Jonathan Band and Jonathan S. Gowdy
Morrison & Foerster LLP
D-Lib Magazine, June 1997
Throughout Washington, companies and associations are gearing up for a legislative battle over specialized or sui generis intellectual property protection for databases. The outcome of this battle could have a significant impact on the market for information, which in turn could affect activities dependent on that information, such as research for academic and commercial purposes. The article provides an overview of the roots of the current debate and the arguments for and against some of the recent proposals.
The present debate regarding database protection can be viewed simply as an extension of the historical clash between two conflicting models of copyright protection for compilations. The first model advocates that databases and factual compilations receive protection per se, i.e., without any showing of creativity or original authorship. Proponents of this theory, better known as the "sweat of the brow" or "industrious collection" doctrine, justify their position by arguing that protection should be extended to databases as a reward for the hard work and investment required to compile the facts and information contained in the database. Such a reward provides compilers with the incentive to develop new databases. Under this doctrine, protection extends to the otherwise unprotected facts contained in the compilation.
The second model of intellectual property rejects the notion that databases without any originality or creativity should be protected. Instead, advocates of the second model would only extend copyright protection to the "expression" contained in the database, which is limited to the original selection, coordination, or arrangement of facts in the database -- but not the facts themselves.
Prior to 1991, the extension of copyright protection for databases and other factual compilations remained an unsettled issue in U.S. courts. Most courts refused to grant copyright protection for databases that did not contain any "originality" in the selection or arrangement of facts, and Congress adopted this view in the 1976 Copyright Act. There, Congress explicitly stated that a copyright in a compilation extended only to the original selection, coordination in arrangement of material in the compilation.
Nonetheless, a minority of courts before and after the 1976 Act adopted the "sweat of the brow" doctrine and protected databases that lacked any element of creativity or original expression. In the 1991 case of Feist Publications, Inc. v. Rural Telephone Service Co., Inc., the U.S. Supreme Court resolved the issue that had divided the lower courts and unanimously rejected the "sweat of the brow" or "industrious collection" doctrine. Moreover, even though the Court recognized that the selection and arrangement of facts could create the requisite "originality" for copyright protection, it emphasized that the copyright in the compilation would be "thin," i.e., it would extend to the particular selection or arrangement of facts but not to the facts themselves. Thus, by rejecting the notion that databases could be copyrighted without demonstrating originality and emphasizing that facts and ideas are not copyrightable, the Court appeared to settle the long-standing clash between the two conflicting models of compilation protection.
A. The European Database Directive
Soon after the issuance of the Feist decision, the European Commission began consideration of a directive intended to harmonize the disparate intellectual property treatment of databases throughout the member states of the European Union. Ultimately, in the Spring of 1996, it adopted the Database Directive which included a two tier approach. The top tier provided Feist -like protection-that is, copyright protection for original selection and arrangement of facts in the database. A second tier provided sui generis protection, prohibiting the unfair extraction of a substantial part of a database reflecting significant investment. A database could simultaneously receive both types of protection: copyright protection for the expression -the selection and arrangement of the data; and sui generis protection against the extraction of a qualitatively substantial part of the data itself. The sui generis protection lasts 15 years, while the copyright protection lasts for the life of the author plus 70 years. The Directive requires implementation in the member states of the European Union by the beginning of 1998.
B. H.R. 3531
Shortly after the adoption of the Database Directive in Europe, Congressman Carlos Moorehead, then Chairman of the House Subcommittee on Courts and Intellectual Property, introduced H.R. 3531, "The Database Investment and Intellectual Antipiracy Act of 1996." Modeled on the European Database Directive, H.R. 3531 would also have prohibited the unfair extraction of a substantial part of a database. H.R. 3531, however, had fewer exceptions than the Database Directive and a significantly longer term of protection -- 25 years. Chairman Moorhead held no hearings on H.R. 3531, and it died with the close of the 104th Congress.
C. The Database Treaty
On the same day that Chairman Moorhead introduced H.R. 3531 in Congress, the Commissioner of Patents and Trademarks, Bruce Lehman, proposed a virtually identical text to the World Intellectual Property Organization (WIPO) meeting in Geneva. The WIPO had for five years been considering a variety of proposals for inclusion in intellectual property treaties, and the European Union placed sui generis database protection on the WIPO agenda as soon as it adopted the Database Directive. With Commissioner Lehman's endorsement of a sui generis database proposal, the notion of a standalone database treaty gathered momentum, and in September of 1996, Jukka Liedes, the Chairman of the WIPO Committee of Experts, formally proposed a database treaty as one of three treaties to be considered at the WIPO Diplomatic Conference scheduled to occur in December 1996. (Chairman Liedes also included in the proposed Copyright Treaty a provision requiring copyright protection for databases. This provision generated no controversy.)
