4
Assessment and Recommendations

Our nation has a vibrant and demonstrably productive community of scientific and technical (S&T) database creators, disseminators, and users that has led the world. Advances in computing and communication technologies make S&T databases and the facts they contain increasingly valuable for producing new discoveries and for accelerating the growth of knowledge and the pace of innovation. The same technologies that facilitate the effective production, dissemination, and use of data, however, can also expedite their unauthorized dissemination and use, with the potential effect of undermining incentives to create new databases, facilitating unfair competition and wholesale piracy, and in the most extreme cases, exposing the original database rights holder to market failure.

As Chapter 3 points out, the current efforts to enact statutory federal database protection in the United States appear to be stimulated by three principal factors: (1) the possibility for rapid and complete database copying with the potential for instantaneous broad dissemination; (2) the gap in U.S. law created by the Feist1 decision, which invalidated copyright protection on the basis of investment and effort (i.e., "sweat-of-the-brow") investments alone; and (3) the E.U. Database Directive, which requires other nations to pass a similar law in order for their citizens to enjoy the E.U. Directive's protections in Europe, thereby providing a potentially unfair advantage to European competitors of the U.S. private-sector.2

1  

Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991).

2  

Although certain database vendors might be at some competitive disadvantage in the European Union, the committee believes that a less protectionist law in the United States that encourages the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases 4 Assessment and Recommendations Our nation has a vibrant and demonstrably productive community of scientific and technical (S&T) database creators, disseminators, and users that has led the world. Advances in computing and communication technologies make S&T databases and the facts they contain increasingly valuable for producing new discoveries and for accelerating the growth of knowledge and the pace of innovation. The same technologies that facilitate the effective production, dissemination, and use of data, however, can also expedite their unauthorized dissemination and use, with the potential effect of undermining incentives to create new databases, facilitating unfair competition and wholesale piracy, and in the most extreme cases, exposing the original database rights holder to market failure. As Chapter 3 points out, the current efforts to enact statutory federal database protection in the United States appear to be stimulated by three principal factors: (1) the possibility for rapid and complete database copying with the potential for instantaneous broad dissemination; (2) the gap in U.S. law created by the Feist1 decision, which invalidated copyright protection on the basis of investment and effort (i.e., "sweat-of-the-brow") investments alone; and (3) the E.U. Database Directive, which requires other nations to pass a similar law in order for their citizens to enjoy the E.U. Directive's protections in Europe, thereby providing a potentially unfair advantage to European competitors of the U.S. private-sector.2 1   Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). 2   Although certain database vendors might be at some competitive disadvantage in the European Union, the committee believes that a less protectionist law in the United States that encourages the

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases The committee believes, however, that the need for additional statutory protection has not been sufficiently substantiated. The high level of activity in the production and use of digital S&T (and other) databases in the United States serves as prima facie evidence that threats of misappropriation do not constitute a crisis. Nor do the existing legal, technical, and market-based measures provide a chronic state of underprotection for proprietary databases. The almost universal use of licensing, rather than sale, of online databases and other digital information, coupled with technological enforcement measures, on balance potentially provides much stronger protections to the licensors vis-à-vis their customers than they enjoyed prior to Feist and under the print media copyright regime (see Table 3.2 in Chapter 3). While some of the current law providing protection to database rights holders remains uncertain in terms of scope of applicability, the trend in recent years has been to broaden, rather than narrow, applicable intellectual property protections. Moreover, strong statutory protection of databases would have significant negative impacts on access to and use of S&T databases for not-for-profit research and other public-interest uses. Nevertheless, although the committee opposes the creation of any strong new rights in compilations of factual information, it recognizes that limited new federal legal protection against wholesale misappropriation of databases may be appropriate. In particular, a balanced alternative to the highly protectionistic E.U. Database Directive could be achieved in a properly scoped and focused new U.S. law, one that might serve as a model for an international treaty in this area. In this chapter, the committee examines several legislative options and related government activities, and recommends a number of legislative principles and policy actions to help inform the current debate. The chapter concludes with a recommendation directed specifically to the not-for-profit S&T community. ASSESSMENT OF LEGISLATIVE OPTIONS, WITH RECOMMENDATIONS ON GUIDING PRINCIPLES The committee assessed and compared three separate proposals for increased database protection of private-sector databases in the United States that were placed in the Congressional Record at the beginning of the 106th Congress.3 During the 105th Congress, the House of Representatives twice approved a measure establishing a specific statutory scheme for the protection of databases—H.R. 2652 4 and Title V of H.R. 2281,5 which was substantially the same as H.R.     use of factual data for both public interest and commercial purposes will benefit the U.S. economy and society to a greater extent. 3   Cong. Rec., Vol. 106, S. 316 (Jan. 19, 1999). 4   H.R. 2652, the "Collections of Information Antipiracy Act," 105th Congress (1997). 5   H.R. 2281, Title V, the "Collections of Information Antipiracy Act," 105th Congress (1998).

