of state, regional, and federal efforts to facilitate technology transfer was also discussed. Workshop discussions centered specifically on technology transfer within the biotechnology and automotive electronics industries. Participants also examined the federal government's role in facilitating technology transfer.
Four factors emerged as central to the discussions: the pace of technological change in each sector; the structure of the industry in question; private sector needs in spurring the commercialization of new technologies; and the relative competitive position of U.S. industry in international markets.
This summary synthesizes ideas expressed at the workshop. It does not represent a consensus opinion of participants in the discussions, members of the Panel on the Government Role in Civilian Technology, the National Academy of Sciences, National Academy of Engineering, Institute of Medicine, or National Research Council. This summary does not contain conclusions or recommendations.
INTRAFIRM TECHNOLOGY TRANSFER
Technology transfer that occurs within a single company is most commonly associated with large corporations, which often have organizationally and geographically separate research and development groups. Discussions during the workshop made it clear, however, that lessons from intrafirm technology transfer in large organizations may also be relevant to small firms.1 In addition, the methods firms use to transfer ideas and information internally may prove effective for technology transfer between firms and among industry, academia, and government. Whether technology transfer takes place within a single firm or between two companies, the conditions for success often remain the same. Customer-supplier links are key to the process. They are needed both to facilitate technology transfer and to maintain competitive leadership, quality, and financial stability.
Mechanisms for In-House Technology Transfer
Corporations with in-house laboratory facilities face the challenge of moving technology between different operating divisions, as well as adopting technologies developed outside the company. This process, as it occurs at several large electronic firms, was outlined at the workshop. Technology transfer at one firm is aided by a "product and process development team" to a group of staff engineers, and other scientific and technical personnel. These individuals are the channels through which customers provide information on improvements in the firm's products and manufacturing processes.
Even when research laboratories and product or process development groups are in close proximity within a company, transferring technology
between these two groups involves significant effort. A number of formal mechanisms to facilitate technology transfer and diffusion were mentioned at the workshop. Among the most important were:
Consultation: Scientists often spend a considerable amount of time with product development personnel to help solve product-specific problems. This joint activity provides researchers with a view of product and process engineering that can improve their work once they return to their laboratories.
Transferring expertise: One of the most effective ways to transfer technology is to move the individuals who have specialized knowledge to divisions within a firm. Many corporations use short-and long-term "internships" to transfer research personnel to areas of the company involved in product development, and vice versa.
Joint assignments and projects: Removing the formal barriers between research and development can enhance technology transfer. Toward this end, some companies assign staff to both product development and research activities. Others bring full-time researchers and development engineers together on the same project.
Other important, but less effective, ways of fostering technology transfer include distributing research reports and technical memos to development staff and conducting research seminars and product strategy reviews.
Workshop participants stressed that no single mechanism or approach to technology transfer is adequate alone to meet corporate technology development needs. Technology transfer is a complex, chaotic, and dynamic process that requires constant revision and change as the realities of the marketplace change.2 Following the one-dimensional, "pipeline" view of technology transfer is no longer a viable strategy.3 Several participants stressed that it was management's responsibility to encourage communication between groups within a firm. Management should also attempt to foster an atmosphere in which high-risk, innovative work is encouraged. Similarly, researchers should be shielded, to the extent practical, from short-term demands of the market. Employees who fear failure or delay will not take risks that may be critical to successful R&D projects.
An official at one company noted, however, that technology transfer does not happen just because management demands it or company policy calls for it. Individuals involved in research and product development must be motivated to undertake the steps necessary for successful technology transfer. This official also noted that simply because technology transfer is often a chaotic process, firms should not be discouraged from constructing a plan for achieving it. There must be clear objectives in any transfer strategy, with timetables and frequent revisions of original plans.
One method of stimulating technology transfer discussed at the work
shop involves moving development engineers to research units prior to the period in which technology transfer is expected to occur. Over time, the number of manufacturing engineers assigned to projects increases, whereas the number of development engineers decreases. One large U.S.-based chemical company uses "business teams," comprised of individuals from marketing, manufacturing, research, accounting, and personnel planning, to increase information and technology flow within the company. It also has a central research department that, in addition to supporting applied R&D, enhances the company's innovative capacity while at the same time strengthening its ability to negotiate with other firms for important research-related information.
The firm's operating departments each have their own research divisions, but the central facility maintains a long-term perspective on R&D challenges facing the company. The system is referred to as "managed collaboration," in which technical staff and management officers with operating experience are integrated into a single corporate research division. Operating department R&D directors are rotated through the organization as directors of corporate research. This process brings business and management experience to the leadership of the corporate laboratory and increases understanding of corporate research goals in the departmental laboratories. Another company integrates its corporate and operating department research personnel and management in project-focused centers. Technology committees, staffed by R&D directors, ensure that institutional knowledge about specific business areas, such as polymers, energy, and health sciences, is shared by all operating departments.
In all in-house mechanisms to enhance technology transfer, company relationships with customers, vendors, and suppliers remain the key link to successful technology development. It was noted that many firms view these relationships as business partnerships and that companies must often work closely with suppliers, sharing data and personnel, to maximize chances for successful product and process technology development.
TECHNOLOGY TRANSFER BEYOND INDUSTRIAL LABORATORIES
In-house efforts to transfer technology from the laboratory bench into marketable products are vital to most firms that conduct research and development, and are part of a larger process that involves many individuals and organizations outside the company as well. From a corporate perspective, to varying degrees, technology transfer relationships with other firms, universities, and federal laboratories are also desirable. Although this interaction has usually been one-way—into the corporation, not out—many firms are discovering that sharing technologies with other organizations can con-
tribute to technology development goals. New relationships are being formed between and among corporations and their suppliers, through joint R&D ventures, cooperative agreements with federal laboratories, and university-industry partnerships. These new relationships reflect the importance of balancing cooperative and competitive interests in today's global economy.