As the December Diplomatic Conference approached, numerous parties began to raise serious questions about the Database Treaty. Developing countries objected to the insertion of this new topic so late in the consultative process. They also questioned the need for a new form of intellectual property protection.
Similarly, the scientific community in the United States reacted with alarm to the proposed treaty, arguing that it would stifle research (this argument will be discussed below in greater detail). The science agencies within the U.S. government (e.g., the Environmental Protection Agency and the National Oceanographic and Atmospheric Administration) took the lead in persuading the National Economic Council within the White House to oppose adoption of a database treaty.
Thus, Commissioner Lehman arrived in Geneva in early December 1996 with instructions to stop the very database treaty that he had helped set in motion. In the face of the opposition from the developing countries and the United States, the WIPO governing body decided at the outset of the Diplomatic Conference to defer further consideration of the database treaty. The Diplomatic Conference, accordingly, adopted two other intellectual property treaties-one dealing with copyright and the other dealing with performance rights-but not the database treaty. (It should be noted, however, that the Copyright Treaty's provision for copyright protection for databases was adopted without question.)
The WIPO has already hosted several meetings this year at which the database treaty has been mentioned. Both the U.S. government and the developing counties remain opposed to a database treaty at this time, so it is unlikely to emerge in the near future.
D. Next Steps
Although H.R. 3531 made no progress in the 104th Congress, and the WIPO is unlikely to adopt a database treaty any time soon, the 105th Congress is considering database legislation once again. Senator Hatch has requested the Copyright Office to study the issue, and the Copyright Office has begun to meet with affected parties to gather information. It anticipates issuing its report during the summer of 1997. On the House side, Congressman Coble, the new Chairman of the Intellectual Property Subcommittee, has suggested that he might introduce a database bill for discussion purposes.
Continued Congressional interest in sui generis database protection means that organizations and individuals should determine whether it would help or harm them. A review of the arguments for and against sui generis protection will assist this analysis.
A. Arguments in Favor of Sui Generis Protection
The major proponents of sui generis protection include the Information Industry Association, McGraw Hill, Reed-Elsevier (Lexis-Nexis), and Thomson (West). They have three basic arguments. First, they contend that the large investment in compiling and maintaining databases needs additional protection from the digital, on-line world which makes copying databases as easy as logging on to the worldwide web. Second, the proponents argue that existing copyright law -- particularly the decisions implementing Feist -- provides no protection for large comprehensive on-line databases which are used by means of a search engine. The compiler has exercised no selection because the databases are comprehensive. Further, arrangement only occurs when the user conducts a search. In the absence of selection and arrangement, no copyright protection is available. The proponents also claim that the utility of contract law is uncertain because contracts prohibiting the copyrighting of uncopyrightable material might be preempted by the Copyright Act.
Third, they argue that the EU Database Directive, which requires each member nation to implement sui generis database protection by the beginning of 1998, will provide European companies an advantage in the database market. Because the Directive extends sui generis protection to non-EU entities on a reciprocity rather than national treatment basis, the Directive will deny non-European companies the new legal protection afforded databases unless their home countries offer comparable protection. Without comparable legislation in the U.S., sui generis proponents claim there will be an "open season" on U.S. databases throughout the EU.
B. Arguments Against Sui Generis Protection
Opponents of sui generis database protection have marshaled many arguments against the adoption of a new regime. First, the opponents contend that despite the outcry from some database companies that sui generis protection is necessary to protect their investments in compiling databases, adequate protection currently exists within the present intellectual property law framework. For instance, only a small amount of selection or arrangement is necessary to bring a database within the protection of the copyright laws. Despite the "thinness" of this protection, it is sufficient to protect against wholesale copying.
Moreover, contract, trade secret and unfair competition laws provide an additional layer of protection for databases irrespective of whether the compilation is copyrightable. While each of these doctrines has limitations, recent decisions have affirmed them and together they weave a strong web. Even when the intellectual property laws do not extend any protection to a particular database, the compiler can still protect its investment from copiers through technological measures that prevent copying of the database. These technological measures will only grow more powerful in the future. Finally, for dynamic on-line databases, where the information is updated on an ongoing basis, the compiler has an even simpler remedy -- it can deny the copier access to the database. Dynamic databases derive their value primarily from being current. If a compiler discovers that a customer is redistributing the database without authorization, the compiler can simply discontinue service to the customer. By the next day, the database has changed, and the old database in the customer's possession is worthless.