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases 2652.6 Both H.R. 2652 and 2281 were adopted on suspension calendar by the House despite significant opposition from an array of scientific, educational, library, and consumer public-interest organizations and institutions, as well as from a number of commercial publishers and information technology services companies.7 These House bills applied both to databases that qualified for copyright protection and to noncopyrightable "sweat-of-the-brow" databases that did not. The proposed legislation drew the concerned attention of the not-for-profit communities because expanding private property rights in factual databases could interfere with scientific progress and other public-interest uses of data.8 At the same time, some private-sector firms believe that their databases are vulnerable to misappropriation due to a gap in the law.9 Other reactions from the private-sector included the perception that the proposal adopted by the House placed too many impediments to transformative uses of existing databases for commercial 6   The only significant change in Title V of H.R. 2281 was to remove "potential markets" from the ambit of liability for not-for-profit uses in Section 1403, Permitted Acts (a) Educational, Scientific, Research, and Additional Reasonable Uses, which was amended as follows: (1) Certain Not-for-profit Educational, Scientific, or Research Uses.—... no person shall be restricted from extracting and using information for not-for-profit educational, scientific, or research purposes in a manner that does not harm directly the actual [or potential] market for the product or service referred to in section 1402." [words in brackets deleted]. 7   See the testimony of the not-for-profit sector cited in note 8 below, and of the commercial opponents to the legislation in note 10 below. In addition, over 130 organizations and companies signed a position statement critical of H.R. 354 that was placed in the public record by Dr. James Neal, director of the Milton S. Eisenhower Library at Johns Hopkins University and president of the Association of Research Libraries, during the March 18, 1999, Hearing on H.R 354, the "Collections of Information Antipiracy Act," held by the Subcommittee on Courts and Intellectual Property of the Committee on the Judiciary of the U.S. House of Representatives [hereinafter March 18, 1999, Hearing]. A copy of the position statement and the full list of signatories may be found online at <www.databasedata.org>. 8   See testimony by Wm. A. Wulf, president of the National Academy of Engineering on behalf of the National Academies, J.H. Reichman, professor at the Vanderbilt University School of Law, and James G. Neal, director of the Milton S. Eisenhower Library at Johns Hopkins University and president of the Association of Research Libraries, at the October 23, 1997, Hearing on H.R. 2652, the "Collections of Information Antipiracy Act," held by the Subcommittee on Courts and Intellectual Property of the Committee on the Judiciary of the U.S. House of Representatives [hereinafter October 23, 1997, Hearing]. 9   Id. See testimony by Paul Warren of Warren Publishing, Inc., on behalf of the Coalition Against Database Piracy. See also testimony by Robert E. Aber, senior vice president and general counsel, the NASDAQ Stock Market, Inc., on behalf of the Information Industry Association, at the February 12, 1998, Hearing on H.R. 2652, the "Collections of Information Antipiracy Act," held by the Subcommittee on Courts and Intellectual Property of the Committee on the Judiciary of the U.S. House of Representatives [hereinafter February 12, 1998, Hearing].

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases purposes.10 Perhaps most significant, the Administration presented its own consensus critique of the House bills on August 4, 1998,11 and the Department of Justice12 and the Federal Trade Commission13 issued legal memoranda outlining their concerns about the legislation's constitutionality and anticompetitive effects, respectively. Soon after the approval of Title V of H.R. 2281 by the House in July 1998, Senator Orrin Hatch (R-UT), chairman of the Senate Committee on the Judiciary, initiated negotiations among the various interests, which continued from early August until early October.14 While substantial progress was made in this process and the needs of the science, education, and library communities were directly acknowledged, a consensus was not achieved before the 105th Congress adjourned in October 1998.15 Shortly after the 106th Congress convened in January 1999, Congressman Howard Coble (R-NC), chairman of the Subcommittee on Courts and Intellectual Property of the House Committee on the Judiciary, reintroduced as H.R. 35416 the proposal that had twice passed the House in the previous session. H.R. 354 included two changes to respond to concerns of the scientific community and other critics of the original legislative proposal.17 Thereafter, Senator Hatch 10   Id. See testimony by Jonathan Band, partner, Morrison & Foerster LLP, on behalf of the OnLine Banking Association, and by Tim Casey of MCI, Inc. on behalf of the Information Technology Association of America at the February 12, 1998, Hearing. 11   See letter from Andrew J. Pincus, general counsel of the Department of Commerce, to The Honorable Orrin G. Hatch, chairman, Senate Committee on the Judiciary, August 4, 1998, summarizing "a number of concerns" of the Administration with H.R. 2652. 12   See memorandum for William P. Marshall, associate White House counsel, from William Michael Treanor, deputy assistant attorney general, Office of Legal Counsel, Department of Justice, July 28, 1998, regarding "Constitutional Concerns Raised by the Collections of Information Antipiracy act, H.R. 2652." 13   See letter from Robert Pitofsky, chairman, Federal Trade Commission, to The Honorable Tom Bliley, chairman, Committee on Commerce, U.S. House of Representatives, September 28, 1998, regarding potential anticompetitive effects of Title V of H.R. 2281. 14   These negotiations were conducted in closed sessions with representatives of the principal organizations that had previously taken a public position on the House bills. The Intellectual Property Counsel to Senator Hatch, Edward Damich, moderated the negotiation process. 15   For a detailed discussion of the Senate negotiations and the legislative process associated with the database protection legislation in the U.S. Congress through early April 1999, see generally J.H. Reichman and Paul F. Uhlir (1999), "Database Protection at the Crossroads: Recent Developments and Their Impact on Science and Technology," Berkeley Technology Law Journal, Vol. 14, pp. 793-834. 16   H.R. 354, the "Collections of Information Antipiracy Act," 106th Congress (1999). 17   The two changes made in H.R. 354 by the House Subcommittee on Courts and Intellectual Property included an attempt to eliminate the potential for indefinitely prolonging the 15-year duration of protection in section 1408 (c), and expanding the scope of the exemption for certain not-for-profit educational, scientific, and research uses in section 1403 (a), both of which are discussed in more detail later in this chapter.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases inserted in the Congressional Record two other proposals—the first drafted by a coalition of commercial and not-for-profit interests (hereinafter, the Coalition Proposal)18 seeking much more limited protection than H.R. 354, and the second a draft bill that had emerged at the end of the 1998 negotiations sponsored by Senator Hatch (hereinafter, the Senate Discussion Draft).19 In the rest of this section the committee discusses several of the most important provisions of these three proposals and evaluates them in terms of their potential effects on access to and use of S&T data by public-interest users. In doing so, the committee recognizes that these proposals have changed and will continue to change before anyone of them is considered for final adoption. Nonetheless, they serve as models for the types of issues that arise from the perspective of the research and education communities confronted with the prospect of legislative changes that would affect access to data.20 As noted in Chapter 1, because of the complex web of interdependent relationships among public-sector and private-sector database producers, disseminators, and users (see Table 1.1 in Chapter 1 for an indication), any action to increase the rights of persons in one category is likely to compromise the rights of the persons in the other categories, with potentially far-reaching negative consequences. A principal concern of the committee, therefore, is that the development of any new database protection measures aimed at protecting private-sector investments take into account the need to promote access to and subsequent use of S&T data and databases not only by the not-for-profit sector, but by commercial producers of derivative databases as well. Of course, it is in the common interest of both database rights holders and users—and of society generally—to achieve a workable balance among the respective interests so that all legitimate rights remain reasonably protected. Therefore, as a general guiding principle, the committee recommends that any new federal protection of databases should balance the costs and benefits of the proposed changes for both database rights holders and users. 