Statement of Marybeth Peters
The Register of Copyrights
before the
Subcommittee on Courts and Intellectual Property
Committee on the Judiciary

United States House of Representatives
105th Congress, 1st Session

October 23, 1997

Collections of Information Antipiracy Act (H.R. 2652)

Mr. Chairman, members of the Subcommittee, I am pleased to testify today on the “Collections of Information Antipiracy Act.” Collections of information, or databases, are increasingly important to the U.S. economy and culture, both as a component in the development of the global information infrastructure and as a means of facilitating scientific, educational and technological advancement. It is therefore vital to achieve an appropriate balance in shaping the legal incentives for their continued production and dissemination, in order to optimize the availability of reliable information to the public.

The economic and societal interests at stake are large, and the dangers on both sides are significant. Assembling large quantities of information requires the investment of considerable human and financial resources. However, the resulting database can be easily appropriated at a nominal price relative to the cost of its creation. With today's digital and scanning technologies, major investments in both online and hard copy databases can be hijacked with the stroke of a key. Unless producers are confident that their investments will not be destroyed, they will produce and disseminate fewer databases, and the public will lose the benefits of access to the information those databases would have offered.

The risks of over-protection are equally serious. The free flow of information is essential to the development of knowledge, technology and culture. Scientific research in particular depends on full and open access to data. The key to legislation in this area is to ensure adequate incentives for investment, without inhibiting access for appropriate purposes and in appropriate circumstances. The net result of a well-crafted bill should be to make more facts more easily available. We believe this result is achievable, but not easy.

In our view, the goal should be to restore the general level of protection that was available under “sweat of the brow” copyright theories, but under a suitable Constitutional power, while bearing in mind the technological and market changes that have taken place since 1991. Flexibility should be built in for uses in the public interest in a manner similar to the function played by fair use in copyright law. Experience has shown that such a system can work to the public benefit; science, education and technology developed successfully under the “sweat of the brow” regime in the United States for 200 years.

The Copyright Office believes that H.R. 2652 presents a constructive new approach to the issue of database protection. The proposed legislation would fill a gap in the law, by seeking to ensure adequate incentives for the production and dissemination of databases while safeguarding important user interests. As the Chairman has made clear, however, the bill is only a starting point in this endeavor. Various issues remain to be addressed, and changes will need to be made, in order to resolve continuing concerns and achieve an outcome that maximizes the public interest.


The issue of legal protection for compiled information has been the focus of recent attention and debate in the United States and abroad. In this country, compilers express concern about changes in the legal landscape for databases since 1991, when the Supreme Court established a restrictive interpretation of the availability and scope of copyright protection in Feist Publications v. Rural Telephone Service Co., 499 U.S. 340 (1991).

Internationally, the matter of database protection has been raised in ongoing discussions of a possible treaty in the World Intellectual Property Organization and in a directive adopted last year by the European Union. As of January 1, 1998, the member states of the European Union will be required to provide, in addition to copyright, a new form of sui generis protection for the contents of databases. This protection will not be made available to databases from other countries unless those countries offer similar protection in their own legislation.

At the request of the Honorable Orrin G. Hatch, Chairman of the Senate Committee on the Judiciary, the Copyright Office earlier this year prepared a comprehensive Report on Legal Protection for Databases (the “Report”). The Report was the outcome of several months of study and numerous meetings with interested parties. It summarizes the past and present state of domestic and international law regarding database protection, describes past proposals to create a new intellectual property right in databases, and identifies issues and concerns raised by such proposals. In the Report, the Copyright Office sought to synthesize the views expressed in the meetings we held, and to provide a helpful background for a Congressional determination of appropriate legislative policy. In August 1997, the Report was submitted to Senator Hatch, with copies provided to the members of this Subcommittee and made available to the public over the Internet. A copy is submitted with this testimony for the record.

We understand that the Subcommittee staff has also held a series of meetings with interested parties. As the Chairman indicated in his statement, H.R. 2652 represents an effort to begin the process of achieving a satisfactory compromise of their varying needs and concerns.

