: Compulsory licensing and the Coase theorem', Journal of Law and Eco-

nomics, 21, 67-95.

Brennan, T. (1986), 'Harper & Row v. The Nation: Copyrightability and fair use',

Journal of the Copyright Society of the U.S.A., 33, 368-89.

Brennan, T. (1988), 'An economic look at taxing home audio taping', Journal ofBroadcasting and Electronic Media, 32, 89-103.

Brennan, T. (1994), 'Markets, information, and benevolence', Economics and

Philosophy, 10, 151-68.

Brown, R. (1985), 'Eligibility for copyright protection: A search for principled

standards', Minnesota Law Review, 79, 579-609.

Carlton, D. and J. Perloff (2000), Modern Industrial Organization, Boston, MA:Addison Wesley.

Ciminello, D. (2000), 'Business process patents- the road to reform', Internet

Law Journal, 14 November 2000, available at http://www.tilj.com/content/

iparticlelll40001.htm.

Depoorter, B. and F. Parisi (2002), 'Fair use and copyright protection: A price

theory explanation', International Review ofLaw and Economics, 21, 452-73.

Gordon, W. (1982), 'Fair use as market failure: A structural and economic

analysis of the Betamax case and its predecessors', Columbia Law Review,

82, 1600-57.

Gordon, W. (2002), 'Excuse and justification in the law of fair use: Commodi-fication and market perspectives', in N. Elkin-Koren and N. Netanel (eds), The

Commodification of Information, Norwell, MA: Kluwer, pp. 149-92.


Gordon, W. (2003), 'Intellectual property', in P. Cane and M. Tushnet (eds), The

Oxford Handbook of Legal Studies, New York: Oxford University Press,

pp. 617^6.

Johnson, W. (1985), The economics of copying', Journal of Political Economy, 93,

158-74.

Klein, B., A. Lerner and K. Murphy (2002), 'The economics of copyright: "Fair

use" in a networked world', American Economic Review Papers and Proceedings,

92, 205-8.

Landes, W. and R. Posner (1989), 'An economic analysis of copyright law', Journal


Liebowitz, S. (1985), 'Copying and indirect appropriability: Photocopying of jour-

nals', Journal of Political Economy, 93, 945-57.

National Council for Higher Education (2003), 'Of fair use, liabilities, and elec-

tronic toll booths', available at http://www.nea.org/he/nche/fairuse.html, accessed

26 April 2003.

National Research Council Computer Science and Telecommunications Board

(2000), The Digital Dilemma: Intellectual Property in the Information Age,

Washington, DC: National Academies Press.

Novos, I. and M. Waldman (1984), 'The effects of increased copyright protection:

An analytic approach', Journal of Political Economy, 92, 23646.

Ramello, G. (2005), 'Private appropriability and sharing of knowledge: Conver-

gence or contradiction? The opposite tragedy of the creative commons', in

L. Takeyama, W. Gordon and R. Towse (eds), Developments in the Economics of

Copyright: Research and Analysis, Cheltenham, UK and Northampton, MA,USA: Edward Elgar.

Takeyama, L. (1997), 'The intertemporal consequences of unauthorized reproduc-

tion of intellectual property', Journal ofLaw and Economics, 40, 51 1-22.

Watt, R. (2000), Copyright and Economic Theory: Friends or Foes?, Cheltenham,

UK and Northampton, MA, USA: Edward Elgar.

Yoon, K. (2002), 'The optimal level of copyright protection', Information Economics

and Policy, 14, 327^8.

6. Towards a differentiated products

theory of copyright

Christopher S. Yoo

6.1 INTRODUCTION

The economic analysis of copyright is largely founded on the premise that

consumption of copyrighted works is non-rivalrous, a premise that distin-

guishes markets for intellectual property from markets for other types of

goods. In the context of copyright, this is generally taken to mean that the

marginal cost associated with the consumption of an incremental unit of

a copyrighted work is effectively zero. Non-rivalry in turn gives rise to a

well-known economic conundrum. If authors are to break even, the prices

they charge must defray the fixed cost needed to produce the work in the

first place (often called 'first-copy costs') as well as cover the marginal cost

associated with producing an incremental unit. However, pricing above

marginal cost necessarily reduces welfare by excluding some potential users

from consuming the work even though the benefits they would derive

exceed the costs of permitting them to do so.

This has led scholars to frame copyright in terms of the need to balance

two opposing considerations. On the one hand are the benefits flowing

from the efficient dissemination of copyrighted works (often called the

'access' side of the tradeoff). On the other hand is the need to provide

authors with sufficient compensation to support the creation of their works

(often called the 'incentives' side of the tradeoff) (Novos and Waldman,

1984; Johnson, 1985; Liebowitz, 1985; Landes and Posner, 1989). This

tradeoff implicitly posits that markets for copyrighted works are protected

by entry barriers and that particular works do not face substantial compe-tition from substitutes.

The traditional economic approach to copyright assumes that copyright

turns authors into monopolists over their works (Novos and Waldman, 1 984;

Liebowitz, 1985). The power over price conveyed by this legal monopoly

power gives rise to the familiar efficiency losses that occur whenever prices

exceed marginal cost. The concomitant transfer of surplus from consumers

to producers has also led some scholars to raise distributional concerns

103


(Netanel, 1996; Boyle, 2000; Cohen, 2000). Together these considerations

have led to the emergence of a consensus that copyright protection is a nec-

essary evil and that Congress and the courts should calibrate copyright pro-

tection to the lowest level that still provides sufficient return to support

creation of the work.

The problem with the traditional approach is that copyrighted works

do in fact face competition from other works that serve as imperfect sub-

stitutes and that entry is often quite easy. In fact, the doctrine known as

the 'idea-expression dichotomy,' which limits copyright protection only to

those aspects of a work that display the author's originality and leaves

unprotected any facts or ideas contained within the work (17 U.S.C.

102(b)), has the practical effect of guaranteeing that any competitor willing

to undertake the same fixed cost investment as the original author remains

free to create alternative works with the same functional characteristics.

The growing recognition that copyrighted works generally face some

degree of competition from other works has led some scholars to turn

to dominant firm and Cournot market structures to model markets for

copyrighted works (Landes and Posner, 1989; Koboldt, 1995; Watt, 2000).

Although an improvement over the monopoly approach, these models

still fail to capture the full dynamics of entry. Other scholars have drawn

on the approach to imperfect competition best suited to capturing the

key characteristics of markets for copyrighted works; the theory of product

differentiation. Early analyses focused solely on product differentiation in

the limited context of direct copying, in which unauthorized copies serve

as imperfect substitutes for the original (Johnson, 1985; Liebowitz, 1985;

Besen and Kirby, 1989; Koboldt, 1995).1 While helpful, these models still

fail to capture the economic impact of free entry by similar works. Some

initial work applying differentiated product models to model competition

among different works has begun to appear (Lunney, 1996; Abramowicz,

2004). Previous efforts have stopped short of considering the full range of

normative and remedial implications.

This chapter offers a more complete exploration of the economic insights

provided by a shift to a differentiated products approach to copyright (see

also Yoo, 2004). Adoption of a differentiated products approach opens upthe policy space by revealing that access to creative works can be increased

by facilitating entry by new works and allowing the ensuing competition to

reduce the spread between price and marginal cost. At the same time, the

possibility of entry largely alleviates any concerns about overstimulation of

creative activity or sustainable profits generally associated with the incen-

tives side of the tradeoff. It suggests that over the long run the presence of

supracompetitive profits will only serve to stimulate entry, which improves

access via lower prices and increased product variety.

Towards a differentiatedproducts theory of copyright 1 05

Perhaps most important, the differentiated products approach suggests

that the degree of tension between access and incentives implied by the

traditional approach may be overstated. By demonstrating how facilitat-

ing entry can essentially promote both considerations simultaneously, the

differentiated products approach echoes one of the central insights of

classic property theory, which emphasizes how well-defined property rights

can promote economic efficiency (Demsetz, 1967; Hardin, 1968). Indeed, it

shows how access to copyrighted works can be served by strengthening

property rights, which would stimulate entry and drive price closer to mar-

ginal cost as well as increase product variety.2

In contrast to the traditional approach, the differentiated products

approach also provides a basis for distinguishing among the available policy

instruments. Previous models tended to represent the overall level of copy-

right protection with a single variable (Novos and Waldman, 1984; Landes

and Posner, 1989; Koboldt, 1995). Viewing copyright through the lens of

product differentiation makes it possible to isolate the impact of multiple

ways in which copyright protection can be strengthened or weakened.

What results is a reconceptualization of copyright that moves beyondthe relatively static vision of monopoly economics and captures the

dynamics of free entry. The differentiated products approach also accords

better with the institutional capabilities of governmental actors. By pro-

viding a decentralized, market-oriented means for ensuring that authors

capture only enough revenue to cover their fixed costs and no more, it

responds to the growing doubts that courts and legislatures are institu-

tionally capable of striking the proper balance between access and incen-

tives. The government is better suited to promoting access by strengthening

copyright protection than to attempting to strike the proper balance

between access and incentives through the careful calibration of the level

of copyright protection.

6.2 THE NORMATIVE IMPLICATIONS OF THEDIFFERENTIATED PRODUCTS APPROACH

The normative insights provided by applying the differentiated products

approach to copyright can most easily be understood in terms of the

theory of monopolistic competition pioneered by Chamberlin (1933).

Monopolistic competition retains most of the assumptions underlying

perfect competition, including free entry and the presence of a sufficient

number of buyers and sellers to justify ignoring strategic reactions to

pricing decisions. The key difference is that monopolistic competitionrelaxes the assumption that competing works are homogeneous. Instead,


product differentiation gives authors a sufficient degree of power over price

to justify modeling each work as facing a downward-sloping demand curve.

Because each author makes his or her own independent price and quantity

decisions, this approach models competition at the producer level rather

than at the industry level. When the market is analyzed at the producer level

rather than the industry level, total surplus depends not only on the amount

of surplus generated by any particular work, but also on the total number

of works created.

The market power conveyed by product differentiation leads producers of

differentiated works to set short-run prices in precisely the same manner as

a monopolist, which results in deadweight loss and short-run supracom-

petitive profits. However, the possibility of entry recognized by monopolis-

tic competition dictates that the short-run equilibrium is unstable. When

entry is free, the presence of supracompetitive profits attracts entry by

authors offering similar works. The classic Chamberlinian formulation of

monopolistic competition also assumes that consumers' preferences are

symmetric with respect to a cluster of other works. 3 The primary effect of

this assumption is to place each work in equal competition with all other

works in the group rather than in localized competition with a smaller

set of near neighbors. Because all of the works in the market are in equal

competition with one another, new entrants take business equally from each

of the incumbents, which causes the demand curve confronting each incum-

bent work to shift inward. In addition, the demand curve confronting each

author becomes more elastic as a growing number of new works increases

the number of imperfect substitutes available for each consumer.

Entry will continue until no economic profits remain, which, under

Chamberlin's original formulation, will occur when the surplus appropri-

ated by each author is just enough to cover fixed costs.4 The equilibrium

number of works is determined by the size of the market relative to the size

of fixed costs. Indeed, as the size of the market expands or the size of the

fixed costs declines, the number of works asymptotically approaches infin-

ity, and the deadweight loss approaches zero (Hart, 1979; Mankiw and

Whinston, 1986; Jones, 1987; Eaton and Lipsey, 1989).5

A New Perspective on Access

The shift to a differentiated products approach effects two significant

changes to the economic analysis of copyright. First, when products are

homogeneous, authors can compete only on a single dimension -price

-

which also greatly simplifies the welfare analysis by reducing it to total

surplus. When the competing works are differentiated, the value of product

diversity must also be taken into account, making simple price-cost

Towards a differentiatedproducts theory of copyright 107

margins incomplete indicators of economic efficiency. As a result, productdifferentiation raises the possibility that any deadweight losses caused by

non-marginal cost pricing might be offset in part by welfare gains resulting

from product variety.

Even more importantly, the differentiated products approach highlights

the existence of an alternative way to reduce deadweight loss that up until

now has largely been overlooked. Specifically, it illustrates how access can

be promoted not by lowering the degree of protection provided by copy-

right, but rather by facilitating entry by similar works. Entry by near sub-

stitutes causes the demand curve facing each work to become more elastic,

which in turn reduces the spread between price and marginal cost.6

In this manner, the differentiated products approach opens up the policy

space by identifying how entry can promote access, an insight that the

traditional approach is poorly situated to take into account. When non-

rivalrous goods are homogeneous, entry is unnecessarily duplicative and

simply wastes resources. Moreover, the tendency towards natural monopolycreated by declining-cost structures strongly suggests that no such entry

would be viable.

This analysis suggests that access may be promoted by strengthening

copyright protection, because it is the presence of profits that stimulates

entry7(although, per the subsequent discussion, the analysis may require

weakening of other aspects of copyright protection). This stands in stark

contrast to the traditional approach to promoting access, which focuses

solely on lowering the level of copyright protection. Indeed, the differ-

entiated products approach suggests that weakening copyright will only

serve to deter entry by reducing the revenue generated by each work.

Consequently, as will be discussed in greater detail later, it may have the

perverse effect of cementing an excessively concentrated market structure

into place.

The differentiated products approach also mitigates any distributional

concerns raised by increasing the amount of surplus captured by authors.

As noted earlier, entry will continue until any supracompetitive profits are

dissipated. Furthermore, any short-run transfer of surplus from consumers

to producers will largely accrue to consumers' benefit over the long run in

the form of increased product variety.

The foregoing discussion demonstrates how the differentiated products

approach opens up the policy space by revealing how access can be pro-

moted indirectly by stimulating entry rather than directly by lowering the

level of copyright protection. In the process, it reveals that under certain

circumstances, economic welfare might better be promoted by follow-

ing precisely the opposite of the policies prescribed under the traditional

approach to the economics of copyright.


The Formalization of Optimal Incentives

Appropriability as a determinant of optimal incentives

In addition to suggesting an alternative way to promote access, the differ-

entiated products approach also offers a solution to one of the fundamen-

tal limitations of the traditional approach, which is its inability to providea basis for formalizing the optimal level of incentives. In the absence of

such a basis, scholars have employed rough metrics to approximate the

proper balance (Fisher, 1998; Brennan, 2002). These scholars concede that

such metrics provide only a vague sense of how much creative activity

would be optimal.

The present analysis overcomes this shortcoming by offering a basis for

determining the optimal level of entry. It suggests that a work should be

produced whenever the surplus it would create exceeds the costs needed to

produce it. This condition is met whenever the total surplus generated bythe work is larger than the fixed cost (Spence, 1976a, 1976b; Dixit and

Stiglitz, 1977; Spence and Owen, 1977; Koenker and Perry, 1981).

This criterion for determining efficient levels of entry illustrates the

importance of the authors' ability to appropriate the surplus generatedfrom their works. It suggests that a reduction in authors' ability to appro-

priate the surplus created by their works can cause them to forgo creating

marginal works even though doing so would cause total welfare to increase.

The larger the reduction in the authors' ability to appropriate surplus, the

fewer welfare-enhancing works will be created.

This represents a fairly sharp departure from the view of appropriability

taken by the traditional approach, which takes the position that copyrightlaw should carefully calibrate appropriability so that works capture only

enough surplus to support creation of the work. The differentiated prod-ucts approach reveals that such fine tuning would be counterproductive,

since any limitations to authors' ability to appropriate surplus will cause a

suboptimal number of works to be created and entry by competitive works

will help ensure that no work garners excessive returns. 8 As a result, increas-

ing appropriability lacks distributional implications over the long run,

since free entry will dissipate any profits initially accrued and largely cause

them to accrue back to consumers in the form of increased product variety.

At the same time, doctrines that restrict appropriability have the inevitable

effect of exacerbating the access side of the tradeoff by decreasing the

extent to which entry can narrow the spread between price and marginalcost in the long run. Indeed, the foregoing analysis demonstrates how low-

ering the level of copyright protection, rather than promoting efficiency,

may have the perverse effect of entrenching a concentrated market struc-

ture into place by making it impossible for new competitors to enter.


Demand diversion as a countervailing consideration

The analysis advanced thus far would appear to suggest that economic

welfare would best be promoted by maximizing authors' ability to appro-

priate surplus. Indeed, were appropriability the only relevant consideration,

copyright policy would devolve into a simple matter of allowing authors

to capture as much revenue as possible. However, complete appropriationof surplus, which would require perfect price discrimination, is a prac-

tical impossibility. One would thus conclude that markets would exhibit a

systematic tendency towards underproduction of copyrighted works that

could not be rectified no matter how much copyright law is structured to

enhance appropriability.

One countervailing consideration, however, is whether the sales captured

by a new entrant represent incremental sales to new customers or instead

represent sales cannibalized from authors already in the market. FollowingBorenstein (1985), the former effect will be referred to as 'demand creation'

and the latter effect as 'demand diversion.'9

The possibility of demand diversion allows markets to create the optimalnumber of works even when authors are unable to appropriate the entirety

of the surplus created by their works. As stated earlier, the basic welfare

criterion for evaluating the efficiency of entry offered by the differentiated

products approach dictates that a work be produced whenever the benefits

it creates exceed the costs required to produce it. Although no author is able

to appropriate the entire surplus created by his or her work, demand diver-

sion raises the possibility that the incremental surplus that an author is

unable to capture might be replaced by surplus cannibalized from other

incumbents. In other words, because of demand diversion, the fact that

perfect price discrimination is impossible need not lead to a systematic

underproduction of product variety.

In fact, demand diversion creates the possibility of excess entry, in which

authors produce new works even when the costs of doing so exceed the

benefits. As noted earlier, efficient entry requires that authors producenew works only when the surplus attributable to demand creation exceeds

the fixed costs needed to produce the work. The problem is that a profit-

maximizing author will enter whenever the total surplus it captures exceeds

the fixed costs of entry regardless of whether the surplus captured results

from demand creation or demand diversion. Such an author could finance

the fixed costs with surplus cannibalized from other producers already in

the market rather than incremental surplus generated by new consumers.

Under these circumstances, the profitability constraint does not necessar-

ily prevent the waste of resources. In such cases, it may be appropriate to

use copyright to restrict entry by increasing the degree of differentiation

required before a new work does not infringe on existing works.


6.3 THE REMEDIAL IMPLICATIONS OF THEDIFFERENTIATED PRODUCTS APPROACH

In contrast to the traditional approach, the differentiated products

approach also provides a basis for distinguishing among the available policy

instruments. As noted earlier, previous models tended to represent the

overall level of copyright protection with a single variable. Only a handful

of copyright scholars have offered some preliminary attempts to analyze the

interaction among limited aspects of copyright protection (Fisher, 1998,

Liu, 2002; Hughes, 2003).

Product differentiation theory provides a basis for modeling the impactof different aspects of copyright protection, as illustrated by the patent lit-

erature using spatial competition to analyse the tradeoff between a patent's

'length,' determined by the duration of the patent term, and its 'breadth,'

which is most usefully described for present purposes as how close a com-

peting product may come in the characteristics space to a patented productwithout constituting infringement (Klemperer, 1990).

The differentiated products approach to copyright suggests a similar

analysis, but with some important modifications. Specifically, it suggests

that the analysis might be enriched by disaggregating the concept of length

into two distinct concepts. On the one hand is what I will call the 'size' of

the right, as determined by the number of surplus-generating activities

contained within the right. On the other hand is what might be called the

'intensity' of the right, as determined by the author's ability to capture the

available surplus. An increase in the size of the right would be represented

by an outward shift of the demand curve. An increase in the intensity of the

right would be represented by an increase in the proportion of the available

surplus captured by authors rather than consumers, created for example bya change in law that facilitates price discrimination.

Expanding the analysis in this manner highlights the complex inter-

actions among these factors. Interestingly, the policy implications do not

all point in the same direction. Specifically, the differentiated products

approach indicates that the best way to render a market more competitive

is by increasing the number of surplus-generating activities encompassed

by the right, facilitating authors' ability to appropriate surplus, and by lib-

eralizing how close competing products can come without constituting

infringement. In other words, economic welfare would best be promoted bya copyright that is large and intense, but narrow. The differentiated prod-ucts approach is therefore not an unqualified endorsement for strengthen-

ing copyright protection. Instead, by providing a basis for distinguishing

among different aspects of copyright protection, it allows for a more

nuanced approach to copyright policy.

Towards a differentiatedproducts theory of copyright 1 1 1

Determinants of the Differentiated Products Equilibrium

The size of the right

An important determinant of the overall competitiveness of markets for

copyrighted works is the size of the copyright, as determined by the numberof surplus-generating activities encompassed within its scope. As noted

earlier, the overall competitiveness among differentiated products is deter-

mined by the level of entry, which is in turn determined by the magnitudeof the relevant market relative to the fixed costs of entry. Increasing the

ratio of the size of the overall market to fixed costs brings the resulting

equilibrium closer to the competitive outcome.

The primary means for policymakers to increase this ratio is to expandthe size of the copyright by increasing the number of surplus-generatingactivities that fall within the scope of each copyrighted work. One deter-

minant of the size of the right that has received a great deal of attention in

recent months is the lengthening of the copyright term effected by the

Copyright Term Extension Act. But other examples abound, such as the

extent to which copyright allows authors to retain performance rights.

Doing so has the effect of causing the demand curve confronting each

copyrighted work to shift outwards.

The differentiated products approach suggests that increasing the size

of the right can promote both access and incentives by increasing the equi-librium number of works. The suggestion that access would be promotedbest by raising rather than lowering the level of copyright protection mayseem counterintuitive. After all, it implies that the proper policy responseto markets that are too concentrated is to increase the degree of copyright

protection that authors enjoy. This apparent contradiction disappearswhen viewed in light of the traditional approach's inability to capture the

dynamics of entry. So long as entry is free, any strengthening in the level of

copyright protection will not ultimately accrue to the benefit of the incum-

bents. Instead, it will only attract more entry, which will in turn reduce

deadweight loss and bring the number of works closer to the optimum. Anyshort-term profits made possible by the expansion of the size of the right

will largely accrue back to consumers in the form of increased variety.

The intensity of the right

Another consideration that determines the market's ability to promote eco-

nomic welfare is the intensity of the right, which is determined by authors'

ability to appropriate surplus. Unlike increases in the size of a copyright,which cause the demand curve confronting a copyrighted work to shift out-

wards, increases in intensity leave the demand curve intact and simplyincrease the proportion of the area under the demand curve captured by

1 1 2 Developments in the economics of copyright

authors. Intensity is affected by copyright principles such as the first sale

doctrine, which limits authors' ability to engage in price discrimination.

The breadth of the fair use doctrine also shapes authors' ability to appro-

priate surplus.

As noted earlier, any reduction in appropriability tends to reduce the

equilibrium number of works. Such a reduction can harm the incentives

side of the tradeoff by causing the total number of works produced to dropbelow optimal levels. It also harms the access side of the tradeoff by limit-

ing the extent to which entry by new works will reduce deadweight loss in

the long run.

This suggests that increasing authors' ability to capture the surpluscreated by their works can promote both the access and the incentives side

of the copyright tradeoff. Again, this argument may seem counterintuitive

from the standpoint of the traditional approach to copyright, which views

access and incentives as being in inexorable tension. The solution lies in

understanding that access may be promoted as much by increasing the

number of works available for consumption as by mandating access to the

limited number of works that have already been created.

The breadth of the right

Finally, the extent to which consumers regard competing works as sub-

stitutes plays a natural role in determining how many firms will enter at

equilibrium and how robust the competition among those firms will be.

Copyright's breadth is a legal constraint on substitution because infringingworks cannot compete absent permission from the copyright holder. This

analysis then initially suggests that copyright should be kept narrow in

order to promote lower price-cost margins.There are, however, considerations that cut in the other direction. For

example, substitutability also determines the extent to which the surplus

captured by any particular work derives from demand diversion. The higherthe degree of substitutability between the works, the greater the propor-tion of the total surplus captured that demand diversion will represent. In

other words, the more closely related works can be without constituting

infringement, the greater the amount of surplus will come from demanddiversion.

As noted earlier, demand diversion plays a critical role in determininghow closely the total number of works produced will approximate the

optimum. Up to a point, demand diversion is beneficial, as it can replacethe surplus that authors are unable to appropriate because of their imper-fect ability to price discriminate. Beyond that point, demand diversion

creates the possibility of excess entry. The danger of excess entry is the

greatest when goods are the most substitutable.


Policymakers may reduce the impact of demand diversion by using the

standard of copyright infringement to reduce the substitutability between

competing works. Doing so would require a delicate balance. On the one

hand, reducing substitutability limits the impact of demand diversion and

thus brings the number of works closer to optimal levels. On the other

hand, reducing substitutability also reduces the extent to which price-cost

margins are narrowed by entry. Policymakers charged with calibrating the

breadth of copyright protection may have to confront the difficult task of

balancing the welfare gains on the access side of the tradeoff against the

welfare losses on the incentives side of the tradeoff. 10

Interactions Among the Different Aspects of Copyright Protection

The differentiated products approach thus provides a framework that is able

to distinguish among three different ways in which copyright protection can

be strengthened or weakened. Although this degree of nuance enriches the

power of the analysis, it also makes copyright policy considerably more

difficult to implement.One problem is that the available policy instruments are not completely

independent. Changes in the legal regime designed to calibrate the size of

the right may also have an impact on the right's intensity or breadth. The

overlapping nature of these considerations complicates isolating the impactof any particular aspect of copyright policy.

It nevertheless may be possible to simplify the analysis with respect to

particular industries or categories of copyrighted works. If one aspect of

copyright protection can be taken as fixed with respect to certain types of

works, the problem that must be solved becomes much simpler. Empirical

studies may provide additional insights into how to balance these coun-

tervailing considerations. Some empirical studies suggest that any welfare

losses resulting from excessive entry are likely to be relatively small (Yarrow,

1985; Goettler and Shachar, 2001). Another empirical study of entry pat-

terns in the radio industry estimates that the deadweight losses attribut-

able to excess entry may be substantial (Berry and Waldfogel, 1999). The

latter study acknowledges, however, that the radio industry is somewhat

unusual in that it serves two different groups of customers - advertisers and

listeners - only one of which (advertisers) is able to make direct paymentsfor programming. What appears to be excessive entry when measured solely

in terms of benefits to advertisers may in fact be efficient when measured in

terms of both advertisers and listeners. ] l

These studies suggest the possibility of isolating the impact of each of

the three factors identified by the differentiated products approach. Even

if the empirical problem proves intractable, the differentiated products


approach should still provide useful intuitions about the way these factors

interact. It suggests, for example, that excess entry is least likely to be a

problem when a work has few substitutes. Thus, contrary to the conven-

tional wisdom, it is when a work is the most unique that the case for

strengthening the level of copyright protection in terms of the size and the

intensity of the right is the strongest.

6.4 CONCLUSION

In the final analysis, product differentiation offers significant promise as a

way to reconceptualize the economic analysis of copyright law. What

emerges is an approach that demonstrates how stimulating entry can

promote both access and incentives simultaneously. This stands in stark

contrast to the position that dominates existing copyright scholarship,

which views these two considerations as being in inexorable tension.

The differentiated products approach also suggests that the best policy

response to a highly concentrated market might well be to strengthen the

degree of copyright protection in order to stimulate new entry. There is

some irony in the fact that copyright protection might tend to be the

strongest when high fixed costs and the low degree of substitutability cause

the market to become the most concentrated, but this apparent paradoxis resolved once one understands the complex manner in which access

and incentives interact with one another. In this sense, the differentiated

products approach to copyright captures the insights of classic property

theory, which emphasizes the importance that well-defined property rights

can play in ensuring optimal investment and deployment. In so doing, it

corrects for the blind spot that results when markets for copyrighted works

are treated as monopolies, which prevents serious consideration of the role

that short-run profits can play in stimulating entry and in promoting eco-

nomic efficiency. At the same time, the differentiated products approach

incorporates the possibility of excess investment and entry stimulated bydemand diversion.

Although the theoretical implications of this analysis are clear, consid-

erable additional work remains to be done before it can be fully opera-

tionalized. As noted earlier, further work should incorporate elements of

sequential innovation that take into account the extent to which current

works serve as inputs to subsequent works, although, for reasons set forth

in the margin, such considerations are unlikely to prove problematic.12

Future work should also consider the implications of relaxing the

symmetry assumption by applying models that allow the extent to which

particular works serve as substitutes for other works to vary (Kaldor, 1935;


Waterson, 1990). This suggests that copyright might profitably be analyzed

by using the spatial competition models pioneered by Hotelling (1929), in

which works compete by adopting locations across a characteristics space

(Yoo, 2004). Relaxing the symmetry assumption allows for the possibility

that entry by a new work will impact only some of the works rather than

being spread evenly across all of the incumbents. This localization of com-

petition has the effect of dividing the relevant market into subsegments,

with the overall competitiveness of the subsegment determined by the size

of the total surplus of the subsegment relative to the fixed cost, rather than

the size of the total surplus of the entire market relative to the fixed cost.

The lack of robust competition within a subsegment may limit the extent

to which entry can push price towards marginal cost. It can also allow the

'integer problem' to arise simultaneously with respect to multiple portions

of the overall market, as the single 'large economy' is chopped into a series

of 'small economies' that are each capable of supporting sustainable profits.

