Академический Документы
Профессиональный Документы
Культура Документы
Developments
in the
Economics of Copyright
my mom, Margery Takeyama, with love, and in memory of my dad, Wallace Takeyama.
To
loving
LNT
To the memory of Sam Postbrief, and grandson, Theo Sam Rubin- Petrovic.
to the future
of his
WJG
Edited by
Lisa N.
Amherst
Takeyama
College,
University,
USA
Wendy J. Gordon
Boston University,
USA
Ruth Towse
Erasmus
University,
The Netherlands
Edward Elgar
Cheltenham,
UK
Northampton,
MA, USA
Lisa N. Takeyama,
All rights reserved.
Wendy
J.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher.
Published by Edward Elgar Publishing Limited
GlosGLSOlUA
UK
Edward Elgar
Suite 202
Northampton
Massachusetts 01060
USA
available
ISBN
84376 930
Printed and
bound
in
Great Britain by
Contents
List
List
vi
vii
xiv
xvi
2.
3.
Should economics play a role in copyright law and policy? Pamela Samuelson Risk sharing and the distribution of copyright collective income Arthur Snow and Richard Watt MP3s and copyright collectives: a cure worse than the disease?
Stan Liebowitz
Peer-to-peer, piracy and the copyright law: implications for consumers and artists Anne Duchene and Patrick Waelbroeck
'Fair use' as policy instrument
23 37
4.
60
80
103
5.
Timothy
6.
J.
Brennan
differentiated products theory of copyright
S.
Towards a
Christopher
7.
Yoo
Private appropriability
or contradiction?
commons
Giovanni B. Ramello
8.
120
approach
in a
market economy?
142
The
F.
basics matter: at the periphery of intellectual property Scott Kiejfand Troy A. Paredes
59
Index
191
Figures
3.1
3.2
4.
1
4.2
4.3
4.4
5.
1
LP albums sold per capita Album revenues Consumer demand with information-push technology Consumer demand with information-pull technology = Application with 9 0.15 = Application with cp 0.30
Consumer
diversion'
surplus, profits,
38
49
69 72
75
75
85
for
work
n:
'equal
5.2
Consumer
Consumer
surplus, profits,
fair
for
work work
n: profit-
use price
90
n: profit-
surplus, profits,
and
costs for
91
vi
Contributors
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
Timothy
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
and communications policy, with over 75 and chapters in journals and books covering economics, law, energy telecommunications, philosophy and politics. He serves on the
policy,
editorial
boards of the Journal of Regulatory Economics, Information Economics and Policy, and Communications Law and Policy. He is coauthor of Alternating Currents: Electricity Markets and Public Policy. He received his PhD in economics from the University of Wisconsin, Madison
in 1978.
in
economics
at
CERAS
- Ecole
et Chaussees, after graduate studies at Universite de (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
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.
the Boston University School of Law.
Among
Professor Gordon's honors are: being named a Fulbright Scholar; by the US Supreme Court; being elected to the Visiting
St.
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
Journal),
Law
Review),
Law
'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 Legal Studies 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 Campbell and 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,
at
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. He
edited the
(Elsevier),
book Perspectives on the Properties of the Human Genome Project and co-authored the popular treatise and casebook Principles of
Patent
Law (Foundation Press), now in its third edition. His research interamong law, economics, ethics, and crea-
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 decisionmaking 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
he has been called to testify as a legal expert before federal courts and agencies and maintains his connection to the law firm and business communities through his ongoing consulting practice, which focuses
in Chicago,
Contributors
ix
and
at
UCLA (PhD). He
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 Faculty
five
lished over 60 articles in journals including the American Economic Review and the Journal of Political Economy, as well as more popular outlets such
and CIO Magazine. He is director of the Center 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 of Economic 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 impact of 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 Economy was 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 Canada on issues related to technology and intellectual property.
as the Wall Street Journal
for the Analysis of Property Rights
Alessandra Narciso
is
University of Leeds (UK). She holds a University in Rome and an (UK). Her area of specialization
a doctoral researcher in the Department of Law of the first degree in law from La Sapienza
LLM
is
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,
hostile takeovers, executive
such as corporate governance and venture capital in developing countries, compensation, CEO overconfidence, informa-
and decision theory, asset securitization, intellectual property and bankruptcy, and hedge funds. Before joining Washington University, Paredes was a corporate and regulatory attorney.
tion overload
Developments
in the
economics of copyright
Giovanni B. Ramello
is
He
econom-
(PhD) and
in
computer
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: interplay 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 European Journal of Comparative Economics. His main research interests concern industrial economics, antitrust, economic analysis of law and institutions, intellectual property, information goods and knowledge production. He has published in these fields a number of papers in national and international journals and books.
Pamela Samuelson
is
a Chancellor's Professor of
Law and
Information
She is a Director Management 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
at the University
of California
at Berkeley.
new information
technologies pose for traditional legal regimes, especially She is co-author of the Software and Internet
of
The
(jointly with Suzanne Scotchmer), published in 2002 in the Yale Law Journal and of 'Intellectual property rights in data?' (jointly with J.H.
Reichman)
in 1997 in Vanderbilt
Law
large number of law review and computing journal articles. She is a Fellow of the Association for Computing Machinery (ACM) and a Contributing Editor 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 through June 1996 she was a member of the faculty at the University of Pittsburgh
Law School, from which she visited at Columbia, Cornell, and Emory Law Schools. She has been a member of the Berkeley faculty since 1996.
Since 2002, she has been an
Amsterdam.
Contributors
xi
Arthur Snow
is a Professor in the Department of Economics in the Terry at the University of Georgia. He earned a PhD in ecoof Business College 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 of Risk and Uncertainty, the International Economic
Review, the Economic Inquiry, and Economics Letters. He presided as presTheory Society at the 2004 Seminar, and currently serves on the editorial board of the Geneva Papers on Risk and Insurance Theory.
ident of the Risk
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 department of economics at the University of Oregon. Included among her most
well-known publications on piracy and intellectual property are The Welfare Implications of Unauthorized Reproduction of Intellectual
Demand Network Externalities' (Journal and 'The Intertemporal Consequences of of Economics) 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 publisher 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 of Economic 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,
Property in the Presence of
Industrial
.
USA.
Paul L.C. Torremans
Intellectual Property
is
Law at the School of Law of the University of He also holds an appointment as Professor of Private Nottingham (UK). 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 published various articles and books, including Holyoak and Torremans
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.), (Kluwer
Intellectual Property
Law International,
Ruth Towse
is
2004).
in Cultural
Reader
Industries at
is
Erasmus University
economics with special reference to the economics artists' labor markets and copyright in the cultural 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
in cultural
Patrick Waelbroeck
MA
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
in his
Madrid in 1 990. His interest in the economics of copyright is manifested book Copyright and Economic Theory: Friends or Foes! (Edward
Elgar, 2000). He is also actively researching the economics of risk bearing and risk sharing. He has published several papers in international journals
this topic, as well as on the theory of oligopoly. He is currently the General Secretary of the Society for Economic Research on Copyright Issues (SERCI) and the managing editor of the Review of Economics on
on
Contributors
xiii
Christopher S. Yoo is Associate Professor of Law at the Vanderbilt University Law School. graduate of Harvard College, the Anderson
School
at
UCLA, and
cutting-
edge economic theories are transforming the regulation of telecommunications and the electronic media. He is also pursuing research on the
work
application of imperfect competition models to copyright. His principal in these areas has been published in the New York University Law
the Georgetown
Review, the Northwestern University Law Review, the Cornell Law Review, 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
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
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.
copyright issues.
inception, SERCI has held annual international congresses in USA, and Italy. From these congresses, two collections of papers have been published, including the present book. Additionally, SERCI takes great pride in the launch of its new journal entitled, Review of
Since
its
Spain, the
Economic 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. Gordon and 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 contains 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 analyses.
Given the
diversity of topics
this
book should
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
sible,
book pos-
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 SERCI
would not
are also
this
exist.
We
(SGAE) for their continued support of, and belief in, SERCI. Many thanks
owed
to
volume
to press.
Finally, I would like to thank my wonderful co-editors, Wendy Gordon and Ruth Towse, who provided an abundance of invaluable input on this project and with whom I greatly enjoyed working.
in addressing
some of
the institutional
and
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 considering the 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 copyright can reduce transaction costs, the potential risk-sharing benefit of copyright collectives 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
of
its
risk-averse
uncertain, a copyright collective can improve the welfare 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.
Stan Liebowitz carefully examines the benefits and costs of certain proposals that have been put forth as alternative solutions to the
3,
In Chapter
problem of peer-to-peer music filesharing. Under these proposals, a copyright collective would be formed to administer a tax-and-subsidy system
xvi
xvii
whereby the record industry's traditional revenue sources would be replaced (or supplemented) by revenue generated from taxation of ancillary 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
losses induced
by
the benefits of such a system might include, for example, decriminalization of widespread illegal behavior, enhanced incentives for the production of creative works, and increased
unauthorized copying.
Some of
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 proposals (as his title indicates) 'a cure worse than the disease'.
Peer-to-peer music filesharing is also addressed by Anne Duchene and Patrick Waelbroeck in Chapter 4. However, in contrast to Chapter 3's
emphasis on proposals to resolve the problems induced by peer-to-peer music filesharing, Chapter 4 is premised on the idea that firms may benefit from peer-to-peer networks in their capacity to convey information to consumers about new music cheaply, as compared to the more costly traditional
mechanisms
(e.g.,
is
advertising and promotions). Duchene and that increasing copyright protection may have
conveyed to consumers via peer-to-peer networks ('demandor via traditional mechanisms ('demand-push'). They develop a pull')
model from which they derive welfare implications of increasing copyright's criminal penalties under both demand-pull and demand-push mechanisms 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 employing peer-to-peer networks.
The
effect
Timothy Brennan addresses the issue of fair use, the 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
In Chapter
5,
flexible doctrine
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 model from which he derives optimal fair use standards under several different
policy objectives. Specifically, in addition to economic efficiency, he considers four possible alternate policy objectives: maximization of number of
benefit,
and numbers of
xviii
Developments
in the
economics of copyright
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'.
uses.
In Chapter 6, Christopher Yoo criticizes the standard economic analysis of copyright that typically assumes copyright turns creators into monopolists. Yoo argues that the well-known 'access-incentives' tradeoff is largely driven by the assumed monopolization of the market. If one instead considers the impact of copyright within a framework that allows for product
differentiation
significant.
and
entry,
the
access-incentives
tradeoff
may
be
less
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
By
access.
Yoo
makes a broader
set
of policy
instruments potentially available for use, but disentangling the effects of one from another can be quite complicated.
Giovanni Ramello argues that the 'standard' economic is too narrowly focused on individual creation. In particular, 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 copyright
In Chapter
7,
analysis of copyright
restricts access to
output
(i.e.,
ways
to the copyrighted good), they have been in which copyright also restricts access
(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 creative commons, a consequence which Ramello calls an 'opposite tragedy of the creative commons'. For these reasons, Ramello argues for a 'minimal-
to inputs
ist'
8, Alessandra Narciso and Paul Torremans examine the and proper balance between copyright and competition (antitrust) 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 application of competition law to cases in which copyright exists should factor in and avoid undermining the structural system originally put in place by
In Chapter
interaction
copyright. So, for example, in antitrust cases involving copyright, it may be inappropriate to employ as precedent court decisions in non-copyright
without explicit recognition of the special structure intentionput into place by copyright. Narciso and Torremans apply this notion to several other policy issues raised by the IMS Health case, with a focus
cases, at least
ally
on the
In Chapter
Scott Kieff
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 piecemeal fashion by creating specialized approaches and unique doctrines. This approach, they argue, 'selectively emphasizes or alternatively ignores par-
of various legal disciplines in crafting specialized doctrines under their competing 'basics approach,' Kieff and Paredes propose that such matters would be best resolved by applying the
ticular features
basics of the relevant substantive law, such as intellectual property law, antitrust law, and property law. Although Kieff and Paredes focus on patent
law, their analysis applies equally to copyright.
that
the
anti-circumvention
provisions
of
the
Copyright Act are inconsistent with the 'basics approach' since the transactions 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 economists and legal scholars are applying their analytical skills to a broad range of theoretical and applied areas in the economics of copyright. We hope
that the
area.
in this
important
role in
1.1
INTRODUCTION
1
The principal justification for intellectual property (IP) laws in the AngloAmerican tradition is economic. 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 embodiment of an innovation generally requires very costly investments; subsequent copies are generally far less costly to make, and are often trivially 2 cheap and easy. 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. 3
rival
Although the non-excludable, noncharacter of intellectual creations resembles public goods, 4 it is socially desirable for private actors, rather than the government, to make
the investments to overcome this public goods problem. 5 Intellectual property rights are an ingenious device for addressing this problem: the
rights to qualifying innovators, and the rights provide incentives to individuals to invest in innovation because the rights
government grants
costs if granted provide assurance that innovators can recoup their their intellectual creations prove to be valued by the public. 6 Original works of authorship are particularly vulnerable to market-destructive
R&D
copying, which
regime.
for
7
is
the economic
wisdom of a copyright
'a natural field
seem to be
economic analysis of
law.'
It
compared with other fields of economic regulation, 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
Developments
in the
economics of copyright
in this field
of law and
community.
growing
The lack of influence is not attributable to inattention from the scholarly Law and economics scholars have produced a substantial and
literature that analyzes
aspects of copyright law. Among the copyright rules that are widely viewed as economically sound are: the idea/expression distinction (that is, copyright protects only an author's
many
10
11 12 expression, not her ideas), the independent creation defense, the derivative work right (that is, the right to control adaptations of the work, such
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 16 copyright, and the work made for hire rule (which treats employers as the 'author' of works by employees performed within the scope of employ17 ment). The economics of copyright collectives and of compulsory licensas a
18 ing has also garnered scholarly attention. 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
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
is
doctrine of
US copyright law. 21
scholars often view uses of copyrighted works as 'fair' if high transactions costs or other factors have impeded the effective establishment of a market
been Wendy Gordon's Sony Betamax case. Gordon argued that use of Betamax machines to tape broadcast television programs for later
analysis of the fair use issues in the
viewing should be fair because the costs of negotiating rights clearances would greatly exceed fees collectable for this kind of socially beneficial
Because the economic effects of a challenged use are almost always of central importance in judging whether the use will be deemed noninfringing as a fair use, economics is especially useful in providing guidance
use.
23
on
letters
works would be created without economic incentives of copyright. 25 Special exemptions from copyright liability for certain events organized by agricultural cooperatives and veterans' groups seem to be the product of special interest lobbying, rather than rigorous economic analysis. 26 Extension of the duration of copyrights
for copyright protection, given that these
the
role in copyright
'restoration' of foreign copyrights that had 28 long been in the public domain for failure to comply with US formalities (e.g., putting copyright notices on published copies of the work) are also
economically questionable.
even
its
29
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
.3 will
of extending the terms of existing copyrights and the Supreme Court's decision in the Eldred vs. Ashcroft case. 30 Although an initial reading of the
Supreme Court's decision might suggest that economics will have little influence in future court decisions interpreting copyright law, Section 1 .3 explains
why this interpretation of the Eldred decision discuss particular venues in which economics
is
incorrect. Section
.4 will
in the
copyright law and policymaking process going forward. Legislators should (and occasionally do) seek independent economic expertise when considering copyright
find
and related intellectual property proposals. Courts may economic analysis a useful input to sound decision-making, especially
in cases involving
new technology
might
resist the
issues unanticipated
adopted copyright
right professionals
right policy process, and why economics should nonetheless play a important role in copyright policymaking in the future.
.2
One salient factor explaining the low impact of economic analysis on copyand policy is a lack of economic expertise in the relevant policymaking 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 economics for guidance about how the law should be crafted. 31 Unlike the US
right law
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 regulations, the US Copyright Office (as well as its patent and trademark counterpart) has no in-house economic experts who might contribute insights about economic effects of various policy proposals. 32
Developments
in the
economics of copyright
For path-dependent historical reasons, copyright and other intellectual property policy matters in the US have generally been considered by subcommittees of the House and Senate Judiciary Committees, rather than by
Commerce 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
revolving door enables copyright experts to move from the or legislative committee staff positions to law firms representlegislature
ing major copyright firms to industry associations to government agencies 34 This contributes to responsible for framing copyright law and policy.
insularity in copyright policy analysis. The is not alone in its systematic, if perhaps unconscious, exclusion of or inattention to economic analysis in the copyright policymaking
enacts. 33
US
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 historically conceptualized copyright in utilitarian terms.
Although the tight nexus between the copyright industry and the policymaking 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 professionals. Few copyright professionals are able, for example, to penetrate
literature. If
the mathematical expressions that often pervade the academic economic economists wish for their work to inform the copyright policy
process, they will have to learn to translate their insights into a more vernacular form. They may also need to publish policy-relevant work in venues
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 software 36 ) 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
37 A legal and policy audience may find incomplete an economic analysis whose policy implications are not spelled out. If economists want their work to inform the copyright policymaking process, they will
professionals.
have to invest the extra time necessary to articulate the policy implications that flow from their analyses.
The standard economic practice of articulating simplifying assumptions before constructing an economic analysis does not mesh well with the
role in copyright
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, grantmakers 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 economists and economically-informed lawyers to exploring the economics of
copyright.
38
is
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 provisions 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 impossible, to assess with any precision the economic effects likely to flow from
adoption of particular copyright
39 right regime as a whole. It is
rules, let
in their analyses
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 econofile
mists considered
it
market for sound recordings. 41 Policymakers will understandably be reluctant to put much weight on economic expertise
was harmful
to the
if
same phenomenon. Another barrier to influence is the understandable reluctance of copyright lawyers and policy professionals to acknowledge the public choice problems presented by the current copyright law and policy-making
process.
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
42
Developments
in the
economics of copyright
concentrated benefit a legislature can grant them. The public may ultimately have to pay higher costs if copyright lobbyists are successful, but these
costs are diffuse
firms.
Collective action problems make it difficult for parties that will be negatively affected by higher protection rules to organize effective resistance to
copyright industry lobbying. This mix of concentrated benefits and distributed costs is likely to yield the best laws money can buy. The Copyright
(DMCA)
from the
US experience.
.3
WHAT EFFECT WILL ELDRED VS. ASHCROFT HAVE ON THE FUTURE OF ECONOMICS IN
COPYRIGHT?
economic and non-economic
rea-
soning in the copyright law and policy process is evident in the Eldred vs. 47 Ashcroft case decided in 2003 by the US Supreme Court. Eric Eldred, an
online publisher of public domain works, challenged the CTEA, which extended copyright terms an additional twenty years, based on an economic 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
was
Congress from granting exclusive rights to authors without the quid pro quo of a 'progress-promoting' act (i.e., authorship of a newly created original
work)
in return. 49 Stated
more
colloquially,
Constitution by extending the term of existing copyrights for an additional twenty years because the public got nothing in return for that grant. The
Such a quid pro 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 CTEA
could 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 CTEA would impose on subsequent users. 53 Tracking down copyright owners and negotiating with them is costly, even if the owners do not insist on payments
for reproducing or distributing copies of their works.
CTEA
impedes
role in copyright
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 new
derivative works. 54
The economists' brief also pointed out that life of the author plus 70 years was virtually equivalent to the perpetual term of copy55 right that the Constitution forbids.
rejected Eldred's challenge to the rational basis, the Court decided, for believing that
CTEA. Congress
to more investments in distributing and preserving existing works. 56 The Constitution gave to Congress the responsibility of crafting copyright legislation, and even if it exercised this power unwisely, the Court was reluctant to second-guess legislative decisions. 57 Only Justice Breyer would have struck down the on an economic interpretation of the Constitution. 58
CTEA
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 among the more conservative members of the Court. 59 Given how often the Court
has emphasized the economic incentive rationale for the existence of copyright law,
60
it is
striking
Justice Ginsburg,
how little attention was paid to economics in Eldred. who authored the majority opinion, is not, of course, the
Justices.
within the Court about constitutional powers of Congress and about how much deference to give to legislation emanating
is
a deep
rift
Eldred's lawyers drew upon several Supreme Court precedents interpreting the constitutional powers of Congress in a restrictive manner
from
61
and establishing rigorous standards for judging the constitutionality of certain kinds of legislation. 62 Further limitations on Congress' power under
Article
I,
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
Court to further
woo the more conservative members of the 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
other restrictive rulings would
restrictions
lot going on in the Eldred case that had nothing whatever 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.
65
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
Developments
in the
economics of copyright
developed an elaborate chart to distinguish the CTEA from other exten67 this chart may have backfired because it underscored the frequency sions, of these enactments. To comprehend the scope of Congress' power [to
extend terms of existing copyrights] under the Copyright Clause,' said the 68 Court, "a page of history is worth a volume of logic." The Court was also understandably concerned about the implications of
' '
CTEA
was
term
extensions in the Copyright Act of 1976 would almost certainly be challenged, and it was not immediately evident how the Court could find a prin69 Had the cipled basis for distinguishing this extension from the CTEA. 70 Court acceded to the heightened scrutiny for which Eldred argued, it
might have opened the floodgates of constitutional challenges to many 71 forms of intellectual property legislation, not just to copyright rules. 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-
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
right law
will consider several contexts in
likely to
is
be a useful input into the policymaking process for copyright and related forms of intellectual property law.
.4
When legislatures consider proposals to add a new subject matter to copyadopt 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
right or to
ence in the
or information resources. Economic expertise, for example, had some influUS 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 pro73 grams. The European Union too relied on economic analysis in crafting
74 appropriate copyright rules for computer programs. Owing in substantial to protect programs by copyright, part to the decisions of the US and
EU
role in copyright
the
became an international standard in 1 994 after member states of World Trade Organization (WTO) acceded to the Agreement on TradeRelated Aspects of Intellectual Property Rights (TRIPS), which required
75 copyright protection for programs. 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-
when adopting
tection for other innovations. It relied heavily on economic arguments in 1984 a new law to protect original layouts of circuits in
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
in the debate about whether 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-
who
database development with 15 years of exclusive rights to control extraction 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. Had
Commission been more rigorous in its use of economics, the European Parliament might have realized there was less of a need for this sui generis law than the Commission posited, and might have adopted a somewhat
the
narrower form of legal protection for data in databases. 82 Economic input should also be sought when copyright laws are
CTEA,
it is
effects of copyright term extension before enacting conceivable that economic arguments against its adoption
would 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 84 heavily criticized as overbroad and anti-competitive. Copyright rules are also sometimes promulgated in the context of agency rule-making, and economic analysis should be a useful input to decisionmaking in this venue as well. For example, the Copyright Office is required
to
10
Developments
in the
economics of copyright
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
US
to conduct rule-makings every three years to consider whether to exempt certain 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 economic effects of broad anti-circumvention rules.
Once rules have been legislated or issued by an appropriate agency, economics may sometimes have a useful role to play, and sometimes not, in how the law is interpreted thereafter. If a legislature has adopted an openended 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 interpreting the law in an economically sound way. However, if the legislature has precisely crafted certain rules that prove to be economically harmful, it may be more appropriate to invoke economic analysis to support proposed amendments 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 US
judicial experience interpreting copyright law as applied to computer 88 Whelan Associates vs. Jaslow Dental Labs., Inc. was the first programs.
American appellate decision to consider whether the 'structure, sequence, and organization' ('SSO') of computer programs could be protected by 89 Whelan relied upon both doctrinal and economic grounds copyright law. in 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 structure, sequence, and organization of literary works are generally protectable by copyright law, the structure, sequence, and organization of programs
should be protected by copyright law as well. 90
Complementing
that
focused on the need for software developers to have sufficient protection to recoup development costs and on the locus of value in computer
programs:
By
grams
far the larger portion of the expense and difficulty in creating computer prois 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 Info World, 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
role in copyright
1 1
.'). The evidence in this case is program code which implements the product that Ms. Whelan spent a tremendous amount of time studying Jaslow Labs, organizing the modules and subroutines for the Dentalab program, and working out the data arrangements, and a comparatively small amount of time actually
. .
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 development of new computer devices that accomplish the same end.' 93 Between 1986 and 1992, the Whelan decision was influential in subse94 quent US cases, both in its doctrinal and economic reasoning. The first
decision that challenged Whelarfs hegemony was Computer Associates vs. 95 On one important point, the Second Circuit Court Altai, Inc. in 1992.
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
97 program, 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 tech98 niques, none of which was protected by copyright law.
The court in Altai directed that these and other unprotectable elements of programs be 'filtered out' before infringement analysis began 99 to ensure compliance with section 102(b) of the US copyright statute which states
no case shall copyright protection for an original work of authorextend to any idea, procedure, process, system, method of operation, ship concept, principle, or discovery, regardless of the form in which it is
that
'[i]n
100 The court described, explained, illustrated, or embodied in such work.' in attributruled that structural similarities Altai were some ultimately
Altai were
developing programs to interoperate with the same three IBM operating system programs, and other structural similarities were to be expected in
101 Hence, programs of that kind.
it
noninfringement.
12
Developments
in the
economics of copyright
As
in
Whelan,
CA
and
its
amid
for 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.
.
.
into
computer
result in
inherently protectable expression. Thus, Feist implicitly undercuts the Whelan rationale, 'which allow[ed] copyright protection beyond the literal computer
order to] provide the proper incentive for programmers by protectefforts. .' [citation omitted]. We note that Whelan was decided 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
[in
code ...
ing their
most valuable
The Second
Circuit
went on to
say:
[W]e are unpersuaded that the test we approve today will lead to the dire consequences for the computer program industry that plaintiff and some amici predict. To the contrary, serious students of the industry have been highly critical of the sweeping scope of copyright protection engendered by the Whelan rule, in that it 'enables first comers to "lock up" basic programming techniques
as implemented in programs to perform particular tasks.' 103
The Second Circuit warned that if courts heeded purely economic arguments for broadening the scope of copyright protection for computer programs, 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 corrosive effect on certain fundamental tenets of copyright doctrine. If the test we
. .
have outlined results in narrowing the scope of protection, as we expect it will, that result flows from applying, in accordance with Congressional intent, long104 standing principles of copyright law to computer programs.
The Second
suitable
more
way to protecting some program innovations than copyright. 105 If copyright proved to be too 'thin' to provide proper incentives to program
106 developers, this was a matter for Congress, not the courts, to consider. Altai holds that copyright protection is unavailable to elements of programs necessary for interoperation with other programs (that is, program
107
interfaces).