General Comments

The threshold question, which occupied much of the time in the Copyright Office meetings, is whether there is a need for new legislation to protect databases in the United States. The introduction of H.R. 2652 was the result of Chairman Coble's conclusion that legislation is advisable. The Copyright Office agrees with this assessment.

In our view, there has been a gap in U.S. law since the Supreme Court's decision in Feist—a gap which has actual and potential negative consequences from the perspective of the broad public interest. While various bodies of law continue to protect database producers, each falls short in coverage, as described more fully below. The bottom line is that in many circumstances there is no legal recourse for a database producer when the essence of the value of the database, and the core of its investment, are taken without permission or compensation.

This situation is essentially the result of historical accident. For two centuries, compilations of fact were protected under copyright law whenever they were created through substantial investment or effort. See Report at 3-7. As copyright law developed, this line of cases continued, but became more and more inconsistent with general principles of copyright, which viewed creative authorship as the touchstone of protection. Id. While enactment of the 1976 Copyright Act appeared to reject the “sweat of the brow” rationale for protection of compilations, id. at 6-8, it was not until 1991, when the Supreme Court decided Feist, that it became clear that copyright could not be used to protect investment without creativity.

During the era of “sweat of the brow” copyright protection, it was not necessary for other bodies of law to develop specifically to address this kind of investment. Since Feist, the source and extent of legal protection for the commercially valuable contents of databases has been uncertain, requiring reliance on a patchwork of different, individually insufficient legal theories.

After examining the post-Feist case law and listening to the participants in the meetings we held, the Copyright Office is convinced that the theoretical gap in the law is leading to real-world consequences that should be remedied or prevented. First, given the thin scope of protection that copyright can provide, database producers have in several cases been unable to obtain relief from the courts against extensive, competitive copying. See Report at 13-18. In our legal system, such precedents may be relied on by others seeking to minimize costs and maximize profits by benefiting from the prior investments of their competitors.

Second, it appears that the erosion of incentives has already begun to have an impact. We have heard reports of reluctance on the part of many producers to create legally vulnerable database products, or to disseminate them widely to the public, particularly over the Internet. In addition, database producers may be choosing not to make their products available in formats that are appealing to consumers but easily copied, such as CD-ROMs. While we have no way of ascertaining the extent of the impact, it seems to be a logical response to the uncertain protection of U.S. law, and one that is detrimental to the public.

Another possible reaction by producers may be the use of highly restrictive contractual terms and/or extremely protective technological measures. The impact of either on the public could be more detrimental than a balanced form of legal protection.

Finally, after January 1998, the absence of protection in the United States will prevent the extension of sui generis protection to U.S. databases in the countries of the European Union. Unless and until the reciprocity provision of the European Directive is withdrawn, U.S. producers are therefore likely to suffer some degree of competitive disadvantage vis-a-vis their European counterparts. Large producers may be able to avoid the harm by setting up commercial establishments within the territory of the European Union, but smaller ones may not have this option available to them.

The argument has been made that Congress should not legislate in this area without empirical evidence of harm to database producers. Of course, empirical evidence can be extremely helpful in setting legislative policy. The fundamental question, however, is not the extent to which database producers are harmed, but the extent to which legislation can be crafted that will overall lead to greater availability of information to the public. Economic theory suggests that this result is possible,1 and it is not clear to us whether the question can be answered conclusively through empirical evidence. If statistics are gathered showing that the database industry is doing well today, for example, they will not tell us whether more databases would be more widely disseminated to the public under an appropriately designed federal law.2

Congress has not historically required empirical evidence to legislate market conduct. This includes the area of intellectual property, where exceptions as well as rights have over the years been added, expanded, or clarified based on individual cases or on concerns about potential future applications of the law. While the implications of legislation dealing with information are profound, and legislators should therefore proceed with care, a requirement of conclusive empirical evidence may impose an unmeetable burden. Congress should be able to take steps to prevent future harm, before substantial damage is done to particular parties or, more importantly, the public interest generally.