18   See the "Database Fair Competition and Research Promotion Act of 1999," Cong. Rec., Vol. 106, S. 316 (Jan. 19, 1999). 19   Id., "Chapter 14—Protection of Databases," S. 322-326. 20   At the time of this writing, the House Committee on Commerce has introduced and marked up a slightly modified version of the Coalition Proposal. See H.R. 1858, The Consumer and Investor Access to Information Act of 1999, 106th Congress, May 20, 1999. The House Committee on the Judiciary also has marked up H.R. 354, which includes a number of significant revisions. Because the study committee had already written its report, it was not able to consider these additional changes to the proposed legislation. Nevertheless, the committee believes that its analysis and recommendations remain relevant to the ongoing debate concerning this legislation, as well as to any eventual implementation of a statutory database protection regime. Any bill that is finally adopted, if any, most likely will be substantially further modified. For this reason, the committee presents its legislative recommendations as guiding principles, rather than as specific legislative language.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases The Standard of Harm The key provision of the three legislative proposals defines the nature of protection accorded a database rights holder and establishes the standard of harm against which a defendant's liability is to be judged. As introduced in January 1999, H.R. 354 prohibited the "extraction or use" of a substantial part of a database if it results in "harm to the actual or potential market" for any product or service incorporating the database.21 A "potential market'' includes any market the database rights holder "has current and demonstrable plans to exploit" or a market that is "commonly exploited by persons offering similar products or services." The Senate Discussion Draft narrowed the protection of actual markets to those markets commonly exploited by persons offering similar products.22 The Coalition Proposal took a different approach, prohibiting only the "duplication of another's database [and inclusion of those records in]... a database that competes with the original."23 To compete with the original database, the duplicate must be substantially identical to the original, must be shown to displace substantial sales or licenses of the original, and must be offered for sale or digitally distributed in such a manner as to "significantly diminish the incentive to invest" in developing the original database. The latter requirement may be interpreted as threatening the opportunity to recover a reasonable return on the investment in collecting or organizing the original database. None of the three legislative proposals purported to create broad property rights in the original database, as the E.U. Directive does. However, by expanding protection to "potential markets," H.R. 354 would allow the rights holder to foreclose markets or uses beyond the rights holder's actual use. This has the effect of granting exclusive rights to the original database rights holder in uses unknown at the time of the database's creation. The limitation of the H.R. 354 language stating "current demonstrable plans to exploit" is unclear because the time at which "current plans" is to be measured is not stated. Does "current" mean at the time the extraction and use occur, at the time the user develops the new market, or at the time the database rights holder brings suit? A scientific researcher might discover an entirely new application for a database, only to be foreclosed from such use if the original database rights holder were subsequently to develop "current and demonstrable plans" to exploit that application as an additional market. For example, scientist A has a database consisting of human gene sequences potentially useful for locating genes controlling certain diseases, but does not know of any particular sequences that are valuable for this purpose. By extraction and use of A's database, scientist B discovers a set of sequences 21   H.R. 354, section 1402. 22   Section 1301(3). 23   Section 1401. The House Committee on Commerce bill, H.R. 1858, has extended that prohibition to include a "discrete section" of a database.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases that seem particularly valuable for further experimentation and makes this subset of sequences available to the scientific community. In doing so, scientist B could violate the protection provided by H.R. 354. Although protection of original, noncopyrightable databases with a strong, copyright-like property right may encourage additional investment in producing databases, it simultaneously discourages others from investing in discovery of new uses for existing databases and elevates the cost of using them. In principle, the public benefits most from the weakest legal incentives for encouraging such investments, and intellectual property theory has always promoted the open availability of facts. For the creation of legal incentives greater than this, the former chairman of the House Committee on the Judiciary, Robert Kastenmeier, required proponents of new intellectual property rights to meet a very heavy burden.24 The Senate Discussion Draft provided considerably narrower protection by requiring a showing of "substantial" harm to the actual or neighboring market,25 which was defined in the Proposed Conference Report Language as "harm [that] is such as to significantly diminish the incentive to invest in gathering, organizing or maintaining the database."26 The harm test in the Coalition Proposal was similarly circumscribed, requiring both a displacement of substantial sales and a showing that the unauthorized use "significantly diminished the incentive to invest in the collecting or organizing of the protected database. 27 These latter two formulations expressly acknowledged that not all duplications are actionable, even if used for commercial purposes (e.g., in distant markets) or for pro-competitive purposes by honest means. The intent was to recognize that competitors who add value and generate socioeconomic benefits should not incur liability if they do not directly harm the market of the original database rights holder, i.e., if they do not compete unfairly. The committee believes that strong protection based on a broadly framed standard of harm test, such as the one proposed by H.R. 354, poses a number of potential problems for research, education, and other public-interest users, as well as for legitimate private-sector, value-adding database producers. As a general rule, the stronger the statutory protection, the greater the encumbrances will be on the reuse and transformation of data received by second-generation database producers and users. One person's derivative use can be characterized as an infringement on the original database rights holder's product; where the bar is set will determine to what extent database producers and disseminators will be 24   See Robert W. Kastenmeier and Michael J. Remington (1985), "The Semiconductor Chip Protection Act of 1984: A Swamp or Firm Ground?," Minn. L. Rev., Vol. 70, p. 417, establishing a stringent four-part test for assessing the merits of any proposed intellectual property protection for new technologies. 25   Section 1302. 26   Proposed Conference Report Language, Section 1302, at 33. 27   Section 1405 (4).