If these effects arise with respect to multiple subsegments, these adverse

effects may be quite substantial (Eaton and Lipsey, 1976). Relaxing the

assumption that works will distribute themselves evenly across the product

space creates the possibility that first movers will employ preemptive

strategies to lock in sustainable supracompetitive positions (Baumol,

1967; Hay, 1976; Prescott and Visscher, 1977; Eaton and Lipsey, 1980;

Bonanno, 1987; Neven, 1987). The analysis becomes even more complex if

one allows for the possibility of production of multiple works by a single

author (Schmalensee, 1978; Eaton and Lipsey, 1979; Brander and Eaton,

1 984; Judd, 1985).

Finally, the policy instruments that follow from the differentiated prod-

ucts approach are by their nature extremely contextual and do not lend

themselves to simple policy inferences. In addition, the interrelationships

among the available policy instruments make calibrating them simultan-

eously an extremely difficult empirical exercise. The fact that the differ-

entiated products approach is contextual and nuanced should not obscure

its basic analytical power and does not by itself justify rejecting the theory.

Indeed, the intuitions that the theory reveals about the relationship

between access and efficiency and the manner in which the various aspects

of copyright protection interrelate are sufficient to justify further inquiry.

ACKNOWLEDGMENT

This chapter is adapted from material originally published at Yoo (2004). I would like to

thank Michael Abramowicz, Mark Brandon, Tim Brennan, Andy Daughety, Paul Edelman,Shubha Ghosh, John Goldberg, Wendy Gordon, Paul Heald, Larry Heifer, Steven Hetcher,

Jack Hirshleifer, Doug Lichtman, Stan Liebowitz, Glynn Lunney, David McGowan,


Michael Meurer, Tom Nachbar, Mark Nadel, Richard Nagareda, Neil Netanel, Erin O'Hara,Giovanni Ramello, Bob Rasmussen, Jennifer Reinganum, Pam Samuelson, Art Snow, Lisa

Takeyama, and participants in workshops conducted at the 2003 Annual Congress of the

Society for Economic Research on Copyright Issues, the Thirty-First Annual Meeting of the

American Law and Economics Association, the 2003 Works-in-Progress Intellectual

Property Colloquium at Tulane University, Columbia Law School, and the Vanderbilt LawSchool for helpful comments on earlier drafts of this work. Financial support from the

Vanderbilt Dean's Fund is gratefully acknowledged. Responsibility for any errors rests with

the author.

NOTES

1 . Ghosh (2004) employs a differentiated products model to evaluate the related issue of

competition between originals and derivative works.

2. The differentiated products approach can thus be understood as striking a middle groundbetween the traditional analysis of increased copyright protection and the nonprotectionof copyrighted works. The possibility of entry suggests that the tradeoff between access

and incentives is not as direct as suggested by the traditional approach. Furthermore, the

differentiated products approach also redresses the central problem with nonprotection,which can provide efficient access, but which struggles to provide sufficient incentives to

support production of creative works. This conclusion is subject to an important caveat

discussed below. When fixed costs are large compared to marginal costs, it is possible that

strong property rights may induce levels of entry that are excessive.

3. Chamberlin's original formulation also made a number of other simplifying assump-tions, none of which turn out to be central to the analysis. For example, Chamberlin

posited that each producer faced identical cost and demand curves. This allowed him to

employ a single graph portraying the price-quantity response of a representative firm to

model the entire market. Allowing the cost and demand curves to vary across productswould simply cause equilibrium price and quantity to differ with respect to each firm,

which is completely reasonable given the assumption that each product is differentiated.

Firm-to-firm variations in price and quantity would not, however, change any essential

aspects of the equilibrium (Kaldor, 1935; Archibald, 1961).

4. There is, however, a well-known exception to Chamberlin's zero-profit result. It has longbeen recognized that the lumpiness of fixed costs may create a situation in which n works

might earn small profits while n + 1 works would run losses. This so-called 'integer

problem' allows for an equilibrium in which n works each earns sustainable profits. In

large economies (i.e., when n is relatively large), such profits will be negligible. This

integer problem was first identified by Kaldor (1935).

5. This conclusion depends on the convexity of consumer preferences and production sets

(Hart, 1980; Roberts, 1980).

6. Because firms must cover their fixed costs, price will not completely converge to marginalcost.

7. Acknowledging how short-run profits stimulate entry should not be confused with

Schumpeterian competition, in which competitors use innovation to obtain long-run

competitive advantage. The ease of entry dictates that any profits should be transient

rather than sustainable. This implies that there will be horizontal competition within the

market rather than vertical competition for the market. Short-run profits instead stimu-

late entry in the same manner as occurs in perfectly competitive markets that are in tem-

porary disequilibrium.

8. Subject to an important consideration discussed in the following section.

9. Other analyses use different terminology to describe the same effect (Beath and

Katsoulacos, 1991, p. 57 ('cannibalisation'); Mankiw and Whinston, 1986 ('business

stealing effect')).

Towards a differentiated products theory of copyright 1 1 7

10. To give another example, substitutability also affects the proportion of surplus that an

author can capture in another way. Monopolistic competition theory indicates that

works with relatively steep inverse demand functions capture a lower proportion of the

available surplus than do works with relatively flat inverse demand functions. Workswith steep inverse demand functions tend to be products with low own-price elasti-

cities of demand, which in turn tend to be those that have the fewest substitutes

(Spence, 1976a; Spence and Owen, 1977). For a more complete discussion, see Yoo

(2004, pp. 273^4).

11. In addition, the fact that their study assumed that the radio market is composed of

homogeneous products led them to overlook potential welfare benefits resulting from

product differentiation. The existing theoretical literature suggests that this simplifying

assumption can have a fairly dramatic effect on the welfare implications (Mankiw and

Whinston, 1986).

12. For example, just as stimulating entry by close substitutes should promote access to

readers by using increased competition to lower price, it should also promote access to

follow-on authors who seek to build on prior work. Once the market for the work becomes

sufficiently competitive, the problem of cumulative innovation, in which a copyrightedwork is licensed simultaneously to both consumers and to other authors who seek to use

the work as an input in creating other works, becomes analogous to the classic problemof transfer pricing, in which a particular good simultaneously serves as an end productand as an input used in making another product. The transfer pricing literature indicates

that welfare is maximized when the price of the good when used as an input is set equalto the price charged of the good when sold as a final product, so long as the final productmarket is sufficiently competitive (Milgrom and Roberts, 1992, pp. 79-83). Thus, so longas the total surplus in the market is sufficiently large relative to the fixed costs of entry,

there is nothing inefficient about charging the market price to follow-on authors who seek

access to a copyrighted work as an input in creating other works. Should the market for

the works not be sufficiently competitive, the differentiated products approach suggeststhat the problem might be redressed by making the market more competitive by stimu-

lating entry rather than by lowering the price paid by follow-on authors. It is true that

holdout behavior may prevent particular authors from creating particular works.

Competition policy, however, focuses on protecting competition, not particular competi-tors. Thus, unless such refusals create losses for more than just particular individuals, nointervention is warranted. And even if intervention were justified, it would take the form

of a targeted remedy and not a general revision of the scope of copyright protection.

REFERENCES

Abramowicz, M. (2004), 'An industrial organization approach to copyright',

William and Mary Law Review, 46 (1), 33-124.

Archibald, G.C. (1961), 'Chamberlin versus Chicago', Review ofEconomic Studies,

29(1), 2-28.

Baumol, W.J. (1967), 'Calculation of optimal product and retailer characteristics:

The abstract product approach', Journal of Political Economy, 75 (5), 674-85.

Death, J. and Y. Katsoulacos (1991), The Economic Theory of Product Differ-

entiation, Cambridge and New York: Cambridge University Press.

Berry, ST. and J. Waldfogel (1999), 'Free entry and social inefficiency in radio

broadcasting', RAND Journal of Economics, 30 (3), 397^420.

Besen, S.M. and S.N. Kirby (1989), 'Private copying, appropriability, and optimal

copying royalties', Journal ofLaw and Economics, 32 (2), 255-80.

Bonanno, G. (1987), 'Location choice, product proliferation and entry deterrence',

Review of Economic Studies, 54 (1), 37-45.

1 1 8 Developments in the economics of copyright

Borenstein, S. (1985), 'Price discrimination in free-entry markets', RAND Journal

of Economics, 16 (3), 380-97.

Boyle, J. (2000), 'Cruel, mean, or lavish? Economic analysis, price discrimination

and digital intellectual property', Vanderbilt Law Review, 53 (6), 2007-39.

Brander, J.A. and J. Eaton (1984), 'Product line rivalry', American Economic

Review, 74 (3), 323-34.

Brennan, D.J. (2002), 'Fair price and public goods: A theory of value applied to

retransmission', International Review ofLaw and Economics, 22 (3), 347-75.

Chamberlin, E.H. (1933), The Theory of Monopolistic Competition, Cambridge,MA: Harvard University Press.

Cohen, I.E. (2000), 'Copyright and the perfect curve', Vanderbilt Law Review,

53(6), 1799-819.

d'Aspremont, C, J.J. Gabszewicz and J.-F. Thisse (1979), 'On "Hotelling's stability

in competition" ', Econometrica, 47 (5), 1 145-50.

Demsetz, H. (1967), Toward a theory of property rights', American EconomicReview (Papers and Proceedings), 57 (2), 347-59.

Dixit, A.K. and I.E. Stiglitz (1977), 'Monopolistic competition and optimumproduct diversity', American Economic Review, 67 (3), 297-308.

Eaton, B.C. and R.G. Lipsey (1976), 'The non-uniqueness of equilibrium in the

Loschian location model', American Economic Review, 66 (1) 77-93.

Eaton, B.C. and R.G. Lipsey (1979), The theory of market preemption: The persist-

ence of excess capacity and monopoly in growing spatial markets', Economica, 46

(182), 149-58.

Eaton, B.C. and R.G. Lipsey (1980), 'Exit barriers are entry barriers: The durabil-

ity of capital as a barrier to entry', Bell Journal of Economics, 11 (2), 721-9.

Eaton, B.C. and R.G. Lipsey (1989), 'Product differentiation', in R. Schmalensee

and R.D. Willig (eds), Handbook of Industrial Organization, vol. 1, Amsterdam:Elsevier Science, pp. 725-68.

Fisher, W.W., HI (1998), 'Property and contract on the Internet', Chicago-Kent LawReview, 73 (4), 1203-56.

Ghosh, S. (2004), 'Market entry and the proper scope of copyright', SSRN working

paper available at http://ssrn.com/abstract=569 162.

Goettler, R.L. and R. Shachar (2001), 'Spatial competition in the network televi-

sion industry', RAND Journal of Economics, 32 (4), 624-56.

Hardin, G. (1968), The tragedy of the commons', Science, 162 (3859), 1243-8.

Hart, O.D. (1979), 'Monopolistic competition in a large economy with

differentiated products', Review of Economic Studies, 46 (1), 1-30.

Hart, O.D. (1980), 'Perfect competition and optimal product differentiation',

Journal of Economic Theory, 22 (2), 279-312.

Hay, D.A. (1976), 'Sequential entry and entry-deterring strategies in spatial com-

petition', Oxford Economic Papers, 28 (2), 240-57.

Hotelling, H. (1929), 'Stability in competition', Economic Journal, 34 (153), 41-57.

Hughes, J. (2003), 'Fair use across time', UCLA Law Review, 50 (2) 775-800.

Johnson, W.R. (1985), The economics of copying', Journal of Political Economy,93(1), 158-74.

Jones, L.E. (1987), The efficiency of monopolistically competitive equilibria in

large economies: Commodity differentiation with gross substitutes', Journal ofEconomic Theory, 41 (2), 356-91.

Judd, K.L. (1985), 'Credible spatial preemption', RAND Journal of Economics,

16(2), 153-66.


Kaldor, N. (1935), 'Market imperfection and excess capacity', Economica, 2 (5),

33-50.

Klemperer, P. (1990), 'How broad should the scope of patent protection be?',

RAND Journal of Economics, 21 (1) 1 13-30.

Koboldt, C. (1995), 'Intellectual property and optimal copyright protection',

Journal of Cultural Economics, 19 (2), 131-55.

Koenker, R.W. and M.K. Perry (1981), 'Product differentiation, monopolistic com-

petition, and public policy', Bell Journal of Economics, 12 (1), 217-31.

Landes, W.M. and R.A. Posner (1989), 'An economic analysis of copyright law',

Journal of Legal Studies, 18 (2), 325-63.

Liebowitz, SJ. (1985), 'Copyright law, photocopying and price discrimination',

Research in Law and Economics, 8, 181-200.

Liu, J.P. (2002), 'Copyright and time: A proposal', Michigan Law Review, 101 (2),

409-81.

Lunney, G.S., Jr. (1996), 'Reexamining copyright's incentives-access paradigm',Vanderbilt Law Review, 49 (3), 483-656.

Mankiw, N.G. and M.D. Whinston (1986), 'Free entry and social inefficiency',

RAND Journal of Economics, 17 (1), 48-58.

Milgrom, P. and J. Roberts (1992), Economics, Organization and Management,

Englewood Cliffs, NJ: Prentice-Hall.

Netanel, N.W. (1996), 'Copyright and a democratic civil society', Yale Law Journal,

106 (2), 283-387.

Neven, D.J. (1987), 'Endogenous sequential entry in a spatial model', International

Journal of Industrial Organization, 5 (4), 419-34.

Novos, I.E. and M. Waldman (1984), 'The effects of increased copyright protection:

An analytical approach', Journal of Political Economy, 92 (2), 23646.

Prescott, E.C. and M. Visscher (1977), 'Sequential location among firms with fore-

sight', Bell Journal of Economics, 8 (2), 378-93.

Roberts, K. (1980), 'The limit points of monopolistic competition', Journal ofEconomic Theory, 22 (2), 256-78.

Schmalensee, R. (1978), 'Entry deterrence in the ready-to-eat breakfast cereal

industry', Bell Journal of Economics, 9 (2), 305-27.

Spence, M. (1976a), 'Product differentiation and welfare', American Economic

Review (Papers and Proceedings), 66 (2), 407-14.

Spence, M. (1976b), 'Product selection, fixed costs, and monopolistic competition',

Review of Economic Studies, 43 (2), 217-35.

Spence, M. and B. Owen (1977), 'Television programming, monopolistic competi-

tion, and welfare', Quarterly Journal of Economics, 91 (1), 103-26.

Waterson, M. (1990), 'Product differentiation and profitability: An asymmetric

model', Journal of Industrial Economics, 39 (2), 1 13-30.

Watt, R. (2000), Copyright and Economic Theory: Friends or Foesl, Cheltenham,UK and Northampton, MA, USA: Edward Elgar.

Yarrow, G.K. (1985), 'Welfare losses in oligopoly and monopolistic competition',Journal of Industrial Economics, 33 (4), 515-29.

Yoo, C.S. (2004), 'Copyright and product differentiation', New York University LawReview, 19 (\),2\2-W.

7. Private appropriability and sharingof knowledge: convergence or

contradiction? The opposite

tragedy of the creative commons

Giovanni B. Ramello

7.1 INTRODUCTION

One of the many Greek myths tells how Eos, goddess of dawn, fell in love

with a handsome young man of royal Trojan blood, Tithonus. The goddess

was so taken with Tithonus that she asked Zeus to make him immortal, so

that they could be happy together forever. The wish was granted but turned

out to be a double-edged sword, because the gift of immortality did not

come with perpetual youth. So as time went by, and Tithonus withered and

grew old, the passion of the goddess died away.

Now, every myth contains a metaphor designed to teach us something.

The lesson we can draw here is that, when policy design fails to accurately

grasp the nature of the reality, the outcomes of the policy can be very

different from those envisaged. In this chapter we shall argue that, due to

analytical shortcomings, the present-day policy of extending and strength-

ening copyright is liable to produce inefficient outcomes in the knowledge

domain, analogous to those described in the above myth.Note that we are not here disputing copyright as it was, say, up until the

1990s, when it seemed to adequately balance the public need for access to

information with the provision of a private incentive to creators - a state of

affairs which, for the sake of simplicity, I have defined as 'minimal copy-

right'.1 That original form of copyright, to be absolutely clear, was fully

compatible and consistent with its statutory goals, namely, promoting the

creation of new knowledge through limited private appropriation of infor-

mation items - as history has amply borne out. 2

Rather, this chapter calls into question the new form copyright has been

taking following the legal trend toward its extension, whose advent can

probably be dated to the TRIPs Agreements (1994), and which is today

120

Private appropriab ility and sharing of knowledge 121

progressively altering the equilibrium established by the previous legal

framework in balancing knowledge sharing with appropriability. I have

used the term 'maximal copyright' to denote this contemporary, extended

form of the right.

The novel legal dynamics are displacing the traditional vision of copy-

right doctrine, which recognized the public-interest goals of the law and

so constrained the property right in various ways (see for instance the

term of duration, fair use doctrine, etc. 3) in order to simultaneously allow

for both (limited) private appropriability as an instrumental device, and

the return of newly created knowledge to the public domain as its primary

goal.

Utilitarian and neoclassical economic theory have provided the general

paradigm in support of the role of copyright as an incentive to create, but

have rarely recognized the importance of balancing the appropriability

granted by copyright with the necessity of assuring broad access to the

works created as a result. The contributions generally emphasize appropri-

ability over accessibility, with one notable exception being Landes and

Posner (1989). These distinguished law and economics scholars clearly

assert, in their influential contribution, that the welfare-enhancing effects

of copyright are ambiguously dependent on a balance between two oppos-

ite forces: an increase in the supply of new works brought about by the

statutory economic incentive on one hand, and a diminution in the supply

brought about by the exclusionary effect of copyright on the other hand.

The resolution of this trade-off is therefore the key to determining what

the overall consequences of copyright will be. Unfortunately, the authors

do not further pursue the analysis in this direction, but merely rely on some

specific assumptions in this respect.4

This chapter shall take the route of attempting to gain further insight

into this ambiguous relationship, paying special attention to the peculiar-

ities which characterize the production side. Using a wide selection of lit-

erature drawn from the various social science disciplines, it will attempt to

expose some of the shortcomings of traditional economic theory as it is

applied to copyright. The resultant perspective should lead to a better

understanding of the incentive/exclusion trade-off, showing how the

overall effect is likely to be greater than expected. This is so because

maximal copyright affects not just the cost of accessing each individual

protected work, but equally and significantly, it impacts on the productive

contexts (those which we have termed creative commons) by leading to a

depletion of knowledge and its dynamic productivity.

It follows, by mere straightforward application of neoclassical economic

reasoning, that the new extended copyright is liable to determine dynam-

ically inefficient outcomes analogous to those presaged by Hardin (1968) in


his landmark paper, and which, in tribute to this seminal author, we have

termed 'the opposite tragedy of the creative commons'.

The chapter is organized as follows: section 7.2 presents the tragedy of

the commons and describes the idiosyncratic feature of creative commons,in which knowledge is created and exists. The asserted peculiarity is then

supported with a discussion of the special nature of knowledge, which is

intrinsically dependent on the sharing process. Section 7.3 expands further

upon the structure of knowledge, whose two complementary constituent

elements, the tacit and codified dimensions, are equally necessary for

knowledge to exist in its entirety, and both have a productive role. It then

illustrates how real-world policies for promoting knowledge creation are

generally consistent with this representation, thus implicitly acknowledg-

ing the limited ability of proprietary systems to stimulate production of

optimal levels of new knowledge. Section 7.4 provides a test for the sound-

ness of the preceding arguments, through a simple check on whether the

proposed policy indications are consistent with the Lockean paradigm

widely invoked in justification of property rights. Finally, section 7.5 con-

tains the concluding remarks.

7.2 COMMONS, COPYRIGHTS AND TRAGEDIES

The economics literature on intellectual property rights generally presents

information as a public good, and consequently treats copyright as a device

designed to stimulate optimal levels of its production (see, for example,Landes and Poster, 1989; Watt, 2000). In other words, the property right is

introduced to avert a specific market failure, i.e., the production of a sub-

optimal level of new expressions of ideas. From this perspective, copyright-

protected information is considered to be an output, and the market failure

to be avoided is its underproduction.However there exists another, distinct type of market failure, arising

from the fact that information also serves as an input to the production

process and represents a common resource.

From Commons to Creative Commons

Summing up, in the knowledge domain, intellectual property rights should

enable individuals to reap the benefits of the new ideas, or expression of

ideas in the case of copyright, which they contribute to create. But, at the

same time, they must also avoid an altogether different kind of market

failure - namely, the depletion of the special type of common resource that

is knowledge.

Private appropriability and sharing of knowledge 1 23

These are two distinct and in a sense contrasting goals, because, as weshall see below, property rights on outputs can have negative effects on the

preservation of inputs.

In order to explain this assertion, though, we need to examine in greater

depth the nature of the knowledge as a commons. A 'commons' is gener-

ally defined to be a resource held in common, that is to say in joint use or

possession. Typical examples are uncultivated pastures, hunting grounds,

fishery stocks, etc. Now the problem here is that allowing open access to all

members of the community which owns the commons can lead to overex-

ploitation of the resource and its consequent depletion. This dynamic has

been extensively investigated in the scientific literature of complementarydomains. 5

The crucial element of all these contributions is that the production func-

tion of the commons exhibits diminishing returns with increasing number

of users. Some examples of this are pastures grazed by increasing numbers

of animals, or fishery stocks exploited by increasing numbers of fishing

boats. Hardin (1968, p. 1243) dramatically describes the previous situation

in his seminal 'tragedy of the commons': 'Ruin is the destination toward

which all men rush, each pursuing his own best interest in a society that

believes in the freedom of the commons. Freedom in a commons brings

ruin to all'.

The pivotal element of the above reasoning is, as always, the exhaustibil-

ity of the commons, which in turn refers us back to the basic postulates of

'scarcity of resources' in economic theory. In other words, the pasture,

fishery stock, etc. are all scarce resources, and as such need to be allocated

efficiently. This is generally achieved through rationing of access.

However, the specific instance of a knowledge commons is characterized

by some markedly different and contrasting traits. In fact, though much of

the above theory still holds true, knowledge is a special type of collective

resource with a peculiar feature that substantially alters the economic

evaluation - namely, the presence of increasing returns to number of users.6

This specific attribute depends, as we shall see, on the inherently collective

nature of knowledge, which means that an excessive rationing of access

through copyright can, ceteris paribus, produce an inverse tragedy to that

prefigured by Hardin -essentially an 'opposite tragedy of the creative

commons'. In the remainder of this work we shall therefore consider refer-

ences to knowledge as an input- which exists within collective contexts -

as corresponding to the concept of 'creative commons'. 7

Naturally, all the above does not seek to negate the economic method,which retains its validity; it merely challenges the results of scientific work

that fails to factor in the peculiar characteristics of creative commons, and

uncritically applies the hypothesis of diminishing returns. Whereas the


novel feature of the context under study concerns precisely the traditional

concept of scarcity, which as we have seen is no longer brought about by

over-exploitation of the collective resource, but rather by its under-

exploitation.8

The Social Nature of Knowledge Production

This important difference in the nature of knowledge is determined by its

social provenance and the sharing process that unavoidably characterizes

its existence. Knowledge in fact belongs intrinsically to the collective con-

texts where it is created. It is brought to fruition in the symbolic sphere

defined by society and renewed through sharing, which is thus an indis-

pensable feature (one might even say it constitutes an essential and not sub-

stitutable input in its own right) for individual creative activities.

Generally speaking, there is a widely held consensus, within the discip-

lines that study creative and inventive activities, that 'intellectual activity in

general is not ex nihilo creation. Given this vital dependence of a person's

thoughts on the ideas of those who came before him, [consequently] intel-

lectual products are fundamentally social products' (Hettinger 1989,

p. 38).'

Note that, in the domains of scientific and technological knowledge,even mainstream economic theory has explicitly adopted the thesis set forth

by Suzanne Scotchmer (1991, p. 29) in a celebrated article: 'Most innov-

ators stand on the shoulders of giants, and never more so than in the

current evolution of high technologies, where almost all technical progress

builds on a foundation provided by earlier innovators'. This view is sup-

ported by the observation that innovation 'consists to a substantial extent

of a recombination of conceptual and physical materials that were previ-

ously in existence' (Nelson and Winter, 1982, p. 130).10

It is no coincidence that the generic term denoting creative activity is

'composition' (we speak in effect of literary composition, musical compos-

ition, etc.) whose Latin etymology specifically invokes the above-described

process: 'cumponere' means literally 'to put together', suggesting that cre-

ation is first and foremost a novel arrangement of existing elements.

However, this viewpoint only seizes upon the incremental aspect of the

production of ideas,11 whereas there is also a non-hierarchical dimension to

consider, namely that of the sharing process which engenders the cognitive

value of knowledge. This is closely bound up with the semantic nature of

knowledge, which originates and exists exclusively within collective con-

texts. Paraphrasing Geertz (1973, p. 11), knowledge 'is public because

meaning is'.12 Thus there can be no knowledge without meaning. And,

accordingly, there can be no meaning without a human group to share it.


Therefore, even though knowledge fragments are often created by

individuals, this can only happen embedded within the broader context of

the collective semantic space to which the knowledge fragments are inextric-

ably tied.

Anthropologists (Romney, 1999, p. 104) provide us with some further

insight into this idea:

knowledge, found mostly in humans, arises from human inventions, is learned

and handed down from one generation to the next, and usually varies from one

society to another ... it is shared among relevant participants and ... it is

learned as part of our social heritage. The word 'relevant' alerts us to the idea

that there may be small specialized subgroups, such as medical practitioners,

whose members share esoteric knowledge not possessed by the wider cultural

group. In short, careful reflection reveals that the very notion of [knowledge]involves sharing of ideas, concepts, behaviours, etc., by more than one person.

In other words, citing again the enlightening definition formulated byWeber and more recently by Geertz (1973, p. 5), we can say that 'man is an

animal suspended in webs of significance that he has himself spun', where

knowledge constitutes the webs that connect him to other individuals. This

definition will also aid in understanding the following section.

Note that the above idea can be easily transposed into the categories of

economics, by thinking of it as an extreme instance of network externalities;

the semantic, and hence the cognitive and the resultant economic value of

knowledge or its fragments (as copyrighted works) is, at least to a certain

extent, an increasing function of the numbers of individuals that share it.13

Hence, if creative commons are deprived of knowledge sharing- not just

in the reductive sense of static shared consumption of information goods,which is normally termed 'information sharing' and sends back just to the

public goods market failure, but rather the sharing of signs, symbols, cre-

ative processes and, in short, all those elements which constitute knowledgein its broadest sense - then they will inevitably dry up in terms of meaningand, by consequence, of value to individuals. 14

It follows that a necessary existence condition for a knowledge creative

commons is the possibility of sharing and it presents value as an output byreason of this feature. Let us look at some examples of this. Consider,

for instance, a simple fragment of knowledge as a literary text or book.

It is attributed value by individuals - who consequently have a certain

willingness-to-pay for its purchase- due to its profound relation with the

social context in which the individuals live and have developed. A text

written in an incomprehensible or invented language would not have any

meaning for the reader, nor any cognitive utility. A similar observation can

be made at the semiotic level: a collection of words which have no referent


in a shared symbolic universe would have the same effect upon the reader

as a text in an unknown language. It cannot be understood, and therefore

has no meaning and no value. 15 This argument can be applied to all the

semantic dimensions of literature.

If we now move on from text to sounds, the above reasoning proves just

as valid: musical language is related to the social context which produces it,

and has value for the individuals as a result of the social references which

it creates at various levels. This is true for both the syntax and the distinc-

tion made between 'sound' and 'noise'. In fact, to a listener from one

cultural context, the music of another culture may sound tuneless, incom-

prehensible or even distasteful. In this case too, not-sharing of the seman-

tic content conveyed through a musical piece translates into a feeling of

strangeness, or even aversion, in the listener (note that in this case seman-

tic does not necessarily mean textual, but rather functional; in other words,

relating to some social function, for example recreational, religious, hedon-

istic, etc.).16

It is worth noting that a growing body of work in economic theory is

today acknowledging the collective and networked dimension of know-

ledge, albeit in an unconscious and indirect form that focuses on the pro-

duction of specific goods, as in the case of industrial clusters (Koepp, 2002;

David and Foray, 2002), as well as scientific and technological knowledge.With respect to the latter, for example, Nelson and Nelson (2002, p. 732)

observe that 'to students of the subject, there is no escaping the fact that,

while particular individuals may know different things, what they know

they have drawn largely from the culture in which they live, and to be used

effectively that know-how often needs to be part of a coordinated activity

involving several persons. And while particular individuals play a key role

in the advance of knowledge, they do so from a largely common platformof know-how, and a large part of what they have achieved sooner or later

becomes part of the knowledge shared among professionals in the field'.

It should be pointed out, again with reference to the traditional eco-

nomic categories, that sharing therefore plays a twofold role in the produc-tion process, first as a constituent and inalienable feature of knowledge as

an input (no sharing= no meaning = no knowledge), and second because

it brings the added benefit of reducing the transaction costs of interper-

sonal relations by creating a common ground in which relations can more

easily and quickly be conducted. This is tantamount to opening a commu-nication channel that enables the various individual members of social

groups to engage in dialogue and learn more easily.17

The above arguments highlight an additional feature of creative

commons, which is that they constitute a social locus rather than a physical

one as in the case of ordinary commons (i.e. a hub for interactions between


individuals).18 To put it another way, the creative commons largely exists

embedded within the sharing process and in association with a social

grouping; whereas ordinary common resources are spatially identifiable,

separated by individuals and divisible into finite portions whose sum will

correspond to the original amount. An uncultivated pasture, for example,

occupies a physical location, exists irrespective of its potential users, and

can be parcelled into smaller lots, each of which is subject to specific prop-

erty rights and whose sum will always add up to the original resource.

