The Trend Toward Strengthened Intellectual Property Rights: A Potential Threat to Public-Good Uses of Scientific Data
Laws and regulations, both national and international, affect the flow of scientific information through electronic networks. Among them are rules regarding liability for false or misleading information, laws protecting individual privacy rights, and export controls. This chapter focuses on intellectual property policies and their expression in laws and regulations affecting the contents of databases, because changes are now afoot that may erode the relatively privileged position science has held within the existing legal framework.1 By restricting scientists' full and open access to the data on which future advances depend (see Box 5.1), these changes could impede the progress of science and thus limit the contributions that science can make to society, notwithstanding the constitutional mandate that intellectual property rights should be limited in time and should advance science and the useful arts.2
One such change is that governments, including our own, are finding it increasingly difficult to maintain the rate of growth that publicly funded science has enjoyed over the past half century. When scientific research is supported instead by private funding, the end results and perhaps the research itself are likely to be kept proprietary. Furthermore, there are indications that the scientific data management that governments continue to fund may well be carried out as if it were proprietary, in the sense that fees for use of the data may exceed the costs of dissemination. As their tax bases decline and governments come to regard their data collections as possible sources of revenue, they have, in some instances, adopted the same short-term, profit-maximizing strategies as private firms. Yet today, when commercially valuable data of scientific importance are made available in electronic form, they also become available for rapid, inexpen
BOX 5.1 Effects of Government Support on U.S. Research and Data Activities The ability of private-sector technological development in the United States to thrive without the kind of centrally organized institutional framework and industrial policy apparatus typical of the European Community and many other nations1 has stemmed in good measure from the large public investments in basic research and development that were made after the late 1950s, in response to Cold War pressures and national security interests. In retrospect, the success of the U.S. innovation system, despite its apparent anarchical character, can be seen as linked to public funding of academic institutions and specialized laboratories, whose research product has paved the way for private industrial applications.2 In this context, the fact that federal funding also largely defrayed the costs of collecting and disseminating raw and elaborated scientific data merits particular attention. Throughout the Cold War period, and extending into the present, the U.S, government has reinforced its subsidies of fundamental research with a policy of open exchange of scientific data. This policy was promoted internationally through the government's bilateral science and technology cooperative agreements and increasingly in recent years through both bilateral and multilateral agreements concerning various large-scale research programs and projects. None of these agreements, however, has broadly encompassed all scientific research activities. Instead, they typically have been limited to scientific cooperation and related protocols for the exchange of data according to the special interests of a geographic region, scientific discipline or subdiscipline, or specific projects undertaken by the parties to the agreement. In the rise of the United States to become the world's leading producer of technological goods and scientific information, the government's role in ensuring an open supply of data to the scientific community under favorable economic conditions has been a constant stabilizing factor.
|
sive copying and manipulation. While this facilitates value-adding uses from one perspective, from another, it undermines the data provider's ability to recover costs, much less to generate a profit.
A second change is that, in many areas of research, the separation has diminished between basic research, where intellectual property rules are more concerned with attribution of ideas and findings than with the appropriation of published material, and applied research, where intellectual property and proprietary concerns predominate. This conjunction has been especially evident in computer
science and biotechnology, where some basic advances are now virtually inseparable from their industrial applications. The granting of patents or other exclusive property rights in these industrial applications can affect the ability of other researchers to test and extend the theories underlying them.
Third, the revolutionary convergence of digital, computing, and telecommunications technologies has profoundly altered the preexisting status quo.3 The potentially large gains and losses from the commercial exploitation of data under these changing conditions have led to a concerted drive for new and stronger forms of legal protection for publishers of electronic databases in general, including compilations of scientific data that were heretofore treated as components of the public domain. 4
The current trend toward stronger and more enduring intellectual property rights, and fewer limitations on the rights of copyright holders vis-á-vis public-good uses of information, could reduce some of the limitations that have benefited scientists, and on which they have relied. Government studies of the challenges that digital technologies pose for intellectual property law at both national and international levels have stimulated calls for strengthening intellectual property rules. In addition to legislation either adopted or still under consideration in the United States and other nations, proposals to strengthen international copyright and related laws were a major focus of multilateral negotiations sponsored by the World Intellectual Property Organization (WIPO) in December 1996, largely at the urging of the United States and the European Union.
One of the draft treaties currently being considered calls for worldwide adoption of a new form of intellectual property protection for the contents of databases.5 Although this treaty was scheduled for discussion and approval at the WIPO Diplomatic Conference held in Geneva, Switzerland, on December 2-20, 1996, the conference delegates decided that it required further study. Future adoption of a treaty with similar proposals would have such profound consequences for transnational exchanges of scientific data that the committee chose it as a principal focus of this chapter.
This chapter begins by briefly describing the relevant legal infrastructure during the predigital period and by identifying certain destabilizing factors, such as the introduction of electronic photocopying machines. It then outlines digital technologies' role in accelerating the rise of information as a commodity to be bought and sold and in thus upsetting the previous imperfect balance between under-protection and overprotection of the rights of data creators and holders. The discussion that follows examines the emerging legal responses to these phenomena. It describes the current legislative and treaty proposals in detail and explores the implications for science of new proprietary rights in databases. The chapter concludes by proposing actions that groups representing the research and education communities should undertake to stimulate reformulation of the legislative and treaty proposals, with a view to reconciling the need to protect the legitimate
interests of database makers with the need to protect the activity of science and to ensure its ongoing contribution to the public interest.
SALIENT FEATURES OF THE PREDIGITAL STATUS QUO
Because the creation, collection, and dissemination of scientific data in the United States have in large part been subsidized by government funding, they have not depended on the balance between incentives to create and efforts to preserve free competition that intellectual property law normally governs. Within this framework, most academic compilers or generators of scientific data were more concerned about obtaining credit or recognition for their contributions than about securing the economic fruits of their efforts.6 Only in cases where members of the scientific community authored discursive scientific works or otherwise participated in applied technological innovation, or where commercial publishers compiled value-added databases, were they likely to be affected by legal rules governing commercial applications of data. In such cases, existing legal institutions proved relatively stable in the predigital epoch, and the scientific community has taken this stability largely for granted.
In the private sector, by contrast, commercial compilers of data have long suffered from a risk of market failure owing to the intangible, ubiquitous, and, above all, indivisible nature of information goods and to the ease with which free riders may have appropriated the fruits of the compilers' investment, once the information goods were made available to the public in print media. Despite this risk, the domestic and international intellectual property systems responded laconically, if not with indifference, to the compilers' dilemma. 7 This indifference stemmed in part from the inability of the worldwide intellectual property system to match compilations of data to the basic subject matter categories covered, respectively, by the Paris Convention for the Protection of Industrial Property (1883) and the Berne Convention for the Protection of Literary and Artistic Works (1886).8 It also stemmed from a concomitant reluctance to fetter the basic building blocks of scientific and intellectual discourse with legal impediments.9
Notwithstanding these infirmities, the commercial exploitation of nonscientific data and of published compilations of information prospered in some developed countries, notably the United States and the United Kingdom (where copyright protection is sometimes available). Whether there would have been greater commercial exploitation of scientific data in the past if publishers could have invoked stronger proprietary rights is a matter of conjecture. Patents are seldom available for database contents because writings are not patentable subject matter and also because the largely incremental character of database development would typically make it hard to meet eligibility requirements. Even contract law has significant limitations when mass-market information products are sold to persons outside the scope of a contract.10
Data as Know-how Applied to Industry
To obtain some measure of protection, firms engaged in industrial applications of scientific discoveries have entrusted their commercially valuable data to trade secret law or to equivalent laws of confidential information. Trade secret law (or equivalent laws of confidential information) provides innovators and investors with no exclusive property rights at all. Rather, it permits third parties to reverse-engineer any unpatented industrial product by proper or honest means in order to reveal the process by which it was obtained, and to use that process to manufacture equivalent goods.11 To the extent that an innovative product is derived from commercial applications of scientific data kept under actual or legal secrecy, a competitor always remained free to generate the same data and to apply it to similar products or uses.12
Trade secret law thus provides qualifying originators with no legal immunity from direct competition. It merely confers a ''head start," that is, an uncertain period of natural lead time, during which originators seek to recoup their investment in research and development while establishing their trademarks as symbols of quality that consumers recognize. In this and other respects, trade secret law operates as a liability regime that discourages certain types of conduct rather than as an exclusive property right that may create a legal barrier to entry (see Box 5.2).
When scientific data are disseminated to the public in print media, they normally forfeit the protection of trade secret law, or related laws of confidentiality, except insofar as two-party contracts may otherwise provide. Not surprisingly, commercial compilers in such cases have sometimes found it difficult to appropriate the fruits of their investment unless either copyright laws or unfair competition laws afford them a limited shelter against wholesale duplication by third parties.
Copyright Law as a Cultural Bargain
The advent of the printing press created for published literary and artistic works markets that had previously existed only in a rudimentary form, owing to the need to produce each copy of a work by hand from a single original. Paradoxically, to promote markets for information goods and other literary and artistic works, the state intervened by erecting new monopoly rights-intellectual property rights-even as it removed the royal privileges and guild monopolies pertaining to tangible goods that were handed down from the Middle Ages.
Information goods have the properties of so-called public goods-they are nondepletable and nonexcludable. A second comer's use of a new information good does not diminish or exhaust it; once it is disclosed to the world, anyone can use it without the originator's permission and without reimbursing him or her for the costs of research and production. Unless the state limits the ability of third parties to copy a given literary production, for example, and to sell the copied
BOX 5.2 Definitions Liability Rule and Exclusive Property Rights A property right precludes third parties from appropriating the object of protection, whereas a liability rule regulates the means by which they can engage in certain potentially harmful acts on certain conditions. 1 If one has "rightful possession of some thing-—such as an automobile or a home" under an exclusive property right, "another person ordinarily cannot take it without permission"; but a liability rule permits others to engage in acts that "create risks of harm and thus constitute probabilistic invasions of property interests" (such as nuisances), while obligating them to pay damages for harm under specified circumstances. 2 Sul generis Sui generis means "of its own kind or class" (Blacks Law Dictionary 1434, West, 6th ed., 1990). The literature refers to special-purpose intellectual property laws that deviate significantly from the classic patent and copyright paradigms as ''sui generis" regimes. See, for example, Pamela Samuelson (1985), "Creating a New Kind of Intellectual Property Law: Applying the Lessons of the Chip Law to Computer Programs," Minn. L. Rev., 70:471 (discussing the sui generis character of the Semiconductor Chip Protection Act). Subpatentable A subpatentable innovation is novel in the sense of being new, but it represents a step in technical progress that an engineer might be expected to make in due course. By definition, a patentable invention must be "nonobvious" in the sense that it represents a breakthrough beyond the capacity of a routine engineer to make in due course.3 In simpler terms, patents are supposed to reward extraordinary achievements, while subpatentable Innovations are those that proceed in small, incremental steps.
|
good for less than the price charged by the originator, neither the author nor the publisher may have sufficient incentives to create or invest in the dissemination of cultural and information goods. 13
The historical solution to this problem has been the mature copyright system, which charges both authors and their publishers a price for overcoming market failure. In effect, copyright law has enabled the state to impose "portable fences"
that accompany intangible creations and that limit what purchasers can do with them, even though they possess the physical artifacts, such as books or printed tables of numbers, in which these intangible creations are embodied.14 In so doing, the state also has imposed legal constraints on authors and publishers—a cultural bargain—that has attempted to balance incentives to create against the public interest in both free competition and access to the copyrighted culture. 15
For example, although copyright law protects an author's personal expression for a relatively long period of time, it attaches only to "original works of authorship." In principle, this requirement excludes functionally dictated collections of data that fail to manifest a creative selection or arrangement.16 Moreover, copyright law never prevents third parties from independently creating their own versions of another author's unprotectable ideas or of the factual discoveries presented in a given scientific publication. In other words, copyright law protects only a given author's style, not his or her factual or ideological content.
