Page 129

Part 3—
From Laboratory To Marketplace:
The Challenges Of Technology Transfer

Technology transfer, the issue of how to translate scientific discoveries into useful, commercially viable products, is a focal point in the debate surrounding biotechnology. However, as Jerome Schultz, of the Center for Biotechnology and Bioengineering at the University of Pittsburgh, points out in Chapter 10, the issue is neither new nor unique to biotechnology.

In a historical review of relationships between universities (where most basic scientific research has traditionally been carried out) and industry (which tends to favor research oriented toward product development), Schultz notes that cooperation, as well as tension, between the two communities extends back to the earliest days of university involvement in research at the beginning of this century. Schultz argues that the explosion of government support for university-based research after World War II weakened university-industry collaborations because it became easier for academic scientists to apply for government grants than to build the relationships necessary to obtain funding from industry.

Both Schultz and National Science Foundation director Neal Lane, author of Chapter 11, argue that although there are indeed significant differences in the cultures of the academic and business communities, the extent of the gulf between the two groups has been exaggerated. Far from being ivory towers interested solely in the pursuit of knowledge for its own sake, Schultz and Lane say, universities have long recognized the need to connect ''the discovery process" (as Lane calls it) with "the process of putting new discoveries to use."

Page 130

Nevertheless, conflicts of interest are an inherent part of the research system, according to José Trías, vice president and general counsel of the Howard Hughes Medical Institute (HHMI) until his tragic death in May 1994. For example, where scientists value the free exchange of information because it promotes the advance of knowledge, business people see a need to protect product-related information to maintain an advantage over their competitors. In Chapter 12, Trías describes the approach taken by HHMI (the nation's largest private nonprofit supporter of biomedical research) to achieving a balance between the valid but often divergent interests of the academic and business communities.

Since 1980, when Congress passed the Bayh-Dole Act (P.L. 96-517) allowing universities to hold patents on discoveries made through government-sponsored research, the U.S. government has encouraged patenting to promote the commercial development of scientific discoveries. However, in Chapter 13, Rebecca Eisenberg of the University of Michigan argues that although patents may promote technology transfer in some circumstances, the adoption of a single approach to technology transfer fails to fully take into account the complexity and unpredictability of the scientific discovery process.

Page 131

10—
Interactions Between
Universities and Industry

JEROME SCHULTZ

A recent survey of graduate deans revealed that one of the most critical issues facing their graduate programs is the question of university-industry interactions; these concerns are summarized in Table 10-1 (Morgan et al., 1993). There is a partial myth that a university is inherently an open institution, that all information generated at a university should be freely available to everybody. Because of this prevailing presumption, it is not unusual to hear comments to the effect that anything that sequesters information at a university is against the public's interest and therefore "bad."

Another perceived problem is that because of funding pressures, the research agenda of universities may be biased by industrial support to focus on "hot areas," which may be different from intellectually important problems. In the past 10 years, biotechnology, computers, and materials science have been some of the frontier areas in which industry has been supporting research. Sometimes ignored in these discussions is the pressure by other institutions, such as, groups in Congress, to prioritize the research portfolio of universities to meet perceived national needs. An example of this is Senator Mikulski's (1994) new emphasis on strategic funding: "… we should be spending more than half of our basic research dollars in areas we consider strategic. Our investments in science and science policy will become a new super highway of ideas and technology to achieve national goals."

Another area of concern is whether industry-related research efforts inevitably lead to the exploitation of students. The presumption is that

Page 132

students working on projects that are generated by industry would be bound by contractual considerations, thus preventing them from pursuing their own creative research ideas. Does their educational experience suffer because of these factors?

Industry also has concerns. One is the perception that university faculty prefer to do curiosity-driven research that is often very far removed from marketable products. A related concern is that academic researchers drift from stated goals of their research programs to follow interesting leads with the hope of being first to uncover and publish on a new phenomenon.

Because much of university research is carried out by graduate students who have other obligations, such as course work, another problem is the difficulty that academic groups have in meeting project objectives in a timely fashion. Companies fear that their personnel may have to expend extraordinary efforts and time with their university colleagues to make sure that a joint project is moving along expeditiously.

Industry is also concerned with the leakage of proprietary information developed within the university's laboratories. It is in the company's interest to maintain certain information as proprietary, either for a patent application or for production purposes.

There are potential problems as seen from both sides in these interactions. For successful interactions it is absolutely necessary for both groups to be aware of these problems and to put procedures in place at the start of a project to prevent interferences in a productive relationship.

Many discussions of university-industry interactions tend to be colored by the patterns of a comparatively few "research" universities that dominate these scientific activities of the academic community. There are about 1500 colleges and universities in the United States. About 500 of them have graduate-level science departments but only about 200 are

Page 133

doctorate-granting institutions. Yet when one examines the size of the research faculty and the amount of research funding, about 100 schools qualify as research universities.

Research funding is concentrated in a relatively few universities. Eighty-five percent of the research funding—about $13 billion in 1992—goes to the 100 research universities, and 21 percent of that funding ($3 billion) goes to 10 universities. The focus here will be on these 10 universities; some of the research paradigms appropriate for these large institutions may not apply to others. Also, there are signs that even the 100 research universities may not be able to sustain their research activities over the next decade because of the need to replace and replenish existing facilities and equipment in the face of leaner federal budgets.

There is extensive diversity in research programs among universities. With the ever-present competition for all sources of funds—federal, state, industry, foundations—each institution seeks to establish its competitive advantage so as to distinguish itself from it colleagues and to be more attractive to supporting agencies.

Distribution Of University Revenue

University-industry relations should be considered in the context of the budget of a typical university. For example, the University of Pittsburgh, a modest-size university, has many sources of funding (Figure 10-1). The size of these income streams indicates the potential effect of these factors on the pattern of university activities. It is not unusual that tuition and fees do not account for most funds needed to run a university. In this instance the research sector accounts for about 20 percent of the university's income, so although important, it is not the dominant factor in university revenue. Most discussions of university finances focus on tuition, state funds, and research grants as sources of revenue, but as can be seen for the University of Pittsburgh, roughly 30 percent of the money comes from other sources. Some universities are involved in auxiliary enterprises (e.g., real estate businesses and educational and other service facilities).

This brief overview illustrates that universities are business enterprises engaged in a variety of businesses. Although this discussion focuses on research, the importance of university-industry interactions should be considered as only one, most likely minor, component of the overall university budget. Averaged across all academia, present industrial support accounts for only about 8 percent of the total research expenditures at universities.

An additional component of university finances, not included in Figure 10-1, is the operation of the university medical center. Most research

Page 134

FIGURE 10-1 Typical distribution of university revenue.

universities have an associated hospital and medical center. The annual budget of these operations is usually the same size as that for the total academic enterprise and often is much larger. The health-related academic and service budget may be the dominant budgetary item in many universities and can have a tremendous effect on the nonhealth academic programs. Surprisingly, there is very little discussion in academic circles on the influence of the health-service enterprise on university planning and resources whereas industry-related issues are hotly debated.

A History Of University Research

Initially, universities in the United States did not engage in any research. Universities were established basically for general education, i.e., language, religion, arts, and humanities; they were not established for research. As best as I can determine, university research in the United States started in the 1850s when land grant universities were established by Congress in response to society's need to develop technologies useful for agriculture and industry. This tradition of universities responding to society's needs continues today and must be appreciated to understand the evolution of the research agenda of universities.

Page 135

Until the early 1900s the research activities of the land grant universities were slanted toward agriculture rather than industry. This neglect of industry-related research for about 50 years probably stemmed from the lack of funds from state governments to purchase and maintain the equipment necessary to set up industry-like manufacturing units. Universities could not afford the production or pilot plant facilities necessary for these types of efforts. The focus instead was on classroom teaching; the practical training of students was left to industrial employers to provide on the job. Hands-on experience that is required for a position with significant responsibility in the production aspects of industry is still not provided at school. Companies usually provide about 2 years of training in technology to an engineering graduate with a bachelor's degree in science. So although the intent of land grant legislation was to foster industry-university collaboration, the interactions with farmers prospered through the formation of university-based agricultural experiment stations.

The debate and conflict that consistently arises today in discussions on the meaning and value of basic vs. applied research goes back almost 100 years, and it is unlikely that any new cogent ideas are going to surface any time soon. The Massachusetts Institute of Technology (MIT) was one of the first universities to initiate research laboratories in the sciences in the early 1900s and also was the first university to experience the stress generated by divergent views on basic vs. applied research. In 1903 A. A. Noyes introduced the concept of providing research experience to broaden the training of students in university laboratories. He solicited and received foundation grants for the Research Laboratory of Physical Chemistry for conducting research within the chemistry department. W. Walker in the same department apparently considered the research carried out by Noyes as ivory tower research, not truly related to industrial needs. So he established another research laboratory—the Research Laboratory for Applied Chemistry—and obtained money from industry to study research problems. This might have been the first "job shop" in a university.