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases enriched at the expense of all socially and economically valuable downstream uses. As noted in Chapter 1 (and indicated in Table 1.1), many organizations are users as well as producers and vendors of S&T databases, as, for example, when they draw on one or more databases to search for cross-disciplinary associations or to create a derivative or value-added database targeted to a competing or entirely new market. Private-sector creators of derivative databases have conflicting views of protection: protection of source databases might deprive them of access, but insufficient protection for their own creations might make them vulnerable to copying. Protection entails contradictory consequences for creators of derivative databases. A concern of the committee, therefore, is that any new protection judged to be necessary must take into account the need to promote access to and subsequent use of S&T data and databases not only by the not-for-profit sector, but by commercial producers of derivative databases as well.28 A major negative effect of a strong standard of harm test would be to raise the resale prices for value-added or derivative databases, as well as to inhibit their production. Value-adding database producers that use multiple data sources to create new products, as is common in both the private and the public sector, are particularly penalized by a strong standard of harm test.29 Although the consequences would be difficult to measure, strong new rights for database rights holders would probably result in a broad loss of research opportunities.30 If, for example, potential users opted to engage in other professional activities rather than deal with more expensive and onerous restrictions on database use, the probability of subsequent discoveries, innovations, and advances in knowledge would decrease, not only because of the reduced number of users, but also because the remaining database users would be constrained in their activities. Downstream commercial providers who must pay license fees to the rights holders of sole-source databases can recover such fees only if they themselves charge more for access, costs that are passed down the chain of derivative products to all users, including investigators in not-for-profit institutions. By making entry into a market more expensive, greater statutory protection also could increase the likelihood that small or niche markets, which are com 28   Indeed, many commercial entities have expressed concerns about such effects of strong database protection. See, for example, the testimony and position statement cited in note 10. 29   As noted by Nobel laureate Joshua Lederberg in his testimony on behalf of the National Academies and the American Association for the Advancement of Science at the March 18, 1999, Hearing, note 7, the "recent advent of digital technologies for collecting, processing, storing, and transmitting data has led to an exponential increase in the number of databases created and used. A hallmark trait of modem research is to obtain and use dozens, or even hundreds of databases, extracting and merging portions of each to create new databases and new sources of knowledge and innovation." 30   See Reichman and Uhlir (1999), note 15, p. 820.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases mon place for many S&T databases, would be served by sole-source providers. A higher cost of entry typically deters entrants and allows the first entrant to act as a monopolist.31 A sole-source may then use its market power to inhibit the development of derivative databases if these are interpreted as undermining the investments in the original database, even if such derivative uses are in completely different markets or are protected as "permitted acts" under a statute. Monopoly power could be exercised over the data in many areas of research, because most observational databases cannot be independently recreated after-the-fact, and it is economically inefficient and undesirable to require independent, redundant collection of original data in activities that use very high cost systems. As the Federal Trade Commission cautioned in its analysis of the predecessor bill to H.R. 354, ". . . policies that further entrench the market power of single-source data providers could have an unintended, undesirable impact on competition and innovation because of the significant potential for anticompetitive conduct in single-source database markets."32 The law should encourage competition because competition leads to lower prices, resulting in broader use and, hence, further discovery and innovation. Increased license fees or unreasonable restrictions on subsequent uses or redissemination of data would negatively affect both government and not-for-profit database value adders or disseminators in other ways as well. For example, European government meteorological data providers, who are already benefiting from the stronger protections offered by the E.U. Database Directive, are placing various use and redistribution restrictions on the National Oceanic and Atmospheric Administration (NOAA), asking NOAA to enforce these restrictions in the United States, contrary to existing U.S. law and policy. Such encumbrances from private-sector sources would be exacerbated by any database legislation that, similar to the E.U. Database Directive, extended protection to elements not now protected. Government S&T managers, in particular, are concerned that they do not receive enough funding to pay license fees and enforce restrictive provisions, in addition to meeting the costs of data collection and database preparation, and anticipate that they might have to decline data contributed by private-sector sources (as well as public-sector European sources) that carry high royalties or restrictions on subsequent distribution that require enforcement by the user. 31   See Laura D'Andrea Tyson and Edward F. Sherry (1997), Statutory Protection for Databases: Economic & Public Policy Issues, research paper prepared under contract to Reed-Elsevier, Inc. and The Thompson Corporation, and presented as testimony on behalf of the Information Industry Association at the October 23, 1997, Hearing, note 8. Tyson and Sherry, however, generally argue that there are not many instances in the commercial database industry in which sole-sources dominate the market and can prevent or inhibit entry. Although the committee did not analyze the entire database market in this study, it did find that in many S&T areas, including practically all observational databases, the data sources are unique. 32   Federal Trade Commission letter, note 13, p. 2.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases With increased statutory protection for databases and the accompanying higher transaction costs, scientists and educators in the not-for-profit sector might no longer be able to afford access to newly proprietary data sources or to enforce subsequent access and use restrictions on the data obtained from those sources, contrary to existing norms and practice.33 Not-for-profit research institutions tend to be conservative, risk-averse organizations that err on the side of caution, and they would likely institute guidelines prohibiting any database research activity that might potentially expose them to liability under a new legislative regime and to costly litigation. Such a possibility is particularly problematic given the uncertainty about what portions of databases would be deemed "qualitatively substantial" by the rights holders in each case and about what they would view as a "reasonable use" by not-for-profit entities. Such defensive measures would serve to further restrict, perhaps even beyond what might be allowed under the law, what scientists and educators can do with databases that they lawfully obtain. Providing stronger property rights for databases that contain information of high commercial value, such as in the area of genomic research, can have the opposite of the intended effect, because the price of access to these databases is inversely related to the number of users who will access them. Hence, from an S&T perspective, the goal is to encourage the generation of publicly funded, and therefore readily available, collections of data in key scientific areas, where the use of this information is of potentially great commercial value, and to discourage the tendency for private companies to capture this information and restrict access to a limited audience. Promoting broad access to publicly generated databases has the additional benefit of fostering active competition and value-adding activity since all commercial and academic organizations would have access to this information. Moreover, enhancing database protection would also serve as an incentive to both government agencies and not-for-profit organizations to privatize or commercialize their research databases. Such action would have the undesirable outcome of reducing the number of databases in the public domain and thus would have a chilling effect on the full and open data exchange and sharing ethos that benefits so many areas of scientific and engineering research. Since a strong case for significantly greater protection of databases has not been made, primarily because existing protections already go a long way toward protecting database providers, the committee believes that the concerns regarding 33   Examples of this problem are already abundant in the restrictions on experimental research uses of patentable or otherwise protected innovations in the biotechnology sector. See M.A. Heller and R.S. Eisenberg (1998), "Can Patents Deter Innovation? The Anticommons in Biomedical Research," Science, Vol. 280, p. 698. Such a problem would be more insidious in the case of noncopyrightable factual databases, which are subpatentable innovations that do not merit strong property rights and that have been used much more widely and openly in research to date.