or prohibits implementation of the law already on the books, which is itself in concert with OMB Circular A-130. Both the statute and the Circular require agencies to: (1) ensure that the public has timely and equitable access to their public information; (2) regularly solicit and consider public input on their information dissemination activities; (3) provide adequate notice when initiating, substantially modifying or terminating significant information products; and (4) prohibit exclusive, restricted or other types of distribution arrangements that interfere with the availability of public information, including restricting use or resale or charging fees or royalties for such information.

Unfortunately, despite the clarity of this law, it has not been adequately enforced. Agency after agency-including some within the Department of Commerce-has disregarded nearly every one of these tenets.

A database protection statute is not needed to address the concerns regarding inadequate enforcement of the law already on the books. Rather, the government itself can solve this problem through responsible and lawful agency activities and through enforcement ofcurrent law by OMB. Critics of the government information provisions in H.R. 354 are aiming at the wrong target. Their concerns would be more easily allayed if they would join industry efforts in working to strengthen and enforce existing laws and regulations.

A third issue raised in the letter regards the inadequacy of "fair use" under your bill, Mr. Chairman. I have already given the industry perspective on this issue, and will not repeat here the many provisions of H.R. 354 that more than adequately address the needs of nonprofit educational, scientific and research communities, as well as those secondary database publishers who have in the past been vocally opposed to database protection.

Another concern raised regards the harm that misuse of collections of information may cause to "potential" markets. SILA would note that this term has long been present in our Nation's intellectual property laws. Indeed, under the 1976 Copyright Act, the effect of a use upon the potential market is mentioned among the four elements that courts must consider in determining whether fair use has been made of a work. Yet unlike copyright law-where fair use is an affirmative defenseunder H.R. 354 the burden is on database owners to prove harm to their actual or potential market.

Regardless of the wisdom in adopting a long-standing and tested standard from another set of intellectual property law, there is a practical consideration here, Mr. Chairman. Businesses do not today, and will not tomorrow, make major investments in products and services with the belief that they can exploit only the current, actual market in which they desire to offer such products and services. Database producers should not be precluded from meeting new demands and exploiting new uses for their products and services during the limited 15-year period during which this legislation provides them limited means to combat unfair use of their work.

One final concern that continues to be raised is whether a law similar to H.R. 354 would be constitutional. There can never be complete certainty regarding the constitutionality of any law Congress enacts, for that is solely and ultimately up to the Supreme Court to determine. However, in crafting the Collections of Information Antipiracy Act to create limited, commerce-based protection, Congress is relying on its Commerce Clause powers under Article I, Section 8, Clause 3 of the Constitution. In that regard, it will follow the successful and well-founded model that has been used to grant protection to trademarks

SIIA COMMENTS ON NEW PROVISIONS OF H.R. 354

SIIA notes that H.R. 354 contains some new language above and beyond the bill that was passed twice last year by the House. It is incumbent upon the Association to offer its preliminary comments on these provisions. Language was added in Subsection 1408(c) to clarify that the limited protection under this bill will last for only 15 years. SILA supports this clarification, while noting that under traditional doctrines of missappropriation and unfair competition-where protection lasts for as long as the product or service has value in the market-no term limitation is necessary. There are many databases whose value extends far beyond a 15-year period. However, clarifying a limited term will hopefully allay some concerns that this proposed statute creates "perpetual protection" for databases and will have the additional benefit of establishing under U.S. law a limited term of protection comparable to that found in the EU Directive.

Another new provision is found in Subsection 1403(a)(2), “Additional Reasonable Uses." Regrettably, Mr. Chairman, SIIA cannot at this point endorse this change

317 U.S.C. 107

from last year's legislation. While the provision appears at first glance somewhat similar to language found in 17 U.S.C. 107, it opens a door to potential mischief by being much broader than the copyright fair use exception-whether in regard to profit-seeking or nonprofit uses of databases. Therefore, before commenting further on this apparent expansion of the bill's already generous provisions exempting nonprofit activities in relation to databases, SILA wants to carefully listen to and analyze the reasons why proponents of this language feel it is needed. The Association believes firmly, however, that those who wish to make use of a database owner's product or service should not be free-under the pretext of having "transformed" or "added value” to a substantial part of the original database—to then harm the market for the first producer's collection of information.