The conflict between Noyes and Walker was so strong that Noyes eventually left MIT and went to the California Institute of Technology, which marked the beginning of the chemistry department there. G. N. Lewis also left MIT's chemistry department and started the research activities at the University of California, Berkeley. The Research Laboratory of Physical Chemistry at MIT was disbanded in the 1930s and replaced by the Division of Industrial Cooperation, a subscription service that gave companies access to faculty resources. Disputes regarding the quality of research and loss of subscribers led to the guideline that sponsored research "must enhance the prestige of the institution, vitalize teaching, and provide helpful contacts for students."

Page 136

MIT eventually handled this situation by sanctioning faculty consulting with industry, and this policy has been adopted by most universities. The presumption is that through consulting, faculty members learn about industrial needs and bring this appreciation of industrial practice to students through the university's education program.

National Research Council

World War I was the next event that significantly affected the university research agenda. After the war, Vannervar Bush realized that universities were a resource for fundamental science, and the National Research Council was formed to coordinate federal needs with university research. This initiated the era of federally funded research, and since then the government has played the leading role in university research.

On the entry of the government into research funding at universities, a fundamental shift in the paradigm for the selection criteria for the recipients occurred. Whereas previously private foundations decided research support based more on the general reputation of the institutions and individuals, the National Research Council adopted the policy of supporting the best-qualified institutions and individuals. This seemingly innocent and reasonable modification of the selection process was the start of the peer-review concept in research funding, which has become an institution in its own right and has grown to be so cumbersome that today it unnecessarily consumes a major part of the energy and saps at the vitality of the research enterprise.

Effect Of World War II

By the 1930s there were still only about a dozen research universities. World War II was the next national emergency that had a major effect on the research programs within universities. In response to the war effort and the increasing awareness of the importance of technology in that effort, universities were encouraged to form research associations to conduct government-sponsored military research (e.g., the Lincoln Labs at MIT and the Willow Run Labs at Michigan). The government-sponsored military-university complex succeeded because of the rapid infusion of funds during the war years. This contrasts with the previous century, when the attempted government promotion of university-industry activities (through land grants) was ineffective. During the war the federal government had a major influence on the university research agenda affecting physics (nuclear weapons), engineering (radar and communications), and microbiology (antibiotics). In many projects secrecy was essential, requiring drastic procedures, such as the isolation of research laboratories from the academic facilities. Universities accommodated

Page 137

themselves to the conduct of secret research, so that they now have over 50 years of experience in conducting research at various levels of security. Thus the recent concern that proprietary research is foreign to universities is largely unfounded.

After World War II, permanent federally sponsored research institutes at universities were established. For example in 1948 Stanford initiated the Steeples of Excellence program to obtain government (mostly classified) research support and developed government-funded microwave and electronics research laboratories. Excellent research capabilities at universities became attractive to industry, and some institutions seized on this new opportunity to further broaden their support from industry. Stanford was highly innovative by developing the concept of research parks situated near the university. Further, they initiated academic degree programs for part-time students to bring industrial researchers back into the university. Stanford also established some of the first industrial-affiliate programs and now has about 35 such programs.

There was another outgrowth of World War II and the perceived need to have a ready scientific capability for military purposes. Vannervar Bush, who was the head of the National Research Council under President Roosevelt, in 1952 helped to form the National Science Foundation (NSF) to support basic research. The infusion of funds into universities by the NSF focused the attention of the university science community on basic research and, indirectly, pulled academia away from industry collaborations.

The Sputnik Crisis

The initial funding of basic research at universities by NSF was modest, but with the launching of the Soviet satellite, Sputnik, in 1957, Congress became intensely concerned about the capability of the American research establishment. Increases in appropriations to federal agencies for the funding of research in universities rapidly escalated. One year after Sputnik, NSF's basic research appropriation was increased threefold to $137 million. These were the "easy money" times for universities. With this infusion of federal resources into research, the number of research universities rose from about 12 to the 100 we have today.

This expansion of federal funding further weakened university-industry relations. Obtaining industrial support requires a good deal of effort: companies with an interest in a specific research program have to be identified by the university's faculty. In turn these companies then have to be convinced that the potential outcome of the research project would be financially beneficial. For obtaining a federal grant, there are only a handful of funding agencies to approach and each agency directs

Page 138

the grant application to its most relevant program. A good deal of anonymity is involved and the faculty do not have to get involved with face-to-face marketing of their research ideas.

As the number of university positions expanded, graduates with doctorates preferred to go to a university rather than into industrial laboratories to pursue their professional careers. University positions are perceived to be more prestigious, and, additionally, faculty members are thought to have more freedom to pursue their own research agendas.

The growth in federal funding of universities continued unabated for more than 30 years until the present time. However, the justification for this funding has drifted from military preparedness to economic development.

Change In Economic Theory

For the most part, the national concern with economic competitiveness and job creation appeared on the national agenda in the 1970s. During that decade a change in economic theory began to emerge. The major efforts of states to expand their job base traditionally had been directed to attracting divisions of mature companies, such as auto assembly plants. David Birch (1979), an economist at MIT, completed a study that suggested that small companies were the source of economic expansion. Birch studied more than 5 million firms to determine where jobs were created and how job creation could be stimulated. He found that most large companies shrank over the 10-year period of the study at the rate of about 8 to 10 percent a year. What this study showed was that if states wanted to produce more jobs, it would be wiser to look at small companies because 52 percent of the jobs came from small companies created within the previous 4 years. The small companies were, in this particular study, technology companies. Another study showed that small businesses commercialize innovations 24 times better than do large businesses, whatever the criteria, and that jobs grew nine times faster and output was three times larger than for large businesses (Osborne, 1988). The findings of this study, popularized by Pat Choate and Susan Walter (1983), eventually led to federal programs such as the Small Business Administration and the establishment of the Small Business Innovation Research grant programs.

Because of this change of paradigm to create more jobs, state governments and the federal government began to consider the promotion of innovation. Who is creating innovation and how could that innovation result in more jobs? To stimulate job creation through innovation, a number of states initiated financial support programs for technology development which naturally affected universities. Pennsylvania was one of the first with its Ben Franklin program; other programs followed quickly, such as the Edison program in Ohio and the CAT program in New York.

Page 139

The basic premise behind these programs was to stimulate research and development in universities and promote cooperation with companies, particularity small companies, to help foster the transfer of technology to the industrial sector and thus to create jobs. This concept has become so pervasive that now there are more than 500 state technology programs. In the biotechnology discipline alone there are now 60 state biotechnology centers.

The three main technological areas targeted for job creation through innovation were biotechnology, computer technology, and materials science. Many states' programs stimulated businesses in those fields. Many new types of arrangements developed between universities and companies (Table 10-2). In the 1970s the industrial agenda and the university research agenda were forced together primarily by state initiatives to stimulate job creation. A recent survey showed that about 400 companies formed by MIT graduates employed 160,000 people and generated a gross income of $27 billion, which represents 20 percent of the state's gross income.

Birch's findings dominated public policy for the past two decades. Recent studies by Steven J. Davis (1994) of the University of Chicago indicated that the role of big business in job creation is equally important.

Universities Enter The Technology Business

December 18, 1980, was the watershed date that launched the current era of intense university involvement with industry. The Bayh-Dole Act (P.L. 96-517) allowed universities to have patent rights for government-sponsored

Page 140

research. The rationale behind this legislation was that billions of dollars of federal funds had been spent in sponsoring research at universities and few jobs had resulted, and legislators were looking for ways to put technology into practice. The federal government had been particularly ineffective at marketing intellectual property (patents) that came from their research support in universities. By giving universities commercial rights to their research innovations, it was expected that it would be in the best (financial) interests of the universities to promote technology transfer to industry. Thus more effective methodologies for commercialization and job creation would evolve.

Before 1980 the federal government spent about $30 billion per year on research and owned about 28,000 patents, but only 5 percent of these technologies were licensed. In that same era, universities obtained about 280 patents per year. By 1992 universities were obtaining patents at a rate of 1,600 per year and generating about $250 million per year in royalties. Data from the Association of University Technology Managers shows that markets of about $11 billion in product sales and 75,000 new jobs were created by the Bayh-Dole act.

Universities, however, were ill equipped to manage this new potential source of revenue. Universities began protecting their financial positions by encouraging faculty to file disclosures for patents. Because confidentiality and secrecy are essential in the early phases of the patenting process, there was a perceived violation of the ''law" of free information access and communication within universities. In addition, stresses began to emerge between universities and industry because companies had been used to free access to faculty, students, and their research laboratories, particularly at publicly supported universities. Further, there was a perception (mistaken) that because most university research was supported by public money, any product of that research was in the public domain.