Second, opponents of sui generis protection reject the argument that harmonization is necessary to prevent European predation of American databases. Although the EU Directive requires implementation by the beginning of 1998, no member country has yet implemented the Directive. In fact, many countries are concerned over how they will implement the Directive since it is directly contradictory to their current law. More importantly, even though the Directive denies sui generis protection for American companies without a similar law in the U.S., it is questionable whether this will lead to an "open season" on U.S. databases. Since U.S. companies will still be able to assert their copyright and contractual rights in the member states, the level of protection for U.S. databases that now exists will continue to exist even after the Directive is implemented. Additionally, U.S. companies may be able to secure sui generis protection if they establish European subsidiaries. (Ironically, one of the leading proponents of sui generis protection in the U.S. is Reed-Elsevier, a Dutch company.)
Third, opponents have asked for concrete examples of market failure-of examples where a database publisher decided not to develop a product out of fear that the product would receive insufficient intellectual property protection. To date, the proponents have provided not even anecdotal evidence. The opponents contend that no evidence has been produced because none exists.
Besides these general objections to sui generis protection for databases, the opponents had a list of specific concerns with H.R. 3531. It is, of course, too early to tell if these same areas of concern will remain with respect to new legislation, if and when it is introduced. Nonetheless, these concerns continue to animate opposition to the concept of sui generis protection.
One of the most pronounced concerns was the broad scope of protection it provided. H.R. 3531 defined a database to include a collection of "work, data or other materials" in any form, i.e., both electronic and non-electronic. Protection was available for any database which was the "result of qualitatively or quantitatively substantial investment of human technical, financial, or other resources in the collection, assembly, verification, organization, or presentation of the database contents." Although it is unclear what level of investment would have triggered protection, this all-inclusive definition of protectable works would go beyond what is commonly thought of as a database and include the full range of copyrighted works such as novels and films. Because the bill permitted cumulation of copyright and sui generis protection, works that would normally be protected under the copyright laws would have been able to receive additional protection under H.R. 3531.
The protection provided for these works under the bill was equally sweeping in nature. For example, the bill prohibited any person without authorization to "extract, use or reuse all or a substantial part. qualitatively or quantitatively, of the contents of the database" or "engage . . . in the systematic extraction, use or reuse of insubstantial parts, qualitatively or quantitatively, of the contents of the database" that conflicts with the owner's normal exploitation of the database or adversely affects the actual or potential market for the database. The legislation did not define "substantial part"; arguably, any amount of data with a small measure of consumer demand is "substantial" enough for protection. Further, if a database is small enough, one fact could constitute a substantial part. The practical effect of these provisions was that in many instances facts themselves would have received de facto intellectual property protection.
Another problem identified in H.R. 3531 was the term of protection. Although protected databases would technically have been protected for only 25 years from the time they were first made available, many databases, especially those in electronic form, would be able to receive perpetual protection because any significant change or updating of the database would result in the creation of a new database with a new 25 year term of protection. This perpetual protection, when combined with the possibility of the protection applying to works now covered by copyright, may have permitted database publishers to circumvent the term limits in the Copyright Act, leading to a drastic diminution of the public domain.
The extensive protection given to databases by H.R. 3531 was made worse, in the eyes of its opponents, by the paucity of exceptions to the restrictions. Unlike works protected under copyright law, databases under H.R. 3531 were not subject to a fair use defense. Proponents of the legislation claimed that the fair use exception was not necessary because the legislation contained an exception for extracting, using or reusing "insubstantial" parts of the database. The opponents responded that the "insubstantial" defense did not offer nearly the same level of protection as that provided under the fair use exception. Furthermore, given the seemingly broad definition of "substantial," it was doubtful that any information worth using would ever be considered "insubstantial."
Databases made by a governmental entity remained in the public domain and were not protected under the bill, but databases prepared by private firms that used information generated or collected by government agencies were not excluded from protection. Since many government organizations already rely on private industry to create government databases, a substantial amount of government information would not have been freely available to the general public. Moreover, as the U.S. government continues to downsize, the outsourcing of the data gathering function would continue, thereby leading to the continual removal of more information from the public domain.