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases increased encumbrances on access and use, as well as the potential for higher prices and related transaction costs, cannot be ignored. In light of these concerns, the committee recommends that any new federal statutory protection of databases should limit any additional protection to prohibition of acts of unauthorized taking that cause substantial competitive injury to the database rights holder in the rights holder's actual market. The standard of harm should be sufficiently clear to permit good-faith users to know when they are infringing on a database rights holder's rights and should not undermine the nation's capabilities for innovation or competition in the marketplace . Such a formulation would help prevent undue and inappropriate interference with scientific inquiry and with other traditional and customary public-interest uses of data, as well as promote legitimate and socially beneficial commercial competitive activities. Scope of Protection The first section of all three of the legislative proposals considered by the committee defined a database as a "collection of information collected and organized for the purpose of facilitating access to discrete items of the information." All three proposals also provided protection to databases developed through the investment of substantial monetary, or other resources. "Information" was defined to include data, facts, or other intangible material capable of being collected and organized in a systematic way. The Coalition Proposal, however, excluded "works of authorship"—a term applicable to subject matter protected by the copyright system. Such an exclusion would deny protection to copyrightable works such as anthologies of an author's works or a scientific journal that might otherwise be regarded as a database of articles. The H.R. 354 and Senate Discussion Draft proposals included these works, consistent with the subject matter scope of the E.U. Directive. The committee believes that the inclusion of collections of works of authorship, which are already unambiguously protected by copyright, is both unnecessary and unsupportable. If the purpose of this legislation is to fill a purported gap in the legal protection currently available to noncopyrightable databases, then that scope of protection should not extend so broadly as to cover fully copyrightable anthologies, journals, and textbooks. The committee therefore recommends that the subject matter scope of any new federal statutory protection of databases be constrained to databases comprising a collection of discrete facts and items of information, and expressly exclude collections of copyrightable material, which is already protected. Further, protection under any new statute should extend only to a database that is the product of a substantial investment, and not to any idea, fact, procedure, system, method of operation, concept, principle, or discovery disclosed by the database.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases research communities upon which innovation and broad-based economic growth ultimately depend. The committee thus recommends that although private-sector databases derived from government data should be eligible for protection, protection should not be extended to databases collected or maintained by the government. Any new legislation should expressly affirm the need for continuation of existing legal norms for wide distribution of government data and of data created pursuant to a government mandate or funding. ASSESSMENT OF POLICY OPTIONS, WITH RECOMMENDATIONS FOR GOVERNMENT ACTION In this section the committee discusses a number of actions that should be taken by various government institutions to help promote access to and use of S&T databases for the public interest. The areas addressed include promoting availability of government S&T data; maintaining nonexclusive rights in government-funded databases by not-for-profit institutions and their employees; organizing discussions of licensing terms for not-for-profit uses of commercial S&T databases; improving the understanding of complex economic aspects of S&T database activities; and promoting international access to S&T data. Although the committee believes that its recommended actions in these areas ought to be undertaken whether or not any new statutory database protection is enacted by Congress, all of these actions will take on an increased urgency and importance if relatively strong new proprietary rights in databases are established by federal statute. Promoting Availability of Government Scientific and Technical Data Increased proprietary protection for commercial databases could have a significant effect on government data collection and distribution efforts. Because researchers and educators likely would be more constrained in their use of data drawn from commercial databases, they might have to request additional funds for the purchase and administration of proprietary data or ask federal agencies to collect and maintain more S&T data on a nonproprietary basis. Thus budgetary strains could increase for federal agencies trying to meet the data needs of their own researchers as well as those related to fulfilling institutional mandates. Under appropriate circumstances and conditions, government partnering with the private-sector—especially not-for-profit institutions—in accomplishing data collection and maintenance can be highly beneficial and effective. The committee fully endorses the existing policy and practice of the federal government, as expressed through OMB Circular A-130, to make public S&T (and other) databases openly available at the lowest possible prices. It is through this policy of

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases efficient-access pricing that the taxpayer derives maximum value from the government's very substantial investments in its collections of data. Consistent with current practice, government S&T agencies must not make their databases—whether created and owned by them or under their control—available on an exclusive basis. They also should continue to maintain under their own control and archive all S&T databases that have value for research and that are otherwise being disseminated on behalf of the government by a private-sector organization or company. Such control should be maintained through physical possession or by appropriate contractual provisions. The long-term maintenance of public databases and archiving of data in readily accessible formats are essential to ensure their availability for reuse in future research or to confirm the results of research already conducted, among other uses.93 Large quantities of government and government-funded data at all levels are lost, discarded, or rendered inaccessible owing to technological change or defects. Although this constitutes a major information management and policy issue in its own right that is beyond the scope of this report, the trend toward greater private-sector management and dissemination of public data—which the committee believes would increase under stronger statutory protection of databases—makes it even more important for government agencies to pay attention to this issue. Without adequate safeguards to ensure long-term preservation of public data created or disseminated on behalf of the government by private-sector entities, even larger amounts of such data may be lost or become inaccessible over time. Finally, in making its data broadly available, the government should require that all private-sector disseminators or transformative users of its data identify the government source(s) of the data being used. Indeed, the same practice should be followed with regard to all sources of data from the private-sector as well. Identifiers on privately disseminated government data will serve the objectives of providing notice to all users that they can contact the government agency source to obtain the original data, making the public aware of its government's activities, and giving proper credit where credit is due. Improving public awareness is an important objective, because all too often the public lacks a full appreciation of the benefits it derives from taxpayer-funded data-related activities. The committee therefore recommends that the following actions be taken by all government entities. Scientific and technical data owned or controlled by the government should be made available for use by not-for-profit and commercial entities alike on a nonexclusive basis and should be disseminated to all users at no more than the marginal cost of reproduction and distribution, whenever possible. While the private-sector's creation of derivative databases from government data should be encouraged, the source 93   See generally National Research Council (1995), Preserving Scientific Data on Our Physical Universe: A New Strategy for Archiving Our Nation's Scientific Resources, National Academy Press, Washington, D.C.