So a creative commons is a special case, as we have seen from the above

arguments, which exists as a shared entity, inextricably bound up with the

existence of the individuals that make use of it, and can only partially be

broken up into fragments (i.e. expressed in tangible media) whose sum will

therefore not be equal to the whole. These characteristics are discussed

further in the next section.

7.3 THE ATOMIZATION PROCESS

This failure to appreciate the peculiar features of creative commons is due,

to a large extent, to the fragmentary view of knowledge taken by the con-

tributions dealing with the economic analysis of copyright, though not bythe economic theory tout court, as we shall see below. In fact, the represen-

tation implicitly adopted by this literature is that, even if each copyrightedwork constitutes a public good in itself, it is also, from a productive stand-

point, a well-defined entity clearly separable from the others. The incentive

mechanism is consequently designed to act on individual units - n times n

copyrightable works - in order to stimulate the creation of an optimalnumber of new expressions of ideas. In other words, the production processis atomized, with each creator likened to a standard production function,

acting individually and in isolation. The problem then becomes, at most,one of guaranteeing access to previous copyrights to be used as static

inputs, in order to enable an efficient number of creators/production

processes to function and produce their output. However such a represen-

tation is inconsistent with the true nature of knowledge, and fails to take

into account the productive role of the creative commons, which taken as

a whole constitutes an input in its own right.

If we instead adopt a more correct analytical perspective, which takes

into account the productive and constituent functions of the sharing

process within the knowledge domain, we find that rationing of access to

individual information items through copyright will in reality have muchmore far-reaching effects than those traditionally envisaged by the conven-

tional economics literature.


The Dyadic Structure of Knowledge

The arguments presented in the previous section, on the whole, bear out the

unavoidably (or inherently) collective nature of knowledge, and the social

indebtedness and embeddedness of the knowledge, or the commodified

fragments thereof, that are ultimately transformed into copyrighted works.

This knowledge therefore possesses productive potential only by virtue of

being created within, and connected with, the broader sphere of knowledge

sharing.

This is consistent with the observation, long consolidated in the other

social sciences, that 'cultural creation cannot be totally integrated into an

industrial production system' (Morin 1962, p. 25).19 In this connection we

notice how many 'creative' industries, unable to fully incorporate creative

activities into the organization, rely upon continual monitoring of the

social context for new ideas and knowledge that can subsequently be inte-

grated into their new products.20

Now, the analytical weakness in the economic theory of copyright lies in

its failure to recognize this significant feature, adopting instead an 'atom-

ized, undersocialized conception of human action' which 'disallows by

hypothesis any impact of social structure and social relations on produc-

tion, distribution, or consumption' (Granovetter 1985, p. 483). In particu-

lar, this procedure of atomization, according to its critics, has the effect of

cancelling from view all that cannot be directly attributed to single indi-

viduals, and which resides instead in the relationships between them. If

such a simplification can have distorting effects on the economic and social

analysis in general, as some scholars maintain, the misrepresentation

becomes dramatic in the knowledge domain, where it can significantly alter

the outcomes and the attendant policy design.21

Once again, it is the nature of knowledge that justifies this assertion. Anextensive body of scientific literature, including economics, has in fact

pointed out the peculiar structure of knowledge, which simultaneously

draws from and feeds back into the individual and social dimensions.

However this peculiar feature is, as we have seen, entirely discounted by the

traditional economic analysis of copyright and intellectual property rights.

In fact, in all the contributions of that literature, the treatment of infor-

mation as an output which can later serve as an input is based on the

assumption that knowledge corresponds to the sum of information items

produced by creative processes, viewed as discrete entities. 22 Hence, accord-

ing to this assumption, a proper incentive will produce the optimal level of

information, and therefore of knowledge.However such a simplification does not accurately reflect the reality,

because knowledge actually has a different and dyadic structure, made up


of two distinct but complementary components. On one side, there is the

so-called 'codified knowledge', which is able to be articulated and encoded;but on the other side there is also an immanent form of knowledge, not

clearly articulable or codifiable, but which is nevertheless communicated

and therefore exists within relations between individuals, consequentlytermed 'tacit knowledge'.

23

The former is what we usually term information (i.e., a fragment of

knowledge) encoded in any one of the various media made available by

society (and which are themselves a subset of knowledge) such as language,

writing, reproduction technologies and so forth. Because it can be encoded,it is then also able to be 'commodified', giving rise to markets of informa-

tion goods, such as the markets for copyrighted works. On the other hand,the latter (tacit) form of knowledge is dynamic in nature, and as such can

never be entirely encoded or commodified. Consequently, as we shall see

below, it cannot be produced and exchanged on the markets, but dependsinstead upon interpersonal relationships for its production and dissemina-

tion. In effect, tacit knowledge grows out of the dynamic sphere of the com-munication process, whose existence is a necessary precondition for

knowledge to exist, as well as for the sustenance of human groupings.24

Therefore, in view of the multidimensional nature of knowledge, as

opposed to the one-dimensional nature of information, we can say that the

knowledge set encompasses the information set, so that the sum of all infor-

mation is only a subset of knowledge.25

Economic theory, with the exception, of course, of that applied to intel-

lectual property, has long recognized the dyadic nature of knowledge.

Beginning in the middle of the last century, von Hayek (1945, pp. 521-2),

for example, contrasted the portion of scientific knowledge that can be

encoded with 'unorganized knowledge . . .,the knowledge of the particu-

lar circumstances of time and place' which is essentially inalienable anddemands the participation of several individuals in order to be exploited.

26

In more recent times, an analogous concept has been again advanced in

economic theory by Nelson and Winter (1982, p. 1 34) for the case of a pro-

ductive organization (i.e., a relevant group of individuals such as a firm)

where in addition to the knowledge owned by individual workers, there

exists a tacit knowledge, embodied within the organization itself and its

procedures, that is 'not consciously known or articulable by anyone in par-

ticular', but effective and crucial for the functioning of the organization.

Transmission of this latter, tacit form of knowledge can only take place,

as we have said, within a context of social interaction and sharing. Theabove argument explains, for example, the rationale behind the educational

systems of different societies: books are generally used to provide informa-

tion, but the presence of a teacher is necessary to impart learning, and in


general to communicate those aspects of knowledge which cannot be stat-

ically encoded. Accordingly, Nelson (2003, p. 917) asserts that any 'class-

room equipment ... [is a] complement not a substitute for an effective

teacher working with students'. What is missing, in fact, is the imparting of

that tacit knowledge which books alone are unable to convey.27

In the light of this perspective, we can reinterpret all those institutions

and events that bring individuals together, for scientific or cultural pur-

poses, as serving to promote, in more or less stable form, the dissemination

and circulation of tacit knowledge which could not be accomplished by anyother means.

Enforcing Private Rights and Financing Social Activities: Resolving the

Apparent Ambiguity

If public policies in support of the production and dissemination of know-

ledge are to be effective, they must take both its tacit and codified dimen-

sions into account. However, the distinction between these two components,

though possible at the theoretical level, is in practice liable to be misleadingbecause the two are not so easily divided. The decoding of codified know-

ledge generally relies upon possession of its complementary component,which is the corresponding tacit knowledge (Rooney et al., 2003).

28

Therefore, any action exerted on one component may have indirect

effects on the other, while the availability of one does not guarantee acqui-

sition of the other, and hence of knowledge in its entirety. This assertion is

backed up by the findings of various studies on technology transfer. These

studies have underlined the difficulty of transferring the tacit portion of

knowledge as compared with its codified portion, precisely due to the

necessity of moving not just physical goods but also individuals (i.e., estab-

lishing ad hoc social relations). At the same time, these studies have pointedout the unavoidable necessity of promoting both types of transfers, if the

policy undertaken is to have a favourable outcome.29

Now, from this perspective too, the economic theory of intellectual

property displays some analytical flaws that need to be addressed. On one

side, these rights appear effectively able to stimulate the production and

exchange of optimal levels of new information items; they in fact avert a

market failure, which would seem to imply that the market, with the appro-

priate legal correctives, is a sufficient instrument for achieving an efficient

equilibrium in the knowledge domain. But, notwithstanding this, private

and public institutions are not content to rely upon the market for optimumknowledge production, and instead finance significant numbers of collect-

ive initiatives which, in a truly efficient market, would have no real reason

to exist.

Private appropriability and sharing of knowledge 131

Within the scientific community, for example, conferences, workshopsand the like are considered to be indispensable to the work and intellectual

growth of scholars, and for this reason are allocated public or private

funding, may sometimes benefit from tax subsidies, etc. In the same vein,

all the governments of countries with sufficiently developed economies sub-

sidize universities, research centres and other such institutions, in the firm

belief that this expenditure will contribute to the nation's technological

advancement and competitiveness.30 We can therefore say that, alongside

the support provided by property systems, there generally exists a public

and private attentiveness toward activities which promote interaction

between individual members of specific scientific and cultural groups.

At first glance, such an observation might seem at odds with the trad-

itional economic analysis of copyright: if intellectual property rights are

able to secure efficient results in the domain of the production and

exchange of new knowledge, the remaining activities are in effect not only

superfluous, but also constitute a needless waste of production resources.

This apparent schizophrenia, which on the one hand causes private

rights to be instituted, whilst on the other promotes and funds socializing

activities, can instead be resolved precisely in light of the twofold nature of

knowledge. The private appropriability, guaranteed ex lege and which

underpins the reward mechanisms of copyright and intellectual property

rights in general, is only effective with respect to codified knowledge, and

can at best stimulate the production of an efficient level of new items of

information. It functions optimally in markets where it is possible to

produce and exchange items comparable in nature to private goods, such as

those derived from codified knowledge. However, this solution is unable to

also provide a correct incentive for the production and exchange of tacit

knowledge, whose existence relies upon non-market substrates connected

with the process of communication and sharing; in other words, on social-

izing frameworks. So, public and private organizations quite correctly

respond to this need by promoting activities designed to bring together

groups of individuals in a more or less stable capacity.

The two different policies are in effect complementary, and work jointly

towards achieving a common aim. However, it is essential to strike a properbalance in combining these two aspects

- and it is this balance, which the

original copyright protection laws appear to have achieved, which is grad-

ually being eroded by maximal copyright.

For example, consider again educational processes. As mentioned earlier,

these rely on a mix of codified knowledge (i.e., teaching materials such as

textbooks, etc.) and tacit knowledge (i.e., direct teaching), which exist, so

to speak, in a symbiotic relationship, with neither of the two components

being able to fully substitute for the other. As a result, the non-availability


of the one will have the indirect effect of also preventing transmission of

the other - it is not possible to teach 'in practice' something to which access

is precluded. To put it another way, restricting access to a book will also

curtail opportunities for circulation of the tacit knowledge indissolubly

tied to that book and its content. 31Naturally, this effect will be magnified

in direct proportion to the degree of appropriation permitted by the law.

Once again, these arguments are backed up by contributions from the

various social science disciplines. Communication studies, for example,

have shown that knowledge 'as a whole' is what enables us to assign

meaning to experience, shaping our perception of it, as well as contribut-

ing to storing the knowledge of the past and perpetuating that of the

present (McQuail, 1983). Therefore, any action on one dimension of know-

ledge will have inevitable repercussions on the other. Ultimately, rationing

of access to codified knowledge by means of copyright will generally have

more far-reaching consequences than those envisaged by the traditional

economic literature, because it also has effects on the transmission of tacit

knowledge.32 This situation can be described with the metaphor of an

opposite tragedy of the creative commons, as suggested in the preceding

paragraph.It should be emphasized that the previous argument is consistent with the

original rationale behind minimal copyright laws, which consider know-

ledge to be a public resource and put a limit on the duration of property

rights, precisely to avoid exaggerating the privatization process and thereby

impoverishing the creative commons. For the same reason, the above laws

provided for various exceptions to exclusive rights; for instance, in fair use

doctrine, when such rights were deemed to be economically inefficient from

a cost-benefit standpoint (Gordon, 1982), or to constitute an impedimentto related creative activities which imperatively demand access to a specific

copyrighted work for incremental creation, without bringing a significant

economic benefit to the holder (e.g., critical essays, scientific works).33

The structure of minimal copyright was designed to grant individual

incentives through private appropriability, but without significantly

impacting on the public dimension connected with the sharing process.

Because relations between individuals are characterized by a certain degree

of randomness, arising from the dynamic nature of human behaviours, a

weak proprietary system will not appreciably affect the sharing of know-

ledge and the productivity of creative commons; as a matter of fact, a

certain degree of entropy must always by definition characterize knowledgeand communication among human beings.

34 Provided the creative

commons system is sufficiently broad, the interference of a minimal form

of copyright- limited in time and space

- will be marginal, with only imper-

ceptible effects on human relationships and hence on tacit knowledge.


In contrast, the increasingly exclusionary effects of maximal copyright,which continues to dilate the private space at the expense of the public

sphere, are having ever more serious repercussions on human interactions,

on tacit knowledge, and thus on the creative commons as a productiveresource.

Naturally, these effects are not immediately apparent, but emerge grad-

ually over a fairly long-term time horizon. As a matter of fact, the effects

of this new legal regime have up until today been mitigated (though admit-

tedly in an improper manner) by the phenomena of piracy and private

copying, which have preserved a 'de facto'', even if not 'de jure\ form of

minimal copyright.35

Nevertheless, as the balance between tacit knowledge/codified know-

ledge gradually shifts in favour of the latter, under the effects of appropri-ation mechanisms, there will be a gradual impoverishment of the sharing

process and hence of the creative commons.One (extreme) example of such a shift in balance would be the total elim-

ination of social relationships which occurs when all the individuals

belonging to a specific society/culture die. In such an eventuality it wouldbe possible to preserve the codified knowledge (objects and relics, cultural

products and literary texts, and information of all kinds); whereas, the tacit

knowledge (i.e., those elements pertaining to interactions among individ-

uals) would be irretrievably lost, as would be the culture and its knowledgeas a whole.

7.4 BACK TO THE BEGINNING: POLICYGUIDELINES AND LOCKEAN PRESCRIPTIONS

Under the proposed representation of creative commons, the social

dimension is central to the creation and existence of knowledge, in the

sense that sharing is an integral and characterizing part of that knowledge.It follows that the same criterion of dynamic efficiency invoked by the eco-

nomic theory of copyright in justification of the right must also challengethe existence of any institutions that significantly impair the sharing

process.

This is not, however, to cast doubt on intellectual property rights across

the board, but only to question the efficiency of their present-day structure

and enforcement. There is no doubt that, in many instances, the incentive

provided by minimal copyright can truly be effective. But one can with

equal validity maintain that creative commons have throughout human

history demonstrated their dynamic efficiency, even in the absence of copy-

right, with the production of quantitatively and qualitatively elevated levels


of information and knowledge.36 A comparative analysis in this domain

therefore risks raising a difficult debate, which in any case falls outside the

scope of the present work. However, it could justifiably be argued that,

because a market of information goods does exist today- and does succeed

in regulating the behaviour of creators and allocating their energies with

respect to other markets - its total elimination might not only prove unfeas-

ible, but also possibly endanger social stability. For the present purposes,

therefore, it is sufficient to treat copyright as a given- an institutional

outcome of western society, which has the attendant virtue of potentially

providing an incentive for the production of new expressions of ideas. The

real issue, though, is to examine how its extension might interfere with the

endogenous dynamics of creative contexts, and in this light attempt to for-

mulate appropriate policy indications.

One additional interesting test of the soundness of the above conclusions

is to see whether they are consistent and compatible with the totemic para-

digm formulated by John Locke (1690), which has been widely relied upon

by the literature for the justification of property rights (Drahos, 1996). The

reference is particularly apposite because this part of Locke's work focused

precisely on the problem of fairly regulating exploitation of the commons,so as to safeguard both the original resource and private appropriability,

defining rights and limits accordingly.

Now one might say that the solution proposed by the philosopher seems

almost to prefigure, at least from an instrumental perspective, the efficiency

objectives prescribed by modern economic theory, and that it is consistent

with the arguments presented here. In fact, alongside the legitimization of

property rights, the English philosopher formulated two complementary

provisos designed to simultaneously guarantee preservation and access to

the commons.

The first clause, the so-called 'sufficiency proviso', recognizes private

appropriability in reason of the individual contribution, but only admits

appropriation 'at least where there is enough and as good left for others'

(Locke, 1690, section 27). By so doing it places a first, weak limit upon indi-

vidual rights, acknowledging in practice the original collective ownershipof any common resource, and the precedence which this takes over private

appropriability.

The second clause, contained in the so-called 'spoilage proviso', reasserts

with greater force the necessity of preserving the integrity of commonresources, clearly calling for the limitation of property rights when these

can have significant effects leading to the impoverishment and depletion of

said resources. 'The same law of nature, that does by this means give us

property,' Locke wrote,'does also bound that property too. . . . Nothingwas made by God for man to spoil or destroy' (Locke, 1690, section 27).


An appropriate transposition of the above two criteria to the context of

copyright confirms the guidelines suggested by our analysis: a minimal

copyright which remunerates individual contributions can be acceptable

and useful for dynamic efficiency, but by this same criterion, given the pecu-

liar architecture of creative commons, it must leave 'enough and as goodfor the others', carefully avoiding their depletion. It follows that a transi-

tion to a maximal form of copyright, which significantly impairs the know-

ledge sharing process, is not to be desired.

In stark contraposition to the above, the recent international legislative

and jurisprudential trends (which should in theory be informed by the

Lockean categories) have instead been moving toward a strengthening of

exclusive rights through the extension of their duration (see US CopyrightTerm Extension Act 1998 and EU Duration Directive, 93/98/EEC),

through the international extension of the rights and their enforcement (see

TRIPs Agreements), through their extension to new subject matters

(see the EU Directive 96/9/EEC on databases), and through a progressive

erosion of the scope for derogation from the exclusive right (i.e., fair use

doctrine and analogous doctrines in the various national legislations).

In light of the preceding discussion, the present-day trend toward boost-

ing the exclusionary effect, without any counterbalancing correctives, runs

the risk of significantly interfering with the sharing process, and could in

the long run bring about conditions of social inefficiency, both in the

markets of ideas and for the creation of knowledge. What is more, because

sharing and communication are of such crucial importance to human

groupings,37 who may react in various ways to interference in these spheres,

society may well also incur additional and neglected costs of social and

legal conflicts. 38

7.5 CONCLUSIONS

Copyrighted works consist of signs which acquire significance by virtue of

belonging and referring to a broader semantic universe - the knowledgedomain - which is by its nature collective. Even when knowledge fragmentsare able to be encoded and occasionally transformed into commodities (i.e.,

codified knowledge) they remain indissolubly linked to that componentwhich exists and is reproduced within the sphere of human relationships

(i.e., the tacit knowledge).These two dimensions can never be entirely separated, and are equally

necessary for the transmission and production of knowledge. So given this

perspective, and even allowing due importance to individual contributions,

knowledge production emerges as being inextricably embedded within


social systems, with the locus of its production being the creative

commons - a special case of common resource which has the peculiar and

distinguishing trait of providing increasing returns with increasing

numbers of users, thus implying a pivotal role of the sharing process.

In other words, the collective setting is the only possible ecosystem for

the creation of knowledge, and as such needs to be preserved. The alterna-

tive hypothesis, rooted in a romantic view of the artist creating in splendid

isolation from the world, is not backed up by the history of art, nor by the

sciences that study knowledge and its production.

In general, the economic analysis of copyright (though not the economic

theory as a whole) falters in its handling of the above-described dynamic,

because application of its method relies upon a representation of reality

which simply does not hold in the case of knowledge. Nevertheless, this

analytical misrepresentation has led some of the literature to justify and

propose policies that are not applicable to the knowledge domain, and

which may have the unintended consequence of impoverishing knowledgeand creative commons.

The original architecture of copyright, which we have here termed

minimal copyright, appeared to take into account the peculiar features of

the knowledge domain, and was accordingly designed as a bounded prop-

erty right in order to avoid any significant interference with the sharing

process. History has also testified to the possibility of peaceful coexistence

between creative commons and bounded proprietary systems, provided the

latter do not significantly compromise the sharing process and hence, the

efficiency of knowledge production.

Today, however, the indiscriminate extension of copyright at the expense

of the social dimension of knowledge, which we have termed maximal

copyright, is liable instead to have the undesired effect of depleting the cre-

ative commons, to the detriment of knowledge.Given this perspective, it would therefore be socially desirable, as a

general rule, to avoid any policy designed to bolster private appropriability

at the expense of knowledge sharing, and instead maintain appropriability

at a minimal level and adopt (or at least preserve the existing) criteria for

derogation from strong exclusive rights that move in the direction of assur-

ing adequate communication flows.

This is not to imply a negation of the usefulness of copyright and intel-

lectual property in general in stimulating individual creation, but rather to

acknowledge that they play an ancillary role, not with respect to the

market, but certainly with respect to the production of knowledge and

culture.


ACKNOWLEDGMENT

This chapter was originally drafted during a visiting scholarship at SIMS, University of

California, Berkeley. I would like to thank Hal Varian for his kind hospitality there, and Jim

Stockinger for having introduced me, on that occasion, to the literature on the sociology of

culture. I am additionally indebted to a number of colleagues and friends for their helpful

comments, suggestions and criticisms, and in particular Jiirgen Backhaus, Alberto Cassone,Alberto Chilosi, Michael Einhorn, Angela Fraschini, Wendy Gordon, Dennis Khong, AnnaMaffioletti, Carla Marchese, Alain Marciano, Federico Revelli, Francesco Silva, Frank

Stephen, Ruth Towse, and an anonymous referee. I also wish to thank all the participants of

SERCIAC 2003 in Northampton, MA, the SIEP Conference 2003 in Pavia, the Alessandria

MIUR Workshop 2003 and the Erfurt Workshop in Law and Economics 2004. The usual dis-

claimers apply. This work was supported by the MIUR research grant on The Governance of

Intellectual Property'.

NOTES

1 . The adjective 'minimal' is a concise way to describe the nature of copyright in its ori-

ginal form -i.e., a heavily bounded property right designed to take into consideration

and pursue the public interest in the broadest sense, according to the precepts of the

standard theory of property and intellectual property (see Drahos 1996).2. However, this is not a particularly useful argument in favour of the efficiency of copy-

right, because human history has also shown that creative activities flourish equally in

the absence of copyright.3. See, for example, Posner (1992). On fair use doctrine, see Gordon (1982).4. On the subject of inventive activities in the patent domain, Arrow (1962, p. 129) seems

to have also intuited this peculiarity, warning that the production of knowledge can con-

stitute not only a productive output, but also an input for other inventive activities. If

this is the case, Arrow observes, the private appropriation procured by intellectual prop-

erty rights may seriously compromise the incremental accrual of knowledge and, conse-

quently, the collective well-being. Unluckily, here again, the observation is not further

followed up in his analysis.

5. In addition to the general contribution of Hardin (1 968), see for example, in economics,

Dasgupta and Heal (1979) and Comes and Sandier (1983).

6. A similar hypothesis in philosophical debate is advanced by Drahos (1996).7. It should be pointed out that there is also an opinion movement called 'creative

commons', whose aims are to promote access to knowledge and limit the exclusion

brought about by copyright through the introduction of novel contractual forms (see

http://creativecommons.org/). Now, although the working goals of this movement are

certainly consistent with the arguments presented here, there is no specific ideologicalconnection.

8. This idea is also supported by some contributions in the legal literature (see Rose, 1986)

dealing with public policy aspects (Rooney et al., 2003).9. This thesis finds wide support today in the economics literature, in Landes and Posner

(1989), Scotchmer (1991) and Foray (2000). Note that sociology, starting with the

contributions of Max Weber (1967) and Merton (1973), has remarked on the collective

dimension of inventive activities.

10. This assertion is also consistent with the Schumpeterian theory of innovation putforward in The Theory of Economic Development (1934).

1 1 . This is in essence also recognized by Landes and Posner ( 1 989).12. Note that Geertz, and anthropologists in general, often refer to the knowledge of a

specific social group as 'culture'. In this chapter we shall always be using the term

'knowledge'.


1 3. On network externalities see for example Liebowitz and Margolis (1 998). For the appli-cation of network externalities to the knowledge domain with specific reference to the

value of a language, see Church and King (1993).

14. The assertion is supported by growing consensus in contributions dealing with the eco-

nomic aspects of knowledge. Rooney et al. (2003, p. 8), for example, observe that '[a]

connected characteristic of knowledge is that it tends to grow through sharing. As

people exchange ideas in conversation the parties to the conversation tend to increase

their knowledge and in all likelihood create new knowledge. This is particularly so in the

case where ideas are shared between people with complementary knowledge working in

related areas. Furthermore, as knowledge diffuses its social value tends to be enhanced'.

Drahos (1996) also agrees on the peculiarity of creative commons.15. It could of course be appreciated for some other physical characteristic, but in this case

would no longer be an information good.16. On these topics see for instance, Schafer (1977).

17. This, in essence, is analogous to establishing a 'common language' (see Nahapier and

Ghoshal, 1998).

18. Of course this does not mean that the sharing process cannot take place in a physical

space. Universities and research institutions, for example, are physical spaces set up pre-

cisely to encourage the development of creative commons, but they are naturally not cre-

ative commons merely by virtue of being spaces.

19. Some authors, dealing for the most part with innovation, have similarly argued that the

production of knowledge does not follow a conventional linear production process, but

instead exhibits a non-linear and collaborative dynamic that is 'not well understood bypolicymakers' (Rooney et al., 2003, p. 69). Cowan et al. (2000), in emphasizing the speci-

ficity of knowledge production, also speak of an 'enculturation model'.

20. This is the case in the fashion industry, but also in the music industry where the majors

generally profit from external Schumpeterian innovators more closely tied to the social

context (see Silva and Ramello, 2000).21 . Note that, though it was subsequently forgotten, this is a criticism that was first levelled

against the traditional economic approach as early as the middle of the twentieth

century. Paul Sweezy, in 1942 (p. 3), already observed that: 'Society is more than a

number of individuals. It is a number of individuals among whom certain definite andmore or less stable relations exist. The form of society is determined by the character andform of these relations'.

22. In this connection, Cowan et al. (2000, p. 217) point out that '[fjor some considerable

time, economists took little if any interest in the question of separating the notion of

knowledge from their idea of information'.

23. For an overview, see Cowan et al. (2000). The fact that tacit knowledge is physically held

by single individuals does not negate its social nature. In effect, because there is no such

thing as a physical entity called 'social relationships', it follows that individuals must be

the 'bearers' of this type of knowledge. But the expression and transmission of tacit

knowledge nevertheless requires interaction between individuals to take place.24. This statement is consistent with the ideas in the literature on communication theory (see

for example, for the 'Palo Alto' school, Watzlawick et al. 1967).

25. The above-described dichotomy between knowledge and information, which attributes

to the former a social dimension that cannot be completely owned by a single individ-

ual, is widely accepted in the social sciences (see Geertz, 1973; but also Weber, 1967, and

Durkheim, 1973).

26. This theory was subsequently taken up, among others, by Polanyi (1958, 1966), Cowanet al. (2000), David and Foray (2002), and Nelson (2003).

27. Nelson and Winter (1982, p. 78) also argue that 'a trait that distinguishes a good instruc-

tor is the ability to discover introspectively, and then articulate for the student, much of

the knowledge that ordinarily remains tacit'.

28. For example, the execution of a musical score requires the ability to read and interpret

it, as well as to perform it on the instrument (Rooney et al., 2003).29. See for example Williams and Gibson (1 990) and Takii (2004).


30. In general, looking at OECD data, we can easily observe that, in developed countries,

the public sector funds at least one-third of R&D expenditure. Additional public con-

tributions then go to educational systems and cultural activities.

31. Consider, in this connection, a simple example: the preparation of a dish involves two

components: the written recipe (codified knowledge) and the practical ability to execute

it (tacit knowledge). If the opportunity to cook a dish is precluded- in other words, if

its recipe is appropriated in an exclusive manner - this will also preclude handover of the

technical ability to execute it, because the cook who knows how to make it will not be at

liberty to teach it.

32. Cowan et al. (2000, p. 225): 'But information written in a code can only perform those

functions when people are able to interpret the code'. So we are once again referred back

to the social dimension.

33. Note that, in this case too, the law appears to take an instrumental view of copyright. It

therefore provides for derogations designed to avoid static inefficiency (negative balance

of welfare without a significantly increased incentive for the copyright holder) or

dynamic inefficiency (endangering certain incremental creations, e.g., satire which plays

an important social role in warranting freedom of expression).

34. Some relatively recent scientific work dealing with social networks has noted how human

relationships are characterized by an exogenous degree of instability, and that this state

of affairs promotes efficient results (see Jackson and Wolinsky, 1996; Dutta and

Mutuswami, 1997).

35. For a taxonomy of unauthorized reproduction, see for example Silva and Ramello

(2000).

36. For western society, see Weiner (2000). For China, see the intriguing book by Alford

(1995).

37. See note 25 above.

38. Consider for example the Napster case, which, albeit indirectly, involved university stu-

dents, the future intelligentsia of US society, and high school students. If, in a similar

case in the future, a massive law enforcement action were to directly target these cat-

egories of infringers, it could lead to unsustainable social consequences.

REFERENCES

Alford, William P. (1995), To Steal a Book is an Elegant Offense, Stanford, CA:Stanford University Press.

