Databases have always been commodities of both commercial value and social utility, ranging from their early incarnation in the eighteenth century as directories compiled by walking door to door to the late twentieth-century compendiums of millions of items in electronic form. The question of whether and how databases should be protected by the law has never been easy, as it necessarily involves finding a balance between two potentially conflicting societal goals: the goal of providing adequate incentives for their continued production, and the goal of ensuring public access to the information they contain. At different points in time, and in different societies, that balance has been struck in different ways.
In the past few years, the issue has taken on new urgency due to changes in the legal, technological and international landscape. The major landmarks among these changes have been the U.S. Supreme Court’s 1991 decision in Feist Publications v. Rural Telephone Service Co.; rapid developments in the technologies for collecting, organizing, reproducing and disseminating information; and the actions of the European Union in harmonizing the laws of its member states. As a result, 1996 saw the consideration of proposals for a new form of protection for databases, both in the World Intellectual Property Organization and in the U.S. Congress. The discussions sparked a heated debate in the United States, involving a broad spectrum of interests.
In early 1997, the Copyright Office initiated its own examination of database protection by scheduling meetings with several groups that had expressed opposition to the 1996 proposals, in order to gain a clearer understanding of their concerns. In April, Senator Orrin G. Hatch, Chairman of the Senate Committee on the Judiciary, requested that the Office broaden its meetings to include a wide range of interested parties, and report to the Judiciary Committee on the outcome in order to assist in the Committee’s consideration of the subject. Since that time, the House Subcommittee on Courts and Intellectual Property, chaired by Rep. Howard Coble, has also asked to see the report in connection with its own consideration.
The purpose of this report is to provide some background and context to the subject of database protection, to identify and clarify the issues involved, and generally to lay the initial groundwork for a Congressional determination of appropriate legislative policy. The report is divided into seven substantive sections: (1) an historical overview of copyright protection for databases in the United States; (2) a description of database industry practices in securing protection against unauthorized use; (3) a description of Copyright Office registration practices relating to databases; (4) a summary of the relevant international context, focusing on the European database directive and the draft WIPO treaty on the protection of databases; (5) a summary of prior Congressional consideration of the subject; (6) a description of the methodology and substance of the meetings held by the Copyright Office; and (7) a discussion of the issues presented.
The report does not make recommendations on either the advisability or the form of any database protection legislation. At this point, we seek only to present the issues to be addressed, and to offer some options for addressing specific concerns. Their resolution will await public hearings and the presentation of evidence.
I. COPYRIGHT PROTECTION FOR DATABASES IN THE UNITED STATES
A. Before Feist
In the terminology of copyright law, a database is a “compilation”: “a work formed by the collection and assembling of preexisting materials or of data . . . .(1) Compilations constitute one of the oldest forms of authorship protected under U.S. law, dating back to the eighteenth century.(2) Compilations were protected as “books” under the first federal copyright statute.(3)
Over the course of the nineteenth century two rationales developed for protecting compilations under copyright. One rationale, which has come to be known as the “sweat of the brow” doctrine, focused on the effort and investment of the compiler. The other focused on the compiler’s judgment and creativity in the selection and arrangement of the materials comprising the compilation.
The earliest compilation cases that discussed the basis for copyright protection identified the compiler’s effort — “his own expense, or skill, or labor, or money(4) — as the critical contribution justifying protection.(5) These cases, involving works ranging from law reports and legal encyclopedias to compilations of war records, emphasized both the compilers’ effort and the copiers’ “unfair use of the copyrighted work, in order to save themselves the time and labor of original investigation.(6) Contemporary treatises echoed this approach.(7)
During the late nineteenth century courts began to articulate a basis for copyright protection generally that differed from the labor/investment approach taken in cases involving compilations. In a series of decisions from 1879 to 1903, the Supreme Court held that the “writings” that could be protected under the copyright clause of the Constitution included “only such as are original,(8) and indicated that creativity is a component of originality.(9) Under this approach, copyright was described as protecting writings that are “the fruits of intellectual labor,(10) “productions of intellect or genius(11) or “original intellectual conceptions of the author.(12)
The evolving doctrine of originality was applied by some courts in compilation cases, particularly cases involving compilations of textual materials such as law books. These cases identified the author’s critical contribution justifying protection as his judgment in selecting and arranging materials.(13) This approach coexisted with, rather than supplanted, sweat of the brow cases. Sweat of the brow was applied to cases involving purely factual compilations, such as catalogs and directories. Sometimes the two approaches appeared to be melded together in a single case, with the court focusing on the “labor” and “skill” contributed by the author.(14) With very few exceptions, one or the other approach was drawn upon by the court to support the conclusion that a particular compilation was protectible, rather than to deny protection.
On the question of the scope of protection afforded to compilations, there was somewhat greater uniformity in the case law. In compilation cases, regardless of the theoretical framework adopted to justify copyright protection, once the plaintiff’s work was determined to be copyrightable, courts generally held a defendant to have infringed whenever material was copied from the plaintiff’s work. Typically there was no inquiry as to whether the particular material copied was protected by the plaintiff’s copyright. To avoid infringement, a second-comer was required to go to the original sources and compile the material independently, without reference to the earlier work.(15) A common thread running through many of these decisions was the court’s desire to prevent the copier from competing unfairly with the compiler by appropriating the fruits of the compiler’s efforts or creativity. In this sense, courts treated copyright protection for compilations much like a branch of unfair competition law.
The Copyright Act of 1976 included a definition of “compilation” which, for the first time, drew an express statutory connection between compilations and “original works of authorship”:
A “compilation” is a work formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship. The term “compilation” includes collective works.(16)
The definition compels a court to examine the nature of a compilation’s “selection, coordination, or arrangement” in order to determine whether the compilation is “an original work of authorship” protectible under section 102(a). In other words, the same originality requirement applies to compilations as to all other works.
A separate section clarified the scope of protection for compilations, specifying that
The copyright in a compilation or derivative work extends only to the material contributed by the author of such work, as distinguished from the preexisting material employed in the work, and does not imply any exclusive right in the preexisting material. The copyright in such work is independent of, and does not affect or enlarge the scope, duration, ownership, or subsistence of, any copyright protection in the preexisting material.(17)
The 1976 Act also codified the idea/expression dichotomy that had been developed by the courts.(18) Section 102(b) provides: “In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.” This language has been interpreted to exclude protection for facts as well.(19)
Cases under the 1976 Act were divided about the continuing viability of the sweat of the brow doctrine. Some circuits continued to apply sweat of the brow.(20) Other circuits rejected sweat of the brow, requiring instead that compilations contain sufficient creativity in their “selection, coordination or arrangement” to render them “original works of authorship” entitled to copyright protection.(21) On both sides of this doctrinal divide, however, there was a consistent line of cases upholding the copyrightability of directories.(22) The stage was thus set for Supreme Court consideration of the issue when it granted certiorari in a Tenth Circuit case routinely applying the sweat of the brow doctrine to protect a white pages telephone directory against wholesale copying.(23)
B. The Feist Decision
The Supreme Court sounded the death knell for the sweat of the brow doctrine in Feist Publications v. Rural Telephone Service Co.(24) In finding a white pages telephone directory to be uncopyrightable, the Court held that the sole basis for protection under U.S. copyright law is creative originality.
The plaintiff, Rural Telephone Service Co. (Rural), was a local telephone company that produced a white-pages telephone directory covering its service area. Feist Publications (Feist), the defendant, published a directory covering multiple service areas. After Feist sought, and was refused, a license to the listings in Rural’s directory, it copied the listings without authorization. The district court found Feist liable for infringement, and the Tenth Circuit affirmed in an unpublished memorandum decision. The Supreme Court granted certiorari, presumably to resolve the split in the circuits.(25)
The Court reviewed the history of compilation copyright and the development of the sweat of the brow doctrine. It repudiated the doctrine in unequivocal terms:
Originality, the Court held, has two distinct components: “independent creation plus a modicum of creativity.(26) The Court emphasized that the creativity component is extremely modest. “To be sure, the requisite level of creativity is extremely low; even a slight amount will suffice. The vast majority of works make the grade quite easily, as they possess some creative spark, ?no matter how crude, humble or obvious’ it might be.(27)
The Court reviewed the definition of “compilation” in the Copyright Act of 1976 and discerned an intent to overrule the sweat of the brow doctrine by legislation. By defining a compilation as “a work formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship,(28) the Court explained, Congress specifically required originality in order to protect compilations, and described the elements of authorship that are protected in a compilation: the selection, coordination and arrangement of the underlying material.(29)
The Court did not limit its holding to statutory interpretation, however. It held that “[o]riginality is a constitutional requirement.(30) Citing nineteenth-century case law, the Court derived this requirement from the Constitutional terms “Writings” and “Authors” in the grant of authority to Congress to enact copyright laws.(31)
On the facts before it, the Court held that Rural’s white pages telephone directory was uncopyrightable.
The selection, coordination, and arrangement of Rural’s white pages do not satisfy the minimum constitutional standards for copyright protection. . . . Rural’s white pages are entirely typical. . . . In preparing its white pages, Rural simply takes the data provided by its subscribers and lists it alphabetically by surname. The end product is a garden-variety white pages directory, devoid of even the slightest trace of creativity.(32)
Rural’s selection of listings was “obvious,” and its arrangement was “not only unoriginal, it [was] practically inevitable.(33) The Court acknowledged that the telephone white pages were an extreme case, falling in “a narrow category of works in which the creative spark is utterly lacking or so trivial as to be virtually nonexistent.(34) By contrast, it stated, “the vast majority of compilations will pass” the originality test.(35)
Although the holding of Feist relates to copyrightability, the Court acknowledged the impact that its reasoning would have on the scope of copyright protection for compilations. “[C]opyright in a factual compilation is thin. Notwithstanding a valid copyright, a subsequent compiler remains free to use the facts contained in another’s publication to aid in preparing a competing work, so long as the competing work does not feature the same selection and arrangement.(36) This represented a complete reversal of the earlier judicial approach in several circuits that held any substantial taking from a copyrightable compilation to be an infringement, and required second-comers independently to collect material for a competing compilation.
