Copymiddle

FYI - Jonathan Franklin

Jonathan Franklin			801 Monroe Street
Reference Librarian/Foreign Selector	Ann Arbor, MI 48109-1210
University of Michigan Law Library	Tel. 734-764-6150
jafrank[_at_]umich.edu			Fax. 734-764-5863

> Copyright Incentives: Neither Left Nor Right But Center
>
> Marshall van Alstyne
> University of Michigan
> School of Information
> 304 West Hall
> 550 East University Ave.
> Ann Arbor, MI 48109
> mvanalst[_at_]umich.edu
>
> Douglas Lichtman
> University of Chicago Law School
> 100 York Street Suite 12-P
> New Haven, CT 06511
> douglas.lichtman[_at_]yale.edu
>
> In 1983, the Free Software Foundation (FSF) introduced the concept of
> "copyleft", a copyright licensing scheme designed to reshape copyright
> law for application to computer software. Copyleft was motivated by a
> simple observation: software copyrights limit the freedom of developers
> to build on the works of others. The intellectual property protection
> on works of expression - books, plays, and paintings, for example -
> function badly in industries that depend on incremental innovation and
> reuse for product improvement, customer satisfaction, and network
> expansion.
>
> So, copyleft was proposed as the following alternative. Under copyleft,
> programmers render their source code to the public, then use existing
> copyright law to ensure that anyone redistributing the code, even with
> enhancements, does so without adding any restrictions on further
> distribution. Copyleft limits the rights of developers to limit others'
> rights of copying and redistribution. According to the FSF General
> Public License:
>
> "You may modify your copy or copies of the Program or any portion of it,
> thus forming a work based on the Program ... provided that you also
> ..... must cause any work that you distribute or publish, that in whole
> or in part contains or is derived from the Program or any part thereof,
> to be licensed as a whole at no charge to all third parties under the
> terms of this License." (sec. 2.0 - 2.b)
>
> While the copyleft approach has helped several popular UNIX-based
> programs enjoy distributed development and widespread use, it introduces
> a second problem. It limits the economic incentives to make follow-on
> investments. Freedom from copy restrictions can imply freedom from
> income. While, in principle, copyleft permits a software developer to
> charge a distribution fee for the bundled software good and its
> derivative innovation, in practice, the requirement of full disclosure
> prevents an author from deriving income from more than a single sale.
> Having sold the software once, a seller must then compete with source
> code available from the first buyer. Such a seller has no market power
> selling against an identical copy of their own product. For software,
> this results in a price at or near marginal cost which approaches zero.
>
> In this abstract, however, we would like to suggest that hidden in the
> copyleft logic is a powerful and as-yet unrecognized business strategy.
> We call that strategy copymiddle. The intuition is straightforward:
> software companies often find themselves in a race both to rapidly grow
> their user network and to quickly upgrade their software products. Many
> of these companies cannot keep pace on their own and would be willing to
> turn to other firms for assistance. In short, these early firms
> recognize that they are better off winning a small percentage of a
> larger market than they are hoarding control over a smaller one. A 5%
> stake in a $1,000,000,000 network pays more than a 95% stake in a
> $1,000,000 network.
>
> Accordingly, we introduce a profit incentive period into a copymiddle
> license which an innovator uses to distribute their software and promote
> complementary investments in product and network growth. Under the
> copymiddle licensing strategy, an innovator would permit the creator of
> a derivative work to retain interest in the subsequent improvements for
> a reasonable period - seven years, for example - before shifting to
> copyleft. Upon expiration of the copymiddle license, the contract would
> specifically eliminate all restrictions on redistribution, ensuring that
> both customers and third party innovators could enjoy free access. All
> parties would then be free to incorporate second tier innovations in
> their own third tier enhancements without fear of hold-up. The creator
> of any third tier derivative work would then enjoy the same rights to a
> profit incentive period as any investor at the second tier. With
> copyrights offering a 75 year statutory period of protection (Section
> 302(c) of the 1976 Copyright Act), an innovator could oversee 10 or more
> rounds of product and network growth. Whereas copyright maximizes
> one-time innovation through the promise of durable monopoly rents and
> copyleft maximizes perpetual distribution through the promise of free
> access, copymiddle strikes a balance through a combination of rents and
> access.
>
> From a broad societal perspective, copymiddle licenses are attractive.
> First, it harnesses market forces to dynamically shorten the effective
> period of exclusion and to increase investment. This strategy has the
> unusual advantage that an inventor, in this case a first author of
> software, may find it economically rational to forgo monopoly rents
> during the tail of the statutory period on a first generation product.
> In exchange, he or she stimulates demand from an expanded network of
> co-investors and consumers who also represent a larger market for a
> second generation product. Although long periods of exclusion limit
> complementary investments for all copyrightable subject matter, the
> problem is particularly acute for software. New functionality is more
> easily developed on the basis of reusable subroutines. Moreover,
> proprietary interests in the form of copyrights hinder the computer
> community's ability to settle on standards.
>
> Second, copymiddle helps solve the problem of enablement. Under
> existing statutes, a copyright petitioner need only submit portions of a
> software good. He or she may withhold or block out segments of the
> source code thereby keeping the key intellectual property secret. This
> stands in contrast to both the requirements of patent law and the
> expectations of traditional copyright law. For patents, the inventor
> must file a "written description of the invention, and of the manner and
> process of making and using it, in such full, clear, concise, and exact
> terms as to enable any person skilled in the art ... to make and use
> the same." (35 U.S.C. sec. 112). For copyright on traditional
> publications, a work itself establishes on the face of it the nature of
> its contribution. The bargain implicit in most intellectual property
> law is that society grants an innovator a limited monopoly in exchange
> for revealing to society the nature of the innovation. Requirements for
> software are thus unique in that the mechanism of operation is not
> disclosed, neither on the face of it as for works of expression, nor in
> the filing of it as for patents. Copymiddle suggests that an inventors'
> best interest may lie in enabling others to enhance and build upon an
> original work.
>
> Third, consumers improve their access to existing products when the
> product becomes freely distributable. They also benefit from earlier
> introduction of new products. Both effects imply a strict Pareto
> improvement in consumer welfare. The copymiddle strategy might
> therefore represent a more balanced approach to intellectual property
> that accounts for both efficient distribution under a static framework
> and efficient innovation under a dynamic framework.
>
> Branscombe, A. W. (1995). Who Owns Information? Cambridge, MA: MIT
> Press.
>
> Davis, R., Samuelson, P., Kapor, M., & Reichman, J. (1996). A New View
> of Intellectual Property and Software. Communications of the ACM,
> 39(3), 21-30.
>
> Gilbert, R., & Shapiro, C. (1990). Optimal Patent Length and Breadth.
> Rand Journal of Economics, 21(1), 106-112.
>
>
> CREW's Web page is at: http://crew.umich.edu/
>
>
> _______________________
>
> Nathan Bos
> Postdoctoral researcher
> Collaboratory for Research on Electronic Work
> 701 Tappan Street
> Ann Arbor, MI 48109-1234
> (734) 647-7730
> http://www-personal.umich.edu/~serp/
> _______________________
Received on Tue Oct 13 1998 - 18:17:46 GMT