The Concept of Fair Use
The mature copyright paradigm further curbs even this limited monopoly by relaxing the author's control over certain uses of great public interest. Thus, numerous exceptions to and limitations on the copyright owner's bundle of exclusive rights favor face-to-face teaching (e.g., by allowing limited duplication of materials for classroom use), library and archival uses, and selected public interest pursuits, 17 in addition to a general "fair use" exception "for purposes such as criticism, comment, news reporting, teaching ... scholarship, or research."18
While the availability and scope of statutory exceptions usually vary with the nature of the subject matter at issue, the fair use exception applies to all subject matter categories across the board. Even so, overriding the copyright owner's exclusive rights in the name of fair use remains an atypical result contingent on a judicial evaluation of the special "purpose and character" of the use, the "nature of the copyrighted work," the ''amount and substantiality of the portion used," and the "effect of the use upon the potential market for or value of the copyrighted work.”19
In recent years, the advent of new technologies—from photocopying machines to computer programs and optical scanners—has unsettled the doctrine of fair use20 by enabling even copies for private research uses to displace commercial markets,21 and also by making it possible to overcome most of the transaction cost problems that increasingly had been used to justify application of the fair use exception in practice.22
Protection Afforded
Copyright law will not protect the product of a compiler's industrious efforts—i.e., of labor, skill, or investment—if the selection or arrangement it em-
bodies does not rise to the level of an original work of authorship. Moreover, a mature copyright system usually affords protection only against wholesale copying of the original selection and arrangement underlying any eligible compilation of data. In the United States, this doctrine of weak or "thin" protection for factual works has been reinforced by First Amendment concerns, which some courts and commentators viewed as mandating broad access to the disparate facts that result from a compiler's efforts.23 When these doctrines apply, they greatly diminish the value of copyright protection even to database publishers who satisfy the eligibility criteria, because their exclusive reproduction and derivative work rights—as construed by the Supreme Court in Feist Publications, Inc. v. Rural Telephone Service Co.—will not normally prevent unauthorized extractions of disparate data for either competing or value-adding uses.24
Some federal appellate courts, however, have begun to rebel against the Feist decision and to reinstate stronger copyright protection for factual compilations and databases by subtle doctrinal manipulation. 25 Whether state or federal unfair competition laws could also provide some supplementary relief against the unauthorized copying of commercially valuable data that are not protected by trade secret or copyright laws remains an unsettled question, although such laws are sometimes invoked both here and abroad.26 In any event, this cyclical fluctuation between states of underprotection and overprotection is a characteristic trait of borderline subject matter that fits imperfectly within the classical patent and copyright paradigms, such as the contents of databases.27
DIGITAL TECHNOLOGY—DISRUPTING THE BALANCE OF PUBLIC AND PRIVATE INTERESTS
Despite (or perhaps because of) the relatively weak legal infrastructure governing use of data, a thriving market for compiled information has grown up, and U.S. publishers appear to play a dominant role in it,28 although it is important to emphasize that this market has been largely concerned with nonscientific data and information. This industry seems largely characterized by niche marketers who supply and dominate specific market segments. The limited size of these segments and the relatively high startup and servicing costs seem to deter second comers from readily entering such markets.29 In other words, once the threshold level of investment has been crossed, the first comer tends to take the relevant market segment as a whole.
The public sector nonetheless has remained largely immunized from the potential abuses of market power inherent in this situation, owing both to its subsidized status and to the long-standing legal tradition that denied copyright protection to works produced by U.S. government agencies.30 As a result, data provided by federally funded projects have flowed through the domestic innovation system with few legal impediments (see Box 5.1), and legal disputes about
ownership or the exercise of proprietary rights in scientific data as such rarely have been ventilated before intellectual property tribunals.31
By the late 1980s, however, digital technologies and new telecommunications networks had combined to produce "the greatest changes in the way information is distributed since the invention of printing by movable type in the 15th century."32 The use of computers made it economically feasible to collect, store, manage, and deliver huge amounts of data at a time when continuously expanding databases have become ever more prominent building blocks of knowledge, especially in the observational sciences, as discussed in Chapter 3. Electronic databases further blur the line between these collection and application functions by allowing users to make their own tailor-made extractions from the mass of data available in the collection as a whole.33 These tools allow users to "add ... immense value to what would otherwise be masses of incoherent, disparate data."34
Moreover, the latest value-added data products, once disseminated worldwide via the Internet and other media, frequently lead to the rapid production of new technical innovations, which result in the generation of more data.35 Electronic publishing thus broadly advances the revolutionary process that computerization began, and it makes both data and research results potentially available at very low cost all over the world.36
As this digital and telecommunications revolution has created vast new markets for electronic information goods and tools,37 it has outpaced the legal infrastructure, which remains geared to the slower-moving print media.38 This strain manifests itself in two contradictory ways. Sometimes digital technology aggravates the basic market-failure characteristic of information goods and thus deepens a chronic state of underprotection. This can occur, for example, when second comers download the originator's data and enter the market with a competing product that free-rides on the originator's investment.39 At other times, however, digital technology so thoroughly overcomes the threat of market failure that it endows the first to invest with abnormal market power that can result in a chronic state of overprotection. This can occur, for example, when sole-source data providers charge exorbitant prices or oblige libraries and research institutions to accept terms and conditions that effectively waive both the special privileges and the fair use exceptions set out in the Copyright Act of 1976.40
The Vulnerability of Publicly Distributed Electronic Databases
To the extent that government- or university-generated data remain uncommercialized, their vulnerability to technically refined means of accessing, downloading, or duplication is only of relative importance. Presumably, the originators want the broadest possible distribution of their data sets.41 Even in this situation, however, there are some concerns that are likely to grow over time. For
example, government may impose cost-recovery conditions on the use of data that third parties who obtain unauthorized access could avoid. Users also might inadvertently corrupt the original database and cause potential harm.
Moreover, over time, the distinction between basic, noncommercialized data and data applied to industrial pursuits or other downstream uses seems likely to break down, as has already occurred in other disciplines, such as the Earth sciences and biotechnology.42 Universities and other research institutions may view data compilations generated in the course of their research as potential revenue sources, especially in an era of declining government support, just as they have done with patentable inventions. As more scientific data are applied to commercial purposes for one reason or another, the data collectors must necessarily distinguish between sources that are made publicly available without charge and those that are not.43 Otherwise, even the providers that do not charge for data could disrupt contractually controlled applications of their own data downstream, not to mention the risk that the noncharging government or academic generator might inadvertently infringe on third parties' proprietary domains.
A related trend is for some governments to commercialize their data, regardless of whether other governments follow suit. The former will become concerned about the vulnerability of their data even if the latter are not. By the same token, those providers that still choose not to charge for their services will increasingly come into contact with (and, perhaps, conflict with) the legal and technical fences that states bent on commercializing data may erect. As one observer put it, "The division between the two regimes" could become "a dam over which information will not easily flow," to the possible detriment of scientific progress and global economic growth, which seems to require that "[m]ore than perhaps any other commodity, data must be allowed to move without barriers."44
To the extent that databases are commercialized, whatever their origin, the refined digital technologies that enhance the compiler's power to collect and disseminate data will enhance as well the free-riding competitor's power to appropriate the fruits of the first comer's investment.45 The second comer who purchases the originator' s product, say, in the form of a CD-ROM, may electronically extract and recompile the data in question at a fraction of their collection and distribution costs. The second product may then be sold for less than the first, because its publisher has contributed nothing directly or indirectly to the research and production costs. Digital technology also enables second comers to extract and recombine the originator's data into value-added products that improve on the original, or that compete in different and sometimes distant market segments.46 In some cases, third parties may even extract the compiler's data in order to make them available over telecommunications networks, an act that can destroy any residual incentives to invest.47 In such cases, existing copyright laws generally afford little or no relief, as explained above.
Relative Invulnerability of Many Privately Controlled Databases
When the database maker is the sole source of the data in question, and substitute databases cannot readily be compiled from public domain sources, digital technology greatly strengthens a supplier's market power. By restricting access to identifiable, on-line subscribers, for example, and by "placing conditions on access and [using technology] to monitor . . . customer usage," the publisher can largely restore the power of the two-party contractual deal that the advent of the printing press had appeared to destroy.48 In effect, publishers in this position may not need copyright law at all, even if they qualified for protection. They may prefer to reject the state-imposed cultural bargain in order to override both its fair-use provisions and its specific exemptions favoring the public interest in teaching and research.49
Moreover, electronic publishers may have virtually no transaction cost problems to overcome because digital technology now enables them to track and charge for every instance of electronic access, even for browsing and scientific uses that were previously exempt. 50 The resulting market power then enables the publisher to impose monopoly prices and arbitrary terms on users-including libraries, educational institutions, and research centers-and to disregard the social consequences that ensue from the inability of such public organizations to foot the bills.51
How Will the Public Interest Be Served in the Information Age?
While many types of scientific data, like other forms of information, possess economic value under the appropriate circumstances, the sponsors of new proprietary rights explicitly contemplate a level of systematic commercialization of both large and small units of data that is unprecedented. How these impending changes in the legal infrastructure will impinge on the research and educational communities has not been clearly worked out even by the European authorities responsible for the European Union's recently adopted Directive on the Legal Protection of Databases. 52 A bill to enact a U.S. model of the European law, which was recently introduced, is even more cryptic in this regard,53 while the WIPO Draft Database Treaty tried to finesse the issue.54 One can predict, nevertheless, that these legislative initiatives will greatly affect the scientific and educational communities if, as Chapters 3 and 4 of this report have emphasized, they lead to a more market-driven environment with fewer government subsidies than before.
