The Legal and Policy Framework for Global Electronic Commerce

The Legal and Policy Framework for Global Electronic Commerce: A Progress Report

Developments in information technology set electronic data collections apart from all previous regulations and considerations on the use of information. Attempts to protect the publisher include encryption devices, contracts to gain electronic access, and discussions of reworking intellectual property rights. None of these measures provides the user with a satisfactory means of access comparable to existing copyright laws. The European Union's answer to the need for regulation in this area has been a sui generis model, offering strong proprietary interests and hybrid intellectual property rights.

Existing copyright laws protect both users and suppliers while still stimulating economic investment and fostering scientific research. The EU's sui generis model[2] or the US version of that model[3], however, disproportionately protects the interest of the publisher. The EU Directive protects the contents of any noncopyrightable database that is the product of substantial investment; protection lasts for fifteen years. The US provision, as part of the Digital Millenium Copyright Act, protects against use of a protected collection of information that is the product of substantial investment if the use would cause harm to actual or potential markets, although scientists and educators were amended to be responsible only for actual markets. The protection also lasts fifteen years, but would be extended following any substantial new investment. These provisions were revised in the States to include more serious limitation at fifteen years and an exemption for additional, reasonable uses by educators, scientists, and researchers[4].

The effects of this legislation is problematic. Access to data and information would not be available to secondary researchers without licenses at possibly several stages of retrieval and use. The secondary researcher would be further restricted against using the material for study or publication, and would be faced with the task of independently developing the data; regenerating data is often impossible and otherwise prohibitively costly. Cooperation and compilation among researchers would be blocked by administrative, legal, and cost barriers representing the publisher's redistribution rights. Competition and innovation would be stifled by data-holders' reluctance to distribute licenses to information, preferring instead to avoid competition and use the information exclusively. Thus, researchers would need to pay artificially high prices to access commercial databases while the sole-source character of the marketplace would be reinforced. Such regulation effectively offers the strongest scope of protection to the most suspect of area in intellectual property in contravention of existing customs of free access to information in research. The later provisions allowing for public access after fifteen years and exemptions for researchers are a move in the right direction, but don't go far enough. Identifying when protection expires remains contentious, and placing the onus on researchers to demonstrate reasonableness within an ambiguous and narrowly drawn provision provides an inherent impediment to research. The later provisions steer the legislation towards a "fair use" concept of database law without providing the safeguards found in traditional copyright law.

As all science operates on databases, and as every aspect of the natural world is now available from electronic databases, the regime would be potentially very damaging for science and technology. Science builds progressively on raw data, summarizing, synthesizing and processing data into new data sets. As the data become more processed, they are also likely to attract copyright protection; if property protection were assigned to noncopyrightable (raw) data sets, which previously had the least marketability but most value to research, it could disproportionately affect the availability of the data most useful to research. The new database laws place emphasis on protecting original investments while traditionally science has operated successfully on unfettered access to and re-use of data held in public domain, motivated by non-economic (academic and professional) incentives. The lack of licensing and contracting restrictions has the further potential to disadvantage researchers by limiting access to and use of collections of data through high prices and harsh terms for access and use. Transaction and administrative costs would increase as research institutions place additional restrictions on what researchers can do in acquiring and using data in order to avoid costly litigation. Value-adding uses of data collection (recompilation, re-use) would be discouraged as users will fear claims of harm to the originator’s actual or potential markets. Likewise, one of the greatest concerns is that rising transaction costs will inhibit the creation and exploitation of multiple-source data products (complex databases from multiple public and private sources), one of the effective means of addressing difficult problems.

An exclusive property right granted to sole-source data providers would strengthen the legal and economic protection of mini-monopolies. Costs of data acquisition would rise and would either be passed on to the taxpayer in the form of increased research grant request, or simply diminish the resources available to researchers. Downstream users would also be affected, discouraging socially and economically beneficial forms of exploiting factual data. The fifteen year term is both open to attack through constant updating and likely to represent such a long delay as to hamper the progress of science by effectively removing the data sets from comparative analysis. The fifteen year period is arbitrary and should be shortened. Independent creation of equivalent databases is not a solution as such activity is either impossible or prohibitively expensive, and contrary to the principle of research that new knowledge should build on the base of existing information. Duplication of previous factual compilations is socially and economically inefficient; the law should work to protect this interest as well. Access to government data may also be circumvented by failure to expressly require grantees and contractors to provide their data back to the government or make their data publicly available, ending in data, produced as a direct result of public funding, under proprietary control of researchers or their institutions. Universal government vigilance and government restraint from exploiting database protection as an income-generating opportunity would be required to prevent this possibility.