The emergence of this new role for universities prompted many debates, and one extreme position is exemplified by a statement of Derek Bok, president of Harvard University, who said the university's autonomy and integrity was threatened by "introducing to the very heart of the academic enterprise a new and powerful motive — the search for commercial utility and financial gain."

I do not think universities had changed at all. As noted earlier, universities have always been opportunistic and responsive to society. For example, in the 1980s NSF (under Eric Bloch's leadership) realized that its emphasis on pure science subsequent to the Sputnik crisis actually drove a wedge between universities and industry. NSF developed a number of programs to force a university-industry partnership. These programs, such as the engineering research centers and science and technology centers, require that the NSF-supported centers have an industrial component.

Page 141

These programs were based on NSF's previous successful experience with the university-industry cooperative research centers started in 1973. Universities readily subscribed to these initiatives. Hundreds of applications were processed by NSF, and well more than 100 grants were given.

Recent policy directives from university administrations and faculty indicate the extent to which university attitudes have accommodated to the societal and political dependence on universities as the engine of economic growth. In 1987 Pennsylvania State University faculty passed a resolution stating that some of the major roles of the university were to promote a business-sensitive environment, to promote economic development, and to encourage entrepreneurial activities. Today about 150 years after the initial foray of universities as institutions to promote economic growth, some universities, such as Boston University, are boldly acknowledging in public that business development is one of a university's roles in society.

Recent "Mega-Agreements"

Some examples of the extent of university participation in cooperative enterprises with companies are the number of "mega-agreements" initiated in the past 10 years (Table 10-3). These highly sought after (by universities) co-development business agreements were seen by some as the "golden calf" that would significantly support university research in the future.

Noticeably, most of these contracts were in the area of biotechnology. Why was that so? Possibly because of the coincidence that biotechnology was emerging as a technology about the same time universities were allowed to own property rights through the Bayh-Dole Act. Also, universities were the lead institutions in biotechnology research; thus, companies that wanted fast access to these technologies found a ready and willing partner in the university. Many faculty believed that these extensive and long-term agreements were the beginning of a new wave of research support for universities and would provide a stable source of funding without the continued harassment of peer review required by government-funding agencies. In fact, these mega-grants were awarded to only four institutions, so the pattern has not been widespread. Further, the mega-grant era is probably over. Some of these companies, such as Johnson & Johnson and Monsanto, have indicated that these arrangements did not work out for them as expected. They did not benefit as they thought they would, and it is unlikely that they will make such arrangements again.

NSF reports that industrial support of academic research and development

Page 142

Page 143

FIGURE 10-2 University-industry research centers (UIRC) founding, by decade, for
centers active in 1990. Source: Cohen, Florida, and Goe (1991).

grew 50 percent, to $1.5 billion per year from 1989 to 1993. A recent survey of engineering research in U.S. universities showed that despite the disadvantages of industrial involvement with universities, 72 percent of the directors of organized research units in academia wanted more industrial involvement in their research programs (Morgan et al., 1993). Michael Crow at Columbia University said that about 35 percent of all patents issued to U.S. companies arise from collaborative research projects with universities (Hanson, 1994). The university-industry enterprise has reached significant proportions and must be considered a permanent component of university operations, at least for the research universities. Figure 10-2 shows the growth of university-industry research groups over the past 100 years.

The cost to universities of protecting their intellectual property is not insignificant. A 1990 survey of a dozen leading universities by Indiana University-Purdue University at Indianapolis showed the following average costs: intellectual property staff, one professional per $57 million in research support; average cost per staff member, $132,800 (not including patent costs); patent disclosures, four to five per $10 million of research grants and contracts; licensing and royalty income, 0.5 to 1.0 percent of total university research funding; and average income to university per license to industry, $18,000 to $28,000 (depending on maturity of the universities intellectual property office).

Page 144

Contrast Of Research Cultures

Today we have a contrast of research cultures (Table 10-4). There is a perception that there are certain qualities expected of university research and others expected of industrial research and that these expectations are at two opposite poles. The two extreme views are that universities should pursue only pure, unrestricted research and that companies are only trying to get products to market. In fact, the attitudes of universities and businesses are actually coming closer together.

Roland Schmitt (1986), president of Rensselaer Polytechnic Institute, formerly at General Electric, and former member of the NSF board, perceives that the ideal relationship between the university and industry is a two-way flow of information. Schmitt believes it should be an arms-length relationship, where industry gives the university an understanding of the barriers that practice is facing but does not participate in the research. The university, in turn, provides the knowledge and talent needed to overcome the fundamental problems. The main goal is not to drive universities away from research, but to orient them toward the areas of fundamental research that are most needed by industry. There would be a definite separation of laboratories and personnel but an interaction by means of information transfer.

Future Directions

I think that universities and industry will be going in a new direction in the future, not the one proposed by Schmitt. There is a new economic

Page 145

paradigm, represented in part by Robert Reich's (1983) statement that "competitive advantage lies not in one-time breakthroughs, but in continual improvements." In industry the big breakthrough with the information explosion is not going to provide one company with an advantage over another company or one country with an advantage over another country. There will be a critical need for continuous interchange between the innovators and the producers. The innovators, who are at the university, somehow will have to be connected with the producers, who are in industry.

How might that happen? I think that universities will develop internal cooperative business-type organizations that interact with industry. Medical practice plans operating as businesses within the university are an example of this kind of organization. In industry a change in structure is resulting from the realization that passing technology from the research group to the production group does not work. Total quality management requires that the two groups be intimately interactive. The fundamental research groups in industry are disappearing, and industry is developing a need for what might be called "professional temporaries" in project-oriented teams.

John Armstrong, retired vice president at International Business Machines, stated that an essential component of effective technology transfer is the mobility of people (Hanson, 1994). Universities need to maintain a flexible research establishment so that industry can tap into the faculty as members of project teams. A faculty member would maintain a university appointment but would have a temporary involvement—a very close involvement—with business research. There are recent examples of the evolution of such research organizations at the University of Maryland and at Harvard University. The University of Maryland just opened a $53 million medical technology center in Baltimore that will house both academic and industrial scientists.

Summary

Universities have responded to societal pressures and values. This adaptive nature of universities was nicely summarized by Susannah Hunnewell (1994) on the occasion of the announcement of the establishment of the Harvard Institutes of Medicine for developing innovative relationships with industry:

Yet, in the light of the economic realities facing the university, perhaps Harvard's wrestle with its relationship to the private sector—the angel of good and evil—is simply a new stage in the evolution of academic values. A 1643 pamphlet stated that Harvard's mission was "to advance Learning and perpetuate it to Posterity, dreading to leave an illiterate

Page 146

Ministry to the Churches." In 1708, amid much dissent, John Leverett became the first Harvard president who was not a clergyman, moving the University one step further from it theological mission and toward the secular academic code of ethics, which is now regarded as similarly unassailable. Perhaps technology transfer and implied partnership with industry—in Daniel Tosteson's words, "veritas brought to the sick and suffering"—will become a fundamental value of academia and of Harvard. However improbable it may seem now, such connections may even, someday, become inseparable from Harvard's mission as a disseminator of truth.

Universities do adapt to society's needs and they have changed over time. Universities do get involved in businesses and they can manage them very well. Universities and faculty have successfully managed proprietary research and they can do that very well. New organizational relationships with industry will be created. There are a number being tried and these kinds of relationships or organizational structures will be part of the university 20 years from now.

References

Birch, D. 1979. The Job Generation Process. Cambridge, Mass.: MIT Program on Neighborhood and Regional Change.

Choate, P., and S. Walter. 1983. America in Ruins: The Decaying Infrastructure. Durham, N.C.: Duke Press Paperbacks.

Cohen, W. M, R. Florida, and W. R. Roe. 1994. University-Industry Research Centers in the United States. Pittsburgh: Carnegie Mellon University Press.

Davis, D. 1994. Quoted in myth: small business as job engine. Sylvia Nasor, NY Times, March 25, p. CI.

Hanson, D. J. 1994. Quoted in Budget realities usher in new era of research and development collaboration. Chem. Eng. News, Apr. 25, 1994., p. 35

Hunnewell, S. 1994. Harvard Magazine (Jan-Feb 1994), p. 3437.

Mikulski, B. 1994 In Forum on Science in the National Interest: World Leadership in Basic Science, Mathematics, and Engineering. Washington, D.C.: National Academy Press.

Morgan, M. E., D. E. Strickland, N. Kannankatty, and E. T. Rotto. 1993. Engineering research in U.S. higher education: characteristics, trends and policy options. Paper presented at the 27th meeting ASEE Midwest, Rolla, Mo., March 31, 1993.

Nicklin, J. L. 1993. University deals with drug companies raise concerns over autonomy, secrecy. Chronicle Higher Educ. March 24: A25.

Osborne, D. 1988. Laboratories of Democracy. Harvard Business School.