Judged purely
and judgment
required to design them, program interfaces might initially seem to be 'ori108 ginal' enough to be protectable expression as a matter of copyright law.
role in copyright
However, once developed, program interfaces unquestionably constrain the design choices of subsequent programmers seeking to develop software 109 capable of successfully interacting with an existing program. The competition policy significance of copyright protection for interfaces was recognized in Europe in part because the Competition Policy Directorate of the European Commission had taken action against IBM arising from its practice of changing interfaces in a manner that had exclusionary impacts on European developers of computer peripherals. European software developers were going to be able to compete with
110
If
US
software 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
111 The US and EU consencopyright protection for software in Europe. sus on limiting copyright to enable interoperability influenced other
addressed a related interoperSega Enterprises, Ltd. vs. Accolade, Inc. issue as to whether could ability programmers lawfully decompile or disassemble other firms' programs in order to get access to information needed to develop an interoperable program. This was a dicey issue for copyright law 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 on authors to control the reproduction of their works in copies. 115 Yet unless
copyright law recognized at least a limited right to decompile or disassemble programs, the decision not to protect interface information would be
significantly
113
undermined. 116
relied
Accolade
on the
fair
use defense of
US copyright law in
respond-
117 Courts consider four factors in ing to Sega's charge of infringement. determining whether a use is fair: the purpose of the defendant's activ-
the nature of the copyrighted work, the amount and substantiality of the use, and the effects of the use on the market for the copyrighted work. 118 Accolade persuaded the court that decompilation was the only
ities,
way Accolade could obtain access to the unprotectable interface information 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 consequently
decided that intermediate copying of programs for the purpose of achiev119 The court relied on economic reasoning interoperability was fair use.
ing in support of the ruling, although
its
analysis
is
not as systematically
14
Developments
in the
economics of copyright
legal
rules in the
Both Altai and Sega vs. Accolade have been widely accepted as sound US and abroad, 121 and the software industry has continued to
Two
lessons
from the software copyright caselaw are, first, that courts should be wary of unsubstantiated economic claims that without broad protection, the affected industry will be severely harmed, and second, that considering evidence 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, challenges posed by
digital
new
works continue to be troublesome. Perhaps the most potent example has been the entertainment industry lawsuits against the developers of 123 While economic analysis has been peer-to-peer file sharing technologies.
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
125 non-infringing uses.
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
impacts of the challenged use was impressive and bodes well for further uses of economic arguments in copyright cases in Canada.
1.5
CONCLUSION
is
an important form of economic regulation. This chapter has why economic analysis has had relatively little effect on copyright law and policy thus far. It has also argued that economics 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, many in the copyright field, including policymakers, have little or no economic expertise and little inclination to seek it out. Second, copyright industry
Copyright
considered several reasons
role in copyright
much
not
all
little
copyright industry professionals may be wary of embracing economics because although it may sometimes provide support for stronger protection rules, it may not always cut in their favor.
embracing a tool that might provide grounds for policy some major participant in the copydecides that economic analysis would be valuright policy-making process
They may
resist
outcomes they
able as input to the policy process, the inertia of the current situation likely to persist.
is
Even
so,
predict that
economic analysis
will
corpus of policy-relevant copyright research and is articulating the policyrelevant 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 of Economic Research on 128 The increasing importance of copyright Copyright Issues (RERCI). industries to the US and world economy should make policymakers more 129 Moreover, the inclusion receptive to insights that economics can provide. of copyright and other intellectual property rights within the framework for promoting world trade by virtue of adoption of the TRIPS Agreement as part of the agreements establishing the World Trade Organization should lead to greater use of economic learning to shape global intellectual 13 This should lead to greater reliance on economic expertproperty norms.
ise at
and policymaking
process.
Perhaps only a matter of time before the US Copyright Office and sister 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,
effects before endorsing parprinciples can contribute to evolving our understanding of the access/incentives tradeoff so that copyright
Economic
continue - or perhaps return to - its long-standing tradition of promoting the progress of science and the useful arts to the benefit of
will
humankind.
ACKNOWLEDGMENT
I
wish to thank
for inviting
MA
Developments
in the
economics of copyright
NOTES
1
.
Highway: From Gutenberg to the Celestial Jukebox, 168-69 (1994); Mark A. Lemley, The Economics of Improvement in Intellectual Property Law, 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 'innovation' to encompass the works of both authors and inventors.
See, e.g., Paul Goldstein, Copyright's
2.
3.
4.
Gordon, Asymmetric Market Failure and Prisoner's Dilemma in Dayton L. Rev. 853 (1992). 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 cloning of computer program innovations). 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
See, e.g.,
Wendy
J.
Intellectual Property, 17 U.
5.
6.
7.
Reichman, Electronic Information Tools - The Outer Edge of World Law, 24 Int'l Rev. Indus. Prop. & Copyright L. 446 (1993). See, e.g., William M. Landes and Richard A. Posner, An Economic Analysis of Copyright Law, 18 J. Legal Stud. 325, 333-44 (1989) (presenting a formal model of copyright). But see Stephen Breyer, The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs, 84 Harv. L. Rev. 281 (1970) (questioning the economic need for copyright). See also Barry W. Tyerman, The Economic Rationale for Copyright Protection for Published Books: A Reply to Professor
See, e.g., J.H.
Intellectual Property
8. 9.
L. Rev. 1100 (1971); Stephen Breyer, Copyright: Breyer, 18 L. Rev. 75 (1972). Landes and Posner, supra note 7, at 325.
UCLA
UCLA
Rejoinder, 20
See, e.g.,
Law
10.
analysis of many bodies of law). 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,
nomic
&
supra note 7; Arnold Plant, The Economic Aspects of Copyright in Books, 1 Econometrica 167 (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 of Intellectual Property (2002 Vol. I, Parts II-V). Other important, but more recent, work is cited throughout this
article.
11. 12.
e.g.,
7, at
347-53. 354-55.
13.
14.
e.g.,
7, at
17U.S.C.sec.202.
17 U.S.C. sec. 201(d). 17 U.S.C. sec. 302-304. See infra notes 51-55
15.
16.
for
economic argu-
17.
18.
of limits on the duration of copyrights. 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). 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,
in favor
ments
Contracting Into Liability Rules: Intellectual Property Rights and Collective Rights Organizations, 84 Calif. L. Rev. 1293 (1996).
role in copyright
19.
of the Arts
(2003).
20.
in
M. Cherbo and Margaret Jane Wyszomirski, eds. The Public Life America (2000); Ruth Towse, ed., A Handbook on Cultural Economics
21
See Richard Watt, Copyright and Economic Theory: Friends or Foesi (2000). Row vs. The Nation, Inc.: Copyrightability and Timothy J. Brennan, Harper Fair 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
See, e.g.,
&
22.
23.
(1985); Landes and Posner, supra note 7, at 357-61; Robert P. Merges, Are You Making Fun of Me? Notes on Market Failure and the Parody Defense of Copyright, 21 AIPLA L.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). See, e.g., Gordon, supra note 4, at 1652-57. Id. at 1602-22. Gordon argues that fair use should be awarded in a wide range of other
id. at
1627-36 (discussing,
e.g., externalities,
antidissemination motives,
Gordon, Excuse and Justification in the Law of Fair Use: Transaction Costs Have Always Been Only Part of the Story, 13 J. Copyright
and nonmonetizable
Society 149(2003).
24.
interests);
Wendy
J.
work
25.
one of the principal criteria for judging whether a use is fair. See 17 U.S.C. sec. 107. See, e.g., William M. Landes, Copyright Protection of Letters, Diaries, and Other Unpublished Works: An Economic Approach, 21 J. Legal Stud. 79 (1992).
See 17 U.S.C. See Sonny
sec.
1
26. 27.
10(6),
10(10).
panying
28.
text.
17 U.S.C. sec. 104A. See, e.g., Edward Lee, The Public's Domain: Evolution of Legal Restraints 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). 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). Eldred vs. Ashcroft, 123 S.Ct. 769 (2003). See, e.g., Peter A. Jaszi, Toward a Theory of Copyright: The Metamorphoses of Authorship, 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). A decade ago, I recommended that government offices charged with responsibilities for intellectual property policy should hire staff economists. See Pamela Samuelson, Will the
29.
30.
31
32.
in the
Twenty-First Century?, 13
Cardozo Arts
& Ent.
L.J.
34.
Litman, Digital Copyright (2001); William F. Patry, Copyright and the Ent. L.J. 139, 141 (1996). Legislative Process: A Personal Perspective, 14 Cardozo Arts 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,
See, e.g., Jessica
&
the
US
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
&
35.
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 Time Warner. France and Germany are among the countries that protect authors' rights on this basis.
staffer in
AOL
See,
e.g.,
1,
at
Developments
in the
economics of copyright
36.
Copyright: Between History and Nature, in Of Authors and Origins: Essays on Copyright Law, 235 (Brad Sherman and Alain Strowel, eds, 1994). 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 Protection for 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
37.
computer program behavior and other industrial design features of programs). 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 externaliagainst cloning of
38.
An important contribution to the legal literature on network effects is Mark A. Lemley and David F. McGowan, Legal Implications of Network Economic Effects, 86 Cal. L. Rev. 479 (1998) (drawing out several possible legal implications of network externalities). 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.,
ties).
supra note
39.
3.
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 economic analysis in respect of intellectual property law because of inadequacies in empiriSee, e.g.,
40.
Cir. 2002).
41
economic expert David Teece focused on the harmful effects on the plaintiffs' ability to offer competing digital music services. Id. at 1017. Defendants' expert argued that use of Napster for sampling purposes enhanced the willingness of users to buy recorded music. Id. Plaintiffs commissioned two empirical studies of the effects of peerto-peer filesharing on sales of recorded music which concluded that among college students, 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 Theory for Statutory Interpretation, 74 Va. L. Rev. 275 (1988); Daniel A. Farber
Philip
P.
and
43.
Frickey,
S.
Choice:
44.
Lunney, Jr., The Death of Copyright: Digital Technology, Private 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). 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 unpublished works has been controversial. See James Adams, Estates' Rights in Canadian
See, e.g.,
Copyright Re-examined, Globe and Mail, Sept. 23, 2003, available at http://www.theglobeandmail.com/servlet/ArticleNews/TPStory/LAC/20030923/COPY23/TPEntertain
45.
ment/TopStories. Pub. L. No. 105-304, 112 Stat. 2860 (1998), anti-circumvention provisions codified at 17 U.S.C. sec. 1201.
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
46.
47.
http://www.freedom-to-tinker.com/superdmca.html. Eldred vs. Ashcroft, 123 S.Ct. 769 (2003). Justices Stevens and Breyer dissented in separate 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)
dissent).
role in copyright
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.
Id. at 23.
49.
50. 51.
Id.
al. in
Eldred
vs.
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.
53.
Id. at 8.
54.
55. 56. 57. 58.
12- 14.
Id. at 8.
783-84.
801 (Breyer, J., dissenting) ('The economic effect of this 20-year extension - the - is to make the copyright term longest blanket extension since the Nation's founding
Id. at
59.
And more importantly, its practical effect is not not limited, but virtually perpetual. to promote, but to inhibit the progress of "Science" - by which word the Framers meant .'). learning or knowledge. See, e.g., Eastman Kodak Co. vs. Image Technical Services, Inc., 504 U.S. 451 (1992);
.
.
61
No. 2 vs. Hyde, 466 U.S. 2 (1984). 347 U.S. 201, 219 (1954); Sony Corp. of Am. vs. Universal City Studios, Inc., 464 U.S. 417, 429 (1984). 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 dissented 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
Mazer vs.
Stein,
Congress' power and considering their implications for various intellectual property
62.
63.
64.
12-13, 20-21.
65.
In view of Justice Breyer's earlier economic skepticism about copyright, see Breyer, supra on economic grounds note 7, Eldred's lawyers expected him to be skeptical of the and 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
CTEA
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
66. 67.
See
Addendum
to
in
Eldred
at
vs.
Ashcroft (chart
distinguishing
68.
CTEA
available
http://eldred.cc/legal/
69.
supremecourt.html. Eldred, 123 S.Ct. at 778, quoting Eldred, 123 S.Ct. at 790.
New York Trust Co. vs. Eisner, 256 U.S. 345, 349 (1921).
20
Developments
in the
economics of copyright
70.
71.
72. 73.
Brief for Petitioners, supra note 48, at 39-47; Eldred, 123 S.Ct. at 788-90. See, e.g., Pamela Samuelson, The Constitutional Law of Intellectual Property After Eldred vs. Ashcroft, 50 J. Cop. Soc'y 547, 556 (2003). Eldred, 123 S.Ct. at 775.
74.
75.
76.
See Final Report of the National Commission on New Technological Uses of Copyrighted Works 14-26 (1979). The analysis focused on the economics of developing individual programs, but did not consider the appropriate scope of copyright protection for computer programs or vexing issues such as the interoperability and decompilation issues discussed infra notes 98-120 and accompanying text. 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. Copyright Protection For Computer Programs: Are They More Different Than They Seem?, 13 Law Comm. 279, 287-88 (1994) (discussing the concerns of the European J. Commission's competition directorate about the software directive that influenced the rule in favor of decompilation for interoperability purposes). See Agreement on Trade-Related Aspects of Intellectual Property Rights, in The Legal Texts: The Results of the Uruguay Round of Multilateral Trade Negotiations 321, art. 10(1) at 325 (Cambridge 1994) (cited hereinafter as TRIPS'). 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 Committee on 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 appropriate and 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
CONTU
&
77.
78.
79.
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). 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. 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). 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.
Intellectual Property
Rights
80.
EU
Data?, 50 Vand. L. Rev. 51, 80-97 (1997). Database Directive, supra note 79, recitals 1-12.
in
81. 82.
See, e.g.,
79, at
113-36
(criticizing the
EU
sui
84.
See, e.g., 17 U.S.C. sees. 1003-1007. See, e.g., Neil W. Netanel, Impose a Noncommercial File Sharing, 17 Harv. J.L. Tech.l (2003).
86.
Use Levy
to
&
role in copyright
21
87.
17 U.S.C. sec. 1201(a)(l)(C). See Memorandum to James Billington from Mary Beth Peters regarding Recommendations of 2002-4, Rulemaking on Exemptions from Prohibition on Circumvention of Copyright Protection Systems For Access Control
RM
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.
89.
797 F.2d
Id. at
222 (3rd
Cir.
986).
1239-40.
Id. at 1231.
Whelan
Associates, Inc.
(D.
95.
Mass
1990).
Cir. 1992).
96.
97.
702-3.
98. 99.
100.
Altai,
Id. at 707.
101.
102. 103. 104. 105.
106.
Altai,
Id. at 71 1-12.
Id. at Id.
Id.
See
id.
107. 108.
Id. at
707-11.
Arthur R. Miller, Copyright Protection for 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).
See, e.g.,
109.
/4/to/,982F.2dat710.
See, e.g.,
110.
Vinje,
The
EC
Directive
On The
Legal
Protection
J.
Duke
Comp.
111. 112.
Id. at
71-78.
113.
14.
Jonathan Band and Masanobu Katoh, Interfaces on Trial: Intellectual Property and Interoperability in the Global Software Industry 271-97 (1995). 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. See, e.g., Andrew Johnson-Laird, Software Reverse Engineering In The Real World, 19
See, e.g.,
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
some developers regard as unacceptable (e.g., giving up the right to of the same program for other platforms). See Sega, 977 F.2d at 1514.
that
117. 118.
1
make a version
The
fair
19.
use defense is discussed in Sega, 975 F.2d at 977 F.2d 1510, 1527-8.
520-28.
in favor
120.
22
Developments
in the
economics of copyright
121
decompilation for purposes of interoperability, see, e.g., Lawrence Graham and Richard 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. See, e.g., Band and Katoh, supra note 1 12, at 271-82, 294-97; Jonathan Band, Software Reverse Engineering Amendments in Singapore and Australia, J. Internet L. 17, 20
(Jan. 2000).
22.
See, e.g. Price WaterhouseCoopers, Contributions of the Packaged Software Industry to the Global Economy (1999), http://www.bsa.org/usa/globalip/econ/pwcl999.pdf.
,
123.
Records, Inc. vs. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001) (prelimSee, e.g., inary injunction against peer-to-peer software developer as a contributory and vicarious infringer of digital music copyrights because of centralized directory and indexing
functionality); In re Aimster Litigation, 334 F.3d 643 (7th Cir. 2003) (summary judgment against peer-to-peer developer as contributory infringer because of direct assistance 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-
A&M
24.
125.
ware developer because of capability for substantial noninfringing uses). See supra note 4 1 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.
127.
128.
Theberge vs. Galierie d'Art du Petit Champlain, Inc., 2002 S.C.C. 24. 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, Copyright and Product Differentiation, 79 N.Y.U.L. Rev. 212 (2004)). See, e.g., Gordon and Watt, supra note 10. For information about SERCI and its conSee, e.g.,
See, e.g.,
129.
and journal, see http://www.serci.org/. 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
ferences
- in 1996 their value increasing at an impressive rate surpassed that of every other export sector in the United States.)
130.
Pamela Samuelson, Challenges For the World Intellectual Property e.g., 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).
See,
2.
2.1
INTRODUCTION
p.
Hollander (1984,
ives as follows:
200) sums
up the
Copyright collectives are associations to whom authors transfer copyrights for purposes of exploitation. In general, collectives are concerned with the following functions: (1) they grant licenses for the use of works in their repertory,
they negotiate and collect royalties, and distribute them to their members, they take legal action against those who infringe the copyrights to which they hold title.
(2) (3)
Each of these three aspects of collective administration of copyrights (as 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,
well as the
each individual
member can achieve a better utility position than by acting we shall consider the possible ways in which collectimproves 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-
payment for, and protecting license agreements). Clearly, 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
ing, receiving
then, the
between individuals is traded, transactions costs not only exist, but they may even 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
24
Developments
in the
economics of copyright
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
tration.
is
a 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-
been fixed onto a physical support, say a CD Rom, it can easily be 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 privileges is automatically purchased as well. These include the right to listen to
ition has
listened to
most
Rom to a third party. In the music at any time and the right to sell the Rom countries, the consumer has the right to make a copy of the
CD
CD
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
on both the monitoring costs and the costs of whenever such recourse is needed. In this way, the option of collective administration can make feasible a set of transactions that is unprofitable when carried out individually. In economic terms, copyright
likewise they can also save
litigation
1
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
collectives enjoy the benefits
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 situation of its members. If the members are risk-averse (again, a very natural
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 administration
tions
and in real-life cases. Given the diversity in the market value of each composition
it is
copy-
quite
common
25
take great pains in analysing the markets in which their repertories are used order to determine how much use is made of each particular composition in the repertory. This is puzzling since it goes quite against the idea
in
that the songs are bundled in the first place, in a blanket license, in order to 2 save on such transactions costs. The reason why this is done is simply to
final distribution
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 collective'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 distribution rule will affect the number of members that the coalition has (in equilibrium) as well as the welfare of each individual. We shall impose two basic requirements of the distribution mechanism of a coalition's net income. First, we will require that the distribution method be such that individual 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
tive distribution
satisfies these
requirements.
preferred by
all
mechanisms
2.2
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
26
Developments
in the
economics of copyright
all
assumption that
deterministic,
1992).
is
common
will
Moreover, we
not only
identically distributed, but that they are also statistically independent with a probability of success that is common knowledge. 3 Concretely, we assume
value
that each composition can be either a 'success', in which case is x = 1 or a 'flop', in which case its market value is x = 0.
,
its
market
We assume
and
that the probability of success is/?, which is assumed strictly between 1 and we denote the true outcome of composition i by x r
,
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.
instead
we
assumed to be strictly wealth-loving and In order to avoid the need for formally considering take the approach of mean-variance analysis. Concretely,
we
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 marketing
shall
assume that
final
of the composition on an individual basis, then the expected value of his wealth (assuming that he has no wealth outside of what the composis
ition generates)
given by
(2.1)
is
given by
-p),
(2.2)
Now, consider the situation of the coalition. Assuming that the coali< n members, the final market value of the blanket license tion has
m
.
It can take on any of the discrete values on 0, 1, 2, 3, ., m, depending exactly how many of the m member comare successes. the well-known formula for the binomial positions Using
compositions
will
be
m\
27
Given
this,
is
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
is the same as for any particular individual; that is, the entire 4 repertory can be marketed at a (transactions) cost of t. Given this, the coalition (with members) earns a net income of X(m) t for distribution
repertory
to the
collective
indi-
we
fix at
A:,
and then
give
+k
Under
is
x-
(2.3)
with a variance of
x-cClearly,
\-L-c- m \\=E(p-xY=p(\-p\ //
I
\
(2.4)
under
members
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
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
28
Developments
is
in the
economics of copyright
more
easily satisfied
when k
is
small,
when
t is
each of which is intuitively understandable. In short, for any given values of / and A:, we have:
or
is
when m
large,
distribution rule under which a collective learns (at a Proposition 2.1 cost of k) the final result of each independent composition in its repertory, and distributes final income accordingly, is incentive compatible.
= (t + k)/t, beyond which given by m* the distribution rule also satisfies the participation condition. 5
There
is
minimum membership,
Assuming
members, the
learn-then-distribute system satisfies both incentive compatibility and participation 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
averse,
risk
of the members. Given that the creators are riskwould be welfare enhancing if the coalition could offer a better situation as well as transactions costs savings. We now go on to conit
sider if this
is
indeed possible.
2.3
the opportunity that the copyright collective has of offering 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
full
Consider
and
fair
member
the expected value of his composition less his share of the trans-
actions costs with certainty. In this case, each creator receives an expected
wealth of
p~ c ~~
<
2 5>
-
with zero variance. Clearly, this distribution rule satisfies both participation and incentive compatibility. Moreover, it also Pareto dominates the learnthen-distribute rule, since all members get a lower variance of wealth and
a higher expected final wealth. Thus,
we
have:
Proposition 2.2 sharing rule stipulating that each individual copyright holder receives the expected value of his composition with
29
certainty satisfies both participation by all individuals and incentive compatibility. It also strongly Pareto dominates the learn-then-distribute
rule.
The
mechanism
are so
obvious begs the question of why the answer lies in the fact that,
not typically used. Perhaps the collective pays each member his
system
is
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
insurance option may not be feasible. However, there are intermediate options that have full insurance, although not fair insurance, that satisfy
fair
still
Pareto
dom-
sum
Consider the following simple option. The collective distributes a fixed 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-thendistribute. That is, each member is given a fixed payment of
t
+k
p
Once
m =p
mm
again, this rule clearly satisfies incentive compatibility, and satisfies 6 participation under the same proviso as in Proposition 2. 1 Moreover, this
.
system strictly preferred by copyright holders to the learn-thendistribute rule since it offers the same expected wealth with a lower variance.
is
all
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
runs into* a bankruptcy can be reduced problem significantly (although, naturally, the reduction on the size of For if we assume that m = 100, and if we depends k). example, assume that k = 10, then the probability of bankruptcy is under 3 per cent. 7
this system, the probability that the collective
Naturally, the accumulated but not distributed net income of the collective would, at certain times, be distributed to the members as profit (as opposed
to 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.
30
Developments
in the
economics of copyright
Proposition 2.3
lective receives
A sharing rule stipulating that each member of the colwith certainty a payment equal to the expected value of
the
patibility always,
payment under the learn-then-distribute rule, satisfies incentive comand 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 bankruptcy 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
is
8
EX(m) = mp
and the variance of the market value of the repertory
is
(2.6)
E(mp
- X(m)) 2 = mp(\-p).
income
is
/
(2.7)
Now
collective's net
pre-agreed shares, such that member net income of the collective, where
receives a proportion
is,
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-m
mj
I
= \ EX(m)-c-i
= \mp -c-m
with variance
(2.8)
\mp - c - m -
\ X(m)
i
-c-tn
} }
= E(\.(mp =
X(m)))
\ 2 E(mp-X(m)) 2
31
Using
rule
is
(2.7),
we find
\}mp(\-p\
Now,
that
(2.9)
since each composition is, ex ante, identical, let us simply assume this case the expected value of final wealth for each
member of
the collective
is
m
the same as under the full and fair insurance mechanism, and than under the learn-then-distribute rule. On the other hand, the greater variance of the final wealth of each of the collective members is
which
is
~rwhich
is 0),
is
but
greater than any full insurance mechanism (under which variance less than what the learn-then-distribute rule offers (which from
(expected value
is
increasing in
greater
than acting alone because of the transactions costs savings, and the variance 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
and a lower
variance). Thus,
we
have:
A sharing rule stipulating that each collective member an share of the final net income of the collective is both paid equal incentive compatible and satisfies participation. It also strictly Pareto dominates the learn-then-distribute rule.
Proposition 2.4
is
rently used method, and that will wonders if this method can be improved upon. This type of question has been analysed extensively in the literature on efficient risk sharing. Two main results of that literature are worth mentioning here. First, the
if
which the net income of a copyPareto dominates the curnot imply any bankruptcy issues, one
in
strictly
so-called 'mutuality principle' of Karl Borch (see Borch 1960) implies that a group of risk-averse agents each contribute a random variable to a
common
32
Developments
in the
economics of copyright
must make the final wealth of each individual member independent of 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
rule.
arises:
it
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
particular, since
satisfies
improved upon. In
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 member of a mutual should
risk sharing efficiency will require that
each
receive a fraction of the total pooled wealth equal to his individual risk tolerance divided by the sum
all
members.
The sharing rule based on risk tolerances would, almost certainly, imply that members who supply identically valued compositions to the repertory
receive different shares of the final pooled wealth. Besen et al. (1992, pp. 395-7) note that discriminatory practices in the distribution of collective income are often frowned upon by regulatory authorities. 9 Given this,
remains to be seen if the fully efficient sharing rule offered by Wilson indeed feasible in the regulatory scenario of most copyright collectives.
it
is
2.4
Our
When
into play that is satisfied vacuously when probnamely, the distribution rule must satisfy the requirement of 'group rationality.' That is, it must not be the case that a subset of
additional condition
comes
members prefers to defect to an alternative collective. As an example, suppose that half of the compositions in the repertory have a high probability of success while the others have a low probability. Clearly, the equal
the
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
33
many members
However, group rationality can be restored by appropriately increasing the share allotted to those with a high probability of success. As long as the collective can categorize copyright holders according to their probabilities, the benefits of risk pooling described in the previous section can be realized by
viable copyright collectives.