CONCLUSION

In closing, Mr. Chairman and Members of the Subcommittee, SIIA wishes again to express its appreciation for the Subcommittee's interest in considering legislation to protect America's databases. H.R. 354 remains in whole a balanced and fair bill, and a new statute along these lines will provide the necessary incentives to database owners to continue investing time, money and personnel necessary to create and maintain databases. With such incentives, valuable and reliable collections of information will be more widely available in a large number of formats. A new law will also assure that those who use databases fairly can obtain quicker and wider access to these products and services. Finally, it will prevent the erosion of a traditionally strong and vibrant U.S. economic sector and those other sectors that benefit from its existence.

The Association looks forward to working with you to assure that we finally see a fair and balanced database protection law enacted.

Thank you, and I will be glad to answer any questions.

PREPARED STATEMENT OF THE COMPUTER & COMMUNICATIONS INDUSTRY ASSOCIATION, THE INFORMATION TECHNOLOGY ASSOCIATION OF AMERICA AND THE ONLINE BANKING ASSOCIATION

INTRODUCTION

This testimony is being presented by trade associations representing a large number of companies involved in the American database industry. Because databases (1) are items of commerce in their own right, (2) are critical tools for facilitating electronic commerce, research, and educational endeavors, and (3) already enjoy substantial legal (copyright and contract) and technological protections, we support federal legislation carefully tailored to provide database proprietors with focused protection against piracy. Conversely, we have consistently opposed legislation that would grant the compiler of any information an unprecedented right to control transformative, value-added, downstream uses of the collection of information or any useful fraction of that collection.

The Computer & Communications Industry Association, the Information Technology Association of America, and the Online Banking Association are the proponents of this specific testimony. However, we have attached to this testimony a copy of a "position statement" to which a number of additional companies, trade associations, institutions, and representatives of the scientific, research, library, and educational communities have recently subscribed. The parties subscribing to the position statement include entities, companies, and trade associations that represent companies that create, sell access to, and use a broad variety of databases. The activities of these companies and entities are vital to American and global commerce. Serious concerns regarding the potential harm to the American database industry and to the economy more generally threatened by the bill under consideration in the House Judiciary Committee, H.R. 354, have caused these companies join together to participate collectively and constructively in the legislative and public policy process that we hope can lead to a balanced and equitable solution to the issue of database protection. We greatly appreciate the opportunity to submit our views through this written testimony.

SUMMARY OF POSITION

As a general matter, we believe that any federal database protection legislation ought to be focused only on protecting database businesses against harmful parasitic conduct of competitors (and, of course, against malicious acts of vandals). Such protection should not-and under the Constitution, may not-extend to facts and other public domain material, as such, whether or not such data are contained in a database. Compilers of information should not be given a statutory right to control

transformative, value-added, downstream uses of information. Unless a taking of information is of a kind that (a) infringes upon existing rights under contracts or Copyright law or (b) genuinely threatens the economic viability of a database business, it should not be illegal.

In order to continue the robust growth and development of the "Information Age", information must continue to be a readily accessible commodity. We are concerned that H.R. 354 would most likely empower those in control of information to impose new private taxes on access to information, which surely would not enhance access to information. Database proprietors should not have a statutory right to prohibit (or charge a royalty for) uses of information contained in databases that do not harmfully compete with or displace the original compiler's actual business. The goal of laws in this area should not be to lock up factual data and information but rather to promote products and services that make such data intelligible and useful-socalled "transformative" and "value-added" services.

Evaluated against this template, H.R. 354 goes far beyond its stated goals and broadly prohibits access to facts and public domain material. As the Federal Trade Commission, the Commerce Department, and the Justice Department recognized with respect to largely identical legislation in the 105th Congress, H.R. 354 presents significant competitive and Constitutional concerns that could and should be avoided by approaching the issue of database protection in a more targeted manner.