Arrow, Kenneth J. (1962), 'Economic welfare and the allocation of resources for

invention', in NBER, The Rate and Direction of Incentive Activity: Economic andSocial Factors, Princeton, NJ: Princeton University Press, pp. 609-24.

Church, J. and I. King (1993), 'Bilingualism and network externalities', Canadian

Journal of Economics, 26, 337-45.

Cornes, R. and T. Sandier (1983), 'On commons and tragedies', American Economic

Review, 73, 787-92.

Cowan, R., PA. David and D. Foray (2000), 'The explicit economics of knowledgecodification and tacitness', Industrial and Corporate Change, 9, 21 1-53.

Dasgupta, Parta and Geoffrey Heal (1979), Economic Theory and Exhaustible

Resources, Cambridge: Cambridge University Press.

David, PA. and D. Foray (2002), An introduction to the economy of the knowledge

society', International Social Science Journal, 54, 9-23.

Drahos, Peter (1996), A Philosophy ofIntellectual Property, Dartmouth: Aldershot.

Durkheim, Emile (1973), On Morality and Society: Selected Writings, Chicago andLondon: University of Chicago Press.


Dutta, B. and S. Mutuswami (1997), 'Stable networks', Journal ofEconomic Theory,

76, 322^4.

Foray, Dominique (2000), L'economie de la connaissance, Paris: La Decouverte.

Geertz, Clifford (1973), The Interpretation of Cultures, New York: Basic Books.

Gordon, WJ. (1982), 'Fair use as market failure: A structural and economic analy-

sis of the Betamax case and its predecessors', Columbia Law Review, 82, 1600-57.

Granovetter, M. (1985), 'Economic action and social structure: The problem of

embeddedness', American Journal of Sociology, 91, 481-510.

Hardin, G. (1968), 'The tragedy of the commons', Science, 162, 1243-8.

Hayek, F. von. (1945), 'The use of knowledge in society', American Economic

Review, 35, 519-30.

Hettinger, B.C. (1989), 'Justifying intellectual property', Philosophy and Public

Affairs, 18,31-52.

Hippel, Erik von (1988), The Sources of Innovation, Oxford and New York: Oxford

University Press.

Jackson, M.O. and A.Wolinsky (1996), 'A strategic model of social and economic

networks', Journal of Economic Theory, 71, 4474.

Koepp, R. (2002), Clusters of Creativity: Enduring Lessons on Innovation and

Entrepreneurshipfrom Silicon Valley and Europe's Silicon Fen, Chichester: Wiley.

Landes, W.M. and R.A. Posner (1989), 'An economic analysis of copyright law',

Journal of Legal Studies, 18, 325-63.

Liebowitz, Stan J. and Stephen E. Margolis (1998), 'Network effects and external-

ities', in P. Newman (ed.), The New Palgrave's Dictionary of Economics and the

Law, vol. II, London: Macmillan, pp. 671-5.

Locke, John (1690), Second Treatise on Government, URL: http://www.ilt.colum-

bia.edu/academic/digitexts/locke/second/locke2nd.txt.

McQuail, Dennis (1983), Mass Communication Theory: An Introduction, London:

Sage.

Merton, Robert K. (1973), The Sociology of Science. Theoretical and Empirical

Investigations, Chicago and London: Chicago University Press.

Morin, Edgar (1962), L 'esprit du temps, Paris: Grasset.

Nahapier, J. and S. Ghoshal (1998), 'Social capital, intellectual capital and the

organizational advantage', Academy of Management Review, 23, 253.

Nelson, R.R. (2003), 'On the uneven evolution of human know-how', Research

Policy, 32, 909-22.

Nelson, K. and R.R. Nelson (2002), 'On the nature and evolution of human know-

how', Research Policy, 31, 719-33.

Nelson, Richard R. and Sidney G. Winter (1982), An Evolutionary Theory ofEconomic Change, Cambridge, MA: Belknap Press.

Polanyi, Michael (1958), Personal Knowledge: Towards a Post-Critical Philosophy,

London: Routledge.

Polanyi, Michael (1966), The Tacit Dimension, New York: Doubleday.

Posner, Richard A. (1992), Economic Analysis ofLaw, Toronto: Little Brown & Co.

Romney, A.K. (1999), 'Culture consensus as a statistical model', Current

Anthropology, 40, 103-15.

Rooney, David, Greg Hearn, Thomas Mandeville and Richard Joseph (2003), Public

Policy in Knowledge-Based Economies, Cheltenham, UK and Northampton,MA, USA: Edward Elgar.

Rose, C. (1986), The comedy of the commons: custom, commerce, and inherently

public property', University of Chicago Law Review, 53, 71 1-81.

Private appropriability and sharing of knowledge 141

Schafer, Murray R. (1977), The Tuning of the World, Toronto: McLelland and

Stewart.

Schumpeter, Joseph (1934), The Theory of Economic Development, Cambridge,MA: Harvard University Press.

Scotchmer, S. (1991), 'Standing on the shoulders of giants: Cumulative research

and the patent law', Journal of Economic Perspectives, 5, 29-41.

Silva, F. and G.B. Ramello (2000), 'Sound recording market: the ambiguous case of

copyright and piracy', Industrial and Corporate Change, 9, 415-42.

Sweezy, Paul M. (1942), The Theory of Capitalist Development, Oxford and NewYork: Oxford University Press.

Takii, K. (2004), 'A barrier to the diffusion of tacit knowledge', Review of

Development Economics, 8, 81-90.

Watt, Richard (2000), Copyright and Economic Theory: Friends or Foes?,

Cheltenham, UK and Northampton, MA, USA: Edward Elgar.

Watzlawick, Paul, Janet Bavelas Beavin and Don D. Jackson (1967), Pragmatics ofHuman Communication, New York: Norton.

Weber, Max (1967), // lavoro intellettuale come professione, Einaudi, Turin (origi-

nally published as Wissenschaft als Beruf).

Weiner, Robert P. (2000), Creativity and Beyond, New York: State University of

New York Press.

Williams, Frederick and David V. Gibson, (eds) (1990), Technology Transfer: ACommunication Perspective, Newbury Park, CA: Sage.

8. IMS Health or the questionwhether copyright still deserves

a specific approach in a market

economy?Alessandra Narciso and

Paul L.C. Torremans

8.1 INTRODUCTION

The ongoing IMS Health saga has yet again highlighted the apparentclash between copyright and competition law and its essential facilities

doctrine. The IMS Health case is a complex one that is still unfoldingbefore the EU Court of Justice. IMS is a major supplier of marketingdata to pharmaceutical and other healthcare companies. In the German

market, it has established a structure of local geographic segments, 1860

in total, called bricks, each containing a comparable number of phar-macies. Such a structure enables the collection of standardized data

without violating data protection laws that prohibit the identification of

individual pharmacies' data. That structure is protected by copyrightunder German copyright law. The brick structure has de facto becomethe industry standard and IMS's competitors had asked for it to be

licensed to them in order for them to be able to compete. The grant of a

licence was refused, which subsequently prompted IMS's competitors to

file a complaint with the European Commission for abuse of dominant

position.

The facts and developments in the IMS Health case will be introduced

subsequently, as will some of the detailed legal issues that arise in the case,

with a focus on the essential facilities doctrine. However, the detailed issues

need to be seen against the background of the apparent clash between

copyright and competition law.

142

IMS Health or does copyright deserve a specific approach? 143

8.2 COPYRIGHT VS. COMPETITION LAW

Copyright

The starting points

Two leading studies on the economic analysis of copyright support the con-

tinued existence of copyright. The starting point indicates therefore that the

debate should not develop along the lines copyright or competition law, but

should rather focus on the role played by both sets of rules in a market

economy (Lemarchand et al., 2003, p. 17).

Two studies highlight the conundrum we are trying to solve, the solution

of which we believe should underpin any detailed ruling in cases such

as IMS. In their study, Cooter and Ulen (1988, p. 145) summarize it as

'[p]ut succinctly, the dilemma is that without a legal monopoly too little

of the information will be produced, but with the legal monopoly too little

of the information will be used'. Landes and Posner (1989, p. 326) empha-size the matter in even clearer terms when they state that 'Copyright

protection- the right of the copyright's owner to prevent others from

making copies- trades off the costs of limiting access to a work against the

benefits of providing incentives to create the work in the first place. Striking

the correct balance between access and incentives is the central problem in

copyright law.'

Historically, copyright is concerned with the protection of literary and

artistic works such as books and sculptures. It is important to note in this

respect that copyright never set out to protect ideas. What was protected

was the way in which the individual author used these ideas creatively and

the resulting expression of his creativity. Only that expression, the individ-

ual non-copied expression by the author of a certain idea was protected.

A lot of emphasis was therefore put on the link between the author and

his work (Torremans, 2001, p. 176).

Justifying copyright

This approach can in part be justified by means of John Locke's labour

theory, which is in essence the combination of two concepts. According to

the first concept, everyone has a property right in the labour of his own

body and brain; the second concept builds on that by adding that the appli-

cation of human labour to an unowned object gives whoever applies that

labour to it a property right in the previously unowned object. If one

applies this to copyright (Nozick, 1974, pp. 181-2), one can see why the

starting point of any copyright law is that the author gets the copyright in

the book or sculpture. However, the theory does not go a lot further than

the issue of the allocation of the (copy)right. It does not explain why an


immaterial right needs to be created and why such a right is necessarily an

object for the purposes of the theory. There is therefore a need for an add-

itional element that justifies the existence of copyright and explains why an

immaterial right in the author's individual expression of an idea needs to

be created (Spector, 1989, p. 273).

That other element is purely economic in nature (Torremans, 2001,

pp. 1421). In an efficient free market structure, competition takes place on

both the production and creative/innovation levels. As was noted above, in

the copyright sphere we are dealing with works that are the expression of

ideas. Starting from these ideas, one has to recognize that they are by their

nature public goods and can therefore freely be accessed and used by

anyone. The way these ideas enter the public domain is through their

expression by individual authors, as such expression is required for the

transmission of the ideas. From an economic point of view, it is also

important to keep in mind that such access is non-exhaustive in nature. The

consumption of the expression does not necessarily make the expressionand its material support unsuitable or unavailable for further consumption.It is also the case that in light of modern (digital) technological advances,

reproduction and distribution of the expression of ideas can be readily and

quickly achieved, and at a very nominal cost. There is therefore plenty of

room for free-riders. In the absence of copyright, the situation is therefore

entirely in favour of competition at the production level. At the innovation

and creation level, there is very little incentive to create. The creator maynot be able to recoup the cost of production, as there is no tool to reap anysubstantial benefit from such creative activity. There is therefore no efficient

market for the author's expression of ideas (Ramello, 2003, p. 8).

Copyright is the tool that is created to give authors a right in their expres-sion of ideas, hence securing for them appropriate profits deriving from the

act of creation. Copyright leads to the creation of an immaterial property

right in the expression of an idea by the author, a right which the author

can use to secure appropriate profit from his or her act of creation on the

market (Maskus, 2000, pp. 28-32). This will enhance creation by providing

incentives; therefore, competition on the innovation and creation level will

be stimulated, whilst any such right will inevitably limit competition at the

production level, as competitors are no longer free to copy the copyrightwork. A restriction on competition is put in place in furtherance of com-

petition (Lehmann, 1989, p. 12).

Competition Law

Competition law aims to maximize social welfare by avoiding or correcting

specific market failures. Consumer welfare should, as a result, be enhanced


and innovation should be encouraged. Competition law attempts to

achieve this objective through the elimination of anti-competitive practices

and forms of behaviour, as these lead to inefficient outcomes.

The agreement between lawyers and economists in this area clearly

covers the point that competition law is an instrument to combat monop-olistic behaviour that results in the exclusion of a segment of consumers

from the market, relative to the competitive context. But when it comes to

the practical application of these principles to the facts of a specific case,

economic theory is incapable of providing the legal community and the

courts with specific tests to detect and correct such anti-competitive behav-

iour (Neumann, 2001), especially in relation to the new information

economy. Economic theory is still largely focused on static efficiency and

the dynamics of the manufacturing industries. The information economyhas quite different dynamic processes. Since copyright operates increasingly

in such an information economy, when it is used to protect information

goods, it is evident that the application of competition law in such cases

may not be clear-cut.

Copyright vs. Competition Law

As copyright confers an exclusive right, there is by definition a possibility

that such an exclusive right will be used to restrict competition. IMS Health

shows this clearly. The exclusive right in the brick structure can be used to

restrict access to the information supply market by refusing to license the

vital elements to potential competitors. There may therefore be a conflict

between competition law and copyright law in a given situation. Copyright

preserves the exclusive right and competition law prima facie has a problemwith whatever restricts competition.

In order to understand the nature of such a potential conflict better, one

needs to see how both regulatory systems, i.e., copyright and competition

law, operate. Copyright operates at a structural level. It puts in place a

rewards and incentives structure that applies to all relevant cases and aims

to enhance competition at the creation and innovation level by granting an

exclusive right in the copyright work, which, at the same time, restricts to

some extent competition at the production level. Copyright, therefore,

defines rights. Competition law, on the other hand, operates at the behav-

ioural level and it does so on a case by case basis. It is therefore not the case

that there is a conflict between two sets of rules that operate in the same

way and at the same level (Ramello, 2003, p. 28). Copyright and competi-

tion law share the same objective, but they have different ways of achiev-

ing it and they operate at different levels (Lemarchand et al., 2003,

pp. 17-19).


Competition law does not operate at the structural level and as such, it

does not interfere with the structure created by copyright to enhance

creation and innovation. However, competition law may interfere at a later

stage with some of the behaviour that flows from the grant of copyright.

What can be derived at this stage is that a system that is designed to regu-

late behaviour should not operate in a way to undo the structural system

put in place by a system such as copyright. The legislator has made a policy

decision to put a structure such as copyright in place to enhance and reward

creativity and innovation and that decision should be respected. In other

words, competition law should restrict itself to its proper role which is to

address anti-competitive behaviour, such as any use of copyright that is not

pro-competitive as was envisaged when the structural system was set up(Torremans, 2001, pp. 302-9).

8.3 THE IMS HEALTH CASE

Background

TheIMS Health case is a complex saga that is still unfolding before the Court

of Justice. IMS is developing along two parallel tracks on the basis of the fol-

lowing facts. IMS is a major supplier of marketing data to pharmaceuticaland other healthcare companies. In the German market, it has established a

structure of local geographic segments, 1 860 in total, called bricks, each con-

taining a comparable number of pharmacies. Such a structure enables the

collection of standardized data without violating data protection laws that

prohibit the identification of individual pharmacies' data. That structure is

protected by copyright under German copyright law. The brick structure has

de facto become the industry standard and IMS's competitors had asked for

it to be licensed to them in order for them to be able to compete. The user is

indeed only interested in data that are presented in a comparable fashion.

The grant of a licence was refused and that resulted in a complaint to the

European Commission for abuse of dominant position. In due course, the

Commission issued its decision, in which first-stage interim measures were

imposed. That Commission decision was appealed, and resulted in the two

Orders of the President of the Court of First Instance 1 and one Order from

the President of the Court of Justice2 that suspended those interim measures.

At the same time though, IMS's competitors had decided to make use of

the brick structure anyway, without waiting for the grant of the licence.

IMS therefore brought copyright proceedings against them in the Germancourts. The German courts referred the case to the Court of Justice for a

preliminary ruling.3


The Commission Decision

General points

When IMS's competitors NDC and AzyX applied for a licence covering the

brick structure, the grant of such a licence was refused by IMS. NDC and

AzyX subsequently filed a complaint with the European Commission for

abuse of dominant position.

The Commission issued a decision4 in which it argued that there was

indeed an abuse of a dominant position under Article 82 and that the

urgent nature of the matter required the imposition of interim measures to

avoid irreparable damage being done. 5

Let us consider the process followed by the Commission in giving this

decision and try to find the justifications (if any) for the need to introduce

interim measures to address the abuse of dominant position by IMS. The

Commission first of all identifies the relevant geographical market,

Germany,6 whereas the relevant product market is the regional sales data

services. 7

As regards the position of dominance, IMS is found to be dominant after

examining several factors, but the Commission primarily refers to the fact

that IMS holds a high market share not only in Germany but also in

Europe;8Germany, being the country that has 'the largest market for

regional sales data services in Europe'9 may therefore be regarded as a 'sub-

stantial part of the common market' in this respect.10

In order to understand the background to the case, it is useful to know

what regional sales data services are. Pharmaceutical companies rely on the

brick system because it enables them to gather data on sales of a particular

drug and compare sales figures for each of their products with those for the

products of their competitors. It also offers them a tool to measure the per-

formance of their sales representatives.] l In the brick structure, these data

are collected in each brick for a small number of pharmacies in a local area.

This allows the pharmaceutical companies to obtain a detailed picture

without infringing German data protection laws, under which the use of

data relating to individual doctors or pharmacies would be prohibited.

The Commission concludes that IMS occupies a position of quasi-

monopoly12 and this is, according to the Commission, due to the fact that

IMS owns the copyright in a unique structure for recording sales data for

pharmaceutical products and services and that in the relevant market there

was simply no competition before the arrival of NDC and AzyX. The

Commission argues that IMS obtained this quasi-monopoly position

through a close collaboration with the German pharmaceutical industry,

which led them to the creation of the brick structure. 13 The exact shape of

the brick structure is therefore largely the outcome of that collaborative


effort. That involvement of the primary customer in the establishment of

the structure also helps to explain why this structure seems to have been

indispensable for other competitors that found it impossible to build a new

structure; the pharmaceutical industry was unwilling14 to switch to another

structure, given the significant additional costs they would incur from

doing so. 15

The Commission is further sustaining that the IMS brick structure is

becoming a 'de facto' industry standard. For the Commission, the 'I860

brick structure is becoming a common language for communicating the

information' 16 and all the pharmaceutical companies which assisted with

its creation are in the end, effectively 'locked in' that structure on a volun-

tarily basis. The 1860 brick structure is especially a great and unique source

of information 17(and here the Commission sees parallels with the Magill

case, where broadcasters' TV listings were a unique source of informa-

tion 18).

The Commission concluded, therefore, that IMS should grant all its

competitors a licence upon payment of a reasonable royalty for the use of

its copyright work. The Commission justifies its conclusion as follows:

there is an abuse that is due to the dominant position occupied by IMS and,

moreover, IMS owns an essential facility that is impossible to substitute.

Reliance on the Oscar Bronner case

In its decision, the Commission refers repeatedly to the essential facility

doctrine, as set out by the Court of Justice in Oscar Bronner. 19

Bronner was not a case concerned with intellectual property rights, but it

may be important because it is clearly a case about essential facilities.

Bronner could therefore shed light on the correct interpretation of the

essential facilities doctrine in European competition law. Bronner dealt with

a system for the house-to-house distribution of newspapers in Austria set

up by the major national newspaper. A smaller competitor argued that it

should be allowed to make use of the system in order to be able to compete.In its view, the system had become an essential facility. The Court of Justice

followed the suggestion of its Advocate General to interpret the essential

facilities doctrine restrictively and then ruled that it did not apply here.

Important for our current purposes is the establishment of a test that a firm

violates Article 82 where:

'the refusal of access to the facility is likely to eliminate all competi-tion in the relevant market;'

'such refusal is not capable of being objectively justified;' and

'the facility itself is indispensable to carry on business, inasmuch as

there is no actual or potential substitute in existence for that facility'.20


However, the Commission forgets to mention that Bronner is a case that

can only show how the essential facilities doctrine can be applied in a

European competition law context in general, being a case of pure compe-tition rules application. It is indeed vital to recall that in a normal situation

competition rules are the only set of rules used to regulate market behav-

iour. In copyright cases, there is not just one set of rules that affects the

market behaviour regulation point. Both copyright and competition rules

set out to regulate competitive conditions on the market. The presence of

such a second set of rules must have an impact on the test that is used,

otherwise its vital role is completely ignored. It is therefore submitted that

it is not possible to transfer the test set out in the Bronner (Treacy, 1998 and

Stothers, 2001) case to copyright cases without modifying it. This must

mean that any strict reliance by the Commission on the Bronner case in its

decision must be misguided.

Reliance on the Magill case

To be fair to the Commission, it also argues that its decision follows from

the judgments in Magill21 and Ladbroke 22

It is by now well known that the Magill case started when Magill was

refused a licence covering the copyright in the TV listings. In its decision

the Commission23 ruled that the broadcasters had abused their dominant

position in exercising the exclusive right awarded to them by copyright.

The Commission ordered that advance programme listings be supplied to

Magill. The broadcasters then took the case to Court of First Instance, but

they were disappointed.24

By then the case had become one of principle. On appeal,25 the Court of

Justice's judgment did not, however, change a great deal, the Court reach-

ing the same conclusion as the Court of First Instance. Moreover, the

wording of the judgment was oriented almost exclusively towards compe-tition law principles and the resemblance with the ICI and Commercial

Solvents vs. Commission26approach, which relies on elements of the US

essential facilities doctrine. One should be reminded though, that in Magill,

the Court first of all confirmed that the mere ownership of an intellectual

property right, such as the copyright in the programme listings, does not as

such confer a dominant position.27

Furthermore, even a firm holding a

dominant position can refuse to grant a licence without it constituting an

abuse of the dominant position. The Court then went on to say that in

exceptional circumstances, the exercise of an exclusive intellectual property

right by a proprietor, e.g., a refusal to grant a licence, can involve abusive

conduct that is prohibited by Article 82 of the Treaty.28 Such a prohibition

can be enforced by the Commission through the imposition of an obliga-

tion to grant a licence.29


The exceptional circumstances referred to by the Court in relation to the

broadcasters' refusal to licence the programme listings in Magill comedown to two elements. By refusing to make programme listings available to

Magill under licence, the broadcasters prevented the appearance of a new

product that they did not offer, but for which there was a clear consumer

demand. It is important to add that as a result of their exclusive copyright

in the listings, the broadcasters were the only source for any such informa-

tion. The second element is that by denying access to that vital material,

without which the new product could not see the light of day, the broad-

casters were effectively reserving the secondary market for weekly TVguides to themselves. The Court clearly emphasizes the secondary market

point by indicating that the broadcasters' main activity is broadcasting.

At this stage, one clearly sees the appearance of the essential facilities

doctrine. Magill can therefore be seen as an application of the essential

facilities doctrine in a copyright case, whereas Bronner applies the same

doctrine to a 'normal' non-intellectual property case, as discussed above

(Woolridge, 1999; Treacy, 1998; and Stothers, 2001). It is also important to

highlight the fact that the Court emphasized that there was a clear and

unmet demand on the consumers' side. Like all US essential facilities cases,

Magill also links the essential facilities doctrine to the benefit that arises for

the consumer from its application (Areeda, 1989).

When the Commission refers to the Court's judgment in Magill, it sum-

marizes the position adopted by the Court as follows:

The Court therefore recognised that in exceptional circumstances the exercise of

an exclusive right deriving from an intellectual copyright may be abusive even in

the absence of additional abusive conduct when, inter alia, it prevents the

appearance of a new product.30

This summary is causing problems in as far as the use of the term 'inter alia'

is possibly a misrepresentation of the Court's judgment in Magill. Indeed,

the Commission seems to suggest that the appearance of a new product is

not a necessary element and that other elements might replace it. It is sub-

mitted that it is wrong to use the term 'inter alia? in this way and to give this

interpretation to thejudgment of the Court. Instead, it is submitted that the

'inter alia' type of exceptional circumstances which the Court withheld in

Magill amounts to the fact that the information concerned had become an

essential facility for the publishers of TV guides (Torremans, 2001, p. 307).

The judgment in Magill supports the point of view that the application of

the essential facilities doctrine in relation to intellectual property (i.e., for

there to be an abuse of the dominant position, the possession of which is

not an abuse per se), requires the presence of all elements set out above.

IMS Health or does copyright deserve a specific approach? 1 5 1

Reliance on the Ladbroke case

The Commission also argues that its view is supported by the judgment of

the Court of First Instance in Ladbroke. 31 This case dealt with the question

of whether a refusal by PMU to license sound and pictures of French horse

races to Ladbroke's betting shops amounted to an abuse of a dominant

position to which the approach adopted in Magill should be applied

(Fitzgerald, 1998).

The Court of First Instance came to the conclusion that the facts of the

Ladbroke case did not warrant the application of the rules set out in

Magill. That conclusion was based almost entirely on the fact that the

images of the horse races were not an essential element for Ladbroke.

Ladbroke did not need the images to enter the market. Despite the fact that

PMU copyright controlled the only access to the images, the essential facil-

ities doctrine did not operate here. PMU could also not be accused of

reserving any market for themselves, as they were not even active on the

betting shop market. The real issue was that Ladbroke wanted to offer its

clients these images as an additional service to its main betting activity in

its shops.

One could argue that the main reason for the Court of First Instance not

to examine the point of the emergence of a new product is the fact that if

one condition is not met, the case was bound to fail as the conditions apply

cumulatively. But one could go further, as the Commission seems to be

doing in its IMS Health decision, and highlight the Court of First Instances

statement that the:

refusal to supply the applicant could not fall within the prohibition laid down

by Article [82] unless it concerned a product or service which was either essen-

tial for the exercise of the activity in question, in that there was no real or poten-tial substitute, OR [our emphasis] was a new product whose introduction mightbe prevented, despite specific, constant and regular potential demand on the part

of the consumer.32

Taken on its own, this quote could suggest that there is no need to establish

that the abuse takes place on a downstream market before Magill can be

applied and, even more importantly, that the conditions found in Magill

might be alternatives, which excludes the need for their cumulative appli-

cation. It is submitted though, that this is not a correct approach. The

whole approach hinges on the literal interpretation out of context of the

single word 'or'. If one puts it in context, that interpretation becomes

difficult. The Court of First Instance felt hardly any need to investigate the

second point (Hull et al., 2002, p. 37). They dismissed the case on the first

ground. Anything they said about the second point is therefore obiter (i.e.,

only incidental).


The Commission's interpretation also disregards the economic analysis

set out above. The approach may well work in a non-intellectual property

context, such as Bronner. Here, the requirements do not need to be applied

cumulatively. But we already explained that the presence of intellectual

property rights and their competition regulative function change the

picture dramatically. It is therefore submitted that the Commission is mis-

guided in its reliance on Ladbroke for its views in the IMS decision (Hullet al., 2002, p. 37).

The Commission's overall conclusion in its IMS decision - that there wasindeed an abuse of a dominant position, based on the fact that it saw the

brick structure as an essential facility before applying its liberal interpreta-

tion of Magill to this case, must therefore become open to serious doubts

as to its correctness.

8.4 IMS IN COURT

The proceedings before the president of the Court of First Instance andbefore the President of the Court of Justice in the IMS case33 were up to

now primarily concerned with the interim measures issue.

From the discussion above it has become clear that it is at least arguablethat Magill had left open several issues. 34 These were primarily open to

argument because it was not clear whether the decision on that point in

Magill depended entirely on the facts of the case or whether they were nec-

essary elements from a legal point of view. In fairness to the Commission,all it was trying to do was to fill in the gaps and clarify the matter, albeit in

a somewhat peculiar way.

One of the issues left open is the question as to whether or not excep-tional circumstances required cumulatively the fact that the intellectual

property right was linked to essential inputs for secondary markets and that

a new product had to be introduced in that market for which there was sig-

nificant and unmet demand. The Commission answers that question clearly

in the negative, widening the application of Article 82 significantly.35 The

Commission flatly stated, in addition, that 'there is no requirement for a

refusal to supply to prevent the emergence of a new product in order to be

abusive'. 36Similarly the Commission dismisses the requirement that the

dominant firm's refusal to licence and grant access to the copyright workmust enable it to restrict competition on a second (downstream) market. 37

IMS counters these arguments with references to other aspects of the

Court's case law. It points out that in order for the use of intellectual prop-

erty rights to become an abuse, there must be exceptional circumstances.

These can be found where the refusal to licence is combined with other


unlawful behaviour or where there is an essential facility at stake. 38 The

notion of essential facilities can, however, in its view only apply where two

distinct markets are involved and the product or service supplied in one

(usually an upstream) market is a necessary input for the production of

goods or services in the second (usually downstream) market.

In his Order of 26 October 2001, the President of the Court of First

Instance summarizes IMS Health's argument further as follows:

There are two essential aspects to the claim advanced by the applicant on the

basis of this case-law. First, it contends that the essential facility doctrine pre-

supposes the use by a dominant undertaking of its market power on the market

where it is dominant in order to preclude or undermine competition on a down-stream or neighbouring market where it already operates or wishes to operate,

or which it wishes simply to prevent, as in Magill, from emerging as a newmarket. Secondly and referring specifically to Magill ... IMS contends that it

was the rightholders' use of the copyright acquired as a result of their activities

on the market where they were dominant (broadcasting) in order to extend that

dominance into a downstream market (weekly television guides) that amountedto the 'exceptional circumstances' justifying the characterisation of their refusal

to license as an abuse. IMS Health's refusal, in this case, to license its competi-tors to use the 1860 brick structure is justified because it is not seeking to exploit

its market power in a separate market. It cannot be abusive, the applicant insists,

for a copyright holder to refuse to license competitors who wish to have access

to its right in order to compete against it on the very market where the exclusiv-

ity granted by that right is used as the central feature of the rightholder's busi-

ness and where its dominance may, at least potentially, turn on the maintenance

of its exclusive right.39

This is indeed important. In Magill, the broadcasters were the sole suppli-

ers of the information on the TV programmes, as only they knew what they

were going to broadcast. Magill could not develop alternative listings. In

IMS, however, competitors in the data services market can develop other

formats. Here too, the Commission's approach would not simply be an

application of Magill, but a massive extension of it.