Whether contractual attempts to reduce users' access to scientific and cultural products that was promoted by copyright laws in the past will survive legal challenges on such grounds as federal preemption of state law, or doctrines of misuse of copyrights (allied to antitrust law), remains controversial.55 Another question is whether the economic and cultural bargain embodied in copyright law remains appropriate for the digital environment (see Box 5.3), given that trade-driven economic policies have otherwise weakened the consensus on which that
BOX 5.3 Copyright Law in the Information Age—Two Perspectives The application of domestic copyright laws to digital technologies that did not exist at the time of drafting undoubtedly means that the statutory language will not always fit the cyberspace dimension. The resulting ambiguities could give rise to legal uncertainty in any number of real or hypothetical situations. Yet, the Copyright Act of 1976 was deliberately drafted with a view to accommodating future technologies, while the manner in which one seeks to resolve ambiguities that impartial observers find genuinely troublesome, rather than merely pretextual, depends on one’s allegiance to the “economic and cultural bargain” thought to be at the core of prior law.1 For example, bookstores do not charge customers for the privilege of browsing through the stacks prior to purchasing specific books, and this use would not violate the copyright owner’s exclusive reproduction rights, irrespective of any contractual relations between booksellers and customers. Once the work in question was converted to digital form and transmitted over telecommunications networks, however, publishers could monitor and charge for analogous uses if they fell within their exclusive rights to reproduce, adapt, publicly perform, distribute, or display copyrighted works, or if a new “exclusive right of transmission” were enacted, as the Information Infrastructure Task Force’s White Paper proposes. Enactment of the White Paper’s proposed “transmission” right could then explicitly or implicitly remove uses analogous to “browsing” from the “fair use” tradition, whereas reliance on existing law would leave the issue to case-by-case judicial determination. However, the WIPO Copyright Treaty that was opened for signature on December 20, 1996, created in Article 8 a new right to make available “to the public…[these] works in such a way that members of the public may access these works from a place and at a time individually chosen by them.” Moreover, the agreed statements concerning this treaty explicitly carry over the preexisting fair uses recognized in state practice and allow for new instances of fair use for digital transmissions of copyrighted works over telecommunication networks. The Copyright Act of 1976 also allows one who purchases copies of most copyrighted literary works to give or lend them to others or even to resell them second-hand for profit without owing additional royalties to the owners of the copyrights in question.2 Whether analogous acts are, or should be, permitted with regard to digitally transmitted works is open to question,3 as is the right of certain users to |
study or download digitally transmitted works for private research and other noncommercial purposes without payment of royalties to the copyright owners. While publishers of print media had more or less grudgingly come to terms with such private or fair uses, they do not believe the same exceptions apply (or ought to apply) to on-line or other forms of digital transmission.4 While the advent of new technologies has always created a degree of legal uncertainty In intellectual property law, the tendency in the past was to allow the law slowly to catch up, despite the risk of some short-term obsolescence. in contrast, the supporters of the currently proposed changes to the law contend that the opposite course of action is needed with respect to digitally conveyed knowledge and information goods, and the major reforms set out in the White Paper adhere to this view. If one believes that the federal courts can apply existing copyright and unfair competition laws to the new technologies with relatively little friction, then one has implicitly opted for a wait-and-see approach or at least for a minimalist approach, based on case-by-case judicial decisions and a minimum amount of tinkering with the statute as it stands.5 This approach leaves the traditional exemptions for scientific and educational users intact, but subject to case-by-case evaluation.6 If, in contrast, one believes that gaps in the law leave on-line publishers too much at risk, then proposals for statutory reform easily escalate into a campaign to rid the emerging Information infrastructure of allegedly anachronistic vestiges of the cultural bargain that, had heretofore protected users and second comers of works in print and other media.7
|
cultural bargain previously depended.56 Equally uncertain is the role that libraries will play once information providers "can connect directly to the user" via digital transmission over telecommunications networks.57 Some observers see the changing role of libraries as grounds for allowing publishers virtually unfettered discretion to impose contractual conditions on library access to networked transmissions. 58 Others see the dependence of users everywhere on digital transmissions for the future acquisition of scientific knowledge as grounds for generalizing some of today's library and fair use privileges to all on-line users.59 The real question, then is how to recreate a "fair use" zone in cyberspace60 that protects the strong public interest in ensuring that certain uses and certain users, notably the research and educational communities, are not priced out of the market or forced to cut back on the kind of basic research that has heretofore played a crucial role in U.S. economic and technological growth.61
THE DRIVE FOR LEGAL PROTECTION OF NONCOPYRIGHTABLE DATABASES
In response to the perceived gap in the worldwide intellectual property system,62 proposals are being put forward to protect noncopyrightable databases by means of ad hoc or sui generis intellectual property regimes—that is, by special intellectual property laws that deviate significantly from the classical patent and copyright models63 that underlie the Paris and Berne conventions of 1883 and 1886, respectively. The impetus for a sui generis database law has come from the Commission of the European Communities, whose member countries have adopted, to varying degrees, a policy of commercializing government-generated data.64 That policy is contrary to the traditional policy of the United States, which has favored full and open access.
Starting in the late 1980s, the Commission of the European Communities began to reevaluate the legal status of databases, especially electronic databases, in the process of formulating an overall strategy for information technologies known as the Information Market Policy Action (IMPACT) program.65 The Commission found that European database producers had to overcome several comparative disadvantages in order to expand their share of the world market and to catch up with U.S. industry, which dominated the market and was growing at a faster rate than its European counterpart. To overcome these disadvantages, the Commission stressed the need for a single, integrated market, undistorted by differing regulatory approaches, and for higher levels of intellectual property protection, tailored to the needs of potential investors in database production, that might stimulate additional investment in this sector.66 Another likely premise in the Commission's thinking was that privatizing the government's role in the collection and distribution of data might also generate income streams that could help to offset the shrinking availability of public funds for research and development.
The Commission decided both to harmonize the domestic copyright laws
insofar as they applied to compilations of data and to require that the member states adopt sui generis intellectual property laws to protect the contents of noncopyrightable electronic databases, a proposal that was subsequently extended to databases in print media as well.67 In this regard, comparative law revealed that the Nordic countries were already experimenting with short-term, copyright-like protection of noncopyrightable compilations (known as the Nordic catalogue rule), with a view to curbing commercial piracy without extending full copyright protection to borderline literary productions that lacked creative authorship.68 Accordingly, in 1992, the Commission proposed an innovative Directive to protect such databases, "loosely modeled on the Nordic catalogue rule, [and that] more directly and strongly protects electronic information tools."69 A greatly amended version of this proposal was adopted by the Council of Ministers and the European Parliament in July 1995,70 which became the final European Directive on the Legal Protection of Databases of March 11, 1996.71
While the precise mesh of the 1996 Directive's two-tiered provisions in administrative and judicial practice remains to be seen, its highly protectionist attributes are unmistakable. Moreover, if all the pending legislative projects are implemented as explained below, similar changes could be introduced into U.S. law and—via a WIPO Database Treaty—into international law as well.
Development of the European Approach to Noncopyrightable Databases
Collections of data, including those relatively unstructured or unprocessed collections of primary interest to scientists, have never fit comfortably within the romantic notion of authorship that once dominated European copyright law, or even within the more pragmatic conceptions of "originality" that pervade modern copyright laws, such as that of the United States. Behind this conceptual resistance there lies a profound concern, often expressed in judicial decisions and the writings of jurists, that facts and ideas constitute building blocks of intellectual discourse that should not (and perhaps cannot constitutionally) be removed from the public domain. In this context, the scientific community's own commitment to the full and unrestricted flow of data represents an important subchapter in a larger discourse that, in this country, at least, is rooted in the First Amendment. 72
The Commission of the European Communities initially addressed with commendable caution the perceived need for legal incentives to spur investment in electronic database production. The Commission affirmed its preference for a regime based on modified liability principles, that is, one that would deter certain types of socially undesirable conduct without vesting exclusive property rights in data as such (see Box 5.2).73 Unfortunately, even the Commission's earliest proposals along these lines were flawed by contradictory elements drawn from the exclusive rights model, while the final version became a much less balanced and potentially anticompetitive exclusive property right.74
The European Commission's Initial Project
The European Commission's initial approach was premised on the "absence of a harmonized system of unfair competition legislation" to safeguard "the investment of considerable human, technical and financial resources" in the making of databases that "can be copied or accessed at a fraction of the cost needed to design them independently."75 The logical solution, therefore, was to codify a new type of unfair competition law. Such a sui generis law, loosely modeled on existing laws that already protected trade secrets or confidential information, would repress conduct amounting to the ''misappropriation" of an electronic database producer's investment without imposing either legal barriers to entry or the social costs of actual or legal secrecy.76 To this end, the Commission proposed simply to forbid the "unfair extraction" of data from an electronic database for commercial purposes without the second comer's having expended independent effort to collect and verify similar information. The first draft Directive accordingly provided a 10-year period of lead time in which the database maker could recoup his or her investment in a noncopyrightable electronic database while preventing copiers from engaging in for-profit extraction or reutilization of the factual contents, in whole or in substantial part.77
The Commission's "unfair extraction" criterion seemed to invite case-bycase judicial distinctions between procompetitive activities, especially independent investment in the generation of a competing electronic database (which was roughly analogous to reverse-engineering by honest means), and market-distorting forms of electronic copying (which were roughly comparable to industrial espionage, commercial bribery, and other types of "parasitic" or free-riding behavior that unfair competition laws interdict). It also may have opened the door to case-by-case judicial evaluation of unauthorized extractions deemed "fair" because they advanced noncommercial educational and scientific pursuits, although neither the draft Directive nor the first Explanatory Memorandum specifically endorsed this proposition. In any event, the drafters further diluted the database maker's new right against "unfair extraction" by engrafting some express user's rights upon it and by adopting explicit measures to safeguard the public interest in free competition.
For example, the drafters apparently envisioned that lawful users of an electronic database could make a limited reuse of its contents even for some commercial purposes, as might occur in value-adding uses. There was also no clear means for database creators to extend the duration of control over the initial compilation by making subsequent changes to it, although the extent to which this omission resulted from a drafting oversight remains unclear. Moreover, price competition was directly encouraged. Second comers could choose between independently compiling their own databases from scratch or invoking a statutory compulsory license against any sole-source provider of data in electronic databases, with a view to competing against that provider while paying
reasonable royalties for the use of the data thus extracted.78 The liability principles loosely embodied in the first draft of the Directive thus created no legal barriers to entry. Arguably, these principles may even have lessened existing economic barriers to entry by empowering would-be competitors to borrow data at reasonable rates when the cost of independently regenerating them appeared too costly or otherwise inefficient as a business strategy.79
Absent from this framework were any explicit exceptions favoring educational and scientific users (assuming these were not implicitly "fair" uses under the basic "unfair extraction" criterion of the draft Directive), an omission that the European Parliament singled out for criticism. Although the legislative history does not explain why the drafters rejected this criticism,80 the reasonable inference from all the evidence is that the Commission believed further exceptions and immunities would unduly weaken the publishers' incentives to invest under a regime that already implemented a procompetitive strategy. If so, the Commission appears to have erred in at least two respects.
First, it seems to have assumed that a competitive market would intrinsically satisfy the needs of the scientific and educational communities, whereas this report shows that basic science has organizational and operational needs that often differ from those a competitive market is geared to meet.81 Indeed, experience demonstrates that basic science may not be able to pay the market rate even when it is competitively determined. Important research projects consequently may languish for lack of affordable data unless nonmarket mechanisms (such as subsidies) or legal constraints on publishers (such as fair use exceptions and compulsory licenses) close the gap.
Second, the drafters apparently assumed that their concern for the public interest in free competition was still a paramount legislative value in developed market economies. Unfortunately, such protection of the public interest that was implicit in their rudimentary liability framework ultimately gave way to a potent exclusive property right in which public interest safeguards were deemphasized in favor of the protection of private interests.
The European Union's Final Product—The 1996 Directive on the Legal Protection of Databases
Although the European Commission's initial project had undergone transformation by the time that the Amended Proposal was put forward in 1993,82 its wholesale conversion from a relatively weak liability regime to a strong exclusive property right occurred during the closed proceedings of the European Council of Ministers, which produced the Common Position of July 10, 1995.83 This version, with minor technical alterations, became the final Directive on Databases, adopted on March 11, 1996, which the European Union member states must now convert into domestic intellectual property laws and regulations. 84
The Directive as finally adopted may be subdivided into five parts: (1) a list of 60 "recitals" or premises that underlie this legislation; (2) a small group of
definitional articles that apply across the board (Articles 1-2); (3) a set of provisions harmonizing the treatment of databases under the member states' domestic copyright laws (Articles 3-6); (4) a set of provisions requiring these same states to provide the new, sui generis intellectual property right for noncopyrightable databases (Articles 7-11); and (5) a final group of "common provisions" that apply to both copyright and the sui generis laws (Articles 12-16).
Of these, the broad definition of "database" in Article 1(2)85 constitutes an important feature. Whereas earlier proposals had limited sui generis protection to electronic databases that were deemed particularly vulnerable to rapid duplication, the broadened definition now includes databases in print form that are accessible to the human eye. In other words, no database circulating within the European Union will escape the regulatory effects of the 1996 Directive, regardless of the medium in which it appears or the nature of its compilers.
As finally enacted, the sui generis right conferred on qualifying database makers is no longer couched in terms of "unfair" or even "unauthorized" acts or uses.86 Rather, the database maker obtains an absolute exclusive "right to prevent extraction and/or reutilization of the whole or of a substantial part, evaluated qualitatively and/or quantitatively, of the contents of that database."87 This two pronged exclusive right, which now applies to both electronic and nonelectronic databases,88 lasts for an initial period of at least 15 years. Any compiler who makes a database available to the public, however, may continually renew the right for additional 15-year terms with every additional investment in the database.89 This renewal right covers the contents of the entire database, and not just the new matter (as would occur under the derivative work right of copyright law).