The Copyright Office therefore concludes that legislation is desirable to address the shortcomings that have been identified in the current state of the law. At the same time, we recognize that this is an extremely sensitive area; there are substantial dangers inherent in establishing legal rights involving the use of facts. It is important not to inhibit or raise the cost of existing uses in the public interest. Care is needed to ensure that incentives are provided to collect and disseminate facts in value-added form, while avoiding making access for legitimate purposes more difficult or expensive.

The Copyright Office believes that H.R. 2652 represents a constructive first step toward this goal. The bill takes into account the major issues and concerns identified in the Report. It adopts a misappropriation approach, founded not on the subject matter of what is taken, but on acts that are harmful, destroying the value of investment. It recognizes valuable categories of conduct that should be permitted, and provides exclusions to safeguard important public policies. At same time, the bill maintains the freedom of the marketplace in setting terms for use, allowing the continuation and further development of multi-tier flexible pricing mechanisms.

Because of the complexity and sensitivity of the issues, however, we are continuing to study the bill's provisions. We believe that careful analysis and consultation are needed, with particular attention to the scope of the permitted acts and exclusions, and the issue of duration. We look forward to discussing specific responses and suggestions with those affected, with a view toward amendments that avoid possible negative consequences. We would be pleased to assist the Subcommittee as it seeks to calibrate an appropriate balance.

The Shortcomings of Existing Law

As discussed in the Copyright Office Report at pages 73-85, various existing bodies of law protect certain aspects of databases, or protect them in certain circumstances. None, however, squarely addresses the core threat to their value in today's technological environment. What is lacking is a reliable, predictable means to prevent the taking, by parties with no relationship to the producer, of the contents of a database produced through substantial investment.


Recent changes in copyright law have impeded the ability of database producers to obtain satisfactory legal relief in many circumstances. As described in more detail in Section I of the Report, beginning in the late 18th century American courts protected compilations against copying based on the compiler's investment of time, effort and/or money. They found infringement even where the copier had taken nothing but facts.

It was not until 1991 that this “sweat of the brow” basis for compilation copyright protection was definitively rejected, when the Supreme Court decided Feist Publications, Inc. v. Rural Telephone Service Co. Feist held that a compilation is eligible for copyright protection only if it evinces a modicum of creativity in the selection, coordination or arrangement of its contents. 499 U.S. 340, 357 (1991). Moreover, the scope of protection extends only to those original elements of selection, coordination or arrangement that make the compilation as a whole protectible. The raw facts can be copied at will. Id. at 349.

While Feist's standard of creativity is minimal, some of the most useful databases may not meet it. The value of many databases to users lies in their inclusion of an entire universe of relevant facts in the most obvious and easily accessible format possible. Without choices made as to which of many items to include, or judgment exercised as to how to arrange those items, the threshold for copyright protection will not be met. Thus, the very comprehensiveness and user-friendliness of such a database may account both for its commercial value and its lack of meaningful copyright protection. In particular, the status of the large electronic databases that play an increasingly important role in the Internet environment has been left in doubt, given their inclusive coverage and lack of a fixed arrangement.

Even if a database is protectible, the scope of the copyright may be too narrow to prevent substantial competitive copying of its contents. Courts since Feist have taken to heart its teaching that copyright in a compilation is “thin.” 499 U.S. at 349. A number of district and circuit courts have found wholesale takings from copyrightable compilations to be non-infringing. See cases cited in Report at 13-17.

Given the limited scope of copyright for compilations as set out in Feist and interpreted by the lower courts, copyright law alone is insufficient to protect the investment of a compiler.


Protection may also be available under the state law doctrine of misappropriation. The tort of misappropriation is not well-defined, however, and varies from state to state. See Report at 82-85. There is no certainty that the doctrine will be consistently interpreted or applied throughout the country.

Furthermore, in order to avoid preemption by copyright law, the state doctrine of misappropriation may be limited to protecting “hot,” or time-sensitive, information. See National Basketball Association v. Motorola, Inc., 105 F.3d 841, 845 (2d Cir. 1997). It therefore would not protect the full range of databases, including those whose value lies in the historical, comprehensive nature of their contents.