Additionally, the Commission is unable to point towards any new

product for which there was a significant unmet demand. The President of

the Court of Justice approved, in this respect, of the ruling by his counter-

part in the Court of First Instance. Serious consideration was therefore

given to IMS Health's view; that is, existing case law required that the

refusal to licence must prevent the emergence of a new product in a market

that is separate from the one in which the dominant undertaking refusing

to licence is operating.40

This suggests that doubts still surround the Commission's ultra-liberal

interpretation of Magill, Ladbroke and Bronner in relation to copyright.

These doubts are re-enforced by the fact that the President of the Court of


First Instance confirms the importance of copyright, which points towards

the conclusion that intellectual property rights may deserve and require

some form of special treatment.

Sticking with the latter case for a second, one can also raise doubts as to

whether the Commission interpreted it correctly on another point. ThePresident of the Court of First Instance ruled that:

[i]t cannot ... be excluded that the balance of interests effected in the contested

decision by the Commission, which seems to equate the interests of NDC and

AzyX with the interests of competition . . . ignores the primary purpose of

Article 82 EC, which is to prevent the distortion of competition, and especiallyto safeguard the interests of consumers, rather than protect the position of par-ticular competitors.

41

8.5 CONCLUSION

Where does all this lead us and what should be the way forward? It is hard

to say where the Court of First Instance and the Court of Justice will even-

tually take European law when they render theirjudgments in the IMS case.

We feel a lot of sympathy for the approach taken and defended by IMS.

This sympathy is based on sound economic arguments. But this debate is

clearly not limited to the facts of the case at hand. What we are really con-

cerned with is the fundamental interaction between copyright and compe-tition law, which, as could be derived from the structure of the chapter,

requires a heavy reliance on fundamental principles in order to understand

both concepts. That interaction cannot be simplified to the question as to

which of the two prevails. That is exactly the wrong approach and cannot

be justified. Intellectual property rights have a positive role to play in a

modern free market society. Like competition law, they serve the purposeof promoting innovation and enhancing consumer welfare. It is therefore

imperative that the presence of an exclusive right, such as copyright or anyother intellectual property right, be taken into account as an additional

factor when cases such as Magill and IMS (and other cases where an intel-

lectual property right allegedly leads to competition law infringements) are

examined. Competition law has indeed as one of its characteristics that all

elements of the market in which competition takes place are taken into

account when its rules are applied. All elements of the competitive struc-

ture of the market are relevant in this respect.

How is that balance to be achieved and how are all these elements to be

taken into account? It is submitted that the approach in Bronner cannot be

applied unreservedly, as it only looks at competition law when establishing

a test for essential facilities cases. Bronner addresses cases where there are no

IMS Health or does copyright deserve a specific approach? 1 55

additional elements of the competitive structure of the relevant market that

are to be taken into account. That is not the market situation with which we

are concerned here. As an additional right that regulates competition by

imposing restrictions at one level of competition in furtherance of competi-

tion at another level, copyright changes the competitive framework; a more

complex approach is therefore warranted. As such, intellectual property

rights as exclusive rights justify a degree of monopolization of a primarymarket. After all, it is their role to protect intellectual creations embodied in

certain articles. Articles such as books, CDs and databases should be pro-

tected in relation to their commercial exploitation on the primary market, as

copyright is given to the works contained in them to reward the creative

activity of the author. That reward does not come as a lump sum, but as a

chance to market any such articles on their normal market. Things are

different when downstream markets are also affected. The reward for the

positive contribution made through the creation of intellectual property is

situated on the primary market and does not justify its abusive use for other

purposes on a secondary market. Monopolization of a secondary market

must therefore be required to apply the essential facilities doctrine.

But that is not all; an abuse in market terms also involves the fact that

another positive development is stopped. This means that stopping the

emergence of a new product for which there is an unmet demand is also

required in a cumulative way. Otherwise, there is no justifiable need to

curtail an intellectual property right that in general fulfils a positive role.

On the other hand, copyright exists to protect the article in which the copy-

right work is included. That protection is not served and does not include

therefore the blocking of the emergence of a new product or article that is

not in direct competition with the first article, and especially not if there is

no alternative for the producer of the new product due to the dominant

position that is occupied by the copyright-holder (Stothers, 2001, p. 93).

By adopting this stricter approach in applying competition law and the

essential facilities doctrine, both areas of law can fulfil their proper role and

serve their common purpose even better. One should also insist on evidence

of the blocking of a new product for which there is a demand and on the

monopolization of a secondary market if the President of the Court's views

are to be followed, and the consumer needs to benefit as well before the

essential facilities doctrine is to be put into operation.

The question is therefore by no means who wins: copyright or competi-

tion law. That is a false debate. The relevant question is instead whether or

not we are able to strike the correct competitive balance by combining all

tools and elements. We have suggested in this chapter that that is indeed

possible and the components and requirements of such a test have been set

out above. That test is necessarily different from the standard competition


law case, as the presence of copyright adds another relevant element to the

competition law equation. The test needs to reflect that, and just as any

competition law test, needs to take account of all the competitive circum-

stances on the relevant market.

NOTES

1 . Order of the President of the Court of First Instance of 1 August 200 1 in case T- 1 84/0 1

R and Order of the President of the Court of First Instance of 26 October 2001 in case

T-l 84/01 R both available at www.Curia.eu.int.

2. Order of the President of the Court of Justice of 1 1 April 2002 in case C-481/01 P(R),

available at www.Curia.eu.int.

3. Case C-48 1/01 IMS Health vs. NDC Health, pending.4. Commission Decision 2001/165/EC of 3 July 2001 in Case COMP D3/38.044 - NDC

Health/IMS Health: Interim Measures [2002] OJ L59/18.

5. The Commission refers in paragraph 38 of its decision to the judgement of the Court in

Case T-44/90 La Cinq [1992] ECR II- 1, where at paragraph 28 the power of the

Commission to impose interim measures finds its justification on the basis that 'protect-

ive measures may be granted only where the practices of certain undertakings areprima

facie such as to constitute a breach of the Community rules on competition'.

6. Decision, paragraph 55.

7. Decision, paragraph 5 1 .

8. Decision, paragraph 59, referring to Case C-62/86 AKZO Chemie BV vs. Commission

[1991] ECR 1-3359.


10. Ibid.

1 1 . Decision, paragraph 93.


1 3. Decision, paragraphs 83, 86, 87 and 1 85.

14. Decision, paragraph 86 infine.

1 5. Especially, it has been argued by the competitors of IMS that a change to the brick struc-

ture involves a change in working conditions under German labour law. The additional

costs therefore include the fact that the contract of service of the sales representatives must

be renegotiated using the German system of co-determination. Any such change to the

brick structure used will therefore mean a long and costly procedure, due to the involve-

ment of the workers' council in the renegotiation procedure. Decision, paragraph 115.


17. Decision, paragraphs 102, 103 and 104.

18. The crucial fact in Magill is that the monopoly was a monopoly over information and

that information happened to be the raw material required by a third party. Therefore,

Magill is now not any longer the only case to be remembered for its unusual facts; that

is, the unusual protection of copyright of factual information.

19. Case C-7/97 Oscar Bronner GmbH & Co KG vs. Mediaprint Zeitungs- und Zeitschriften-

verlag GmbH & Co KG [1998] ECR 1-7791, [1999] 4 CMLR 112.

20. Ibid.

21. Case T-69/89 Radio Telefis Eireann vs. Commission (Magill TV Guide Ltd intervening)

[1991] 4 CMLR 586; case T-70/89 British Broadcasting Corporation andBBC Enterprises

Ltd vs. Commission (Magill TV Guide Ltd intervening) [1991] 4 CMLR 669; and case T-

76/89 Independent Television Publications Ltd vs. Commission (Magill TV Guide Ltd

intervening) [1991] ECR 11-575, [1991] 4 CMLR 745.

22. Case T-504/93 Tierce Ladbroke SA vs. Commission (Societe d'Encouragement et des

Steeple-Chases de France intervening) [1997] ECR 11-923, [1997] 5 CMLR 309.


23. Magill TV GuidellTP, BBC and RTE decision [1989] OJ L78/43.

24. See above, note 21.

25 . Joined cases C-24 1 /9 1 P and C-242/9 1 P Radio Telefis Eireann and Independent Television

Publications Ltd vs. Commission [1995] ECR 1-743, [1995] 4 CMLR 718.

26. IC1 and Commercial Solvents vs. Commission [1974] ECR 223.

27. See also Philips Electronics NV vs. Ingman Ltd and the Video Duplicating Company Ltd

[1999] FSR 112, where Laddie J. applied the Magill approach in a UK case.

28. Joined cases C-241/91 P and C-242/91 P Radio Telefis Eireann and Independent Television

Publications Ltd vs. Commission [1995] ECR 1-743, [1995] 4 CMLR 718, at paragraphs46, 49, 50 and 54.

29. Ibid.


31. See above, note 22.


33. Order of the President of the Court of First Instance of 10 August 2001 in case T- 184/01

R, Order of the President of the Court of First Instance of 26 October 2001 in case T-

1 84/01 R and Order of the President of the Court of Justice of 1 1 April 2002 in case C-

481/01 P(R), all available at www.Curia.eu.int.

34. See the discussion in paragraphs 88-105 and the conclusion in paragraph 106 that there

is 'a very serious dispute' concerning these in the Order of the President of the Court of

First Instance of 26 October 2001 in case T-l 84/01 R. This Order was upheld by the

Order of the President of the Court of Justice of 1 1 April 2002 in case C-481/01 P(R).35. Paragraph 100 of the Order of the President of the Court of First Instance of 26 October

2001 in case T-l 84/01 R and paragraph 17 of the Order of the President of the Court of

Justice of 1 1 April 2002 in case C-481/01 P(R).36. Paragraph 180 of Commission Decision 2001/1 65/EC.

37. Paragraphs 85, 86 and 102 of the Order of the President of the Court of First Instance

of 26 October 2001 in case T-l 84/01 R.

38. Paragraph 79 of the Order of the President of the Court of First Instance of 26 October

2001 in case T-l 84/01 R.

39. Paragraph 8 1 of the Order of the President of the Court of First Instance of 26 October

2001 in case T-l 84/01 R.

40. Paragraph 1 7 of the Order of the President of the Court of Justice of 1 1 April 2002 in

case C-481/01 P(R) with reference to paragraph 105 of the Order of the President of the

Court of First Instance of 26 October 2001 in case T-l 84/01 R.

4 1 . Paragraph 1 45 of the Order of the President of the Court of First Instance of 26 October2001 in case T-l 84/01 R.

REFERENCES

Areeda, P. (1989), 'Essential facilities: An epithet in need of limiting principles',

Antitrust Law Journal, 58, 841-53.

Cooler, R. and T. Ulen, (1988), Law and Economics, 1st edn, New York: HarperCollins (3rd edn, 2000), Harlow and Reading, MA: Addison-Wesley.

Fitzgerald, D. (1998), 'Magill revisited: Tierce Ladbroke SA v The Commission',

European Intellectual Property Review, 20 (4), 154-61.

Hull, D., J. Atwood and J. Perrine (2002), 'Intellectual property- Compulsory

Licensing', European Antitrust Review, 36-9 (A Global Competition Review

Special Report).

Landes, W. and R. Posner (1989), 'An economic analysis of copyright law', Journal



Lehmann, M. (1989), 'Property and intellectual property-

property rights as

restrictions on competition in furtherance of competition', International Review

of Industrial Property and Copyright, 20 (1), 1-15.

Lemarchand, S., O. Freget and F. Sardain (2003), 'Biens informationnels: entre

droits intellectuels et droit de la concurrence', Proprietes Intellectuelles, 6, 1 1-23.

Maskus, K. (2000), Intellectual Property Rights in the Global Economy,

Washington, DC: Institute for International Economics.

Neumann, M. (2001), Competition Policy. History, Theory and Practice,

Cheltenham, UK and Northampton MA, USA: Edward Elgar.

Nozick, R. (1974), Anarchy, State and Utopia, Oxford: Basil Blackwell.

Ramello, G. (2003), 'Copyright and antitrust issues', in Wendy Gordon and

Richard Watt (eds), The Economics of Copyright: Developments in Research and

Analysis, Cheltenham, UK and Northampton, MA, USA: Edward Elgar,

pp. 118-47.

Spector, H. (1989), An outline of a theory justifying intellectual and industrial

property rights', European Intellectual Property Review, 11 (8), 270-3.

Stothers, Ch. (2001), 'Refusal to supply as abuse of a dominant position: Essential

facilities in the European Union', European Competition Law Review, 22 (7),

256-62.

Torremans, P. (2001), Holyoak and Torremans Intellectual Property Law, 3rd edn,

London: Butterworth.

Treacy, P. (1998), 'Essential facilities - is the tide turning', European CompetitionLaw Review, 19(8), 501-5.

Watt, R. (2000), Copyright and Economic Theory: Friends or Foes!, Cheltenham,

UK and Northampton, MA, USA: Edward Elgar.

Woolridge, F. (1999), The essential facilities doctrine and Magill II: The decision

of the ECJ in Oscar Bronner', Intellectual Property Quarterly, 2, 256-64.

9. The basics matter: at the peripheryof intellectual property

F. Scott Kieff and Troy A. Paredes

9.1 INTRODUCTION

Controversies often arise at the interfaces where intellectual property ('IP')

law meets other topics in law and economics, such as property law, contract

law, and antitrust law. }

Participants in the debates over how to mediate these

interfaces often view each interface as a special case deserving unique treat-

ment under the law. 2 The doctrines of copyright and patent misuse are cases

in point: they graft select antitrust principles onto copyright or patent law,

even though there is an entirely distinct body of law - antitrust law -

designed to deal with the putative concerns about competition that

allegedly give rise to misuse. We argue that such specialized approaches to

IP are built by selectively exalting and ignoring particular aspects of the

positive and normative frameworks from distinct substantive areas of law -

IP law, antitrust law, property law, and contract law. Overlooking the total-

ity of these frameworks frustrates the nuanced equilibria to which they each

have evolved, as well as the full complement of important dynamic forces

that affects each legal framework as it continues to develop.3

Instead, we argue that the better approach focuses on the 'basics' - or

core principles and features - of each distinct area of law.4 Our approachavoids specialized frameworks for analysing IP law and the interfaces it

shares with other bodies of law. To do so, the 'basics approach' has both a

procedural and a substantive component.On procedure, our approach emphasizes that the analysis in any particu-

lar case should apply carefully whatever legal regimes the issue at hand

implicates- IP law, antitrust law, contract law, etc. In short, courts should

not create new doctrines and approaches unique to IP when other bodies

of law already apply. At a minimum, courts should take better care to con-

sider accurately and fully the framework of the existing positive law

regimes, as well as the history of the normative debates leading to their evo-

lution. Courts, then, should expressly identify the perceived failures of

these regimes, if any, before creating new doctrines and approaches to

759


resolve the putative failures. The mere formality of identifying and rigor-

ously considering the different bodies of law that apply to some IP-related

matter can result in greater respect for the dignity of such bodies of law and

ultimately can bring important discipline and restraint to judicial decision

making.On substance, our approach emphasizes judicial adherence to the full

range of established positive and normative frameworks within each bodyof law that the court applies. Consequently, as we understand the basics of

the various bodies of law we consider in this chapter, our proposed

approach will in some cases yield different substantive outcomes than if

judges take what we see as a more activist stance toward IP.

When it comes to IP law in particular, it is regrettable that courts and

commentators have demonstrated a surprising willingness to abrogate, if

not ignore, the express language of the statutes Congress has passed in this

area, as well as the reasons for these legislative enactments. Two examples

from patent law are demonstrative: the doctrine of misuse and the doctrine

of non-obviousness. 5 Before the present institutional framework for the

patent system, the 1952 Patent Act, both of these doctrines of patent law

were at best unpredictable, and at worst so predictably anti-patent that no

patent benefited from either, causing the expected value of patents to

plummet.6 The 1952 Act statutorily overruled both of these aspects of

patent law. 7

Yet, on the issue of non-obviousness, over ten years passed after imple-

mentation of the 1952 Act before the Supreme Court in the famous Graham

case instructed lower courts to apply the framework of the Act's new

Section 103 requirement of non-obviousness. 8 And then soon afterwards,

the Court re-injected confusion into the non-obviousness doctrine by con-

triving new requirements for 'synergism' and 'combination' patents.9It then

took until the creation of the Federal Circuit, a full 30 years after the

passage of Section 103 in the 1952 Act, before these 'innovations' in apply-

ing the law of Section 103 were eliminated and the decisional framework of

the 1952 Act was applied consistently according to its own terms. 10

Even more strikingly, on the issue of misuse, almost 30 years passed after

the 1952 Act before the Supreme Court issued an opinion instructing the

lower courts to apply the framework of the new Section 271 provisions

about what does not constitute misuse and what does constitute indirect

infringement.11 And even after this Supreme Court case, Congress acted

again in 1988 to add subsections 4 and 5 to Section 271(d) to be emphat-

ically clear that subsections 1 through 3 were to be applied according to

their terms. 12

Although this chapter emphasizes judicial decision making, the 'basics

approach' provides guidance for policymakers deciding what laws and

The basics matter 1 6 1

regulations to promulgate.13 Similar to courts applying existing legal

regimes, legislators and regulators setting new policy also should carefully

take into account existing positive law regimes and their normative and his-

torical underpinnings before adopting new laws or regulations that mightrun afoul of the substantive basics of existing legal institutional frameworks.

Fundamentally, we argue that the dignity of each separate and distinct

area of law can and should be respected and applied on its own terms to

settle disputes involving IP. Under our approach, the basics matter, to a

very large extent. The basics matter in the sense that they are where the

analysis of any dispute or transaction involving IP should begin. The basics

also matter in that they are where the analysis should in fact, end.

For example, in our 'basics approach,' there is no need to create special

doctrines or approaches to address matters such as price discrimination or

restrictive licensing arrangements involving IP.14

Rather, analyzing the

legality of such arrangements simply requires one to look to the basics of

each applicable substantive law regime: IP law, antitrust law, and what some

people call the 'general law' - property law, contract law, and the like. The'basics approach' gives us a workable - and more predictable

- frameworkof analysis than creating one-off doctrines that are unique to IP at the

periphery of the law of IP where it intersects with other areas of the law.

The 'basics approach' yields rules for resolving disputes that are easier to

apply and that transacting parties can better understand and rely on in

advance than using more specialized approaches tailored for IP - such as

the doctrines of copyright or patent misuse. Misuse doctrines are unpre-dictable in several respects. First, they include various limitations onrestrictive licensing arrangements beyond what antitrust law or contract

law would prohibit. Second, misuse doctrines do not even impose such

additional limitations in a predictable fashion because the decisional

frameworks themselves for misuse are unpredictable-except, of course, to

the extent the doctrines become so firmly entrenched as essentially to evis-

cerate entire areas of IP law. 15By reducing legal uncertainty, the 'basics

approach' facilitates the ex ante coordination necessary to promote innov-

ation through the commercialization of the inventions, symbols, and cre-

ative works that are protected by patents, copyrights, and trademarks - the

primary goal of IP law and an important goal of antitrust law.

We proceed in section 9.2 to discuss the broad framework of the 'basics

approach,' using the topic of price discrimination as a representative

example. Section 9.3 reviews the basics of the core substantive areas of law

that IP typically implicates: IP law itself, as well as antitrust law andso-called general law, which includes property law and contract law. Sec-

tion 9.4 shows how to solve various problems at the periphery of IP law

by employing the 'basics approach,' as opposed to an approach, such as


copyright or patent misuse, that selectively emphasizes or alternatively

ignores particular features of various legal disciplines in crafting special-

ized doctrines for IP. By focusing on the basics, our approach suggests an

important way to reconceptualize IP law with important implications for

bringing new ideas to market. Section 9.5 concludes.

9.2 THE THEORETICAL FRAMEWORK

IP rights generally operate as rights of exclusion. 16 As a result, many worry

that their enforcement will result in too little use of whatever they cover.

Further, the subject matter IP rights cover generally is understood to show

prototypical attributes of public goods in that it is non-rival and non-

exclusive. Classic work by Demsetz, however, has shown that private pro-

ducers can produce and sell an efficient level of public goods under

appropriate conditions, and that price discrimination can advance a com-

petitive equilibrium outcome for public goods, resulting in little, if any,

deadweight loss.17 When an owner of IP rights is permitted to price dis-

criminate, the owner may adopt a pricing regime and licensing scheme that

increases output, eating into any deadweight loss otherwise associated with

market power and the underproduction of public goods.18

Yet, an IP owner's use of price discrimination may not always lead to this

welfare-enhancing outcome. Recent works by Gordon, Lunney, and

Meurer have shown that while price discrimination by IP owners might lead

in theory to more use in certain instances, in practice some price discrim-

ination strategies-

depending, in part, on the licensing arrangements

employed to discriminate among users - can result in less output than if

such price discrimination were prohibited.19 Put simply, price discrimin-

ation has its own shortcomings, and sometimes results in less, not more, use.

The indeterminate results of price discrimination caused us to think

more critically, and more broadly, about IP and price discrimination and

ultimately about the interface IP law shares with other disciplines, such as

antitrust law and the general law, including property law and contract law.

More specifically, there are different types and strategies of price discrim-

ination with different potential consequences- both positive and negative

from the perspective of social welfare. Price discrimination can be done by

the explicit use of different stated prices, in the extreme case by charging

each user his or her reservation price. Price discrimination can also be

done through more complex licensing arrangements, like tying, which can

allow each user more specifically to reveal his or her own demand for

the tying good by how much of the tied good they use. The 'basics

approach' is particularly useful for analyzing the legality of each form of

The basics matter 163

price discrimination since each implicates aspects of IP law, antitrust law,

contract law, and property law.

As discussed more fully below, approaching IP from the basics of IP law,

antitrust law, and the general law of property and contracts enables trans-

acting parties to know better ex ante how to structure transactions that will

be enforced later. In addition to reducing legal uncertainty, when a court

disciplines itself to using an analysis that applies each body of law on its

own terms, there is less opportunity for courts to fashion new and uniquedoctrines of IP law that undercut private ordering by effectively rewritingex post the parties' contract, let alone the legislature's actions. 20 Courts are

ill-equipped to second-guess the substance of contracts entered into by

sophisticated parties merely because the courts believe that some different

arrangement would promote better the use of the underlying IP rights.

Such judicial meddling is particularly troublesome when its downstreamincentive effects on parties

-including owners of IP rights and financiers,

such as venture capitalists- are taken into account. Not only do special-

ized doctrines such as misuse and preemption create uncertainty, but moretimes than not they have the effect of eroding the legislatively created prop-

erty rule protection for IP rights, further compromising commercialization

and private ordering by restricting an IP holder's rights both to use his IP

and to exclude others from having access to the subject matter it covers. 21

There are at least three additional advantages to the 'basics approach,'besides facilitating private ordering and predictability. First, each substan-

tive area of law provides a more informed forum for debate of the issues

that arise in that field. Courts, for example, should not reach out to 'solve'

perceived shortcomings in antitrust law or contract law through the law of

IP, which itself has specific statutory components passed to overturn

similar court action in the past.22

Second, as the product of a long historyof adjudication, lawmaking and academic debate, each area of law pre-

sumably reflects a relatively efficient framework and set of principles that is

actually workable, having stood the test of time. Such longstanding bodies

of law are in contrast to special approaches thatjudges certainly can employto deal with IP, but that are untested and that might simply reflect a particu-lar normative viewpoint that is not satisfied when more appropriate legal

regimes are applied. Third, good cases can be made for each legal regime to

continue to evolve, and they certainly will. However, the one-off, sui generis,

or specialized approaches courts have used at the interface where IP law

meets these other regimes have the effect of skirting many of the diverse

views present in the vibrant debates that persist over how each such regimeshould develop. Put differently, these special judicial approaches to IP

subvert the open and constructive debate that exists within each body of law

regarding whether and how it should evolve further.


Given our view of the present substantive basics of antitrust law, prop-

erty law, and contract law, the 'basics approach' shows greater respect for

private ordering than the IP-specific approaches that we question.

However, to be clear, we would urge courts to follow the 'basics approach'

so as to apply each applicable body of law on its own terms when consid-

ering matters involving IP, even if each such body of law was more restrict-

ive presently or in the future than we understand it or might prefer it to be.

That is, the virtues of the procedural component of the 'basics approach'

are independent of our or any other particular interpretation of the sub-

stantive basics of each body of law that courts would apply.23

Courts that adopt special approaches to address matters at the periphery

of IP law run the risk of crafting judicial doctrines that inappropriately

override well-established bodies of law that are informed by longstanding

judicial and scholarly thought and consideration of each area. Put simply,

when considering disputes and transactions at the periphery of IP law

where it intersects other bodies of law, courts often take select principles

from each body of law out of their larger context and legal framework,

while ignoring other basic features and principles of relevant legal regimes.

For example, the misuse doctrine overlooks a number of considerations

involving vertical restraints of trade that drive the present conclusion under

antitrust law that few vertical restraints are anticompetitive and that manyare in fact procompetitive. Such selective picking and choosing not only

creates uncertainty, but, as suggested, often gets it wrong. In part, the

'basics approach' reflects humility toward the complexity and values

embodied in each area of law.

The kind of respect for private ordering associated with the 'basics

approach,' together with the corresponding benefit of greater predictabil-

ity, promotes the commercialization of IP and the subject matter it

protects. Our approach is in contrast to the approaches offered elsewhere

by academics such as Baxter, Bowman and Kaplow, who each offer

analytical tools that only can be applied ex post to evaluate the validity

of any particular licensing arrangement, and as a result have limited

utility ex ante for parties seeking to structure their affairs in a mutually

beneficial way.24

9.3 THE BASICS

Antitrust law, IP law and the general law of property and contracts are each

well-established disciplines and bodies of law. To be sure, numerous debates

exist within each field, and the law continues to develop. But we believe that

general consensus can be found on the broad positive legal frameworks of


each field and the core principles that undergird them. Although further

development within each discipline may be advantageous, it will be realized

best if reached through a debate that is fully informed of all diverse views

by occurring within the context of that field.

Approaching from the basics embraces the established frameworks and

principles, even as they may evolve in the future, and affords each area of

the law equal dignity. The 'basics approach' applies each area of the law

according to its own terms, and leaves the debates within each legal field to

be had and resolved within such field. In other words, questions about

restraints of trade are left to the field of antitrust law, and questions of con-

tract validity are left to the field of contract law. More to the point, focus-

ing on the basics avoids the fashioning of new doctrines within IP law that

skirt the basics of IP law, antitrust law, or contract law, such as happenswhen some licensing arrangement that does not violate the antitrust laws

or that is otherwise a valid contract is held invalid as a matter of some form

of sui generis IP law like misuse.

To help frame the 'basics approach,' the following discussion highlights,

at a general level, what we understand to be the core of each discipline-

antitrust law, IP law, property law, and contract law. The discussion is

designed to be a summary, by nature; and so does not attempt to review

fully the entirety of each discipline, which in each case fills volumes.

Nevertheless, this summary discussion does endeavor to represent fairly the

basic principles and positive framework of each body of law.

Antitrust Law

Antitrust law is designed to root out unreasonable restraints of trade and

transactions that substantially lessen competition or tend to create monop-oly.

25 But it is well established that antitrust law does not prohibit market

power as such. Nor does antitrust law prohibit a monopoly, if it is achieved

by having lawfully outcompeted other competitors. As Judge Learned

Hand famously put it, The successful competitor, having been urged to

compete, must not be turned upon when he wins.'26 And increasingly,

antitrust law takes account of dynamic efficiency, as well as allocative

efficiency. Even specific types of conduct that are often associated with

restraining trade and that partly drove the passage of the federal antitrust

laws - such as price discrimination, tying, and exclusive dealing- are not

prohibited in every instance. Rather, such conduct generally is prohibited

only to the extent it unreasonably restrains trade. Indeed, many such prac-

tices are procompetitive. The usual test for unreasonableness in this context

is highly fact-dependent and generally is based on a 'rule of reason' analy-

sis as opposed to treating such conduct as an antitrust violation per se.27


Furthermore, antitrust law generally allows unilateral refusals to deal. 28 AsJustice Holmes and then-attorney Rich, who later was the chief architect

of the Patent Act of 1952 and a Federal Circuit judge, also pointed out, it

makes no sense to tell a property owner that she can absolutely exclude

others on the one hand but that she cannot on the other hand be more gen-erous and allow limited access to her property, without giving away the

entire store.29 Accordingly, restrictive licensing arrangements also generally

are permitted.30 To use a simple analogy, as a homeowner, I have the right

to exclude you entirely from my house or to sell you my house, lease you a

room for a limited period of time, or grant you a limited easement across

my front yard.31 Even though refusals to deal and restrictive licenses might

technically restrain trade, they generally do not do so unreasonably and

may by procompetitive.