WE SUPPORT DATABASE PROTECTION LEGISLATION THAT FOLLOWS THE JUDICIALLY CREATED MISAPPROPRIATION DOCTRINE

The associations are willing to support a carefully tailored federal codification of the judicially created and constitutionally consistent misappropriation doctrine. A recent decision of the United States Court of Appeals for the Second Circuit, National Basketball Association v. Motorola, Inc., 105 F.3d 841 (2d Cir. 1997), provides a coherent restatement of the essential principles of the traditional tort of misappropriation and a useful starting point for potential federal legislation directed against misappropriation. In NBA, the Second Circuit considered the National Basketball Association's claim that a service providing sports scores to fans through paging devices misappropriated the NBA's rights in its basketball games. The defendant got the scores from reporters who watched or listened to broadcasts of the games on television or radio and keyed information about the games into personal computers. No information proprietary to the NBA was taken and the NBA was not in the pager sports score business (although it said that it might one day want to be in that business). The court defined misappropriation under New York law (derived initially from the Supreme Court decision in INS v. AP) as incorporating the following elements:

1. The plaintiff generated or gathered information at a cost;

2. The information is time-sensitive;

3. The defendant, by using the plaintiffs information, is free-riding on the plaintiffs efforts;

4. The defendant is a direct competitor of a product or service offered by the plaintiff; and

5. Free-riding by the defendant and others on the plaintiffs efforts would so reduce the incentive to produce the product or service in question that its existence or quality would be substantially threatened.

Applying these criteria, the Second Circuit found that nothing that the defendant did inflicted any harm on the NBA. Thus, there was no basis for a claim of misappropriation. The court noted that the result might well have been different if the defendant had been taking its data from a competing sports score pager service and thereby free-riding on the efforts of the first service.

To some extent, the limitations set out in NBA arise from the need to harmonize misappropriation claims under state law with the preemption of state law mandated by Section 301 of the 1976 Copyright Act. To that extent, the requirements could be adjusted by enactment of a federal misappropriation law. However, such adjustments would have to conform to the constitutional holding in Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991), which generally precludes legislation taking facts out of the public domain but expressly condones the traditional misappropriation doctrine.

H.R. 354 HAS A NUMBER OF PROBLEMS THAT PROBABLY CANNOT BE CURED BY

AMENDMENTS

Inherent in the structure and approach embodied in H.R. 354 are certain problems and unintended consequences that we believe cannot be effectively cured without shifting to a new template. Conceptually, those problems are the product of an approach that starts with a broad and monolithic prohibition on access to and use of information and then tries to carve out piecemeal exceptions to that prohibition. We believe that many of these problems could be avoided if the issue were approached by a law prohibiting only specific harmful practices against the background of a general presumption of unfettered access to information. Since the only articulated need for legislation is to prevent parasitic undermining of incentives to make new investments in databases, a database protection law ought to be focused on the prevention of that particular harm.

We believe that H.R. 354 does not adequately address some key issues of concern to participants in the commercial database industry, including the following:

• It applies to small amounts of information. H.R. 354 prohibits the extraction, or use in commerce, of "a substantial part, measured either quantitatively or qualitatively, of a collection of information. . . ." By allowing the database publisher to prevent reuses of “qualitatively” substantial parts of a database, H.R. 354 effectively prevents the reuse of any information. The second-generation publisher has no way of knowing which bits of information the first generation publisher considers qualitatively substantial. Moreover, if the first generation publisher chose to litigate over even a very minor reuse of information, that action could never be resolved on summary judgment because the qualitative substantiality of the taking would always be a question of fact. • It can prohibit legitimate reuse of information. H.R. 354 prohibits the reuse of information "so as to harm the actual or potential market" of the publisher. Any harm, even one lost sale, would suffice to establish liability. Moreover, "potential market" is defined to mean any market the person claiming protection "has current and demonstrable plans to exploit or that is commonly exploited by persons offering similar products or services. ." Since it is common to reuse information in new products, almost any reuse of information could arguably meet this test. As a result, publicly useful products and services could be kept off the market for years by an inattentive or monopolistic database proprietor.

• The new reasonable use exception is too narrow. H.R. 354 includes a new provision for "reasonable uses" which did not appear in H.R. 2652. This provision is certainly a step in the right direction, particularly because it considers "the extent to which and the manner in which the portion used or extracted is incorporated into an independent work or collection, and the degree of difference between the collection from which the use of extraction is made and the independent work or collection."

However, the new reasonable use exception is only available for extractions "done for the purpose of illustration, explanation, example, comment, criticism, teaching, research, or analysis. Presumably an extraction for inclusion in a commercial database, even one used by others for teaching, research or analysis, would not qualify.