Reich, R. 1983. Pp. 324 in The Next American Frontier. New York: Penguin Books.

Schmitt, R. 1986. Pp. 19-27 in The New Engineering Research Centers: Purpose, Goals, and Expectations. Washington, D.C.: National Academy Press.

Page 147

11—
National Science Foundation's
Perspective on University-
Industry Interaction

NEAL LANE

In a rapidly changing field such as biotechnology, which touches the lives of people in so many ways, it is appropriate that we should devote a great deal of attention to the ethical concerns of industry-university relationships. All of us are aware of the increasing trend of establishing closer connections between the academic and industrial communities. The National Science Foundation (NSF) has served as a catalyst in developing models for how universities can do a better job of getting the best and the newest information out where it will do the most good.

There is a growing awareness that it is in the best interest of the research community to connect the discovery process, which is at the core of the academic world, to the process of putting new discoveries to use, which is the hallmark of successful industries. At NSF, we have encouraged this approach for quite some time with positive results. Nevertheless, developing partnerships between academe and industry, between discovery-driven research and market-driven technology development, is not without problems.

Partnerships

One frequently noted source of problems is the different cultures of the two systems. The academic culture values openness and views itself as primarily driven by the pure desire for new knowledge, unswayed by prospects of personal financial gain (although a few academics have made some personal financial gain). The marketplace in its pure form has little

Page 148

interest in investing significantly in knowledge for its own sake. Indeed, it has little room for other than economic concerns, such as maximizing profit, managing competition, and gaining market share, all understandable goals of business.

Both of these views fail to capture the remarkable complexity of the real world. Science and technology, universities and the private sector, have worked together for decades in mutually beneficial ways. Industry has a constant need for new information and for knowledgeable people familiar with the latest techniques. Academic researchers often find new scientific challenges arising from "real-world" problems, and it has been that way for a long time. More importantly, academia and industry have worked together in ways that have often strengthened the more positive aspects of the values of both: broadening the horizons of industry and focusing research on areas of practical consideration.

NSF-Sponsored Centers

The possibility of mutually beneficial and noncompromising relations between academic researchers and industry is a basic assumption that underlies the success of more than 100 NSF-sponsored centers connecting universities and industry. Around the country, there are more than 1,000 centers and other collaborative organizations of this kind (Cohen et al., 1993). Research at the NSF centers ranges from telecommunications to steel making, from hazardous waste management to biotechnology.

So it is actually a myth that NSF has supported only the pure pursuit of knowledge, independent of any possible outcomes, and has shied away from encouraging its researchers to interact with industry and with states. It has a long-standing tradition of several decades of encouraging strong interaction with industry.

In the industry-university centers supported by NSF, both partners are beneficiaries of new information. Just as important, inasmuch as the primary responsibility of the university partner is education, the centers provide undergraduate and graduate students with a solid academic background plus hands-on experience with the practical problems of industry.

The umbrella provided by a center enables collaboration that might not otherwise occur. It creates an atmosphere where researchers can devote their energies to a small but important part of a larger problem related to their discipline or a joint project with another institution. These collaborations probably would not get started if they were left to a conversation at lunch or a corridor discussion at a scientific conference. They require much effort. They require champions on both sides. They require

Page 149

time and energy, both of which are precious, and a careful nurturing of the organizations involved in order to succeed.

It would have been comfortable for everyone never to have ventured beyond the university laboratories of independent investigators into the complicated world of collaborative centers. Many conflicts could have been avoided by serving only the interest of the academic researchers. However, more than 20 years ago, NSF took a chance on finding common ground between different cultures, and we have many successes to show for it. NSF provided the framework of incentives that enabled the different participants to come together. Despite the conflicts that do exist, these programs are having a profound effect on the culture of science and engineering. Multidisciplinary work is getting the attention it deserves, and many of the long-standing barriers between the industrial sector and the universities have begun to fall.

Most important, students are able to see the opportunities available in industry. There is another myth, at least in many disciplines of science, that universities actively discourage students from taking an interest in industry. It is suggested that what most professors do is attempt to replicate themselves so that they can be proud of their offspring, who will go out and do things similar to what they have done: teach and carry out research in major research universities around the country. Some data suggest that this does occur, but there are many examples of professors who have encouraged their students to work in a variety of careers.

Some Challenges

Universities could do more. We at universities and colleges could look harder at the educational opportunities we are providing. Are there situations in which we could help change student attitudes? Could we make clearer to students how their energies, abilities, and innovative skills could do more for society in jobs other than academic positions? Students are so sensitive to the views of their research advisors that it would not take much to show them that they would not be considered failures if they choose not to enter academia but to contribute to society in other ways.

Larger conflicts occur in these industry-university interactions, conflicts concerned with ownership of knowledge and equitable distribution of the benefits of academic research. These are difficult issues. The National Science Board's policy encourages maximum openness, reflecting the academic view that scientific information must be made widely available within the community. That ideal can run directly counter to the interest of industrial partners who may seek to protect new discoveries that their money has helped fund. NSF-sponsored centers have had success

Page 150

developing lasting intellectual property agreements between universities and industry and even between potentially competing industries working in the same center. It is not easy, but it can be and is being done.

Our experience has been that although intellectual property is a complex issue, it is an issue that usually can be addressed to the satisfaction of all partners. NSF has for the past two decades developed university-industry partnerships that have successfully dealt with the different values and objectives of the partners, addressed concerns about the ownership of intellectual property, and addressed potential conflicts of interest. Still, there are many possible pitfalls in future industry-university collaborations.

Conflict Of Interest

In the area of conflict of interest, NSF is in the process of publishing its policy on investigator conflicts of interest (NSF, 1994). The policy gives the primary responsibility for addressing potential financial conflicts of interest to the academic institutions, and NSF by and large would play no part. Institutions would establish their own policies, following these minimal requirements:

In the areas of intellectual property and technology transfer, the basic NSF policy is federal law, namely the Bayh-Dole Act (P.L. 96-517 [35 U.S.C. 18]). This law encourages the commercialization of new products developed with federal funds. NSF grantees are expected to seek patent protection and take steps to speed commercialization of inventions that result from NSF funding. If the awardee or the investigator does not wish to seek patent protection, NSF will make the opportunity available first to any other federal agency with an interest in the technology and then to the public.

The academic research community, as well as the industrial community, increasingly recognizes the benefits of carefully crafted collaborative efforts, but the university's primary mission is education, and that is challenging in itself. Whatever else the university does, it must not fail to educate well. We at NSF are confident that future partnerships between universities and industry, if they are based on mutual respect and careful

Page 151

consideration of each partner's responsibilities, can tremendously benefit both partners and the country as a whole.

References

Cohen, W., R. Florida, and W.R. Goe. 1993. University-Industry Research Centers in the United States: Final Report to the Ford Foundation. Pittsburgh: Carnegie Mellon University.

National Science Foundation. Fed. Regist. June 28, 1994.

Page 152

12—
Conflict of Interest in Basic
Biomedical Research

JOSÉ E. TRÍAS

The great judge Learned Hand once said that justice is ''the tolerable accommodation of the conflicting interests of society, and I don't believe there is any royal road to attain such accommodations concretely." What Judge Hand said of doing justice applies with equal force to protecting the integrity of basic biomedical research. Just as society must balance the rights of victims and the accused, of a free press and privacy, of free enterprise and public safety, and of free speech and national security, so we must balance competing interests in basic biomedical research. These conflicts are inherent to our research system. There is no practical way to avoid them nor is there only one way to resolve them. The best that universities and other research institutions can aspire to is, as Learned Hand said, "a tolerable accommodation." No royal road exists to attain those accommodations concretely.

Few challenges are greater than that posed by the complex and conflicting demands we place on our scientific enterprise. Consider, for example, the opening section of the Bayh-Dole Act of 1980. It calls for balancing our interest in promoting rapid utilization of inventions arising from federally funded research with our interest in promoting small business, U.S. manufacturing, collaborations between commercial concerns and nonprofit organizations, and free competition and free enterprise. Our interest in the integrity and vitality of academic research is not specifically mentioned but presumably is implicit throughout. We should also add competing institutional and individual interests to this competition of national interests. Singly, each interest represents a strongly held

Page 153

value. Assembling them into a coherent whole makes for hard choices. None of us has a monopoly on wisdom or integrity in making these choices. We should regard the range of our approaches as a mark of healthy pluralism, not as a sign of weakness.

Core Principles Of Research

Those of us who conduct basic biomedical research in academic settings can agree on a few core principles. One is that the directions of basic research should be determined by sound scientific judgment, unencumbered by monetary concerns. Another is that the results of basic research should be published promptly and made available widely. Scientists who engage in basic research also should be free to talk with other scientists in both academic and industrial settings about new ideas, current or future research efforts, and other aspects of their work.