A new legal regime that encourages commercial exploitation of research would undermine the scientific cooperation. Science and technology would suffer as researcher would either exploit their colleagues or cut off cooperation altogether. The most profound impact of the proposed legislation is in the lost opportunity costs as researchers opt to repeat work performed elsewhere to avoid administrative time, legal hurdles, and prohibitive expenses. Thus, an overprotective database regime would seriously impede the use, reuse, and transformation of factual data that are the lifeblood of science and technology.

Several organizations have worked to block efforts to regulate databases according to the unbalanced EU Directive. Early efforts to slow legislation in the U.S. House of Representatives were not as successful; H.R. 2652 was pushed through the house and then attached to H.R. 2281 (Digital Millennium Copyright Act) in July 1998. More recently, the Administration has produced position papers supporting the interests of the scientific and educational communities, addressing issues of increased transaction costs, disruption of non-commercial research, sole-source provider burdening, constitutional obstacles, and deleterious effects on competition and innovation. The concerns of the U.S. Academies were presented and clarified to the Senate Committee under Senator Hatch in late 1998, including the need to protect publishers from free-riding and to preserve the incentive to invest through a true unfair competition regime without creating a strong property right in data collections. In response in part to the perceived over-empowerment of publishers, Senator Hatch advocated adopting the Hatch Database Discussion Draft of October 5, 1998, which largely reflected the Academies’ position.

Under the Draft, the strong proprietary right approach moved closer to a “misappropriation” (unfair competition) approach by conditioning liability on acts that “cause substantial harm to the actual or neighboring markets”, leaving courts to determine “substantial harm”. A full exemption immunized scientific activities instead of the “fair use” approach previously recommended, lifting the burden on researchers to show fair use applied. Thus, liability would only apply if non-profit researchers caused substantial harm by using unreasonable amounts of the collection, or if they produced a market substitute, or if they sought to avoid payment for the research tools; the onus remains on publishers to demonstrate infraction. Exemptions were extended to library and instructional uses. Concerns of permanent capture by private information of government data were addressed. The wording of the duration clause diminished the risk of perpetual protection beyond the fifteen year period. The need for regulation of licensing terms and conditions was recognized. The definition of databases was clarified to exclude ordinary literary works.

The Draft thus moved the Administration away from the strong exclusive property right approach adopted in the EU Directive towards a more balanced unfair competition approach, both protecting publishers and avoiding harm to research. Value-adding issues still need to be addressed. The Draft was not adopted under time pressures and the database component of the legislation was stripped from the Act; new negotiations began under the new Congress in 1999. The Draft nevertheless represents a balanced model of database protection. In the meantime, the scientific and technical communities will need to preserve the sharing ethos in scientific data, encourage price discrimination in favor of the scientific and technical communities, and develop differentiated products for the non-profit sector. These communities require fair and reasonable access to data, the ability to use the data, and freedom from contractual or technical interference; meeting these objectives will mean close collaboration with government.

Most proponents of a strong proprietary interest in database protection fail to consider the dual nature of data and information. While data function as raw materials that are fashioned by researchers into products, data and information together are bundled into downstream products that attract property rights. The property rights that are well suited to downstream applications are mistakenly applied to the upstream data as components of the innovation. This leads to balkanization of the public domain, increased transaction costs, and impeded access to the building blocks of knowledge. The lessons of the industrial revolution should be realized in allowing principles of unfair competition law to guide the information economy – allowing the reuse of ideas rather than monopolies on products of routine innovation. We should address the risk of market failure with underprotection; there will be no potential shortage of investment and, for reasons of public policy, we do not want to monopolize sources of data or retard the progress of science.

The placing of strong property rights in upstream collections contravenes our constitutional heritage. While it is claimed that the “harm to actual or potential markets” test was drawn from Copyright Laws, in the database arena it indirectly creates an exclusive property right in noncopyrightable collections. Facts and ideas that the copyright law must leave to unrestricted public use cannot constitutionally be withdrawn from public use under the First Amendment by a database law that protects against extraction and use. The protection of noncopyrightable data against use on both primary and secondary markets thus offends the constitutional culture; H.R. 354 institutes copyright-like protection on noncopyrightable material under the “harm to actual or potential markets” test, which effectively creates exclusive property rights. The broad definitions of “information” and “collection of information” further complicate the issue.