Although we have focused on the case of independent composition lotcorrelated lotteries would yield the same qualitative results, except in the unrealistic case of perfect positive correlation where all compositions either succeed or fail together and there are no insurance possibilities.
teries,
collective's
Negative correlation introduces hedging opportunities that improve the own risk situation. Indeed, with perfect negative correlation, the
it
collective
can attain a riskless position, just as it can whenever to a perfect reinsurance market. In these cases the collective
full
has access
offer its
can
members
insurance for their composition lotteries. As correlation among the lotteries and reinsurance opportunities become imperfect, the insurance benefits a collective can offer become less valuable, but
fair
and
rule.
2.5
ASYMMETRIC INFORMATION
far
Thus
we have
abstracted from the informational asymmetries associand adverse selection. The learn-then-distribute
rule provides the strongest incentive to create a profitable copyrightable 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 composer's costly effort to produce a successful piece cannot be monitored and rewarded directly. However, the rule is an inefficient response to the
problem, since
failure
it
composer bearing
all
the risk of
when, as we have seen, insurance opportunities are available. If moral hazard poses a significant problem for the membership of a collective, then it would be inefficient to take full advantage of the available insurance 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 productive effort against the risk-bearing costs that this necessarily entails.
An
optimal balance would call for a distribution rule that ties royalty payments partially to a composition's realized success, offering some incentive for
tory, offering
productive effort, and partially to the realized success of the entire repersome insurance against failure. Thus, moral hazard would
34
Developments
in the
economics of copyright
mitigate, but not eliminate the advantage rules that incorporate risk pooling.
we have identified
for distribution
Adverse
selection,
collective
have different perceptions of the likelihood of success, poses a different problem. In a heterogeneous population, when individual composers are
classification
better informed about the probability of success than the collective, risk through categorization and screening, as commonly practiced
in insurance markets,
would allow the collective to provide insurance to its as effectively as when information is symmetric. 10 not members, although 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. 11 Indeed, it is possible for the divergence in beliefs to be so great that the insurance terms
offered
by the
members. In
learn-then-distribute rule
copyright collective.
2.6
CONCLUSIONS
we have proposed that copyright collectives may not be 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
In this chapter
offering their
on a
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
rule. First,
a partial explication can be found in the existence of asymmetric information - adverse selection and/or moral hazard. In these cases, it is well known
that the agent (the informed party, here, the creator) must retain risk in order to satisfy incentive compatibility. Secondly, copyright holders may be unrealistically optimistic
35
full
members could be
members were
NOTES
1
.
We note that in some instances, some of the monitoring costs are borne by the purchasis required to report usage rates to the collective. In any case, the colneeds to suffer some costs in processing and administering this information. 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
2.
3.
4.
5.
6.
opportunity to exercise monopoly power. In sections 2.4 and 2.5 we consider the implications of relaxing these assumptions. 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 requirement = 1 In fact, the more the underlying is that q < m, and in the chapter we have assumed q 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. 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. Indeed, participation can be achieved with a lower membership than under the learnthen-distribute rule, since each member gets a better risk-bearing situation. Exactly how the limit value of m is affected is impossible to determine unless we introduce an analysis based on utility functions.
' .
7.
is
These
results are
989) for
9.
American Society of Composers, Authors and Publishers (ASCAP) 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
in particular
licences'.
0.
1 1
The
their
may
12.
work was discussed by Watt (2000, p. 89), and also by Besen (1987). 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,
sis',
S.
RAND Report N
(1987),
intellectual property:
An
economic analy-
36
Developments
in the
economics of copyright
Besen, Besen,
S. and S. Kirby (1989), 'Compensating creators of intellectual property: R-3751-MF. Collectives that collect', Report
RAND
Salop (1992), 'An economic analysis of copyright collectives', Virginia Law Review, 78, 383-41 1. Borch, K. (1960), The safety loading of reinsurance premiums', Skandinavisk
S., S.
Kirby and
S.
Aktuarietskrift, 153-84.
New
York:
McGraw-Hill.
Crocker, K.J. and A.
(ed.),
Snow
(2000),
Handbook of
Insurance, Boston,
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,
Wilson, R. (1968),
Econometrica, 36,
13-32.
3.
MP3s and
copyright collectives:
3.1
INTRODUCTION
The advent of massive unauthorized copying by individuals using peer-to-peer systems and MP3 files, besides generating an enormous amount 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 frustrated in
attempt to have the courts shut down these latter firms. Simultaneously, the record industry has been covertly distributing on
its
1
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 unclear exactly how this would be done.
MP3
files,
although
it is
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 hundreds 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
38
Developments
in the
economics of copyright
6.0
5.5
5.0
4.5
4.0
3.5
3.0 2.5
2.0
1.5
1.0
Figure
3.
sharp decline of about 30 per cent in the number of subscribers to major 2 filesharing networks after announcement of these law suites.
legitimate grounds for concern. As Figure 3.1 demonstrates, sales since 1999 have been on a steady decline. The decline is considerably larger than any other decline in the last 30 years, although
is still
introduction of
radio in 1921
smaller than the declines brought about from the 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
from vinyl records. For many analysts, the behavior of the copyright industries appears misguided. 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 3 with copyright, do not effectively address the interests of creators or users.
Partly as a backlash to the behavior of the copyright owners, many academics 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
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 proposals would probably be to call them variations of a 'copyright collective'.
fact
39
they appear to have 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
if
only
we have
the courage to scrap the current system. While I will discuss these proposals in
is
more
that a pool of money would be generated in a secondary market (presumably 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 some commentators see a copyright collective as a supplement to the current copyright system, it is also viewed, particularly by its more passionate
While such a system does offer some theoretical advantages, I argue below that the defects of such a system have not been sufficiently examined.
is obviously imperfect, as any system must be, unlikely that a copyright collective would meet even the modest goals of a net positive impact, to say nothing of the claims of virtual perfection
it.
and survive does not mean that it is from an economic perspective. It is the purpose of this chapter to
efficiency characteristics
3.2
It is
important to note that the compulsion aspect of the 'compulsory 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
license' refers to
may
in principle
is
it is
membership essential to receive copyright payments. In that case, distinction without much of a difference.
One analogy
performing
casters.
that
is
often brought
up
is
rights societies
The blanket
40
Developments
in the
economics of copyright
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
board).
is to be contrasted with a compulsory license favorably put forward Lawrence Lessig which, for example, compels the copyright owners of by 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
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.
me illustrate how this works with an historical example. Suppose Jimmy Webb writes a song with incomprehensible lyrics called 'Mac Arthur Park'. He can sell or give the song to the highest bidder (the song was actuLet
hit
sung by deceased actor Richard Harris). If a song appears likely to be a ('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
ally
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
tributed. 4
one half of one cent per minute, per record disthe 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
systems. Further, these new systems are more likely to 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
41
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
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
copyright as
(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 generates most 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 compulsory
license
makes
sense.
Similarly
for
the
compulsory
license
in
records,
the
transactions
surrounding the first recording allows the market to determine and generate 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 system would be to compensate copyright owners for the damage brought about by MP3 downloads. The system as usually envisioned would provide immunity from copyright violations to private individuals engaging in MP3 downloads 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 put forward by Neil Netanel. I refer to this as a quid pro quo system, where
file
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
42
Developments
in the
economics of copyright
(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 copytape
5 right violation. This
restitution.
is
Not Exogenous
systems, a tax and distribution 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-
scheme
is
set
up
to
ditional rights of the copyright owners). The quid pro trast, essentially creates new rights to the traders of
quo system,
files
in con-
MP3
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 MP3 traders if they are discovered violating copyright. The recent crackdown by the RIAA on downloaders is an example of increasing this cost. Not only is
but
MP3 downloaders must bear the risk of being charged with copyright
predict that
if
infringement.
One can
these costs to
MP3
a larger usage of
MP3
files
of CDs.
It is
unclear
how much
of 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 downloading that occurs before the system is in place, although that is exactly
3.3
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 MP3 downloads to be a perfectly acceptable form of behavior. The number of
downloaders
is
43
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, ish scenario that has been suggested.
is
the
more nightmar-
Under
these circumstances, society could well decide that the cost of MP3 downloading under the current copy-
too great - in particular, greater than the costs of switching to some form of copyright collective. This is particularly true given the well-
known
summarize below.
The Imperfection of
System
in
and of them-
selves physical commodities. Instead, these products generally must be put into some physical representation (records, tapes, books, etc.) before they
can 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 and sell copies of the creator's work, and any producer of reproductions who
did not pay the creator could undercut the price that was charged by producers who did pay the creator. This state of affairs would be expected to eliminate from the market any producers who paid the creator, leaving no
revenue 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 expect
build homes, or janitors to clean floors), economic analypredicts that without copyright (and patent) laws, the production of intellectual products would be far below levels that would be considered
sis
8 socially desirable.
home builders to
Even with the existence of intellectual property protection, the quantity of intellectual products created and consumed will be less than the ideal level. This is the unavoidable underproduction and underconsumption of
from intellectual products being 'non9 goods (sometimes called public goods). Non-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.
intellectual properties that results
rival'
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 no
diminution in the possible consumption choices of anyone
else.
Therefore,
44
Developments
in the
economics of copyright
is
an economic requirement for efficient consumption of a non-rivalrous good 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 reproduction cost. Note that the only single markup over the reproduction cost
would be a markup of zero. 10 A markup of 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
that could achieve this requirement
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 individuideal
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. Nonrivalrous goods would be expected to be produced at less than the ideal
level.
surpasses
In an ideal world, every product that has a value to consumers that its cost of production will be produced. For any single intellec-
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
tual work, even a
out of the market. Thus, some intellectual products will not be produced, even where the potential value is greater than the cost of production. Only a perfect price discriminator would be able to appropriate sufficient revenues 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 of non-rivalrous goods because the markets for any particular - books, movies, and so forth - consist of type of non-rivalrous good
in the case
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
individual,
than
it is
to imagine the
exist at all.
These market
inefficiencies are
intellectual
45
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 copying is allowed to consume the product. As explained above, under a copyright regime, successful projects must have prices that deter some consumers from generating surplus because the
creator
must
receive
therefore, will be
below
This logic also argues for a copyright term, if copyright is the chosen parais only long enough to allow the creator to receive sufficient payment to voluntarily provide the work so as to minimize the time during which consumption remains below the ideal level. These consumption inefficiencies in the copyright system can provide ammunition to those arguing for changes. If only we were willing to forgo
digm, that
providing any system of payment to the creators of works, they suggest, that consumption of these works would
Yet few appear willing to argue that copyright owners should not be compensated at all for their efforts. It has sometimes been suggested, therefore,
that alternatives to copyright, such as
The current proposals for copyright collectives suggest that a tax be placed on 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 consumption
in the ancillary
tax.
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 by the 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 revenues 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
taxes, as
the performers making CDs were to be paid through general opposed to taxes on ancillary products, there would still be the
if
46
Developments
in the
economics of copyright
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
It is
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. few simple examples can
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 particular songs. How would an outside analyst determine the proper payment for 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 pay for the CDs that they play. Quite the opposite - there is a history of record
be used to
illustrate cases
producers paying radio stations in order to play their records. Further, radio is a substitute for listening to recorded music, and the recorded music
13 industry would likely benefit if radio could be restricted or eliminated. Although payola is a practice that governments have tried to ban, it is in
between record companies. television and movie studios for product has recently done to have its automobiles and motorplacement, as Mars is thought to have made cycles featured in James Bond movies. one 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 Cups enjoyed the benefits of global publicity from having an immensely popular
Similarly,
BMW
M&M
extraterrestrial enjoying
its
products.
How would any third-party entity know whether the producers of James
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.
47
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 generate 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 automobile was represented as being unreliable, movie studios will need to pay for the automobiles they use and are likely to use fictitious vehicles to escape the wrath of the automobile company so characterized. The point here is to demonstrate the difficulty of determining the most - whether a rudimentary aspect of pricing product receives a positive or negative price. Obviously, if the sign of the price is difficult for an analyst
to determine, the
is
even more
difficult
still.
Since copyright
workings of some for 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).
the subject of this chapter, it is useful to examine the copyright markets. For example, should authors be paid
Academics are willing to pay to be published because their reputations and 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
gratifying to have a
book
to
would like to be published authors. It is show to friends and neighbors. There has
who pay
publishers to print
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 book market (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
a government 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?
solution.
How would
48
Developments
in the
economics of copyright
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.
results that
I
cannot over-emphasize
how
difficult
it is
to emulate the
transactions.
3.5
COLLECTIVE WORK?
The overriding determination that must be made in a copyright collective system is determining the amount of revenue (the royalty) to be raised. The two secondary decisions that must be made are: (1) how the money is going to 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.
(e.g.,
performing rights
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
increasingly more difficult for the CLB to choose the quantum as time goes on since the initial market values will recede into the past.
become
Initially, the market values will be well known since we have been in a period where the market was the mechanism generating values. If, for example, the current system were to be scrapped today, the CLB would have a fairly good idea of what the quantum should be if it wanted to match the revenues of
49
8000-
70006000
-
1
'
c
5000-
400030002000
O> O^ ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON ON
vo r- oo ON ON ON ON ON ON ON ON
Figure 3.2
Album
revenues
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 MP3 downloading, 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 movements 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 changes would be countenanced, approved, or understood. Demographics were largely unchanged. Income changes did not follow this pattern. Prices of
variables that experts look at
substitute products did not suddenly rise or fall. Thus, the usual explanatory 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 importance
of portable devices,
difficulties are at
it have known to double the revenues and 40 say, per cent, or 250 per cent? These very prosaic the heart of the mispricing that is rampant in non-market
how would
CLB
dominated by
50
Developments
in the
economics of copyright
CLB will
and
have to wade through reams of testimony that is often outside the realm of the expertise of
the
members of the
I
not discuss in detail the actual performance of such boards Although in this chapter, there are numerous instances of decisions indicating that these boards, for whatever reason, have reached decisions that are inconwill
sistent
think
it
agreed by
error in
it would be would be an
economic
CLB
to set the
quantum
quantum
constant. This
particularly egregious error if the inflation-adjusted copyrighted goods was rapidly growing at the time. Yet there have been
licenses.
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-
CDs, making audiophiles less likely to participate in MP3 downmarket 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 indicative of an overall increase in demand, however. Thus, the problems discussed in this section would still exist, even though they might be mitigated to some extent.
relative to
loads. If the
Using
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
the quantity of records sold? there also might not be. Even
get accurate statistics
Although there might be such a relationship, there were, we do not know whether we can
on
total
MP3 downloads.
51
Not every download is actually a download. Many downloads are unsucone reason or another. Not all downloaders are Some downloaders have the capability to purchase CDs and others equal. do not. If a large number of five-year-olds are convinced to download a
cessful or aborted for
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.
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 songs than they otherwise would? It would be relatively inexpensive to manipulate MP3 download figures. Additionally, as we shall see below, individual creators have incentives to inflate their downloads, also skewing measured
Finally,
downloads.
As a general statement,
it is
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 MP3
downloading, and not by individuals having no association with MP3 download activity. It is difficult to make a serious case that MP3 downloading 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
Blank
this as a
CDs
backdrop, the choices available are not terribly appealing. 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 MP3 downloaders. 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 CD writers. For this reason, one might argue that a tax on audio equipment would 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
Some
individuals are
more
DVDs
52
Developments
in the
economics of copyright
There
is
more
tinues to exist.
serious problem in the case where the CD market conA tax on audio equipment would be paid not only by MP3
downloaders 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.
is currently enjoying the most support is one 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?
to tax ISPs.
Although
with the
this is
an interesting rhetorical
two problems
$5 per month figure is based on the current estimate of losses suffered by the record industry, and those losses are likely to
logic. First, the
increase
when everyone gets the green light to stop paying for CDs.
in the order of
A more
$20 per month. Second, plenty of individuals would balk at paying these amounts. Many individuals using the Internet do not buy CDs and see no reason they should subsidize CD consumers.
money among One possibility is that a single copyright collective would be formed, covall
to be divided
copyright owners, and that the collective would be charged with distributing the revenues. Even if this were to be the case, however, the colering
lective
its
would still have to decide how the revenues were distributed among members, and it would face the same decisions that I am going to ascribe
in the following paragraphs. Further, when there are different involved in the creation of a product (musicians, composers, recordparties
to the
CLB
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
example, the Copyright Arbitration Review Panel (CARP) determines the relative amounts received by the sports claimants, the
license, for
programming (news and information), Hollywood producers and others. In other words, the Board determines the revenues going
creators of local
to 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.
likely
Regardless of how the various parties are aligned, how would revenues be distributed? The suggestion that has most often been made is to
53
MP3
downloads as a basis
ologies could be created to measure downloads, this view holds, could award the dollars based upon these measurements.
methodand we
The comparison
is
often
sell
made
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
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.
lation, as
One fly in this ointment is that MP3 downloads are amenable to manipuI have mentioned when discussing the determination of the
It is
quantum.
possible for
an individual
artist to
massive numbers of
tributed revenues.
files in
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
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
address, but
pretext for creating a copyright collective system to begin with. Gaming the system in this manner is not a problem under current
com-
these licenses
are placed are purchased in the market (for example, in the case of performing 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-
programs containing their music. A second problem in using download statistics is that downloads may not match up well with purchases. Bands could convince their fans to download songs for the sake of downloading. It is likely easier to get young
tions to play
14
fans with
much
time but
little
songs to help out their favorite band than it might be to get them to buy more 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
sites, I
am
not sure
how
54
Developments
in the
economics of copyright
shrouded in
3.6
When
frequently
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 industry 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
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
and
front-end market arrangements that would serve to limit the
damage done
performing rights royalties which are not available in the case of a replacement for the record industry. I will take each of these points
by
inefficient
in turn.
Efficiently Priced?
Can we
take the history of blanket license rates as a harbinger that a copywork well as a replacement for the record industry?
rights efficiently priced? The answer is that we have no 15 earthly idea. There are several markets where blanket licenses are used. In the case of
Are performing
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 by the 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
television
55
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 tribunals, one finds no clear method to tell whether rates are too high or too
low.
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
payments will change as the industry grows or declines. This assures some modest linkage between them and is likely to keep the royalty payments from getting too far out of line. 16
royalty
If the current
lective,
is
be impossible to tie royalty payments to some industry 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.
however,
will
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
For sound recordings, the compulsory license only covers addisong 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
3 per cent.
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
set in
markets in the
first place.
In
many of
these
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
the value of the right in the broadcast of a single program. Imagine the
56
Developments
in the
economics of copyright
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,
appears to be far better for everyone involved to have something like a blanket license.
it
In principle, you could have market negotiations between ASCAP/BMI and broadcasters to set the blanket tariff rate, but a concern that the performing 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 requirement that either party could appeal to the appropriate regulatory body to have the rates set.
It is
is
set
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 17 is created since they are not the consumer of the music. Further, MP3s substitute for the purchase of a record, and the suggestion 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.
is
No Safety Valve
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
There
made
is a back-end payment that is payments that are most often paid to composers when their compositions are commissioned for a television program.
some extent, for errors in the royalty rate. The television blanket license, for example,
in addition to front-end
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
57
will largely
copyright collective system that provided the revenues for the entire recording industry enterprise has no 'front-end' that might be used to compensate for an error in the royalty rate. If a mistake is made in this market,
is
very
likely,
there
is
no
3.7
CONCLUSIONS
difficulty
The current
with enforcing copyright in the face of MP3 downloads 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
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
choice of rules.
revenues are the very questions that any economic system answers by its 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 and control methods.
lective will
command
There are good reasons to believe that mispricing by the copyright colbe a more serious problem for the record industry than it might have been for other markets that have adopted some form of copyright collective.
collective to limit
Before
we throw out
we need
to investigate
more
from
carefully the arguments that are used to support a movement away 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 how onerous enhanced enforcement of current copyright laws will turn out to be, or whether such enforcement can feasibly protect the industry. Only as
58
Developments
in the
economics of copyright
we
NOTES
An earlier version of this article was first published by the Progress and Freedom Foundation
on IPcentral http://www.ipcentral.org/review/ vln2intro.html.
1
.
victory in the Napster case, the industry hit a roadblock in its attempt to shut Grokster, Streamcast and Kazaa. judge has ruled that those non-centralized filesharing systems were little different than VCRs and thus were not
its
After
2.
3.
behavior of their customers. See for example, http://www.pewinternet.org/PPF/r/124/report_display.asp. 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
4.
5.
6.
7.
8.
siteGrammy.com: http://www.grammy.com/features/2003/0725_complicenses.html. There are some other notification requirements that I ignore here. Also, the price is actually the higher of eight cents per song of five minutes or one-and-a-half cents per minute. The 1992 Act also required that manufacturers of DAT machines restrict the machine from being able to make copies of copies. The DAT technology never was very successful in the US whereas CD burners have become extremely common. For evidence on the impact of MP3 downloads on the sales revenues in the recording industry through 2002 see Liebowitz (2004a). Estimates suggest that 40 million Americans have downloaded at least one copyrighted work. See 'Downloads to save music biz' by Jane Weaver, MSNBC, 12 August, 2002. 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 optimal duration of the copyright is less than the time advantage that one gets from being first.
See Plant (1934). 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
difficulty
9.
includes 'non-excludability' (the inability to prevent users from consuming the good). The with the latter definition is that non-excludability is a function of the law and the
the subject of this chapter. The two attributes are not necessarily related and I have always thought it was counterproductive to treat non-excludability as an inherent attribute of intellectual products, since it is not.
10.
I
am
in the
different valuations for the product, with these valuations arrayed down to zero. Any positive price then excludes some potential users. If multiple prices are allowed, however,
1 1
12.
then charging each consumer just slightly below what he is willing to pay can achieve optimal results. This is known as perfect price discrimination. 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 government intervening in the market. Do not confuse the claim that markets do not produce the ideal output with the claim
13.
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). In a recent paper, I examined two natural experiments regarding the impact of radio on record sales - the introduction of radio in the US and the introduction of commercial radio 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
that markets
59
14.
5.
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). 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 payments. 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 combination 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. Some percentage of bandwidth will be devoted to music downloads, and ISPs wouldn't
17.
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 customers 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 of Law and Economics, 22, 269-328. Demsetz, Harold (1969), 'Information and efficiency: Another viewpoint', Journal
12, 1-22.
W.W.
Entertainment', draft. Ku, Raymond (2002), 'The creative destruction of copyright: Napster and the new economics of digital technology', University of Chicago Law Review, 69, 263-324. Lessig, Lawrence (2001), The Future of Ideas, New York: Vintage Books.
Liebowitz, Stan
in
J.
(ed.), Copyright: Administrative Institutions, Copyright Administrative 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, and Economic 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:
Ysolde Gendreau
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 May, 167-95.
in books',
Economica,
4.
consumers and
artists
INTRODUCTION
of original
voices in the music industry claim that Internet piracy has reduced CDs and that illegal MP3 files have become a substitute to
Many
sales
legal
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 purchases,
sales. The goal of this chapter is which could eventually boost to analyse the effect of increasing copyright protection on the pricing and protection strategies of the firms and its consequence on consumer surplus.
CD
4.2
Digital
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
61
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'
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
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
also in uploading
In addition, users can transfer songs from an original format and upload them in a similar way. digital
1
user).
CD
into
Most P2P technologies include a priority rating system that provides information about the material shared by users. For instance, the Kazaa
is a measure of how many megabytes have been uploaded how many megabytes have been downloaded over a given compared of time. It is clear that such a system benefits users who share large period
priority rating
to
popular
such as recent theatrical and pornographic movies. with P2P, the Internet also gave birth to audio streaming. Audio Along can take two forms. On the one hand, Internet radio stations streaming
files
sites independent from major technological distribution combroadcast licensed music. On the other hand, specific streaming panies owned by large software and content providers (such as technologies
owned by
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
2 by the Royalty Panel (CARP).
with other non-digital components such as a printed booklet (with lyrics, case for music CDs. pictures, song and artist information, etc.) and a Recent digital technological protection tools that we will discuss later on
CD
can also make digital music files close to private consumption goods. Finally, music is a typical experience good. When facing new music, consumers like to spend time getting recommendations from music magazines, listening to music on the radio or in a record store before making their
62
Developments
in the
economics of copyright
purchase decision. The traditional way for record companies to provide information to consumers on the existence and style of new CDs and
spend large costs on advertising and promotion. However, Napster experience, it has become clear that there is a cheaper way for consumers to obtain this information: by searching, downloading and testing digital music files made available through P2P or other fileartists is to
after the
swapping technologies. Acquiring information is costly and time consumdownloaded files could differ from the one users expected, mainly because 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 opporing:
dents.
tunity cost of spending time online, i.e. mainly teenagers and college stuTo sum up, we consider traditional marketing and promotions as
an information-push technology, and new filesharing networks (and to a where 3 consumers, not firms, spend time and resources.
lesser extent audio-streaming), as information-pull technologies,
The Law
at protecting authors of 'original works of authorship', including literary, dramatic, musical and artistic works. The protection is available for published and unpublished works.
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 amount of the work used; the effect of the use on the market for or value of the
original work. Unlike traditional
means of copying,
is
copy can be downloaded by any user across the world. So far, public new form of end-user piracy by strengthening existing copyright laws, as illustrated by the 1 998 US Digital Millennium 4 Copyright Act (DMCA).