The final 1996 Directive does not make sui generis protection contingent on the showing of a creative achievement or of a novel contribution to the prior art, the classical bases for justifying legal derogation from free competition in the past. Rather, it requires the database maker to prove that "there has been qualitatively and/or quantitatively a substantial investment in either the obtaining, verification or presentation of the contents" or in "any substantial change resulting from the accumulation of successive additions, deletions or alterations." 90 Because the Directive itself provides no further guidelines for evaluating the requisite level of investment in either case, this threshold will remain uncertain, pending decisions by European courts applying the still to be drafted domestic database laws. Nevertheless, there are no limits to the number of quantitative or qualitative changes that will thus qualify for such extensions, and any publisher who continues to make a substantial investment in updating, improving, or expanding an existing database can look forward to perpetual protection.
Even though the sui generis right depends on mere investment rather than on some palpable creative contribution, the scope of protection that the 1996 Directive affords investors in noncopyrightable databases now appears roughly equivalent to that which it elsewhere affords full-fledged authors of copyrightable compilations (and greater than that traditionally afforded to authors of literary works; see Box 5.4).91 This conclusion follows both from the definitions of the exclu-
BOX 5.4 The 196 European Directive's Broad Protection for Database Investors The investor's scope of protection under the hybrid right to prevent extraction appears paradoxically to exceed even that afforded authors of traditional literary and artistic works under the classical copyright paradigm of the Berne Conventionl in at least three important respects. First, the basic idea-expression dichotomy underlying U.S. copyright law2 (which the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) applies universally to all copyrightable works, including such borderline works as computer programs and factual compilations3) doesn’t apply to noncopyrightable databases covered by the sui generis regime. For this and other reasons explained below, in the universe of data generators, there is no public domain substratum from which either research workers or second comes are progressively entitled to withdraw previously generated data4 without seeking licenses that may or may not be granted. On the contrary, every independent generation of data, however mundane or commonplace, will obtain protection if it costs money, and every regeneration or reutilization of the same data in updates, additions, and extensions that cost money will extend that protection without limit as to time.5 As a consequence, third parties wilt rarely be able to avoid the expense of regenerating preexisting data—in the way that they can always use previously generated ideas, however much it cost to develop them—unless the originator of the relevant database has abandoned it, or declined to exercise his or her proprietary rights, much as occurs under trademark laws, 6 To be sure, data providers (including, where feasible, members of the scientific community) could decide not to exercise proprietary rights in certain databases, for example, those funded by government agencies; but this would not change the legal situation with respect to scientifically important data located in privately owned databases or in those funded by public agencies, especially foreign agencies, that had opted to commercialize their data. The absence of any equivalent of the idea-expression doctrine under the new sui generis regime means that investors, in effect, obtain proprietary rights in data
|
as such, a type of ownership that the copyright paradigm expressly precludes. The drafters of the sui generis right play this down by insisting that third parties always remain free to generate their own databases. But this opportunity exists only for data that are legally available from public sources and whose cost of independent regeneration is not prohibitively high in relation to the gains expected from the exercise, if any. As for proprietary data not legally available for second comers to exploit, there is no opportunity to avoid the originator's exclusive rights to prevent extraction or reuse of existing data. In such cases, the investor's exclusive rights necessarily vest in the data as such. A deeper point is that, regardless of whether it Is possible n theory to regenerate the data from publicly available sources, investors in database production can always deny 1thir parties the right to use preexisting data in value-added applications,8 even when the third parties are willing to succumb to royalty-bearing licenses;9 and there is no escaping such licenses unless the database publisher either declines to exercise his or her rights, or engages in an abusive exercise of market power. In other words, except when the new proprietary rights are abandoned or misused, the concept of incremental or "cumulative and sequential innovation," which is central to the development of modern technological paradigms,10 has been banished from the universe of database production, despite the economic waste and inefficiency inherent in such policies. A second; and closely related way In which the database investor's scope of protection under the 1998 European Directive exceeds that of authors under the classical copyright paradigm is to be seen in the treatment of derivative works. Under copyright laws, the scope of an authors exclusive right to make a derivative work extends only to the original, expressive matter added to the underlying work. It cannot protect either ideas or preexisting expressive matter, including any matter that has entered the public domain,11 But the 1996 Directive recognizes no such legal distinction and, as just explained, it harbors no working conception of a public domain whatsoever. Unless local European courts applying the domestic laws that implement the Directive take pains to limit this omission, the upshot will be that each new extension of the database makers exclusive rights by dint of his or her "substantial investment” updates, additions, and revisions12 will, in effect, requalify that investor for protection of the database as a whole, for additional 15-year periods, and not just for the revised or added matter—the “derivative work”—as would occur under the copyright laws. This, in turn, reinforces the monopolistic
|
effects inherent in the originators ability to deny third parties the right to build Incrementally and sequentially upon preexisting scientific technical knowledge. A third way in which the scope of protection for investors in database production exceeds. that afforded authors of copyrightable literary and art works results from the much narrower range of public interest exceptions applicable to investors.13 In effect, the sole important exception available to all users of noncopyightable electronic databases under the Directive is the fight to extract or reutilize "insubstantial parts of the database?”14 Even this exception applies only to "awful users" of the database (i.e., presumably subscribers to an online service or purchasers of a CD-ROM), which suggests that in most cases, the except merely validates acts incidental to obtaining the value for which Tone has paid. Moreover, a lawful user of a noncopyrightable database cannot extract or reuse even substantial parts of its contents in 'repeated and systematic' ways that conflict with a normal exploitation of that database or .. unreasonably prejudice the legitimate interests of the maker.15 Arguably, this could preclude most value-added uses of an insubstantial par of the database, regardless of the commercial or noncommercial purpose of the users.
|
sive rights set out in the Directive itself and from the Council of Ministers' closed-door decision to delete from its 1995 Common Position the initial draft Directive's compulsory license requirement facing sole-source providers.92
As defined in article 7(2) of the 1996 Directive, the investor's sui generis extraction right covers even temporary transfers of data to on-line receivers, much like the author's broadened rights to prevent reproduction in copyright law under article 5(a).93 The investor's reutilization right covers on-line use or transmissions of data, including those in value-added or derivative formats, much like the author's broadened "communications" rights under article 5 (b), (d), (e).94 In this and other respects, including the omission of any requirement for a compulsory license against sole-source providers, the drafters of the 1996 Directive have integrated the sui generis regime into the broader regulatory framework for national and international information infrastructures that the European Union and U.S. intellectual property authorities are now jointly promoting.95
Potential Effects of an Exclusive Property Rights Approach
Unlike earlier versions, the final 1996 Directive does give European Union member states the option of allowing authorized extraction of a substantial part from a noncopyrightable database "for the purposes of illustration for teaching or scientific research, as long as the source is indicated and to the extent justified by the noncommercial purpose to be achieved."96 This exception is available only to a "lawful user" and only for the purpose of "extraction" but not for that of reutilization, and it will exist only in those member states that opt to enact it.97
If a member state enacts this provision, the scientific or educational user's exempted extraction must satisfy both the noncommercial purpose test and the general obligation "not [to] perform acts which conflict with normal exploitation or [that] unreasonably prejudice [the maker's] legitimate interests."98 Because the normal use of a scientific database in an academic institution typically is to serve as a research and educational tool, this exemption could merely permit illustration of conclusions reached, but not uses for other scientific or educational purposes, such as browsing or even extractions from and use of the collected data for the purpose of reaching the conclusions that one may then freely "illustrate."99
Of course, local legislators might manufacture loopholes through which to widen this exception,100 and database publishers might refrain from imposing harsh or oppressive terms and conditions that unduly impinge upon scientific and educational uses. The fact remains, however, that nothing in the Directive as finally enacted requires such accommodations. Its sui generis provisions contain no real equivalents of the private use, fair use, and related exceptions that traditional copyright laws afford scientific and educational users of core literary and artistic works. Moreover, database publishers who acquired market power through restricted on-line transmissions reportedly have recently imposed questionable contractual conditions on libraries and academic subscribers.101
It follows that under the 1996 Directive, the most borderline and, in the sense that they are basic building blocks of knowledge, questionable of all objects to receive intellectual property protection—compilations of data and facts, scientific or otherwise—paradoxically obtain the strongest scope of protection available from any intellectual property regime except, perhaps, for the classical patent paradigm itself. 102 Nor are the breadth of protection and the monopolistic power it tends to breed likely to be offset by greater competition in the market for electronic databases, especially now that the 1996 Directive as finally adopted no longer contains the compulsory license requirement that had initially been devised for this purpose.
Formally, of course, third parties still remain free to compile a database exactly like one already in commerce, because independent generation of the relevant data at one's own time and expense is always permitted. In practice, this option ignores the economic realities of the database industry, in which start-up
costs can be relatively high, the prospects for market sharing have seldom been realized, many valuable data sets are unavailable from public sources, and the existence of one complex database seems empirically to constitute a de facto barrier to entry that is seldom overcome. Moreover, as discussed in previous chapters, many databases in the natural sciences contain unique, nonreproducible observations that are by definition available only from a sole source. This lack of effective competition, with its inherent possibilities for discouraging add-on products and for engaging in abuses of market power, was downplayed by the European Council of Ministers in 1995, even though it had been uppermost in the minds of the European Commission's own drafters a short while earlier. Article 16 of the final Directive thus merely calls for 3-year reviews to determine whether existing antitrust laws prove inadequate to deal with the "abuse of a dominant position or other interference with free competition," in which case proposals for "non-voluntary licensing" may once again be considered.103
The fear of market failure and of chronic underprotection that initially motivated the quest for a sui generis regime to protect electronic databases has thus given way to the creation of "mini-monopolies over information"104 and to an underlying logic that is inconsistent with the public interest in the full and open flow of scientific data. The original goal of providing some incentives to augment the publishers' investment in compiling electronic databases has generated a set of norms that could render many scientific and technological undertakings prohibitively expensive. As explained below, the short-term social benefits of this so-called "extraction right" may thus conceal the long-term social costs of diminished research and development capabilities at scientific and educational institutes, including public and semipublic institutions that are already indirectly subsidizing private research and development.105
Overlapping U.S. and European Union International Models
When the European Commission began its deliberations concerning database protection, the climate in which intellectual property policy discussions at both the national and the international level took place differed from that prevailing today. The fate of the Uruguay Round of multilateral trade negotiations and its intellectual property component, the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), remained uncertain. The U.S. intellectual property authorities had not yet begun to survey the issues posed by widespread transmission of digitized information over telecommunications networks. The Supreme Court had just denied copyright protection to telephone directories in Feist and had recently invalidated state protection of subpatentable industrial designs.106 These decisions proclaimed renewed faith in a 19th-century vision of the competitive ethos without recognizing the unresolved problems of gaining returns from investments in subpatentable information goods under 21st-century
conditions.107 It also seems noteworthy that a few years earlier, the chairman of the House Subcommittee on Intellectual Property had set very high standards that would have to be met before Congress would consider proposals for additional forms of sui generis intellectual property protection that deviated from the classical patent and copyright paradigms.108
Against this background, the European Commission's early drafts for a Directive on databases adopted a defensive posture with respect to foreign publishers, which proposed a strict criterion of material reciprocity. Databases made in countries that did not enact sui generis legislation akin to that envisioned by the Directive would consequently remain vulnerable to wholesale copying within the European Union itself.109 This decision to discriminate against foreign nationals operating in nonharmonizing states was modeled on the earlier and equally controversial decision by the United States to impose a material reciprocity clause under the Semiconductor Chip Protection Act of 1984.110 Although both decisions rested on dubious legal grounds even before the TRIPS Agreement was adopted, and even though that agreement rejects this approach, at least in spirit,11l a version of the reciprocity provision nonetheless entered the European Commission's final Directive on the Legal Protection of Databases, as adopted in 1996.112
A Coordinated High-Protectionist Strategy
By 1995, however, when the European Union's Council of Ministers met to adopt its Common Position on the pending database Directive, the climate surrounding worldwide intellectual property policymaking had profoundly changed. Universal intellectual property standards embodied in the TRIPS Agreement had become enforceable within the framework of the World Trade Organization,113 largely as the result of sustained pressures by a coalition of powerful manufacturing associations in Europe, the United States, and Japan.114 The success of this venture presages further alignments of interests by U.S. and European Union officials with a view to forging a common, strongly protectionist strategy for intellectual goods in the post-TRIPS environment.115
Included within this strategy was a packet of complementary proposals for amending or expanding the Berne Convention. Known as the "Digital Agenda," these proposals were considered in a December 1996 Diplomatic Conference hosted by WIPO.l16 Some of the proposals, which the European Union's own intellectual property authorities placed on the agenda for that conference, would have conformed international copyright law to the regulatory framework for a global information infrastructure that the U.S. Information Infrastructure Task Force's (IITF) White Paper on the national information infrastructure recently endorsed.117 The White Paper took the view that on-line providers are, or should be, strictly liable for digital transmissions of copyrighted works, even if this obliges providers to serve as "copyright police" without regard to their ability to
perform such functions.118 It also proposed a battery of measures that would prohibit the decoding of encrypted transmissions (or the tampering with other electronic safeguards) as copyright infringement and that would forbid altering "copyright management information, including the terms and conditions for access to on-line transmissions.'' 119
Despite the innocuous appearance of these and related proposals, they are broadly drafted and may result in indirectly overruling numerous judicial precedents, including some that permit reverse-engineering of the noncopyrightable components of computer programs.120 These proposals might also help to immunize copyright owners from claims of misuse for imposing harsh or oppressive conditions on users in the form of nonnegotiable electronic contracts.121 The single most troubling aspect about the White Paper (and the legislative proposals it has spawned) is, as so many qualified observers have concluded, that it favors "reducing the application and scope of the fair use doctrine in cyberspace."122
Another proposal, the WIPO Draft Database Treaty,123 called for worldwide protection of noncopyrightable databases under sui generis intellectual property regimes. A Committee of Experts prepared the WIPO Draft Database Treaty, which draws from both the European Union's model and from draft U.S. legislation. It was placed on the December 1996 Diplomatic Conference agenda at the behest of the U.S. Patent and Trademark Office (PTO).124 The Diplomatic Conference was thus asked to convert the WIPO Draft Database Treaty into international law, even though the United States lacked any corresponding domestic regime as of the time of its writing.125 In addition, there has been no empirical test of the controversial final 1996 European Directive in actual practice,126 and no preliminary reports or studies evaluating even the economic justification for such measures have been issued by WIPO or by any other reputable international institution.127 The Diplomatic Conference postponed action on this proposal and charged WIPO to set a timetable for further deliberations.