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases of the original government data must ensure that those original data remain openly available. Any information product derived from a government database also should be required to carry an identifier stating the government source(s) used. Maintaining Nonexclusive Rights by Not-for-Profits in Government-Funded Databases Science best advances by promoting a culture of openness and sharing, whereas individual commercial companies best advance by maintaining control and secrecy. The tension between these two cultures has been attenuated through the ''first sale" doctrine, whereby a purchaser of a book, journal, or other intellectual work is free to use the facts and ideas in the work and the publisher is not able to prevent the purchaser from placing the intellectual work in a library or passing it on to another person for similar uses, regardless of the medium in which the work is presented.94 As discussed in Chapter 3, this right is being superseded increasingly by licensing arrangements in the online digital environment. In the academic sector, such institutional licenses typically permit users to share with collaborators or colleagues on other campuses, so long as the sharing is not systematic. In fact, in some cases, universities are now able to negotiate licenses that "meet or exceed" their user's needs.95 Nevertheless, in order to maintain a reasonable balance between the scientific and education communities' interests in openness and sharing, on the one hand, and the commercial community's interests in exclusivity on the other, some minimal constraints ought to be placed on the commercial community to guarantee researchers and educators access to and unfettered use of facts, data, and intellectual works published by their peers. The existing practice in the publication of research results has been for researchers to pay page charges and to contractually give up exclusive copyright in their works in order to have their articles published in the primary journals read 94   17 U.S.C., section 109. 95   Personal communication from Ann Okerson, Yale University, September 1999. An example of this type of licensing language is contained in Academic Press's IDEAL license (for 200+ journals):     Copying and storing is limited to single copies of a reasonable number of individual items. Downloading an entire issue of a journal is not permitted. However, digital or print copies may be included in coursepacks and reserves, or in internal corporate training programs and drug application materials. Authorized Users may transmit downloaded copies of individual items to persons who are not Authorized Users for the purpose of scholarly communication, so long as the transmission is not done on a systematic basis.     At the same time, the traditional user rights under the first sale doctrine are in danger of being significantly eroded by the Uniform Computer Information Transactions Act, which is currently being considered for enactment at the state-level. See generally, "A Guide to the Proposed Uniform Computer Information Transactions Act" at <www.2bguide.com>.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases by their peers. Because authors transfer exclusive copyright in their work, they may be legally obligated to ask publishers for permission to distribute copies of their authored articles to their own students and to their close research associates. As one participant in the committee's January 1999 workshop wryly noted, "We have a great system: we pay to publish and we pay to get it back." In addition, the pricing structure that seems to maximize profits for commercial scientific publishers is one that limits acquisition of journals to the elite academic libraries and researchers that can afford them.96 Many academic institutions now have difficulty paying library subscription rates, and their researchers, professors, and students thus lack convenient access to many journals, even to those in which they publish. Such concerns will only increase as electronic publishing becomes more widespread. It is already common practice in electronic publishing—and one of its tremendously productive features—to link electronic articles to the data sets upon which the research results depend. If legislation protecting databases is enacted, the current practice of requiring scientific authors to give up exclusive rights in their research articles on a take-it-or-leave-it basis could be extended to the data sets underlying the results reported in the research articles. Ceding control of databases created in the not-for-profit sector, especially those created with taxpayer support, to private-sector vendors that can establish their own terms for access to and use of the underlying research data is thus a major concern. In scientific disciplines where marketplace competition is highly constrained or absent, there is a need to provide safeguards against monopolistic practices. The committee therefore believes it important to initiate a "safety net" approach in the digital database context to help preserve the balance previously provided by the "first sale" doctrine. This approach will help ensure public access to data and databases developed in whole or in substantial part at federal government expense. Databases developed primarily with government funds should not fall under the exclusive control of private parties such that dissemination of the data to the public or other scientists is limited. Nor should public access to government-funded databases be highly constrained. The economic basis for funding science from governmental funds is that the research produces public goods. One researcher's use of these public goods does not decrease the value and benefits to others of the public goods. Specifically, for any research accomplished wholly or in substantial part with federal funds, universities and not-for-profit organizations should be re 96   See Association for Research Libraries (1999), ARL Statistics: 1997-98, Martha Kyrillidou et al., eds., Association of Research Libraries, Washington, D.C., also available online at <www.arl.org/stats/arlstat/1997-t2.html>, showing trends in average rise in costs of serial (journal) subscriptions between 1986 and 1998, pp. 8-9. For a retrospective look at these issues see <www.lib.virginia.edu/mellon/mellon.html>.