Clarisa Long*

Biomedical research is undergoing a paradigm shift from gene-centric biology to genome-centric biology that changes the way we think about biological innovation, genetic information, and the way biologists do their work. The shift to genomic-centric biology is creating a hybrid model of discovery and innovation in which the creation of new ideas and products remains the dominant strand but the supply of new inputs such as databases and software are becoming increasingly important to the process of innovation.

A genome is the complete set of an organism's genetic material, found in the nucleus of each cell. It contains the master blueprint for all cellular structures and activities for that organism. Whereas until recently, biomedical research was limited to examining single genes, technology is now allowing scientists to study entire genomes. The Human Genome Project, which aims to map and sequence the entire human genome by 2003, is one example of how the genomic revolution is accelerating the pace and redirecting the nature of biomedical innovation.

The genetic information being derived from research on the human genome and the genomes of other organisms is extremely valuable, both for basic research and for commercial purposes, even if few commercial products have resulted. This information is critical for cost-effectively creating accurate diagnostics for many inherited diseases, for developing animal models for human disease research, and for facilitating our understanding of gene function in health and disease.

Thus, the core business of an increasing number of new market entrants in genomics is information, not the sale of drugs or diagnostics. The business model of many genomics companies starkly contrasts with those of integrated pharmaceutical firms or first generation biotechnology companies. The entrance of a host of gene information firms, such as Human Genome Sciences, Incyte, and Sequana, along with new entrants in other parts of the discovery pipeline, has shifted the market structure and upset many of the traditional economic assumptions about innovation in new drugs. These new entrants are less capital intensive, exhibit much faster time to market, and offer different risk-reward models for investors.

The growing intersection of this new genomics framework with intellectual property rights already is profoundly reshaping the balance struck among the interests of biomedical research, private sector market participants, and the public good. The policy challenges presented by a genomic-centered approach to biology in which information rather than a tangible commodity is the crown jewel greatly complicate the debate about the related questions of who owns what information, and who can access it on what terms. This trend blurs further the faint, erratic dividing line between the public domain and proprietary interests.

The application of advanced computing power to genomic data and information is becoming a driving force in both biomedical research and the development of new genomic products and services. New information technologies have created tools and methods that are transforming the research enterprise and becoming central to the process of innovation. Bioinformatics and combinatorial chemistry are revolutionizing what constitutes genetic research and how innovation takes place. Most significantly, genomic information and research tools based on information technologies -- not only the products derived from their use -- have themselves become marketable.

The first big problem, therefore, is assuring a sufficient supply of genomic information through research. Here a dilemma arises. On the one hand, intellectual property rights provide a necessary incentive to innovate. On the other hand, economists have found that excessive protection early in the innovation process can impede innovation at later stages. Under such circumstances, increasing or expanding intellectual property protection does not necessarily lead to greater amounts of socially beneficial innovation. In other words, without legal protection not enough information will be produced but with legal protection not enough of the information will be used.

The second overarching problem, often overshadowed by the initial issue of producing enough genomics knowledge, involves the diffusion or distribution of genomic information. A state of the world in which the innovations that result from genomic incentives are not widely disseminated may be less socially beneficial than one where there are fewer innovations but they are distributed more broadly. The problem manifests itself most acutely in trying to allocate fair compensation to the creators of valuable information assets, including databases, while assuring that other stakeholders have access to information, as the intellectual property incentives also have been designed to provide.

Intellectual property rights do not operate in a vacuum. They influence and are influenced by the character of the technology and the structure of scientific research and innovation in that sector. Below, I highlight a few context conditions that need to be considered in assessing the likely effects, positive and negative, of strengthening intellectual property protection for genomic information.

Genomics is a "cumulative systems technology"

Genomic research is a "cumulative systems technology" marked by an interactive and cumulative process of discovery and innovation that builds on an accumulated stock of genetic knowledge developed from numerous sources. The manipulation and recombination of existing knowledge by multiple researchers constitutes an essential part of genomics today. Knowledge creation is not only more interactive than in the past, but also more collective. Like multimedia, it builds on a broad range of research and interacts with multiple innovations to create an entirely new technology. Many of these follow-on creations generate social benefits comparable to or surpassing the initial discovery.

Genomic innovation is also characterized by its growing divergence from either the single inventor/absolute novelty approach to innovation that continues to dominate patent law or the individual author/original expression model that continues to guide copyright. Many of the traditional presumptions underlying the discrete invention or expression models do not apply to cumulative systems technology such as genomics. Innovation is more dependent on the timely and efficient distribution of information than ever before. The initial discovery usually is only the starting point because it often opens up a broad range of follow-on improvements, many of which are not immediately foreseeable. These trends make it increasingly difficult to separate the innovation from earlier or parallel advances, to identify who really deserves credit as the initial inventor or creator, and to gauge very precisely which incentives have what effect.