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. Following
filed by the Record Industry Association of America (RIAA), was found guilty of copyright infringement. The court found that Napster
a lawsuit
Peer-to-peer, piracy
and
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
argument according to which sampling leads to higher purchases was supported by an empirical study, but ruled out as flawed and non-objective by the court. After monitoring activities over
distribution. Napster's
CD
new
lawsuits charging music uploaders with copyright infringein September 2003. These legal actions are also based on the arguthat music
5
industry.
downloads are causing substantial damage to the music Under the current legislation, any user of P2P risks legal attacks.
legal protection against Internet piracy increases 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
The
DMCA
piece of software that can detect, monitor and block (unauthorized) use of copyrighted material. There are two types of existing technologies:
DRM
watermarking
and
finger-
printing (converts the files content into a unique identification number). Despite a promising future, companies that create such technological protection tools struggle to survive. 6
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
Samuelson (2003) argues the goes beyond the Copyright Act. can protect any digital content even if it is not protected by Indeed, the 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, sometimes stands for 'digital restrictions management'. Moreover, it can potentially protect over an infinite amount of time, which is contrary to
DRM standards. Finally, DRM software threatens the privacy of users. DRM DRM
DRM
DRM
DMCA
is
that prohibits users bypassing technological measures of protection. Many and its people involved in the copyright law have argued that
DMCA
corollary
DRM
fail
64
Developments
in the
economics of copyright
in addition,
impede
artistic
and
scientific progress.
allows some form of Liebowitz (2002) challenges that view. Because first-price discrimination, he argues, the use will not be reduced (however,
his analysis fails to notice that transformative
DRM
and
authorized under fair use could be greatly reduced by DRM). Moreover, first-price discrimination leaves no surplus to consumers and is questionable from a social point of view if the 'intensive' music listener is a lowincome individual who would be charged more than, say, a wealthy individual who only listens to music occasionally. Liebowitz (2002) also
on previous work
that
DRM will
argued by Landes and Posner 1989). Here he argues not hinder such innovation because copyright only protects
(as
the expression of ideas and not the ideas themselves. In addition to these arguments, we would like to discuss three other
can be important economic dimensions of DRM. First, because 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 compatible cannot be neglected. For that reason, some industry observers say that
DRM
DRM DRM
software and
stands for 'down-right messy'. Second, technological protection its inventors create an additional layer between artists and
inefficiency. Third, if
and software, this could raise the barrier of entry in the distribution and the software protection markets and increase market power.
legally
mandated
in players
4.3
Despite the technological breakthrough brought by filesharing technologies, 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
ial,
the quantity of copies likely to result from the purchase of original materhe or she can indirectly appropriate revenues by charging a higher price
for the original (see Liebowitz 1985;
et al.
1999).
and
the
related to the pricing of a club membership the nature of the cost to copy. The second argument is related to the
first
The
argument
is
literature
demand of
on bundling and how club formations can reduce the variance of the club as a whole compared to individual demands.
Peer-to-peer, piracy
65
Potentially, indirect appropriation could arise if users of filesharing technologies 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 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 noneconomic 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 consumers' 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 we have 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 copies can benefit copyright owners as shown by Conner and Rumelt (1991), Takeyama (1994) and Shy and Thisse (1999). There is a case for weak network 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 on both 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
the
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
66
Developments
loss of
in the
economics of copyright
and the
computer
resources.
However,
if
there
is
only a small
users sharing a large number of files, the extent of network effects will be limited (for a discussion see Peitz and Waelbroeck, 2004b).
number of
can provide information on the value of original an experience good. Takeyama (2003) shows how copies with information on the characteristics of a durable good can solve adverse selection problems. Peitz and Waelbroeck (2003b) show how a multi-product firm can benefit from better matching consumers to
Finally, digital copies
due to better sampling on P2P networks, despite the negative competition effect due to the availability of digital copies. We believe 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.
their ideal products
4.4
We propose a simple model of the music industry to analyse welfare implications of an increase in copyright protection against end-user piracy when firms can implement technological protection. The model includes three
original
First,
and relevant
features.
protection (set by public authorities) and technological protection (implemented by copyright owners). Copyright owners can implement costly technological protection (DRM) on original material in order to prevent
However, 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
their duplication (as
files).
MP3
also required. Legal protection consists of legal proceedings against copiers who must internalize the risk of being caught and the resulting penalty. 8
of copyright laws
is
Second, music is an experience good. In addition to the traditional way of 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
lieu
consumers bear the cost of acquiring information in of record companies. 9 In short, we view traditional distribution as
an information-push technology in which the firm pays to provide information 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
Peer-to-peer, piracy
67
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
file
available
Some
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
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 protection gives record companies using the information-push technology
incentives 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 producing
small-audience music.
4.5
We
analyze the role of an increase in copyright protection when only two 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-
sell
CD box,
original versions of their records (with 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
consumers. On the other hand, the information-pull (P2P) allows consumers to search and test digital copies that we technology
product to
all
68
Developments
in the
economics of copyright
assimilate to
tures).
downgraded versions of originals (with a limited set of feaThus 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 10 cost of the original are normalized to zero. and legal protection. Regardless of the distribution techTechnological
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
difficult to create
of originals in order to make digital copies of his or her product more and to find online. 11 Technological protection, a, can be
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
DMCA.
being caught downloading unauthorized digital copies. Music experience. Consumers incur an opportunity cost s of spending time searching and downloading digital music files using the P2P technology.
We
1
.
mass
but
assume that s is uniformly distributed on the segment [0, s\ of That opportunity cost not only reflects a direct cost of spending
time on the Internet (related to the user's level of Internet sophistication), 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 the function
utility
what they have listened to obtain a utility represented by v(l a), whereas buyers of the original product obtain a for all 0<a<l. The differential between copies and v>v(l a)
like
of the original over the copy, such as 12 lyrics, booklet, pictures, song information, and so on. Consumers who do not like the music get utility, regardless of whether
originals, av, represents the value
they learned the existence and the characteristics of the music through advertising or through a digital copy, and regardless of the quality differential
latter
between copies and originals. The implicit assumption behind the requirement is that consumers do not care about additional features of music they do not like.
Peer-to-peer, piracy
and
69
To sum up, the total expected cost of copying is s + 9. It is worth noting that this cost could have taken the form of 5(1 + <p) (as in Novos and
However, we do not consider that copiers with a higher more from legal protection than others. downloading Thus, we assume that legal protection creates an additional burden that is
Waldman,
1984).
Consumers' expected
utility is
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 13 marketing and advertising cost K. We assume that advertisement
all
fixed
reaches
consumers who are thus aware of the existence and the char-
of the product. Hence, consumers take their purchasing decisions in a perfect information environment. Consumers who like the
acteristics
music
(in
graded version.
the utility of
s
of purchasing the original version is w(l,0) = v-/?, and downloading a digital copy is w(0,l) = v(l -a) -s- 9. Let = p-q>-av denote the opportunity cost of the indifferent consumer
The
utility
between the original and the copy. Consumers purchase the original provides a higher utility than copying or not purchasing at all:
if it
If the first inequality holds, consumers who like the product (in proportion p) purchase it if their opportunity cost of using the P2P technology is such that s s. This is illustrated in Figure 4. 1
>
very low, all consumers who like the product (in pro< 0. Similarly, when the price portion p) purchase it. This occurs when s
the price
is
When
Figure
4. 1
70
Developments
in the
economics of copyright
is
>
s or/?
>
v).
d(p, a)
p^^ S
otherwise), and maximizes its profit function to set the optimal level of copyright protection a and the optimal price/?:
if
Max IT = pp
<*./>
s -K c(a)
s.t.p
<
v.
The optimal
price/?*
that
and
z,s
z*
As a
result,
a*
is
increasing in
<p:
da* -
>
O.
14
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. 15 Proposition 4.1 describes the impact of a higher legal protection on a firm's profit and consumers' surplus:
Proposition 4.1 Increasing legal protection Increases the firm's optimal profit IT*; (1)
(2)
cp:
Proof
(1)
Peer-to-peer, piracy
71
(2)
First,
consumers who
still
after
an increase
Indeed, (0, 1) = v(l -a) -5 -9. Since > 0, v(l -a) decreases as 9 increases, so that w(0, 1) also da*/d(p decreases. Second, the utility of consumers who still purchase the
in
9 have a lower
utility.
Third, consumers
product after an increase in 9 decreases because of the higher price. who switch from the copy to the original have a lower utility. Previously they had a higher utility of copying. After
Proposition 4. 1 states that a reinforcement of legal protection (as a consequence of the Digital Millennium Copyright Act, for instance) makes
firms increase technological protection. Firms who distribute their products 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
its
nology. We assume that consumers are not initially informed about the characteristics of the product, but they can acquire information by downloading digital copies. Consumers who have copied the product and
= 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 purw(0, l)
enjoyed the music have the possibility to directly purchase the original from the firm. The expected utility of downloading the digital copy is
utility of purchasing the original product would be-/?-s-<p<0. On the other hand, consumers who have copied the product (of mass pv(l - a) - 9 /s) and who liked it (in pro- a) + va - - 17 portion p) have an expected utility w(l, 1) = p(v(l -p) s 9. It is clear that copiers who like the music will purchase the original provided
who like what they listened to have an extra of utility av-p purchasing the original instead of just copying. This is illustrated in Figure 4.2.
that /?< av. Indeed, copiers
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:
72
Developments
in the
economics of copyright
Figure 4.2
if
tion program:
Max TT =
a,P
-5
c(a)
a).
s.t.p<av
Here the
potential
and
<p
<
pv(l
level
of technological protection, a,
18
cp.
copyright protection
demand
when copyright
buyers need to download a digital copy first and this activity is more costly 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
(1)
(2) (3)
An increase in copyright protection: Decreases the profits of firms distributing their products using P2P; Decreases the social loss due to the downloading cost;
Decreases buyers' surplus.
Proof
(1)
~T~ fl(p
(2)
ova ~--'
5
S^
*
acp
Peer-to-peer, piracy
and
73
ds
da
v/5
Substituting the expressions found in note 19 and rearranging reinforcement of the copyright law from shows that ds/dy < 0.
(3)
- .?(<P copying without enjoying the product (1 pX-stoj) 2 )]. The surplus of copiers who become purchasers is given by
cp 2 (with <p 2 >(pj) makes a proportion sfo^ s(<p 2 ) stop copycpj There is a reduction of the number of copiers, which means that ing. there is a smaller proportion of copiers who waste resources on
to
/ 6
= <;(pv[l-a'
o
After rearranging,
we obtain
. ;-
that
- <x*(p)] -
pv
The expression
ing in
<p.
in
between brackets
is
negative.
Thus
Ub is decreas-
intuition behind Proposition 4.2 is that an increase in legal prodemand for copies and thus the potential demand for originals. Since the firm extracts all buyers' surplus, consumers only obtain
The
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 consumers' surplus. The reason is that increasing 9 reduces the number of
loading music they do not like and the number of copiers buyers of the original product.
74
Developments
in the
economics of copyright
In this section
we assume
that C(OL)
its
= ca 2/2.
When
ogy,
it
a price
+
cs(y
+ s)
v p
2
2cs
and a
level
of technological protection
2cs
v p
(the
denominator
is
positive
yields a
profit of
cp(9
1
+ s) 2
2v 2 p
4cs
using the information-push technology. When the firm uses the information-pull technology, p* = a*v and a level of technological protection
^
it
sets
a price
vp(vp
9)
It is
is
~ <P) 2
2
2s(cs
2v 2 p 2)<
In Figures 4.3 and 4.4, we compare the profits using both distribution tech= 0.15 and <p = 0.30). The nologies for two different legal environments (<p = 1 19 The flattest curve = = 1 v 1 other parameters are set to s 0.45, c
, ,
K=
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 the potential audience of the music, p.
lies in
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.
Peer-to-peer, piracy
75
Information-push
0.2Information-pull
0.1
-
0.2
0.4S
0.6
0.8
-0.1
-0.2Potential audience
Figure
4.
Application with
cp
= 0. 75
0.6i
0.5
-
Information-push
0.4-
0.20.1
-
Information-pull
-0.1
0.6
0.8
Potential audience
Figure 4.4
Application with
q>
= 0.30
with
For instance,
itable.
0.8<p<0.9
pull technology
although the traditional distribution technology was profFinally, firms producing music with a large audience prefer the
<p
information-push technology.
4.4). First, the
from 0.15 to 0.3 has several noticeable effects (see Figure range of p that supports the information-push distribution technology increases. Second, more firms prefer the information-push technology. Third, the minimum audience which makes it profitable to enter
Increasing
the market with the information-pull technology increases too. Thus there are fewer firms in the industry and a more important proportion chooses
76
Developments
in the
economics of copyright
the information-push technology. The aggregate effect of an increase in copyright protection at the industry level depends on the distribution of
is a high proportion of superstars, an increase in leads to an increase of industry profits. However, copyright protection when there are proportionally more small-audience artists, industry profits
audience
p.
When
there
might
fall.
4.6
We
entry through cheap distribution channels, and open the market to new artists 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 producers could use legal protection as a foreclosure tool against new products.
However it is foreseeable that as a growing number of new products becomes available, consumers will ignore the existence of one particular product in which they could have a potential interest. We believe that new distribution technologies could open the market for informational intermediaries. Indeed, in our model, both distribution technologies are inefficient at transmitting relevant information. On the one hand, produc-
who use information-push technologies waste the fixed cost of advertisement and promotion on reaching consumers with no interest in their music. On the other hand, some P2P users waste resources on downloaders
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), targeting only consumers who have the highest likelihood of purchasing the original. Some of the new emerging business models are based on crossplatforms, like Kazaa for instance, who work together with record companies as a new vehicle to reach mass markets. As P2P plays the part of an advertiser, record companies are becoming simple distributors. MP3.com
has been using their data on consumers to provide record companies information 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 containing lyrics, pictures, or an access code to online chat rooms, forums, makingoffs, additional music and a discount on live performances, etc. Indeed, the possibility of offering an Internet connection with login and password to a
with additional features for purchasers of original products reduces the value of a digital copy. In fact, the emerging new business models are
site
Peer-to-peer, piracy
77
ketplace
beginning to take into account other assets than just the music. The maris 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
wish to thank B. Caillaud, G. Llobet, P. Pereira, J. Pouyet, M. Bourreau, and session 2nd 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.
participants of the
We
ZEW
NOTES
1
.
2.
For instance, eMule automatically shares files that are being downloaded. 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 roycopyright owners. 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-INSIGHT reports). Moreover, a survey of IPSOS of 2002 finds that 73 per cent of music downloaders '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 changed the genre of music they typically listened to, mainly because they were able to experiment with different forms of music. See Peitz and Waelbroeck (2004b). We believe that these numbers understate the potential of file-sharing systems for sampling purposes because cross-recommendations and profiles of downloaders could be greatly improved. The makes it a crime to circumvent anti-piracy measures built into most commercial software (except for research purposes, non-profit libraries, etc.). While the limits liability of copyright infringement of Internet service providers and institutions of higher education, it requires webcasters and commercial broadcasters to pay licensee fees. Finally, the does not affect conditions of copyright infringements, including fair use. The lays the legal foundation of pay per use even for material that is no longer protected by copyright and so on a perpetual basis. This argument is only partly validated by cross-country data analyzed by Peitz and Waelbroeck (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 playing
alties to
3.
4.
DMCA
DMCA
DMCA DMCA
5.
6.
7.
an important role. MIT Technology Review (March 2002) comments on the struggles of ContentGuard and TrustTechnologies, two important producers of solutions. alone is not sufficient per se to secure digital content, and has to be coupled with software or hardware solutions. Eckersley (2003), citing sources from IBM's Thomas J. Watson Research 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
DRM
DRM
78
Developments
in the
economics of copyright
8.
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. Our model also applies to audio streaming where we analyze the effect of increasing royalties
less
less
consumer's
surplus.
9.
its 'New Artist program', which enabled starting bands to set up a 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). 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
10.
negligible
11.
In another
compared to the fixed marketing cost. more elaborated version, we allow technological protection
to decrease the
12.
value of the original as well, through a reduction of its fair use value. The results presented in this simpler version are qualitatively similar. 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. For simplicity, we do not make a distinction between the artist and his label. The implicit
3.
14.
assumption is that it is possible to write a contract that eliminates Applying the implicit functions theorem to
vertical inefficiencies.
we
get
~
da _
pv/2s
is
'**
d<f
2\
d(f>
dcp
2\
d<p
16.
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
17.
and promotion K have not been spent. In other words, products which 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. 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. At the optimum, the effect of increasing a on profit
is
8.
given by
F(a,<p)
= -_ [pv(pv(l - a) - <p)] -
2 2 P av
c'(a)
pv/s
d(f
dF/dot
Peer-to-peer, piracy
79
is
positive,
it
negative.
We
= _ 2v2
da
19.
- c"(a) <
0.
Thus
dq>
<
0.
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 of Law and Economics, 42 (1), 1 17-55.
(1989), 'Private copying, appropriability, and optimal Journal of Law 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,'
Besen,
S.
copying
working paper.
Landes, W. and R. Posner (1989), 'An economic analysis of copyright of Legal Studies, 18, 325-63.
law,'
Journal
Liebowitz, S. (1985), 'Copying and indirect appropriability: Photocopying of journals,' Journal of Political Economy, 93, 945-57.
Liebowitz,
S.
Networked Economy,
Novos,
An
Peitz,
(1984), 'The effects of increased copyright protection: analytic approach,' Journal of Political Economy, 92, 236-46.
I.
and M. Waldman
P.
M. and
M. and
of the economics
Peitz,
P.
Waelbroeck (2003a), 'Piracy of digital products: A critical review literature,' CESIfo working paper #1071. Waelbroeck (2003b), 'Making use of file-sharing in music distribusales:
1 (2),
mimeo. Peitz, M. and P. Waelbroeck (2004a), 'The effect of Internet piracy on music Cross-section evidence,' Review of Economic Research on Copyright Issues,
tion,'
71-9.
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,' Communications of the ACM, 46 (4), 41-5. Takeyama, L.N. (1994), The welfare implications of unauthorized reproduction of
Peitz,
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: and Northampton, Developments in Research and Analysis, Cheltenham,
intellectual property in the presence
UK
5.
Brennan
5.1
INTRODUCTION
framed as between providing more
access for the public to created works and giving profits to those who created the works. To this observer of copyright policy discussions, grant-
ing any weight to the latter side of this tradeoff often seems to come grudgingly. It is as if only a minimal degree of fairness justifies consideration of
creators in the balance at
all.
between access rights and exclusion rights take place on many intellectual property fronts. Seemingly innocuous or obvious business practices have been patented, leading to concerns that they will be
battles
These
and promote collusion (Ciminello 2000). Cases involving MP3 music filesharing and DVD decoding cases suggest a conflict between listeners and producers regarding access to music and films and on broadband deployment. 2 The DVD cases also reflect controversies regarding the Digital Millennium Copyright Act's restrictions on access to decryption technology. 3 The Supreme Court ruled against contentions 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 4 Right to their respective Writings and Discoveries.' One facet of these controversies is whether viewing copyrighted works
unnecessarily monopolized
l
as 'cultural' goods, should change copyright doctrines, par'fair use.' Fair use is defined statutorily as the right to copy or use copyrighted works in ways that would otherwise lead to liability for infringement, based on four factors: purpose and character of
e.g.,
as special,
ticularly to
expand
amount of it used, and effect on its market The operational content of these factors continues to evolve. 6 Policy goals other than economic efficiency, e.g., maximizing access (Brown 7 1985), '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
the use, nature of the work,
or value. 5
80
'
81
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 8 grounds' associated with schools, libraries, and the courts.
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 valuations. We look at fair use when there may be 'diversion' in that some uses,
which users could pay and could be identified, may nonetheless be 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 9 allowing uses for which there is low willingness to pay. The motivating question is whether different policy objectives should
for
obtained for
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, arguments about 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
Still,
narrowly
structured, inherently unrealistic models may yet offer insight into how a 10 change in objective may lead to a change in policy. 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
Formal models of the design of copyright have focused primarily on the extent to which copying should be made more or less costly (Novos
82
Developments
in the
economics of copyright
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 consumers 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. Brennan (1986) takes a more narrowly economic approach, treating fair use as a compulsory license where
1 1
the fee
is
zero, building
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 important limitation of Klein, et al.'s insight, however, is that it ignores the other three
12 statutory factors, beside effect on market value, associated with fair use. holders and user could whether the nature and Copyright groups dispute
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
holding
more
strategic approach.
Depoorter and
can
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
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
available.
14
in a multiperiod setting.
commit not
tion prices.
to offer a
in later periods to
amount
fall, the rationale for fair use defenses against accusations 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 might not have been worth identifying in the analog environment (National Research Council 2000, p. 136). Advocates of fair use view such develop-
works below what users have come to and research settings (National Council for Higher Education, 2003). These discussions only highlight that there might be 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
ments as reducing
free access to
83
of those works to help equalize educational opportunities for models below do not incorporate such distributional factors
all.
The
directly,
although some
fair
use settings
we do examine,
welfare, or
5.3
We
look at copyrighted works as a series of monopolies. Each has separindependent demands, created at a fixed cost independent of actual use. As more works are supplied, the 'marginal' work is valued less (has a
ate,
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
Y. 15
We
mentarity in consumption or production, e.g., that availability of work reduces the cost of proincreases demand for Y, or that access to work
ducing work
16
and
As
the point
to suggest the relative position of optimal amounts of allowed (fair) use under different policy goals rather than to estimate specific quantitative fair use standards or to look at changes in external con-
of the models
ditions
(e.g.,
new
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
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 copyprinciple,
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
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.
84
Developments
in the
economics of copyright
5.4
We
that
eter n.
work
at price/? declines,
i.e.,
Z)/J
that
,
constant across
for each
all
demand
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),
where
(5.1)
pDp (p,n)+D(p,n)=Q. 19
(5.2)
let
Let /?*() be the profit-maximizing price for work n as defined by (5.2), and = n*Cz) II(/?*(),) be the maximum profit received by the copyright
n.
20
From (5.1),
11*' CO
(5.2)
use
is
Dn <Q. The number of works n* with no determined where the profits just cover the cost of production,
= c.
Use/Number of Works Tradeoff
fair
i.e.,
where n*(*)
The
Fair
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
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
the
is
maximizes
profits will
be determined by
(5.2), inde-
pendent of
Thus, the extent of fair use has no effect on prices of copyworks that are sold. The amount of uses of work n with fair use k, righted
USE(n,k),is
USE(n,k)
(1
-k)D(p*(n\n) +
fc/>(0,/i).
(5.4)
85
At
fair
use intensity k, works will be created up to the point where profits 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
is
fraction is negative because demand is positive and the denominator negative by assumption. Equation (5.6) not only defines N'(k) but provides the basis for the assumption in the derivations below that N'(k)<Q,
that fair use reduces the number of works created. The magnitudes of N' and D n are inversely related. If reducing the number of works just a little will bit would show a large increase in demand for the marginal work, only fall a small amount for a given increase in the degree of fair use.
i.e.,
Figure
work n. The key for any case will be to recognize tradeoffs. Increasing k may
cl(\-k)D(p*(n),n)
(\-k)D(p*(n\n)
(1
-)D(0,
n)
n)
Uses of work n
Figure
5. 1
Consumer surplus,
profits,
'equal diversion
86
Developments
in the
economics of copyright
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
is
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 1 and it. Were the supply of work n guaranteed, we would want to set k
,
set
Area 2 to the
entire
now
look at setting
of
We can
Number
Works
Let k N 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
From
with
fair
Figure 5.1, the contribution to net economic welfare from work n use k is the sum of Areas 1, 2, and 3. To reduce some of the clutter
n) of
work n
CS(p,n)="JD(p,n)dp.
p
Area
is (1
c.
k)CS(p*(n\n)\ Area 2 is &CS(0,); Area 3 is p*(n)(\ -k) Aggregated over all N(k) works supplied, net economic
is
N(k)
NEW(k) =
I
o
[(1
-k)CS(p*(n\n) + kCS(0,n)
+ /?*()(! -k)D(p*(n\ri)]dn-cN(k).
Recalling from (5.5) that the profits of the marginal work just cover its creation costs, the degree of fair use, k NEW that maximizes is that which satisfies the first-order condition
,
NEW
J/ -\r
ri TT/- \ ;
^(K)
f o
dk
[CS(0,/i)
CS(p*(n),n)
P *(n)D(p*(n\n)}dn
0.
(5.7)
'
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 consumers 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,
I
IN(k)p*(n)(l
- k)D(p*(n),n)dn -
cN(k)
cost,
^
The term on
increasing in k.
,
/}WW ^^.
is
(5-8)
on
profits
to inframarginal works. This was subtracted out is maximized, nomic welfare. At k NElv, where
NEW
ACS
will thus
be
Accordingly and not surprisingly, the level of fair use that maximizes consumer surplus, kACS will be greater than k NEW The lost profits to copyright work holders would not matter. The difference between k ACS and
,
.
k NEW 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 aggregate consumer surplus is the objective, fair use should be stronger the
higher are prices for copyrighted works.
Gross Benefit
and 4
in Figure 5.1,
88
Developments
is
in the
economics of copyright
GB(k)
defined by adding costs back into the expression for net economic
i.e.,
welfare,
GB(k)
Hence,
= NEW(k) + cN(k).
d(GB) dk
d(NEW)
dk cN'(k\
(5.9)
The first-order condition for gross benefit is given by (5.7) for NEW, adding back cN', which is negative. At kNEW the level of fair use that maximizes
,
d(GB)/dk<0, implying that the maximizes gross benefits, k GB is less than kNEW
,
.
level
of
fair
use that
This
is less
surprising than
it
may
initially
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
Fifth,
of Uses
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
other 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
o
[(1
k)D(p*(n),n)
+ kD(Q,n)]dn
and
cKfJSE)
m
o
dk
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 extending fair use reduces the number of works. No general comparison between k USE the degree of fair use that maximizes access, and other values of fair use found above is apparent.
,
89
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
o
[CS(0,/i)
CS(p*(n),n)]dn
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 contribution at k USE suggesting that
,
CS
and hence
fair use.
that
i.e.,
is plausible. Increasing the amount of fair use will tend to uses of increase inframarginal works that are relatively highly valuable and creation of works that are valued less. If these relative values are discourage
This result
ignored and
all
justification for
uses counted equally, this difference will not matter making fair use stronger will not apply.
and this
Summary
Recalling that the optimal fair use regimes are kN for number of works, k NEW for net economic welfare, k cs for consumer surplus, k GB for gross value of uses, and k USE for number of uses,
k ACS > k
mw >k GB >kN =
first
The degree of
inequality depends on the prices at which copyrighted works are sold. The degree of the second and third inequalities depends on the sensitivity of the supply of copyrighted works to price. In
the
be able to come up with specifications for this model in which k NEW >k USE 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
addition, one
may
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.