Against this background, the changes to the European Commission's Database Directive made in the Council of Ministers' Common Position of 1995, including deletion of the compulsory license provision and other measures that strengthened the exclusive rights apparatus,128 reflect the coordinated strategies that the European Commission and the U.S. intellectual property authorities are now jointly pursuing. If implemented as proposed, these strategies could gradually extend international norms concerning the legal protection of databases from the Berne Convention (or related instruments) to the TRIPS Agreement, which empowers the Council for TRIPS to "undertake reviews in the light of any relevant new developments which might warrant modification or amendment of this Agreement."129 This, in turn, could eventually obviate the long-term effects of the reciprocity clause in the 1996 European Directive by replacing it with a set of harmonized norms binding on all World Trade Organization member states, like those already adopted for semiconductor chip designs in articles 35-38 of the TRIPS Agreement.130
H.R. 3531—The Proposed U.S. Database Investment and Intellectual Property Antipiracy Act
The bill proposing a U.S. law to protect noncopyrightable databases, which was introduced to the House of Representatives at the behest of the PTO,131 articulated a still more protectionist strategy than that of the European 1996 Directive. Under H.R. 3531, as under the final 1996 Directive, a compiler would qualify for exclusive rights to prevent extractions and reuses of the whole or substantial parts of any database by dint of his or her having made substantial investments in the collection, assembly, verification, organization, or presentation of its contents.132 These exclusive rights would attach automatically upon the expenditure of resources, and if the owner continued to invest in updating or otherwise maintaining the database in question, its 25-year initial term of protection could be renewed continually, without limit,133 in all the contents of that database. This provision thus ignores the constitutional enabling clause, which requires intellectual property rights to be limited in time.134
Furthermore, when one scrutinizes the details of the pending U.S. proposal, one finds that its definition of ''database" is much broader than that of the 1996 Directive. It contemplates, for example, that noncopyrightable components of computer programs could qualify for protection as databases, and it provides no apparent criterion for excluding even facts or data compiled for scientific and historical works.135 Moreover, the database maker's exclusive rights to extract, use, or reuse all or a substantial part of the contents are reinforced by allowing database makers to control any use that "adversely affects the actual or potential market for that database" in addition to uses that otherwise "conflict with the database owner's normal exploitation." 136 This specification, which is not found in the 1996 Directive, has the potential for impeding virtually any judge-made exceptions analogous to "fair use" under copyright laws, because any such exception would almost certainly affect the "potential market" for any given database.137 At the same time, the database owner's potentially perpetual "derivative work" right (flowing from continuing updates), which is subject to no public domain exceptions whatsoever, becomes even easier to obtain than under the 1996 Directive, because the U.S. bill would condition the renewal right merely upon "any change of commercial significance" to the database contents and not solely on additional "substantial investments."138
The pending bill then expands the U.S. database owner's scope of protection well beyond that of the 1996 Directive's sui generis regime by introducing an array of measures that, when read together, could produce formidable anticompetitive effects. For example, the 1996 Directive's principal concession to users—the exception for extraction of insubstantial parts139—is ostensibly broadened in H.R. 3531 to permit uses or reuses of insubstantial parts,140 but it is then significantly narrowed in at least two ways. First, there is a new provision that not only forbids "repeated or systematic use or reuse of insubstantial parts"
(like the comparable provision of the European Directive),141 but also expressly forbids extraction or uses even of insubstantial parts "that cumulatively conflict . . with ... normal exploitation ... or adversely affect ... the actual or potential market."142 This latter clause acquires further strength by means of still other provisions that seem to outlaw extraction or reuse of even insubstantial parts of a protected database in any product or service that directly or indirectly competes with the database from which it was extracted in any market, however distant.143 Also forbidden are extraction, use, or reuse of even insubstantial parts "by or for multiple persons within an organization or entity in lieu of... authorized additional use or reuse ... by license, purchase, or otherwise."144
Given such restrictions, one is hard-pressed to imagine unauthorized uses of an insubstantial component that the drafters of the U.S. bill would deem legitimate. To forestall even this remote possibility, the bill allows publishers contractually to override even the formal right of lawful users to extract or use insubstantial parts, in contrast with the express nullification of similar contractual provisions in the 1996 Directive.145 One knowledgeable source reports that some U.S. database publishers, opposed to this constraint in the 1996 Directive, expressed an intent to exercise permissible contractual overrides in practice.146 A similar intention seems manifest in the clause allowing publishers to impose separate licenses for networked use of a database within organizations, including nonprofit academic and scientific institutions, which can be construed as covering the extraction, use, or reuse even of insubstantial parts.147
Taken together, these and other provisions of the proposed H.R. 3531 reinforce the single most disturbing aspect of the 1996 European Directive, namely, that it precludes formation of an evolving public domain from which third parties can freely draw.148 To this end, the bill expressly confines permissible acts of "independent creation" to data or materials not found in a database subject to the proposed sui generis regime.149 This restriction applies regardless of whether the unauthorized extraction or use is made for purposes of noncommercial scientific endeavor or for commercially important value-added products that build incrementally on existing compilations of data. Every unauthorized use or reuse of existing data thus potentially violates the database owner's unbounded derivative work right. Furthermore, the existence of this potential violation is determined without regard to the substantiality of the second comer's own expenditure of effort or resources, to the similarity or differences of the latter's product or service, or to the public-good aspects of the activities undertaken. 150
The monopoly conferred on database owners under the pending U.S. legislation is then perfected by recognizing no public interest exceptions whatsoever. Even the weak exception for extraction (but not reuse) "for the purposes of illustration for teaching or scientific research" that the 1996 Directive allowed European Union member states to enact 151 is omitted from both H.R. 3531 and the U.S. submission to WIPO. 152
The sole concession to science and education in H.R. 3531 is a provision not
found in the European Directive that expressly denies coverage to "a database made by a [federal] government entity."153 Because most databases of primary importance to science are funded by federal government agencies, this provision appears to recognize that such databases merit different treatment from those normally covered by the proposed sui generis regime. The message is rendered ambiguous, however, because it does not expressly apply to databases funded by government and also because of language in the same provision to the effect that "any database otherwise subject to this Act ... is not excluded here from because its contents have been obtained from a governmental entity."154 This language appears to allow private firms that invest in data obtained from federal government sources to qualify for protection. However, it also can be read as implicitly inviting federal governmental agencies to derogate from the traditional U.S. position, reiterated in a recent Office of Management and Budget (OMB) directive, which limits such agencies to the recovery of dissemination costs from commercial applications of government-funded data by the private sector.155 If government agencies were to move beyond the cost-recovery threshold, the continued ability of scientists to access such data on favorable terms, which current policy seeks to guarantee, would then be called into question.
Disregarding the status of databases made by governmental entities, H.R. 3531 would render virtually any act of "collecting, assembling, or compiling ... data ... from ... a database subject to this Act" a prohibited or infringing act. The perpetrator could never justify such acts as incidental to other acts of independent creation, or as incidental to recognized public interest exceptions, or even as a legitimate means of building on preexisting data sets.156 Nor does H.R. 3531 express any concern that application of its exclusive rights might lead to abuse of a dominant position or to other anticompetitive acts that might require "nonvoluntary licensing" at some point in the future.157
Ancillary provisions of H.R. 3531 also embody some of the current administration's most controversial proposals concerning the regulation of national and global information infrastructures. For example, one provision, following a proposal from the IITF White Paper, would outlaw making or distributing any technical device (or performing any technical service) the primary purpose of which is to circumvent self-help technological security measures that publishers rely on to help protect the contents of their databases.158 Another provision, also inspired by the White Paper, would forbid tampering with database management information attached to copies of database contents or otherwise distributing copies in a form that bears false information about ownership or other aspects of managing the relevant proprietary rights.159
Self-help measures, such as encryption for networked transmissions, often serve valid commercial purposes, and they may be indispensable for the protection of privacy.160 However, such measures also may unduly reinforce the publisher's power to impose harsh contractual terms in two-party deals,161 a prospect that H.R. 3531 completely ignores. There are concerns as well that
publishers will use these provisions to fend off legitimate public interest challenges to the scope of protection obtained under the proposed sui generis regime. If, for example, decrypting a coded transmission is necessary to extract part of a database for noncommercial scientific purposes, and that act of decryption itself constitutes a tort,162 researchers are unlikely to explore the potential availability of judge-made public interest exceptions to the exclusive rights conferred by the new regime.163
In sum, by providing a longer period of protection, more powerful exclusive rights, no public interest exceptions or privileges, harsh criminal penalties,164 and ancillary rules reinforcing self-help policing of on-line transmissions, the proposed U.S. law H.R. 3531 would grant database owners greater monopoly power than that emanating from the European Union's 1996 Directive. In so doing, the drafters of H.R. 3531 take no notice of the important role that affordable, unrestricted flows of data have traditionally played in supporting U.S. research and education, or in other sectors vital to economic development. The proposed regime thus risks triggering a chain of unintended consequences that could ultimately compromise both the foundations of basic science and the technological superiority of the national innovation system.165
The Rush to Legislate
In mid-1996, the PTO (which now speaks for the copyright office in international affairs) presented proposals similar to those set out in the IITF White Paper before WIPO, with a view to amending the Berne Convention and to adopting new instruments related thereto at the December 1996 WIPO Conference.166 However, the Diplomatic Conference rejected many of these proposals and adopted a more socially balanced approach to the protection of copyrighted works digitally transmitted over telecommunications networks. Whether WIPO member states—including the United States—will enact this more balanced approach into their domestic laws remains to be seen. In May 1996, the U.S. negotiators at WIPO also presented the draft of an international treaty to protect noncopyrightable databases,167 and on August 30, 1996, the chairman of the Committee of Experts published a WIPO Draft Database Treaty. The conference postponed immediate action on this treaty, but at a meeting of the WIPO Governing Body in March 1997, a new timetable was established for further work.