These concerns are particularly acute in trying to decide whether databases deserve additional protection. A major tension exists in trying to reconcile the incentives needed to obtain the creation and early release of these valuable works with those that will permit and facilitate the creation of subsequent innovations through use of the database and its contents.

Unlike many other innovative sectors, the nature of genomics innovation requires "systematic access to the state of the art."(1) There is a legitimate concern that strong protection of specific discoveries too early in the evolution of genomics will retard the pace of overall innovation, or redirect genetic research in less beneficial directions. The semiconductor industry in its formative years was marked by similar rapid, multidirectional progress in both the underlying basic research and the cumulative technology that grew out of it. This highly beneficial burst of scientific, technological, and economic advancement with broad social consequences, only partially glimpsed at the time, would not have been possible in a legal regime that strongly protected intellectual property rights in many of the early innovations.

Appropriating value from genomic innovation

With the shortening of innovation cycles and the increasing cost of innovation, creating adequate incentives is essential for developing the next generation of research, products, and services that will ensure international competitiveness and economic growth. Basic research and new technologies often fail to attract sufficient private sector investment because the results are not subject to exclusive ownership -- in other words, they have limited appropriability. Market failure in databases, for example, occurs not because the work is copied but because the cost of copying the information using new digital technologies is less than the investment in developing the database that the developer must recoup.

The race to patent or create a cognizable right also may have desirable effects by accelerating the rate of innovation. Competition between genomic information companies to identify and patent gene sequences, by speeding up new discoveries, has created spillover benefits for basic research.

On the other hand, the creation of new ideas and products may be retarded by stronger protection in certain industries for a number of reasons. The promise of strong intellectual property protection may result in redundant or duplicative research in the race to capture the prize of monopoly rights.(2) It also may deter research if the field is considered too crowded. Also, excessive incentives may lead to the premature exploitation of new knowledge. Scientific research can be analogized to fishing, where the property rights go to the person who catches the fish.(3) A fisherman can wait for the fish to grow larger but risks losing it if someone else catches it first. This leads to too many little fish being caught. At the moment genetic research is yielding many little fish.

The effects of transaction costs

The markets for genomic information are subject to a number of significant transaction costs. This introduces another major consideration in deciding whether the incentives created by strong intellectual property rights for genomic information will produce additional innovation at acceptable costs.

Without strong rights, groups that develop new innovations may limit their dealings to a few customers with whom they have strong contractual relations. Strong intellectual property protection makes it possible to disseminate innovations more broadly. The intellectual property right reduces the transaction costs of dealing with numerous buyers on an individual basis through contracts and, therefore, makes possible many more transactions in the information. The incentives to innovate from strong protection facilitate the sale of rights and thereby reduce the transition costs of licensing. Only by protecting intellectual property rights can there be an efficient market for trade in genomic information.

On the other hand, genomics information such as that compiled in a complex database may not be able to be shared efficiently by contract with all the users who wish to access it. Strong protection for an expressed gene sequence or a database arguably may create such high transaction costs that it impedes rapid diffusion of timely information to both consumers and to researchers who could use the information to improve on it, or fit it together with other information, for significant follow-on discoveries and insights. Also, broad rights covering research tools rather than marketable products are likely to promote cross-licensing arrangements that can be used to deter entry to the network by other users.(4)

From a legal perspective, granting private property rights to genomics information may increase transaction costs because researchers and others users confront the problem of determining who owns what, how to secure rights in a timely and efficient way, and at what price. Presumably, many of the transaction costs can be avoided by the use of collectives similar to those used in other contexts such as ASCAP and BMI in the recording and multimedia sectors.

In genomics, the licensing issues will be further complicated because the industry is comprised of a changing collection of many different industries and institutions, public and private, each of which has its own informal norms and standards about licensing intellectual property rights. The likelihood of finding common ground may prove difficult.

Litigation costs also are likely to impose significant transaction costs in genomics. The courts and Congress will need years to settle many of the incipient issues. Uncertainty about ownership rights and access to genomic information during the period in which the legal and political system tries to catch up to fast-moving and novel technological changes in genomics also can have a chilling effect on innovation. In the context of another dynamic cumulative technology -- computer software -- the Office of Technology Assessment concluded that: "[u]ncertainty about ownership of a component or the scope of intellectual property rights could discourage the development of programs composed of components from different sources."(5)

Some specific characteristics of the genomics industry

The effects of intellectual property incentives on market behavior depend on the specific characteristics of the genomics market and its underlying technologies. Patents have proved to be a particularly effective and necessary tool in encouraging innovation in the pharmaceutical and first generation biotechnology industries. These sectors are characterized by enormous costs in research and in obtaining regulatory approvals, by unusually high degrees of risk and uncertainty in bringing products to market, and by final products that can be imitated at very low cost. Innovation in the drug industry frequently has been found to be particularly dependent on strong patent rights because it is the only effective way to permit firms to appropriate sufficient returns from their investment. The benefits of research are captured largely in its final use value by the firm responsible for the development.