90
Developments
in the
economics of copyright
5.5
We now model
Gordon and
eter
r,
use
more
in line
where
enforced,
i.e.,
work valued at a reservation price below r are assume throughout this section that this restriction is that there is no diversion of higher valued uses into the fair
uses of a
We
use regime.
is the copyrighted work for which profits where p*(n*) = c/D(p*(n*),n*). Obtaining results in this setting requires the further assumption that as n increases, /?*() falls, i.e., 22 < p*' < O. If/?*' < 0, fair use will not discourage any creation of works for r that the For all holder >r, implying copyright /?*(*). w<*,/?*(w) >/?*(*) 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
As
in the earlier
model, n*
i.e.,
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,
n: profit-
91
D(r, n)
Z>(0, n)
Uses of work n
Figure 5.3
Consumer
surplus, profits,
n: profit-
Area 2 represents the consumer benefits from fair use, to those with reservation prices below r obtaining the work for free. Area 5 represents consumers unwilling to pay p*(n) but willing to pay more 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 below p*(ri). The second case is where p*(n) ^ r. There, the copyright holder of work
n
is
value the
unable to charge the nominal profit maximizing price, since users who 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
right holder;
under the fair use policy. Area 3 is the profit to the copyArea 4 is the cost of creating this work.
92
Developments
in the
economics of copyright
will
be created. The
fair
9
is r
use level that maximizes output, indicated again N <r*. To avoid discouraging uses that would leave
output and prices unaffected, rN =r*. By any policy criterion, the optimal be no less than r*.
with results as seen in Figure 5.2. Net economic welfare NE M^with fair use r is consumer surplus plus profits
aggregated over works in each of these two categories,
less creation cost.
NEW(r) =
o
[CS(p*(n),n)
CSF(r,n) +p*(n)D(p*(n),n)]dn
N(r)
+
J
N*(r)
CS(0,) i)dn-cN(r}.
(5.11)
CS(p, n) is consumer surplus for those who pay price p to 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, and Areas 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
before,
As
use work
n.
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 The value of r that maximizes NEW, rNEW must then be at most/?*(0), making
,
N*( rNEw)
anc*
all
NEW is
the
value of r satisfying
d(NEW) _
CFS(r,N*(r))
+ p*N*(r)D(p*(N*(r)\N*(r)) -
CS(0, JV*(r))]
'
93
Because
sum of Areas
1,
2,
and
(r))-cN
The
for
(r).
(5.12)
first term is the increase in surplus from allowing more uses of works 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 D r and N' are negative, it will be useful to use absolute values and change signs as appropriate. Let
\[D(^n)-D(r,n)]dn
N
be the average amount of fair use of works vation price r. We can rewrite (5.12) as
through TV at
maximum reser-
d(NEW} =
The
N*(r)rD r (r N*(r))
9
\N'(r)\[CS(0 9 N(r))
c].
(5.13)
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 surplus
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
as the supply of
rD(r,N*(r))
.
[CS(0, N(r))
-c]~
6 USE JVOF
N(r)
N*(r)
where
^M-W^*'-"*
is
....'..,..
the elasticity of the quantity of fair uses for works where the price above the fair use reservation price, and
is
94
Developments
in the
economics of copyright
is
the absolute value of the elasticity of supply of works with respect to the
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. The denominator is the total net surplus of works that would be excluded if fair
use were marginally enhanced.
fair use.
of total works created to works created where price exceeds the marginal
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*
;
LXI i n r*D(r,n*)
.
[CS(0,*)-c]
If the elasticity
ticity
USE <e,
^W
sufficiently exceeds the elas-
of
minimum
will
As before, the difference between net economic welfare NEW and aggregate
consumer surplus
ACS is that the latter does not include profits for works
by Area 3 in Figures 5.2 and 5.3. As consumer surplus 1 and 2 in the cases portrayed by those figures, aggreconsumer over all works gate surplus supplied is
created, represented
N(r) Nr)
ACS(r)=
Because
[CS(p*(n\n)
+ CSF(r,n)]dn+
f
N*(r)
[CS(r,n)
CSF(r,n)]dn.
CSr(r,n) =
-D(r,n), maximizing
N*(r)
D(r,n)dn
+
At N(r), the prices above r
N'(r)(CS(r,N(r))
CSF(r,N(r))}.
combined consumer surplus for customers with reservation who pay for the work and for those who with reservation
95
prices
below
who
is just the total available surplus cover cost at N(r) by definition. The
CS(Q,N(r))
- c,
d(ACS) = d(NEW)
dr
dr
[-rD
(r,n)-D(r,n)]dn.
(5.16)
W3
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.,
the inte-
The degree of
exceeds
r
fair
would increase ACS. increasing use that maximizes aggregate consumer surplus, r ACS
r
At r NEW
further
NEW
The result is qualitatively the same as in the 'equal diversion' case; profits when fair use is increased do not count against aggregate surplus as they do against net economic welfare. The difference here depends on the
lost
convexity of the profit function, which determines the degree to which the
because (5.14) holds, rACS may the Changing policy objective from economic welfare to consumer surplus, ignoring profits to creators, need not lead to a more generous fair use regime.
integral in (5.16)
is
equal r* as well.
Gross Benefit
as the gross consumer surplus, neglecting use reservation price is less than the profitmaximizing price, as portrayed in Figure 5.2, the gross benefit is represented by Areas 1,2,3, and 4. For works where the reservation price exceeds
GB
costs.
fair
the profit-maximizing price, the gross benefit will be the consumer surplus obtained if uses were free to all. In both cases, gross benefit equals net eco-
added back in. Accordingly, as with holds, adapted for the low reservation price implecosts
mentation,
i.e.,
(r) .
(5 , 7)
At r NEW where
, ,
the
first
is
(N
gross benefit, r GB will be less than r NEW If creation costs are disregarded, the social calculus will tilt toward increasing the number of works and away
relative to
is
maximized.
96
Developments
in the
economics of copyright
to which r GB falls below r NEW depends on the sensitivity of the supply of works to the level of fair use. If N' is small, these will differ little. As rNElv may equal r*, r GB will be even more likely to equal r*, the value of
The degree
fair
Number
For works where the profit-maximizing price exceeds the fair use reservar, the number of uses will be those purchased at the profit maximizing price, D(p*(n\ ), plus free uses under the fair use regime, D(r,n). For works where the fair use reservation price exceeds Z>(0,)
tion price
all
The
total
number of
uses,
USE(r\
7r)
USE(r)
is
[D(p*(n),n)
f
N*(r)
D(0,n)dn.
maximized
at r USE, defined
first-order condition
^^=
MT7V1?\
dr
N*W
f J
-D(r,n)dn + N'(r)D(Q,n) = Q.
(5.18)
The
first
is
term
is
second
making more of them 'fair;' the from the reduced output when more generous fair
use reduces the supply of works. To compare r USE to other fair use standards, multiply the derivative in (5.18) through by r. Comparing the result with (5.12) gives
dr
dr
N'(r)[rD(Q,n)
(CS(0,#(r))
r
c)].
(5.19)
>
i.e.,
atr =
NEW
Dividing both sides by 0(0, n) implies that r USE > rNEW if and only if rNEW is greater than the average net surplus per use of the marginal work at r NEW
.
Even if r USE >rNEW r USE may be below ACS This ambiguity 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
,
is
under a
The
added uses
in
In the 'equal diversion' this case from at the average willingness to pay over all setting, the added uses are valued than the works, which is likely to be greater average surplus associated with
increasing fair
'
97
(5.17)
and
(5.19) to find
d(USE)
dr
d(GB)
dr
N'(r)[rD(0,n)-CS(0
N(r))].
This implies that r USE >r GB if and only if r/)(0, n)<CS(Q, N(r)) at r GB 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 preserves many of the results under equal diversion. Fair use is strongest if
works
uses,
aggregate consumer surplus is the objective and weakest if numbers of is the objective. In between those two, the level of fair use that maxilevel that
>r r
where
r*
is
>r >r r r ~ r r*
would not
the
maximum
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
r
NEW depends on the extent to which increasing the level of fair use reduces
works for which the fair use price exceeds the price the copyowner would prefer to charge. The difference between r NEW and r N depends on how the ratio of the supply elasticity of works compares to the = r*, this ratio may be small enough so that elasticity of fair uses. At r N * = = r NEW 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,
profits for
right
>
we could have
strict
equality
among
all
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 copyrighted works 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.
98
Developments
in the
economics of copyright
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
according to numbers of works supplied, net economic welfare, aggregate consumer surplus, gross benefit, and numbers of uses.
Many
results held in
marginal benefit of more fair use is the increase in surplus from added uses of works produced. This should equal the marginal cost of
spective, the
more fair use, which from an efficiency perspective is lost net surplus (consumer 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 influenced 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
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 a less generous fair use standard than that maximizing aggregate consumer surplus. Neglecting willingness to pay uses would tend to
likely lead to
more in favor of additional, less economworks 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 maximizing uses of copyrighted works could be less generous than one designed
to
welfare.
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,
'
99
fair
use
level
of
ing supply of works, and both standards maximizing net economic welfare.
use would be between that maximizing uses and that maximizmay be less generous than that
One
the availability of copyrighted works in the future, e.g., by stimulating creativity and fostering a sense that creators are enriching an ongoing
stream 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 complementarity
affects fair use policy
generated. If
depends very much on how that external benefit is 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 tradeoff between 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 products (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 may inhibit their ability to inform judgments on the design of copyright policy.
ACKNOWLEDGMENT
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.
Thanks
NOTES
1.
Bank & Trust v. Signature Fin. Group, 149 F.3d 1368 Amazon.com v. Barnes & Noble, 73 F. Supp. 2d 1228 (1999).
State Street
2.
A&M
Records,
al. v.
Inc., et al. v.
Studios, et
2002).
100
Developments
in the
economics of copyright
12 Stat. 2860 (Oct. 28, 1998). 2003), interpreting U.S.
3.
4.
Digital Millennium Copyright Act, Pub. L. No. 105-304, Eldred et al. v. Ashcroft (U.S. Sup. Ct. 01-618, Jan. Constitution, Article I, Sec. 8.
15,
5.
17US.C.107.
See Napster,
n.
6.
Harper
& Row
The Nation
v. Universal City Studios, Inc., 464 U.S. 417 (1984); Enterprises, 471 U.S. 539 (1985); Campbell v. Acuff-Rose v.
9.
10.
11.
thank Wendy Gordon for this reference. Brennan (1994) reviews the limits of markets in providing social goals. 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 transactions costs. We show that the fair use policy maximizing net economic welfare is typically greater than zero. A fair use policy focusing on economic considerations may then be more generous than that which minimizes transactions costs. 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). See also National Research Council (2000, p. 138). For an exclusively economic approach, see Landes and Posner (1989).
12.
From
17 U.S.C.
is
whether a use
fair
use are:
the purpose
use, including
is
of a commercial
nature or
2. 3.
is
the nature of the copyrighted work; the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and 4. the effect of the use upon the potential market for or value of the copyrighted work.
Klein et
three.
13.
al.
(2002) deal only with the last of these, the potential market, and not the
this
first
Ramello (chapter 7 of
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. 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. 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. One could regard the demand curve for work n as net of the marginal cost of producing
copies.
6.
17.
8.
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
20.
books, or theatrical release compared to video rental for films. '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.
'
101
21
22.
23.
no fair use might also maximize producer surplus, which undoubtedly development of copyright policy in a political environment. This assumption does not follow from D n <Q. For any n, the value of p* is defined by = 0. Consequently, p*' = - H IH Since the denominator is negative,/;*' <0 if ry/7, n) = pn pp and only if O^ < 0. From (2), 11^ pDpfl + D n The second term is assumed negative, but the first could be positive and outweigh it. We have/?*' < when D < 0, i.e., the demand pn curve becomes steeper at any price as n increases. The consumer surplus of works created exceeds the creation cost c for r>r*. At r*, the * just equals costs. As consumer surplus exceeds profits, consumer profit of work surplus exceeds costs. It thus will be true for all <*, since profits increase as n falls.
policy to have
will affect the
.
.
24.
From Figure
tor
is
5.2,
CSF(r,n)
= CS(0,)-
CS(r,n)-rD(r,ri).
The
- D(r, n) - rD
r (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)
NEW.
REFERENCES
(1989), 'Private copying, appropriability, and optimal Journal of Law and Economics, 32, 255-80. Besen, S., W. Manning and W. Mitchell (1978), 'Copyright liability for cable television: Compulsory licensing and the Coase theorem', Journal of Law and Economics, 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 of
Besen,
S.
copying
Broadcasting 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
iparticlelll40001.htm.
Depoorter, B. and F. Parisi (2002), 'Fair use and copyright protection: price theory explanation', International Review of Law and Economics, 21, 452-73. structural and economic Gordon, W. (1982), 'Fair use as market failure: 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: Commodification and market perspectives', in N. Elkin-Koren and N. Netanel (eds), The Commodification of Information, Norwell, MA: Kluwer, pp. 149-92.
102
Developments
in the
economics of copyright
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
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 of Legal Studies, 18, 325-63. Liebowitz, S. (1985), 'Copying and indirect appropriability: Photocopying of journals', Journal of Political Economy, 93, 945-57. National Council for Higher Education (2003), 'Of
and
elec-
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: Convergence or contradiction? The opposite tragedy of the creative commons', in L. Takeyama, W. Gordon and R. Towse (eds), Developments in the Economics of and Northampton, MA, Copyright: Research and Analysis, Cheltenham,
UK
6.
Towards a
Christopher S.
differentiated products
theory of copyright
Yoo
6.1
INTRODUCTION
The economic
analysis of copyright is largely founded on the premise that of consumption copyrighted works is non-rivalrous, a premise that distinmarkets for intellectual property from markets for other types of guishes
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.
If
Non-rivalry in turn gives rise to a authors are to break even, the prices
they charge must defray the fixed cost needed to produce the work in the 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 competition from substitutes. The traditional economic approach to copyright assumes that copyright turns authors into monopolists over their works (Novos and Waldman, 1 984;
The power over price conveyed by this legal monopoly gives rise to the familiar efficiency losses that occur whenever prices exceed marginal cost. The concomitant transfer of surplus from consumers
Liebowitz, 1985).
power
some
103
104
Developments
in the
economics of copyright
(Netanel, 1996; Boyle, 2000; Cohen, 2000). Together these considerations have led to the emergence of a consensus that copyright protection is a necessary evil and that Congress and the courts should calibrate copyright protection to the lowest level that
still
The problem with the traditional approach is that copyrighted works do in fact face competition from other works that serve as imperfect substitutes 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
on the approach
to capture the full dynamics of entry. Other scholars have drawn 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). 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
1
different works has begun to appear (Lunney, 1996; Abramowicz, 2004). Previous efforts have stopped short of considering the full range of
among
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 up the 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 incentives 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.
05
promote both considerations simultaneously, the 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 maring entry can essentially
differentiated products
2 ginal cost as well as increase product variety. 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 copyright 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 beyond the relatively static vision of monopoly economics and captures the dynamics of free entry. The differentiated products approach also accords
By providing 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 institutionally capable of striking the proper balance between access and incentives. 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
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 competition
relaxes the assumption that
106
Developments
in the
economics of copyright
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 of surplus generated by any particular work, but also on the total
amount number
of works created.
differentiated
The market power conveyed by product differentiation leads producers of works to set short-run prices in precisely the same manner as
a monopolist, which results in deadweight loss and short-run supracompetitive profits. However, the possibility of entry recognized by monopolistic competition dictates that the short-run equilibrium is unstable. When
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
entry
3 symmetric with respect to a cluster of other works. 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 incumbent work to shift inward. In addition, the demand curve confronting each
elastic as
a growing
increases
number of imperfect
Entry will continue until no economic profits remain, which, under Chamberlin's original formulation, will occur when the surplus appropri4 ated by each author is just enough to cover fixed costs. The equilibrium
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 infinity,
loss
Mankiw and
changes to the
a differentiated products approach effects two significant 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
107
margins incomplete indicators of economic efficiency. As a result, product differentiation 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
it
illustrates
how
access can
be promoted not by lowering the degree of protection provided by copyright, but rather by facilitating entry by similar works. Entry by near substitutes causes the
demand curve
facing each
work
to
become more
elastic,
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 nonrivalrous goods are homogeneous, entry is unnecessarily duplicative and simply wastes resources. Moreover, the tendency towards natural monopoly created by declining-cost structures strongly suggests that no such entry
which
would be
viable.
This analysis suggests that access may be promoted by strengthening copyright protection, because it is the presence of profits that stimulates
(although, per the subsequent discussion, the analysis may require weakening of other aspects of copyright protection). This stands in stark
entry
contrast to the traditional approach to promoting access, which focuses solely on lowering the level of copyright protection. Indeed, the differentiated products approach suggests that weakening copyright will only serve to deter entry by reducing the revenue generated by each work.
7
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,
dissipated. Furthermore, any short-run transfer of surplus from consumers to producers will largely accrue to consumers' benefit over the long run in
the
variety.
The foregoing
discussion demonstrates
how
approach opens up the policy space by revealing how access can be promoted 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 following precisely the opposite of the policies prescribed under the traditional approach to the economics of copyright.
108
Developments
in the
economics of copyright
Incentives
Appropriability as a determinant of optimal incentives In addition to suggesting an alternative way to promote access, the differentiated products approach also offers a solution to one of the fundamen-
of the traditional approach, which is its inability to provide a basis for formalizing the optimal level of incentives. In the absence of such a basis, scholars have employed rough metrics to approximate the
tal limitations
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.
analysis overcomes this shortcoming by offering a basis for the determining optimal level of entry. It suggests that a work should be whenever the surplus it would create exceeds the costs needed to produced
The present
produce
the
Stiglitz,
it.
This condition
work
is
is met whenever the total surplus generated by than the fixed cost (Spence, 1976a, 1976b; Dixit and larger
1977; Spence
Perry, 1981).
This criterion for determining efficient levels of entry illustrates the importance of the authors' ability to appropriate the surplus generated
their works. It suggests that a reduction in authors' ability to appropriate the surplus created by their works can cause them to forgo creating
from
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 copyright
law should carefully calibrate appropriability so that works capture only enough surplus to support creation of the work. The differentiated products approach reveals that such fine tuning would be counterproductive,
any limitations to authors' ability to appropriate surplus will cause a suboptimal number of works to be created and entry by competitive works
since
will help
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
of exacerbating the access side of the tradeoff by decreasing the extent to which entry can narrow the spread between price and marginal cost in the long run. Indeed, the foregoing analysis demonstrates how loweffect
ering the level of copyright protection, rather than promoting efficiency, may have the perverse effect of entrenching a concentrated market structure into place
by making
it
impossible for
new competitors
to enter.
109
The
diversion as a countervailing consideration analysis advanced thus far would appear to suggest that economic welfare would best be promoted by maximizing authors' ability to appropriate surplus. Indeed, were appropriability the only relevant consideration,
Demand
copyright policy would devolve into a simple matter of allowing authors to capture as much revenue as possible. However, complete appropriation
of surplus, which would require perfect price discrimination, is a practical 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. Following Borenstein (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 optimal number of works even when authors are unable to appropriate the entirety
of the surplus created by their works.
dictates that a
As
by the differentiated
work be produced whenever the benefits products approach 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 diversion 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
underproduction of product
In fact,
variety.
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 produce new works only when the surplus attributable to demand creation exceeds the fixed costs needed to produce the work. The problem is that a profitmaximizing 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 necessarily 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.
10
Developments
in the
economics of copyright
6.3
In contrast to the traditional approach, the differentiated products approach also provides a basis for distinguishing among the available policy
instruments.
As noted
earlier,
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
Liu, 2002; Hughes, 2003). Product differentiation theory provides a basis for modeling the impact of 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-
may come in the characteristics space to a patented product without constituting infringement (Klemperer, 1990).
peting product
approach to copyright suggests a similar with but some important modifications. Specifically, it suggests analysis, that the analysis might be enriched by disaggregating the concept of length
differentiated products
The
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
into
'intensity'
of the
right, as
available surplus.
An increase in
determined by the author's ability to capture the 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 by a change in law that facilitates price discrimination.
in this manner highlights the complex interthese factors. among Interestingly, the policy implications do not in the same direction. point Specifically, the differentiated products
approach indicates that the best way to render a market more competitive by increasing the number of surplus-generating activities encompassed by the right, facilitating authors' ability to appropriate surplus, and by liberalizing how close competing products can come without constituting infringement. In other words, economic welfare would best be promoted by a copyright that is large and intense, but narrow. The differentiated products approach is therefore not an unqualified endorsement for strengthenis
ing copyright protection. Instead, by providing a basis for distinguishing among different aspects of copyright protection, it allows for a more
policy.
1 1 1
The
An
important determinant of the overall competitiveness of markets for copyrighted works is the size of the copyright, as determined by the number of surplus-generating activities encompassed within its scope. As noted
earlier,
among
is
differentiated products
is
deter-
mined by the
of entry, which
to
expand
the size of the copyright by increasing the number of surplus-generating activities that fall within the scope of each copyrighted work. One determinant 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 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
size
of the right can promote both access and incentives by increasing the equilibrium number of works. The suggestion that access would be promoted
by raising rather than lowering the level of copyright protection may seem counterintuitive. After all, it implies that the proper policy response
best
to markets that are too concentrated
is
protection that authors enjoy. This apparent contradiction disappears when viewed in light of the traditional approach's inability to capture the
is free, any strengthening in the level of not ultimately accrue to the benefit of the incumbents. Instead, it will only attract more entry, which will in turn reduce deadweight loss and bring the number of works closer to the optimum. Any short-term profits made possible by the expansion of the size of the right
dynamics of
entry.
So long as entry
will largely
The intensity of the right Another consideration that determines the market's ability to promote economic 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 outwards, increases in intensity leave the demand curve intact and simply increase the proportion of the area under the demand curve captured by
1 1
Developments
is
in the
economics of copyright
authors. Intensity
doctrine,
affected
which
The breadth of
priate surplus.
As noted
earlier,
any reduction
equilibrium number of works. Such a reduction can harm the incentives side of the tradeoff by causing the total number of works produced to drop below optimal levels. It also harms the access side of the tradeoff by limiting 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 surplus created 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
stitutes plays
the right
which consumers regard competing works as suba 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 infringing
Finally, the extent to
works 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
captured by any particular work derives from demand diversion. The higher the degree of substitutability between the works, the greater the propor-
works can be without constituting infringement, the greater the amount of surplus will come from demand
diversion.
tion of the total surplus captured that other words, the more closely related
demand
As noted
earlier,
how
optimum.
Up
to a
demand diversion plays a critical role in determining number of works produced will approximate the point, demand diversion is beneficial, as it can replace
the surplus that authors are unable to appropriate because of their imperfect ability to price discriminate. Beyond that point, demand diversion
creates the possibility of excess entry. The danger of excess entry greatest when goods are the most substitutable.
is
the
113
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
welfare losses
balancing the welfare gains on the access side of the tradeoff against the on the incentives side of the tradeoff. 10
Interactions
Copyright Protection
implement.
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
One problem
overlapping nature of these considerations complicates isolating the impact of 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 counempirical 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 pattervailing considerations.
Some
terns in the radio industry estimates that the deadweight losses attributable to excess entry may be substantial (Berry and Waldfogel, 1999). The
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 payments
latter
for
in
programming. What appears to be excessive entry when measured solely terms of benefits to advertisers may in fact be efficient when measured in
listeners.
]
These studies suggest the possibility of isolating the impact of each of the three factors identified by the differentiated products approach. Even
if
14
Developments
in the
economics of copyright
approach should
still
way
these factors
is
least likely to
be a
problem when a work has few substitutes. Thus, contrary to the conventional 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
an approach that demonstrates how stimulating entry can access and incentives simultaneously. This stands in stark both promote contrast to the position that dominates existing copyright scholarship, which views these two considerations as being in inexorable tension.
emerges
is
differentiated products approach also suggests that the best policy response to a highly concentrated market might well be to strengthen the
The
new
entry.
There
is
might tend to be the when fixed costs and the low of strongest high degree substitutability cause the market to become the most concentrated, but this apparent paradox
irony in the fact that copyright protection
is
some
and
resolved once one understands the complex manner in which access incentives interact with one another. In this sense, the differentiated
nomic
that short-run profits can play in stimulating entry and in promoting ecoefficiency. At the same time, the differentiated products approach
diversion.
demand
erable additional
tionalized.
Although the theoretical implications of this analysis are clear, considwork remains to be done before it can be fully opera-
As noted earlier, further work should incorporate elements of innovation that take into account the extent to which current sequential 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
for other
1 1
by using the spatial competition models pioneered by Hotelling (1929), in which works compete by adopting locations across a characteristics space
that entry
(Yoo, 2004). Relaxing the symmetry assumption allows for the possibility by a new work will impact only some of the works rather than
being spread evenly across all of the incumbents. This localization of competition 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
to
the size of the total surplus of the entire market relative to the fixed cost. lack of robust competition within a subsegment may limit the extent
cost. It
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.
effects
with respect to multiple subsegments, these adverse 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
If these effects arise
may be
strategies to
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,
1967; Hay, 1976; Prescott
984; Judd, 1985). Finally, the policy instruments that follow from the differentiated products approach are by their nature extremely contextual and do not lend
1
among
themselves to simple policy inferences. In addition, the interrelationships the available policy instruments make calibrating them simultan-
eously an extremely difficult empirical exercise. The fact that the differentiated products approach is contextual and nuanced should not obscure
its
basic analytical
the theory.
Indeed, the intuitions that the theory reveals about the relationship between access and efficiency and the manner in which the various aspects
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,
116
Developments
in the
economics of copyright
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 Law School 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
.
2.
Ghosh (2004) employs a differentiated products model to evaluate the related issue of competition between originals and derivative works. The differentiated products approach can thus be understood as striking a middle ground between the traditional analysis of increased copyright protection and the nonprotection of 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.
it is
possible that
3.
induce levels of entry that are excessive. Chamberlin's original formulation also made a number of other simplifying assumptions, 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
strong property rights
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 products would 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
4.
It has long been 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). This conclusion depends on the convexity of consumer preferences and production sets
aspects of the equilibrium (Kaldor, 1935; Archibald, 1961). There is, however, a well-known exception to Chamberlin's zero-profit result.
5.
6.
(Hart, 1980; Roberts, 1980). Because firms must cover their fixed costs, price will not completely converge to marginal
cost.
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 stimulate entry in the same manner as occurs in perfectly competitive markets that are in tem-
8.