C. Concerns of Specific Communities
Support for sui generis protection appears limited to the Information Industry Association and some of its larger members, including Thomson (West) and Reed-Elsevier. In contrast, many diverse communities have registered concern with sui generis protection, at least in the form of H.R 3531 and the proposed database treaty.
1. Scientists and Researchers
Most forms of research require the use of large amounts of data. Some forms of research, such as that concerning global warming, require systematic use of entire databases. If databases which are now freely available fall under sui generis protection, the cost of research will inevitably increase. Moreover, the culture of science, which involves the sharing of data among institutions, will change as institutions come to treat their own databases as profit centers. Significantly, sui generis protection will raise the cost of research not just for academic institutions, but also for businesses and government agencies which conduct research. For this reason, many professional societies of scientists, engineers and educators (e.g., the National Research Council, the National Academy of Sciences, the National Academy of Medicine and the National Academy of Engineering) strongly oppose sui generis protection. Likewise, most library associations objected to H.R. 3531 and the database treaty.
2. Interoperable Software Developers
Although H.R. 3531 specifically stated that computer programs would not be subject to the Act, it provided that databases contained within computer programs would receive protection. Thus, look-up tables and command sets necessary for interoperability -- precisely the elements found unprotectable in recent software copyright decisions -- could receive protection as a database. Because such protection would hamper software interoperability, the American Committee for Interoperable Systems (including companies such as NCR, 3Com and Sun Microsystems) have objected to sui generis protection.
3. Internet Companies
Routing tables and directories necessary to the functioning of the Internet fall within the definition of a database in H.R. 3531. sui generis database protection, therefore, could lead to the concentration of market power in the Internet. Additionally, the unauthorized transmission of a database could lead to vicarious liability for the on-line service company which unknowingly provided the hardware and software facilities by which the transmission occurred. Several telecommunications companies providing online services, including both RBOCs and long distance carriers, have expressed concern with sui generis protection for these reasons.
4. Value Added Database Publishers
There are many legitimate firms which take data from existing databases and add value to them by inserting new information or arranging the information in a different way. Sui generis protection could put this entire industry out of business. The degree of a publisher's concern often turns on where in the data "food chain" the publisher appears. If it is at the top -- if it generates the data itself (e.g., a sport league) or it has an exclusive license to the data from a government agency (e.g., West) -- then the publisher is more likely to support sui generis protection. If, however, the publisher is further down the food chain, and it must rely on databases compiled by other companies, it is more likely to oppose sui generis protection because its input prices will increase. Groups with members down the food chain, who have opposed sui generis protection, include the Association of Legal Publishers, the Association of Directory Publishers, Donnelly Marketing, Bloomberg, and Fantasy Football Index.
5. Businesses Which Rely on Customer Lists and Other Data
In the Information Age, companies have become increasingly more reliant on information obtained from other sources for the conduct of their own business. This information is as diverse as customer lists, financial data, or gene sequences. Much of this information is already propriety, and businesses pay for it willingly. Sui generis protection will impose an additional layer of protection on this information, and over time will increase its cost. But while the additional protection will increase the cost of the information, it is not clear that there will be any off-setting benefit in the form of additional information generated. Groups which have opposed sui generis or sweat-of-the-brow protection on this basis include the Direct Marketing Association, Haines, the Online Banking Association, and the Biotechnology Industry Association.
No business or organization can afford to ignore the issue of sui generis database protection. Depending on where a company falls in the data food chain -- and almost every business is somewhere in the food chain -- sui generis protection will either add to the bottom line or take away from it. Accordingly, companies would be well advised to study last year's H.R. 3531, and new legislation if introduced, so that they can determine their position on this controversial issue and act to support or oppose it.
1 See, e.g., Miller v. Universal Studios, Inc. 650 F.2d 1365 (5th Cir. 1981); see also Patterson & Joyce, Monopolizing the Law: The Scope of Copyright Protection for Law Reports and Statutory Compilations, 36 UCLA L.Rev. 719 (1989).
2. See, e.g., Leon v. Pacific Telephone and Telegraph Co., 91 F.2d 484 (9th Cir. 1937); Jeweler's Circular Publishing Co. v. Keystone Publishing Co., 281 F. 83 (2d Cir. 1922).
3. 499 U.S. 340, 111 S. Ct. 1282, 113 L. Ed. 2d 358 (1991).
Jonathan Band is a partner and Jonathan Gowdy is an associate in the Washington, D.C. office of the law firm of Morrison & Foerster LLP.