Trade Secrets and Trademarks

A database producer harmed by the extraction of contents from his database may also seek a remedy under trade secret or trademark law. But these doctrines provide quite limited protection for most databases. Various necessary elements of a trade secrecy claim would be lacking in the case of the ordinary use of a typical database. See Report at 76-77. Trademark law is similarly limited, extending only to acts that threaten consumer confusion as to the source of origin of “brand name” databases. Id. at 77-78.


The enforcement of contractual terms is a useful means to prevent unauthorized uses of databases in many circumstances. The utility of contracts is limited, however, since they bind only the parties to the contract. Moreover, in the context of database protection, contract law raises unsettled issues of enforceability and national uniformity. See Report at 78-82.

Technological measures

Technological measures, such as encryption software, have the potential to be extremely effective as a practical matter in protecting databases. For a discussion of the current state of development and level of use of technological protection in the database industry, see Report at 26-28. Nevertheless, the availability of such self-help mechanisms has never been considered a substitute for the provision of meaningful legal rights. The ability to erect electrified fences, for example, has not led to the abolition of real property rules. Technology is not foolproof; it can be circumvented and may not prevent unauthorized use once someone gains access. Nor is it effective for the many databases still published in print form. Moreover, relegating database producers to strong technological measures may not be the ideal solution for consumers.

In sum, existing forms of protection provide some recourse to database producers, and have allowed them to continue to function in today's marketplace. The shortcomings of these alternatives, however, are significant, and leave producers vulnerable to many types of unauthorized uses or offer options that may be less than optimal from a consumer perspective.

Analysis of the Provisions of H.R. 2652

The Core Prohibition

Section 1201 of the bill defines the prohibited acts. It essentially adopts a misappropriation approach, prohibiting conduct that causes market harm to one who has invested in producing an information-based product or service. In this respect, it is related to unfair competition law, and draws on concepts underlying state law doctrines of misappropriation.

The choice of a misappropriation model avoids or ameliorates a number of problems that have been identified as caused by an exclusive property rights approach, as embodied in last year's H.R. 3531. By seeking to prohibit inappropriate and damaging behavior, rather than providing ownership rights in information products, this model has the following advantages: it allows many types of beneficial uses to proceed unchecked, without requiring reliance on a defense; it makes clearer that facts themselves cannot be the subject of private ownership; it may obviate the need for certain specific statutory terms that have proved difficult to craft; and it minimizes possible Constitutional concerns.

As drafted, the prohibition in section 1201 covers only acts that cause market harm to a commercial product or service. It therefore would not apply to most instances of ordinary use by a member of the public, or by a student, scholar or journalist. Nor would every commercial use necessarily result in market harm. This differs from copyright, where liability for infringement does not require a showing of harm.

Some have urged that any prohibition be limited to uses by a competitor. Doing so would eliminate potential grey areas and rule out any possibility of liability for nonprofit uses. Such a statute, however, would not fully prevent damage to the value of information products. Non-competitive uses can seriously harm the market for a database, most obviously where a database is produced for a particular nonprofit market, and members of that market utilize it without payment. Because of similar concerns, Congress has rejected the idea of absolute exemptions for nonprofit or educational uses under copyright law. See, e.g., H.R. Rep. No. 1476, 94th Cong., 2d Sess. 66-67 (1976). The Copyright Office therefore recommends a more tailored approach to avoiding inappropriate liability for nonprofit uses, as discussed below in connection with section 1202, “Permitted Acts.”

Another question is whether the harm that gives rise to liability should be limited to actual markets rather than including potential markets, as specified in section 1201. We believe that looking only to actual markets would be too restrictive; those who invest in creating information products should have some leeway to recoup their investment over time by exploiting those products in various markets. On the other hand, the relevant markets must have some limitation if the market harm element is to have meaning.