A true unfair competition approach, by contrast would only assign liability when harm was done by improper, unfair, or dishonest means, allowing honest “harm” and true transformative uses that promote competition, science, and education. The “harms to markets” test is too broad in that any successful competitor would by definition harm a prior entrant’s market; other industries might apply for similar protection, and the cumulative anti-competitive effects could undermine the U.S. position in the global marketplace. Traditional intellectual property law and tradition have mandated unfettered use of noncopyrightable facts and favored unbridled competition among the products of mere investment. While publishers are deserving of protection for their investments, enforcing strong monopolies in collections of data will balkanize the information economy. Instead, we should err on the side of caution and underprotect the building blocks of knowledge, thereby balancing public and private interests, and ensuring a high degree of competition and maximum dissemination of information. In the event of real economic harm, we can always adjust the level of protection upwards, but acquired rights cannot easily be eradicated

Reichman's article is largely a history of the legislative attempts to regulate data and information from the perspective of the concerns of education and research. He reviews the concerns surrounding the database portions of the Digital Millennium Copyright Act with respect to hearings, proposals, concerns, and benefits. He compares the legislative attempts to the existing EU Directive; he advises strongly against following the Directive’s model of providing a strong proprietary interest in noncopyrightable data collections. He advocates instead underprotection of data collection, arguing that if the fears of the proprietary interest proponents are realized in market effects, the regulations can be adjusted accordingly; whereas imposing overregulation may lead to stagnation of research that can't be corrected by later removal of regulation.

The article provides a thorough review of the development of the American legislation in the light of the EU Directive. Reichman approaches the review from the perspective of protecting the interest of researchers; his concerns regarding their access to information and the free availability of data are well founded within the framework of the legislation. However, his solutions and proposals seem rather narrowly conceived.

At times, the author seems to forget that the legislation does not have research materials and scientific databases as its target; rather, it means to protect literary works, acts of creativity, and collections that would normally and rightfully be allowed protection under copyright. The culture of science and research, as Reichman points out, has always been one of mutual and free sharing and access to information. On the one hand, the act is designed to provide a strong proprietary interest to databases which, but for their electronic form, would otherwise fall under existing copyright schemes, and on the other hand, the research community depends on its academic, scientific or research exemption from the scheme.

Reichman is trying to fit a square peg into a round hole. He’s trying to adapt the database regulations to fit the research exemption (in advocating underprotection), while forgetting the purpose of the regulation in providing protection for publishers. Even his support for the exemption provisions of the Draft don’t meet with his approval as the thrust of the majority of the document seems to allow for a loophole to the provisions. One gets the impression that the author would rather see database regulation tailored to the research community and provide for exceptions for the publishing community.

Perhaps a better solution, beyond the already extensive exemptions made for the academic and scientific communities under the Draft, would be a complete exemption for these communities, and separate legislation for each of the database purposes: commercial/publisher and academic/research. Commercial databases would be subject to strong proprietary rights, and access to databases would be subject to the publisher’s whim; academic databases would be regulated by separate legislation enforcing free access for academic and research purposes, and separate access barriers for commercial purposes. This approach would protect the publisher, both academic and commercial; allow for unimpeded research activity; permit the free market to reign in the commercial world; obligate adherence to government regulation in the academic world; and allow research for scientific and commercial purposes to operate concurrently. To a large extent, the Draft, through its exemptions for academics and research, accomplishes much of this separate legislation approach; however, as the author points out, the Draft lacks the positive obligations to regulate the academic access and availability of information, a problem solved in the separate legislation solution.

The Draft and the developments leading up to it are aimed at the commercial realm of data and information, and make valid and workable exemptions for science and research. Reichman’s focus on the effect of the legislation on research is unfair in that it presupposes that the legislation is primarily aimed at protecting the culture of academic research. The issues raised in his article can be addressed by way of exemption to legislation whose focus is commercial; the legislation reviewed to a large extent includes these exemptions and addresses his concerns. Further exemptions would satisfy his remaining concerns, but, in my opinion, concurrent legislation that treats academic and commercial databases separately would best satisfy the problems database legislation pose to research.

[1] 14 Berkely Tech. L.J. 793

[2] Directive 96/9/EC of the European Parliament, 1996 O.J. (L 77) 20.

[3] HR 2281, 105^th Cong. (1998).

[4] HR 354, 106^th Cong. (1999).