Association of Research Libraries (ARL®)

Peter Jaszi
Washington College of Law

Librarians have been enthusiastic early adopters of digital information technology--and it is clear that the library community and the larger public have benefited as a result. My message today, however, is that, in at least one respect, new technology is a source of danger rather than promise in that its introduction is being taken as an occasion to rewrite some of the most fundamental rules of American Intellectual Property (IP) law--to the benefit of the information industries and the detriment of information consumers, including library patrons. Indeed, we face today unparalleled threats to the continued existence of a commons in information!

Looking out into the audience, I am pleased to see the familiar faces of several decorated of last year's legislative campaigns. At the same time, I'm somewhat daunted, since I know that some of you will already have been subjected to one of my harangues on disturbing trends in late 20th century American IP law. But so be it.

I'm reminded, in this connection, of a story about the early 20th century Anglo--Irish barrister, Sergeant Sullivan (the last Sergeant--at--Law who distinguished himself in his defense of Roger Casement). Once asked by an impatient judge to summarize his best argument on legal point at issue. He did so, and the tetchy jurist responded: "Well, I don't think much of it." To which Sullivan rejoined: "It'll grow on you m'lud, it'll grow on you."

As, I hope, may some of the points I'm going to make today about the topic of "database protection"--perhaps the most consequential issue left unresolved when a weary Congress wrapped up its work on the somewhat grandiosely designated DMCA last fall.

I'd like to start with two historical propositions. First, for more than two hundred years, with a few trivial exceptions, copyright law has been the exclusive mechanism by which United States federal law has balanced proprietary claims in information with user interests in access to it. Second, it is a long--standing principle of copyright that there are no rights in factual data as such, even when someone has gone to trouble and expense to collect or compile them. I want to stress that this principle is truly a venerable one. The Supreme Court did not, as has sometimes been suggested, invent it in its 1991 Feist decision. Rather, that decision confirmed what had been the clear trend of U.S. copyright law for all or most of the 20th Century.

Taken together, the traditional exclusivity of copyright law and the settled principle that the contents of factual compilations (as distinct from the manner of their presentation) aren't protected by copyright have meant that facts have been free for use. Regardless of their source, they have been considered available--as far as IP law is concerned--without any requirement of permission or payment, whether for academic, scholarly, and educational purposes or for commercial ones.

This state of affairs, I would suggest, is no doctrinal quirk or historical accident. Rather, it represents the instantiation, in the domain of legal rules, of what has been a broadly shared vision about the social value of information--that the existence of a robust public commons of primary data is essential to our collective cultural, educational, and scientific enterprises, and that enclosures of that commons, though they may be yield short--term benefits for some, ultimately will redound to the detriment of all.

My unhappy task today is to confirm that this shared vision is currently under siege, and that in the current session of Congress it is (I would estimate) more likely than not that legislation will be enacted which will fundamentally alter the legal position of facts, and will, by commodifying data, transform our information use practices in profound and deleterious ways.

The legislation to which I refer is H.R. 354, the "Collections of Information Antipiracy Act," now pending in the House of Representatives--a bill strongly supported by large information conglomerates like Reed Elsevier and Thompson, and also by firms with more specialized interests in data ownership, such as the New York Stock Exchange. Please note that H.R. 354 is not--repeat not--copyright legislation. Nor, despite the claims of its proponents, is it merely an extension of familiar common law principles of "misappropriation" law. Rather, it represents an effort to devise what would be, in effect, a new form of intellectual property protection, applicable to subject matter which is (according to the Supreme Court) beyond the reach of copyright itself as a constitutional matter.

Before discussing the provisions that are at the heart of the bill--those which relate to the scope of protection it proposes--some other features of the legislation are worth noting:

• Although H.R. 354 is characterized as part of the content proprietors "digital agenda" of IP law reform, its coverage would not be limited to compilations in digital formats, or to the uses of information in the network environment; instead, it would apply with equal force to print--on--paper collections of information;

• H.R. 354's definitions are so broadly drawn that many kinds of information products would be swept within its scope--statistical compilations, directories, and electronic "databases" qualify as "collections of information," but so do charts and tables, anthologies of copyrighted works, volumes of academic journals, and much, much more.

• The protection H.R. 354 would afford to compiled information would be, in many cases, effectively perpetual. The bill's notional 15--year limit on the term of protection for items of data is both legally ambiguous and likely to be practically ineffective (at least where dynamic electronic databases are concerned).

• Protection under H.R. 354 would be retroactive, sweeping in compilations prepared during the 15--year period preceding enactment.

What, then, is the nature of this "protection"? H.R. 354's central provision--the prohibition around which the rest of the bill is organized--is well worth a dramatic reading:

Any person who extracts, or uses in commerce, all or a substantial part, measured either quantitatively or qualitatively, of a collection of information gathered, organized, or maintained by another person through the investment of substantial monetary or other resources, so as to cause harm to the actual or potential market of that other person, or a successor in interest of that other person, for a product or service that incorporates that collection of information and is offered or intended to be offered for sale or otherwise in commerce by that other person, or a successor in interest of that person, shall be liable to that person or successor in interest for the remedies set forth in section 1406.