This unnecessarily fast pace has so far allowed little time for public hearings or for national science groups and other interested parties to organize, analyze the proposals, and contribute to their shaping. 168 If successful, it would convert the IITF White Paper's "reform proposals" for a sui generis law to protect noncopyrightable compilations of data into international minimum standards of intellectual property protection binding on all signatories to the Berne Convention.169 This "whiplash effect" would then oblige the United States to implement these same standards in its domestic laws, as a matter of international law, even if
Congress had not already adopted similar legislation in the interim. 170 The haste with which both the U.S. and the European Union authorities have moved to implement these measures at the international level 171 thus raises still further questions about the extent to which the public interest has been compromised.
The case for moving so far and so fast rests largely on the supposed difficulties of enforcing territorially grounded intellectual property rights in cyberspace172 and on the fear of "detaching information from the physical plane, where property law of all sorts has always found definition."173 From a legal perspective, these developments raise daunting problems of conflicts of law, a field that has never found it easy to accommodate intangible property.174 Yet, it will not do to exaggerate these difficulties while ignoring the harmonizing effects of the TRIPS Agreement, which requires all countries that belong to the World Trade Organization to adopt both the universal minimum standards of the Berne Convention (whether or not they adhere to that convention) and the additional standards concerning computer programs, compilations, and related subject matter set out in the TRIPS Agreement itself.175 Regardless of whose law applies, in other words, digitally transmitted information goods will eventually become subject to the same international minimum standards of protection in all developed countries and in most developing countries as matters stand.176
To be sure, these standards harbor "gray areas" that are open to different interpretations, notably with respect to the scope of copyright protection afforded borderline works, such as computer programs and those databases that otherwise meet the domestic criteria of
eligibility.177 But the developed countries have only just begun to grapple with these issues, and there is no basis for an empirically grounded consensus even with regard to computer programs or industrial designs,178 let alone databases and other electronic information tools. As previously demonstrated, moreover, a solid body of scholarly opinion holds that "a combination of technological restrictions (such as encryption), contractual arrangements and criminal sanctions (for unauthorized decryption)" constitutes overprotection that raises a far greater risk than the risk posed by underprotection.179 Taking the IITF White Paper's controversial proposals to a premature diplomatic conference without allowing time for adversely affected users, especially the scientific and educational communities, to participate in the discussions thus lacked justification in terms of sound public policy.
CHARTING A WELL-CONSIDERED COURSE IN THE NEW ERA
Putting Science into the Picture
At the time of this writing, neither the scientific nor the educational community has played any part in the relevant deliberations concerning the legal protection of databases, and they have not been consulted on any official basis. If matters proceed without adequate input from researchers and educators, Con
gress could enact the proposed sui generis database regime, despite the risk that "it would allow a limited group of database creators to control the dissemination of information" and that the "resulting restrictions on the transfer of knowledge would be detrimental to society, as information lies at the core of social advancement."180
In most cases the proposed sui generis regime will simply engraft a strong legal monopoly onto the preexisting natural monopolies that are typical of the database industry. As noted in Chapter 4, it is the social costs of these double-barreled monopolies to the public at large that must actually be taken into account, along with their overall impact on a scientific community whose leading role in world technological production is linked in still unexplored ways to the traditional funding of scientific data by government.
Clearly, U.S. policy makers should not incur such risks without evaluating in advance the possible repercussions that sui generis database laws might have on the nation's scientific and technological capabilities and future progress, and without taking measures to alleviate them before embarking on such an uncharted course. By the same token, the scientific community has a vital stake in the formulation of new database laws, in order to ensure that legal incentives to stimulate investment in the production and distribution of data do not end by impeding the full and open flow of those same data to basic science.
The scientific community can ill afford to remain indifferent to these proposals for database protection, if only because its whole modus operandi has been based on the principle of full and open exchange of data. This principle is indirectly undermined by the pending proposals concerning legal regulation of the national information infrastructure and directly threatened by the drive to institute sui generis intellectual property rights in the contents of electronic and other databases. As regards the database laws in particular, the foregoing analysis suggests that science and education have two paramount concerns that need to be pursued in the course of future legislative deliberations:
- Sui generis laws to protect databases should, on the whole, reflect a proper balance between public and private interests, including the public interest in free competition, that is, between public goods and private intellectual property.
- Such laws should contain measures specifically designed to preserve and promote the scientific and educational enterprise, including the need to facilitate and encourage the establishment and maintenance of databases essential to the work of science.
Reconciling the Needs of Science with Those of a Competitive Market
The advent of new proprietary rights where none previously existed will influence the collection and distribution policies of all data providers, including
government-funded providers and other sources that currently follow nonprofit pricing policies. As funding sources shrink and foreign governments operating under the European Union's 1996 Directive shift to profit-oriented policies, more and more data of interest to science will be covered by proprietary rights, and fewer data will be made available to science on a cost-of-dissemination basis. The tensions already reflected in the recent OMB circular181 will become more generalized, even though different disciplines will experience different degrees of hardship.
The experience with Landsat is indicative of transnational problems likely to arise when states adopting different policies invoke their sovereign rights to buttress their respective positions. For example, some of the international ground stations that receive Landsat data reportedly object to the traditional U.S. policy of making data available at the cost of fulfilling a request. They want to continue to charge whatever the market will bear, and they are pressing the United States to change its policy and laws. If the cost-recovery approach is not extended to ground station agreements in other countries, this would leave academics and other nongovernmental users to pay prices that they simply cannot afford.
As discussed in Chapter 4, the adverse effects of Landsat commercialization on the scientific community were easy to document, although the value of lost research opportunities remains hard to quantify in terms of objective social costs. In other cases, however, it will prove harder to show the effects on science, especially if a commercialized database has many private downstream users who are better able to afford the rates, and there is no powerful upstream user community—akin to the global change research users of Landsat data—capable of voicing its distress in terms that cannot be ignored. In such cases, the high cost of data may simply inhibit project formulation when there is no realistic possibility of funding that cost. Yet, because academic scientists are relatively few in number and not typically a presence in day-to-day decision making at the policy level, their lost research opportunities may simply go unreported and unrecorded.182 Although such lost research opportunities are difficult to predict or quantify accurately, the areas likely to be most adversely affected include data-intensive research in the observational sciences that rely on unique, multiple, or continuously updated data sources.
Of course, the law of diminishing returns also applies, and commercial providers may find that no one will access their files if they charge too much. Before this point is reached, however, the more likely result—as suggested by the Landsat example—is that the provider may determine that the price/volume point on the demand curve at which the service is expected to be viable can be afforded by only a few well-financed scientists. This does not provide general access for those unable to pay at that level, including both students and scholars with limited grant funds, not to mention scientists and other potential users in poorer communities.
More generally, such an approach ignores the contribution of basic science to the ability of U.S. firms to predominate in markets for technology and informa-
tion goods. Despite a general consensus on the need for sustained levels of investment in research and development, the proposed database laws could change the status quo—without anyone's wanting it to happen—by elevating the price of the one raw material to which U.S. researchers have always had ready access. If less available scientific information were to translate to fewer applications of economic importance, the end result would be a loss of U.S. technological competitiveness in an integrated world market.183
Preserving the Public-Good Aspects of Science—What Is Needed?
The negative prospects outlined above do not mean that the interests of research and education are best served by the absence of legal protection for the contents of databases. As Chapter 4 suggests, a socially balanced, procompetitive database regime might indirectly help science to contain costs by bringing market forces to bear on some of the pressure points. It would provide a greater stimulus to third-party investors who might compete with sole-source data generators or distributors (when the market segment in question can feasibly support multiple providers), or who might adapt sole-source data sets to applications of particular interest to science. While this stimulus might not change the overall market structure or significantly reduce the formation of natural monopolies, in the short term at least it could help to trigger countervailing tendencies and thus lead to lower prices and fewer restrictions on access, particularly if novel, value-added products become of greater importance to science over time.
Conversely, if a socially imbalanced, overly protective database law converts existing impediments into insuperable legal barriers to entry, the adverse effects on science—absent offsetting legal safeguards—would soon make themselves felt (see Box 5.5). In this context, the scientific and educational communities—like value-adding users and second comers in general184—would arguably fare better either under a simple unfair competition law that prohibits wholesale copying or under a sui generis regime built on more refined liability principles than under any regime based on exclusive property rights.
A liability model creates no legal barriers to entry in its own right, nor need it significantly strengthen the sole-source data provider's market power. A liability regime also can eliminate the "refusal-to-deal" problem, by addressing the serious concerns of those who fear the power of sole-source data providers to restrict access to data on a variety of grounds. When an automatic license is built into a modern liability regime, it tends inherently to solve the problem of abuse without recourse to antitrust law. For these reasons, the European Commission's initial preference for a liability regime, rather than an exclusive property right,185 merits careful consideration by the U.S. scientific community as a possible response to the overall challenge posed by the drive for sui generis database laws.
BOX 5.5 Possible Consequences of an Overly Protectionist Database Regime Under an exclusive property rights model, a database owner's absolute monopoly could disincline him or her to allow scientists access to certain files, especially if the owner feared that the uses in question could lead to value-added products that diminished his or her market power.1 Providers and distributors would also be likely to charge higher prices for all uses, to demand payment for certain uses that were previously free, and to resist pressures for price discrimination favoring scientific users. As matters stand, the electronic publishers' growing capacity to charge for each and every use of on-line data (or at least for every documented access to the database) and to track and monitor every user potentially liable for these charges means that it becomes increasingly capable of serving "as its own collection society, subject to no consent decrees, no membership controls and no external regulation." 2 In this milieu, even blanket licenses can be set unrealistically high for large-scale nonprofit users, such as libraries, universities, and research institutions, and the net impact of the licensing fees will further depend on other contractual conditions that accompany the licenses. Even when a blanket license fee is relatively low, for example, if the institution is obliged to purchase many licenses for different researchers or groups of researchers, the total cost may still become prohibitive. The existing tendencies of some publishers to approach academic and scientific users one by one and to impose harsh or oppressive terms3 could only be strengthened by the enactment of a new and powerful intellectual property right covering the contents of electronic databases as such. |
With or without the more procompetitive conceptual framework of a liability model, a socially balanced database law should preserve and promote the public-good aspects of science and education. This goal requires careful crafting of its technical legal machinery, as well as the inclusion of safeguards that address the specific needs of the scientific and educational communities.
Ensuring Legal Safeguards for Access
Exclusive control over data, like exclusive control over ideas, raises serious concerns, including First Amendment concerns, that are particularly germane to open scientific inquiry.186 While meeting these concerns does not necessarily imply that data should become available without charge or proprietary interests, it does mean the following:
- The law itself should define the parameters of an evolving public domain from which investigators can freely extract and use data for certain purposes.
- The law must also guarantee scientific and educational users access to that domain on reasonable terms and conditions.
- The definition of a protectible database should be narrowed so as to exclude ideas and contents of scientific theories.
- Database owners should never possess the right to preclude access to otherwise publicly available data when sought for purposes of basic scientific research.
The terms of access would then depend in part on the size and scope of any free use and fair use zones built into a proper sui generis law for the benefit of scientific and educational users.187
Publishers are likely to oppose such exceptions because they represent a de facto subsidy to educational and scientific users, which in an on-line environment can no longer be hidden behind the ancillary need to overcome transaction costs.188 Nevertheless, the case for maintaining such exceptions is even stronger with regard to the contents of databases than to other objects of intellectual property protection. As in other cases, publishers require state intervention in the marketplace to enforce the fictitious portable fences on which the protection of intangible literary productions depends. In this case, however, the objects of protection—data—are functionally determined elements or particles of knowledge that fall well below the "grain size" threshold of existing intellectual property laws.189 While database publishers need not contribute any intellectual achievement for which a reward is justifiable in terms of social costs, they have now staked a claim to subject matter that world intellectual property law had left unprotected as a building block of scientific and technological progress.