C. Subsequent Judicial Interpretation of Feist
Feist’s teachings have proved important for lower courts both in determining copyrightability and in assessing scope of protection.
1. Copyrightability
Subsequent cases have confirmed that the category of works lacking the requisite level of creativity is small. A series of three Second Circuit decisions rendered shortly after Feist is illustrative.
In Key Publications, Inc. v. Chinatown Today Publishing Enterprises Inc.,(37) the Second Circuit sustained the copyrightability of the yellow pages of a telephone directory for New York’s Chinese-American community. The court found that the selection of entries in Key’s directory was original.(38) In addition, the arrangement of the directory into categories (e.g., Accountants, Bridal Shops, Shoe Stores, Bean Curd & Bean Sprout Shops) was, when “viewed in the aggregate,” original, because it “entailed the de minimis thought needed to withstand the originality requirement.(39)
In Kregos v. Associated Press,(40) the court found the plaintiff’s “pitching form” — a form comprised of nine statistics about a pitcher’s performance — copyrightable. Kregos’ selection of those nine statistics from the universe of statistics that can be used to describe a pitcher’s performance could be original, according to the court. Reversing the district court’s grant of summary judgment to the defendant, the Second Circuit held that “[i]t cannot be said [as a matter of law that] Kregos has failed to display enough selectivity to satisfy the requirements of originality.(41)
By contrast, in Victor Lalli Enterprises, Inc. v. Big Red Apple, Inc.,(42) the Second Circuit found insufficient creativity to support a copyright. The compilation at issue in Lalli was comprised of “lucky numbers” used in gambling, arranged in a grid with months along the vertical axis and days of the month along the horizontal axis. The numbers were computed according to a formula that was standard in that industry. The court found no originality in either the selection or arrangement of the data: “Lalli exercises neither selectivity in what he reports nor creativity in how he reports it.(43) The compilation was therefore held uncopyrightable.(44)
Among works that are particularly vulnerable to a finding of uncopyrightability are comprehensive factual databases covering an entire universe of information, where the element of “selection” is lacking and the “arrangement” is obvious.(45) The very comprehensiveness and ease of use of such a database may account both for its commercial value and its lack of protection under copyright.
2. Scope of Protection
The Feist statement that “the copyright in a factual compilation is thin” has been borne out in case law subsequent to the Feist decision. In both Key Publications and Kregos, the Second Circuit’s holding that the work was sufficiently original to be copyrightable was followed by a finding of noninfringement.(46) Although the court had stated in Key Publications that, while compilation copyright is thin, “we do not believe it is anorexic,(47) the scope of protection adopted in Kregos was quite narrow. There, the Second Circuit held that the defendant’s compilation would not infringe if it “differs in more than a trivial degree” from the plaintiff’s work, essentially creating a “virtual identity” standard for infringement.(48) The Ninth Circuit had also applied a virtual identity standard in compilation cases predating the Feist decision.(49) Without necessarily articulating a virtual identity standard, a number of district courts have adopted a similarly narrow scope of protection.(50)
Other courts have accorded even thinner protection to compilations in which copyright was conceded, failing to discern originality in any particular acts of selection or arrangement. In a pair of post-Feist cases, the Eleventh Circuit found that copying of significant portions of copyrightable compilations was not infringing because the material copied did not rise to the level of creative authorship.
In Bellsouth Advertising & Publishing Corp. v. Donnelley Information Publishing, Inc. (“BAPCO”),(51) the Eleventh Circuit held that the defendant’s entry into a computer of all of the names, addresses and telephone numbers of advertisers in the plaintiff’s yellow pages telephone directory, together with business type and type of advertisement, did not infringe.
Since the parties had stipulated to the copyrightability of the plaintiff’s directory, and agreed that “the only elements of a work entitled to compilation copyright protection are the selection, arrangement or coordination as they appear in the work as a whole,(52) the court focused on the elements of selection, coordination and arrangement that the plaintiff claimed were infringed, and found each to be either unprotectible or not copied. For example, the plaintiff claimed (and the district court held) that it selected the listings by determining the geographic scope of the directory, establishing a closing date for changes, and limiting listings to subscribers to its business telephone service, as well as through a variety of marketing techniques. The court found that these elements did not meet the level of creativity required by Feist.(53) Moreover, the court did not consider these elements to be “acts of authorship, but techniques for the discovery of facts . . . . The protection of copyright must inhere in a creatively original selection of facts to be reported and not in the creative means used to discover those facts.(54)
The court also found the arrangement of the directory “in an alphabetized list of business types, with individual businesses listed in alphabetical order under the applicable headings” to be unoriginal, and to have merged with the idea of a business directory.(55) As to the headings used in the plaintiff’s directory, the court did not rule on protectibility, finding as a factual matter that there was insufficient evidence to establish that defendant had copied them.(56)
The Eleventh Circuit reached a similar result in Warren Publishing, Inc. v. Microdos Data Corp.(57) As in BAPCO, the copyrightability of Warren’s compilation — a hard copy directory of cable television systems and their owners — was conceded and therefore not an issue before the court.(58) Microdos, the defendant, marketed an electronic database of information on the cable television industry. Warren claimed infringement as to the communities chosen and the designation of certain of them as “lead communities” in circumstances where a cable operator owns systems in multiple communities.(59) The district court had found that “?the selection
of . . . communities was creative and protectible because Warren uses a unique system in selecting the communities that will be represented in the Factbook.(60) The Eleventh Circuit held that, to the extent that the district court was correct in characterizing Warren’s claim as relating to a system of selecting communities, section 102(b) of the Copyright Act would bar protection.(61) Even if that characterization were incorrect, the court held that Warren’s selection was not original and thus unprotectible. According to the Eleventh Circuit, Warren “did not exercise any creativity or judgment in ?selecting’ cable systems to include in its Factbook, but rather included the entire relevant universe known to it.(62) As to the selection of principal communities, the court held that since Warren made this determination by contacting cable operators and asking them, “the selection is not its own, but rather that of the cable operators.(63) Consequently, the court found no infringement and vacated the district court’s injunction.
BAPCO and Warren appear to equate a compiler’s criteria for selection and organization, respectively, with ideas—which are by definition unprotectible. Taken together, these two cases represent a different approach from the doctrine of “soft facts” or “soft ideas” articulated by the Second Circuit in CCC Information Servs., Inc. v. Maclean Hunter Market Reports, Inc.(64) In CCC, the Second Circuit posited that there are facts or ideas that are “infused with the author’s taste or opinion,” as opposed to explaining phenomena or furnishing solutions to problems.(65) The court recognized that using the merger doctrine(66) to rule out protection for the compilation itself by characterizing as “ideas” the criteria used to select or arrange its contents would render copyright for compilations “illusory.(67) This is because “virtually any independent creation of the compiler as to selection, coordination, or arrangement will be designed to add to the usefulness or desirability of his compendium for targeted groups of potential customers, and will represent an idea.(68) The approach taken by the court was, “[i]n cases of wholesale takings of compilations, a selective application of the merger doctrine, withholding its application as to soft ideas infused with taste and opinion . . .(69)
In summary, very few of the post-Feist compilation cases have held entire works to be uncopyrightable. In fact, copyrightability of the entire work is seldom even contested. Disputes tend to focus instead on the scope of protection. Consistent with Feist’s pronouncement that copyright affords compilations only “thin” protection, most of the post-Feist appellate cases have found wholesale takings from copyrightable compilations to be non-infringing. This trend is carrying through to district courts as well.(70)
Many of the issues discussed in this report relate in some way to the manner in which the database industry operates today: how databases are protected against unauthorized use, and how they are licensed. A basic overview of industry practices is useful in examining the adequacy of existing protection and the impact of any changes in the law.
The information in this section is derived from a number of sources, including comments made at the Copyright Office meetings, public documents and submissions prepared by industry members, including a 1995 position paper of the Information Industry Association,(71) informal queries by Copyright Office staff, and third party publications. While details can be difficult to obtain, since database producers are hesitant to make public their proprietary information and business strategies, it is possible to identify certain common or standard practices as well as general trends.
The limitations in the coverage of copyright, described above, have motivated database producers to be creative in protecting their products. They have developed a variety of legal and business strategies, which are typically implemented in combination. In recent years, three main strategies have emerged. Producers have (1) sought to enhance their copyright protection by altering the structure or content of their databases to incorporate greater creativity; (2) increased their reliance on contracts to restrict the use of databases; and (3) employed technological safeguards to prevent unauthorized access and use.(72)
A. Enhancing Copyright Protection
Some producers have altered the content or structure of their databases, with the goal of obtaining or increasing meaningful copyright protection. One technique is to enhance the contents of the database by adding copyrightable material. Another is to make the database more creative through subjective selection and unusual arrangement of its content. Databases that incorporate one or more of these techniques are often referred to as “value-added databases.” While some companies first began enhancing their factual content in response to Feist, other companies have built their businesses on this practice for market reasons.
It is difficult to obtain specific examples of “enhanced” databases, since producers do not wish to identify databases that are vulnerable to copying, or point to which aspects of their databases they believe to be copyrightable.
Some database producers have purposely added copyrightable text to their databases in response to the Feist decision. Databases that feature copyrightable text have a stronger likelihood of copyright protection than ones that are purely factual. Such text may take the form of descriptive bibliographies, abstracts, profiles or annotations connected to database entries. A competitor interested in copying only the noncopyrightable portions would face two obstacles: (1) the process of separating facts from text may be labor intensive (though possibly less so as technology improves); and (2) the copyrightable material may be integral enough to give the first producer a competitive edge in the marketplace.
