As usually defined, “databases” include numerical data, text, images, and any other “organized collection of information.” Because enormous numbers of products fit this description, it is sometimes hard to think about such apparently straightforward questions as, “Is existing legal protection adequate?” or, “Could it be improved?” This paper tries to make matters more concrete by examining existing databases and how they are produced. The results are then used as a benchmark to evaluate potential legislation. Special attention is paid to features and problems that set scientific/technology databases apart from other products.

The world of scientific and technology databases is already extremely rich and well-developed. Since the U.S. government has never enacted database legislation, this presents a paradox: If existing databases can be freely copied, why do firms continue to invest in them? The answer is that database providers have devised a bewildering number of unofficial (“self-help”) methods for protecting their investments. These include but are not limited to (1) bilateral agreements with users, (2) “shrink-wrap” or “click-wrap” language, (3) bundling with copyrighted materials, (4) continual updating and improvement that leaves would-be copiers “out of date, ” (5) search-only Web sites where the underlying database cannot be downloaded, and (6) passwords and encryption. The fact that rich and diverse databases exist in today's world shows that such protection can be extremely robust. At the same time, self-help strategies may cause undesirable distortions in the economy, particularly when they discourage database suppliers from sharing products with a wider audience. Even more insidious, lack of statutory protection may mean that some databases are never created in the first place.

Scientific and technology databases present unique needs and problems. These include

• The need to assure private firms that they can profitably invest in commercializing and extending government databases for use by a broader audience;

• The need to keep database prices within the reach of academic users, who have traditionally driven most advances in basic knowledge;

• The scientific community's need for value-added or edited databases that not only collect but also update, cross-check, comment on, and try to reconcile reported results;

• The fact that virtually all scientific databases have historically been created by combining and extending earlier data sets; and

• The scientific community's need for full and unrestricted access to data, which inevitably conflicts with self-help strategies based on secrecy or partial disclosure.

The modern history of database reform begins with the U.S. Supreme Court's 1991 decision in Feist Publications, Inc. v. Rural Telephone Service Co., which restricted “sweat-of-the-brow” protection under copyright in the United States. This was followed by the European Union's (E.U.) 1996 Directive on Databases, which required member countries to expand their statutory protection of databases. The E.U. Directive also contained a controversial threat that citizens of countries (including the United States) that did not adopt E.U.-style statutes would not be protected by the new laws when they took effect. Because of the E.U. Directive, the U.S. Congress introduced European-style legislation in 1996 and again in 1997-1998. Scholars have also suggested alternatives to the European model.