Intellectual Property Law

Intellectual property law is designed to and indeed does facilitate the

downstream commercialization or realization of the protected subject

matter. 32 While intellectual property law does positively reward, and

thereby encourage, invention and innovation,33

it is not adaptable to being

finely tuned to this goal. It is quite difficult to figure out how appropriately

to reward invention and innovation, and it turns out that a great deal of

inventive and innovative activity would not predictably be responsive to

direct rewards. 34 In practice, IP law facilitates commercialization by

forcing parties to negotiate with each other under the threat of suits for

infringement.35

IP law recognizes that limiting the property owner's causes of action to

be against only those who directly infringe, would unduly undermine or

even eviscerate the role of IP rights in important cases. As a result, the doc-

trines of indirect infringement- induced and contributory

- arose to

capture those activities that, at the time conducted, clearly cause essentially

the same economic effect as direct infringement.36 In the patent context, for

example, by requiring the IP owner to prove not only that his IP rights have

been directly infringed by the one induced, but also that the alleged inducer

intended to induce the infringement, the inducement doctrine captures onlythose who clearly intend to induce infringement and who are successful in

doing so. 37 The contributory infringement doctrine operates similarly. It

requires proof of direct infringement and proof that the alleged contribu-

tor knew that the allegedly contributing conduct was 'especially made or

especially adapted for use in an infringement,' although broad safe harbor

is given to those who provide something that is 'a staple article or com-

modity of commerce suitable for substantial non-infringing use.' 38 Indirect


infringement is not accidental. If so desired, it can be avoided relatively

easily through ex ante consideration of known patent rights in view of these

basic legal rules, which are expressly provided by statute and thereby know-

able relatively easily.39

The ability of an IP owner to elect to sue or license those who would

otherwise be guilty of direct or indirect infringement facilitates both price

discrimination and coordination among complementary users. For this

reason, the 1952 Patent Act expressly provides- and, indeed, this was a

major impetus for the writing of the Act - that neither efforts to price dis-

criminate nor the granting of a restrictive or unrestrictive license to a

potential infringer shall constitute misuse.40 This provision was ignored by

many courts until the 1980 Dawson decision,41 which finally recognized its

impact. To be certain this was clear, Congress acted again in 1988 by adding

subparts 4 and 5 to Section 271(d) of the Patent Act to provide expressly

that neither a refusal to license nor a tying arrangement in the absence of

market power is patent misuse.42

Importantly, because the doctrines of copyright misuse43 and trademark

misuse44 are based on the doctrines of patent misuse and patent law's indir-

ect infringement, our discussion has focused on patents. The lessons

learned from the 'basics' view of patents are also applicable throughoutIP law.

Under the basics of IP law, it is clear that contracts facilitating price dis-

crimination or imposing restrictions on a licensee are allowed - indeed,

they are contemplated - at least to the extent they are otherwise properly

formed and enforceable under the general law of contracts, as explored

below. IP rights only give IP owners rights of exclusion, not rights to use.45

The uses to which an IP owner can put their IP or the subject matter pro-

tected by it is (or at least should be) determined by other areas of law. IP

law does not limit the rights of an IP owner to use his or her IP or the sub-

ject matter covered by it in any way that otherwise would be permissible

under other areas of law, including antitrust law, property law, and contract

law. To rely on the express statutory rights of exclusion against others that

IP law grants to IP owners as a basis for restricting the IP owner's rights to

use, conflicts with the basics of IP law; an owner of IP should enjoy similar

rights to use as an owner of tangible property enjoys. On the other hand,

the ownership of IP rights does not magically immunize the owner from

whatever limitations on use of IP or the subject matter it covers are imposed

by other areas of law, including antitrust law and the general law of prop-

erty and contracts. For example, an IP owner's exercise of his IP rights

should (and does) remain subject to the antitrust laws, and a restrictive

licensing arrangement should not be enforced if it is not validly entered into

under contract law.46


The General Law: Property and Contracts

Property law and contract law operate to facilitate private ordering, a keyto commercialization of IP assets and to the exploitation of their value.

While property law generally eschews restraints on alienation and, throughits numerus clausus principle, seems to recognize only certain estates in land,

these doctrines only operate as default rules in practice, and a nearly infin-

ite range of dealings can be carried out through contract.47 Moreover, even

within the traditional forms of property, transferors and transferees have a

great deal of flexibility to carve up interests in property along the dimen-

sions of time, use, and the number of property owners.48 For example, whenit comes to real property, highly particularized defeasible fees can be

created and will be enforced, and a real property owner can create anynumber of leasehold interests in his property. All of these transactions are,

of course, facilitated by a general regime of property rule protection, as

opposed to liability rule protection, for rights in both real and personal

property.49

To be sure, when parties order their affairs through contract, this requires

a valid contract, which in turn requires compliance with proper formation

details, such as consideration and no unconscionability. With very few

exceptions, positive contract law does not regulate the substance of the

parties' arrangement, focusing instead on the contracting process.50

Whatever strictures property law and contract law impose on private

ordering, parties are generally free to order their affairs and to carve uprights, duties and obligations as they see fit. These basics of the general law

of property and contracts should extend to the use and licensing of IP

rights, just like they do to other types of property. Nothing under propertyor contract law provides any particular reason to be skeptical about IP con-

tracts that facilitate price discrimination, exclude certain parties from

having access to IP rights, or impose restrictions on licensees. What is more,

special approaches to disputes and transactions involving IP rights often

ignore or intentionally override purposeful normative and positive features

of antitrust law, IP law, or the general law and, in so doing, risk upsetting

well-developed frameworks without adequately accounting for competingconsiderations.

9.4 WHEN APPLIED, THE BASICS SOLVETHE PROBLEM

Applying the 'basics approach' to prototypical cases at the periphery of IP

law, including price discrimination, restrictive licensing arrangements, and

The basics matter 1 69

suits against indirect infringers, provides a set of rules that are useable

ex ante by all market participants in a way that helps them order their

affairs while at the same time being more fair and efficient. The 'basics

approach' has important normative implications, in that judicial fidelity to

the basics ultimately allows market actors to have greater freedom and

ability in structuring their interactions in welfare-enhancing ways, in addi-

tion to reducing legal uncertainty.

The cases we explore are appropriately viewed as prototypical for several

reasons. They involve fact patterns that are representative. They have actual

historical significance through their particularly important roles in the

body of case law. And the primary architect of the present patent system-

the 1952 Patent Act - wrote a five-part series of articles about these cases

before drafting the statute designed to change fundamentally the waycourts applying the law would look at the issues raised by the cases. 51

As discussed more fully below, the cases can be divided into two sets.

Importantly, a review of both sets of cases shows that the 'basics approach'is not merely a veiled effort to promote pro-patent or pro-copyright

- or

more generally, pro-business-

positions. Rather, the basics framework, as

a method of judicial decision making, is offered as a coherent approachthat more predictably can be engaged ex ante and that reflects fidelity to,

and respect for, separate areas of the law. Although we focus on patents,

since the core features of other areas of IP law largely derive from patent

law, the basics framework and the essence of the following analysis extend

to copyrights and other forms of IP as well.

Indirect Infringement vs. Breach of Contract

The first set of cases involves the tension between indirect infringement and

indirect participation in a breach of contract. Indirect infringement may be

actionable as a matter of IP law, as discussed earlier. Indirect participation

in a breach of contract may be actionable as a matter of contract law under

doctrines such as tortious interference with contract, as in the famous

multi-billion dollar judgment from the Texaco, Inc. vs. Pennzoil, Co. litiga-

tion. 52However, the happenstance that a contract relates to patents should

not transform interference with that contract into patent infringement. The

facts that need to be proven are different under these different frameworks.

The potential remedies are different as well. 53

Wallace, the classic case of indirect infringement, involved a patent on

an oil lamp having a new burner, together with a standard fuel reservoir,

wick, and chimney.54 In the case, a competitor of the patent owner had sold

a rival product, which included the new burner and other lamp parts but

not the chimney. The court reasoned that the defendant had contributed to


infringement on the part of its customers, because they would inevitably

add a chimney. A judgment of contributory infringement makes sense

under the 'basics approach' because the intended and actual impact of the

competitor's efforts were to make sure that its customers acted in an

infringing manner. Indeed, Wallace is the case that gave rise to the entire

doctrine of indirect infringement throughout all of IP law. 55

By way of comparison, if the plaintiff-patentee in Wallace instead had

entered into arrangements with its customers obligating them to buy chim-

neys from the patentee, the analysis under the basics of IP law would be

different. A rival seller of chimneys might be liable for tortious interference

with contract, or the tying arrangement might violate the antitrust laws.

However, the competing chimney seller would not be liable for contributory

infringement under the basics of IP law. 56

The Heaton case involves an example of just this type of arrangement.

Heaton involved a patentee who sold a patented machine with a label

license57 under the patent that restricted the machine's use to certain

unpatented inputs (staples, literally)- a tying arrangement

- and a defend-

ant who sold competing inputs.58 The court seemed to reason that by pro-

viding its staples for use in the machine, the defendant was contributing to

breach of the label contract, which had given the permission through the

label patent license to use the machine. Once that license under the patent

was gone, the use of the machine became infringing. Rather than sue for

interference with the contract, the plaintiff sued the competing seller of

staples for indirect infringement of the patent under patent law. The court

decided that the defendant was, indeed, committing contributory infringe-

ment of the patent. But this turned a case about indirect participation in

breach of contract into patent infringement. By deciding the case the wayit did under IP law, the court in effect extended inappropriately the scope

of IP rights. A collateral inappropriate consequence of the court's reason-

ing in Heaton, of course, would be to immunize potentially anticompetitive

licensing arrangements from the antitrust laws.

The 'basics approach' rejects the analysis of Heaton. Under the 'basics

approach,' and as pointed out by Rich, this decision was inappropriate

because it 'transformed the law of contracts into "patent law".' 59It may

have been appropriate for the plaintiff to consider an interference with con-

tract argument, if sufficient facts could be proven to substantiate the claim

under contract law. 60 It may even have been appropriate for the defendant

to consider an antitrust tying argument, if the case could be proven under

antitrust law.61

By not addressing these contract and antitrust arguments head on, cases

like Heaton allow parties, and judges, selectively to mix features of various

bodies of law and to extrapolate from them to forge new hybrid doctrines


of law that run afoul of the basics of each area. In many instances, such

selective application of the law leads to doctrines, such as misuse, that erode

IP rights. In other cases, such as Heaton or those cases in which courts have

subjected transactions involving IP to less scrutiny under antitrust law, the

new doctrines can work to expand IP rights. What is more, in all cases, the

courts fail to give any meaningful test for determining when those IP rights

should be so eroded or expanded.The Supreme Court applied the same approach as Heaton in the A.B.

Dick case, which involved a patent on a mimeograph machine sold with a

label restriction limiting the brand of unpatented ink that could be used in

the machine. 62 As in Heaton, the Court agreed with the plaintiff-patentee

in A.B. Dick, and held that there was contributory infringement of the

patent. Because this was a Supreme Court case, its reasoning had a longer

lasting impact in pushing IP law in a direction contrary to the 'basics

approach.'63

The 'basics approach' rejects the reasoning of A.B. Dick for the same

reason it rejects the reasoning of Heaton. Indeed, eventually, these cases

were effectively overturned.64 As Rich pointed out later in his testimony

before Congress concerning the provisions he drafted on indirect infringe-

ment in the 1952 Patent Act, any effort to follow this inappropriate body of

law 'would kill itself in time.'65

An understanding of the basics suggests why Heaton, A.B. Dick, and

their progeny were not sustainable over the long run. The problem is not

merely one of courts going too far one way (e.g., effectively extending the

scope of IP rights to anything connected to IP and simultaneously immun-

izing all transactions involving IP from serious antitrust scrutiny) or the

other (e.g., eliminating the doctrine of indirect infringement, thereby

eroding IP rights). The problem is more fundamental. Namely, cases like

Heaton and A.B. Dick ignore the basics of each implicated body of law -

IP law, antitrust law, and the general law of property and contracts. As a

result, they lead to unpredictable results and, in the name of IP law,

encroach upon the boundaries of other well-established bodies of law that

reflect more nuanced and time-tested doctrines and rules that have staying

power and that are perfectly capable of resolving the disputes on their ownterms.

Infringement under IP Law vs. Sui Generis Law

The second set of prototypical cases involves the question of what body of

law should govern determinations of infringement: the body of organic

IP law - patent, copyright, or trademark - or some special sui generis bodyof law. In many of the cases involving charges of indirect infringement


and misuse - which are admittedly somewhat difficult doctrines - too manycourts and commentators have not followed the 'basics approach' and have

instead tried to re-hash the normative case for IP to develop new special-

ized approaches in these doctrinally difficult cases that they hope will get

IP scope just right. The fundamental problem with these specialized

approaches is that they re-cast the entire legal institutional framework for

IP in a way that has pernicious ripple effects throughout IP law by ignor-

ing the many choices that have been made over IP law's development.One basic trap into which these courts and commentators have fallen

when adopting such sui generis approaches to IP is focusing on the wrongparty when considering whose behavior should matter in cases of possible

indirect infringement. The behavior of the putative indirect infringer to

facilitate or encourage direct infringement is relevant to the analysis under

both inducement of infringement and contributory infringement. Thebehavior of the patentee

- in the sense of the patentee engaging in conduct

that leverages his IP rights with the goal of extracting value - is not rele-

vant to inducement or contributory infringement. Indeed, in such

instances, the patentee is simply exercising his rights to exclude and to use,

as the basics of IP law and the general law anticipate. Put simply, the ques-tion of a putative defendant's infringement should not turn on whether or

not the patentee was trying to get as much out of the patent as possible

through some restrictive licensing arrangement, tie-in, or otherwise. If the

patentee, or any property owner for that matter, behaves in a way that

antitrust law or contract law properly prohibit, then that is a matter of

antitrust law or contract law.

The modern trend towards sui generis analysis of infringement- as com-

pared with an analysis based in the basics of IP law - has its most visible

roots in the Supreme Court's Leeds & Catlin decision. Just like the classic

indirect infringement case of Wallace, discussed previously, Leeds & Catlin

involved a patentee's competitor selling something that had no substantial

non-infringing use. 66 In Leeds & Catlin, the defendant-infringer sold spe-

cially grooved records that could only be used in a patented record player

known as a 'Victrola.' The Supreme Court reasoned that the defendant's

selling of the records was infringement because the records were the 'dis-

tinction [or key element] of the invention.'67 This reasoning is flawed.

Although the 'basics approach' might reach the same result - a finding of

contributory infringement- it would do so for an entirely different reason

than offered by the Court. Under the basics of patent law, there is no 'dis-

tinction,' or key element, of subject matter claimed under the patent. The

patent system operates using what is known as 'peripheral claiming'- as

distinct from 'central claiming'- in which the function of the patent claim

is not to set forth the heart of the protected subject matter but rather to set


forth its outer bounds. 68 Direct infringement is measured against these

outer bounds. Indirect infringement is premised upon some occurrence of

direct infringement. But the reach of indirect infringement does not turn

on whether the putative defendant is targeting some key element of the

claim. Rather, as discussed earlier, for a proper analysis of contributory

infringement under the basics, a key question is instead whether there were

any substantial non-infringing uses for the grooved records. Because there

were no such uses in Leeds & Catlin, and because the other elements of

contributory infringement were established (i.e., direct infringement and

knowledge of the patent), applying the basics would have resulted in a

finding of contributory infringement.

Hanging determinations of indirect infringement on the factors outlined

earlier in our discussion of the basics - such as intent for induced infringe-

ment and absence of non-infringing substitutes for contributory infringe-

ment - may seem like an effort to exalt form over substance. After all, the

reasoning the Court adopted in Leeds & Catlin seems to strike at the heart

of substance by focusing on the key element. But the Court fails to give anyinstruction on how to determine which element is key, and neither has anyother court or commentator of which we are aware. The tests for indirect

infringement have the essential advantage of being comparatively easy to

administer. They look to facts well within the control of the putative

infringer and are strongly biased in favor of the putative infringer in the

types of errors one would expect the tests to generate.

The intent requirement under an inducement analysis and the broad and

readily identifiable safe harbors under a contributory analysis ensure these

important biases and that the doctrines are relatively easy to administer. 69

Importantly, the improper reasoning of the Court in Leeds & Catlin is not

mere harmless error. It matters which approach is adopted by courts - espe-

cially the Supreme Court, even if the results are the same in a particular

case. By suggesting in Leeds & Catlin that the case turned on the heart of

the invention, the Court advanced a line of precedent that focused on the

wrong issues in patent cases. One of the most pernicious cases in this line

of precedent was Carbice, in which the Court denied relief to a patentee

after reasoning that the patentee was trying to extend the patent beyond the

key elements of the claim. 70 The plaintiff-patentee in Carbice had a patent

on a packaging method that used dry ice. What troubled the Court was that

the patentee had a practice of entering into licensing arrangements oblig-

ating the licensee to use only certain containers for packaging productswith the dry ice. The facts of Carbice are somewhat similar to those of

Leeds & Catlin with one important difference: the defendant sold a product-dry ice - that was a staple article of commerce usable in many non-infring-

ing manners other than in the patented method of ice-cream packaging.


The 'basics approach' would again yield the same result as the Court's

analysis- in this case, no contributory infringement

- but again for a

different reason. Instead of focusing on the patentee's alleged extension of

the patent beyond its key elements, the 'basics approach' would turn on the

many non-infringing uses for dry ice.

As Rich emphasized, it is the behavior of the putative contributory

infringer that is relevant to a determination of contributory infringement,

not that of the patentee.71 Under the 'basics approach,' it makes sense that

the organic IP law- in this case patent law

- evolved so that as implementedin the 1952 Act's provision of Section 271, focus is on the behavior of the

putative infringer precisely because it is comparatively easy to judge.

Furthermore, an IP holder should not be denied relief for contributory

infringement- or even direct infringement

-simply because the IP holder

is exercising his rights to exclude and to use through a tying arrangementor restrictive license. Such conduct is properly a subject for antitrust law

and contract law, but should have no bearing on a court's analysis of indir-

ect (or direct) infringement under patent law. Courts should not recast such

conduct as an effort by the IP holder to 'extend' his patent rights for the

purpose of transforming a matter for antitrust and contract law into a

matter for some new version of IP law.

To be sure, the Court did not always reach the right result, as it did in

Leeds & Catlin and Carbice. Because the Court continued to misplace its

focus on the putatively key elements of patent claims, by the time of the

Mercoid cases, the doctrine of indirect infringement had been almost entirely

eliminated as a result of judicial reasoning that precluded any action for

indirect infringement. In essence, because by its nature every indirect

infringement case involves a defendant who is not triggering at least one

element of the patent claim - direct infringement occurs when all elements

are satisfied - the focus on 'key element' in the Court's reasoning allowed

every putative indirect infringer to argue that the missing element was the one

that was 'key' and therefore no action for indirect infringement could lie.72

In response, Rich drafted what became Section 271 of the 1952 Patent

Act to overrule statutorily cases like Mercoidand to revive indirect infringe-

ment. 73 Under this established basic framework of patent law after the 1952

Act, the essential inquiry for indirect infringement is on the comparatively

easy to administer framework discussed earlier. 74 While it may be appro-

priate to debate the benefits and costs of allowing actions for indirect

infringement, the above review is designed to show at least two important

things. First, sui generis attempts to re-hash the proper scope of an organic

IP right when addressing cases of misuse or indirect infringement will yield

a test that is comparatively more difficult to administer, that eliminates the

doctrine, or both. Second, unlike prior approaches commentators have


offered for addressing issues at the periphery of IP law - many of which

urge a nearly impossible ex post balancing of dynamic and allocative

efficiency that inappropriately emphasizes trying to achieve some optimal

reward to inventors as opposed to commercialization - the 'basics

approach' provides a set of clearer rules and doctrines that market partici-

pants can rely better on ex ante in structuring their affairs.

Other Pernicious Ripple Effects

The 'basics approach' has important implications for resolving matters

involving at least three current and controversial issues found at the periph-

ery of IP law: patent and copyright misuse, restrictive licensing arrange-

ments, and preemption. Applying the basics to these and other tough cases

that simultaneously implicate IP law, antitrust law, and contract law avoids

a host of pernicious ripple effects - namely, undercutting innovation and

the commercialization of IP - that arise from more specialized approachesto disputes and transactions involving IP.

Concerning the misuse doctrine, the 'basics approach' is not compatiblewith the Federal Circuit's present view of patent misuse, which seems to

leave a broad and vaguely defined space for misuse. 75 In Virginia Panel, the

Federal Circuit suggested the following test for determining whether a

patentee has misused his patent: 'When a practice alleged to constitute

patent misuse is neither per se patent misuse nor specifically excluded from

a misuse analysis by Section 271(d) [of the Patent Act], a court must deter-

mine if that practice is reasonably within the patent grant.'76 But import-

antly, the patent statutes make no provision for per se misuse. 77Rather,

Section 271(d) provides specific safe harbors for conduct that is not misuse.

Further, it is inappropriate to suggest that some use of a patent is not within

its scope, since patents only give a right to exclude. The right to use is

derived from sources external to IP law.

When the 'basics approach' is employed, other bodies of law, such as

antitrust law, provide the proper legal lens through which to inspect a paten-

tee's use of a patent and the subject matter it covers. For example, the basic

thrust of putative misuse is that an IP holder should be denied relief for

infringement when he has used his IP in some allegedly anti-competitive

way. Yet, if the challenged conduct is indeed anti-competitive, it ought to

trigger the antitrust laws. As discussed earlier, patentees and copyright

holders, like other property owners, are subject to antitrust law, because

patents and copyrights give only a right to exclude, not a right to be free

from the constraints of other laws. In brief, the pernicious effect of the

misuse doctrine is that it erodes IP rights, at least at the margin, and risks

rooting out procompetitive and competitively-neutral behavior that the


antitrust laws recognize as such and permit.78 If the antitrust laws are too

lax, the appropriate remedy is to fix the antitrust laws. As Rich pointed out

in commenting on the unfortunate habit of courts to treat potential

antitrust concerns as some how more serious and in greater need of polic-

ing when IP is involved:

The patent right is not the only form of property subject to such misuse. But it

is so little understood, as compared to other forms of property, that much

mystery attaches to it and much confusion surrounds it. ... [Practices that

restrain trade are] not due to the patent law .... [They are] due to failure to

enforce the anti-monopoly laws. The advocates of reform would do well to

restrict the attack to the latter aspect and not confuse the issue by abortive

attempts to emasculate the patent law . . . ,

79

Concerning restrictive licensing arrangements, the 'basics approach' sug-

gests that courts generally should enforce restrictive licenses involving IP as

long as they are enforceable under contract law and do not run afoul of the

antitrust laws. Indeed, affording IP holders the right to carve up interests

in their IP and the subject matter it covers is consistent with the basics of

property law and the right to use enjoyed by owners of tangible property.

Courts adopted the 'basics' reasoning in considering the validity of restrict-

ive licenses of copyrights in the ProCDSQ case and of patents in the

Mallinckrodt81 case. 82 Even when a potential or actual IP owner tries to

extract payments for activities that fall outside the protection of IP,83 courts

should enforce these contracts to pay as long as the arrangement- which

may amount to little more than an effort to ease either the risk burden or

the financing burden of the transaction - is properly enforceable under con-

tract law. 84

By way of contrast, in practice, courts are skeptical of contracts that

happen to be tied to royalty payments beyond patent term, even though the

economic justification for this skepticism is lacking.85

Further, courts that

do not stick to the basics will often err by finding that restrictive licensing

arrangements, including tie-ins, constitute some sort of impermissible

extension of IP rights.

Finally, concerning preemption, the 'basics approach' is not compatible

with the approaches adopted by the Supreme Court, which generally can

be 'seen as efforts to place limits on the ability for [IP owners] to avail them-

selves of various State laws.' 86 The Court's approaches make no sense in

part because IP rights confer rights of exclusion on IP owners, not add-

itional restrictions on use on IP owners and not additional rights to use on

non-owners. 87

For example, in Bonito Boats, the Court's approach is premised on the

contrivance that federal patent law creates a right to copy.88 The Court in


Bonito Boats decided that this right to copy would be frustrated by the state

law at issue, which regulated one particular form of copying boat hulls,

called 'plug molding.'89 The case arose because a party seeking to engage

in plug molding in violation of the state statute argued that federal patent

and copyright law preempted the state statute based on the SupremacyClause of the US Constitution. The Court invalidated the state statute

under the doctrine of conflict preemption.90

There are several problems with the Court's reasoning. There is no right

to copy - indeed, no affirmative right at all - that is conveyed on the public

by patent law, or for that matter by copyright or trademark law. 91 These IP

regimes only create under certain situations specific rights of exclusion

vested in the IP holder, as explained above. Although the plug molding

activity was not covered by any of these federal IP rights, the mere absence

or expiration of any such right of exclusion says nothing about a third

party's affirmative right to use the subject matter such right of exclusion

might have covered. Indeed, use of IP rights, whether by the IP holder or

some third-party licensee, is often restricted, if not outright blocked, byother IP rights and by regulation, but this should not be grounds for finding

that such restrictions are preempted by IP law.

The Bonito Boats Court essentially rejected, or at least glossed over, these

arguments by suggesting that the purpose of the state statute somehow con-

flicted with a purported 'strong federal policy favoring free competition in

ideas which do not merit patent protection.'92 But this analytical framework

is unworkable in that it would seem to extend to block any state law or legal

enforcement of contracts that interferes with a right to use or copy.

Consider, for example, a state law against cheating on exams or, to be closer

to the case, a contract term against plug molding. Also consider a contract

term making a promise to do or to abstain from doing any activity in a waythat allegedly conflicts with the IP law regime putatively doing the pre-

empting.93 Under the reasoning of Bonito Boats, each such state law or

contract term would not be enforceable. In addition, the Bonito Boats rea-

soning may eviscerate the rights of exclusion that the patent, copyright, and

trademark statutes are designed to create and that are both properly justi-

fied and authorized.94 That is, every IP right and every form of market reg-

ulation or other exercise of police power will impact, to some extent,

competitive economic concerns of the type that also underlie each of the

federal IP regimes, thereby providing under the Bonito Boats rationale an

extensive basis for preemption and interference with private contracting.

What is more, to anyone informed by public choice theory, every IP right,

market regulation, and other exercise of police power can be seen as motiv-

ated at least in part by its impact on these same competitive economic con-

cerns. Put differently, the reach of the preemption analysis in Bonito Boats


would allow any potential defendant to select a federal IP regime that does

not reach such party-say patent law - and then use that federal IP regime

to justify a finding of preemption against any other state (or even federal)95

law that impacts competitive economic concerns. What is more, if the

federal IP regime that is said to have this preemptive effect is itself weaker -

such as when it does not reach indirect infringement or is limited bymisuse - then the overall power of the Bonito Boats preemption approachis even greater.

In contrast, the 'basics approach' recommends the Federal Circuit's

alternative analytical framework for conflict preemption called the 'extra

element test,' which does not suffer the shortcomings of Bonito Boats and

which, indeed, facilitates the smooth operation of each IP and competitionlaw regime, federal and state alike.96 That test simply asks whether the basic

legal elements of the cause of action that is putatively preempted are

exactly the same as those of the cause of action that is putatively doing the

preempting; the presence of extra elements means no preemption. The state

statute at issue in Bonito Boats would be analyzed easily under this test

because liability under it turns on a host of elements that are unconnected

to patent law, including copying via the plug mold technique. Even closer

cases, such as those involving state laws regulating statements about patents

themselves, can be decided using the extra element test.97

The 'basics approach' also provides guidance for policymakers evaluat-

ing whether to promulgate IP laws that interface with other state law

regimes. For example, in derogation of principles of state contract law and

trusts and estates law, the copyright regime gives authors a non-transferable

right to terminate transfers98 of their copyrights and even sets forth its owntrusts and estates provisions governing who gets this termination right

upon the author's death so as to expressly preempt other arrangementsauthors might make by testamentary will as well as state default rules

of intestacy.99By way of another example, the anti-circumvention provi-

sions of the recently promulgated Digital Millennium Copyright Act

('DMCA'),100 which prohibits manufacture or distribution of any technol-

ogy, product, service, or device that circumvents copy protection technol-

ogy, does not make sense under the substantive component of the 'basics

approach' because the transactions it regulates are better governed byeither state contract law or by the IP law of indirect infringement, discussed

earlier. 101 In short, the 'basics approach' would urge policymakers consid-

ering each such positive law framework instead to leave these issues to be

decided as a matter of state law. That having been said, even if the wisdomof some legislative or regulatory action can be challenged under the 'basics

approach,' once policymakers have acted, courts should apply the law on

its terms and refrain from crafting new doctrines or from engaging in other


creative decision making, which would have the effect of end running the

legislative or regulatory body.

9.5 CONCLUSION

Like the Supreme Court in Dawson and earlier work by commentators such

as Baxter, Bowman, Gordon, Kaplow, Lunney, Meurer, and Rich, we strike

a balance between a view of IP that is too restrictive and one that goes too

far. But following more closely the writing of Rich, who after all drafted

the statutory framework that Congress adopted for patent law, we look not

only to the direct impact that applying the basics at the periphery of IP law

has on the commercialization of the subject matter it protects. We sepa-

rately believe that it is important to respect the different legal institutional

frameworks of the various bodies of law that are involved - IP law, antitrust

law, and the general law - each of which strikes its own balance among the

competing needs of those who interact with these disciplines. We do not

dispute here that it is important for doctrine to develop over time toward

more efficient and equitable outcomes. Such doctrinal developments,

though, should occur within the context of the applicable body of law. IP

law, for example, should not be a vehicle for restructuring contract or

antitrust law collaterally from outside those legal regimes. Although in

practical terms, the 'basics approach' often reaches results that are similar

to the outcomes of other approaches, we provide a normative justification

for a positive law framework that is more predictable and that captures the

distinct and important balances that are struck within each separate bodyof law that is implicated.