The choice of the term “potential” seems appropriate, carrying over a concept that has proved useful in the context of the copyright fair use doctrine, which requires courts to weigh as the most important factor in the fair use analysis “the effect of the use upon the potential market for or value of the copyrighted work.” 17 U.S.C. § 107(4); Harper & Row, Publishers, Inc. v. Nation Enterprises, 471 U.S. 539, 566 (1985). In determining which markets are closely enough related to be considered “potential,” courts could examine the producer's business plans as well as customary industry practices, as they have done under copyright law. See generally 4 Melville Nimmer and David Nimmer, NIMMER ON COPYRIGHT § 13.05[A][4] (1997) (”Nimmer”). The mere possibility that a use could be licensed should not be sufficient, or the term would become circular.

We also note that in order to be unlawful, the act of using information, in addition to the prerequisite of causation of market harm, must be made in commerce, further limiting the scope of section 1201 and exempting various private uses.

The language of section 1201 avoids terms that have come to have established meanings in other contexts, or have been criticized as ambiguous or overbroad. Thus, for example, the bill refers to “collections of information” rather than the defined copyright term, “compilations,” or last year's term, “databases.” It may be advisable to define some of the provision's terms, in order to provide greater clarity or address some of the concerns about subject matter coverage described in the Report. Where, however, the meaning of words is clear as a matter of plain English, definitions can sometimes raise more questions than they resolve.

In particular, we believe that the phrase “gathered, organized or maintained” covers those acts involved in collecting and presenting information that require major investment, and does so in straight-forward language.

The bill retains the concepts of “all or a substantial part” and the “investment of substantial . . . resources” that were the subject of discussion during the Copyright Office meetings earlier this year. It is difficult to see how these concepts could be omitted, without the undesirable result of protecting individual facts or trivial efforts. The term “substantial,” serving as a dividing line between the legally cognizable and not cognizable, probably cannot be more precisely defined in a statute. It is a term that courts in a common law system such as ours are well-equipped to flesh out in examining the facts of cases, and handle routinely in the context of other statutes. See generally 4 Nimmer § 13.03[A] (citing cases defining how much similarity is “substantial” in determining copyright infringement).

At the same time, the bill avoids some of the problems that have been identified with a pure unfair competition or misappropriation approach. It is not so narrowly written as to be limited to conduct between competitors or engaged in for profit, and it recognizes the possibility of transfers and licenses of protected interests. Nor is it limited to acts involving “hot” or time-sensitive information--a limitation which may be necessary to save state misappropriation doctrines from preemption by copyright law, see NBA v. Motorola, 105 F.3d 841 (2d Cir. 1997), but which eliminates protection for those who make major investments in producing valuable collections of historical or timeless information.

The bill's approach does have one clear disadvantage. The choice of a misappropriation model creates uncertainty as to whether such a law in the United States will trigger reciprocity for U.S. databases from the countries of the European Union. The EU’s reaction, of course, cannot be predicted at this time. The argument can certainly be made that the bill provides adequate protection against the threats to investments in databases that have been identified by the EU, and addresses the critical problems in a manner that is substantially equivalent in effect. The differences in approach, while meaningful, should be viewed as reasonable accommodations to differing domestic traditions and circumstances.

Permitted acts

Section 1202 provides a list of five types of permitted acts, some clarifying the scope of the prohibition in section 1201 and others exempting uses that would otherwise be covered. This list contains several items not included in prior proposals, which relate to concerns raised in Copyright Office and Subcommittee staff meetings earlier this year. Each of the five subsections constitutes a statutory recognition of a topic with important policy implications. In the view of the Copyright Office, such express treatment of each of these topics is highly advisable, although their scope of coverage and precise formulation may need further consideration.

The message of paragraph (a) is critical to any legislation dealing with database protection, as a matter of general policy as well as First Amendment concerns. The statement that an individual item of information is always free for the taking serves a policy function parallel to the idea/expression dichotomy in copyright law. 17 U.S.C. § 102(b). It provides statutory assurance that one is free to take any particular fact from a database, or any larger grouping of facts below the threshold of a “substantial part” of the whole collection. Thus, a student could obtain information and use it to prepare a class report; a scientist could find the answer to a question raised in the course of her research.