In seeking an unprecedented level of state intervention, therefore, it seems only logical that publishers should exchange a measure of support for the public-good uses of scientific data for lessened risk aversion and for a measure of artificial lead time in which to recoup their investments and turn a profit. This logic is reinforced by the fact that much, if not most, of the data likely to be commercialized under the proposed sui generis regime will, at some stage, have been a product of public-good undertakings funded largely by governments. Requiring publishers to further the public-good aspects of scientific data hardly
seems unreasonable in this context, especially in view of the potential for rent-seeking that inheres in a market structure dominated by sole-source providers.
At the same time, one cannot push the concept of fair use to the point of requiring the private sector to make up for diminished government support of scientific research in general and of the generation of data in particular. Policymakers must, indeed, take pains to avoid a worst-of-both-worlds outcome, in which government support for the production of scientific data declines, while private investment in the generation, distribution, and application of data languishes for lack of adequate incentives. To the extent that private industry develops electronic information tools specifically to promote scientific investigation or other educational endeavors, the imposition of a subsidy favoring science becomes harder to justify and even counterproductive, given that scientific and educational institutions must pay for the many tools they use. This said, data nonetheless constitute a unique kind of tool, and no amount of investment can justify their greatly diminished availability for scientific inquiry.
Appropriate fair use provisions should thus be seen as part of a new cultural bargain that responds to serious concerns about the ability of data publishers to control access to scientific data as such. Implementing this bargain will require careful distinctions between uses that are ''free" and those that providers must permit, but on fair and reasonable terms and conditions.
For example, scientists must freely be able to use the data underlying existing scientific theories to verify or challenge those theories and to develop new ones. Similarly, researchers should have completely free use of their own notes and working files in the conduct of their investigations, regardless of whether these files are embodied in electronic or print media. By the same token, a scientist who creates a new database while using another lawfully obtained database covered by a sui generis law, along with other data, should owe nothing but reasonable use or access charges to the proprietary right holder if he or she did not reproduce a substantial component of the protected data in the new database. Indeed, a sui generis law should never prevent anyone, including scientists, from reproducing or using an insubstantial part of the contents of a protected database for virtually any lawful purpose.
Ascertaining fair uses that database owners must permit the scientific and educational communities to make on more favorable terms than those applicable to ordinary commercial users constitutes a more delicate task. As discussed throughout this report, whenever a given database is funded by government, a bedrock concept of fair use should require that the scientific and educational communities have access to its contents at no more than the marginal cost of reproduction and dissemination.190
In other words, data generated by public funds should come freighted with a built-in, cost-based discount for scientific research and education as a condition of further commercialization by others. This principle mainly preserves the status quo, at least for U.S. scientists, without shifting the costs of generating and
distributing raw or processed scientific data onto private publishers. As long as enlightened government policy continues to favor substantial funding of the data-generating processes, this principle would promote science by preserving the public-good aspects of the data thus collected, without unduly inhibiting private incentives to invest. By the same token, it would prevent the private sector from displacing (or appropriating) the public-good aspects of government-funded data merely because sui generis legislation had been enacted to stimulate investment in distribution or value-added applications. While such a policy may conflict with the 1996 European Directive, depending on how the European Union member states choose to implement the relevant provisions, its adaptation in the United States could influence other countries, including even some European Union member states, which might decide to exercise their implementing option in precisely this way.191
Conversely, when the private sector or other nongovernmental entities fund the generation or distribution of data, a different fair use calculus should come into play. Here the problem is that the ability of science to pay the going, commercial rates is not commensurate with its resources or with the public interest in a strong, basic scientific establishment. The solution is not to shunt the problems of science onto publishers, who have their own business risks to manage, but to ensure that publishers charge scientific and educational users fair and reasonable prices that take account of the overriding public interest at stake.
Achieving this goal, however, is complicated by the difficulties of weaning sole-source providers from the rent-seekers' mentality if market forces themselves do not compel more favorable treatment of scientific and educational users. The appropriate response is to incorporate legal standards into the database law that can create sufficient leverage for scientific and educational users to obtain such treatment. The gentlest and least market-distorting form of leverage, in turn, is the legal uncertainty with which legislatures can endow the relevant fair use provisions. This strategy gives both sides the maximum incentives to negotiate their own licenses providing for price discrimination, product differentiation,192 and other forms of relief on terms that seek to reconcile the different interests at stake.
Various technical devices, adopted singly or in combination, can be employed to bring about this result. For example, a general clause governing licenses in the database law can expressly provide that all licensing and distribution agreements effected under such a regime must be made "on fair and reasonable terms, with due regard for the needs of the scientific and educational communities, for the public interest in preserving competition, and for the needs of national economic development."193 Such clauses, which have already been used in some database transactions,194 would then be construed in the light of other provisions favoring publishers, so as not unduly to impair the commercial value of the database or the owner's return on his or her investment.195 This approach should at least induce publishers to develop favorable subscription rates
for academic and research institutions rather than insisting on per-use charges that may or may not apply in other circumstances.
The use of compulsory licensing can also increase the bargaining power of privileged users. As previously indicated, a properly crafted liability regime protecting investment in databases (which some investigators recommend) could itself incorporate an automatic license favoring second comers and value-adding users, which would kick in after an initial period of guaranteed lead time.196 A refinement of this mechanism could then allow the scientific and educational communities to trigger a special compulsory license for essential needs in the event that publishers fail to provide reasonable terms and conditions.197 If Congress ultimately adopts an exclusive rights regime for database owners, rather than an unfair competition model or a more refined liability model, such a regime could nonetheless include non-voluntary license provisions to meet the needs of these communities. In theory, such a clause permits either side to seek a judicial decision triggering, or blocking, the compulsory license for privileged uses. In practice, a built-in duty to negotiate before seeking such a license,198 coupled with the uncertainty inherent in the applicable legal standards (and the well-known limits of judicial capability), should almost invariably lead to an accommodation between publishers and the research and educational communities that would remove the bone of contention.
Assessing Long-term Effects
Even if the research and educational communities persuade Congress to take their needs into account when fashioning a sui generis database law, their attention must remain focused on its operational consequences, assuming that the above-mentioned issues have been resolved. The introduction of new legal instruments, and a shift toward the commercialization of data, may profoundly change current institutions, especially those bearing on the funding of scientific research. The effects of these changes certainly will need to be assessed and monitored over time.
Where necessary, steps must be taken to ensure that new institutions suited to the maintenance of scientific progress are set in place before existing institutions are undermined or eliminated. The public funding of basic scientific data collection and dissemination should remain sufficiently robust to support the level of technological applications that has enabled U.S. firms to retain a competitive edge in the global marketplace. In this connection, if conciliatory efforts fail to dissuade foreign government science agencies from overcharging researchers and educators (i.e., charging more than marginal cost) for essential scientific data, countervailing pricing strategies and other tactics may become unavoidable. In any event, government agencies, the research and education communities, and database makers will have to cooperate at the implementation stage, with a view
to reconciling the greater role of the private sector with the public-good aspects of national science policy.
From a long-term perspective, the research and education communities must face up to the fact that new technologies for generating, evaluating, and distributing data will continue to change many institutions on which basic science has traditionally relied. While certain to be disruptive, such changes, suitably guided, could foster and enhance vast new opportunities for the sciences, provided that the scientific community moves to meet the challenges in a timely and sustained fashion.
For example, new modes of organizing and distributing the funds needed to generate data may be devised, while the prospects for internalizing transmission and publication costs through the use of electronic communications networks merit careful study. To the extent that widespread commercialization or privatization of data does result from the adoption of new intellectual property rights and other factors, it could stimulate the scientific community to organize its own institutions or mechanisms for the management and dissemination of scientific data, which could operate outside the commercial arena. Because the research and educational communities are both producers and consumers of data, collective action along these lines could make science itself an increasingly important player in the market for databases generally, as well as a stabilizing force in determining the balance between public and private interests.199 How to organize such large-scale undertakings will require careful thought and study in order to avoid sacrificing or compromising other goals of scientific endeavor, or undermining traditional norms of science emphasizing objective pursuit of knowledge based on full and open exchange of data.
Meanwhile, the adoption of different legal regimes to protect database makers by countries with different agendas and at varying stages of economic development could further complicate the full and open flow of scientific data across international frontiers. Measures to harmonize the domestic database-protection laws, or at least their effects on the transnational flow of scientific data, will therefore require intergovernmental study, as will measures and proposals affecting the regulation of national and global information infrastructures. Pressures to integrate these and other international intellectual property standards ever more deeply into the global trade apparatus will certainly mount as countries move to implement and expand the TRIPS Agreement and related international conventions within the framework of the World Trade Organization and that of WIPO, which continues to administer the Paris and Berne conventions.
The ensuing tensions and conflicts will make it more necessary than ever to develop a framework treaty to safeguard the full and open exchange of scientific data in an increasingly commercialized environment. The difficulty of regulating both public and private interests within such a treaty should not be underestimated, however, while the developing and least-developed countries are likely to play a more conspicuous role in intellectual property policymaking as time goes on.
RECOMMENDATIONS REGARDING LEGAL DEVELOPMENTS AFFECTING ACCESS TO DATA
The new proposals supporting an overly protectionist property rights regime for the contents of databases and for on-line transmissions of data and other scientific information have reached an advanced stage of legislative consideration at both the national and the international levels. The committee believes that these legislative changes do not reflect adequate consideration of the potential negative impacts on scientific research and education and that they have been proposed for implementation at an unnecessarily precipitous pace. The committee therefore recommends that the Office of Science and Technology Policy, leaders from the science agencies and professional societies, and all those concerned with sustaining the health of the scientific enterprise should immediately take the following actions:
- Present to all relevant legislative forums the principle of full and open exchange of scientific data resulting from publicly funded research, and clarify the importance of sustaining such exchange to the nation's future whenever these forums consider laws that would apply to exchange of scientific data.
- Demand that national and international legislative processes now in progress slow to a rational pace, and that the deliberations become more public to allow the scientific and educational communities to present their views and concerns to lawmakers.
- Advocate the incorporation of equivalents of "fair use" as part of any regulatory structure applying to databases as such, or to on-line storage and transmission of data and other scientific information. As a corollary, ensure that the public-good aspects of scientific data are preserved and promoted in laws and regulations governing intellectual property on the Internet and in any future electronic networked environments.
- Work with Congress and the official U.S. representatives to the World Trade Organization and the World Intellectual Property Organization to ensure that the nation's interests in maintaining preeminence in science and technology are not undermined.
- Pursue these issues not only within the United States, but also internationally through international scientific organizations and U.S. foreign-policy channels as they deal with trade and other agreements affecting intellectual property protection.