Apart from the copyright implications, a market clearly exists for databases combining text with facts or other public domain materials. EMBASE, for example, is an on-line database from Elsevier Science; its index of titles from international biomedicine, pharmaceutical research and related disciplines includes bibliographic information. American Statistical Index, from Congressional Information Inc., provides abstracts as well as indexing of all federal statistical publications. The reports of judicial opinions published by West Publishing Company include synopses of the cases.
There is a simpler, although less effective, alternative to incorporating copyrightable text throughout the database. Some producers insert text only in designated places, such as the foreword or afterword to a directory, or a section describing community services. As a business matter, such additions may provide consumers with a useful component that other compilations lack. From a legal standpoint, this strategy has limitations. It will make the work as a whole protectible, but will not extend protection to otherwise unprotectible material it contains. It therefore could prevent unsophisticated, verbatim copying, but would not serve as a meaningful barrier to copying of the factual component. The producer will, however, be able to obtain a registration for the database with the Copyright Office. Although the copyright still covers only the copyrightable components of the work, the existence of the registration may deter a would-be copier.
As a variant of this strategy, database producers might insert quasi-factual material such as approximations or “soft facts” into the database. In addition to its utility to consumers, material of this sort may have a greater claim to copyright protection, and may therefore help to ensure copyright protection for the database as a whole.(73) The scope of the protection, however, will remain thin.
In the wake of Feist, practitioners advised database producers to increase the likelihood of copyright protection by incorporating a more subjective selection of facts or a more creative arrangement.(74) The utility of this strategy depends on the nature of the database. For some databases, such added value may enhance the desirability of the product. Lawyers have long found the West Publishing Company’s indexing system of cases by key number to be a valuable research tool. Consumers may prefer a listing of restaurants that weeds out those not worth visiting.
This is not the case, however, for the many databases whose market appeal lies in the availability of comprehensive and easily accessible, unadorned facts. From the point of view of an individual user, added textual information may be superfluous or irrelevant, and may make the database too large or unwieldy. Subjective selection may destroy the database’s value as a resource when the user’s goal is to examine all the relevant facts. And creative arrangement, to the extent that it is possible,(75) will almost by definition make the facts more difficult to locate. Moreover, adding these elements entails time and money. This may make the database more expensive to create, which in turn may make it more expensive for consumers.
For many database producers, contracts provide a major source of protection, either complementing copyright law or picking up the thread where it falls short. Although contracts in the database industry were common prior to Feist, in both the digital and the print worlds, companies report that they have reviewed and strengthened them in recent years. These include form contracts as well as negotiated agreements tailored for individuals or institutions. They may appear in traditional print, in shrink-wrap form, on a computer screen as part of software or on-line, or in a combination of these formats. For example, a user may first encounter license terms through shrink-wrap packaging, and then receive the same or additional terms on his computer screen.
Though terms vary from company to company and from product to product, the core coverage of database contracts tends to be similar: contracts restrict access, specify permissible conditions of use, and set terms for enforcement and remedies. They may also contain language designed to educate the consumer about legal rights and limitations.
For databases other than those made freely available to the public (such as telephone directories), contracts are generally the condition of access for a user. Even for a noncopyrightable database, they can also offer users the benefit of timely, updated information.(76)
One common use of contracts is to restrict or limit the manner of use of a database. An on-line license typically dictates the parameters of acceptable downloading and redissemination, as in the following excerpts from the agreements of two major database producers, Dun & Bradstreet and Lexis-Nexis:
You are granted a nonexclusive, nontransferable limited license to access and use for research purposes the Online Services and Materials from time to time made available to you . . . you are prohibited from downloading, storing, reproducing, transmitting, displaying, copying, distributing, or using Materials retrieved from the Online Services. You may not print or download Materials without using the printing or downloading commands of the Online Services.(77)
Customer shall not . . . use Information in connection with providing advice or recommendations to others, publish Information in the news media, incorporate or use Information in any kind of database or marketing list to be provided to a third party, or produce Information in judicial or administrative proceedings, including discovery proceedings, without D & B’s prior written consent, unless required by law. (78)
Customer shall not copy, download, upload or in any other way reproduce Information or Software except . . . Customer may create for internal use online and offline printouts of materials received in electronic form.(79)
These agreements limit users’ ability to use the contents of databases in ways that the law would otherwise allow.
Other agreements used in connection with databases in CD-ROM format make explicit reference to fair use. For example, a Lexis-Nexis contract for CD-ROMs allows users to “create a printout of an insubstantial portion of material retrieved from the Licensed Databases,” and reproduce them “to the extent permitted under the fair use provisions of the Copyright Act.”
Contracts may also establish enforcement procedures and remedies. Such terms can include the ability to terminate a subscriber’s access, suspend or discontinue services, or pursue any other legal remedy.(80)
Terms may be more restrictive for particularly valuable or sensitive information. Dun & Bradstreet, for example, has strict practices for its sensitive information, such as information relating to bankruptcy filings. For these products, it restricts third party distribution and exercises extreme caution in its licensing practices. By keeping direct control over distribution, the company is always in a position to recall or expand earlier data. It also conducts thorough background checks on potential patrons and extends licenses only to those who are creditworthy and risk-free.
Despite their usefulness, database producers report practical limitations on the effectiveness of contract restrictions, primarily as a result of the privity requirement of contract law.(81) A CD-ROM product, for instance, is physically out of a company’s control once delivered to the client. The contract accompanying the product binds only the initial parties; it would not bind third parties who come into possession downstream. As a safeguard, companies may limit CD-ROM licenses to institutions that are not seen as grave risks for piracy.
As might be expected, different companies provide different types of price structures. Some charge users a flat fee; some charge by byte or by minute for databases made available on- line. Others provide free unlimited access as part of a subscription.
Whatever the system used, it is fairly standard to differentiate prices among users. Companies may offer two-, three-, or even four-tier pricing, contingent on the nature of the use that will be made of the database.(82) Some do not charge at all for non-profit or academic uses, and then have graduated rates for different commercial uses. In general, commercial producers report that they make their databases available to educational institutions at greatly reduced fees—often a small fraction of the fees to commercial users. Many will charge at least some fee, in order to cover costs, and to make clear that there is value to their product and that it should be treated accordingly.
Recently, educational institutions have tended to seek “buffet-style” terms, allowing unlimited use by as many users as desired for a flat fee. Various mechanisms are utilized, with the fees to such institutions sometimes priced according to the number of users at that institution. A choice of features may also be available, with higher fees charged for better services or software, such as more sophisticated search engines.
In contrast to commercial producers, some science agencies have found price differentiation impractical, since it can be more expensive for them to keep track of who is entitled to which price than to make the information available for free.
C. Technological Safeguards
Technological safeguards, while offering great promise to producers in supplementing legal protection, are still in the early stages of development. Despite the frequent mention of cryptographic software as a technological solution to the protection of intellectual property, such sophisticated protections are not yet in common use. Few if any database producers today rely upon technology as a sole means of protection, utilizing such safeguards only in combination with licensing and enforcement of legal rights.
The technological safeguards in use today are in large part simple, or low-end, measures. Standard measures in the on-line world include the requirement of user-passwords for dial-up services and the ability to disconnect a user whose behavior appears suspect. Lexis-Nexis, for example, will automatically cut off users if contractual limits for downloading are exceeded. In these circumstances, a user’s password can be temporarily disabled until further information about his or her actions can be obtained.
Nevertheless, some companies believe that encryption is an option whose time is just around the corner. At present, many users of large databases are institutional clients who obtain access through closed system networks. As more database providers consider making their products available for open exploitation over the Internet by individual consumers, encryption needs will expand. For these providers, the goal is to provide a commercial channel secure enough to allow database content and user payments to be exchanged electronically. They believe the growth of commercial products on the World Wide Web will lead to greater use of access keys and other forms of decryption software. Such options are, in the prediction of one company, a year or so away. For now, the combination of relatively high costs and still undeveloped on-line markets make them impractical. It is likely that any on-line dissemination of databases will involve a combination of encryption, secure electronic transfer of funds, and “click-wrap” licenses requiring users to agree to terms before accessing the database.
Meanwhile, however, a few companies have begun to experiment with encryption for their off-line products. For example, one Dun & Bradstreet product, Business Solutions in a Box, is a CD-ROM package of information for small business entrepreneurs. Although the entire database is included on the disk, it appears in encrypted form with restrictions that prevent users from accessing and downloading more than 300 of a possible ten million records unless they pay an additional fee. In keeping with industry trends, however, the encryption is not Dun & Bradstreet’s sole protection. Users must assent to these terms in a “click-wrap” license before accessing any of the records.
Database producers consider technological safeguards, like contracts, to be a useful but imperfect solution. Such safeguards cannot protect databases in print form, which still represent a large proportion of the market. Moreover, high-end measures like encryption are, in their current form, expensive to maintain and inconvenient for users. As technology develops, their cost may decrease and their ease of use may increase. Nevertheless, producers report that they will never rely solely on technological measures to protect their products. First, there are security problems. Such measures, like physical locks, can be broken into; encryption keys, for example, can be shared by users in some circumstances.(83) Second, once a database has been lawfully decrypted, the producer cannot control subsequent access to and use of the decrypted version.
III. COPYRIGHT OFFICE REGISTRATION PRACTICES
The question of whether and to what extent databases are copyrightable is basic to the discussion of database protection generally and is an underlying theme throughout this report. The Copyright Office, as the agency of the U.S. government responsible for registering copyright claims in works of authorship, makes determinations of copyrightability on a daily basis. Its work includes the task of assessing the copyrightability of databases, which are often dynamic works of authorship and can be fixed in many forms, from print to electronic media. These assessments have important consequences. A registration certificate issued within five years of publication is prima facie evidence of copyrightability and of the facts stated in the certificate.(84) Moreover, courts generally give deference to the Register of Copyrights with respect to copyrightability and her decision to allow or deny registration.(85)
Over the years, the Office has developed practices and procedures to assist in determining the copyrightability of the works it examines, including databases. This section discusses registration generally and outlines the practices governing the registration of databases both before and after Feist.