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases quired by the funding agency to retain nonexclusive rights in any resulting databases. Under OMB Circular A-110, federal grant recipients have initial control over the intellectual property and databases that have been produced from their federally funded projects.97 The primary concern under a new statutory regime is the inequality in bargaining power between large publishers and individual researchers and scientific authors. Based on past practices, the committee is concerned that many researchers may be required to give up exclusive rights in the databases produced at federal expense in return for having their research results published. If new database legislation is enacted, publishers may request rights in both the intellectual work (i.e., typically the journal article), as well as rights to the collected data sets from which the intellectual work arose or upon which the work may depend and which might usefully be linked to the article in electronic publishing environments. If the negotiated contract provides for reasonable access to and use of the government-funded data for further scientific work, it is unlikely that the right of the researcher to independently provide access to the government-funded data would ever have to be invoked. However, the proposed provision provides a safety mechanism. The retained right of the researcher to distribute the data is likely to be invoked only in the unusual situation in which data gathered through a federally funded project or grant have been transferred through contract from an academic institution to a commercial entity with highly constrained access to and use of the data. This approach has a relatively narrow application. The proposed requirement would not apply generally to copyrighted works that may have been produced using federal funds (e.g., research articles), nor would it apply to state or local government databases or to databases generally. The requirement also would not automatically apply to databases created with only partial (e.g., less than half) funding from the federal government, unless specifically agreed to by the parties. The provision would allow not-for-profit institutions and researchers to share underlying federally funded data with others regardless of contract provisions with the private-sector, but would impose no affirmative requirement on them to share such data. Universities and other not-for-profits, of course, could not distribute any value-added features provided by the private-sector publisher to the government data unless agreed to by contract with that entity or otherwise permitted by law. Based on the foregoing discussion, the committee recommends that federal funding agencies should require university and other not-for-profit researchers or their employing institutions that use federal funds, wholly or in 97   Office of Management and Budget (1997), Circular A-110, "Uniform Administrative Requirements for Grants and Agreements with Institutions of Higher Education, Hospitals, and Other Not-for-profit Organizations," revised November 19, 1993; as further amended August 29, 1997.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases substantial part, in creating databases not to grant exclusive rights to such databases when submitting them for publication or for incorporation into other databases. Organizing Discussions of Licensing Terms for Not-for-Profit Uses of Commercial Scientific and Technical Databases Whether or not new database protection legislation is adopted, the committee believes that representatives from the not-for-profit research and education communities should engage in a series of discussions with commercial database publishers and vendors in different market segments in order to achieve a better understanding of their respective needs and concerns and thus foster the development of mutually acceptable licensing terms that can reduce uncertainty and transaction costs. Such discussions would be especially important in the months and years immediately after enactment of a any new federal database statute, since there would be many definitions and concepts that would not have been fully defined and that would be subject to broadly divergent interpretations by different parties. One person's legitimate derivative use may be another's harmful infringement. Previously established guidelines or understandings concerning copyrighted works will not in most cases be transferable to the database context and therefore will most likely confuse user communities, without the benefit of a fresh set of clarifying discussions. Further complications will arise if currently copyrighted works, such as journals, textbooks, reference books, and other anthologies, are also included in the definition of protected databases or "collections of information" under any new U.S. legislation, as they already are in the European Union. In addition to promoting some mutual understanding regarding licensing terms, clarifying discussions might help prevent unnecessary conflicts and litigation. It is unrealistic to assume that a model contract or even standard individual contract terms could be developed to cover all or perhaps even most such transactions. As discussed in Chapter 1, a key characteristic of S&T data is the heterogeneity of data types, sources, and uses. The expectation of developing a one-size-fits-all approach would be not only illusory and impossible, but also ultimately harmful. To avoid becoming futile, discussions among stakeholders must be founded on realistic and well-focused objectives that would have a reasonable chance of success. In establishing such discussions, it is essential that representatives of all major stakeholders be involved so that all relevant interests and viewpoints can be considered. For example, the committee would not endorse a process such as the one that resulted in the "Agreement on Guidelines for Classroom Copying."98 98   For a history of academic fair use and classroom guidelines, see Kenneth D. Crews (1993), Copyright, Fair Use and the Challenge of Universities, University of Chicago Press, Chicago, IL.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases That agreement has perhaps been workable for campus administrators, campus libraries, and the photocopying centers on campuses, but not for students and faculty, who were not involved as stakeholders in the discussions. Examples of both classroom guidelines99 and of existing digital licensing terms and phrases, and their evaluation from the not-for-profit perspective,100 may be found online on the World Wide Web already. Private-sector S&T database producers and disseminators should remain cognizant of the social value of their products, particularly for not-for-profit research, education, and other public-interest uses. Database vendors whose primary source of revenue lies outside the not-for-profit S&T communities should endeavor to provide public-interest users access to their databases on favorable terms. Database vendors whose primary source of revenue is in the S&T research and education community should be encouraged to provide access on favorable terms once a reasonable return on investment has been achieved. Indeed, as noted in Chapter 2, all pricing inhibits access, especially for those researchers who do not have adequate and strong institutional funding, whether academic, research institute, or industrial. Of course, the limitations are a matter of degree, depending on level and pattern of pricing. The goal should be to bring all sectors into a cooperative system in which data are made widely and readily available for scientific and educational use at as low a total cost (to the user population and society as a whole) as possible, and to do that within an environment that encourages, rather than inhibits, the inquisitiveness and inventiveness of the user while encouraging the entrepreneurship of suppliers. It is in the common interest of both database rights holders and users—and of society generally—to achieve a workable balance among the respective interests so that all legitimate rights remain reasonably protected. The participants in these discussions would be primarily representatives of commercial S&T database disseminators and their government agency and not-for-profit-sector users. The committee makes its recommendation to the administration, rather than directly to those communities, however, because it believes that the discussions should be held under a convener such as the Copyright Office, which has the greatest subject matter expertise in these issues within the government. Such a focused venue would not only help stimulate progress on important issues, but would also mitigate the potential for accusations of collusion or conspiracy under federal antitrust laws. The committee therefore recommends that the Copyright Office sponsor discussions between the representatives of private-sector producers of databases and user stakeholder representatives from government agencies and not-for-profit groups to help 99   Some examples of classroom guidelines may be found online at <fairuse.stanford.edu/library/index.html>. 100   See <www.library.yale.edu>, and type "licensing" in the "Search" box.