There also can be no argument about the importance of bringing the results of biomedical research expeditiously from the laboratory to clinical settings. When academic institutions and private industry work together to conduct biomedical research or to commercialize its results, the public stands to benefit not only through improved health, but also with new jobs and enhanced international economic competitiveness.

Support For Research

Industrial support for basic biomedical research can be invaluable in times of national budgetary retrenchment. Traditional sources of support for academic research these days are very tight. In recent years, the ratio of research project grant applications funded by the National Institutes of Health has been about one in four. In some institutes, fewer than one in five applications is supported. Funding from foundations and other nonprofit organizations is limited. Industrial support and collaboration, therefore, are essential to our biomedical research efforts.

Collaboration between industry and nonprofit organizations also has become a significant national objective, especially under the Clinton administration. The Bayh-Dole Act specifically authorizes use of the patent system to promote such collaborations, and federal income tax laws now clearly permit them.

Many prominent officials gathered in early 1994 at the National Academy of Sciences for a conference sponsored by the Office of Science and Technology Policy. Representative George Brown, who chaired the House Science, Space, and Technology Committee, said this at the meeting: "We cannot have a research system running on its own predetermined track and hope that it will intersect serendipitously with the needs of a dynamic,

Page 154

changing society. We must have, instead, a research system that arches, bends, and evolves with the society's goals." Senator Barbara Mikulski, who oversaw appropriations for several federal agencies that fund research, warned that "the U.S. is losing ground. The U.S. is losing time. And the U.S. is losing opportunities. To regain the ground we have lost over the last two decades, we must seek new models of collaboration between our universities and the private sector." Other speakers echoed the same theme.

The importance of academic-industrial collaboration also was made clear at the January 1944 meeting at the National Institutes of Health on conflict-of-interest issues inherent in the commercial sponsorship of academic research. In many ways this trend is welcome and overdue. Strengthening the ties between academic and industrial research will promote innovations of social value. Within the medical arena, the result could be needed funding of academic biomedical research and faster development of products that ease suffering and save lives. However, this trend also will accentuate and make it harder to balance this complex array of competing and conflicting national, institutional, and individual interests.

Competing And Conflicting Interests

Any restraint on the publication of research results impinges on academic freedom, and yet some restraint is required to protect patent rights and legitimate commercial incentives. Similarly, any commercial sponsorship agreement that lays out a particular research goal detracts from the principle that academic research directions should be determined solely by sound scientific judgment, but commercial sponsorship is an indispensable support for academic biomedical research. Moreover, an uncompromising insistence on protecting research directions from financial interest in every case would lead us to forbidding our faculties to hold stock in biotechnology companies, share royalties in licensed discoveries, and consult with industry. We are not prepared to do this.

Another conflict involves the public interest, so forcefully embraced by the Bayh-Dole Act, in having a research discovery brought to practical application by the company best able to do so. Because that determination is in most cases possible only after the discovery has been made, this interest is potentially threatened by any preexisting commitment that favors one company over another. Research arrangements between commercial sponsors and academic biomedical research labs almost always provide the sponsor with a preference over its competitors in commercializing discoveries that arise from the lab's research. The preference is often in the form of an option to license future discoveries, regardless of

Page 155

their commercial values, on terms negotiated before the discoveries were made. In some cases a very large fraction of the funding provided by the commercial sponsor is attributable to a preference of this kind. Where federally funded research is involved, this makes for a lively tension between two stated policies of the Bayh-Dole Act: to achieve the best practical application of a discovery and to promote academic-industrial partnerships. The resolution is not in sight.

One last conflict concerns restrictions on academic biomedical researchers discussing their work with others. Such restrictions may diminish academic freedom and retard scientific progress, but they also may be necessary to protect legitimate commercial interests in an academic-industrial collaboration. It is unclear whether the resolution of this conflict should rest with the individual scientists or with the collaborating institutions. Nothing short of a resolve by the government to free academic research from dependence on commercial sponsorship would avoid the conflict.

All of the conflicts just discussed, and not just this last one, are unavoidable given our institutional arrangements for funding and conducting basic biomedical research. We cannot practically change these arrangements or wish these conflicts away. We have to confront the conflicts and search for a balance that is appropriate for individuals, institutions, and the nation as a whole.

If Judge Hand is right, that means we must find tolerable accommodations. We must recognize that there is no easy, single way to arrive at these accommodations in every case, and we should welcome both a diversity of approaches and a disparity of accommodations. Perhaps our divergent solutions will never converge, but we should not fail, on that account, to exchange impressions.

Howard Hughes Medical Institute

In this vein, I would like to describe the role of the Howard Hughes Medical Institute (HHMI) in conducting basic biomedical research and the approach that it has taken in trying to balance these competing interests. HHMI is the nation's largest private nonprofit supporter of biomedical research by far, with an endowment nearing $8 billion. Last year it spent $268 million on biomedical research, an average of more than $1 million for each of its 225 investigators.

HHMI does not award research grants. It carries out basic biomedical research by employing its own scientists and their staffs and equipping their laboratories at more than 50 universities, teaching hospitals, and other academic settings (called host institutions) across the country. HHMI investigators concurrently hold faculty appointments at the host

Page 156

institutions. Like their faculty colleagues, in addition to conducting research they teach, mentor young scientists, and discharge other academic duties. For instance, Tom Caskey, Savio Woo, and seven other of our investigators have laboratories at Baylor University. There is an HHMI investigator at Rice, and there are 12 investigators in Dallas at the University of Texas Southwestern Medical Center.

HHMI investigators are appointed for terms of 3, 5, and 7 years, subject to reappointment after peer review of their work near the end of their terms. The attrition rate due to this process is less than one in five. This low attrition rate reflects the excellence of the scientists involved. HHMI's current 225 investigators include a number of Nobel laureates and more than 40 members of the National Academy of Sciences. Our philosophy is to support outstanding scientists for extended periods, not specific projects with limited time horizons.

We plan to add another 49 investigators, among them some of the nation's brightest young stars in biomedical research. We also employ and support research teams for each investigator. Altogether, there are nearly 2,000 people on our scientific staff. They do basic research focusing on cell biology, genetics, immunology, neuroscience, and structural biology. In addition, HHMI awards more than $50 million each year through its grants program, mainly for science education.

Contributions Of HHMI Investigators

HHMI investigators are making substantial contributions to elucidating fundamental processes that underlie health and disease. Ray White's participation in identifying the gene for an inherited form of colon cancer; Francis Collins's collaboration in the discovery of the gene for cystic fibrosis; Lou Kunkel's identification of the gene for muscular dystrophy; Bob Horvitz's participation in identifying a gene associated with familial amyotrophic lateral sclerosis, the inherited form of the disease that killed baseball star Lou Gehrig; Tom Caskey's participation in cloning an altered gene believed to be responsible for the most common form of inherited mental retardation, fragile X syndrome; Wayne Hendrickson's and Stephen Harrison's production, together with their colleagues, of the first three-dimensional image of the site on white blood cells (T cells) that acts as a target for human immunodeficiency virus, an image that could be helpful in designing drugs to prevent the virus from getting its critical foothold; and Mario Capecchi's development of gene targeting, a technique that has revolutionized genetic engineering and is now being used in laboratories across the country—all these and many other examples of extraordinary work—have been supported by HHMI.

This work does not go on in isolation from industry. HHMI encourages

Page 157

its investigators to form scientifically productive collaborations with private companies. We strongly encourage our investigators, through our host institutions, to commercialize the results of their research.

As of February 1944, HHMI investigators had recorded 469 discoveries with our office. Patent applications had been filed for 134 of these discoveries, and patents had been granted for 10 of them. HHMI also is involved in 65 licensing agreements. For example, one of our scientists at Yale, Richard Flavell, was among the leaders of a research team that made important discoveries about Lyme disease. Through a licensing agreement, SmithKline Beecham is now applying that research to develop a Lyme disease vaccine. Research by other HHMI investigators had led to diagnostic tests now used for sickle cell anemia and other genetic disorders. Mike Welsh, one of our investigators at the University of Iowa, is collaborating with a private firm on gene therapies for cystic fibrosis. David Williams, one of our investigators at Indiana University, is helping to develop therapies that will reduce the side effects of radiation treatment in cancer patients. HHMI investigators elsewhere are working with companies on research related to acquired immune deficiency syndrome, tuberculosis, and many other diseases.

This is a substantial record. It shows the importance that HHMI places on collaborating with industry and in transferring discoveries quickly from the research bench to the hospital bedside. In fact, our collaborations with industry extend well beyond licensing patented discoveries. Many HHMI investigators consult for private companies, hold stock in biotechnology ventures, or serve on scientific review boards of industrial concerns. A few investigators are founders of biotechnology concerns; most investigators regularly exchange ideas and biological materials with their counterparts in industry. HHMI investigators are engaged in scientifically important collaborations with biotechnology and pharmaceutical concerns. All of these interactions with industry are accommodated, usually successfully, by HHMI's policies on conflict of interest.