One of the primary roles of the Copyright Office is to register copyright claims in works of authorship.(86) Although registration is not a condition of copyright protection, it provides many benefits,(87) and is therefore routine practice for many commercial copyright owners, including database producers.
In order to be registered, a work must comprise original authorship. When there is a genuine question about the copyrightability of a work, the Office notes its uncertainty by registering under its “rule of doubt.(88) This means that although the work will be registered, “there is a reasonable doubt about the ultimate action which might be taken under the same circumstances by an appropriate court with respect to whether (1) the material deposited for registration constitutes copyrightable subject matter or (2) the other legal and formal requirements of the statute have been met.(89)
Databases may be collections of works (for example, journal articles) or of data (facts). In examining a database for registrability, the Copyright Office must determine whether it is a protectible “compilation” as defined in the Copyright Act.(90) The Office has prepared guidelines to assist its examiners in determining the copyrightability of databases. General guidelines are set out in the Compendium of Copyright Office Practices.(91) More specific guidelines for databases, including those fixed in automated form, are set out in a series of memoranda issued to the examining staff before and after Feist (discussed below).
Where the contents of the database represent new copyrightable subject matter, there is no question that a claim in the database may be registered. Therefore, the Office focuses on whether the claim has been appropriately stated so as to identify that new subject matter. Often, however, the claim is limited to compilation authorship because the contents of the database consist of preexisting materials, whether facts, public domain materials or works that have been previously published. In such cases, the Office must determine whether the selection, coordination or arrangement is copyrightable, making the database registrable.
Where a compilation lacks a certain minimum amount of original authorship, registration will be refused.(92) In general, the greater the amount of material from which to select, coordinate, or order, the more likely it is that the compilation will be found registrable.(93) There is also a basic de minimis quantity test: “Any compilation consisting of less than four selections is considered to lack the requisite original authorship.(94)
B. Pre-Feist Practices
As a general matter, the Copyright Office has always applied an originality standard in examining works for registration. Until the late 1980's, however, it also registered compilations based on “sweat of the brow.(95) Such compilations included but not were not limited to white pages telephone directories and other factual databases. Beginning in 1987, the Office began to question the copyrightability of works where sweat of the brow was the only basis for registration. By 1989, it had abandoned this standard for most compilations, continuing to apply it only to works like telephone directories in which some courts were still upholding copyrightability based on sweat of brow.
Prior to Feist, the Office registered a number of compilations under the rule of doubt. The “doubt” was primarily based on the co-existence in case law of the sweat of the brow standard with the 1976 Act’s explicit originality standard. One such registration involved the bibliographic database of the On-line Computer Library Center (“OCLC”), which consisted of a collection of numerous member libraries' catalogue entries, where the order was determined by the contributing libraries and the arrangement was chronological. Another case involved an application from the National Republican Congressional Committee for registration of its donor lists, arranged by zip code and alphabetically within each code. The Office registered the list as a compilation, but in correspondence with the applicant noted its uncertainty and the need for judicial guidance on the copyrightability of compilations of data.(96)
2. 1988 and 1989 Interim Guidelines for Database Registration
In 1987, the Office reviewed its database practices and considered abandoning its practice of registering sweat of the brow claims. “Interim Guidelines” were issued in 1988 with a memorandum to the staff stating that the courts had concluded that sweat of the brow might still apply to telephone directories but should not be extended to other factual compilations. Examiners were directed to register telephone directories as “a separate category of copyrightable works,” and to reject registration for certain other compilations.(97) The Guidelines specified that criteria used to consider the copyrightability of telephone directories should not be applied to other compilations, with the exception of street directories.(98)
Among the types of claims recommended for refusal were 1) in-house directories that contain an entire universe based on a single source arranged mechanically or alphabetically; 2) “parts” catalogues and price lists where the catalogue or update represents an exhaustive list of inventory and the arrangement is numerical; and 3) membership lists containing the entire universe of members arranged alphabetically, by state or zip code.(99) Because these guidelines rejected sweat of the brow as a basis for protection for most categories of work at a time when some jurisdictions still accepted it, there was dissatisfaction with them both in the Copyright Office and within the private sector, and they were never fully implemented.
In 1989, the Office issued “Guidelines for Registration of Fact-Based Compilations.” The guidelines advised examiners to register commercial telephone, street and business directories, and parts catalogues and inventory lists that were not “clearly de minimis.” (100) Examiners were advised to reject standard organization charts and any compilations containing fewer than four items.(101) More difficult claims, such as mailing lists and subscriber lists, were to be rejected unless “the compilation represents a modicum of selection and/or arrangement authorship and the quantity of material compiled is not de minimis.(102) The Guidelines stated that telephone directories continued to be treated differently by all federal courts (i.e., even those that otherwise rejected sweat of the brow), and therefore should not be examined under "the usual Copyright Office criteria."(103) Thus, in 1989 the Copyright Office moved almost entirely to an originality standard, rejecting sweat of the brow for all compilations except telephone books and similar directories.
During the period from 1985 to 1989, the Office considered some special issues posed by automated databases. The Compendium of Copyright Office Practices defines “automated database” as a “body of facts, data, or other information assembled into an organized format, suitable for use in a computer and comprising one or more files.(104)
One of the major issues posed by automated databases is the status of ongoing updates or other changes.(105) The Copyright Office has been faced with the need to determine when a modified database qualifies as a new work of authorship subject to a separate registration. To the extent that each update of a database contains copyrightable subject matter, it may be registered.(106) Each registration for a published, updated database covers only the additions that were published on the date specified in the application as the date of publication.(107)
Automated databases may be updated frequently; it is not unusual for a database to be updated several times a day. Database producers on many occasions informed the Office that it was impossible as a practical matter to register and deposit the “new” work each time revisions were made available to the public. Those who did register updated versions adopted certain practices designed to ensure that the bulk of their databases was covered by a registration, generally making regular, but periodic, registrations. OCLC, for example, chose to register its updated database once a month, on a day when many additions and revisions were made. However, because such a registration covered only the new material added on the given date of publication, many published updates and additions were not registered.
To address this problem, in 1989 the Copyright Office adopted a regulation allowing group registration for both published and unpublished automated databases.(108) The regulation allows three months worth of updates to be registered at one time, with a deposit consisting of identifying material from one representative day.
C. Post-Feist Practices
The Feist decision did not have a major impact on the Copyright Office’s registration practices for compilations. It did, however, give the Office the clear authority to reject works for which protection was claimed solely on the basis of “industrious collection” or “sweat of the brow.”
Immediately following Feist, the Copyright Office revisited its examining practices for compilations, and issued new guidelines in 1991.(109) In general, the Office concluded that most compilations would continue to meet the standard of originality required by Feist.(110) Nevertheless, examiners were instructed to give extra scrutiny to five types of works: (1) telephone directories; (2) street directories, cross-directories and other directories; (3) periodically updated directories; (4) annual cumulations; and (5) parts catalogues and inventory lists. Specifically, the guidelines advised examiners to reject registration applications where the claim was limited to “white pages,” “listings,” or “revised listings” in phone books, and to continue to question claims where the nature of the contribution was not clear. More specialized or feature-heavy directories, such as business profiles or annotated membership periodicals, were contrasted as compilations that clearly involve sufficient selectivity to be copyrightable.(111)
In practice, the Office continued to exercise considerable judgment in applying the guidelines. It did not categorically refuse all business directories, for example, registering those it believed to contain copyrightable authorship. The Office sent letters to remitters of compilations, citing the Feist holding and the corresponding change in Office practices. In the cases where applicants continued to submit telephone directories, examiners suggested statements of the claim in terms of copyrightable authorship, such as “new text in foreword, “ or “revisions and additions to yellow pages.(112) Such correspondence over claim specifications and disclaimers prompted some concern in the private sector that the Office’s new practices were akin to a “patentization” of the copyright system.(113)
The Copyright Office’s heavy post-Feist correspondence lasted for a few months. Within two years of Feist, the Office was corresponding with applicants only on occasion. Many applicants sent letters with their submissions, specifying what they believed constituted the originality in their selection, coordination and arrangement.
Despite industry concerns that the Office would interpret Feist too broadly, establishing strict bright line rules, the Office continued, and continues today, to accept most compilations submitted for registration. It is impossible to know, however, how many compilation claims are not submitted because their owners are concerned that the Office will question copyrightability or refuse registration.
Occasionally, the Office still receives an application to register white pages telephone directories. In 1991, Southwestern Bell submitted a claim for copyright registration in certain features of the St. Louis White Pages. The company claimed that the work was copyrightable either as a compilation or as a graphic work. The latter claim was based on the typeface and layout of the page, which included various “user-friendly” features. When the Copyright Office denied registration, Southwestern sued under the Administrative Procedures Act.(114) The court affirmed the Copyright Office’s denial of registration, finding no abuse of discretion.(115)
Databases have also been the subject of attention in the international arena. Their status as copyrightable subject matter is guaranteed by the two major multilateral treaties relating to copyright. The Berne Convention for the Protection of Literary and Artistic Works since 1948 has required member countries to protect “[c]ollections of literary or artistic works such as encyclopaedias and anthologies which, by reason of the selection and arrangement of their contents, constitute intellectual creations.(116) The coverage of databases of fact was confirmed in 1995 by the TRIPs Agreement,(117) which states: “Compilations of data or other material, whether in machine readable or other form, which by reason of the selection or arrangement of their contents constitute intellectual creations shall be protected as such.(118)
Since January 1996, developed country members of the World Trade Organization have been bound by this obligation; the obligation takes effect for all other members over the next few years.(119) The TRIPs Agreement also specifies that the copyright protection for compilations “shall not extend to the data or material itself,(120) and contains another provision stating that “[c]opyright protection shall extend to expressions and not to ideas, procedures, methods of operation or mathematical concepts as such.(121)
Pursuant to the action plan set out in its 1991 “Follow-up to the Green Paper,(122) the European Commission proposed in 1992 to harmonize the national laws within the European Union regarding the protection of databases. The Commission proposal was adopted in a modified form as a directive to the member states on March 11, 1996 (attached to this report as Appendix C).(123) The directive is required to be implemented by the member states by January 1, 1998.