9.
porary disequilibrium. Subject to an important consideration discussed in the following section. Other analyses use different terminology to describe the same effect (Beath and Katsoulacos, 1991, p. 57 ('cannibalisation'); Mankiw and Whinston, 1986 ('business
stealing effect')).
1 1
10.
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. Works with steep inverse demand functions tend to be products with low own-price elasticities 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
11.
(2004, pp. 273^4). 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
12.
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). 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 copyrighted work 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 problem of transfer pricing, in which a particular good simultaneously serves as an end product and 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 equal to the price charged of the good when sold as a final product, so long as the final product
market
is sufficiently competitive (Milgrom and Roberts, 1992, pp. 79-83). Thus, so long as 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
that the
the works not be sufficiently competitive, the differentiated products approach suggests problem might be redressed by making the market more competitive by stimulating entry rather than
holdout behavior
may
Competition policy, tors. Thus, unless such refusals create losses for more than just particular individuals, no intervention 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.
by lowering the price paid by follow-on authors. It is true that prevent particular authors from creating particular works. however, focuses on protecting competition, not particular competi-
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 of Economic Studies,
29(1), 2-28.
Baumol,
Economy, 75 (5), 674-85. Death, J. and Y. Katsoulacos (1991), The Economic Theory of Product Differentiation, 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 of Law and Economics, 32 (2), 255-80. Bonanno, G. (1987), 'Location choice, product proliferation and entry deterrence', Review of Economic Studies, 54 (1), 37-45.
The
W.J. (1967), 'Calculation of optimal product abstract product approach', Journal of Political
and
retailer characteristics:
1 1
Developments
S.
in the
economics of copyright
Borenstein,
Boyle,
J.
RAND Journal
of Economics, 16
380-97.
(2000), 'Cruel,
mean, or
lavish?
Economic
digital intellectual property', Vanderbilt Law Review, 53 (6), 2007-39. Brander, J.A. and J. Eaton (1984), 'Product line rivalry', American Economic
and
Review, 74
(3),
323-34.
Brennan,
and public goods: theory of value applied to retransmission', International Review of Law and Economics, 22 (3), 347-75.
D.J. (2002), 'Fair price
MA:
Cohen,
Harvard University
I.E. (2000),
Press.
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 Economic Review (Papers and Proceedings), 57 (2), 347-59.
A.K. and I.E. Stiglitz (1977), 'Monopolistic competition and optimum product 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 persistence of excess capacity and monopoly in growing spatial markets', Economica, 46
Dixit,
(182), 149-58.
Eaton, B.C. and R.G. Lipsey (1980), 'Exit barriers are entry barriers: The durability 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
(eds),
Handbook of Industrial
Organization, vol.
1,
Amsterdam:
Review, 73
1203-56.
(2004), 'Market entry and the proper scope of copyright', working paper available at http://ssrn.com/abstract =569 162. Goettler, R.L. and R. Shachar (2001), 'Spatial competition in the network televi-
Ghosh,
S.
SSRN
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 competition', 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),
large economies:
The
Commodity
Economic Theory, 41 (2), 356-91. Judd, K.L. (1985), 'Credible spatial preemption', 16(2), 153-66.
RAND
Journal of Economics,
1 1
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?', Journal of Economics, 21 (1) 1 13-30.
RAND
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 competition, 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: proposal', Michigan Law Review, 101 (2),
409-81.
Lunney, G.S.,
Vanderbilt
Jr.
Law
Review, 49
(3),
483-656.
(1986), 'Free entry
and
social inefficiency',
17
(1),
48-58.
and
J.
Englewood Cliffs, NJ: Prentice-Hall. Netanel, N.W. (1996), 'Copyright and a democratic civil
106
(2),
Law Journal,
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 foresight', Bell Journal of Economics, 8 (2), 378-93. Roberts, K. (1980), 'The limit points of monopolistic competition', Journal of
Economic 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.
B. Owen (1977), 'Television programming, monopolistic competiwelfare', Quarterly Journal of Economics, 91 (1), 103-26. Waterson, M. (1990), 'Product differentiation and profitability: asymmetric model', Journal of Industrial Economics, 39 (2), 1 13-30.
Spence,
tion,
M. and
and
An
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 Law
Review, 19 (\),2\2-W.
7.
Private appropriability
and sharing
7.1
INTRODUCTION
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
One of
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
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
ening copyright
analytical shortcomings, the present-day policy of extending and strengthis 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 copyright'. 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 information 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
1
120
121
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
right doctrine,
dynamics are displacing the traditional vision of copywhich 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
goal.
domain
as
its
primary
economic theory have provided the general the role of copyright as an incentive to create, but of paradigm have rarely recognized the importance of balancing the appropriability granted by copyright with the necessity of assuring broad access to the
Utilitarian
and
neoclassical
in support
works created as a
result. The contributions generally emphasize appropriover accessibility, with one notable exception being Landes and ability Posner (1989). These distinguished law and economics scholars clearly
welfare-enhancing effects of copyright are ambiguously dependent on a balance between two opposite 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
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.
This chapter shall take the route of attempting to gain further insight
ambiguous relationship, paying special attention to the peculiarwhich characterize the production side. Using a wide selection of literature drawn from the various social science disciplines, it will attempt to expose some of the shortcomings of traditional economic theory as it is
into this
ities
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
maximal copyright
affects
be greater than expected. This is so because 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
It
its dynamic productivity. mere follows, by straightforward application of neoclassical economic extended copyright is liable to determine dynamnew the that reasoning, outcomes inefficient analogous to those presaged by Hardin (1968) in ically
122
Developments
in the
economics of copyright
landmark paper, and which, in tribute to this seminal author, we have termed 'the opposite tragedy of the creative commons'.
his
as follows: section 7.2 presents the tragedy of 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
the
dependent on the sharing process. Section 7.3 expands further the structure of knowledge, whose two complementary constituent upon the tacit and codified dimensions, are equally necessary for elements,
intrinsically
and both have a productive role. It then real-world policies for promoting knowledge creation are generally consistent with this representation, thus implicitly acknowledging the limited ability of proprietary systems to stimulate production of
knowledge to
exist in its entirety,
illustrates
how
optimal levels of new knowledge. Section 7.4 provides a test for the soundness 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 contains the concluding remarks.
7.2
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 suboptimal level of new expressions of ideas. From this perspective, copyrightprotected information
to be avoided
is its is
underproduction.
However
from
there exists another, distinct type of market failure, arising the fact that information also serves as an input to the production
common resource.
Commons
From Commons
to Creative
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
23
distinct
and
we
effects
on the
preservation of inputs. In order to explain this assertion, though, we need to examine in greater 'commons' is generdepth the nature of the knowledge as a commons.
to say in joint use or possession. Typical examples are uncultivated pastures, hunting grounds, the problem here is that allowing open access to all fishery stocks, etc. that
is
ally defined to
be a resource held in
common,
Now
members of
ploitation
community which owns the commons can lead to overexof the resource and its consequent depletion. This dynamic has
the
been extensively investigated in the scientific literature of complementary domains. 5 The crucial element of all these contributions is that the production function of the commons exhibits diminishing returns with increasing number
users. Some examples of this are pastures grazed by increasing numbers of animals, or fishery stocks exploited by increasing numbers of fishing
of
boats.
Hardin (1968,
all
p.
which
ruin to
men
believes in the
all'.
rush, each pursuing his own best interest in a society that freedom of the commons. Freedom in a commons brings
pivotal element of the above reasoning is, as always, the exhaustibility 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,
The
efficiently. 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
is
This
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 references 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;
merely challenges the results of scientific work the peculiar characteristics of creative commons, and
it
Whereas the
124
Developments
in the
economics of copyright
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
exploitation.
8
its
under-
own right)
Generally speaking, there is a widely held consensus, within the disciplines that study creative and inventive activities, that 'intellectual activity in
not ex nihilo creation. Given this vital dependence of a person's on the ideas of those who came before him, [consequently] intelthoughts
general
is
lectual products
p. 38).'
1989,
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 innovators stand
on the shoulders of
giants,
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 previ10 ously in existence' (Nelson and Winter, 1982, p. 130). It is no coincidence that the generic term denoting creative activity is 'composition' (we speak in effect of literary composition, musical compos-
whose Latin etymology specifically invokes the above-described process: 'cumponere' means literally 'to put together', suggesting that creation is first and foremost a novel arrangement of existing elements.
ition, etc.)
this viewpoint only seizes upon the incremental aspect of the 11 of ideas, whereas there is also a non-hierarchical dimension to production consider, namely that of the sharing process which engenders the cognitive
However,
is
closely
bound up with
con-
Paraphrasing Geertz (1973, p. 11), knowledge 'is public because 12 meaning is'. Thus there can be no knowledge without meaning. And, accordingly, there can be no meaning without a human group to share it.
Private appropriability
25
Therefore,
knowledge fragments are often created by happen embedded within the broader context of the collective semantic space to which the knowledge fragments are inextricindividuals, this can only
even though
ably tied.
p.
some
further
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
be small specialized subgroups, such as medical practitioners, 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.
that there
may
In other words, citing again the enlightening definition formulated by 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 13 extent, an increasing function of the numbers of individuals that share it. 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, creative processes and, in short, all those elements which constitute knowledge in its broadest sense - then they will inevitably dry up in terms of meaning 14 and, by consequence, of value to individuals. It follows that a necessary existence condition for a knowledge creative commons is the possibility of sharing and it presents value as an output by reason 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 - due to its willingness-to-pay for its purchase 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
126
Developments
in the
economics of copyright
in
as a text in
a shared symbolic universe would have the same effect upon the reader an unknown language. It cannot be understood, and therefore
all
has no meaning and no value. 15 This argument can be applied to semantic dimensions of literature.
the
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
is true for both the syntax and the distinc'sound' and 'noise'. In fact, to a listener from one cultural context, the music of another culture may sound tuneless, incomprehensible or even distasteful. In this case too, not-sharing of the semanit
tion
made between
content conveyed through a musical piece translates into a feeling of strangeness, or even aversion, in the listener (note that in this case semantic tic
in other words,
relating to
istic, etc.).
some
16
recreational, religious,
hedon-
It is worth noting that a growing body of work in economic theory is today acknowledging the collective and networked dimension of knowledge, albeit in an unconscious and indirect form that focuses on the pro-
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
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 platform
effectively that
of 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
nomic
should be pointed out, again with reference to the traditional ecocategories, that sharing therefore plays a twofold role in the produc-
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 interpersonal relations by creating a common ground in which relations can more easily and quickly be conducted. This is tantamount to opening a commu-
tion process,
members of social 17 groups to engage in dialogue and learn more easily. 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
nication channel that enables the various individual
Private appropriability
18
27
To put it another way, the creative commons largely exists embedded within the sharing process and in association with a social
individuals).
common
separated by individuals and divisible into finite portions whose sum will correspond to the original amount. An uncultivated pasture, for example,
lots,
each of which
erty rights and whose sum will always add up So a creative commons is a special case, as we have seen
its potential users, and subject to specific propto the original resource.
is
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
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 contributions dealing with the economic analysis of copyright, though not by the economic theory tout court, as we shall see below. In fact, the represen-
work
tation implicitly adopted by this literature is that, even if each copyrighted constitutes a public good in itself, it is also, from a productive standpoint, a well-defined entity clearly separable from the others. The incentive
is
consequently designed to act on individual units n times n in order to stimulate the creation of an optimal copyrightable works number of new expressions of ideas. In other words, the production process
mechanism
is
acting individually
atomized, with each creator likened to a standard production function, 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 representation
is
into account the productive role of the creative a whole constitutes an input in its own right.
If
we
instead adopt a
more
correct analytical perspective, which takes and constituent functions of the sharing
domain, we find that rationing of access to individual information items through copyright will in reality have much more far-reaching effects than those traditionally envisaged by the conventional
economics
literature.
128
Developments
in the
economics of copyright
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 cannot be totally integrated into an
(Morin 1962,
19
p. 25).
In this connection
we
notice
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'creative' industries,
how many
20 grated into their new products. 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 production, distribution, or consumption' (Granovetter 1985, p. 483). In particular, this
cancelling
viduals,
procedure of atomization, according to its critics, has the effect of from view all that cannot be directly attributed to single indi-
and which
between them.
If
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. An extensive 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.
is,
as
we have seen,
entirely discounted
by the
rights.
In fact, in all the contributions of that literature, the treatment of information as an output which can later serve as an input is based on the assumption that knowledge corresponds to the sum of information items 22 produced by creative processes, viewed as discrete entities. Hence, according to this assumption, a proper incentive will produce the optimal level of information, and therefore of knowledge.
However such a
made up
Private appropriability
29
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
which is nevertheless communicated and therefore exists within relations between individuals, consequently termed '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,
clearly articulable or codifiable, but
then also able to be 'commodified', giving rise to markets of information 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
it is
never be entirely encoded or commodified. Consequently, as we shall see below, it cannot be produced and exchanged on the markets, but depends
instead
upon
interpersonal relationships for its production and disseminaknowledge grows out of the dynamic sphere of the comexist, as well as for
is a necessary precondition for the sustenance of human groupings. 24
knowledge to
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 information is only a subset of knowledge. 25
Economic
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 particular circumstances of time and place' which is essentially inalienable and demands 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 productive 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 particular', but effective and crucial for the functioning of the organization.
Transmission of
as
we have
form of knowledge can only take place, of social interaction and sharing. The
above argument explains, for example, the rationale behind the educational systems of different societies: books are generally used to provide information,
is
30
Developments
in the
economics of copyright
room equipment
that tacit
encoded. Accordingly, Nelson (2003, p. 917) asserts that any 'class... [is a] complement not a substitute for an effective
What is missing,
in fact,
is
the imparting of
27 knowledge which books alone are unable to convey. 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 any
other means.
Apparent Ambiguity
If public policies in support of the production and dissemination of knowledge are to be effective, they must take both its tacit and codified dimensions into account. However, the distinction between these two components,
though possible at the theoretical level, is in practice liable to be misleading because the two are not so easily divided. The decoding of codified knowwhich
ledge generally relies upon possession of its complementary component, is the corresponding tacit knowledge (Rooney et al., 2003). 28 Therefore, any action exerted on one component may have indirect on the other, while the availability of one does not guarantee acqui-
effects
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., establishing ad hoc social relations). At the same time, these studies have pointed out the unavoidable necessity of promoting both types of transfers, if the 29 policy undertaken is to have a favourable outcome. 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 appropriate 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 optimum knowledge production, and instead finance significant numbers of collective initiatives which, in a truly efficient market, would have no real reason
to exist.
Private appropriability
131
community, for example, conferences, workshops considered to be indispensable to the work and intellectual growth of scholars, and for this reason are allocated public or private
scientific
Within the
and the
like are
funding, may sometimes benefit from tax subsidies, etc. In the same vein, all the governments of countries with sufficiently developed economies subsidize universities, research centres
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 traditional 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
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
activities,
rights in general,
is
can
only effective with respect to codified knowledge, and production of an efficient level of new items of
It functions optimally in markets where it is possible to and produce 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 socializing frameworks. So, public and private organizations quite correctly
information.
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 proper balance 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
As
132
Developments
will
it is
in the
economics of copyright
of the one
the other -
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
content. 31 Naturally, 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
tied to that
book and
its
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 contributto storing the knowledge of the past and perpetuating that of the ing 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,
32
because
it
on the transmission of
tacit
This situation can be described with the metaphor of an knowledge. 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 knowledge 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 impediment to 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
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 knowledge and the productivity of creative commons; as a matter of fact, a certain degree of entropy must always by definition characterize knowledge and 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 imperceptible effects on human relationships and hence on tacit knowledge.
Private appropriability
33
In contrast, the increasingly exclusionary effects of maximal copyright, dilate the private space at the expense of the public sphere, are having ever more serious repercussions on human interactions,
which continues to
on
tacit
commons
as a productive
resource.
Naturally, these effects are not immediately apparent, but emerge gradually over a fairly long-term time horizon. As a matter of fact, the effects
legal regime have up until today been mitigated (though admitan 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
this
of
new
tedly in
',
Nevertheless, as the balance between tacit knowledge/codified knowledge gradually shifts in favour of the latter, under the effects of appropriation 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 elimination of social relationships which occurs when all the individuals
belonging to a specific society/culture die. In such an eventuality it would be 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 individuals)
would be
irretrievably lost, as
its
knowledge
as a whole.
7.4
Under
dimension
an integral and characterizing part of that knowledge. same criterion of dynamic efficiency invoked by the economic theory of copyright in justification of the right must also challenge
sense that sharing It follows that the
the existence of any institutions that significantly impair the sharing
process.
This
is
doubt on
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 copyright, with the production of quantitatively and qualitatively elevated levels
34
Developments
in the
economics of copyright
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
behaviour of creators and allocating their energies with respect to other markets its total elimination might not only prove unfeasible, but also possibly endanger social stability. For the present purposes,
in regulating the
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-
therefore,
mulate appropriate policy indications. One additional interesting test of the soundness of the above conclusions
and compatible with the totemic paraformulated John Locke (1690), which has been widely relied upon digm by for the the literature justification of property rights (Drahos, 1996). The by
is
reference
is
precisely so as to safeguard
particularly apposite because this part of Locke's work focused fairly regulating exploitation of the commons,
defining rights
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
it is
consistent
The
first
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 individual rights, acknowledging in practice the original collective ownership of any common resource, and the precedence which this takes over private
appropriability.
with greater force the necessity of preserving the integrity of common resources, 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
Locke wrote,'does also bound that property too. Nothing was made by God for man to spoil or destroy' (Locke, 1690, section 27).
property,'
.
Private appropriability
35
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
An
and useful
for
liar architecture
dynamic efficiency, but by this same criterion, given the pecuof creative commons, it must leave 'enough and as good
significantly impairs the
for the others', carefully avoiding their depletion. It follows that a transi-
tion to a
know-
ledge sharing process, In stark contraposition to the above, the recent international legislative and jurisprudential trends (which should in theory be informed by the
not to be desired.
Lockean categories) have instead been moving toward a strengthening of exclusive rights through the extension of their duration (see US Copyright
EU
their extension to new subject matters on databases), and through a progressive the Directive 96/9/EEC (see erosion of the scope for derogation from the exclusive right (i.e., fair use doctrine and analogous doctrines in the various national legislations).
EU
In light of the preceding discussion, the present-day trend toward boosting 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,
society
37
may
social
and
38 legal conflicts.
7.5
CONCLUSIONS
Copyrighted works consist of signs which acquire significance by virtue of - the knowledge belonging and referring to a broader semantic universe domain - which is by its nature collective. Even when knowledge fragments are able to be encoded and occasionally transformed into commodities (i.e.,
codified knowledge) they remain indissolubly linked to that component which exists and is reproduced within the sphere of human relationships
(i.e.,
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,
36
Developments
in the
economics of copyright
social
systems,
its
commons - a
special case of
trait
common
production being the creative resource which has the peculiar and
of providing increasing returns with increasing thus numbers of users, implying a pivotal role of the sharing process. the collective In other words, setting is the only possible ecosystem for
distinguishing
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
because application of
its
in the case of knowledge. Nevertheless, this has led some of the literature to justify and analytical misrepresentation not that are applicable to the knowledge domain, and propose policies
which may have the unintended consequence of impoverishing knowledge and 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 property 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,
ative
have the undesired effect of depleting the credetriment of knowledge. to the commons, Given this perspective, it would therefore be socially desirable, as a general rule, to avoid any policy designed to bolster private appropriability
is
liable instead to
expense of knowledge sharing, and instead maintain appropriability minimal level and adopt (or at least preserve the existing) criteria for derogation from strong exclusive rights that move in the direction of assuring adequate communication flows. This is not to imply a negation of the usefulness of copyright and intelat the
at a
lectual property in general in stimulating individual creation, but rather to acknowledge that they play an ancillary role, not with respect to the
Private appropriability
37
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, Anna Maffioletti, 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 disclaimers apply. This work was supported by the MIUR research grant on The Governance of
Intellectual Property'.
NOTES
1
.
2.
3.
4.
is a concise way to describe the nature of copyright in its oria 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). However, this is not a particularly useful argument in favour of the efficiency of copyright, because human history has also shown that creative activities flourish equally in the absence of copyright. See, for example, Posner (1992). On fair use doctrine, see Gordon (1982). 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 constitute not only a productive output, but also an input for other inventive activities. If
The
adjective 'minimal'
ginal
form -
i.e.,
this
is
the case,
Arrow
erty rights may seriously compromise the incremental accrual of knowledge and, consequently, the collective well-being. Unluckily, here again, the observation is not further
5.
(1
6.
7.
Dasgupta and Heal (1979) and Comes and Sandier (1983). A similar hypothesis in philosophical debate is advanced by Drahos (1996). 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 ideological
connection. This idea is also supported by some contributions in the legal literature (see Rose, 1986) dealing with public policy aspects (Rooney et al., 2003). This thesis finds wide support today in the economics literature, in Landes and Posner
8.
9.
Scotchmer (1991) and Foray (2000). Note that sociology, starting with the Max Weber (1967) and Merton (1973), has remarked on the collective dimension of inventive activities.
(1989),
contributions of
10.
This assertion is also consistent with the Schumpeterian theory of innovation put forward in The Theory of Economic Development (1934). This is in essence also recognized by Landes and Posner ( 1 989).
12.
Note
that Geertz,
and anthropologists
specific social
group as
'culture'. In this
chapter
we
knowledge of a term
'knowledge'.
38
Developments
in the
economics of copyright
(1 998). For the applispecific reference to the
3.
Margolis
domain with
14.
The assertion is supported by growing consensus in contributions dealing with the economic 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 would no longer be an information good.
16.
On
17.
18.
19.
20.
'common language' (see Nahapier and Ghoshal, 1998). 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 precisely to encourage the development of creative commons, but they are naturally not creative commons merely by virtue of being spaces. 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 by policymakers' (Rooney et al., 2003, p. 69). Cowan et al. (2000), in emphasizing the specificity of knowledge production, also speak of an 'enculturation model'. 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).
these topics see for instance, Schafer (1977). This, in essence, is analogous to establishing a
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 and
more or less stable relations exist. The form of society is determined by the character and
form of these
22.
relations'.
In this connection,
time, economists
Cowan
et al. (2000, p.
'[fjor
some considerable
took
little if
any
23.
24.
25.
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. 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). The above-described dichotomy between knowledge and information, which attributes to the former a social dimension that cannot be completely owned by a single individual, is widely accepted in the social sciences (see Geertz, 1973; but also Weber, 1967, and
knowledge from their idea of information'. For an overview, see Cowan et al. (2000). The
Durkheim,
26.
1973).
27.
28.
29.
This theory was subsequently taken up, among others, by Polanyi (1958, 1966), Cowan et al. (2000), David and Foray (2002), and Nelson (2003). Nelson and Winter (1982, p. 78) also argue that 'a trait that distinguishes a good instructor is the ability to discover introspectively, and then articulate for the student, much of the knowledge that ordinarily remains tacit'. 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). See for example Williams and Gibson (1 990) and Takii (2004).
Private appropriability
39
30.
In general, looking at data, we can easily observe that, in developed countries, the public sector funds at least one-third of expenditure. Additional public contributions then go to educational systems and cultural activities.
OECD
R&D
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 - in other words, if it (tacit knowledge). If the opportunity to cook a dish is precluded is in an exclusive manner - this will also its handover of the
recipe appropriated technical ability to execute liberty to teach it.
preclude
it,
not be
at
32.
Cowan
functions
33.
'But information written in a code can only perform those interpret the code'. So we are once again referred back
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
34.
35.
dynamic inefficiency (endangering certain incremental creations, e.g., satire which plays an important social role in warranting freedom of expression). 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). For a taxonomy of unauthorized reproduction, see for example Silva and Ramello
(2000).
36.
For western
(1995).
society, see
37. 38.
See note 25 above. Consider for example the Napster case, which, albeit indirectly, involved university students, 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 categories of infringers, it could lead to unsustainable social consequences.
REFERENCES
Alford, William
P.
Book
is
Arrow, Kenneth
invention', in
J.
NBER,
'Economic welfare and the allocation of resources for The Rate and Direction of Incentive Activity: Economic and
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 knowledge codification 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,
Drahos, Peter (1996), A Philosophy of Intellectual Property, Dartmouth: Aldershot. Durkheim, Emile (1973), On Morality and Society: Selected Writings, Chicago and London: University of Chicago Press.
140
Developments
in the
economics of copyright
Dutta, B. and
76,
S.
Mutuswami (1997),
(2000), L'economie de la connaissance, Paris: La Decouverte. Geertz, Clifford (1973), The Interpretation of Cultures, New York: Basic Books. structural and economic analyGordon, WJ. (1982), 'Fair use as market failure: sis of the Betamax case and its predecessors', Columbia Law Review, 82, 1600-57.
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.
Granovetter,
Hettinger, B.C. (1989), 'Justifying intellectual property', Philosophy
Affairs,
and Public
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
Entrepreneurship from 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(ed.), The New Palgrave's Dictionary of Economics and the London: Macmillan, pp. 671-5. Locke, John (1690), Second Treatise on Government, URL: http://www.ilt.columities',
in P.
Newman
Law,
vol. II,
bia.edu/academic/digitexts/locke/second/locke2nd.txt.
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 knowhow', Research Policy, 31, 719-33. Nelson, Richard R. and Sidney G. Winter (1982), An Evolutionary Theory of
MA:
Belknap
Press.
London: Routledge.
Polanyi, Michael (1966), The Tacit Dimension,
New York:
Doubleday.
Co. Posner, Richard A. (1992), Economic Analysis of Law, Toronto: Little Brown 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 and Northampton, Policy in Know ledge- Based Economies, Cheltenham,
UK
of the commons: custom, commerce, and inherently Law Review, 53, 71 1-81.
Private appropriability
141
MA:
Harvard University
S.
Press.
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 New York: Oxford University Press.
Scotchmer,
(1991), 'Standing
Takii,
tacit
knowledge', Review of
Friends
or
Foes?,
Cheltenham,
Theory:
Watzlawick, Paul, Janet Bavelas Beavin and Don D. Jackson (1967), Pragmatics of Human Communication, New York: Norton. Weber, Max (1967), // lavoro intellettuale come professione, Einaudi, Turin (originally published as Wissenschaft als Beruf). Weiner, Robert P. (2000), Creativity and Beyond,
New
New York
Press.