Nor is the exception available for a second-generation product which "is likely to serve as a market substitute for all or part" of the first generation product. In other words, even if the second database is completely different from the first, in that it is far more comprehensive and far better organized, the publisher of the first database can pursue damages and injunctive relief because the second publisher extracted an allegedly qualitatively substantial part of the first database.

Not only is the "reasonable use" exception of H.R. 354 inapplicable to commercial value-added uses of databases, but this exception and the preexisting exception for nonprofit scientific and research use, carried over from H.R. 2652, provide no real comfort to the scientific researcher. In most cases, a scientist who reuses existing data to create a new database would not qualify for the preexisting research exception because the new database would "harm directly the actual market" for the first database. Indeed, the more revolutionary the new database, the more likely it is to harm the market for the first database. The new reasonable use exception also would not apply because the first database, from which the researcher extracted the information used in his or her new database, would probably qualify as one "primarily de

veloped for and marketed to persons engaged in the same field or business as the person making the use," thus falling outside the scope of the exception. • H.R. 354 does not adequately respond to the significant threat posed by sole source databases. To be sure, H.R. 354 permits the independent collection of information. In many instances, however, there is no feasible way for another person to collect the information independently. For historical information, for example, one cannot go back in time to gather the information; one must rely on existing databases. In other instances, the first generation publisher is also the original source of the information. The publication of the information is incidental to its main activity. In still other instances, the publisher may build up its database incrementally over many years, incurring modest costs along the way that it has already recovered.

The only other relief H.R. 354 provides with respect to sole source providers is to leave intact the antitrust laws. Because proving that a publisher has monopoly power in a relevant market is extremely difficult, and because antitrust enforcement and litigation have not generally been effective tools in restraining monopolists, however, the antitrust laws are unlikely to rein in sole source suppliers.

• H.R. 354 applies retroactively. Because H.R. 354 applies retroactively, any database created within the past fifteen years receives its protection. Databases already in existence, however, do not require an incentive to ensure their creation. The retroactive aspect of H.R. 35J provides publishers of existing databases with an unwarranted and unnecessary windfall profit.

• H.R. 354 as a practical matter grants perpetual protection. H.R. 354 attempts to correct a problem identified last year: that additional investment in the "maintenance" of a database could lead to a new term of protection for the database, thus resulting in perpetual protection. The new language, however, does not cure a more fundamental problem noted last year: with dynamic online databases, there will be no accessible hard copy record of the database as it existed on a certain date. Accordingly, the second publisher will have no way of knowing which portions of the database are more than fifteen years old, and thus no longer subject to protection. Theoretically possible private solutions to this problem, such as "tagging" each individual data element with a date code, would be prohibitively costly. The public and private costs of a registration or deposit system are likewise large and unjustifiable. Fifteen years is too long. Traditional misappropriation law recognizes that the value of information is linked to its recency or freshness. While a term of protection longer than that encompassed by the traditional "hot news" cases ought to be constitutionally permissible, a fifteen-year term is way too long for facts, public domain information, and most other kinds of information. The very concept of a fixed term is probably unworkable, because the period of protection necessary to protect a database proprietor's incentives to invest probably differs widely from database to database. Moreover, the inclusion of a fixed term-along with a fair ("reasonable") use exception-evidence that the proposed legislation would effectively give a form of copyright protection to databases. The Supreme Court's Feist decision holds that such protection would be unconstitutional.

Having given these problems careful thought, we believe that this bill as drafted would lock up the contents of databases rather than provide targeted protection to those engaged in database businesses against truly harmful parasitic or malicious behavior. Thus, despite considerable thought and effort, we have been unable to formulate amendments to H.R. 354 that both work within its framework and effectively and economically cure these problems.

CONCLUSION

We respectfully urge the Subcommittee to take full account of the fact that most database proprietors are also users of information contained in databases compiled by others. Many respectable and responsible companies engaged in compiling, selling access to, and using databases do not support H.R. 354 as currently drafted because they fear disruption of a healthy status quo in the database business.1 These

1 Between 1991 and 1997, according to statistics offered by proponents of H.R. 354, the number of databases increased 35 percent, the number of files contained within these databases increased 180 percent, and the number of online searches increased 80 percent. During that same period, major publishing companies spent billions of dollars to acquire US database businesses. These data attest to the healthy state of the US database industry under current law.