A number of factors appear to have led the European Union (EU)(124) to harmonize the law regarding database protection. The rapid expansion of the Internet raised the EU’s awareness of “the exponential growth, in the Community and worldwide, in the amount of information generated and processed annually in all sectors of commerce and industry,” and the important role of databases “in the development of an information market within the community.(125) The EU also expressed concern about the “very great imbalance in the level of investment in the database sector both as between the Member States and between the Community and the world’s largest database-producing third countries.(126) In addition, the Feist decision in the U.S. Supreme Court galvanized concern regarding the adequacy of copyright protection for databases within the EU.(127)
The directive covers compilations of data in any form, and thus includes hard copy compilations as well as electronic databases.(128) The Commission’s original proposal was limited to electronic databases, but in the course of deliberations this approach was found unworkable, because it would subject the identical material to differing legal standards based solely on the medium employed. As one of the participants is reported to have stated, “making use of a scanner should not be decisive in granting legal protection.(129) In addition, technologies such as scanning and optical character recognition render even hard-copy databases vulnerable to unauthorized copying and commercial reuse in both hard-copy and electronic form.(130) Moreover, the TRIPs Agreement makes no such distinction.(131)
As adopted, the directive establishes a dual system for protection of databases. One component is copyright protection for the “structure” of the database.(132) The other is a sui generis (“of its own kind” — i.e., not falling within existing categories of legal protection) intellectual property right in the contents of the database.
The copyright portion of the directive, Chapter II, applies only to the structure or schema of a database, without prejudice to any existing protection under copyright for the database contents.(133) It seeks to harmonize the scope of copyright protection for databases throughout the European Union. It does so in two major respects: First, it sets a uniform standard of originality. Second, it establishes a uniform list of “restricted acts” (i.e., exclusive rights) and exceptions to restricted acts.
Prior to the directive, copyright protection for databases in the member states could be divided into two general groups. In the U.K., Ireland and the Netherlands, the threshold for protection was quite low. In particular, Anglo-Irish common law incorporated a “sweat of the brow” doctrine that developed from the same line of eighteenth and nineteenth century English cases that were cited in early U.S. compilation cases.(134) In the remaining European countries, however, copyright imposed a fairly high threshold of originality to qualify for protection.(135) This is in keeping with the “author’s right” approach that prevails throughout most of Continental Europe, which defines originality as an expression of the author’s individual personality.(136)
The standard established by the directive requires the database to, “by reason of the selection or arrangement of [its] contents, constitute the author’s own intellectual creation.(137) This language was incorporated verbatim from the EU’s 1991 directive on the protection of computer programs.(138) It was originally adopted to override the very high standard of originality mandated by the German Supreme Court in the “Inkasso Programm” case and other decisions.(139) At the same time, by requiring an “intellectual creation,” the database directive imposes a higher standard of originality than that required under current law in the U.K., Ireland and the Netherlands. The directive thus charts a middle course on the level of originality required. Although the directive’s standard of originality has not been tested in practice, the formulation appears to be quite similar to the criteria for protection under U.S. law, as set out in the definition of “compilation” in the Copyright Act and interpreted by the Supreme Court in Feist.(140)
The “restricted acts” (exclusive rights of the copyright owner) under the directive are reproduction (temporary or permanent), adaptation, distribution, and communication, display or performance to the public.(141) Authorization is not required for a lawful user to engage in any restricted act “which is necessary for the purposes of access to the contents of the database and normal use of the contents.(142) Any contractual provision to the contrary is “null and void.(143)
In addition to this mandatory exemption, the directive permits member states to provide for limitations on the restricted acts in the following cases:
(a) in the case of reproduction for private purposes of a non-electronic database;
(b) where there is use for the sole purpose of illustration for teaching or scientific research, as long as the source is indicated and to the extent justified by the non-commercial purpose to be achieved;
(c) where there is use for the purposes of public security o[r] for the purposes of an administrative or judicial procedure;
(d) where other exceptions to copyright which are traditionally authorized under national law are involved, without prejudice to points (a), (b) and (c).(144)
Such exceptions are subject to an overall economic harm limitation, ensuring that they cannot “unreasonably prejudice[ ] the rightholder’s legitimate interests or conflict[ ] with normal exploitation of the database.(145)
As a supplement to copyright, Chapter III of the directive establishes a sui generis form of protection for the contents of databases. The stated justification for this protection is that “in the absence of a harmonized system of unfair-competition legislation or of case-law, other measures are required in addition [to copyright] to prevent the unauthorized extraction and/or re- utilization of the contents of a database,” the making of which “requires the investment of considerable human, technical and financial resources while such databases can be copied or accessed at a fraction of the cost needed to design them independently.(146)
Some of the EU member states originally advocated leaving the protection of the contents of databases to unfair competition law, and the initial Commission proposal described the sui generis right as a “right to prevent unfair extraction from a database” for commercial purposes.(147) By mid-1993, however, “an increasing majority of interested parties” were reportedly favoring the creation of a property right along the lines ultimately adopted.(148) The rationale, at least in part, was the perceived difficulty in harmonizing unfair competition law throughout the European Union. In addition, the Commission has noted that “unfair competition rules only come into play once an act has taken place. They do not provide an economic right with clear scope which can be freely transferred.(149)
In some respects the sui generis right is similar to the “catalogue rule” existing in the Nordic countries, which provided a model for the Commission. That rule establishes a “related right” for factual compilations, in addition to copyright protection. The catalogue rule provides to the producer of a catalogue, table, or similar matter “in which a large number of information items have been compiled” a right against unauthorized reproduction.(150) Originality is not a requirement for protection, and the term of protection for such “catalogues” is fairly short: 10 years from publication or 15 years from creation, whichever expires sooner.
The essential features of the database directive’s sui generis right are:
a. Protection for “substantial investment.” The sui generis right is available for “the maker of a database which shows that there has been qualitatively and/or quantitatively a substantial investment in either the obtaining, verification or presentation of the contents . . .(151) “Substantial investment” is not defined in the directive. However, the recitals leading up to its provisions indicate that “such investment may consist in the deployment of financial resources and/or the expending of time, effort and energy.(152)
b. Protects against acts of extraction and re-utilization. The rights accorded under the directive are the rights to “prevent extraction and/or re-utilization of the whole or of a substantial part . . . of the contents of that database.(153) “Extraction” is defined as “the permanent or temporary transfer of all or a substantial part of the contents of a database to another medium by any means or in any form.(154) “Re-utilization” is defined as “any form of making available to the public all or a substantial part of the contents of a database by the distribution of copies, by renting, by on-line or other forms of transmission.(155)
c. “Insubstantial parts” excluded from protection. The maker of a database “may not prevent a lawful user of the database from extracting and/or re-utilizing insubstantial parts of its contents . . . for any purposes whatsoever.(156) Any contractual provision to the contrary is “null and void.(157) The directive does not attempt to define “insubstantial parts,” but does state that substantiality is to be “evaluated qualitatively and/or quantitatively.(158)
d. Exceptions for certain uses. The directive permits member states to adopt exceptions from the sui generis right for lawful users in three specific categories: (a) extraction for private purposes of the contents of a non-electronic database; (b) “extraction for the purposes of illustration for teaching or scientific research, as long as the source is indicated and to the extent justified by the non-commercial purpose to be achieved”; and (c) “extraction and/or re- utilization for the purposes of public security or an administrative or judicial procedure.(159) These exceptions are similar to those permitted under copyright, but without the additional reference to “other exceptions to copyright which are traditionally authorized under national laws.” Nevertheless, the recitals indicate that existing exemptions to any existing similar sui generis rights are grandfathered under the directive.(160)
The exceptions must be read in conjunction with provisions in the directive on “obligations of lawful users,” prohibiting lawful users of databases that have been made available to the public from “performing acts which conflict with normal exploitation of the database or unreasonably prejudice the legitimate interests of the maker of the database,” or “caus[ing] prejudice to the holder of a copyright or related right in respect of the works or subject matter contained in the database.(161)
e. Fifteen year term of protection. The term of protection for the sui generis right is fifteen years.(162) This was an increase from the ten-year term that was originally proposed in 1992.(163) Any qualitatively or quantitatively “substantial change,” including one resulting from an accumulation of small changes, “which would result in the database being considered to be a substantial new investment,” qualifies the resulting database for its own fifteen-year term of protection.(164)
f. Available to non-EU nationals only on the basis of reciprocity. The sui generis right is available only to database makers who are EU nationals or habitual residents.(165) For purposes of the directive, this would include business entities that have a business presence in the EU (defined as a central administration or principal place of business in the EU, or a registered office in the EU plus a genuine, ongoing operational link with the economy of a member state).(166) The EU can conclude agreements to extend the right to databases made in third countries.(167) Although the provisions of the directive themselves are silent as to the basis for such agreements, the recitals make clear that protection will be offered only on the basis of reciprocity— i.e., where the third country offers “comparable protection” to EU databases.(168)
The original proposal for the directive also included a compulsory license, requiring database vendors who are the sole source of any given information to license that information to competitors on “fair and non-discriminatory terms.(169) This provision proved controversial. It was dropped after the European Court of Justice imposed a similar licensing requirement under existing principles of EU competition law in the “Magill case.(170) At the same time, apparently as part of an overall compromise, changes were made in the scope of the right and the exceptions, as well as the provision on rights of lawful users.(171)
The recitals acknowledge the important role of competition policy in the database area.(172) In addition, the directive establishes a procedure for review every three years to determine, among other things, “whether the application of [the sui generis] right has led to abuse of a dominant position or other interference with free competition which would justify appropriate measures being taken, including the establishment of non-voluntary licensing arrangements.(173)
Member states are obligated to implement the directive by January 1, 1998. Germany has already enacted implementing legislation, and we understand that most of the other member states expect to meet the deadline. All are actively preparing implementing legislation, and several are at an advanced stage in their internal processes of adoption.