Williams, Frederick and David V. Gibson, (eds) (1990), Technology Transfer: Communication Perspective, Newbury Park, CA: Sage.
8.
economy?
Alessandra Narciso and
INTRODUCTION
The ongoing IMS Health saga has yet again highlighted the apparent clash between copyright and competition law and its essential facilities doctrine. The IMS Health case is a complex one that is still unfolding
before the
EU
Court of
Justice.
IMS
is
data 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 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
licence
file
them
in order for
them
to be able to compete.
The grant of a
refused, which subsequently prompted IMS's competitors to a complaint with the European Commission for abuse of dominant
facts
was
position.
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
The
need to be seen against the background of the apparent clash between copyright and competition law.
142
IMS
8.2
143
Copyright
The
starting points
Two leading studies on the economic analysis of copyright support the continued 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
IMS. In
their study,
it
as
dilemma
little
little
of the information will be used'. Landes and Posner (1989, p. 326) emphasize the matter in even clearer terms when they state that 'Copyright
protection
- the
owner
to prevent others
from
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
law.'
copyright
artistic
is concerned with the protection of literary and 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
Historically, copyright
the resulting expression of his creativity. Only that expression, the individual non-copied expression by the author of a certain idea was protected.
lot
his
and
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
body and
cation of
concept, everyone has a property right in the labour of his own brain; the second concept builds on that by adding that the appli-
human
it
labour to an
unowned
labour to
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
It
144
Developments
in the
economics of copyright
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-
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).
itional
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
consumption of the expression does not necessarily make the expression and its material support unsuitable or unavailable for further consumption.
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
It is
room
is
therefore
At the innovation and creation level, there is very little incentive to create. The creator may not be able to recoup the cost of production, as there is no tool to reap any substantial benefit from such creative activity. There is therefore no efficient
entirely in favour of competition at the
production
level.
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 expression 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 copyright work. A restriction on competition is put in place in furtherance of competition
(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
IMS
145
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 monopolistic 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 behaviour (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 economy has 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
that such
exclusive right, there is by definition a possibility 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 problem with 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 behavioural level
that there
and
it
basis. It is therefore
is
way and
ing
it
at the
same level (Ramello, 2003, p. 28). Copyright and competisame objective, but they have different ways of achievat different levels
(Lemarchand
et
al.,
2003,
pp. 17-19).
146
Developments
in the
economics of copyright
Competition law does not operate at the structural level and as such, it does not interfere with the structure created by copyright to enhance
and innovation. However, competition law may interfere at a later with some of the behaviour that flows from the grant of copyright. stage What can be derived at this stage is that a system that is designed to regulate behaviour should not operate in a way to undo the structural system
creation
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
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
was
set
up
8.3
Background
The IMS 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 following
facts. IMS is a major supplier of marketing data and other healthcare companies. In the German market,
it
structure of local geographic segments, 1 860 in total, called bricks, each containing 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
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
it
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 and one Order from
1
the President of the Court of Justice 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 German
courts.
Court of
Justice for a
preliminary ruling.
IMS
147
When IMS's competitors NDC and AzyX applied for a licence covering the
brick structure, the grant of such a licence
NDC and
subsequently filed a complaint with the European Commission for abuse of dominant position.
AzyX
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
The Commission
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,
6 Germany, whereas the relevant product market is
services.
examining several
that
As regards the position of dominance, IMS is found to be dominant after factors, but the Commission primarily refers to the fact
IMS
8
Europe;
holds a high market share not only in Germany but also in Germany, being the country that has 'the largest market for
9 regional sales data services in Europe' may therefore be regarded as a 'sub10 stantial part of the common market' in this respect.
it is
useful to
know
what regional
Pharmaceutical companies rely on the enables them to gather data on sales of a particular
each of their products with those for the It also offers them a tool to measure the perof their competitors. products formance of their sales representatives. In the brick structure, these data are collected in each brick for a small number of pharmacies in a local area.
sales figures for
] l
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 quasi12 monopoly 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
outcome of
that collaborative
148
Developments
in the
economics of copyright
in the establishment of
effort.
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
structure, given the significant
15
was unwilling 14 to switch to another additional costs they would incur from
doing
so.
is further sustaining that the IMS brick structure is a 'de facto' industry standard. For the Commission, the 'I860 becoming brick structure is becoming a common language for communicating the information' 16 and all the pharmaceutical companies which assisted with
The Commission
creation are in the end, effectively 'locked in' that structure on a voluntarily basis. The 1860 brick structure is especially a great and unique source
its
of information 17 (and here the Commission sees parallels with the Magill case, where broadcasters' TV listings were a unique source of information 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
19 by the Court of Justice in Oscar Bronner. 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 smaller competitor argued that it up by the major national newspaper. should be allowed to make use of the system in order to be able to compete.
In
its
followed the suggestion of its Advocate General to interpret the essential facilities doctrine restrictively and then ruled that it did not apply here.
is
'the refusal
of access to the
competi-
'such refusal
is
on
business,
inasmuch as
IMS
specific
approach?
149
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 competition 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 behaviour. In copyright cases, there is not just one set of rules that affects the market behaviour regulation point. Both copyright and competition rules
set
such a second
otherwise
it is
out to regulate competitive conditions on the market. The presence of set of rules must have an impact on the test that is used,
its vital role is completely ignored. It is therefore submitted that 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
in its
To be
it
its
22 the judgments in Magill21 and Ladbroke It is by now well known that the Magill case started
refused a licence covering the copyright in the TV listings. In its decision the Commission 23 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
24 they were disappointed. 25 By then the case had become one of principle. On appeal, the Court of Justice's judgment did not, however, change a great deal, the Court reach-
as the Court of First Instance. Moreover, the was oriented almost exclusively towards compeof the wording judgment tition law principles and the resemblance with the ICI and Commercial Solvents vs. Commission 26 approach, which relies on elements of the US
ing the
same conclusion
One should be reminded though, that in Magill, 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
the Court
first
of
all
abuse of the dominant position. The Court then went on to say that in exceptional circumstances, the exercise of an exclusive intellectual property
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 obligaright
29 tion to grant a licence.
50
Developments
in the
economics of copyright
The exceptional circumstances referred to by the Court in relation to the broadcasters' refusal to licence the programme listings in Magill come down 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
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 broadcasters were effectively reserving the secondary market for weekly guides to themselves. The Court clearly emphasizes the secondary market
TV
point by indicating that the broadcasters' main activity is broadcasting. At this stage, one clearly sees the appearance of the essential facilities
an application of the essential doctrine in a copyright case, whereas Bronner applies the same doctrine to a 'normal' non-intellectual property case, as discussed above
doctrine. Magill can therefore be seen as
facilities
(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,
facilities
consumer from its application (Areeda, 1989). When the Commission refers to the Court's judgment marizes the position adopted by the Court as follows:
The Court therefore recognised
an exclusive
right deriving
in Magill,
it
sum-
from an
that in exceptional circumstances the exercise of intellectual copyright may be abusive even in
it
prevents the
This
is
summary is causing problems in as far as the use of the term 'inter alia'
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 submitted that it is wrong to use the term 'inter alia? in this way and to give this
interpretation to the judgment of the Court. Instead,
'inter alia'
it is
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
The judgment
in
of TV guides (Torremans, 2001, p. 307). 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
se),
all
IMS
The Commission
its
view
is
the Court of First Instance in Ladbroke. 31 This case dealt with the question to license sound and pictures of French horse of whether a refusal by
PMU
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
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 facilities 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
its
its
main
betting activity in
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
was no
real or poten-
tial substitute,
quote could suggest that there is no need to establish 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-
Taken on
its
own,
this
cation.
It is
is
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).
152
Developments
in the
economics of copyright
The Commission's interpretation also disregards the economic analysis 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 misguided in its reliance on Ladbroke for its views in the IMS decision (Hull
set
The Commission's overall conclusion in its IMS decision - that there was indeed an abuse of a dominant position, based on the fact that it saw the
brick structure as an essential facility before applying its liberal interpretation of Magill to this case, must therefore become open to serious doubts
as to
its
correctness.
8.4
IMS IN COURT
the president of the Court of First Instance and before the President of the Court of Justice in the IMS case 33 were up to
now primarily concerned with the interim measures issue. From the discussion above it has become clear that it is at least arguable
34 open several issues. 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 necessary 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.
that Magill
had
left
One of
the issues
left
open
is
tional circumstances required cumulatively the fact that the intellectual property right was linked to essential inputs for secondary markets and that
new product had to be introduced in that market for which there was sigand 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'. 36 Similarly the Commission dismisses the requirement that the
a
nificant
dominant firm's refusal to licence and grant access to the copyright work must enable it to restrict competition on a second (downstream) market. 37
IMS
counters these arguments with references to other aspects of the It points out that in order for the use of intellectual prop-
become an abuse, there must be exceptional circumstances. These can be found where the refusal to licence is combined with other
IMS
specific
approach?
153
is
an
The
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
notion of essential
facilities
goods or services in the second (usually downstream) market. In his Order of 26 October 2001, the President of the Court of Instance summarizes IMS Health's argument further as follows:
First
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 presupposes 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 downstream 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 new market. 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 amounted to the 'exceptional circumstances' justifying the characterisation of their refusal to license as an abuse. IMS Health's refusal, in this case, to license its competitors 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 exclusivity granted by that right is used as the central feature of the rightholder's business and where its dominance may, at least potentially, turn on the maintenance of its exclusive right. 39
This
ers
is indeed important. In Magill, the broadcasters were the sole suppliof the information on the TV programmes, as only they knew what they were going to broadcast. Magill could not develop alternative listings. In
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 counterpart in the Court of First Instance. Serious consideration was therefore
given to
that
is
IMS
refusal to licence
must
existing case law required that the prevent the emergence of a new product in a market
is,
to licence
40
still
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
154
Developments
in the
economics of copyright
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. The President 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 and AzyX with the interests of competition ignores the primary purpose of Article 82 EC, which is to prevent the distortion of competition, and especially
NDC
to safeguard the interests of consumers, rather than protect the position of particular competitors. 41
8.5
CONCLUSION
all this lead us and what should be the way forward? It is hard where the Court of First Instance and the Court of Justice will even-
Where does
to say
tually take
European law when they render their judgments in the IMS case. We feel a lot of sympathy for the approach taken and defended by IMS.
is
This sympathy
this
debate
is
What we
cerned with
is
the fundamental interaction between copyright and compecould 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 purpose of promoting innovation and enhancing consumer welfare. It is therefore imperative that the presence of an exclusive right, such as copyright or any other intellectual property right, be taken into account as an additional factor when cases such as Magill and IMS (and other cases where an intellectual 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 structure of the
in this respect.
How is that balance to be achieved and how are all these elements to be
taken into account?
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
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
restrictions at
level,
additional right that regulates competition by one level of competition in furtherance of competicopyright changes the competitive framework; a more
therefore warranted.
complex approach
is
As
rights as exclusive rights justify a degree of monopolization of a primary market. After all, it is their role to protect intellectual creations embodied in
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 articles on their normal market. Things are
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 copyright 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 not in direct competition with the first article, and especially not if there
is 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
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
serve their
essential facilities doctrine
is
The question
tion law.
is
That
is
to be put into operation. no means who wins: copyright or competiThe 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
156
Developments
in the
economics of copyright
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
stances
test,
all
NOTES
1
.
2.
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. 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.
IMS Health
vs.
4.
5.
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. 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 'protective
6.
7. 8.
measures may be granted only where the practices of certain undertakings areprima Community rules on competition'.
AKZO
Chemie
BV
vs.
Commission
ECR 1-3359.
9.
10.
1 1
.
12.
1
3.
14.
1
5.
IMS that a change to the brick strucworking 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 involveture involves a
Decision, paragraphs 83, 86, 87 and 1 85. Decision, paragraph 86 in fine. Especially, it has been argued by the competitors of
change
in
paragraph 115.
Decision, paragraph 89. Decision, paragraphs 102, 103 and 104. The crucial fact in Magill is that the monopoly was a
that information
19.
Magill is the unusual protection of copyright of factual information. Case C-7/97 Oscar Bronner GmbH & Co KG vs. Mediaprint Zeitungs- und Zeitschriftenis,
monopoly over information and happened to be the raw material required by a third party. Therefore, now not any longer the only case to be remembered for its unusual facts; that
verlag
20. 21.
Ibid.
GmbH &
Co
112.
Eireann
vs.
Commission (Magill
TV Guide
Ltd intervening)
Ltd vs.
CMLR 586; case T-70/89 British Broadcasting Corporation and BBC Enterprises Commission (Magill TV Guide Ltd intervening) [1991] 4 CMLR 669; and case Tvs.
Commission (Magill
TV
Guide Ltd
ECR
11-575, [1991] 4
CMLR 745.
Commission (Societe
d' Encouragement et des
Ladbroke
SA
vs.
ECR 11-923,
[1997] 5
CMLR 309.
IMS
23.
24.
157
25
Magill TV GuidellTP, BBC and RTE decision [1989] OJ L78/43. See above, note 21. 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] 4 718.
CMLR
26. 27.
28.
Commission [1974] ECR 223. 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. 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 paragraphs 46, 49, 50 and 54.
vs.
29.
30. 31. 32.
33.
Ibid.
34.
35.
36. 37.
38.
39.
40.
41
Decision, paragraph 67. See above, note 22. Decision, paragraph 68. 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 T1 84/01 R and Order of the President of the Court of Justice of 1 1 April 2002 in case C481/01 P(R), all available at www.Curia.eu.int. 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). 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). Paragraph 180 of Commission Decision 2001/1 65/EC. 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. 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. 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. 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. Paragraph 1 45 of the Order of the President of the Court of First Instance of 26 October 2001 in case T-l 84/01 R.
REFERENCES
Areeda,
P.
An
Antitrust
Law Journal,
58, 841-53.
Cooler, R. and T. Ulen, (1988), Law and Economics, 1st edn, New York: Harper Collins (3rd edn, 2000), Harlow and Reading, MA: Addison-Wesley.
Fitzgerald, D. (1998), 'Magill revisited: Tierce Ladbroke European Intellectual Property Review, 20 (4), 154-61.
SA
The Commission',
-
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 of Legal Studies, 18, 325-63.
158
Developments
in the
economics of copyright
- property rights as (1989), 'Property and intellectual property in furtherance of on restrictions competition', International Review competition of Industrial Property and Copyright, 20 (1), 1-15. Lemarchand, S., O. Freget and F. Sardain (2003), 'Biens informationnels: entre
Lehmann, M.
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, and Northampton MA, USA: Edward Elgar. Cheltenham, 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 and Northampton, MA, USA: Edward Elgar, Analysis, Cheltenham,
UK
UK
pp. 118-47.
Spector, H. (1989),
An outline of a theory justifying intellectual and industrial 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),
property
rights',
256-62.
Torremans,
Treacy,
P. (2001), Holy oak and Torremans Intellectual Property Law, 3rd edn, London: Butterworth. P. 'Essential facilities - is the tide turning', European Competition
(1998),
19(8), 501-5. Watt, R. (2000), Copyright and Economic Theory: Friends or Foes!, Cheltenham,
Law Review,
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
F.
of intellectual property
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
IP law, antitrust law, property law, and contract law. Overlooking the totality 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 approach
avoids 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
our approach emphasizes that the analysis in any particushould apply carefully whatever legal regimes the issue at hand - IP law, antitrust law, contract law, etc. In short, courts should implicates
lar case
On procedure,
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-
and fully the framework of the existing positive law regimes, as well as the history of the normative debates leading to their evolution. Courts, then, should expressly identify the perceived failures of
sider accurately
these regimes,
if
new
759
60
Developments
in the
economics of copyright
resolve the putative failures. The mere formality of identifying and rigorously 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 body of 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 we see as a more activist stance toward IP.
if
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 6 plummet. The 1952 Act statutorily overruled both of these aspects of
7 patent law. Yet, on the issue of non-obviousness, over ten years passed after implementation 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 con9 triving new requirements for 'synergism' and 'combination' patents. It 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 applying the law of Section 103 were eliminated and the decisional framework of 10 the 1952 Act was applied consistently according to its own terms. 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 11 And even after this Supreme Court case, Congress acted infringement.
again in 1988 to add subsections 4 and 5 to Section 271(d) to be emphatically 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
61
13 Similar to courts applying existing legal regulations to promulgate. regimes, legislators and regulators setting new policy also should carefully
take into account existing positive law regimes and their normative and historical underpinnings before adopting new laws or regulations that might
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
Rather, analyzing the arrangements involving IP. 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
restrictive licensing
14
people
'basics approach' gives of 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.
- property law, contract law, and the like. The us a workable - and more predictable - framework
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 unpredictable in several respects. First, they include various limitations
restrictive licensing
on
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
By reducing
approach'
facilitates the
ation through the commercialization of the inventions, symbols, and creative 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 and
tion 9.4
so-called general law, which includes property law and contract law. Secshows how to solve various problems at the periphery of IP law
162
Developments
in the
economics of copyright
copyright or patent misuse, that selectively emphasizes or alternatively ignores particular features of various legal disciplines in crafting special-
By focusing on the basics, our approach suggests an to reconceptualize IP law with important implications for important way to market. Section 9.5 concludes. ideas new bringing
ized doctrines for IP.
9.2
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 nonexclusive. Classic
of public goods under can advance a comdiscrimination and that price appropriate conditions, for outcome public goods, resulting in little, if any, petitive equilibrium 17 of IP rights is permitted to price disan owner When loss. deadweight
an
efficient level
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
criminate, the
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
Yet,
more use in certain instances, in practice some price discrimination 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 discriminin theory to
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
ation has
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 discrimination 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
more
licensing arrangements, like tying, which can specifically to reveal his or her own demand for
tied
good they
use.
The
'basics
approach'
form of
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,
trans-
acting parties to
be enforced
know better ex ante how to structure transactions that will later. In addition to reducing legal uncertainty, when a court
is less
own
terms, there
an analysis that applies each body of law on its opportunity for courts to fashion new and unique
doctrines of IP law that undercut private ordering by effectively rewriting ex 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 downstream incentive effects on parties - including owners of IP rights and financiers, such as venture capitalists - are taken into account. Not only do specialized doctrines such as misuse
and preemption
more
times than not they have the effect of eroding the legislatively created property rule protection for IP rights, further compromising commercialization
and and
private ordering by restricting an IP holder's rights both to use his IP 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 substantive 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
which itself has specific statutory components passed to overturn similar court action in the past. 22 Second, as the product of a long history
IP,
of adjudication, lawmaking and academic debate, each area of law presumably 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 that judges certainly can employ to deal with IP, but that are untested and that might simply reflect a particu-
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
lar
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 regime should 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.
164
Developments
in the
economics of copyright
Given our view of the present substantive basics of antitrust law, propand 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 considering matters involving IP, even if each such body of law was more restrictive presently or in the future than we understand it or might prefer it to be.
erty law,
That
the virtues of the procedural component of the 'basics approach' are independent of our or any other particular interpretation of the sub23 stantive basics of each body of law that courts would apply.
is,
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 many are 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 predictabilmatter it ity, promotes the commercialization of IP and the subject
protects.
Our approach
is
in contrast to the
by academics such as Baxter, 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
Bowman and
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
each field, and the law continues to develop. But we believe that can be found on the broad positive legal frameworks of consensus general
exist within
165
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 contract validity are left to the field of contract law. More to the point, focusing 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 happens
each
when some
or that
is
licensing
arrangement that does not violate the antitrust laws is held invalid as a matter of some form
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
To help frame
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
basic principles
Antitrust
and
law.
Law
is
Antitrust law
transactions that substantially lessen competition or tend to create monop25 But it is well established that antitrust law does not prohibit market oly.
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
26 And increasingly, compete, must not be turned upon when he wins.' 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 practices are procompetitive. The usual test for unreasonableness in this context
is
sis
highly fact-dependent and generally is based on a 'rule of reason' analyas opposed to treating such conduct as an antitrust violation per se. 27
166
Developments
in the
economics of copyright
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
Justice
others on the one hand but that she cannot on the other hand be more generous 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
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 33 thereby encourage, invention and innovation, 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
and innovative activity would not predictably be responsive to direct rewards. 34 In practice, IP law facilitates commercialization by
inventive
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 docof indirect infringement - induced and contributory - arose to capture those activities that, at the time conducted, clearly cause essentially
trines
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
the
intended to induce the infringement, the inducement doctrine captures only those who clearly intend to induce infringement and who are successful in
doing
The contributory infringement doctrine operates similarly. It requires proof of direct infringement and proof that the alleged contributor knew that the allegedly contributing conduct was 'especially made or
so.
37
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
167
infringement
is
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-
The
ability
who would
otherwise be guilty of direct or indirect infringement facilitates both price discrimination and coordination among complementary users. For this
- and, indeed, this was a reason, the 1952 Patent Act expressly provides major impetus for the writing of the Act that neither efforts to price discriminate nor the granting of a restrictive or unrestrictive license to a 40 potential infringer shall constitute misuse. This provision was ignored by
many
Dawson decision, 41 which finally recognized its impact. To be certain this was clear, Congress acted again in 1988 by adding
courts until the 1980
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
42 patent misuse. 43 Importantly, because the doctrines of copyright misuse and trademark 44 misuse are based on the doctrines of patent misuse and patent law's indir-
market power
is
ect infringement,
learned from the 'basics' view of patents are also applicable throughout
IP
law.
Under
it is
crimination or imposing restrictions on a licensee are allowed - indeed, - at least to the extent they are otherwise properly they are contemplated
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 protected 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 subject 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 property 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
if it is
168
Developments
in the
economics of copyright
to commercialization of IP assets
and
ite
these doctrines only operate as default rules in practice, and a nearly infin47 range of dealings can be carried out through contract. 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 dimensions of time, use, and the number of property owners. 48 For example, when comes to real property, highly particularized defeasible fees can be created and will be enforced, and a real property owner can create any
it
number of leasehold
opposed to
property.
49
of these transactions
are,
both
real
and personal
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'
Whatever
50 arrangement, focusing instead on the contracting process. strictures property law and contract law impose on private
ordering, parties are generally free to order their affairs and to carve up rights, 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
they do to other types of property. Nothing under property or contract law provides any particular reason to be skeptical about IP contracts that facilitate price discrimination, exclude certain parties from
rights, just like
having access to IP
rights,
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
9.4
Applying the
IP
arrangements, and
69
ex ante by
all
market participants
way that helps them order their same time being more fair and efficient. The 'basics
in a
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 addition 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
through their particularly important roles in the of case law. And the body primary architect of the present patent system the 1952 Patent Act wrote a five-part series of articles about these cases
historical significance
before drafting the statute designed to change fundamentally the courts 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
way
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 approach that 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,
law, the basics
since the core features of other areas of IP law largely derive from patent framework and the essence of the following analysis extend
to copyrights
well.
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-
However, the happenstance that a contract relates to patents should not transform interference with that contract into patent infringement. The
facts that
tion. 52
The
an
need to be proven are different under these different frameworks. 53 potential remedies are different as well. the of classic case indirect Wallace, infringement, involved a patent on
oil
lamp having a new burner, together with a standard fuel reservoir, 54 and wick, chimney. 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
170
Developments
in the
economics of copyright
infringement on the part of its customers, because they would inevitably add a chimney. judgment of contributory infringement makes sense
competitor's efforts
and actual impact of the 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 chimneys from the patentee, the analysis under the basics of IP law would be
under the
'basics approach' because the intended
different.
with contract, or the tying arrangement might violate the antitrust laws. However, the competing chimney seller would not be liable for contributory
56 infringement under the basics of IP law. The Heaton case involves an example of just this type of arrangement. Heaton involved a patentee who sold a patented machine with a label
license 57
- a tying arrangement - and a defendant who sold competing inputs. The court seemed to reason that by providing 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
unpatented inputs
(staples, literally)
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 infringement of the patent. But this turned a case about indirect participation in
breach of contract into patent infringement. By deciding the case the way it did under IP law, the court in effect extended inappropriately the scope
of IP
rights.
would be to immunize potentially anticompetitive licensing arrangements from the antitrust laws. The 'basics approach' rejects the analysis of Heaton. Under the 'basics
ing in Heaton, of course,
approach,' and as pointed out by Rich, this decision was inappropriate because it 'transformed the law of contracts into "patent law".' 59 It may have been appropriate for the plaintiff to consider an interference with contract 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
these contract and antitrust arguments head on, cases Heaton allow parties, and judges, selectively to mix features of various bodies of law and to extrapolate from them to forge new hybrid doctrines
By not addressing
like
171
many
instances, such
IP
rights.
of the law leads to doctrines, such as misuse, that erode 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 same approach as Heaton in the A.B. 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
Dick
approach.'
63
The
reason
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 infringerejects the
ment in the 1952 Patent Act, any effort to follow this inappropriate body of
law 'would
kill itself
in time.' 65
understanding of the basics suggests why Heaton, A.B. Dick, and 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 immuntheir
An
from serious
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 -
other
(e.g.,
IP law, antitrust law, and the general law of property and contracts. As a they lead to unpredictable results and, in the name of IP law, encroach upon the boundaries of other well-established bodies of law that
result,
more nuanced and time-tested doctrines and rules that have staying power and that are perfectly capable of resolving the disputes on their own
reflect
terms.
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 body
of law. In
many of
172
Developments
in the
economics of copyright
and misuse - which are admittedly somewhat difficult doctrines - too many courts and commentators have not followed the 'basics approach' and have
instead tried to re-hash the normative case for IP to develop new specialized 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-
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 wrong party when considering whose behavior should matter in cases of possible
ing the
The behavior of the putative indirect infringer to or encourage direct infringement is relevant to the analysis under both inducement of infringement and contributory infringement. The
indirect infringement.
facilitate
behavior 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 relevant 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 question 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 compared with an analysis based in the basics of IP law - has its most visible
roots in the
&
Leeds
&
Catlin
involved a patentee's competitor selling something that had no substantial 66 In Leeds & Catlin, the defendant-infringer sold spenon-infringing use.
cially
known
grooved records that could only be used in a patented record player 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
tinction,'
than offered by the Court. Under the basics of patent law, there is no 'disor key element, of subject matter claimed under the patent. The
from
'central
claiming'
known
not to
set
73
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
its
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
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.
earlier in
Hanging determinations of indirect infringement on the factors outlined our discussion of the basics - such as intent for induced infringe-
ment and absence of non-infringing substitutes for contributory infringement - 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 any
instruction
on how
to determine
which element
is
key,
we
are aware.
infringement have the essential advantage of being comparatively easy to administer. They look to facts well within the control of the putative
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
infringer
readily identifiable safe harbors
69 important biases and that the doctrines are relatively easy to administer. Importantly, the improper reasoning of the Court in Leeds & Catlin is not
matters which approach is adopted by courts - especially 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
It
wrong
of precedent that focused on 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 70 key elements of the claim. 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 obligthe invention, the
Court advanced a
line
One of
the
ating the licensee to use only certain containers for packaging products 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 non-infring-
many
ing
in the patented
174
Developments
'basics approach'
in the
economics of copyright
The
would again yield the same result as the Court's no contributory infringement - but again for a analysis 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
in this case,
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 implemented in 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 arrangement
is
should have no bearing on a court's analysis of indirect (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
and contract
law, but
purpose of transforming a matter for antitrust and contract law into a matter for some new version of IP law.