The pharmaceutical and biotechnology industries also have depended almost exclusively on the discovery of new knowledge and the development of completely new products. As a result, patents tended to correspond to only one product. Until now, this innovation model contrasted sharply with most sectors where innovation focuses more on incremental product improvements or advances in system design.

Technologies, such as genomics, that are heavily dependent on basic research have a set of characteristics that strongly affect the nature and type of protection that should be provided. They tend to have multiple researchers with the same base of knowledge who tend to perceive the same opportunities and, therefore, are more inclined to race -- whether it is a race to patent or to publish, or both. In the next few years, biomedical advances and new drug products will depend as much on incremental, ongoing improvements in databases or redesigns of existing software systems as on blockbuster breakthroughs. Also, innovation will depend to an unprecedented degree on new relationships to genomic information that are dynamic and not static or fixed like most information we are accustomed to dealing with.

Another important force affecting the role of intellectual property rights in genomics information is the need to attract private investment capital. Strong intellectual property rights attract investment capital that start-ups need to enter and compete. This is particularly the case with genomic companies which have uncertain market prospects and often unproven management teams led by former academic stars with no substantial business background. The only tangible assets that venture capitalists or early-stage institutional investors can use to value their potential investments in companies are the economic returns made possible by the company's potential intellectual property rights.

Finally, a broad consensus exists that intellectual property rights are most effective in creating the right incentives as new product developments near the market.(6) Conversely, rights granted on basic research at precompetitive stages are problematical. Depending on the circumstances, they can provide the basis for the orderly development of the full range of possibilities or they can preempt broad areas of future research, or direct it to less beneficial areas of inquiry. As a result, there are concerns about granting patents or other rights to discoveries or creations which occur relatively early in the research process long before a market exists for a product or process.

Moreover, the line separating precompetitive and commercial development in genomics has blurred with the compression of time and the changing nature of the innovation process. It becomes progressively more difficult to agree on what constitutes precompetitive research as opposed to commercial development. The various phases of research and development no longer fall into tidy logical and highly ordered sequences, nor do they necessarily proceed in any linear fashion.

The traditional bifurcations along which the production and ownership of genetic research have divided –basic /applied research, academia/industry, and public/private sector -- are blurring in the field of genetic research. The Human Genome Project has evolved into a joint public and private effort, academics are coordinating their efforts with commercial genomic firms, and government routinely collaborates with the private sector. As genetic research progresses and information is produced in new ways, new legal issues arise, the most fundamental of which is: Can the intellectual property system continue to maintain a delicate balance between information production and dis-semination? Can the legal system create reasonably clear, bright-line tests determining whether the results of genetic research are subject to private rights of protection? In the absence of bright line rules, how will the scientific community strike the production/dissemination balance?

Shifts in U.S. science and technology policy reflect the powerful effects attributable to the globalization of technology and markets. The United States' biggest successes in international markets have come in those products and services in which U.S. firms have been vigorous technological innovators. American pharmaceutical, biotech- nology, and information service firms are three of those world leaders. Looking forward, the most successful economies will be those that can best harness their brain power to generate economic growth.

*Clarisa Long is a Research Fellow at Harvard University, the Abramson Fellow at the American Enterprise Institute for Public Policy Research in Washington, D.C., and a Vice Chairman of the Federalist Society Intellectual Property Practice Group.

1. Organization for Economic Cooperation and Development, Production and Distribution of Knowledge in the New Systems of Innovation: The Role of Intellectual Property Rights (1996).
2. See Patha Dasgupta & Joseph Stiglitz, "Uncertainty, Industrial Structure and the Speed of R&D," 11 Bell J. Econ. 1 (1980).
3. See Patha Dasgupta & Yoram Barzel, "Optimal Timing of Innovations," 50 Rev. Econ. & Stat. 348 (1968).
4. John Barton, "Patent Scope in Biotechnology," 26 ITC 605 (1995).
5. Office of Technology Assessment, Pharmaceutical R&D: Costs, Risks and Rewards 154 (1993).
6. Rebecca Eisenberg, "Proprietary Rights and the Norms of Science in Biotechnology Research." 97 Yale L.J. 177 (1987).