In February 1996, the European Union submitted a proposal to the World Intellectual Property Organization (WIPO) in Geneva for a treaty on the subject of legal protection for databases, for consideration by the WIPO Committees of Experts that had been meeting on a regular basis to discuss two other proposed treaties in the field of copyright and neighboring rights.(174) The European treaty proposal would have required countries adhering to the treaty to establish a new, non-copyright form of protection for databases, referred to as sui generis protection. It was based on essentially the same concept contained in the directive, but framed in the simpler, more bare-bones style of international treaty language. The European proposal was distributed and briefly discussed at the Committee of Experts meeting that took place the week of February 5-9.
The next Committee of Experts meeting took place in May 1996. For consideration at that meeting, the United States submitted its own treaty proposal on the subject of database protection. This proposal differed from that submitted by the EU in several respects, primarily relating to the protectibility of government databases (the U.S. proposed allowing countries to deny protection), the term of protection (the U.S. proposed 25 years), the requirements for protecting foreign databases (the U.S. proposed national, i.e. non-discriminatory, treatment), and the ability to vary rights by contract (the U.S. proposed complete freedom of contract). Again, there was a brief debate of the database issues, during the course of a week-long discussion including the two other proposed treaties. The meeting concluded with a decision by the Committees to recommend to the Governing Bodies of WIPO that a Diplomatic Conference be convened in December 1996 to consider the adoption of treaties in all three areas.
On August 30, 1996, WIPO distributed draft texts of three treaties prepared by the Chairman of the Committees of Experts, Jukka Liedes of Finland, based on the various treaty proposals made by governments and the Committees of Experts’ discussions of those proposals. One of the three, entitled “Basic Proposal for the Substantive Provisions of the Treaty on Intellectual Property in Respect of Databases to be Considered by the Diplomatic Conference” (the “ Draft Database Treaty” or “draft treaty”), dealt with the proposed sui generis right in databases.(175) A copy is attached as Appendix D.
The draft treaty combined elements of both the European and the U.S. proposals. As is standard in WIPO treaties, it set out the basic concepts of the subject matter of protection and the nature and duration of the rights, but left to individual countries the freedom to flesh out the details through exceptions and limitations.
The draft was structured as follows: A preamble described the importance of databases in the global information infrastructure and the need to provide effective legal protection for them on an international basis. Article 1 set out the scope of protection, requiring contracting parties to protect databases in any form or medium, based on the criterion of “a substantial investment in the collection, assembly, verification, organization or presentation” of the contents, irrespective of any protection under copyright or other legal rights in the database under national law.
The rights to be granted were defined in Article 3 as “the right to authorize or prohibit the extraction or utilization” of the database’s contents. They were to be granted to the maker of the database, and thereafter be freely transferable.(176) No formalities could be imposed on the enjoyment and exercise of the rights.(177)
Article 2 contained definitions of the critical concepts of “database,” “maker of the database,” and “substantial investment,” as well as the acts of “extraction” and “utilization” that constituted the rights to be granted, and the term “substantial part” appearing in the definition of those acts. “Database” was defined as “a collection of independent works, data or other materials arranged in a systematic or methodical way and capable of being individually accessed by electronic or other means.” The “maker of the database” was “the natural or legal person or persons with control and responsibility for the undertaking of a substantial investment in making a database.” The requisite “substantial investment” was defined as “any qualitatively or quantitatively significant investment of human, financial, technical or other resources in the collection, assembly, verification, organization or presentation of the contents of the database.”
An act of “extraction” was “the permanent or temporary transfer of all or a substantial part of the contents of a database to another medium by any means or in any form,” while an act of “utilization” was
the making available to the public of all or a substantial part of the contents of a database by any means, including by the distribution of copies, by renting, or by on-line or other forms of transmission, including making the same available to the public at a place and at a time individually chosen by each member of the public.
Finally, the “substantial part” referred to in both of these definitions was itself defined as “any portion of the database, including an accumulation of small portions, that is of qualitative or quantitative significance to [its] value.”
Exceptions were dealt with in Article 5. Using the long-accepted language from the Berne Convention(178) and the TRIPs Agreement,(179) this article allowed individual countries to provide exceptions or limitations to rights “in certain special cases that do not conflict with the normal exploitation of the database and do not unreasonably prejudice the legitimate interests of the rightholder.” It also left to national determination the treatment of databases made by governmental entities, agents or employees.
Two alternatives were given for the term of protection: either 15 or 25 years.(180) A separate paragraph required a new term of protection when a database is changed through a new investment sufficient to qualify for protection in itself:
Any substantial change to the database, evaluated qualitatively or quantitatively, including any substantial change resulting from the accumulation of successive additions, deletions, verifications, modifications in organization or presentation, or other alterations, which constitute a new substantial investment, shall qualify the database resulting from such investment for its own term of protection.(181)
This term of protection would apply not only to databases produced after the entry into force of the treaty, but also those that were already in existence at that time.(182) The resulting new protection for existing databases would not affect, however, any “acts concluded or rights acquired” before the treaty’s entry into force, and countries could allow continued distribution of previously lawfully made copies for a period of up to two years.
The draft treaty made clear that the new form of protection would not affect in any way other laws relating to databases or their contents, including copyright, antitrust law, data protection and privacy, access to public documents and the law of contract.(183)
Each country would be required to protect makers of databases who were nationals of other contracting parties, on a national treatment basis—i.e., providing the same rights in respect to sui generis protection that it provided to its own nationals.(184) Finally, the draft contained the same provisions on technological protection measures and enforcement of rights as were contained in the two other draft treaties.(185)
The treaty proposal proved controversial within the United States. Numerous comments were submitted to the Patent and Trademark Office, with the overwhelming majority opposing conclusion of a treaty at that time, expressing concerns about the impact of such protection and/or urging delay until there had been an opportunity for full domestic debate. This group of commentators included members of the scientific, library and educational communities, as well as individual members of the public and at least one major database producer. Those commenting in favor were companies and trade associations from the publishing and database industries.
The draft treaty on databases was never reached or discussed in substance at the December 1996 Diplomatic Conference. The negotiation of the other two treaties continued throughout the entire three weeks of the Conference, concluding only in its final hours on the last day. During the Conference, a number of delegations expressed the view that the database treaty was premature, and that they were not ready to negotiate its provisions. Ultimately, the Conference adopted a recommendation that the Governing Bodies of WIPO should convene in March of 1997 and decide on the course of future work on the issue.
On March 20-21, the Governing Bodies determined that the subject of legal protection for databases should be taken up again in a meeting in Geneva on September 17-19, 1997. This will be an informational meeting, where delegations will discuss the treatment of databases under the laws of their respective countries, and their experiences in dealing with the subject. The draft treaty itself is not scheduled to be the topic of debate.
During 1996, the possibility of legislation providing a new form of protection for databases was raised in the respective Congressional committees dealing with intellectual property.
In February 1996, the chief intellectual property counsel for the Senate Committee on the Judiciary attended the meeting of the WIPO Committees of Experts in Geneva, and reported to the delegates that the Senate was examining the issue and considering the introduction of legislation.
On May 23, 1996, during the next WIPO Committees of Experts meeting, Congressman Carlos Moorhead, then Chairman of the House of Representatives Subcommittee on Courts and Intellectual Property, introduced H.R. 3531, entitled the “Database Investment and Intellectual Property Antipiracy Act of 1996" (attached as Appendix E).(186) Like the subsequently-prepared WIPO draft treaty, the bill would have protected databases that result from a substantial investment against various acts of unauthorized extraction or use. As is usual with national legislation, the bill differed from the treaty primarily in containing more detail, particularly about the scope of rights and exceptions, as well as in spelling out remedies.
Section 2 was the definitional section. It defined “database” as “a collection, assembly or compilation, in any form or medium now or later known or developed, of works, data or other materials, arranged in a systematic or methodical way.” The terms later relied on to delineate the prohibited acts, “extraction” and “use and reuse,” were defined as follows:
“Extraction” means the permanent or temporary transfer of all or a substantial part of the contents of a database or of a copy or copies thereof. Such transfer may be to an identical or different medium, and by any means or in any form, now or later known or developed.
“Use” and “reuse” means making available all or a substantial part, qualitatively or quantitatively, of the contents of a database, or access to all or such substantial part, whether or not for direct or indirect commercial advantage or financial gain, by any means now known or later developed, including any of the following: (i) marketing, selling, or renting; (ii) in the form of permanent or temporary copies; or (iii) by distribution, any online or other form of transmission.
The bill did not define “substantial part,” but defined its opposite, “insubstantial part,” as “any portion of the contents of a database whose extraction, use or reuse does not diminish the value of the database, conflict with a normal exploitation of the database or adversely affect the actual or potential market for the database.”
Section 3 set out the standards a database would have to meet to qualify for protection. A database would qualify
if it is the result of a qualitatively or quantitatively substantial investment of human, technical, financial or other resources in the collection, assembly, verification, organization or presentation of the database contents, and (i) the database is used or reused in commerce; or (ii) the database owner intends to use or reuse the database in commerce.(187)
Specifically excluded from protection were databases made by a governmental entity, whether state or federal, but not databases whose contents were obtained from such an entity.(188) Another subsection ruled out protection for computer programs.(189)
The prohibited acts were set out in Section 4. The bill would have made it unlawful to perform the following acts without authorization:
(1) extract, use or reuse all or a substantial part, qualitatively or quantitatively, of the contents of a [protected] database . . . in a manner that conflicts with the database owner’s normal exploitation of the database or adversely affects the actual or potential market for the database;
(2) engage . . . in the repeated or systematic extraction, use or reuse of insubstantial parts, qualitatively or quantitatively, of the contents of a [protected] database . . . in a manner that cumulatively conflicts with the database owner’s normal exploitation of the database or adversely affects the actual or potential market for the database; or
(3) procure, direct or commission any [of the foregoing] act[s].(190)
The bill further provided examples of circumstances in which acts of extraction, use or reuse would be considered to conflict with a normal exploitation of the database or adversely affect its actual or potential market.(191)
Exceptions were dealt with in Section 5. One paragraph stated that a lawful user of a database could extract or use insubstantial parts of its contents for any purpose, subject to the “repeated or systematic” test of section 4(2), set out above.(192) The other made explicit that anyone was free independently to collect, assemble or compile from other sources any of the material contained in a database.(193)
Section 6 established the duration of protection. It provided a basic term of protection of 25 years, but with the ability to obtain a new term upon “any change of commercial significance.”