As to the qualification that such a taking is not a prohibited act “in itself,” the prohibition would have no meaning without it, since every “substantial part” of a collection of information is necessarily made up of many individual items. Moreover, a user would be able to evade the statute by engaging in ongoing, systematic takings of insubstantial parts that together add up to a substantial part.

Paragraph (b) is equally critical to any form of database protection. Its express inclusion in the bill makes clear that the producer of an information product cannot monopolize information simply by placing it within the product. Anyone is free to obtain the same facts elsewhere, without taking them from that particular collection. Thus, another party could create a competing database from the same or different sources, even if it were identical to the first. In the area of government data, the data could be obtained directly from the government, just not freely extracted from a value-added version that has required substantial investment to prepare. Members of the public would have the same access to the underlying government data that they would have had if the value-added version had not been created.

Paragraph (c) establishes a safe harbor for the use of databases for verification purposes. In doing so, it continues the treatment of verification under the old “sweat of the brow” line of copyright cases, which permitted as a form of fair use the good faith verification of facts gathered through independent effort. See, e.g., G.R. Leonard & Co. v. Stack, 386 F.2d 38 (7th Cir. 1967). This permitted act may be of particular importance to scientists and other researchers, ensuring that they can check the results of their research, despite the fact that doing so may entail the use of an entire database. At the same time, by limiting the verification to that done “within an[] entity or organization,” the bill avoids the potential harm that could be caused by permitting a competitor to circulate another's collection of information in the marketplace.

Paragraph (d) deals with nonprofit uses with a public interest component: education, science and research. As drafted, it simply restates in the affirmative that such uses are permitted as long as they do not cause market harm. In the view of the Copyright Office, the primary significance of this paragraph as written is to provide an express mention of this category of uses and thereby indicate legislative recognition of their value and importance. In analyzing the appropriateness of their treatment in the bill, it would be extremely helpful to hear from these communities as to specific examples of uses that could still be problematic. This would allow Congress to tailor the scope of the permitted use in such a way as to ensure that uses in the public interest will not be harmed.

The uses covered by paragraph (d) are the types of uses that are dealt with in the copyright law primarily through the fair use doctrine. 17 U.S.C. § 107. Although the bill is based on different goals and a different legal theory, and does not confer exclusive property rights, the policies behind fair use, which are intended to protect the public interest in access to information, are appropriate to consider here as well. The factors that courts are required to weigh in assessing a copyright fair use defense are not all mentioned in paragraph (d), but similar concepts are incorporated in various places throughout the bill.

The focus in the first copyright fair use factor on the “purpose and character of the use” encompasses paragraph (d)'s reference to “not-for-profit educational, scientific, or research purposes.” The second factor, “the nature of the copyrighted work,” cannot be a relevant criterion in the bill for determining permitted uses, since collections of information are by definition utilitarian rather than creative. The third and fourth factors--”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”--are not elements of an affirmative defense but are built into the section 1201 prohibition itself, through its prerequisites of the taking of a “substantial part” and the causation of harm to the “actual or potential market.”

The last “permitted act” is a broad safe harbor for news reporting in paragraph (e). Again, the Copyright Office believes it is advisable and appropriate to provide such statutory recognition to the value and importance of news reporting, especially given the First Amendment implications. This paragraph would cover the reporting to the public of financial information and sports statistics, about which great concern was expressed in the Copyright Office meetings. As drafted, however, the paragraph appears overbroad in its coverage, and could significantly undercut the purpose of the prohibition in section 1201. First, it should be made clear that the permitted activities must be legitimate news reporting, so that this paragraph cannot be used to excuse every individual who decides to disseminate substantial parts of databases over the Internet. Second, the coverage of news reporting should not be read to permit newspapers to reproduce a collection of information in such a way as to substitute for its purchase in the market. One possible solution might be to add a clause reading: “to the extent justified by the news reporting purpose.” Finally, the paragraph should be modified so as to exclude at least some directly competitive uses, where one news reporting entity competes in the market for news reporting by taking timely information gathered by another such entity. Otherwise, the bill would allow precisely the activity that the Supreme Court held unlawful in International News Service v. Associated Press, the case that established the misappropriation doctrine in American law. 248 U.S. 215 (1918).