HHMI Policies On Conflict Of Interest

HHMI investigators are expected to spend at least 75 percent of their time on research. The remaining time is available for teaching classes, serving on campus committees, and performing other academic duties. If professional time remains after discharging those duties, it can be used for consulting and for serving on scientific review boards of companies.

Investigators are required to sign a statement that assigns to HHMI all rights to any invention or discovery developed through research financed by HHMI. We work closely with our host institutions to commercialize these advances. If one of our researchers at Baylor makes a discovery,

Page 158

for example, we assign to Baylor all of the intellectual property rights. Baylor then takes the lead in seeking a patent and looking for licensees. We share with Baylor the costs of obtaining a patent. If there are royalties, we share those, too. The investigator's share is determined by the formula that Baylor uses for that purpose. The host institution takes the lead in commercializing the property, usually through a technology-transfer office, but HHMI monitors the process. We get copies of disclosure forms, initial patent applications, and so forth. We review license agreements with special care, because we want to be sure that the licensee's access to future research conducted in HHMI labs as a result of the license is limited to what is commercially reasonable to bring the specific discovery in question to practical application.

Collaborations between HHMI investigators and for-profit companies are welcome as long as they are driven primarily by scientific considerations. Our policy is that a company cannot simply pay for research that it hopes to commercialize. It must make a direct and substantial scientific contribution of its own, such as a process, a compound, or a laboratory technique. In short, HHMI's policies embrace scientific partnerships with industry but rule out commercially sponsored research arrangements.

Companies that enter into scientific collaboration agreements with HHMI reasonably expect some commercial benefit in return. When the situation warrants, HHMI will grant a company a time-limited option to license the fruits of the collaboration, but only on terms to be negotiated after the discovery is made. The firm also may receive a limited right of prepublication review to protect intellectual property, and its scientists may be allowed to work in an HHMI laboratory while supported by company funds. However, the company may not leverage the collaboration into broader rights to HHMI research.

We require our investigators to obtain written approval before entering into a collaboration with a for-profit company, even if the arrangement is proposed as a result of collegial and informal ties. Our experience is that the more informal the arrangement, the more likely it is that disputes will arise down the road, even if everyone starts off with the best of intentions.

We also require HHMI investigators to obtain prior approval for consulting agreements. Our researchers may not consult for a company in which they hold a significant equity interest. Generally this means 5 percent or more of the company's outstanding equity, but it depends on the circumstances. We sometimes consider a smaller percentage to be significant.

Consulting is limited to the exchange of ideas. An HHMI investigator may not conduct research for a company or direct others in doing so. Similarly, compensation for consulting may include fixed amounts of cash and equity but may not include incentive features, such as bonuses based

Page 159

on performance. We scrutinize high levels of compensation to ensure that payment is made only for the exchange of ideas and not to gain inappropriate access to HHMI research.

We consider the sharing of our biological materials to be an essential aspect of scientific citizenship. Our system enables our researchers to transfer materials to other academic and nonprofit laboratories with a minimum of paperwork. Transfers to for-profit labs require a bit more documentation but nothing burdensome. Our biggest concern in this area comes when one of our scientists accepts a transfer from a for-profit lab. Sometimes the agreements that accompany these transfers are too broad, claiming rights to almost anything that results. HHMI policies require that these agreements meet a test of reasonableness.

HHMI investigators are encouraged to seek outside funding from the government and from private philanthropic organizations. They may accept gifts from for-profit companies as long as they make no commitment to the sponsor in return.

Finally, HHMI investigators are required to conform to the policies of the host institution in addition to the HHMI policies just described. The result is to give dispositive effect in any case to the more restrictive policy. HHMI's policies resemble in many ways those of our host institutions. Indeed, our requirement that HHMI investigators conform to both HHMI and host institution policies works tolerably well precisely because, in many cases, the policy differences are not great. By delegating commercialization of HHMI discoveries largely to host institution policies and procedures, HHMI adopts, de facto, the principles and policies of the Bayh-Dole Act, thus indirectly embracing national interests such as the preference for small business, the U.S. manufacturing requirement, and the promotion of free competition and free enterprise. However, HHMI policies do differ from those of our host institutions in certain significant respects.

For example, as mentioned earlier, we have an explicit upper limit of permissible stock ownership in a company for which a HHMI investigator wishes to consult. We specifically disallow incentive compensation in consulting without exception. We restrict consulting to the exchange of ideas and flatly prohibit the conduct or direction of research for an industrial concern. In scientific collaborations with an industrial partner, we may agree to grant a preference to the industrial partner to commercialize the fruits of the collaboration, but we will limit the preference to an option to license on terms to be negotiated after the discovery is made. Some of our host institutions draw these lines in somewhat different ways. The sharpest difference is in our approach to commercial sponsorship of research. Academic research institutions, by and large, accommodate it. HHMI policy rules it out.

Page 160

HHMI's Philosophy

Our philosophy is to support outstanding scientists for extended periods, not specific projects with limited time horizons. We believe that the nature and extent of our support frees HHMI investigators to pursue basic biomedical research along directions determined by their own scientific judgment and to take unusual scientific risks. This freedom is expensive, but we believe it contributes to the vitality of our biomedical research enterprise and is worth the cost. We believe that the pressures inherent in commercial sponsorship could undermine this freedom. We rule out commercial sponsorship of HHMI research to protect this freedom.

Like our host institutions, however, we continue to evaluate and reevaluate our policies on conflict of interest. Academic-industrial interaction is increasing. The pattern of interaction is evolving. It would be sheer happenstance if policies fixed yesterday served us well tomorrow. Along with our host institutions, we believe that our labors in this area are not near an end.

Page 161

13—
Patents: Help or Hindrance to
Technology Transfer?

REBECCA EISENBERG

"Intellectual property" is a broad heading used to refer to a wide variety of rights associated with inventions, discoveries, writings, artistic works, product designs, and designations of the source of goods and services. Patents and trade secrets are the most important of these sorts of intellectual properties in the field of biotechnology.

One aspect of intellectual property that distinguishes it sharply from other forms of property—and for some people makes it harder to justify—is that intellectual properties may be possessed and used by many people simultaneously. This is not so for tangible property. If someone borrows my car, I cannot use it—nor can anyone else—until the car is returned to me, but if someone borrows my secret manufacturing process or my backup copy of my word processor, I can keep on using it while someone else is using it. In fact, no matter how many people I share my word processor with, as long as everybody can make a copy, it is not going to interfere with my ability to keep on using it. This capacity for simultaneous possession by many people is a feature that is common to all sorts of intellectual property, including computer programs, musical recordings, lists of customers, and self-replicating cell lines or genetically engineered organisms. Many people intuitively feel that they are doing nothing wrong when they make unauthorized use of intellectual property as, for example, when they borrow and copy other people's computer programs.

What, then, justifies a system of exclusive legal rights to ideas and

Page 162

information that others could benefit from without depriving the owners of their use? In the United States, intellectual property is usually justified in instrumental terms, although some advocates of intellectual property try to justify it in moral or natural-rights terms. The instrumental justification for a patent system is that inventions and discoveries are often costly to make as an initial matter but cheap and easy to copy once someone else has made them. Because the public benefits from new inventions and discoveries, we want to encourage people to invest in research and development, but they might not be willing to do so unless they have some means of preventing competitors from reaping the benefits of their investment without sharing in the initial risk and cost.

Secrecy As A Way To Protect Intellectual Property

One way of keeping inventions and discoveries out of the hands of competitors is to keep them secret. As long as no one else knows the company's formula, the company does not have to worry about competition from outsiders who did not share in the cost of developing it. However, secrecy only works for certain types of inventions, such as manufacturing processes, that can be exploited commercially without disclosure. Many inventions and discoveries are self-disclosing once you sell a product that incorporates them. Even if secrecy is feasible, it might not be desirable. We might want to promote disclosure of new inventions and discoveries in the interest of furthering continuing technological progress in the field.

Patent Protection

An alternative to secrecy for some inventions is patent protection. A patent gives an inventor the right to exclude others from making, using, and selling the invention for a limited term: 17 years from the date the patent is issued under current U.S. law, 20 years from the application filing date in many other countries. The inventor may choose to make, use, and sell the patented invention; license others to do so exclusively or nonexclusively; or suppress the use of the invention entirely. The one thing the inventor cannot do is to keep the invention secret. To obtain a patent it is necessary to file an application that includes a full disclosure of the invention and of how to make it and use it. In many parts of the world, this disclosure is made public 18 months after the application filing date. In the United States it is made public as soon as the patent is issued. Under either system, an inventor who wants to disclose the invention earlier can do so as soon as the patent application is on file without jeopardizing the prospects for getting a patent. So in addition to requiring

Page 163

disclosure, patents promote disclosure by providing a property alternative to trade secret rights that survives even after disclosure.