Sections 7 and 8 established remedies, both civil and criminal.
Section 9 explained the relationship of the proposed protection to other bodies of law. It stated that copyright protection would not be affected, and that parties would remain free to enter into contractual agreements with respect to databases or their contents.(194) It also made clear that
[n]othing in th[e] Act shall prejudice provisions concerning copyright, rights related to copyright or any other rights or obligations in the database or its contents, including laws in respect of patent, trademark, design rights, antitrust or competition, trade secrets, data protection and privacy, access to public documents, and the law of contract.(195)
Sections 10-13 dealt with the circumvention of technology used to protect databases against unauthorized acts, and with the integrity of database management information. Their language paralleled similar prohibitions contained in the then-pending bills proposing a National Information Infrastructure Copyright Protection Act.(196)
Section 14 contained a three-year statute of limitations.
Section 15 made the date of enactment the effective date of the act, and barred liability for the use or reuse of database contents lawfully extracted from a database prior to that date.
The House bill was introduced as an indication to the international community that Congress was interested in pursuing the subject of database protection. No hearings were held, and no corresponding bill was introduced in the Senate.
In the months following introduction of H.R. 3531, many of the groups that had opposed conclusion of the draft WIPO treaty expressed similar concerns about the bill relating both to substance and to timing and process. They urged that all interested parties be given an opportunity to provide input and that a thorough analysis of the issues be undertaken. Their substantive concerns are reflected in the discussion of the issues in Section VII below.
In order to provide Congress with complete and balanced information, the Copyright Office sought to meet with as many interested parties as possible. The Office initiated the process by scheduling a series of five meetings with the major groups that had already been vocal in indicating their interest in the subject of database protection: (1) the library community; (2) science agencies and organizations; (3) educational groups; (4) database producers who favor legislation; and (5) database producers who oppose legislation or do not favor it at this time. The goal was to start by ascertaining the shared views and concerns of each of these identifiable groups.
These meetings took place in March, May and June of this year. The participants were selected as follows: the Office identified those entities and individuals whose interest we had learned of through prior contacts and discussions. We added the names of organizations and associations who had submitted comments to the Patent and Trademark Office on the WIPO draft database treaty, or who had contacted the staff of the Congressional committees to communicate their concerns. We then asked representatives from each group to suggest any additional parties who should be invited. Finally, some participants contacted us directly with requests to attend.
Subsequent meetings were scheduled with persons or entities whose interests were distinct from the larger groups, or who had scheduling problems making it difficult for them to attend the large meetings. The Office also met with several academics and lawyers with particular expertise on the subject, who shared their own analysis of the issues presented. Finally, we made ourselves available to meet with anyone else who wished to communicate views or concerns. In total, the Office held sixteen meetings, as well as receiving a number of additional communications, including by mail or telephone.
The meetings were structured to provide an informal environment conducive to focused, productive and open discussion. All were led by Marybeth Peters, the Register of Copyrights, with the assistance of the staff of the Office of Policy and International Affairs. Each participant was given an opportunity to present its specific views, and then an unlimited time period was devoted to general discussion of the issues. While there were no formal presentations or questions, Copyright Office staff occasionally asked questions to clarify facts or positions. The discussions were not transcribed, and written statements were not required, although some participants chose to submit them during or after the meetings.
A list of those attending the meetings is attached to this report as Appendix F. We note that the number of participants on any side of an issue was purely the result of the selection process described above. The Office made no attempt to achieve a numerical balance or to evaluate the relative size or importance of any interest group or position.
This section gives a general overview of the views expressed in the Copyright Office meetings. Inevitably, it cannot constitute a complete or perfectly accurate description of any one party’s or group’s views, but represents our best effort at communicating the essence of each position. There will, of course, be numerous opportunities at later stages in the process of legislative consideration for additional presentations and submission of materials. We do not identify particular parties, except where necessary to describe a distinct point of view.
From the outset, the Copyright Office made clear that it was starting from first principles and working from a clean slate, rather than assuming that any of the proposals from last year would be the starting point for Congressional consideration. During the meetings, however, elements of prior proposals were frequently discussed, and these discussions are reflected below where useful and relevant.
The meetings indicated that core elements of agreement exist as to certain principles, where the difficulty lies primarily in determining how to implement those principles (whether by legislation or the absence of legislation). Thus, participants generally agreed on the following points: (1) databases are vulnerable to copying, and adequate incentives are needed to ensure their continued creation; (2) individual facts should not be the subject of private ownership; (3) anyone should be free to obtain facts independently from original sources, even after they have been incorporated in a database; (4) government databases should not be protected; (5) it is important not to harm science, research, education and news reporting; and (6) “free riding” in the form of substantial copying for commercial, competitive purposes should not be permitted.
In other areas, there is intense disagreement as to fundamental principles. The participants sharply differed, for example, on the adequacy of existing means of protection for databases; whether additional statutory protection or its absence is more likely to diminish access to data or raise its cost; and whether non-competitive uses that may harm the market for a database should be permitted.
Some participants in the Copyright Office meetings held strong views either in favor of new legislation or in opposition. In general, many members of the library and scientific communities, as well as some educational groups, telephone companies and Internet-related businesses, expressed opposition, while a majority of database producers, including producers of a variety of scientific and scholarly databases, and the owner of a major on-line retrieval service advocated legislation. It must be stressed, however, that positions were not uniform within all of these communities. Some commercial database producers, including one of the largest in the global marketplace, oppose legislation at this time; many scientific researchers, particularly those working for industry, favor it. The reasons for the differences among those who appear to be similarly situated were not always clear. In some cases, it may simply be that they hold differing perceptions of the law or the potential dangers posed.
A large number of the participants were undecided, or took neutral or intermediate positions. Many have interests on both sides of the issue, as they both produce databases and rely on information obtained from the databases of others. They generally expressed a desire to ensure adequate incentives, along with concerns about the possible negative impact of new protection. For some of these participants, their view of any legislation would turn on the form and scope of protection it provided. Others were still analyzing the issues, and had not yet formed an opinion.
The Office also met with several groups or entities with no position on the advisability of legislation generally, but with a specific concern about how some aspect of any such legislation might affect their activities. Most of these participants stated that if a need were established for new protection, and it was possible to provide adequate protection without harming legitimate user interests, they would either support or not oppose legislation.
This section briefly summarizes the positions of those with clear views either pro or con. More detail will be given in the discussion of the issues below.
Proponents
Proponents of new legislation make the following principal points:
(1) Databases are increasingly important to the U.S. economy and to science, and will be a key component of content on the Internet. They often provide information not otherwise available from a single source in a usable form, and ensure that the information is reliable and timely. Given the acceleration of developments in communications, storage and retrieval technologies over the past five to ten years, vast quantities of information are made available today much more quickly, and users have much greater capabilities to access and manipulate it. In addition, markets and science have evolved to demand increasing levels of comprehensiveness, accuracy and timeliness.
(2) Large investments of time and money are necessary to produce and maintain many databases. Voluminous information must be collected, placed in a usable format, and kept accurate and up-to-date.
(3) While it is expensive to collect and verify large numbers of facts, it is increasingly cheap to copy and disseminate them. Databases are therefore vulnerable to acts of piracy that threaten to destroy or significantly reduce their markets. This threat has been growing with the evolution of technology. With today’s digital and scanning capabilities, major investments in both online and hard copy databases can be hijacked with the stroke of a key.
(4) Existing law is insufficient to protect against this threat. Although various forms of protection are available today, both legal and technological, there has been a gap in the law since the Supreme Court’s decision in Feist Publications v. Rural Telephone Service Co., resulting in an inability to obtain satisfactory legal relief in many circumstances. Problems have already been experienced by a number of database producers. It is critical to restore the protection against piracy that existed under the “sweat of the brow” theory of copyright law.
(5) Unless adequate protection is available for databases that require substantial resources to produce and maintain, such investments will significantly diminish. The result will be a loss not only of commercial profits, but of the public benefits accruing from the creation of databases and access to the information they contain. On the international level, markets for databases have become global, and the United States must provide adequate protection if it is to avoid competitive disadvantage with other regions of the world such as the European Union.
(6) The needed protection can be provided through appropriately crafted new legislation without harming the legitimate interests of the science community and other user groups. Indeed, scientific research will benefit from such protection, since researchers rely heavily on the private sector to make the high levels of investment necessary to produce and maintain reliable, up-to-date and comprehensive collections of scientific data.
Opponents
Generally, those that oppose new protection or are doubtful about its advisability do not contest the proponents’ assertions as to the importance of databases, the changes brought about in their creation, dissemination and use by developments in technologies, and the need to provide adequate incentives. They disagree as to the conclusions to be drawn, however, and make the following points:
(1) Proponents have not produced sufficient evidence that a problem exists that requires a legislative solution. Their arguments about the need for additional protection are based on theory, isolated anecdotes, and speculation about possible future harm. International developments, particularly the outcome of the European directive, will not cause them serious detriment. It would be premature for Congress to legislate without more extensive factual evidence, or without expert economic analysis.