To be
Leeds
sure, the
&
Catlin
Court did not always reach the right result, as it did in 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
all
elements
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 infringement. 73 Under this established basic framework of patent law after the 1952 Act, the essential inquiry for indirect infringement is on the comparatively 74 easy to administer framework discussed earlier. While it may be appropriate 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
75
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 participants can rely better on ex ante in structuring their
affairs.
The
important implications for resolving matters and controversial issues found at the periph-
and copyright misuse, restrictive licensing arrangements, 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 approaches
to disputes
is not compatible with 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
and transactions involving IP. Concerning the misuse doctrine, the 'basics approach'
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
that practice is reasonably within the patent grant.' 76 But import77 Rather, antly, the patent statutes make no provision for per se misuse. Section 271(d) provides specific safe harbors for conduct that is not misuse.
if
Further,
its
it is
scope, since patents only give a right to exclude. derived from sources external to IP law.
employed, other bodies of law, such as which to inspect a patentee'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
the 'basics approach'
is
When
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
laws. In brief, the pernicious effect of the erodes IP rights, at least at the margin, and risks rooting out procompetitive and competitively-neutral behavior that the
that
it
176
Developments
in the
economics of copyright
78 If the antitrust laws are too antitrust laws recognize as such and permit. lax, the appropriate remedy is to fix the antitrust laws. As Rich pointed out
commenting on the unfortunate habit of courts to treat potential antitrust concerns as some how more serious and in greater need of policing when IP is involved:
in
The patent
is
right is not the only form of property subject to such misuse. But it 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 79 attempts to emasculate the patent law
so
little
'basics approach' sugshould enforce restrictive licenses that courts involving IP as gests generally contract law and do not run afoul of the are enforceable under as they long
in their
antitrust laws. Indeed, affording IP holders the right to carve up interests 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 restrictive licenses
of copyrights in the ProCD SQ case and of patents in the Mallinckrodt 81 case. 82 Even when a potential or actual IP owner tries to
83 payments for activities that fall outside the protection of IP, courts
extract
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
tract law. 84
- is
contrast, in practice, courts are skeptical of contracts that to royalty payments beyond patent term, even though the to be tied happen economic justification for this skepticism is lacking. 85 Further, courts that
By way of
do not
by finding that restrictive licensing constitute some sort of impermissible arrangements, including tie-ins,
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 themselves 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 additional 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
77
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 Supremacy Clause 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 91 by patent law, or for that matter by copyright or trademark law. 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, by other 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 conflicted
ideas which
with a purported 'strong federal policy favoring free competition in 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 way that allegedly conflicts with the IP law regime putatively doing the pre93 Under the reasoning of Bonito Boats, each such state law or empting. contract term would not be enforceable. In addition, the Bonito Boats reasoning may eviscerate the rights of exclusion that the patent, copyright, and trademark statutes are designed to create and that are both properly justified 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 motivated at least in part by its impact on these same competitive economic concerns. Put differently, the reach of the preemption analysis in Bonito Boats
78
Developments
in the
economics of copyright
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 by
misuse - then the overall power of the Bonito Boats preemption approach is even greater.
In contrast, the 'basics approach'
alternative analytical
recommends
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 competition law 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
framework
for conflict
because
to
would be analyzed easily under this test on a host of elements that are unconnected liability via the plug mold technique. Even closer patent law, including copying
under
it
turns
cases,
such as those involving state laws regulating statements about patents 97 themselves, can be decided using the extra element test.
'basics approach' also provides guidance for policymakers evaluatwhether to promulgate IP laws that interface with other state law ing For regimes. example, in derogation of principles of state contract law and
The
and estates law, the copyright regime gives authors a non-transferable to terminate transfers 98 of their copyrights and even sets forth its own right trusts and estates provisions governing who gets this termination right
trusts
upon
the author's death so as to expressly preempt other arrangements authors might make by testamentary will as well as state default rules of intestacy. 99 By way of another example, the anti-circumvention provi-
sions of the recently promulgated Digital Millennium Copyright Act 100 which prohibits manufacture or distribution of any technol('DMCA'),
ogy, product, service, or device that circumvents
ogy, does not make sense under the substantive component of the 'basics approach' because the transactions it regulates are better governed by
by the IP law of indirect infringement, discussed In short, the 'basics approach' would urge policymakers considering 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 wisdom
either state contract law or
earlier.
101
can be challenged under the 'basics once have acted, courts should apply the law on approach,' policymakers its terms and refrain from crafting new doctrines or from engaging in other
legislative or regulatory action
of some
179
effect
9.5
CONCLUSION
Dawson and earlier work by commentators such Bowman, Gordon, Kaplow, Lunney, Meurer, and Rich, we strike
is too restrictive and one that goes too 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 sepait is important to respect the different legal institutional frameworks of the various bodies of law that are involved - IP law, antitrust - each of which strikes its own balance law, and the general law 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,
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
antitrust law collaterally
for a positive law
distinct
framework that is more predictable and that captures the and important balances that are struck within each separate body
is
of law that
framework for understanding IP law and the broader interfaces that IP law shares with a number of bodies of law, such
offer, in
We
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 transactions 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. A further 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 institutional capability of courts to
80
Developments
in the
economics of copyright
for the
to facilitate 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,
what
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 contributions 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
Tim Bresnahan, Richard Epstein, Paul Goldstein, Wendy Gordon, Mark Lemley, Larry Lessig, Stanley Liebowitz, Glynn Lunney, Charles McManis, Michael Meurer, Pam Samuelson, 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).
Barton,
NOTES
1
.
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.
See, e.g.,
Carrier, Unraveling the Patent- Antitrust Paradox, 150 U. Pa. L. Rev. 761 (2002); Wendy J. Gordon, Intellectual Property as Price Discrimination: Implications for Contract, 73 Chi.-
L. Rev. 1367 (1998); Louis Kaplow, The Patent- Antitrust Intersection: A Reappraisal, 97 Harv. L. Rev. 1813 (1984); Glynn S. Lunney, Copyright and the Supposed Efficiency of First-Degree Price Discrimination (2002) (working paper); Michael J. Meurer, Copyright Law and Price Discrimination, 23 Cardozo L. Rev. 55 (2001).
Kent
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 spawn new sub-specialties, such as 'copyright and antitrust.'
3.
at the core
Donald
S.
4.
5.
6.
F. Scott Kieff, Principles of Patent Law 1066-1 155 (2001); Troy Paredes, Copyright Misuse and Tying: Will Courts Stop Misusing Misuse?, 9 High Tech. L.J. 271 (1994). We do not use the word 'basics' pejoratively, such as in the sense of an unduly simple characterization 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 may be useful in elucidating relative strengths and weaknesses of different strategies to shaping IP and other commercial laws outside the US as well. See generally Chisum et al. supra note 3, at 514-19, 1066-99 (discussing evolution of the non-obviousness and misuse doctrines). Although there was some variation in the way courts treated patents under these doctrines, a sufficient number of powerful courts (including the Supreme Court) were treating the patents that came before them so poorly under these doctrines that the expected
and
81
all patents plummeted. On non-obviousness, courts applied a tautological and unpredictable subjective decisional framework then called the 'requirement for invention': to be patentable, an invention had to constitute an 'invention.' This standard
value for
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' Jurgensen
vs. 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
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 decisional frameworks, see F. Scott Kieff, The Case for Registering Patents and the Law and
On
and 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
9.
employed a subjective standard. See Kieff, supra note 7, at 88-95. 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 treating so-called 'combination' patents differently is that all patent claims in the present US patent claiming system can be viewed as combinations of 'old elements.' See Kieff, supra
7, at
1 1 1
10.
how a claim operates as a simple logical list of elements and only found when each and every element on that list is present in the allegedly infringing product or process). 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 importnote
(explaining
is
that infringement
11.
ance of the Federal Circuit's creation for application of the Section 103 framework). 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 patent on 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 contributory infringement).
2.
13.
14.
15.
See infra note 42 and accompanying text. is, we do not suggest that legislative promulgation is itself infallible. See also infra note 23. 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). 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
That
misuse).
16.
IP rights are rights to exclude others from doing something. IP rights are not rights to 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
do
82
Developments
in the
economics of copyright
17.
18.
19.
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. 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). See generally F. Scott Kieff, Property Rights and Property Rules for Commercializing Inventions, 85 Minn. L. Rev. 697, 727-32 (2001). 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
tion,
would bring about the same output as under competibecause no customer willing to pay the seller's marginal cost would be turned
away. But perfect price discrimination is infeasible, and imperfect price discrimination 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 Competition 188-95 (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 (Richard
eds, 1989).
connecting, judicial analysis to established positive and normative decisional frameworks. 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 and accompanying text. For more on the importance of property rule protection of intellectual property and the legislative history of the present statu-
22. 23.
tory regimes, see Kieff, supra note 18. See infra notes 70-71 and accompanying text.
24.
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). Baxter, supra note 1 would require that the licensing arrangement be confined 'as nar,
and administration permit.' Baxter, rowly and specifically as ... technology 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 compresumably rejecting the views of parties petitively superior to other available options 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 many
factors.
For an interesting comparative institutional analysis that stresses the importance of and predictability in judicial decision making, see Cass R. Sunstein and Adrian Vermeule, Interpretation and Institutions, 101 Mich. L. Rev. 885 (2003).
certainty
83
25.
26.
27.
See generally Phillip Areeda and Louis Kaplow, Antitrust Analysis: Problems, Texts, Cases 174-250, 447-77, 785-806 (1997). United States vs. A luminum Co. of America, 1 48 F.2d 4 1 6, 430 (2d Cir. 1 945). For more on the rule of reason generally, see Areeda and Kaplow, supra note 25, at
28. 29.
id. at 663-784. 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 undisputed, 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, and if that consequence is one that the owner may impose unconditionally, he may impose The domination [over a material used in a it conditionally upon a certain event. patented device] is one only to the extent of the desire for the [patented device].
. . .
vs. Universal Film Mfg., 243 U.S. 502, 519-20 (1917) (Holmes, 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
J.,
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.
31
32.
See generally Areeda and Kaplow, supra note 25, at 413-44, 686-784. 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 mentioned here, which, of course, they are not. As discussed infra in Section 9.4 (third subsection), 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 restrictions 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 constrained, advised by counsel, and fully possessed of contractual intent (and therefore not a good candidate for the contract law defenses to formation of unconscionability, adhesion, 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 type of one-two punch in the case of IP, see infra note 85 and accompanying text. 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 commercialization 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 commercialization theory and its associated focus on coordination because at a minimum
What
84
Developments
in the
economics of copyright
theory did motivate the shaping of present IP regimes as a historical fact and because we see the commercialization effect as the most important in that the regimes can 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
this
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.
tive
theories
Rev. 267 (1996) (same). For recent reviews of the copyright literature on incenand a collection of sources, see Michael Abramowicz, Copyright
Redundancy, George
Mason Law
http://papers.ssrn.com/sol3/results.cfm (reviewing and collecting sources and highlighting the opportunity cost issues discussed by Lunney as well as showing how additional
33.
34.
35.
36.
37.
works on the margin may contribute little while at the same time causing rent dissipation); 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. Innovation is a broader term than invention and is generally understood to include the downstream dissemination of inventions. It is sometimes also called commercialization. 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). For a thorough model of the commercialization goals of IP law, see Kieff, supra note 18. 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 27 1 of the Patent Act of 1952, 35 J. Pat. Off. Soc'y 476 (1953). 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)).
35 U.S.C.
271(c) (contributory patent infringement).
38. 39.
To be
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
sure, the ease
patent right to exclude hinges on the body of law governing the field called 'claim construction,' which is presently the topic of substantial debate because it is considered by
many
more on claim
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 Records, Inc. vs. Napster, Inc., 239 F.3d 1004, 1020 (9th Cir. 2001); In re Aimster Copyright Litigation, 334 F.3d 643 (7th Cir.
A&M
2003).
40.
41.
42.
952 Act, courts had used the misuse doctrine to erode the ability for intelowners 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). Dawson Chem. vs. Rohm and Haas Co., 448 U.S. 176 (1980). See 35 U.S.C. 271(d)(4^5) (added by Pub. L. No. 100-703, 201, 102 Stat. 4676 (1988)).
Before the
1
lectual property
In
its entirety,
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
85
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.
342 F.3d 191, nor this Court has affirmatively recognized the copyright misuse doctrine [tjhere is ... a wellestablished patent misuse doctrine, and other courts of appeals have extended the
203-04 (3rd
Entertainment, Inc., See, e.g., Video Pipeline, Inc. vs. Buena Vista Cir. 2003) (noting that although '[njeither the Supreme Court
. .
.
Home
44.
doctrine to the copyright context.'). See generally Carl W. Schwarz, The Intellectual Property IAntitrust Interface, in 1 No. 6 7 No. 6 ANIPLR 15 (copyAndrews Intell. Prop. Litig. Rep. 15 (2000), available at
WL
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:
45.
is claimed in the patent. Copyrights give the copyright holder the right to restrict copying of the creative expression 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, Patents for 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). See supra note 45. 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).
Licensing and Litigation (2003). Patents give the patentee the right to restrict use of what
46.
47.
48. 49.
Id.
50.
51
52. 53.
For more on property rules versus liability rules in the context of IP, see, e.g., Kieff, supra note 18, at 732-33. 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). 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. Texaco, Inc. vs. Pennzoil, Co., 729 S.W2d 768 (Tex.App. 1987). 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
(i.e.,
54.
Wallace
vs.
86
Developments
in the
economics of copyright
history of contributory infringe-
55.
See Chisum et al., supra note 3 at 950-55 (discussing ment doctrine and the role of the Wallace case).
56.
Because the chimneys are usable with non-infringing lamps and are not specially adapted for infringing uses, their sale falls within the safe harbors of Section 271(c). See supra notes 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. 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. 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'). 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. They may 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 customer of both the plaintiff and the defendant for both breach of contract and patent
57.
That
58.
59.
&
60.
sufficient facts to
mount such an
argument.
61.
62. 63.
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). Henry vs. A.B. Dick Co., 224 U.S. 1 (1912). These cases supported the improper view that causes of action for patent infringement could 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 unduly stretch 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. 65.
See infra note 65 (reviewing history of these cases). As the Supreme Court later pointed out in Dawson, when Rich was testifying in support of 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 Subcommittee No. 4 of the House Committee on the Judiciary, 81st Cong., 1st Sess., 17 (1949) (testi-
'
mony
As
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 on
a film projector and a restrictive label contract limiting use to certain film)). Compare Motion 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
187
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 14. Not surprisingly, in United States vs. United Shoe unpatented.' 15 U.S.C. 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) (75 AT), the Court found a set of complex leasing arrangements 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 historical context. A significantly more complete treatment is provided in Rich, supra note
51, at
66.
24 1-283.
Leeds
&
Catlin Co.
vs.
Victor Talking
67. 68.
Id. at 335.
69.
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). A determination 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 something outside 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 manner that is not only less costly to the patentee but also to all third parties by simply drafting a 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)). 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 was followed 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').
71
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.
73.
Mercoid Corp. vs. Mid-Continent Investment Co., 320 U.S. 661 (1944) ('Mercoid T), and Mercoid 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 generally as Warfield} (patent on use of lecithin as an emulsifier in chocolates to improve its properties by, for example, preventing 'whitening' after only a few days). See, e.g., Dawson Chem. vs. Rohm and Haas Co., 448 U.S. 1 76, 214 (1980) ('Respondent's
'
method 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 subsections (a), (b), and (c) those acts that would constitute direct, induced, and contributory
infringement, respectively; while at the
in
it
188
Developments
would not be misuse
subsections
(a), (b),
in the
economics of copyright
74. 75.
76.
for a patentee to sue or license anyone who could be sued under or (c). See supra notes 36-39 and accompanying text. 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 by merging the Court of Claims with the Court of Customs and Patent Appeals and transferring 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). Panel Co., 1 33 F.3d 860, 869 (Fed. Cir. 1997) (internal citaVirginia Panel Corp. vs.
MAC
77.
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., Morton
Salt Co. vs. G. S. Suppiger Co., 314 U.S. 488, 491 (1942), and arrangements in which a patentee effectively extends the term of its patent by requiring post-expiration royalties, 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 arrangement does not constitute patent misuse. See 35 U.S.C. 271(d)(5) (1994) (added by Pub.L. No. 100-703, 201, 102 Stat. 4676 (1988)).
78.
133 F.3d at 869 (internal citations shortened). For an expanded discussion of this point in the context of copyright misuse, which derives from patent misuse, see Paredes, supra note 3.
Rich, supra note 29, at 245.
79. 80.
ProCD,
in
Inc. vs. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996) (non-commercial use restriction shrink-wrap copyright license for computer program held valid and enforceable as a
81
contractual limit on use). 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
82.
unauthorized acts may support suit for infringement). 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 and
the
83.
84.
along the lines of the procedural aspects of our 'basics approach' discussed supra notes 1 3 and 23 and accompanying text. 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 protected by some particular piece of IP. 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 payoff of the different states of the world under which patent rights might or might not materialize. Aronson vs. Quick Point Pencil Co., 440 U.S. 257 (1979) (contract to pay royalty on a technology was enforceable even though no patent ever issued on the technology where at the time the contract was entered into the technology might have been patented and the contract provided a low royalty rate for the case where no patent issued and a higher rate for the case where a patent did issue).
if at all,
189
85.
See, e.g., Scheiber vs. Dolby Labs., Inc. 293 F.3d 1014 (7th Cir. 2002) (Posner, J.) (discussing 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 hoppicking 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. Chisum et al., supra note 3 at 1 1 55. See generally id. at 1 1 55-96 (reviewing preemption). Nor do IP rights give IP owners any affirmative right to use. 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
entitled 'The
Wm.
symposium
89.
90.
91
United States Supreme Court's Effect on Intellectual Property Law This Millennium' at William Mitchell College of Law held April 24, 2004). 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 preemption 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); Compco Corp. vs. Day-Brite Lighting, Inc., 376 U.S. 234 (1964). As reviewed in detail throughout Chisum 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. Bonito Boats, 489 U.S. at 168. 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.'). 489 U.S. at 168. Also consider the real estate analogy discussed supra note 3 1 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
,
.
190
Developments
in the
economics of copyright
95.
96.
97.
98.
yard has sufficient nexus to interstate commerce that it may be regulated by Congress using commerce clause power). 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 Century Fox Film, 539 U.S. 23 (2003) (holding that the Lanham Act does not prevent unaccredited copying of uncopyrighted work and expressing concerns that otherwise the Lanham Act 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 evidentiary 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). 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). 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 publicizing an 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). 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 Copyright Act, 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. See 17 U.S.C.
99.
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.
101. See supra Section 9.3,
1201(a)(lH2), (b)(l) (2000). second and third subsections (discussing indirect infringement
respectively).
and contract,
Index
A&M Records, Inc.
A.B. DickcasQ 171
access
differentiated products theory of copyright 106-107, 114 side of tradeoff 103, 105 advertising revenues 55
vs.
Napster, Inc. 5
radio 54
television 54, 56
Abramowicz, M. 104
Bonanno, G.
1 1
and
Borch, K. 31
Borenstein, S. 109 Boyle, J. 104
agency rule-making 9-10 Agreement on Trade Related Aspects of Intellectual Property Rights
(TRIPS) 9, 15, 120,135 American Society of Composers, Authors and Publishers (ASCAP)
38
ancillary products, tax
on
Brennan, D.J. 108 Brennan, T. 82 broadband 80 Broadcast Music, Inc. (BMI) 38 Brown, R. 80
cable operators 40, 41 cable transmission 55
56
antitrust law 159, 161, 165-6, 175-6
121, 131
Canada
14
Audio
81
Coase theorem 23
codified
knowledge
I.E.
47
Cohen,
authorship
commons
104 134
knowledge as 123
Bakos, Y. 64
'tragedy of the
commons'
122, 123
Baumol,W.J. 115
Berry, S.T. 113
Besen, S. 26, 32, 64, 82, 104 blanket licenses 39, 40, 50, 53
competition, and copyrighted works 104 competition law 144-5 vs. copyright 142
collectives
797
192
Developments
in the
economics of copyright
income 23-36
mispricing by 57
8,
10-14
'structure, sequence,
and
1
mechanisms 28
to replace current system
40^1
among
different aspects
113-14
letters
and
diaries 2
markets 47
(CTEA)6,
111
maximal
minimal 120, 132, 133, 136, 137 misuse 167, 175 models of design 81-2 nexus between industry and policy makers 4
open-ended rules 10 perpetual term 7 and policy 3-4
special exemptions 2
and Eldred 6-8 Cournot market structures 104 Cournot oligopoly 99 Court of First Instance 151, 152,
154
creation costs 98
'creative
153,
commons'
cultural differences 4
competition law 142, 143-6 weakness in economic theory 128 welfare-enhancing effects of 121 Copyright Act 62 Copyright Arbitration Royalty Panel
vs.
cultural
economics 2
(CARP)
48-50
as coalitions 25
database development, and copyright law 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
comparison to blanket licenses 54-7 distribution mechanisms 25 division of revenue 52^1 and efficient consumption 45-6
functions of 23
copyright 103-19
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
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 849 gross benefit 87-8, 95-6 net economic welfare 92-4
9,
63
offs 83,
management'
(DRM)
57,
63, 64,
77-8
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 effects 65-6
fingerprinting 63
62
network
64
watermarking 63
distributional concerns 104
Dixit, A.K. 108 dominant firm structures 104
Fitzgerald, D. 151
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
differences in 5
45
M.
128
hacking 63
Hardin,G 105,121
Hart, O.D. 106 Hay, D. A. 115 Hayek, F. von 129
123
Heatoncase 170-71
Hettinger, E.C. 124
194
Developments
in the
economics of copyright
intellectual property rights 130, 133
Hughes,! 110
Hull, D. 151, 152
Internet piracy
IBM
13
vs.
and
CDs
60
profits
74-6
142-58
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
,
151,
technological
and
legal aspects
153
reliance
60-64
technological protection 66, 68, 72 ISPs, tax on 52, 56, 59
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 technologies 67-8 infringement under IP law, vs. sui generis\aw 171-5 innovation 161, 184
'integer problem' 115 intellectual creations, as public
Kazaa 76
Kirby, S.N. 64, 82, 104 Klein, B. 82
Klemperer,
P.
110
knowledge
anthropologists on 125
codified
knowledge
goods
133
collective
1,23,43,58,127
intellectual products intellectual property
43^
5
economic aspects
US
166-7
24-7
knowledge
3,
knowledge creation,
136
collective setting
87
framework 162^
Index
195
Ladbrokecasc
Landes,
149, 151-2
multiple works
1 1
W.M.
122, 143 learn-then-distribute rule 28, 29, 30, 33 Leeds & Catlin case 1 72-3
Lehmann, M. 144
Lemarchand,
Lessig, L. 40
letters
S.
143, 145
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
and
diaries,
copyright protection
Liebowitz,
S.J.
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
network
effects, file-sharing
Mankiw,N.G. 106
market failure 122, 130 market inefficiencies 44 market power 106 markets 46 copyright 47 Maskus, K. 144
technologies 65-6
network externalities
4,
138
Neumann, M.
non-authorized audio-streaming
technologies 60, 61
maximal copyright
monopolies, as copyrighted works 83 monopolistic competition theory 105 monopolists, authors as 103 moral hazard 33 Morin, E. 128 MP3.com 76, 77
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
MP3s
and 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
65
10
196
Developments
in the
economics of copyright
Ramello, G. 144, 145 Rate Courts 54
real-world difficulties 46-8
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
545
performing rights societies 39^0, 53, 54,56 performing rights tariffs 38, 54-5 Perloff, J. 99 perpetual term copyright 7
Perry,
(R&D)
costs
1,10
restrictive licensing
arrangements
176
M.K.
108
(RERCI)
15
policy guidelines 133-5 Posner, R.A. 64, 103, 104, 105, 121, 122, 143
record industry 55
Rumelt, R.P. 65
preemption 176-7
Prescott, E.G. 115
Samuelson,
scarcity 124
P.
63
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
vs.
Accolde, Inc.
duration of 1 32 and Locke's totemic paradigm 134 on outputs 123 public choice problems 5-6
public
on
good
1,
information as 122
intellectual creations as
127
models
1 1
Spector, H. 144
Spence,
status
M.
108 108
quo systems 42
blanket licenses 54
excess entry to industry 113
Stiglitz, I.E.
Index
197
sui generis 1 \ -2
Judiciary
knowledge
3,
New
8
Takeyama, L.N. 65, 66, 82 tax, on ancillary products earlier twice 41,45,51-2,56
tax-and-subsidy system 41
television
(CONTU)
47
115
P.
Visscher,
tariff 55, 59 Pennzoil, Co. 169
M.
performing rights
Texaco, Inc.
vs.
Waelbroeck,
Waldfogel,
J.
64, 66
113
65, 69, 82, 103, 105
Thisse, IF. 65
Waldman, M.
watermarking,
Torremans, P.L.C 143, 144, 150 'tragedy of the commons' 122, 123
transactions costs 23, 24, 25, 27, 32, 34,
management'
Waterson, M. 115 Watt, R. 9, 104 Weber, M. 125
(DRM)
63
82
and
fair
Treacy,P. 149,150
Whelan Associates
vs.
Jaslow Dental
US
Congress
constitutional powers 7
9, 15
Copyright Office
3,
9-10
Stanford
Law
Libre
Iflb