We offer, in the end, a framework for understanding IP law and the

broader interfaces that IP law shares with a number of bodies of law, such

as antitrust law, property law, and contract law. Our framework is in large

part animated by a property rights perspective that places priority on

ensuring the appropriate ex ante incentives to facilitate the complex trans-

actions needed to ensure wider use of the subject matter IP rights cover,

such as through information dissemination and commercial sales of

embodiments. Our framework is equally motivated by attention to the

basics of each other body of law we discussed with an understanding that

only through coherent discussion of each area in a piece-wise fashion can

the right progress be made on both positive law and normative fronts. Afurther advantage of the 'basics approach' is that it should reduce legal

uncertainty, which itself is a source of inefficiency.

Finally, the 'basics approach' reflects a general skepticism about the insti-

tutional capability of courts to make ex post determinations regarding how


to facilitate the complex commercialization process that must occur for the

public to derive the benefits of the various works protected by IP rights. Put

differently, we believe that private ordering and markets are more effective

than courts, all things considered, at solving what in essence are industrial

organization matters.

ACKNOWLEDGMENT

The authors gratefully acknowledge financial support from the John M. Olin Foundation, the

Hoover Institution, and the Washington University School of Law, as well as intellectual con-

tributions from participants in the 2003 Society for Economics Research on Copyright Issues

Annual Conference held June 19-20, 2003, in Northampton, Massachusetts. The authors also

gratefully acknowledge more detailed comments provided by Michael Abramowicz, John

Barton, Tim Bresnahan, Richard Epstein, Paul Goldstein, Wendy Gordon, Mark Lemley,

Larry Lessig, Stanley Liebowitz, Glynn Lunney, Charles McManis, Michael Meurer, PamSamuelson, Joel Seligman, Henry Smith, Polk Wagner, and John Witherspoon. A revised

version of this chapter has been published in the George Washington Law Review (2004).

NOTES

1 . See, e.g., Symposium, The Interface Between Intellectual Property Law and Antitrust Law,87 Minn. L. Rev. 1695 (2003); William F. Baxter, Legal Restrictions on Exploitation of the

Patent Monopoly: An Economic Analysis, 76 Yale L.J. 267 (1966); Ward S. Bowman, Jr.,

Patent and Antitrust Law: A Legal and Economic Appraisal, at xii (1973); Michael A.

Carrier, Unraveling the Patent-Antitrust Paradox, 150 U. Pa. L. Rev. 761 (2002); Wendy J.

Gordon, Intellectual Property as Price Discrimination: Implicationsfor Contract, 73 Chi.-

Kent L. Rev. 1367 (1998); Louis Kaplow, The Patent-Antitrust Intersection: AReappraisal, 97 Harv. L. Rev. 1813 (1984); Glynn S. Lunney, Copyright and the SupposedEfficiency of First-Degree Price Discrimination (2002) (working paper); Michael J.

Meurer, Copyright Law and Price Discrimination, 23 Cardozo L. Rev. 55 (2001).2. Indeed, these debates often take on status as their own specialized disciplines bearing

new 'and'-based names, such as 'intellectual property and antitrust,' which in turn spawnnew sub-specialties, such as 'copyright and antitrust.'

3. For earlier articulations of the views at the core of the 'basics approach' developed below,

see, e.g., Donald S. Chisum, Craig Allen Nard, Herbert F. Schwartz, Pauline Newmanand F. Scott Kieff, Principles of Patent Law 1066-1 155 (2001); Troy Paredes, CopyrightMisuse and Tying: Will Courts Stop Misusing Misuse?, 9 High Tech. L.J. 271 (1994).

4. We do not use the word 'basics' pejoratively, such as in the sense of an unduly simplecharacterization of the law or legal process. In addition, we recognize that there is

sufficient path dependency and context dependency to the development of the 'basics'

that our discussion here is most applicable to the regimes that have evolved in the US.That being said, we do think that the comparative institutional analysis offered here maybe useful in elucidating relative strengths and weaknesses of different strategies to

shaping IP and other commercial laws outside the US as well.

5. See generally Chisum et al. supra note 3, at 514-19, 1066-99 (discussing evolution of the

non-obviousness and misuse doctrines).

6. Although there was some variation in the way courts treated patents under these doc-

trines, a sufficient number of powerful courts (including the Supreme Court) were treat-

ing the patents that came before them so poorly under these doctrines that the expected


value for all patents plummeted. On non-obviousness, courts applied a tautological and

unpredictable subjective decisional framework then called the 'requirement for inven-

tion': to be patentable, an invention had to constitute an 'invention.' This standard

became so vague and yet so difficult to satisfy that Justice Jackson remarked, '[T]he only

patent that is valid is one which this Court has not been able to get its hands on' Jurgensenvs. Ostby & Barton Co., 335 U.S. 560, 572 (1949) (Jackson, 1, dissenting). On misuse,

courts applied such a broad definition of misuse that for all intents and purposes patents

could no longer be asserted against indirect infringers. See generally infra notes 40-41

and accompanying text (discussing evolution of the misuse doctrine in relation to the

doctrines of contributory infringement and inducement of infringement).

7. On non-obviousness, Congress passed Section 103 of the statute in the 1952 Act to

replace the requirement for 'invention' with the requirement for 'non-obviousness.' Far

more than a different word, this test for patentability set forth a much clearer and more

objective decisional framework. For a comparative institutional analysis of these deci-

sional frameworks, see F. Scott Kieff, The Casefor Registering Patents and the Law andEconomics of Present Patent-Obtaining Rules, 45 B.C. L. Rev. 55 (2003) (comparativeinstitutional analysis of patent-obtaining rules). On misuse, Congress passed Section

271 of the statute in the 1952 Act to revive the doctrines of contributory infringementand inducement of infringement and to make clear what does not constitute misuse. See

infra notes 40^1 and accompanying text (discussing evolution of the misuse doctrine in

relation to the doctrines of contributory infringement and inducement of infringement).

8. Graham vs. John Deere Co., 383 U.S. 1, 17-18 (1965) (describing new framework).

Although there is some language in the Graham opinion to suggest that the 1952 Act did

not change the law, it is important to note that the opinion ties the statutory objective

standard of non-obviousness to eighteenth century case law that employed a similar

objective standard while specifically rejecting the nineteenth century case law that

employed a subjective standard. See Kieff, supra note 7, at 88-95.

9. See Sakraida vs. Ag Pro, Inc., 425 U.S. 273, 282 (1976) (holding patent invalid because it

was a mere combination of old elements and had no synergistic effect); Andersons-Black

Rock, Inc. vs. Pavement Salvage Co., 396 U.S. 57, 61 (1969) (holding patent invalid

because '[n]o such synergistic result is argued here'). Of course, the problem with treat-

ing so-called 'combination' patents differently is that all patent claims in the present USpatent claiming system can be viewed as combinations of 'old elements.' See Kieff, supranote 7, at 1 1 1 (explaining how a claim operates as a simple logical list of elements andthat infringement is only found when each and every element on that list is present in the

allegedly infringing product or process).

10. See George M. Sirilla, 35 U.S.C 103: From Hotchkiss to Hand to Rich, the Obvious

Patent Law Hall-of-Famers, 32 J. Marshall L. Rev. 437, 543 (1999) (describing import-ance of the Federal Circuit's creation for application of the Section 103 framework).

11. Dawson Chem. vs. Rohm and Haas Co., 448 U.S. 176 (1980) (setting forth history of

Section 271 and then applying the statute to hold no misuse where the holder of a patenton a method of using a chemical as a herbicide charges customers of the herbicide above

market price for the chemical itself and sues competing chemical company for contribu-

tory infringement).1 2. See infra note 42 and accompanying text.

13. That is, we do not suggest that legislative promulgation is itself infallible. See also infra

note 23.

14. For the most influential articulation of the 'basics approach' we explore in this chapter,

and the earliest we could find, see Giles S. Rich, The Relation Between Patent Practices

and the Anti-Monopoly Laws (pts. 1-5), 24 J. Pat. Off. Soc'y 85, 1 59, 241 , 328, 422 (1942).

15. See infra notes 70-71 and accompanying text (discussing evisceration of the areas of

contributory infringement and inducement of infringement by the doctrine of patent

misuse).

16. IP rights are rights to exclude others from doing something. IP rights are not rights to

do that something. Their impact is more precisely viewed as being exclusionary than

exclusive. The impact of IP rights is only properly viewed as being exclusive in those cases


where the one exercising the right to exclude happens otherwise to be free (such as from

other rights of exclusion or other regulations) to do the excluded activity.

17. Harold Demsetz, Toward a Theory of Property Rights, 57 Am. Econ. Rev. 347, 354

(1967); Harold Demsetz, The Private Production of Public Goods, 13 J.L. & Econ. 293

(1970). For a basic overview of the economics of price discrimination, see Jean Tirole,

The Theory of Industrial Organization 133-68 (1997).

18. See generally F. Scott Kieff, Property Rights and Property Rules for Commercializing

Inventions, 85 Minn. L. Rev. 697, 727-32 (2001).

19. See generally Gordon, supra note 1; Lunney, supra note 1; Meurer, supra note 1. For

more on the debate over the impact of imperfect price discrimination on output, see

Richard A. Posner, Antitrust in the New Economy, 68 Antitrust L.J. 925, 933 n. 10 (2001):

Perfect price discrimination would bring about the same output as under competi-

tion, because no customer willing to pay the seller's marginal cost would be turned

away. But perfect price discrimination is infeasible, and imperfect price discrimina-

tion can result in a lower or higher output than under competition, or the same

output. See F.M. Scherer and David Ross, Market Structure and Industrial

Performance 494-96 (3rd ed. 1990); Paul A. Samuelson, Foundations of Economic

Analysis 42-45 (1947); Joan Robinson, The Economics of Imperfect Competition188-95 (1933). Many economists believe that even crude discrimination is more likely

to expand than to reduce output, see, e.g., Robinson, supra, at 201 ;Scherer and Ross,

supra, at 494-96; Peter O. Steiner, Book Review, 44 U. Chi. L. Rev. 873, 882 (1977),

but there does not appear to be a firm basis for this belief. See Hal R. Varian, Price

Discrimination, in Handbook of Industrial Orgnanization, at 597, 629-33 (RichardSchmalensee and Robert D. Willig eds, 1989).

20. The court must discipline itself in several respects. In part, this means limiting, or at least

connecting, judicial analysis to established positive and normative decisional frame-

works. In part, this means expanding the analysis sufficiently to include the panoply of

established positive and normative decisional frameworks of both IP and non-IP areas

of law.

21 . For an example of the powerful effect of one such specialized doctrine - patent misuse-

see infra notes 70-71 and accompanying text. For more on the importance of propertyrule protection of intellectual property and the legislative history of the present statu-

tory regimes, see Kieff, supra note 18.

22. See infra notes 70-71 and accompanying text.

23. Although, again, we do urge those promulgating IP policy at any level to integrate the

procedural aspects of our approach into their decision making processes. See supra note

13 and accompanying text (describing procedural aspect of our approach when applied

to substantive analysis of potential new policies).

24. Baxter, supra note 1, would require that the licensing arrangement be confined 'as nar-

rowly and specifically as ... technology . . . and . . . administration permit.' Baxter,

supra note 1, at 313. Bowman, supra note 1, would endeavor to determine the extent to

which the arrangement deals with something that a court later determines to be com-

petitively superior to other available options-presumably rejecting the views of parties

to the particular arrangement under scrutiny who must have elected to enter into it over

other options available at the time of entering the arrangement. Kaplow, supra note 1,

would examine the ratio between the reward the patentee receives if the arrangement is

enforced and the monopoly loss that would result. More recently, Carrier, supra note 1,

has argued to look even more broadly to ex post data about how particular industries

have experienced innovation to determine whether it has tended to be driven more by

competition or by innovation, without offering devices for measuring any of these manyfactors.

For an interesting comparative institutional analysis that stresses the importance of

certainty and predictability in judicial decision making, see Cass R. Sunstein and Adrian

Vermeule, Interpretation and Institutions, 101 Mich. L. Rev. 885 (2003).


25. See generally Phillip Areeda and Louis Kaplow, Antitrust Analysis: Problems, Texts,

Cases 174-250, 447-77, 785-806 (1997).

26. United States vs. Aluminum Co. of America, 1 48 F.2d 4 1 6, 430 (2d Cir. 1 945).

27. For more on the rule of reason generally, see Areeda and Kaplow, supra note 25, at

203-50.

28. See generally id. at 663-784.

29. I suppose that a patentee has no less property in his patented machine than any other

owner, and that, in addition to keeping the machine to himself, the patent gives him

the further right to forbid the rest of the world from making others like it. In short,

for whatever motive, he may keep his device wholly out of use. So much being undis-

puted, I cannot understand why he may not keep it out of use unless the licensee, or,

for the matter of that, the buyer, will use some unpatented thing in connection with it.

Generally speaking, the measure of a condition is the consequence of a breach, andif that consequence is one that the owner may impose unconditionally, he may imposeit conditionally upon a certain event. . . . The domination [over a material used in a

patented device] is one only to the extent of the desire for the [patented device].

Motion Picture Patents Co. vs. Universal Film Mfg., 243 U.S. 502, 519-20 (1917) (Holmes,

J., dissenting) (citations omitted); Giles S. Rich, The Relation Between Patent Practices

and the Anti-Monopoly Laws (pt. 4), 24 J. Pat. Off. Soc'y 328, 330 (1942) (citing same and

providing English translation from Latin for the Justinian Maxim cited by Holmes: '[one]

to whom the greater is lawful ought not to be debarred from the less as unlawful').

30. See generally Areeda and Kaplow, supra note 25, at 413-44, 686-784.

3 1 . What is particularly troubling about the approaches we criticize is that they would have

striking implications if applied in analogous fashion to the real estate transactions men-tioned here, which, of course, they are not. As discussed infra in Section 9.4 (third sub-

section), we criticize in the IP context approaches that treat restrictive contractual

arrangements as illegal. As a result, they are not only unenforceable, but also efforts to

use them would be viewed as misuse and so would lead to the property to which those

transactions relate to be essentially forfeited. See Morton Salt vs. Suppiger, 314 U.S. 488

(1942) (a finding of misuse renders the IP right unenforceable). Consider the implication

of this reasoning for a real estate transaction involving the sale of half of a parcel in

which the half that is sold is encumbered by a negative easement - such as a promise not

to build a factory that produces smelly emissions. The reasoning of the preemption

approach we criticize would allow the buyer to argue that the proper domain of restric-

tions on emissions is the body of federal environmental law and that, therefore, under

the doctrine of conflict preemption the contract term to limit use, which is a matter of

state law, is preempted and thus not enforceable. What is more, the reasoning of the

misuse approach we criticize would further allow the buyer to argue that it is a misuse of

that property right to attempt to extract - or extort - such a promise, and as a result the

property right in the entire parcel itself is forfeited. Put simply, the one-two punch of the

approaches we criticize would allow even a buyer who is sophisticated, not resource con-

strained, advised by counsel, and fully possessed of contractual intent (and therefore not

a good candidate for the contract law defenses to formation of unconscionability, adhe-

sion, duress, mistake, etc.) effectively to take possession of the entire parcel of land

without paying a cent by simply waiting for the seller to offer half the parcel encumbered

by the negative easement at a price lower than for the whole. For an example of this typeof one-two punch in the case of IP, see infra note 85 and accompanying text.

32. Although there are a number of incentive-based theories for IP that are mentioned in the

literature - including 'incentive to invent,' 'incentive to disclose' or 'teach,' 'incentive to

innovate,' and 'incentive to design around' - there are essentially three dominant theories

today: (1) some version of the 'incentive to invent' and 'incentive to disclose' theories

treated together under the rubric of 'reward;' (2) the 'prospect' theory; and (3) the com-mercialization theory. IP law certainly does have a number of important effects, and each

of these theories of IP is useful in elucidating these effects. We emphasize here the com-mercialization theory and its associated focus on coordination because at a minimum


this theory did motivate the shaping of present IP regimes as a historical fact andbecause we see the commercialization effect as the most important in that the regimescan be and in many cases are most easily and effectively adapted to achieve that goal. For

a recent review of the patent literature on incentive theories and a collection of sources,

see Chisum et al., supra note 3, at 58-90 (reviewing various incentive theories for the

patent system); Rebecca S. Eisenberg, Patents and the Progress of Science: Exclusive

Rights and Experimental Use, 56 U. Chi. L. Rev. 1017, 1024-46 (1989) (same); A. Samuel

Oddi, Un- Unified Economic Theories Of Patents - The Not-Quite-Holy Grail, 71 Notre

Dame L. Rev. 267 (1996) (same). For recent reviews of the copyright literature on incen-

tive theories and a collection of sources, see Michael Abramowicz, Copyright

Redundancy, George Mason Law & Economics Research Paper No. 03-03, available at

http://papers.ssrn.com/sol3/results.cfm (reviewing and collecting sources and highlight-

ing the opportunity cost issues discussed by Lunney as well as showing how additional

works on the margin may contribute little while at the same time causing rent dissipa-

tion); Glynn S. Lunney, Jr., Reexamining Copyright's Incentives-Access Paradigm, 49

Vand. L. Rev. 483 (1996) (reviewing and collecting sources and suggesting that incentives

may draw efforts away from other productive activities). It should be noted, though, that

the 'basics approach' would hold even if the incentive-based theories for IP were stressed.

33. Innovation is a broader term than invention and is generally understood to include the

downstream dissemination of inventions. It is sometimes also called commercialization.

34. For a discussion of the problems with efforts to reward inventive activities, see, e.g.,

Chisum et al., supra note 3 at 70-72 (reviewing so-called 'incentive to invent' theory of

patents and criticisms thereto); Kieff supra note 18, at 707-17 (reviewing problems with

reward alternatives to patents).

35. For a thorough model of the commercialization goals of IP law, see Kieff, supra note 18.

36. For an overview of contributory and induced infringement and their history in the

patent context, which is representative for the rest of IP, see, e.g., Giles S. Rich,

Infringement Under Section 271 of the Patent Act of1952, 35 J. Pat. Off. Soc'y 476 (1953).

37. See, e.g., Hewlett-Packard Co. vs. Bausch & Lomb Inc., 909 F.2d 1464 (Fed. Cir. 1990)

(inducement of patent infringement requires proof of both intent to induce and actual

direct infringement by the one induced) (citing 35 U.S.C. 271(b)).

38. 35 U.S.C. 271(c) (contributory patent infringement).39. To be sure, the ease of predicting outcomes of indirect infringement is attenuated by the

uncertainties in other aspects of IP law on which indirect infringement may depend, such

as the basic scope of IP subject matter. For example, in patent law the basic scope of the

patent right to exclude hinges on the body of law governing the field called 'claim con-

struction,' which is presently the topic of substantial debate because it is considered by

many to be too uncertain. For more on claim construction, see the recent important

empirical work by Polk Wagner at www.claimconstruction.com. For another example,the extent of the home recording and person sharing exemptions in copyright law caused

a great deal of the uncertainty surrounding the indirect infringement claims in the

famous Napster and Aimster cases. See A&M Records, Inc. vs. Napster, Inc., 239 F.3d

1004, 1020 (9th Cir. 2001); In re Aimster Copyright Litigation, 334 F.3d 643 (7th Cir.

2003).

40. Before the 1 952 Act, courts had used the misuse doctrine to erode the ability for intel-

lectual property owners to price discriminate or engage in restricting licensing. Section

271(d) expressly states that such conduct shall not be misuse. See 35 U.S.C. 271(d)(l-3)

(added by the 1952 Patent Act); see also Kieff, supra note 18, at 736-38 (discussing

history of Section 271 of the 1952 Patent Act).

41. Dawson Chem. vs. Rohm and Haas Co., 448 U.S. 176 (1980).

42. See 35 U.S.C. 271(d)(4^5) (added by Pub. L. No. 100-703, 201, 102 Stat. 4676 (1988)).

In its entirety, augmented Section 271(d) provides:

No patent owner otherwise entitled to relief for infringement or contributory

infringement of a patent shall be denied relief or deemed guilty of misuse or illegal

extension of the patent right by reason of his having done one or more of the


following: (1) derived revenue from acts which if performed by another without his

consent would constitute contributory infringement of the patent; (2) licensed or

authorized another to perform acts which if performed without his consent would

constitute contributory infringement of the patent; (3) sought to enforce his patent

rights against infringement or contributory infringement; (4) refused to license or use

any rights to the patent; or (5) conditioned the license of any rights to the patent or

the sale of the patented product on the acquisition of a license to rights in another

patent or purchase of a separate product, unless, in view of the circumstances, the

patent owner has market power in the relevant market for the patent or patented

product on which the license or sale is conditioned.

43. See, e.g., Video Pipeline, Inc. vs. Buena Vista Home Entertainment, Inc., 342 F.3d 191,

203-04 (3rd Cir. 2003) (noting that although '[njeither the Supreme Court nor this Court

has affirmatively recognized the copyright misuse doctrine . . . [tjhere is ... a well-

established patent misuse doctrine, and . . . other courts of appeals have extended the

doctrine to the copyright context.').

44. See generally Carl W. Schwarz, The Intellectual PropertyIAntitrust Interface, in 1 No. 6

Andrews Intell. Prop. Litig. Rep. 15 (2000), available at WL 7 No. 6 ANIPLR 15 (copy-

right and trademark misuse are each derived from the law of patent misuse) (citing Juno

Online Servs., L.P. vs. Juno Lighting, Inc., 979 F. Supp. 684 (N.D. III. 1997) (trademark

misuse)). For a review of intellectual property misuse, including trademark misuse, see

generally American Bar Association Antitrust Section Intellectual Property Misuse:

Licensing and Litigation (2003).

45. Patents give the patentee the right to restrict use of what is claimed in the patent.

Copyrights give the copyright holder the right to restrict copying of the creative expres-

sion embodied in the protected work. Trademarks give the trademark owner the right to

restrict use of symbols that are confusingly similar to (and in some cases also those that

dilute) the protected mark. For none of these IP systems does the IP right give its holder

some affirmative right to use. Indeed, rights to use are entirely controlled by other areas

of law. For example, a patent on a drug does not allow the patentee to avoid FDA or

EPA restrictions on the drug's use. Similarly, various criminal and other public safety

laws would restrict the holder of a patent on a gun's right to use that gun. See generally

F. Scott Kieff, Patentsfor Environmentalists, 9 Wash. U.J.L. & Pol'y 307, 307-08 (2002)

(invited symposium piece for National Association of Environmental Law Societies'

annual meeting entitled 'Sustainable Agriculture: Food for the Future,' held 15-17

March, 2002, at Washington University School of Law) (discussing how the right to

restrict use conferred by IP law does not interfere with other restrictions on use).

46. See supra note 45.

47. See Thomas W. Merrill and Henry E. Smith, Optimal Standardization in the Law of

Property: The Numerus Clausus Principle, 1 10 Yale L.J. 1 (2000).

48. Id.

49. For more on property rules versus liability rules in the context of IP, see, e.g., Kieff, supranote 18, at 732-33.

50. See generally Russell Korobkin, Bounded Rationality, Standard Form Contracts, and

Unconscionability, 70 U. Chi. L. Rev. 1203 (2003) (reviewing debates in contract law

about the applicability of the unconscionability doctrine).

5 1 . See Giles S. Rich, The Relation Between Patent Practices and the Anti-Monopoly Laws

(pis. 1-5), 24 J. Pat. Off. Soc'y 85, 159, 241, 328, 422 (1942). As suggested supra note 14,

this is one reason why Rich's views have been so influential.

52. Texaco, Inc. vs. Pennzoil, Co., 729 S.W2d 768 (Tex.App. 1987).

53. The transformation of breach of contract into patent infringement is significant. At least

one essential difference between patent infringement and breach of contract is that the

remedies for infringement include a right to exclude (i.e., property rule protection),

whereas a contract is generally viewed as little more than a promise either to perform or

to breach and pay actual damages (i.e., liability rule protection).

54. Wallace vs. Holmes, 29 F.Cas. 74 (No. 17,100) (C.C.D. Conn. 1871).


55. See Chisum et al., supra note 3 at 950-55 (discussing history of contributory infringe-

ment doctrine and the role of the Wallace case).

56. Because the chimneys are usable with non-infringing lamps and are not specially adaptedfor infringing uses, their sale falls within the safe harbors of Section 271(c). See supranotes 36-38 and accompanying text (discussing safe harbors of Section 271(c)). Put

differently, the patent could not be asserted against the sale of the chimneys as a matter

of direct or indirect infringement.57. That is, the contract for sale of the machine included a set of contract terms relating to

the patent that were written on the label that was affixed to the machine itself.

58. Heaton-Peninsular Button-Fastener Co. vs. Eureka Specialty Co., 77 F. 288 (C.C.A. 6

1896) (opinion by Lurton, C.J.) (also known as the 'Button Fastener Case').

59. Rich, supra note 51, at 251. The successful argument in Heaton - offered by Frederick

P. Fish, founding partner of the law firm formerly known as Fish, Richardson, & Neave,which later became the firms of Fish & Richardson and Fish & Neave - held out the sales

of the staples as proxies, or counters, for measuring use of the patented machine. Theymay have been, and such an arrangement would likely have been efficient. But the cause

of action against the defendant, if any, would then be some form of interference with

contract, not patent infringement. Depending on the ultimate interpretation of the label

contract, the plaintiff may have had a cause of action against the party who was a cus-

tomer of both the plaintiff and the defendant for both breach of contract and patent

infringement.60. The court opinion suggests there may have been sufficient facts to mount such an

argument.61. The court opinion does not discuss these facts, but it is likely there was no evidence of

market power. It is curious that the court did not discuss the antitrust argument, because,

as Rich pointed out, the opinion was written against a background in which antitrust

law was recently enacted: The Sherman Act had been passed six years before!' Rich,

supra note 51, at 254 (punctuation emphasis in original).

62. Henry vs. A.B. Dick Co., 224 U.S. 1 (1912).

63. These cases supported the improper view that causes of action for patent infringementcould be maintained in situations where the basics would only allow a cause of action

for some form of contractual business tort, at most. This led pro-patent courts to undulystretch the reach of patent law and to the inevitable response by anti-patent courts that

the entire body of indirect infringement should be eliminated. See infra note 65 and

accompanying text.

64. See infra note 65 (reviewing history of these cases).

65. As the Supreme Court later pointed out in Dawson, when Rich was testifying in supportof what became Section 271 of the 1952 Patent Act, 'Rich warned against going too far

[and] took the position that a law designed to reinstate the broad contributory infringe-

ment reasoning of [A.B. Dick] "would kill itself in time."' Dawson Chem. vs. Rohm and

Haas Co., 448 U.S. 176, 208 (1980) (citing Hearings on H.R. 3866 before SubcommitteeNo. 4 of the House Committee on the Judiciary, 81st Cong., 1st Sess., 17 (1949) (testi-

mony of Giles Rich)).

As the Court also pointed out in Dawson, A.B. Dick 'was followed by what may be

characterized through the lens of hindsight as an inevitable judicial reaction.' Dawson,448 U.S. at 191 (citing Motion Picture Patents Co. vs. Universal Film Mfg. Co., 243 U.S.

502 (1917) (reaching a result opposite to A.B. Dick on similar facts involving a patent ona film projector and a restrictive label contract limiting use to certain film)). CompareMotion Picture Patents, 243 U.S. at 519-21 (Holmes, J., dissenting) (arguing that the

patentee should be entitled to capture all the market generated by the invention and

expressing concerns about the transactions that had been entered in reliance on the rule

of A. B. Dick).

The law continued to fluctuate after Motion Picture Patents. In United States vs.

United Shoe Machinery, 247 U.S. 32 (1918) ('Shoe Machinery P), a case also argued for

the patentee by Frederick P. Fish, the Court returned to reasoning similar to that in

A.B. Dick to permit a complex leasing arrangement. Soon thereafter, the Clayton Act


was passed, in part, in response to cases like A.B. Dick and Shoe Machinery /, and its

Section 3 was directed to sales and leases of articles of commerce 'whether patented or

unpatented.' 15 U.S.C. 14. Not surprisingly, in United States vs. United Shoe

Machinery, 258 U.S. 451 (1922) ('Shoe Machinery //), the Court found that the leases

violated the Clayton Act. Similarly, in International Business Machines Corp. vs. United

States, 298 U.S. 131 (1936) (75AT), the Court found a set of complex leasing arrange-ments accompanied by sales of punch cards to violate the Clayton Act.

This brief review of the evolution from A. B. Dick to IBM is provided here only for his-

torical context. A significantly more complete treatment is provided in Rich, supra note

51, at 24 1-283.

66. Leeds & Catlin Co. vs. Victor Talking Mach. Co., 213 U.S. 325 (1909).

67. Id. at 335.

68. For more on peripheral claiming, see F. Scott Kieff, Perusing Property Rights in DNA, in

F. Scott Kieff, Perspectives on Properties of the Human Genome Project 135 (2003). Adetermination of infringement under a central claiming system requires the court to

determine the heart of the invention and whether the putative infringement is close

enough to that heart to justify a judgment of infringement. A determination under

peripheral claiming requires the court to determine only the outer bounds of the claim.