(2) The combination of means of protection that exist today appears to be adequate. Copyright law continues to protect databases even after Feist. The Supreme Court in Feist explicitly stated that most databases would still qualify for copyright protection, and subsequent cases have borne this out. Moreover, recent case law has made clear that meaningful protection is available outside of copyright through contract and the common law of misappropriation. Technological means of protection are also available and effective. Proponents can come to Congress if this situation changes, for example if the case law begins to develop in an unsatisfactory direction.
(3) The U.S. database industry today is an example of market success, not market failure. The industry is thriving under the current legal regime, and has become the leader in the global marketplace. Databases continue to be created and marketed, and businesses are paying record sums of money to purchase database producers.
(4) In this area, it is critical to proceed with great caution, especially in a time of rapidly evolving technologies and uses of data, since it is hard to predict future implications. New protection could result in negative consequences, even if unintended. A perceived trend toward commercialization of data, particularly data produced by government funding, could be exacerbated. Information could as a practical matter become less accessible or more expensive; concern about potential liability could have a chilling effect on uses of information that are in the public interest, such as scientific research and education. New legal protection could raise a new barrier to market entry for second comers, decreasing rather than increasing already low levels of competition and driving up prices.
(5) Copyright law embodies an appropriate balance between incentives for creation and the free flow of information, by granting rights but leaving ideas and facts in the public domain and providing leeway for public interest activities through the doctrine of fair use and other exceptions. This balance furthers Constitutional policies and should not lightly be disturbed. New rights should not be provided, especially if they give equivalent or greater protection than copyright, without the justification of creativity; facts should be left free for all to use.
In addition to these general points, government science agencies have raised concerns about the impact of any new protection in this area on the policy of full and open access to data that the United States has strongly pursued in the international arena.(197)
During the course of the meetings, several key issues began to emerge. At each meeting, different ones were stressed, different concerns expressed, and different subsidiary issues identified. Overall, however, most of the substance of the discussions can helpfully be grouped under one of six topics. It should be noted that the issues are interrelated in many respects, so that they cannot each be resolved in a vacuum.
The first, and threshold question, is whether additional legal protection for databases is needed. Several participants made the point that even if a need were shown for additional protection, Congress should not enact legislation without performing a cost-benefit analysis to determine whether the need outweighed the harm that would be caused by any such legislation. Of course, the ultimate test of any proposed legislation is whether its benefits outweigh its costs, and this test would have to be met in order for any form of database protection to be enacted. We have not addressed it as a separate issue in the report, however, because the question of what harm might be caused is dependent on how all of the other issues are resolved. Accordingly, our discussions of potential harm occur in the context of specific issues. For example, certain types of harm might be avoided by adopting certain exceptions, or otherwise framing the scope of protection in a certain way.
If the threshold question of need is answered in the affirmative, the next question is what type of protection would be preferable—a new form of property right, or a tort concept closer to unfair competition. The remaining issues also would only need to be addressed if Congress decides that some form of legislation is desirable. These are: (1) definitional issues—how should the concept of “database” be defined, what should be the criteria for a database to qualify for protection, and what degree of taking should be actionable? (2) how can it be ensured that uses of information in the public interest, such as for scientific, educational and news reporting purposes, are not harmed? (3) what should be the duration of any such new protection? and (4) how should “sole source” data be handled—i.e., situations where the data contained in a protected database is not available elsewhere? One additional issue was not discussed at any length in the meetings, but would need to be examined: what constitutional constraints may limit Congress’s ability to legislate in this area. We discuss each of these issues in turn, describing the main points raised in the meetings.(198) In the process, we note those points where the issues are most obviously inter-related.
B. Is There a Need for Additional Protection?
All agree that the proponents of a new form of statutory protection have the burden of establishing the need for such protection. Some participants in the Copyright Office process chose to refrain from discussing any other issue pending the resolution of this first issue. They felt that they could not analyze what models of protection might be appropriate before identifying clearly the nature and scope of the problem.
Establishing such a need is the threshold question for any legislative initiative. Traditionally, the proponent of any change in the law, whether new rights or new limitations on rights, has borne the burden of convincing Congress of the need for the change.
Views diverge sharply, however, as to the type and degree of proof required to satisfy this burden. The options proposed ranged the gamut from a threat of future harm, to evidence of individual real-world problems, to empirical data generated through broad-scale studies. In the past, changes in intellectual property law have often been based on evidence of one of the first two types. Those arguing for economic studies believe that a higher standard is necessary here, either because it is preferable for the establishment of any new right, or because of the special nature of protection for collections of data in particular.
Various forms of protection against piracy do exist today for databases in the United States. Chief among these are copyright law, contracts, state misappropriation doctrine, trade secrecy, trademark law, and technological means of protection. The question is whether the combination of these existing sources of protection is sufficient to provide adequate incentives to produce a suitably wide variety of databases.
Much of the time at the meetings was devoted to debating this question. The remainder of this section summarizes the arguments on both sides.
Proponents argue that existing forms of protection are not sufficient. Some forms protect only certain limited aspects of databases, insufficient to reward the investment required to produce them; others are not well-defined and established, or uniform in geographic application. They perceive a gap in protection since the Supreme Court in Feist ruled out copyright for the “sweat of the brow” involved in producing a database, and believe that gap has had real-world negative consequences. They argue that Congress should stay ahead of the curve and prevent more serious harm from occurring.
In particular, proponents describe several cases where database producers have been unable to obtain relief from the courts against substantial, competitive copying.(199) The economic significance of such losses, they assert, has been evidenced by the effect on the producers’ stock prices.(200) They report that piracy has been a problem for others as well, but has not yet led to litigation. They express an understandable reluctance, however, to draw attention to particular databases as possibly uncopyrightable or subject to thin protection.
In addition, proponents assert that uncertainties in U.S. law have begun to affect investment decisions, with producers choosing not to create particularly vulnerable databases, or not to disseminate them broadly, because of a perception that the risks are too great. At least some large database producers in the United States and some European producers have reportedly been unwilling to make their databases available on-line in this country, despite the potential for substantial profit from that form of exploitation. One producer has even decided not to make its print database available to libraries because of a fear of piracy by library patrons.
Opponents, in contrast, view such evidence as insufficient, either because it is isolated, or because it is based on speculation as to future harm. They believe that existing law is adequate, and that the courts are generally drawing appropriate lines between protection and free use. Moreover, they point out that legal protection can be supplemented by technological protection, and argue that database producers should avail themselves fully of their existing options before seeking a legislative solution.
As discussed in section I above, databases are copyrightable subject matter under U.S. law. Proponents value the benefits of copyright protection, and rely on it to the extent possible, but point out that copyright provides only limited protection for databases. While most databases remain copyrightable after Feist (and indeed the Court was careful to state that the white pages directory before it represented an extreme case),(201) it is precisely those databases that require the greatest amount of investment and may be the most valuable to users whose copyright status is most doubtful: the massive, comprehensive database covering the entire universe of a given field, produced in electronic form with the arrangement of the data not fixed by the producer but chosen by each individual user. For such databases, it may be difficult for a court to discern any acts of selection or arrangement on the part of the compiler rising to the level of creative authorship.
One rational response of database producers to Feist’s analysis has been to add as many copyrightable elements as feasible to their databases (whether additional text or creative methods of selection or arrangement).(202) The result of such changes or additions may be to make the information in the databases less easily accessible to users, or less complete. As the capabilities of personal computers and mass marketed software such as search engines increase, consumers are more and more interested in products that offer comprehensive raw data in electronic form for their own selection and arrangement.
A database of meteorological, environmental or medical information, for example, must be comprehensive, accurate, and up-to-date, or the results could be injurious to health or safety. And it is most useful when organized in the most logical, obvious way possible. Subjective selection or a unique arrangement may impede the database’s utility or ease of access. Even where this is not the result, resources are diverted from the task of collecting and disseminating information to the task of satisfying copyrightability thresholds—a diversion that may not be in the best interests of the public, as it is likely to lead to less production of content or higher prices.
At least of equal concern to proponents is the question of the scope of protection for copyrightable databases. In Feist, the Supreme Court made clear that the copyright in a compilation is “thin”—that it will not prevent the copying even of all the material contained in the compilation, if the copier does not take the creative elements of selection, coordination or arrangement that made the compilation copyrightable.(203) In other words, if the copyright in a database is based solely on its unusual arrangement of the data, a competitor may download and reuse all of the data that was collected at great expense, as long as the format is different.
Judicial developments since Feist have augmented rather than allayed these parties’ concerns about the adequacy of copyright protection for databases. They read many of the subsequent cases as applying Feist’s teachings broadly, resulting in an extremely narrow and almost meaningless scope of protection. These cases, they say, hold that very little in any database is protectible, and that virtually everything of value is free for the taking. In particular, they point to the decisions in BAPCO v. Donnelley and Warren Publishing, discussed above in Section I.B., both of which allowed commercial competitors to extract substantial amounts of the contents of expensive-to-produce databases, finding that various acts of selection and arrangement did not meet Feist’s standard of minimal creativity.
Opponents believe that existing copyright law is adequate and appropriate, as supplemented by the other forms of protection described below. While they focused more on other issues during the meetings, some stressed the fact that very few databases have been held unprotectible, and viewed the thin scope of protection provided by the courts to be appropriate in light of the public interest in access to information.
While not discussed much at the meetings, trade secrecy law may also provide protection to certain databases. Compilations of data are one of the types of material that can be protected as trade secrets.(204) Several of the necessary elements of a trade secrecy claim, however, make such a claim unlikely for the typical database. First, the data must not be common knowledge, and must have been kept secret. Disclosure through sale, display, or circulation of goods embodying the compilation, for example, will forfeit trade secret status.(205) As a result, only those databases produced for internal use, and not made available to the public or exploited commercially, will be eligible for protection. Second, a claim for breach of trade secrecy requires a relationship between the owner of the secret and the defendant, involving either a contract or a confidential relationship,(206) or the use of improper me