The bill also contains two explicit exclusions in section 1203, barring application of its prohibition to two types of subject matter: government collections of information and computer programs. The Copyright Office supports both of these exclusions.

Much of the discussion in the meetings held by the Copyright Office centered on the need to leave government information in the public domain, free for all to use. We believe that a clear and absolute exclusion of such information from any database legislation is critical. We therefore support the approach of section 1203(a), which excludes a wider range of government information than does the Copyright Act (cf. 17 U.S.C. § 105), and represents a significant broadening from the corresponding provision in last year's bill. Section 1203(a) rules out protection for state and local governments as well as federal, and not only for collections of information prepared for a government entity by its employees, but also for those prepared by independent contractors and exclusive licensees.

As to the exclusion for computer programs in section 1203(b), such works differ fundamentally from collections of information in that the data they contain is put together not to be accessible to a researcher but to achieve a particular result. See 17 U.S.C. § 101 (definition of “computer program”). Moreover, computer programs are adequately protected by copyright law. Because confusion is possible, however, particularly because computer programs are often used to produce or operate databases, it is important to clarify the distinction in statutory language.

Additional exclusions could be designed to address other specific concerns as to the bill's coverage. See Report at 92-94.


H.R. 2652 contains only two definitions, in section 1204. In our view, both are necessary. The definition of “information” clarifies that a collection subject to the bill's prohibition can be a collection of copyrightable works as well as of facts.3 Databases of copyrighted works may become an important element of the Internet environment, providing sources of entertainment to the public; incentives should be ensured for their production as well as for the production of collections of pure fact. The definition of “commerce” provides the Constitutional basis for Congress to legislate under the Commerce Clause.

As noted above, additional definitions may be advisable to clarify the scope of the prohibition or the permitted acts, but should not be included unless they can shed more light rather than create new ambiguity.

Relation to other laws

The final substantive provision of the bill is section 1205, which sets out the relation of the prohibition to other laws. This too is an important section, since numerous bodies of existing law relate to certain aspects of the use of information. The bill adopts the approach of preempting equivalent state laws, but leaving untouched all other laws relating to information.

The Copyright Office supports this approach. We believe that federal preemption of state misappropriation laws is advisable, given their unclear outlines and variation from state to state. One of the chief benefits of a federal statute in this area is the ability to provide greater certainty and uniformity.4 It is in the interest of both producers and users of databases to be able to know what acts are permitted and what acts are not everywhere in the country. This is particularly important in an era where databases are often disseminated on-line, and can be simultaneously available in every state.

We also agree that other laws dealing with information should not be affected. Some of these laws protect producers of information; some protect users or the public interest generally; some offer certain benefits to all. It seems appropriate to leave in place the status quo of other bodies of law, which involve their own distinct policies and balances.

The bill singles out two areas for explicit savings clauses: contractual freedom and the Telecommunications Act of 1996. Such separate treatment may be called for in order to address the particular controversies that have arisen in these areas. Paragraph (c) provides that nothing in the bill “shall restrict the rights of parties freely to enter into licenses or any other contracts with respect to the use of information.” It therefore leaves for resolution elsewhere the ongoing debate over the extent of the ability to vary legal rights by contract. See Report at 80-81.5 Paragraph (d) appears to respond adequately to concerns expressed in the Copyright Office meetings that any legislation not overturn or otherwise interfere with the compromises struck in the recently enacted Telecommunications Act.

Other Issues


H.R. 2652 is a short bill, providing the basic elements of a misappropriation regime for collections of information. Further consultation and discussion may indicate the desirability of clarifying or expanding some of its provisions, or adding provisions to deal with other issues. For example, additional exclusions in section 1203 could resolve problems with the coverage of particular material that is not intended to be, or should not be, subject to the prohibition of section 1201.

We would also like to comment on three issues raised in the Copyright Office Report that are not expressly addressed in any particular section of the bill: duration, sole source databases, and Constitutional concerns.