Advocates of patents believe that they promote technical progress both by providing economic incentives to make new inventions and to develop them into commercial products and by promoting disclosure of new inventions to the public. The extent to which the present patent system achieves these goals is not known. Few people would argue that invention and technical disclosure would come to a standstill in the absence of a patent system. Firms that introduce new technologies into the market might find some research and development profitable even without patent rights. The lead-time advantage over competitors gained by being first in the market with an innovation, for example, might be enough to justify continued expenditures on research and development.

Costs Of The Patent System

The prospect of obtaining patent rights undoubtedly increases incentives to invest in research and development, at least in some fields, but the social costs associated with having a patent system have to be weighed against these benefits. The most obvious social cost associated with patents is that they create monopolies that increase the price and reduce the supply of products that are covered by patents. This may be a tolerable cost for socially useful inventions that would not have been made without the incentives of the patent system—we might choose to have these inventions at a high price rather than not to have them at all, but it is a high price to pay for inventions that would have been developed even without patent rights.

It is therefore important to formulate rules of patent law that exclude from protection inventions that would have been made even without the added incentive of the patent system. Most patent systems attempt to do this by requiring that an invention have a certain level of importance before it can be patented. In the United States we require that an invention be new, useful, and nonobvious to be patented. The nonobviousness requirement is a mechanism for distinguishing between inventions that would come about without the patent system and those that need its added incentive. These rules are very difficult to administer and result in a lot of uncertainty in the patent system.

Patent systems also entail considerable administrative costs. These include the costs of determining which inventions are patentable, an activity that consumes the time and energy of technically trained people who might otherwise be adding to the knowledge base more directly. Patent applications also incur costs in procuring and enforcing patents,

Page 164

and their competitors incur costs in avoiding infringement (including the costs of research efforts aimed at inventing around patented inventions).

Patents may also inhibit inventive activity of people other than the patent holders in fields that are dominated by patents. Patents may distort social priorities by diverting resources toward invention and away from other social problems. They may distort research in favor of making patentable inventions and away from areas in which patent protection is not available, such as basic research or discoveries in fields that are excluded from patent protection but might nonetheless be socially beneficial.

The Patent System In Biotechnology

These costs of the patent system should be remembered when considering the role of patents in biotechnology, particularly the role of patents in publicly funded biotechnology such as the Human Genome Project (Eisenberg 1994a,b). The Human Genome Project is an interesting example because it involves extensive government funding directed toward generating vast amounts of information in the hope that that information will ultimately be put to use in developing new products and processes for the diagnosis and treatment of human disease. Much of this information is generated in government and university laboratories that are not in a position to undertake the research and development necessary further downstream to translate basic research discoveries into commercial products.

Patenting As A Way To Promote
Technology Transfer

Technology transfer to the private sector is a prerequisite for the development of genome-related products, but how to achieve technology transfer in such a project is a complex matter. U.S. policy since 1980 has reflected an increasingly confident presumption that patenting discoveries made in government-sponsored research is the most effective way to promote technology transfer and commercial development of those discoveries in the private sector. Policy makers of prior generations may have thought that the best way to achieve widespread use of the results of government-sponsored research was to make them freely available to the public. Advocates of the new patenting strategy stress the need for exclusive rights as an incentive for industry to undertake the further costly investment necessary to bring new products to market.

This new strategy is justified in terms of both trade policy and technology policy. The trade policy argument is that although the United

Page 165

States leads the world in basic research and the creation of new technologies, other nations do a better job of commercializing and adopting new technologies in the private sector. As a result, U.S. firms lose sales to foreign manufacturers of goods that are based on technologies pioneered in the United States. The competitiveness of U.S. firms in world markets might be enhanced by leveraging U.S. strengths in research into a stronger position of dominance in applied technology.

The technology policy argument is that government-sponsored basic research discoveries that have been left in the public domain have not been picked up by the private sector and developed into commercial products, or at least not at the rate that one would hope to see. If the economy needs a steady infusion of new technologies to grow and to improve worker productivity, many argue that we need to induce the private sector to tap into the wealth of new information emerging from government and university research. This rationale presumes that inventions made freely available are languishing in government and university archives rather than being actively exploited by all.

New Patent Policy Of The 1980s

The solution to these twin concerns, in keeping with the privatization ethos of the 1980s, was to offer up the results of government-sponsored research for private appropriation by U.S. industry through the mechanism of licenses under government- and university-owned patents. Exclusive patent licenses from a government agency or university would make it profitable for U.S. industry to develop products that would be too risky or costly to pursue if the discoveries were left in the public domain and competitors were therefore free to enter the market once it was established. Technology transfer facilitated by patent rights would generate new products for U.S. consumers and create jobs for U.S. workers while protecting U.S. firms from foreign competition.

Curiously, although the primary motive behind this patent policy appears to have been a desire to benefit U.S. industry, the primary impetus to get it enacted into law seems to have come from the government, with the support of universities, rather than from the private sector. Although industry has been slow to go for the bait of patent licenses to government-sponsored research discoveries, the government has not wavered from the patenting strategy but has instead fortified it by extending it to cover more discoveries made in a wider range of research settings.

Starting in 1980 the presumption in favor of patenting research discoveries was applied to small business and nonprofit organizations making research discoveries with federal funding under the Bayh-Dole Act. In 1983 these provisions were extended by a Presidential memorandum to

Page 166

large businesses doing government-sponsored research. They were extended by statutory amendments to discoveries made at government-owned, contractor-operated facilities in 1984, then to intramural research and research performed under agreements between government agencies and the private sector under the Federal Technology Transfer Act of 1986. Subsequent legislation and executive orders have continued to broaden and fortify this policy.

Today we have a system that virtually guarantees that wherever federally sponsored inventions are made—whether in government, university, or private laboratories—anyone involved in the research project who wants the discovery to be patented may prevail over the objection of anyone who thinks the discovery should be placed in the public domain. Thus, for example, if a university is reluctant to patent a discovery made in its laboratories with federal funds, the sponsoring agency may insist on obtaining a patent. If a government agency or university has no interest in pursuing a patent, the investigator who made the discovery may step in and claim patent rights. If anyone sees money to be made through patenting a government-sponsored research discovery and has the resources and sophistication to pursue patent rights, chances are it will be patented.

Questioning Federal Patent Policy

Now all of this makes sense if we want all government-sponsored research discoveries to be patented. But do we? Since 1980 federal patent policy has assumed that discoveries left in the public domain will not be used and that granting exclusive rights to discoveries to businesses will ensure their commercial exploitation for the benefit of consumers, taxpayers, and the economy. Our present statutes come close to reflecting a conclusive presumption that this is so. But is it so in fact?

The answer probably will vary from one field to the next and from one discovery to the next. The strong pro-patent tilt of current policy seems like a vast oversimplification of the enormously complex task of achieving technology transfer across the broad spectrum of discoveries emanating from federally sponsored research.

One reason for the complexity is that technology transfer requires extensive back-and-forth communication among different types of institutions and among researchers and technology users who speak to each other across significant cultural divides. The extent of this problem varies among fields. In some fields, researchers in government and university labs share norms of openness that conflict with commercial interests in secrecy. Patent rights may sometimes reduce this difficulty by providing intellectual property rights that survive disclosure. At other times, concerns about preserving the ability to patent future discoveries might fortify

Page 167

commercial incentives to maintain secrecy and thereby aggravate the conflict between the cultures of academic research science and industry. Any policy that promotes widespread patenting of the results of government-sponsored research would thus need to take into account and manage the effect of patents on the research enterprise.

Even setting aside the culture of academic research and focusing exclusively on the perspective of industry, current policy seems to oversimplify a complex problem. Patents may make sense as a means of facilitating technology transfer for some government-sponsored discoveries, but there are reasons to suspect that they make little sense for others. The course of scientific discovery and product development is infinitely complex, variable, and unpredictable. Uniformity in technology transfer policy may therefore be a false ideal. Neither the old-fashioned approach of leaving all new discoveries in the public domain nor the newer approach of assigning exclusive rights in such discoveries to private parties should be uniformly applied across the entire range of publicly supported discoveries. In our eagerness to avoid the inadequacies of the public domain approach, we may have moved too quickly and too emphatically in the opposite direction to the point where patent rights in some government-sponsored discoveries may actually be undermining rather than supporting incentives to develop new products and bring them to market.

NIH Applications For cDNA Patents

One sign of trouble in paradise for federal technology transfer policy was the reaction of industry trade groups a few years ago to the filing of patent applications by the National Institutes of Health (NIH) on thousands of partial cDNA sequences of unknown function identified in government laboratories (Eisenberg, 1992). The research that led to the controversial patent applications consisted of taking randomly selected cDNAs from a human brain tissue cDNA library and finding the DNA sequence for small portions of those genes without knowing what proteins or functions are associated with the genes. Beginning in the summer of 1991, NIH filed patent applications claiming the partial cDNA sequences as well as the full genes of which they are a part, which NIH claimed could be readily obtained with the partial sequence information. An avowed purpose of seeking these patent rights was to be able to offer licenses to firms to promote the development of products related to the sequences.