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases develop a common understanding and optimal terms for the licensing of S&T databases and data products. Improving the Understanding of Complex Economic Aspects of Scientific and Technical Database Activities Although a detailed economic analysis is well beyond the scope of the committee's charge for this study, this report raises significant questions throughout regarding the adoption of statutory database protection; the economic underpinnings of different types and mixes of provisions, and the potential effects of both an overall statutory regime and specific provisions on various segments of the database industry and on the relationships among the different parties involved in creating, disseminating, and using S&T databases. Certainly, at a minimum, the questions raised in the E.U. Database Directive and in the Senate Discussion Draft, as well as any other questions that are ultimately identified in the course of the legislative process, should be the subject of more detailed study in advance of any legislatively mandated report on effects of increased protection. Such a study would help provide a comprehensive base of knowledge with which to officially evaluate the effects of new statutory protection of databases. In addition to these broad economic issues, the committee suggests that research be devoted to, among others, the following specific issues affecting the creation, dissemination, and use of S&T databases by the government and by the not-for-profit and for-profit sectors: Is there an increase in the number of databases available that can be related to (caused or affected by) an increase in database protection? Is there an increase in the costs (or a reduction in the amount) of scientific research that can be related to an increase in database protection? What is the role of licensing restrictions on access to and use of databases for research and education activities and what are the effects? What are the effects of sole-source databases on S&T niche markets and on the level of scientific research? What are the trends in privatization of government S&T data, and what effects has privatization had on access to and use of such data? What have been the effects of the E.U. Database Directive on access to and use of European S&T data by the European research and education community, and by the U.S. government and by the U.S. research and education community? The committee recommends that the Congressional Research Service, the National Science Foundation, the Department of Commerce, and other federal science agencies, as considered appropriate, should undertake and fund external research that investigates the changing and complex economic

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases aspects of S&T database activities, particularly in the context of any new legislative database protection measures that may be enacted and in support of the legislative principle recommended above regarding the conduct of periodic assessments of the effects of any new statutory protection of databases. Promoting International Access to Scientific and Technical Data It is a well-known truism that science knows no boundaries and that practically all research that is conducted on an open basis also involves international collaboration to some degree. Some research, such as that in observational space and Earth sciences, is inherently international and cannot be conducted successfully without either the collection of global data or access to foreign databases.101 As a result, the U.S. government science agencies in recent decades have concluded thousands of bilateral and multilateral general S&T cooperation and specific research program agreements.102 These agreements will take on added significance with the implementation of the E.U. Database Directive and with possible adoption of restrictive database protection legislation in the United States and elsewhere, since the negotiated terms of those agreements can specify the terms under which databases related to the research in question can be accessed and used. As the world's largest producer and disseminator of S&T data, the U.S. government has significant leverage in negotiating appropriate terms for the exchange and use of public data with other nations. At the same time, the committee agrees with the Administration's concerns regarding the E.U. Directive's reciprocity provision and supports the U.S. Trade Representative's (USTR's) placement of that topic on the Administration's 1998 Special 301 Review.103 The committee would go one step further, however, and suggest that the USTR and other appropriate entities within the Administration negotiate with the Commission of the European Communities to review and revise its E.U. Directive, based on the substantial criticisms of that new legal regime in this report and in other position statements and articles cited above. If the U.S. Congress enacts a new database protection statute based on properly balanced unfair competition principles, the committee urges the USTR, the U.S. Patent and Trademark Office, and other appropriate administration officials to 101   For example, for a comprehensive listing of most internationally available data sets from space missions, see the NASA Goddard Space Flight Center's National Space Science Data Center home page online at <nssdc.gsfc.nasa.gov>. For a listing of many international Web sites covering all aspects of Earth science data, see the NASA Global Change Master Directory online at <http://gcmd.gsfc.nasa.gov/pointers/pointwais.html>. 102   For a discussion of some of the large international research programs, see National Research Council (1997), Bits of Power, note 49, pp. 58-61. 103   See Pincus statement, March 18, 1999, Heating, note 7, p. 33.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases promote that statute as a model for international database protection within the World Intellectual Property Organization. The committee recommends that all departments and agencies of the federal government should continue to adopt international S&T agreements that include provisions to facilitate access to S&T data across national boundaries and should conduct periodic reviews of international policies and agreements to promote conformity to the above principles. In addition, the committee recommends that the U.S. government should negotiate with the Commission of the European Communities to revise its highly protectionist E.U. Database Directive. RECOMMENDED APPROACH FOR THE NOT-FOR-PROFIT SCIENTIFIC AND TECHNICAL COMMUNITY Finally, there is the question of what the research and education community should do in the event that highly restrictive statutory protection of databases is enacted by Congress. Certainly, leaders in all the major not-for-profit research, higher education, and library associations and in many individual institutions have voiced their concerns about the legislative proposals that have been introduced in the Committee on the Judiciary in the House of Representatives in both the 105th and 106th Congress.104 As the various critics and this report point out, such a new statutory regime could have many negative effects, among them significant changes in the terms for access to and use of databases sold or licensed by the commercialsector, the possibility of increased economic exploitation on a proprietary basis of heretofore openly available S&T databases by not-for-profit researchers and educators and their institutions, and the stimulation of further privatization of such public good activities by the government. The question thus raised is what actions the not-for-profit community should take on its own behalf if restrictive new provisions are enacted that encourage or exacerbate these negative effects. Proponents of new legislation rightly point out that any new law would not require individuals or organizations to make use of the new protections, and that if the not-for-profits are concerned about these legal developments as a matter of principle, they should resist the temptation to adopt proprietary restrictions on their databases. The committee agrees that much of the responsibility for maintaining a policy of full and open data availability in the academic community rests with the community, which will have to act to ensure continuation of the broad sharing of data and research results. Nevertheless, the 104   See the testimony given by Wulf, Reichman, and Neal at the October 23, 1997, Hearing, note 8; by Stewart at the February 12, 1998, Hearing, note 9; and by Lederberg, Phelps, and Neal at the March 18, 1999, Hearing, note 29. See also the position statement signed by representatives of many of these organizations, note 7.

OCR for page 73
A Question of Balance: Private Rights and the Public Interest in Scientific and Technical Databases committee also believes that the pressures to commercialize and privatize currently open data sources would increase inevitably under a regime such as the E.U. Directive or the one proposed by H.R. 354, and that fully maintaining the customary or traditional approaches to data exchange would prove to be difficult. In the event that the universe of public domain S&T data is found to be shrinking unacceptably, additional defensive measures may have to be taken to reinvigorate a robust public-interest sector for such data. Therefore, as its last recommendation, the committee urges that the not-for-profit S&T community continue to promote and adhere to the policy of full and open exchange of data at both the national and international levels.