**Note the many uses of Orand Either -- inescapable signs of expansive tendencies in legislative drafting!

In plain English, then, H.R. 354 would provide a legal regime in which items of data themselves would be legally protected against a wide range of uses not authorized by their compiler. Moreover, proponents of the bill have made it clear that the "harm to the market" referred to in the quoted passage is not limited to competitive injuries, but includes any loss of potential revenue from the licensing of data use. In other words, one could violate the prohibition merely by using without payment, in one's research or writing, information for the use of which a data provider would prefer to charge.

H.R. 354 has a long, if not particularly distinguished pedigree--the language I've just read marks it as the direct descendant of the first major legislative experiment in the field of database protection: the EU Database Directive of 1996. In fact, proponents of the legislation argue now that a principal benefit of its enactment would be to give U.S.--made databases protection in the countries of the European Community, although they are less clear about what concrete benefits such protection would bring.

Since 1996, there have been a series of bills in Congress--H.R. 3531 in the 104th and H.R. 2652/S. 2991 in the 105th--which followed the approach of the EU Directive: incorporating a broadly--defined prohibition against unauthorized "extraction" or "use" of data from proprietary compilations, subject to certain narrowly defined exceptions. As database legislation has evolved, the terms in which these exceptions have been defined have grown relatively--and I stress relatively--more generous in an attempt to reconcile the various groups that have been vocal in their opposition to broad new IP rights in data: schools, libraries, archives, scientific associations, other learned societies, commercial producers of value--added databases, and many more.

Even in its most evolved forms to date, however, these domestic proposals for database legislation are remarkably grudging in the scope of the exceptions they recognize--H.R. 354, for example, includes nothing that remotely begins to approach the "fair use" doctrine of copyright law as a safeguard for users' interests.

To illustrate, consider H.R. 354's treatment of educational, scientific, and research interests in access to data. The sorts of uses teachers, scientists, and scholars make of data, we are told, will be permissible so long as they do not "harm directly the actual market" for the database in question. The problem, of course, is that where the research community are the intended consumers of a particular commercial factual compilation--of whatever kind--any unauthorized or uncompensated research use of data from that compilation will, by definition, harm the market for it! Elsewhere, H.R. 354 immunizes from liability certain other passive uses of information: "an individual act." If that isn't restricted enough, however, we are told that this immunity from liability applies only if the "individual act" was not "part of a pattern, system, or repeated practice ... with respect to the same collection of information." In effect, this is a "one bite at the apple rule" which leaves researchers who wish to make repeated references to material from a particular compilation without legal recourse.

Let me be clear: These exceptions are not nothing, but (on close examination) they really do not amount to much. Under the regime of H.R. 354, researchers engaged with primary data from non--public sources (including public data which is offered by private vendors) will face a difficult practical choice: to comply with the emergent regime of metered pay--per--use information consumption that the new legislation inevitably would help to usher in and adopt more parsimonious practices of data use to help keep the bills within reasonable limits or to risk exposure to swinging civil (and in some cases criminal) penalties.

And, I would add, H.R. 354 suffers from another significant measure--it includes no provisions which would deal effectively with the problem of assuring access on fair terms to the contents of "sole source" databases--including "de facto sole source" commercial databases consisting in significant part of government information.

Having said this, I am willing to concede that, in fact, there may be a need for some database protection--protection, that is, against unfair and parasitic competitive practices such as the wholesale stripping of the contents of one commercial database to create a new product that will compete with it in the marketplace. Indeed, some of the entities represented in this room today are ones which might benefit from such a limited new legal regime. But, I want to emphasize, this is not the model which is now before the Congress in H.R. 354.

Happily, the outlines of such an alternative do exist--and there may still be time to redirect the congressional momentum for database protection towards this alternative.

In an important statement of January 19, Sen. Hatch (R--UT), chair of the Senate Judiciary Committee (which has jurisdiction over IP issues) set out various different "models" of database legislation. One of them, the so--called "Database Fair Competition and Research Promotion Act of 1999," would take an approach to the problem which is, in essence, the inverse of the approach of H.R. 354. Where H.R. 354 contains a broad prohibition against unauthorized "extraction" and "use"--which must then be qualified by various exceptions and limitations to achieve some semblance of balance--the alternative legislation would define prohibited conduct narrowly, to reach only certain forms of unfair commercial competition, and thus putting most of the enterprises with which we are concerned in the academic, research, and scientific communities beyond the reach of the prohibition, as matter of definition. It addresses the most pressing concerns of those who deal commercially in compiled data, while leaving the interests of most end users of information unaffected.

To date, of course, this alternative legislation is only a hypothetical possibility--it is has not yet been introduced in either House. But the concept it represents is beginning to attract interest from some members of Congress. It can succeed, however, only if it receives massive support from all the communities whose interests would be adversely affected by enactment of H.R. 354. [CITE to databasedata.org.] The communities represented in this room are beginning to be heard--and I hope their voices will grow every more prominent in the months to come.

Thank you.

Copyright © 1999 by Peter Jaszi