Some of these patent applications were rejected by the U.S. Patent and Trademark Office, and the new leadership of NIH decided not to appeal the rejections and to withdraw the remaining claims. Although the immediate controversy was thereby resolved, it is nonetheless worth-while

Page 168

to reflect upon this controversy as a case study of the role of patents in technology transfer. These patent applications generated considerable controversy among scientists throughout the world who charged that the human genome represents the universal heritage of humanity and should be dedicated to the public domain. They argued that intellectual property rights could undermine scientific collaborations and thereby retard progress in the Human Genome Project. Much of the controversy within the scientific community has been a reprise of an old debate about the effect of intellectual property rights on research science norms.

What was striking about this controversy was that the patent applications were also opposed by trade groups from the industry that NIH intended to benefit through technology transfer. These trade groups are not composed of naive, idealistic scientists who have limited experience with patents and limited interest in product development. Their members are the same hard-nosed, profit-maximizing firms that the government is trying to entice into developing products out of government-sponsored inventions. Position statements from the Pharmaceutical Manufacturers Association and from two biotechnology trade groups that have since merged, the Industrial Biotechnology Association and the Association of Biotechnology Companies, expressed views on the NIH patent applications that contradict the hypothesis that patent protection for those particular discoveries was necessary to protect the interests of firms that might develop related products in the future.

The Pharmaceutical Manufacturers Association and the Industrial Biotechnology Association urged that NIH not seek patent protection on DNA sequences with unknown biological function but instead place such sequences in the public domain. The third group, the Association of Biotechnology Companies, supported the NIH decision to seek patent protection, but only as a means of generating revenues for the government and not as a means of ensuring the availability of exclusive rights to those sequences. Indeed, even the Association of Biotechnology Companies urged that the patents be licensed on a nonexclusive basis so as not to block development projects in industry. So although this position is nominally consistent with current federal patent policy, it contradicts its underlying rationale by conceding that, at least in this particular case, exclusive rights to the discoveries could interfere with their effective commercial development.

Industry Opposition To NIH Patenting cDNA

Why might U.S. industry object to NIH's pursuit of these patent rights and what does that tell us about the role of patents in technology transfer? First, an easy explanation is that the firms may not want NIH to be in a

Page 169

position to grant or deny licenses to develop genome-related products. There is an essential irony in using government-owned patents to achieve technology transfer. This strategy places a government agency in a licensing role for the purpose of promoting privatization. If NIH holds patent rights to a significant portion of the human genome, it may use its position as licensor to regulate the development of genome-related products, which is the last thing that industry wants.

Exclusive licenses under NIH patents until recently included reasonable-pricing clauses that permit NIH to monitor the reasonableness of prices charged for licensed products. Exclusive and nonexclusive NIH licenses include domestic manufacturing clauses requiring the licensee to manufacture products in the United States or at least granting a preference for U.S. manufacture. Such provisions tie the hands of industry and limit the profitability of products developed even under an exclusive patent license.

Firms may be particularly wary of NIH as a licensor in view of its recent role in authorizing a generic drug manufacturer to pursue NIH's claim against Burroughs Wellcome to patent rights in the use of the drug azidothymidine (AZT) in the treatment of acquired immune deficiency syndrome. This episode highlights the ambivalence of NIH toward profit maximization in the marketing of health-related products.

Second, the patent rights that NIH sought may have seemed unnecessary as a means of protecting the profit expectations of industry. The current government patent policy is based on a simple model of technology transfer in which a patent on a government-sponsored invention is the only source of exclusivity on the horizon for firms seeking to develop related commercial products. However, commercial products in industries that make use of patents, such as the pharmaceutical industry, typically embody multiple patented inventions. If a firm has its own patent rights in a product that are adequate to protect its market position, NIH patent rights covering the same product or covering inventions that are necessary to develop or market the product may be nothing more than an annoyance to the firm. If government patents are not only unnecessary to provide market exclusivity, but also come with burdensome restrictions on pricing and place of manufacture, firms may see them more as a regulatory hurdle than as an incentive to innovation.

Third, NIH patents may have seemed ineffective in protecting the market position of innovating firms. Patent rights are most likely to be effective in promoting product development when they ensure the patent holder of a commercially effective monopoly in the relevant product market. Whether NIH patents would have had this effect depends on the scope of patent rights NIH had been able to obtain from the Patent and Trademark Office. Generally, the most effective commercial protection,

Page 170

and therefore the most powerful incentive to invest in product development, is provided by a patent on an end product that is sold to consumers. Partial cDNA sequences of unknown function may turn out to be marketable as end products—perhaps in a diagnostic product—but it is more likely that they will be useful as research tools to find the full-length genes to which they correspond and to make the products for which those genes code.

The NIH patent applications included claims to these full-length genes on the theory that, by disclosing how to use the partial sequences as probes to find the full genes, NIH had provided an enabling disclosure of how to make those full genes. Although the NIH patent application did not disclose either the complete DNA sequence for the genes or the proteins for which they code, it did provide a general description of how to use the partial sequences as probes to find the full genes and how to achieve expression of the gene products once the full genes have been found. But under recent court decisions it is unlikely that NIH would have been able to obtain patents on the full genes without setting forth their full sequences (see, e.g., Fiers v. Sugano, 984 F.2d 1164 [Fed. Cir., 1993]). Thus, NIH patent rights would probably have been limited to narrower claims to the specific partial cDNA sequences than are actually set forth in the applications. Such limited patent rights would probably not have been broad enough to give effective commercial protection to firms seeking to bring related products to market, and the argument for obtaining patents as a means of promoting product development would lose its force.

Patents On Research Tools

The partial cDNA sequences themselves are primarily useful as tools for research and development. Not only is it difficult to detect and prove infringements of such patents, but often the only effective remedy even for proven infringement will be damages, because an injunction against future use of the invention at that stage would not thwart the efforts of a competitor who has already finished using the invention. One could argue for a substantial damage remedy, but as long as the competitor no longer needs to use the patented invention in the manufacturing stage, an injunction against future infringement would not serve to keep the competitor off the market. Firms that are interested in developing end products for sale to consumers are unlikely to see patents on research tools used only during research and development as an effective means of promoting their market exclusivity in the ultimate products. Such patents may generate royalty income for their owners, and the prospect of earning royalties may make it more profitable to develop further research

Page 171

tools in the private sector. However, it is unlikely to enhance the incentives of firms to develop end products through the use of those research tools.

I think there are reasons to be wary of patents on research tools apart from their ineffectiveness in promoting product development. Negotiating licenses for access to research tools may present particularly difficult problems for would-be licensees who might not want to disclose the directions of their research in its early stages by requesting a license. Moreover, a significant research project might require access to many research tools. If each of these tools required a separate license and royalty payment, the costs and administrative burden could mount quickly.

Patent holders, moreover, may find it more lucrative to license research-tool patents on an exclusive rather than nonexclusive basis and in the process choke off other firms' research and development. For years this country has sustained a flourishing biomedical research enterprise in which investigators have drawn heavily on discoveries that their predecessors left in the public domain. Even if exclusive rights enhance private incentives to develop further research tools, they could do considerable damage to the research enterprise by inhibiting the effective use of existing tools. Patents on research tools may offer ineffective commercial protection to firms that use the tools to develop new products for consumers while interfering with research and development within those firms.

The more research that remains to be done to develop a product, the more likely it is that the innovating firm will make further patentable inventions of its own. These subsequent inventions are more likely to be incorporated in the final product, and patents on such inventions are thus likely to be far more important to the firm's profit expectations than exclusive access to any particular research tool.

Conclusion

The present policy of promoting patents on federally sponsored inventions has become rapidly entrenched in U.S. law, although it is not clear that this policy always serves its underlying agenda of furthering the transfer of new technologies to the private sector for commercial development. Patents undoubtedly have a critical role to play in facilitating technology transfer in some contexts, but they can also interfere with technology transfer and with the broader goal of promoting continuing technological process. These goals may sometimes be served by allocating new information to the public domain. Government is uniquely situated to enrich the public domain, and we should be wary of disabling the government from performing this critical function in our eagerness to enhance private incentives to put existing discoveries to use.

Page 172

References

Eisenberg, R. 1992. Genes, patents, and product development. Science 257:903-908.

Eisenberg, R. 1994a. A technology policy perspective on the NIH gene patent controversy. University of Pittsburgh Law Review 55:633-652.

Eisenberg, R. 1994b. Technology transfer and the genome project: problems with patenting research tools. Risk: Health Safety Environ. 5:163-175.