Anything within those bounds infringes and anything outside does not. The so-called

'doctrine of equivalents' ('DOE') that exists under the present patent system, even

though not provided for in the statute, is an odd exception to the peripheral nature of our

present peripheral claiming system because it allows the patentee to capture somethingoutside the claim. Although some commentators like this doctrine because it gives some

flexibility, they fail to see how the patentee can achieve this same flexibility in a mannerthat is not only less costly to the patentee but also to all third parties by simply draftinga better patent disclosure at the outset. F. Scott Kieff, Property and Biotechnology, in

Chisum et al., supra note 3, at 318-323 (showing how as a matter of positive law and

practice the disclosure rules of Section 1 12 of the Patent Act can operate better than the

DOE for both patentees and third parties and citing F. Scott Kieff, The Case for

Registration and the Law and Economics of Present Patent-Obtaining Rules, 45 B.C. L.

Rev. 55, 99-105, 109-1 14 (2003) (discussing the normative case for the disclosure rules

and showing how they are a better institutional choice - in terms of minimizing social

costs - for allowing both patentees and third parties to manage the problem of claim

breadth than other institutional approaches such as the DOE)).69. Again, as discussed supra note 39, the relative crispness of these doctrines can be

muddied in practice by their interaction with other, fuzzier doctrines of each IP law

regime.

70. Carbice Corp. vs. American Patents Corp., 283 U.S. 27, 33 (1931). A similar approach wasfollowed in Lietch Mfg. vs. Barber Co., 302 U.S. 458 (1938) (also known generally as

'Barber'} (Brandeis, J.) (patentee 'attempting ... to employ the patent to secure a limited

monopoly of unpatented material').

7 1 . Rich supra note 5 1, at 345 (describing the opinions of the Court in Carbice and Barber

as revealing 'a very significant preoccupation by the Court with the objective of the

plaintiffs rather than with the doings of the defendant') (emphasis in original).

72. Mercoid Corp. vs. Mid-Continent Investment Co., 320 U.S. 661 (1944) ('Mercoid T), andMercoid Corp. vs. Minneapolis-Honeywell Regulator Co., 320 U.S. 680 (1944) ('Mercoid

//') (patent on new furnace stoker switch). The same approach was used earlier in

American Lecithin Co. vs. Warfield Co., 105 F.2d 207 (C.C.A. 7, 1939) (also known gen-

erally as'

Warfield} (patent on use of lecithin as an emulsifier in chocolates to improveits properties by, for example, preventing 'whitening' after only a few days).

73. See, e.g., Dawson Chem. vs. Rohm and Haas Co., 448 U.S. 1 76, 214 (1980) ('Respondent'smethod of doing business is thus essentially the same as the method condemned in the

Mercoid decisions, and the legislative history reveals that 271(d) was designed to retreat

from Mercoid in this regard.'). Section 271 achieved this result by codifying in subsec-

tions (a), (b), and (c) those acts that would constitute direct, induced, and contributory

infringement, respectively; while at the same time codifying in subsection (d) that it


would not be misuse for a patentee to sue or license anyone who could be sued undersubsections (a), (b), or (c).

74. See supra notes 36-39 and accompanying text.

75. The US Court of Appeals for the Federal Circuit has jurisdiction over most appeals in

patent cases. See Federal Courts Improvement Act of 1982, P.L. 97-164, 96 Stat. 25

(Apr. 2, 1982) (creating a uniform forum for patent appeals in the Federal Circuit bymerging the Court of Claims with the Court of Customs and Patent Appeals and trans-

ferring to the new court jurisdiction over appeals from patent cases that were tried in the

district courts). Patent cases for purposes of making this jurisdictional decision are those

in which the well-pleaded complaint alleges a claim arising under federal patent law.

Holmes Group, Inc. vs. Vornado Air Circulation Systems, Inc., 535 U.S. 826 (2002).76. Virginia Panel Corp. vs. MAC Panel Co., 1 33 F.3d 860, 869 (Fed. Cir. 1997) (internal cita-

tions omitted).77. According to the Federal Circuit in Virginia Panel'.

The courts have identified certain specific practices as constituting per se patent

misuse, including so-called 'tying' arrangements in which a patentee conditions a

license under the patent on the purchase of a separable, staple good, see, e.g., MortonSalt Co. vs. G. S. Suppiger Co., 314 U.S. 488, 491 (1942), and arrangements in whicha patentee effectively extends the term of its patent by requiring post-expiration roy-

alties, see, e.g., Brulotte vs. Thys Co., 379 U.S. 29, 33 (1964). Congress, however, has

established that other specific practices may not support a finding of patent misuse.

See 35 U.S.C. 271(d) (1994); Dawson Chem. Co. vs. Rohm & Haas Co., 448 U.S.

176, 202 (1980) (construing earlier version of 271(d)). A 1988 amendment to

271(d) provides that, inter alia, in the absence of market power, even a tying arrange-ment does not constitute patent misuse. See 35 U.S.C. 271(d)(5) (1994) (added byPub.L. No. 100-703, 201, 102 Stat. 4676 (1988)).

133 F.3d at 869 (internal citations shortened).78. For an expanded discussion of this point in the context of copyright misuse, which

derives from patent misuse, see Paredes, supra note 3.

79. Rich, supra note 29, at 245.

80. ProCD, Inc. vs. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996) (non-commercial use restriction

in shrink-wrap copyright license for computer program held valid and enforceable as acontractual limit on use).

81 . Mallinckrodt, Inc. vs. Medipart, Inc., 976 F.2d 700 (Fed. Cir. 1992) (single use restriction

in label license held valid and enforceable limit on grant of authority so that

unauthorized acts may support suit for infringement).82. Efforts to respond statutorily to these cases and others at the interface between contract

law and IP law, such as the proposed Article 2B of the Uniform Commercial Code andthe Uniform Computer Information Transactions Act ('UCITA'), should be conducted,if at all, along the lines of the procedural aspects of our 'basics approach' discussed

supra notes 1 3 and 23 and accompanying text.

83. For example, the payment may be for an activity that is not protectable by IP generally,

happens not to have been protected by any particular piece of IP, or was formerly pro-tected by some particular piece of IP.

84. Chief Justice Burger, writing for the majority of the Court, even allowed a promise to

pay royalties to reach activity that was never patented so long as at the time the contract

was executed it reflected both parties' reasoned assessment of the likelihood and payoffof the different states of the world under which patent rights might or might not materi-

alize. Aronson vs. Quick Point Pencil Co., 440 U.S. 257 (1979) (contract to pay royalty ona technology was enforceable even though no patent ever issued on the technology whereat the time the contract was entered into the technology might have been patented andthe contract provided a low royalty rate for the case where no patent issued and a higherrate for the case where a patent did issue).


85. See, e.g., Scheiber vs. Dolby Labs., Inc. 293 F.3d 1014 (7th Cir. 2002) (Posner, J.) (dis-

cussing at length the strength of the reasoning of the dissenting opinion of Justice

Harlan in Brulotte vs. Thys Co., 379 U.S. 29, 34 (1964) (Harlan, J., dissenting), but

nonetheless following the majority opinion in that case in refusing to enforce a properly

formed IP licensing contract - indeed, a settlement agreement from prior litigation-

among commercial parties simply because some payments happened to extend beyond

patent term at the request of the licensee). The case at the root of this line of precedent,

Brulotte, involved a patentee who sold a hop-picking machine to farmers and who had

several patents that would be infringed by such a machine. The machines were not sold

for a simple one-shot price. Instead, payment was to be made over time and based on the

actual economic advantage the machine generated for the farmer over alternative hop-

picking approaches. Because this meant that payment would extend beyond the last of

the patent terms, the Court held the contract to be unenforceable beyond that term in an

opinion written by Justice Douglas, who was well known for his dislike of patents. In

dissent, Justice Harlan pointed out that this holding would make unenforceable deals

that were actually advantageous to farmers who either were liquidity constrained at the

time of purchase or who were skeptical of the economic value of such capital equipment.86. Chisum et al., supra note 3 at 1 1 55. See generally id. at 1 1 55-96 (reviewing preemption).87. Nor do IP rights give IP owners any affirmative right to use.

88. For more on the conflict between preemption and the basics of IP law, see generally

F. Scott Kieff, Contrived Conflicts: The Supreme Court vs. the Basics of Intellectual

Property Law, 30 Wm. Mitchell L. Rev. 1717(2004) (invited piece for symposiumentitled 'The United States Supreme Court's Effect on Intellectual Property Law This

Millennium' at William Mitchell College of Law held April 24, 2004).

89. See Bonito Boats, Inc. vs. Thunder Craft Boats, Inc., 489 U.S. 141, 144-^5 (1989) (holding

that state law against so-called 'plug molding' of boat hulls was preempted by federal

patent law and citing Fla.Stat. 559.94 (1987)). To be sure, the reasoning explored in this

chapter is not new and indeed was more thoroughly set forth in the opinion by Judge Rich

in the case that was in conflict with the decision by the Florida Supreme Court in Bonito

Boats. See Interpart Corp. vs. Italia, 111 F.2d 678 (Fed. Cir. 1985) (Rich, J.) (no preemp-tion because patent law says nothing about a right to copy and because the state statute

did not even prevent copying - it merely prevented one form of copying). Similarly, the

reasoning of Bonito Boats is not new either and its roots can be found in the earlier cases

of Sears and Compco. Sears, Roebuck & Co. vs. Stiffel Co., 376 U.S. 225 (1964); CompcoCorp. vs. Day-Brite Lighting, Inc., 376 U.S. 234 (1964). As reviewed in detail throughoutChisum et al., supra note 3, Sears and Compco did not raise as many alarms as Bonito

Boats because these earlier cases came so soon after the 1952 Patent Act. It took the

Court until 1980, for the most part, to recognize the total overhaul in the framework of

patent law that was implemented by the 1952 Act.

90. Bonito Boats, 489 U.S. at 168.

91 . Compare, e.g., Bonito Boats, 489 U.S. at 152 ('[T]he federal patent laws must determine

not only what is protected, but also what is free for all to use. We have long held that after

the expiration of a federal patent, the subject matter of the patent passes to the free use

of the public as a matter of federal law.').

92. 489 U.S. at 168.

93. Also consider the real estate analogy discussed supra note 3 1 .

94. It is well recognized that Congress has the power to promulgate the statutes that create

the institutional framework for the positive law IP regimes. Patent and copyright laws

are promulgated pursuant to express authorization in Article 1 of the U.S. Constitution,

while the trademark laws are promulgated under the general commerce clause power of

Article 1 that is now recognized to be quite expansive. Compare In re Trade-Mark Cases,

100 U.S. 82, 94 (1879) (holding trademark laws to be improper exercise of the power to

promulgate patent and copyright laws and of the commerce clause power because they

regulate activity that is not sufficiently interstate) with Wickard vs. Filburn, 317 U.S. 1 1 1

(1942) (holding that even growing wheat for personal consumption in one's own back


yard has sufficient nexus to interstate commerce that it may be regulated by Congress

using commerce clause power).95. Even though the federalism and Supremacy Clause concerns of the US Constitution are

not applicable, the Court essentially used this same preemption approach against federal

IP law in Dastar and to a lesser extent in TrafFix. See Dastar Corp. vs. Twentieth CenturyFox Film, 539 U.S. 23 (2003) (holding that the Lanham Act does not prevent unaccred-

ited copying of uncopyrighted work and expressing concerns that otherwise the LanhamAct would interfere - or conflict - with the Copyright Act); TrafFix Devices, Inc. vs.

Marketing Displays, Inc., 532 U.S. 23 (2001) (holding that the existence of expired utility

patents in which certain design elements were mentioned created sufficiently strong evi-

dentiary inference of design's functionality that the design was not eligible for trademark

or trade dress protection and suggesting that otherwise there might be conflict between

the Lanham Act and the Patent Act); see also, Kieff, supra note 88, at 7-9 (discussing

Dastar and TrafFix cases).

96. See, e.g., Dow Chemical Co. vs. Exxon Corp., 139 F.3d 1470, 1473 (Fed. Cir.1998)

(exploring interaction between patent law and a state law providing a business tort for

interference with contract).

97. See, e.g., Hunter Douglas, Inc. vs. Harmonic Design, Inc., 153 F.3d 1318, 1336-37 (Fed.

Cir.1998) (no conflict-type preemption of various state law claims based on publicizingan allegedly invalid and unenforceable patent in the marketplace as long as the claimant

can show that the patent holder acted in bad faith in publication of the patent, which is

the 'extra element' beyond patent law).

98. Before the 1 976 Copyright Act, copyright term was shorter, but the copyright owner was

given a non-transferable renewal right. Since the implementation of the 1976 CopyrightAct, copyright term has been very long, but the original copyright owner is given a right

to terminate any transfers of that right during a statutorily-defined period within this

longer term.

99. See 17 U.S.C. 203 (governing grants executed on or after January 1, 1978); see also id.

304 (governing grants executed before January 1, 1978).

100. Digital Millennium Copyright Act, PL. 105-304, 1 12 Stat. 2860 (1998), implemented at

17 U.S.C. 1201(a)(lH2), (b)(l) (2000).

101. See supra Section 9.3, second and third subsections (discussing indirect infringementand contract, respectively).

Index

A&M Records, Inc. vs. Napster, Inc. 5

A.B. DickcasQ 171

Abramowicz, M. 104

access

and differentiated products theoryof copyright 106-107, 114

side of tradeoff 103, 105

advertising revenues 55

agency rule-making 9-10

Agreement on Trade Related Aspectsof Intellectual Property Rights

(TRIPS) 9, 15, 120,135American Society of Composers,

Authors and Publishers (ASCAP)38

ancillary products, tax on 41, 45, 51-2,

56

antitrust law 159, 161, 165-6, 175-6

appropriability 108, 134

and knowledge sharing 121, 131

Areeda, P. 150

atomization, of the production process127-33

audio equipment, tax on 5 1

Audio Home Recording Act (1992) 41,

81

audio-streaming technologies 60, 61

authors

as monopolists 103

payments to journals 47

rights of 4

'vanity' press 47

authorship 1,2

Bakos, Y. 64

Baumol,W.J. 115

Berry, S.T. 113

Besen, S. 26, 32, 64, 82, 104

blanket licenses 39, 40, 50, 53

comparison to copyright collectives

54-7

radio 54

television 54, 56

Bonanno, G. 1 1 5

Bonito Boats case 176-8, 189

book market 47

Borch, K. 31

Borenstein, S. 109

Boyle, J. 104

Brander,J.A. 115

breach of contract, vs. indirect

infringement 169-71

Brennan, D.J. 108

Brennan, T. 82

broadband 80

Broadcast Music, Inc. (BMI) 38

Brown, R. 80

cable operators 40, 41

cable transmission 55

Canada 14

Carbicecase 173

Carlton, D. 99

CDs 24, 42, 45, 51-2

and Internet piracy 60

market 50, 52

sales 47

Chamberlin, E.H. 105

Ciminello, D. 80

clubs 64

Coase theorem 23

codified knowledge 129, 130, 132, 133

Cohen, I.E. 104

commons 134

knowledge as 123

'tragedy of the commons' 122, 123

competition, and copyrighted works

104

competition law 144-5

vs. copyright 142

complementary domains 123

'composition' 124

797


compulsory licensing 2, 38, 39, 40, 41

Computer Associates vs. Altai, Inc. 1 1

computer programmesapplication programming interfaces

(APIs) 21

and copyright law 8, 10-14

'structure, sequence, and

organization' ('SSO') 10-1 1

comScore/Media Metrix 37

Conner, K.R. 65

consumption, non-rivalrous 103

Cooler, R. 143

copyright 1

differentiated products theory103-19

duration of 2-3, 135

extended 120, 121-2

foreign 3

impact of economics 2, 3-6

imperfections of system 43^4

justification 143^4-

length of term 45

and Locke's labour theory 143-4

markets 47

maximal 121, 133, 135

minimal 120, 132, 133, 136, 137

misuse 167, 175

models of design 81-2

nexus between industry and policymakers 4

open-ended rules 10

perpetual term 7

and policy 3-4

special exemptions 2

vs. competition law 142, 143-6

weakness in economic theory 128

welfare-enhancing effects of 121

Copyright Act 62

Copyright Arbitration Royalty Panel

(CARP) 40, 52, 54, 61

copyright collectives 2, 39

advantages of 24

amount of revenue to be raised

48-50

as coalitions 25

comparison to blanket licenses 54-7

distribution mechanisms 25

division of revenue 52^1

and efficient consumption 45-6

functions of 23

income 23-36

mispricing by 57

model 25-8

and MP3s 37-59

placement of tax 51-2

risk sharing distribution

mechanisms 28

to replace current system 40^1copyrighted works

and competition 104

as monopolies 83

copyright law 62-3

and computer programmes 8, 10-13

and database development 9

general applicability 5

literature on 2

Copyright License Board (CLB) 48,

49-50

copyright protectioninteraction among different aspects

113-14

letters and diaries 2

Copyright Term Extension Act

(CTEA)6, 111

and Eldred 6-8

Cournot market structures 104

Cournot oligopoly 99

Court of First Instance 151, 152, 153,

154

creation costs 98

'creative commons' 123, 125-7, 132,

133, 135, 136, 137

productive role 127

cultural differences 4

cultural economics 2

database development, and copyrightlaw 9

David, P.A. 126

Dawsoncase 167, 186

deadweight losses 106-107, 162

demand diversion 109

Demsetz, H. 105, 162

Depoorter, B. 82

derivative work right 2

differentiated products theory of

copyright 103-19

and access 106-107, 114

appropriability 108, 109, 112

breadth of rights 112-13

Index 193

and deadweight losses 106-107

demand diversion 109

determinants of equilibrium 111-13

early analyses 104

incentives 114

intensity of rights 111-12

and market power 106

normative implications 105-109

and optimal incentives 108-109

remedial implications 110-14

and sequential innovation 1 14

size of rights 1 1 1

ways of strengthening or weakening

rights 113-14

Digital Millennium Copyright Act

(DMCA) 6, 6 1,62, 77, 178

anti-circumvention provisions 9, 63

and decryption technology 80

digital music distribution 60-61

digital product, music as 61-2

'digital rights management' (DRM) 57,

63, 64, 77-8

fingerprinting 63

and market power 64

standards 64

vertical inefficiency 64

watermarking 63

distributional concerns 104

Dixit, A.K. 108

dominant firm structures 104

Drahos, P. 134

duration of copyright 2-3, 135

Duration Directive, EU 135

DVD decoding cases 80

Eaton, B.C. 106, 115

Eaton, J. 115

economic analysis 45differences in 5

economic aspects 1 3

economic literature, language problems4

economics, impact on copyright 2, 3-8

educational processes 131-2

Eldred vs. Ashcroft 3

and Copyright Term Extension Act

(CTEA) 6-8

effect on economics in copyright 6-8

envelope theorem 72-3

equal-share rule 31-2

European Union (EU)Court of Justice 142, 146, 149, 153

Directive on Competition Policy 1 3

Directive on sui generis 9

Duration Directive 135

exemptions from copyright 2

extended copyright 121-2

'fair use' 62, 63, 64, 121, 135

aggregate consumer surplus 87, 94-5

definition of 80

doctrine, and US copyright law 2

'equal diversion' setting 849gross benefit 87-8, 95-6

net economic welfare 92-4

number of uses 96-7

number of works trade offs 83,

84-5

optimal standards 97-8

as a policy instrument 80-102

reservation price 90-97, 98

and transactions costs 82-3, 90-97

file-sharing technologies 60, 61, 65, 71

legal actions against 62

network effects 65-6

fingerprinting, 'digital rights

management' (DRM) 63

Fisher, W.W. Ill 108, 110

Fitzgerald, D. 151

Foray, D. 126

foreign copyright 3

free entry 105, 106, 108

Geertz, C. 124, 125

'general law' 161

property and contracts 168

Goettler, R.L. 113

Gordon, W.J. 2, 80, 82, 99, 132

Graham case 160

Granovetter, M. 128

hacking 63

Hardin,G 105,121

Hart, O.D. 106

Hay, D.A. 115

Hayek, F. von 129

Heatoncase 170-71

Hettinger, E.C. 124

Hollander, A. 23

Hotelling,H. 115

123


Hughes,! 110

Hull, D. 151, 152

IBM 13

1CI and Commercial Solvents vs.

Commission 149

'idea-expression dichotomy' 2, 104

illegal copies, welfare implications 67

IMS Health case 142-58

background 146

Commission decision 147

court proceedings 1 52-4

licensing of brick structure 142, 146

reliance on Ladbroke case 151-2

reliance on Magillc&SQ 149-50, 151,

153

reliance on Oscar Bronner case

148-9, 154-5

incentive/exclusion trade-off 103, 105,

121

independent creation defense 2

indirect infringement 1 84

vs. breach of contract 169-71

information, as a public good 122

informational intermediaries 76

information transmission technologies67-8

infringement under IP law, vs. sui

generis\aw 171-5

innovation 161, 184

'integer problem' 115

intellectual creations, as public goods1,23,43,58,127

intellectual products 43^intellectual property

economic research on 5

and general law 168

incentive based theories 183-^

and price discrimination 162

Intellectual Property Clause, USConstitution 6

intellectual property law 159, 166-7

basics approach 161, 163-8

application 168-79

and the Beaton case 170-71

economic justification 1-2

infringement under vs. sui generis

law 171-5

periphery claims 1 87

theoretical framework 162^

intellectual property rights 130, 133

as rights of exclusion 162

Internet piracyand CDs 60

digital music distribution 60-61

effect of increased protection on

profits 74-6

with increased copyright protection,

model 66-76

information-pull technology 71-3,

74,76

information-push technology 69-7 1,

74

legal protection 66, 68, 70-71, 72, 76

technological and legal aspects60-64

technological protection 66, 68, 72

ISPs, tax on 52, 56, 59

Johnson, W. 82, 104

Jones, L.E. 106

Judd, K.L. 115

Kaldor, N. 114

Kazaa 76

Kirby, S.N. 64, 82, 104

Klein, B. 82

Klemperer, P. 110

knowledge

anthropologists on 125

codified knowledge 129, 130, 132,

133

collective and networked dimension

126

as a 'common' 123

dyadic structure 128-30

economic aspects 1 3

sharing process 124-5

social dimension 133

social nature of production 1 24-7

tacit knowledge 3, 129-30, 131, 132,

133, 135, 138

value of 125

knowledge creation, collective setting

136

knowledge sharing 125, 126, 138

and appropriability 121, 136

Koboldt, C. 104, 105

Koenker, R.W. 108

Koepp, R. 126

Index 195

Ladbrokecasc 149, 151-2

Landes, W.M. 64, 103, 104, 105, 121,

122, 143

learn-then-distribute rule 28, 29, 30, 33

Leeds & Catlin case 1 72-3

Lehmann, M. 144

Lemarchand, S. 143, 145

Lessig, L. 40

letters and diaries, copyright protection2

Liebowitz, S.J. 64, 65, 103

Lipsey,R.G. 106, 115

Liu,J.P. 110

Locke, John 1 34

labour theory 143^1

totemic paradigm 1 34

Lunney, G.S., Jr. 104

McQuail, D. 132

Magillcase 148, 149-50, 151, 152

use of term inter alia 1 50

Mallinckrodt case 176

Mankiw,N.G. 106

market failure 122, 130

market inefficiencies 44

market power 106

markets 46

copyright 47

Maskus, K. 144

maximal copyright 121, 133, 135

MercoidcasQs 174

minimal copyright 120, 132, 133, 136,

137

misuse 160, 161, 175-6

monopolies, as copyrighted works 83

monopolistic competition theory 105

monopolists, authors as 103

moral hazard 33

Morin, E. 128

MP3.com 76, 77

MP3sand album sales 49

cases involving 80

and CD sales 60

and copyright 57

and copyright collectives 37-59

downloads as a proxy for the market

50-51

measurement of downloads 53-4, 58

quality of files 65

multiple works 1 1 5

music

as an experience good 66, 68

as a digital product 61-2

musical composition 245, 25-8

heterogenous and correlated

lotteries 32-3

mutuality principle 3 1

Napster 37, 58,62-3,78, 139

National Commission on NewTechnological Uses of

Copyrighted Works (CONTU) 8

National Council for HigherEducation 82

National Research Council 82

Computer Science andTelecommunications Board 8 1

Nelson, K. 126

Nelson, R.R. 124,126,129,130Netanel, N. 41, 104

network effects, file-sharing

technologies 65-6

network externalities 4, 138

Neumann, M. 145

Neven,D.J. 115

Nielsen NetRatings 37

non-authorized audio-streaming

technologies 60, 61

non-obviousness 160, 181

non-rivalrous consumption 103

non-rivalrous good pricing 44

Novos, I. 65,69,81, 103, 105

Nozick, R. 143

open-ended rules 10

optimal incentives 108-109

Oscar Bronner case 148-9, 150, 152

Owen, B. 108

ownership 2

Parisi, F. 82

Patent Act (1952) 160, 166, 167, 169

patents 5, 12, 80, 137, 170, 180-81,185

'combination' 160

misuse 160, 167, 175

'non-obviousness' 160

and spatial competition 1 10

payola 46-7, 59


peer-to-peer (P2P) networks 5, 37, 60,

61,62,67-8

monitoring 65

opportunity cost 69

penalties for use 63

and record companies 76

Peitz, M. 64, 66

perfect price discrimination 44, 58

performing rights

as an ancillary market 55-6

pricing efficiency 545performing rights societies 39^0, 53,

54,56

performing rights tariffs 38, 54-5

Perloff, J. 99

perpetual term copyright 7

Perry, M.K. 108

PEW Internet project 37

piracy 133

see also Internet piracy

policy, and copyright 3-4

policy guidelines 133-5

Posner, R.A. 64, 103, 104, 105, 121,

122, 143

preemption 176-7

Prescott, E.G. 115

price discrimination 162, 167

pricing 47

privacy 63

private rights, enforcement 130-33

ProCD case 176

production, atomization of 127-33

profits, short-run 107, 116

property rights 105, 114, 121, 122

duration of 1 32

and Locke's totemic paradigm 134

on outputs 123

public choice problems 5-6

public goodinformation as 122

intellectual creations as 1, 23, 43, 58,

127

quid pro quo systems 42

radio

blanket licenses 54

excess entry to industry 113

performing rights tariff 55

and record sales 58-9

Ramello, G. 144, 145

Rate Courts 54

real-world difficulties 46-8

record industry 37-8, 42, 57

album revenues 49

and peer-to-peer (P2P) networks

76

royalty payments 55

sales and radio 58-9

Recording Industry Association of

America (RIAA) 37, 53, 62-3

research and development (R&D) costs

1,10restrictive licensing arrangements

176

Review of Economic Research on

Copyright Issues (RERCI) 15

risk sharing 23-36

Romney, A.K. 125

Rooney, D. 130

royalties 48

rates 54, 55

record industry 55

Rumelt, R.P. 65

Samuelson, P. 63

scarcity 124

Schmalensee, R. 115

Scotchmer, S. 124

Sega Enterprises, Ltd vs. Accolde, Inc.

13-14

Shachar, R. 113

sharing process, knowledge as 124

short-run, profits 107, 116

Shy, O. 65

social activities, financing 130

Society for Economic Research on

Copyright Issues (SERCI) 15

software copyright law 8, 10-14

songs, on record 40

Sony Betamax case 2

sound, social context of 126

spatial competition models 1 1 5

Spector, H. 144

Spence, M. 108

status quo systems 42

Stiglitz, I.E. 108

Stothers, Ch. 150, 155

strategic withholding 82

substitutability 112-13, 117

Index 197

sui generis 1 \ -2

European Union (EU) Directive 9

infringement under IP law 171-5

tacit knowledge 3, 129-30, 131, 132,

133, 135, 138

Takeyama, L.N. 65, 66, 82

tax, on ancillary products earlier twice

41,45,51-2,56

tax-and-subsidy system 41

television

blanket licenses 54, 56

performing rights tariff 55, 59

Texaco, Inc. vs. Pennzoil, Co. 169

Thisse, IF. 65

Torremans, P.L.C 143, 144, 150

'tragedy of the commons' 122, 123

transactions costs 23, 24, 25, 27, 32, 34,

82

and fair use 82-3, 90-97

Treacy,P. 149,150TRIPS Agreement see Agreement on

Trade Related Aspects of

Intellectual Property Rights

Ulen, T. 143

unauthorized copying, literature on64

USCongress

constitutional powers 7

extension of copyright 8

Constitution, Intellectual PropertyClause 6

copyright law, fair use doctrine 2

Copyright Office 3, 9-10

Copyright Terms Extension Act

(1998) 135

House and Senate JudiciaryCommittees 4

National Commission on NewTechnological Uses of

Copyrighted Works (CONTU) 8

'vanity' press 47

Virginia Panel case 175

Visscher, M. 115

Waelbroeck, P. 64, 66

Waldfogel, J. 113

Waldman, M. 65, 69, 82, 103, 105

Wallace case 169-70

watermarking, 'digital rights

management' (DRM) 63

Waterson, M. 115

Watt, R. 9, 104

Weber, M. 125

welfare-enhancing effects of copyright121

Whelan Associates vs. Jaslow Dental

Labs, Inc. 10-11

Whinston, M.D. 106

willingness to pay 98-9, 125

Wilson 32

Winter, S.G. 124, 129

Woolridge, F. 150

work made for hire rule 2

World Trade Organization (WTO) 9, 15

Yarrow, O.K. 113

Yoo,C.S. 104,115

Yoon, K. 82

Stanford Law Libre

3 LIDS Db3 ^37 Iflb

takeyama gordon towse developments in the economics of copyright

Documents