The bill does not provide for a set term of protection. This is analytically consistent with the choice of a misappropriation approach. Intellectual property rights are granted for a fixed number of years, as required by the Constitution. U.S. CONST., Art. I, sec. 8, cl. 8. Unfair competition and misappropriation laws, in contrast, since they do not confer ownership rights in property but rather prohibit wrongful conduct, do not contain specified durations. Cf. 17 U.S.C. § 1101 (anti-bootlegging provision added to Title 17 in 1995, which contains no set term of protection). Often, however, they incorporate concepts of ongoing investment, typically as part of the circumstances that make the prohibited conduct wrongful. Thus, trademarks are protected as long as their owners continue to use them in connection with goods or services; failure to use a mark for too long will lead to a finding of abandonment. See J. Thomas McCarthy, TRADEMARKS AND UNFAIR COMPETITION § 23 (1996). The doctrine of misappropriation, at least in some states in its non-preempted form, looks to the timeliness of the information at issue as a measurement of its value and need for protection. See NBA v. Motorola, Inc., 105 F.3d at 845.

The Copyright Office believes that the duration issue is one that requires further study. The aim is to determine an appropriate measure for how long the prohibition should last, consistent with its theoretical grounding in misappropriation rather than property. One should not be able to produce a collection of information, never do anything else with it, and forever prevent others from using it freely. On the other hand, to be as effective as possible in its goals of stimulating production and dissemination, the statute should provide incentives for ongoing investments in maintaining, refreshing and distributing such collections.

“Sole source” databases

The issue of “sole source” databases, where as a practical matter the information is not available elsewhere, is not separately addressed in the bill. See Report at 102-107. Several aspects of the bill do, however, relate to the treatment of sole source databases. The broad exclusion of government collections of information will rule out protection for some--those created by an exclusive licensee of the government. Some of the permitted acts set out in section 1202 are also relevant. Thus, the safe harbor for news reporting will allow important information contained in sole source databases to be communicated to the public. Otherwise, the bill makes the choice to leave the issue to be dealt with through existing mechanisms of antitrust law and general federal regulation and oversight of particular industries.

Constitutional constraints

Finally, the potential Constitutional constraints in this area, described in the Report at 107-111, should be carefully analyzed. We note, however, that the prohibition the bill would create is based on a different model than copyright law. It is intended to stimulate investment rather than creation, and does so by prohibiting commercially damaging conduct rather than creating exclusive property rights. It therefore appears closer to federal unfair competition legislation in its Constitutional underpinnings, and may similarly pass muster under the Commerce Clause.


The Copyright Office believes that H.R. 2652 is a constructive step toward the goal of promoting the availability of more information to the public. We look forward to working with the Subcommittee and all interested parties to achieve a balanced and beneficial result.

1 See, e.g., J.H. Reichman & Pamela Samuelson, “Intellectual Property Rights in Data?,” 50 Vand. L. Rev. 51 (1997); Laura D'Andrea Tyson and Edward F. Sherry, “Statutory Protection for Databases: Economic and Public Policy Issues” (1997).

2 A recent report prepared by Laura D'Andrea Tyson and Edward F. Sherry, based on research funded by Reed-Elsevier, Inc. and The Thomson Corporation, presents statistics indicating that while the database industry has continued to grow in the six years since Feist (measured by the number of both databases and database producers), its rate of growth has slowed. Laura D'Andrea Tyson and Edward F. Sherry, “Statutory Protection for Databases: Economic and Public Policy Issues” at 10 and Table 2 (1997).

3 The existence and scope of copyright in the materials compiled should not be either enlarged or curtailed by the prohibition in section 1201. Cf. 17 U.S.C. § 103(b) (providing similar rule with respect to preexisting works contained in copyrightable compilations).

4 In his dissent in INS v. AP, 248 U.S. 215 (1918), the case that established the doctrine of misappropriation, Justice Brandeis argued that liability for such conduct should for this reason be established by legislation rather than common law.

5 Some participants in the Copyright Office meetings based their positions on the need for legislation in whole or in part on the ability of producers to impose contractual limitations on the use of their databases.