NOTES
2. |
U.S. Constitution, Article I, Section 8, clause 8. |
10. |
See Reichman and Samuelson, note 1, at pp. 58-72, 137 (citing authorities). |
16. |
See 17 U.S.C. §102(a), (b); Feist Publications, Inc. v. Rural Telephone Service Co., 111 S. Ct. 1282 (1991). |
18. |
17 U.S.C. §107 (1994). |
19. |
See, e.g., 17 U.S.C. §107 (1994); Campbell v. Acuff-Rose Music, Inc., 114 S. Ct. 1164 (1994); Harper & Row Publishers, Inc. v. Nation Enterprises, 471 U.S. 539 (1985). |
|
private benefits test of fair use in terms of transaction costs) with Goldstein, Celestial Jukebox, note 21, at pp. 223-24 ("celestial jukebox may reduce transaction costs" and "perceived need for safety valve such as fair use," which would lead to a copyright law with ''no exemptions from liability"). But see Goldstein, Celestial Jukebox at p. 230 (stressing the enduring importance of "exemptions or compulsory licenses for educational and research uses"). |
|
innovation while restricting access to ideas, methods and processes by indirect means and for a very long duration ... Yet, underprotection can result from the inability of copyright-like models to protect the internal dynamic features of technological innovation, in which idea and expression merge, and also from the lack of any exclusive right to control end use ... Similarly, on the industrial property side ..., "overprotection results from the progressive monopolization of ever smaller aggregates of inventive activity, which elevate social costs in return for no clearly equilibrated social benefits. Yet, the nonobviousness standard [of patent law] and its variants can also induce states of chronic underprotection by excluding the bulk of the incremental innovations that underlie today's most promising technologies. |
33. |
Gregory M. Hunsuker (1997), "The European Database Directive: Regional Stepping Stone to an International Model?," Fordham Intell. Prop., Media & Entertain. L., forthcoming. |
|
Antonelli (1992), "The Economic Theory of Information Networks," in The Economics of Information Networks (C. Antonelli, ed.), Elsevier Science Publishing Co. |
36. |
See, e.g., David and Foray, note 35; see also Cristiano Antonelli, note 35, at pp. 5-28. |
41. |
See the introductory discussion in Chapter 3. |
43. |
See, e.g., John Browning (1996), "Cyber View: Playing Facts and Loose," Scientific American (June): 30-32 (warning about unintended effects of legal restrictions on searching and gathering data). |
44. |
Charles von Simson (1995), "Feist or Famine: American Database Copyright as an Economic Model for the European Union," Brooklyn J. Int. L., 20:729, 768. |
|
ers, who do not merely duplicate or "clone" the first comer's product, are hardly free riders even when they do not contribute directly to the first comer's production costs under a licensed royalty transaction. See, e.g., Samuelson, "Missing Foundations," note 34; see generally Reichman, ''Legal Hybrids," note 27, at pp. 2521-23, 2535-39. |
46. |
See Reichman and Samuelson, note 1, at pp. 66-70, 124-30. |
51. |
See note 3 reporting OTA concerns in this regard; Kurtz, note 32. |
53. |
See H.R. 3531, U.S. Congress, House of Representatives, 104th Cong., 2d Sess., May 23, 1996, Database Investment and Intellectual Property Antipiracy Act of 1996. |
54. |
See WIPO Draft Database Treaty, note 5, article 5 (I) (proposing a truncated version of the Berne Convention's fair use clauses). |
57. |
Ginsburg, "Speculations," note 49, at p. 60. |
|
service help-line"). But see Jane C. Ginsburg (1994), "Surveying the Borders of Copyright," J. Copyright Soc'y, 41:322:325-26 (arguing for legal restraints on such contractual conditions). |
60. |
See, e.g., Hamilton, note 59, at pp. 628-29. |
66. |
See, e.g., Rosier, note 28, at pp. 109-10, 133-39. The Commission stressed the vulnerability of database publishers to market failure, but devoted little or no published attention to the |
|
countervailing risk of technologically induced overprotection. It nonetheless attempted to deal with this latter problem by means of a compulsory license, but was foiled by the Council of Ministers at the last moment. |
70. |
See generally Common Position, note 52 at pp. 14-29. |
71. |
See E.C. Directive on Databases, note 52. |
74. |
See, e.g., Samuelson, "Missing Foundations," note 34; Rosier; note 28. |
75. |
Now Preambular Recitals 6-7; First Explanatory Memorandum, note 73. |
|
compiler's factual contents closed a gap in the Nordic catalogue rules, which case law had not yet resolved. See notes 68 and 69 and accompanying text. Already at this first draft stage, however, the language chosen to implement the Commission's "unfair competition" approach was contradicted by other language describing the database maker's "exclusive right to prevent unauthorized extraction and reutilization" of contents. See, e.g., First Explanatory Memorandum, note 73, at p. 53. |
81. |
See, e.g., the discussion in Chapter 4. |
82. |
See Amended E.C. Directive on Databases, note 67; Reichman, "Legal Hybrids," note 27, at pp. 2494-98 (analyzing and criticizing these proposals). |
83. |
See E.C. Common Position, note 52; Hunsuker, note 33 (approving this version); von Simson, note 44 (criticizing this version); Reichman and Samuelson, note 1 (criticizing this version). |
86. |
See text accompanying notes 76 and 77. |
87. |
E.C. Directive on Databases, note 52, article 7(1). |
90. |
See id., article 7. |
91. |
See E.C. Directive on Databases, note 52, articles 3 and 5. |
92. |
See E.C. Common Position, note 52, articles 8-9, 16(3). |
96. |
E.C. Directive on Databases, note 52, article 9(b). |
97. |
See id., articles 9, 9(b). |
98. |
See id., articles 8(2), 9(b). |
101. |
See note 3 (findings of OTA Report). |
102. |
See generally Reichman, Collapse of the Patent-Copyright Dichotomy, note 8, at pp. 488-489 (discussing economic implications and contradictions of such paradoxes). |
103. |
See E.C. Common Position, note 52, article 16(3); E.C. Directive on Databases, note 52, article 16(3). |
104. |
Rosler, note 28, at pp. 138, 140 (stressing tendencies of "[m]onopolists typically [to] charge large premiums for their goods"). |
105. |
See, e.g., Rosler, note 28, at pp. 141-43; Reichman, "Legal Hybrids," note 27, at pp. 2496-98. |
106. |
See Feist Publications, Inc. v. Rural Telephone Service Co., 111 S. Ct. 1282 (1991); Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989). |
107. |
See, e.g., Reichman, "Electronic Information Tools," note 15, at pp. 466-67, 472-75. |
110. |
See 17 U.S.C. §§902(a)(l)(A)-(C), 913, 914 (1994); Jay A. Erstling (1989), '"The Semiconductor Chip Protection Act and Its Impact on the International Protection of Chip Designs," Rutgers |
|
Computer Tech. L. J., 15:303; Charles R. McManis (1988), "International Protection for Semiconductor Chip Designs and the Standard of Judicial Review of Presidential Proclamations Issued Pursuant to the Semiconductor Chip Protection Act of 1984," Geo. Wash. J. Int'l L. Econ., 22:331. |
112. |
See E.C. Directive on Databases, note 52, article 11 and Preamble, Recital 56. |
118. |
See, e.g., U.S. White Paper, note 38, at pp. 114-24; P. Samuelson, "Copyright Grab," Wired 4.01:136, 190-91; McManis, note 64, at pp. 68-70 (criticizing this view). |
|
licenses, see note 55; see also Charles R. McManis (1993), "Intellectual Property Protection and Reverse Engineering of Computer Programs in the United States and European Community," High Tech. L. J., 8:25, 88-96 (concluding that contracts, or at least shrink-wrap licenses, that prohibit reverse engineering are preempted by federal intellectual property law); Julie E. Cohen (1995), "Reverse Engineering and the Rise of Electronic Vigilantism: Intellectual Property Implications of 'Lock-out' Programs," So. Cal. L. Rev., 68:1091. But see Raymond T. Nimmer (Reporter for the Drafting Committee on Uniform Commercial Code, Article 2B (licenses)), U.C.C. Revision: Information A.S.E. in Contracts (April 15, 1996) (arguing that proposed Art. 2B of U.C.C. should make such licenses presumptively valid); Pro-CD, Inc. v. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996) (upholding shrink-wrap license concerning electronic database). |
122. |
McManis, "Emerging Computer Technology," note 64, at p. 67 (citing authorities); Hamilton, note 59, at pp. 628-29. |
123. |
See note 5. |
125. |
Powell, id. |
|
uncertain, and this outcome would depend on the interpretation of various provisions in both the TRIPS Agreement and prior international agreements (see notes 63 and 111 and accompanying text), the goal is clearly to develop "a model in the search for a global solution regarding the protection of databases which is presently discussed at WIPO." Gaster, note 128. |
132. |
See id., sections 2, 3(a). |
133. |
See id., sections 2, 3(a), (b), 6. |
134. |
U.S. Constitution, Article I, Section 8, clause 8. |
136. |
H.R. 3531, note 131, section 4(a)(1). |
137. |
This provision is thus consonant with several other key provisions that greatly strengthen the scope of protection in general. See infra text accompanying notes 138-155. |
138. |
See H.R. 3531, note 131, article 6(b); notes 89-90 and accompanying text. |
140. |
See H.R. 3531, note 131, section 4(a)(2). |
141. |
Id.; E.C. Directive on Databases, note 52, article 7(g). |
142. |
See H.R. 3531, note 131, sections 4(a)(l), (2), 5(a). |
144. |
Id., section 4(b) (4). |
145. |
See E.C. Directive on Databases, note 52, article 15 (expressly voiding contractual provisions to this effect). |
147. |
See H.R. 3531, note 131, sections 4(a) (2), 4(b) (4). |
148. |
See text accompanying notes 91-92 and Box 5.4; Reichman and Samuelson, supra note 1, pp. 103-109. |
150. |
See H.R. 3531, note 131, sections 4, 5, 6. |
154. |
See H.R. 3531, note 131. |
156. |
See H.R. 3531, note 131, sections 4, 5(a)(b); text accompanying notes 140-147 (stressing built-in restrictions on claiming use of an insubstantial part in practice). |
158. |
See, e.g., H.R. 3531, note 131, section 10 ("Circumvention of Database Protection Systems"); U.S. Proposal on Databases, note 124, article 8 (Prohibition of Protection-Defeating Devices). |
159. |
See, e.g., H.R. 3531, note 131, section 11 ("Integrity of Database Management Information"). |
160. |
See, e.g., Branscomb, note 3, at pp. 175-77. |
161. |
See text accompanying note 15. |
162. |
See H.R. 3531, note 131, sections 10, 11. |
163. |
See further infra text accompanying notes 137-138. |
164. |
See H.R. 3531, note 131, sections 8, 13. |
165. |
See David and Foray, note 35. |
166. |
See Perlmutter, note 116; Morton David Goldberg (1996), "The Digital Agenda in the U.S. and WIPO," paper presented to the Fourth Fordham Conference. |
167. |
See note 124. |
|
To this and similar complaints, Visser's report makes the following reply: [M]any limitations are the result of successful "lobbying." Intermediaries and users applying for specific limitations must realize that right owners, who will oppose any limitation as a matter of principle, are generally very well represented at the (national and international) legislative level. Thus, the extent to which copyright limitations will be preserved or extended in the digital environment will eventually be determined by the ability of intermediaries and users to have their voice heard on the political level. Visser, at p. 13 (quoting P.B. Hugenholtz and D.J.G. Visser (1995), Copyright Problems of Electronic Document Delivery: A Comparative Analysis, Report to the Commission of the European Communities (DG XIII), Brussels/Luxembourg, p. 62.) See also Robert J. Hart, "Intellectual Property and the Global Information Infrastructure-The Perspective in Japan, Australia, and Canada," paper presented to Fordham's Fourth Conference (stressing that only Australia's proposals concerning regulation of national information infrastructures have so far reflected concerns for "fair use" and related exceptions). |
172. |
See, e.g., McManis, note 64, at p. 211. |
175. |
See, e.g., Reichman, "Universal Minimum Standards," note 130, at pp. 347-51, 365-73. |
180. |
Rosier, note 28, at pp. 141-42. |
183. |
See David and Foray, note 35. |
184. |
See Reichman and Samuelson, note I, at pp. 113-30, 137-63. |
185. |
See text accompanying notes 73-74. |
186. |
See, e.g., Ginsburg, "Information After Feist," note 23, at pp. 384-387. |
188. |
Goldstein, Celestial Jukebox, note 21, at pp. 220-22. Gordon, "Fair Use as Market Failure," note 22. |
189. |
Samuelson et al., "Manifesto," note 38, at pp. 2385-86 (discussing limits of legal protection for single features of computer programs). |
190. |
See, e.g., OMB Circular A-130, U.S. Government Printing Office. For the problem of leakage as a limit on price discrimination, see Chapter 4. |
191. |
See E.C. Directive on Databases, note 52, article 6(2). |
196. |
See Reichman and Samuelson, note 1, at pp. 145-51. |
199. |
See, e.g., Hunsuker, note 33. |