From the university's perspective, technology transfer—both into and out of the university—can be productive, reflecting, in part, the academic tradition of free flow of information and expertise. Even in universities, however, the desire for professional prestige, the competition for scarce research funds, and the trend toward more extensive relationships with the private sector can reduce the flow of new ideas and technologies. As academic institutions seek to capitalize on the economic benefits from their research activities, these impediments to successful technology transfer may continue to increase in number and complexity. To state and federal governments, technology is viewed as a valuable resource that contributes to economic vitality and public welfare. Enhancing the diffusion and use of new technologies—especially those developed with the assistance of public funds—is a primary goal. Government agencies such as the National Institutes of Health (NIH), the National Science Foundation (NSF), and the Departments of Commerce, Defense, and Energy (DOC, DOD, and DOE, respectively) have also moved to stimulate technology transfer from federal facilities.
Mechanisms for "Cross-Boundary" Technology Transfer
Workshop participants spent a considerable amount of time discussing technology transfer involving "cross-boundaries." A number of the specific mechanisms used to encourage this activity are outlined below.
Industry-university cooperative arrangements; Using this approach, corporations can gain valuable access to the university research community, and students and researchers can gain insight into commercial technology development. Several different types of programs exist, including joint research-industrial parks and the NSF's Engineering Research Centers. These relationships often involve funds from state, federal, and private sources.
Company-supplier relationships: A number of steps can be taken to increase the likelihood of technology transfer between firms and their suppliers. Corporate engineers and engineers working for suppliers can spend time in each other's laboratories, for example. Joint R&D projects, between a firm and its supplier, have the potential for enhancing the success of product and process technology development. Finally, quality improvement programs, undertaken jointly by corporations and suppliers, can lead to more effective technology transfer.
Cooperative R&D ventures between competitors: The benefits of cooperative arrangements include shared capital costs, economies of scale, and reduced risk, and can involve a considerable amount of technology transfer. U.S. corporations are increasingly working within such relationships. Whether they involve private firms only or combine the efforts of public and private entities, cooperative ventures can facilitate technology transfer and can result in technological breakthroughs. Moreover, federal and state governments can leverage investment in research and allow market forces to direct research agendas by requiring matching funds from private participants.
Cooperative Research and Development Agreements (CRADAs): Cooperative Research and Development Agreements were authorized by the Technology Transfer Act of 1986. Under a CRADA, a federal laboratory provides scientists and equipment for a particular industry-based research project; the company provides funding and its own scientists and equipment. CRADAs mentioned at the workshop that appear promising include those between NIH and small biotechnology firms. Many observers believe, however, that these agreements have yet to achieve their full potential to stimulate technology transfer and development. Participants identified a number of problems with CRADAs: they require a high level of technical sophistication on the part of industrial partners, which narrows the field of potential participants; ''cultural" differences (with regard to incentives, recognition, and rewards) between private industry and federal laboratories can reduce the potential effectiveness of CRADAs; and a considerable amount of administrative oversight is associated with their operation. These difficulties contribute to an environment that inhibits technology transfer.
TECHNOLOGY TRANSFER IN AN EMERGING INDUSTRY: BIOTECHNOLOGY
The biotechnology industry is relatively young, having experienced rapid growth in the late 1970s and early 1980s. The founders of this industry, most of whom began their careers in university research laboratories, have been essential to technology transfer. The industry's rapid commercialization of discoveries is due in large measure to its continuing ties to academic research institutions and its openness in sharing the results of basic research. Another factor contributing to the development of the biotechnology industry centers on the availability of venture capital.4
As the industry matures, however, some of these advantages may no longer apply.5 Workshop participants noted, for example, that as the founders of biotechnology companies retire, the industry's openness to academic laboratories is likely to diminish. In addition, some scientists who, in the past, had affiliations with both academia and industry are now less mobile be-
tween these sectors. Another potential problem, according to workshop participants, is conflict of interest. Policies intended to address real or perceived conflicts of interest, whether at the federal or the university level, may make it more difficult for industry and academia to maintain close ties in the future.
Participants pointed out that as investors increasingly focus on anticipated returns to investment in biotechnology, the flow of venture capital to the industry is slowing. Although this may simply reflect the maturity of the biotechnology industry, it may force some firms to reduce their levels of basic research funding and, as a result impair long-term competitive advantages. Participants noted that firms with access to capital markets and government research support, or those that operate under reduced regulatory scrutiny and transparent conflict of interest guidelines, would benefit.
Industry Strategies for Biotechnology Transfer
Representatives from the biotechnology industry discussed a number of strategies to encourage technology transfer. For example, many firms have established scientific advisory boards, whose members are drawn at least in part from academia. Not only do these advisors provide scientific expertise, they also are a link between the firms and university research laboratories. Other steps taken by the biotechnology industry that directly or indirectly result in technology transfer include the following:
Using CRADAs: Involvement with federal laboratories is one method of leveraging R&D funds. In some instances, firms have calculated that problems associated with CRADA are less important than access to research conducted in federal laboratories, one participant at the workshop explained. Several commercially successful products are the result of CRADAs, including the AIDS drugs AZT (azidothymidine) and DDI, and the human immunodeficiency virus (HIV)-antibody tests. Other benefits of collaborating with federal laboratories include access to expensive state-of-the-art equipment and machinery, as well as technical assistance from federal researchers.
Licensing: Several participants at the workshop expressed the view that product licensing should occur at or as close as possible to the time of a breakthrough. Waiting for patents, for the determination of all possible uses or even for the complete understanding of a new technology, can result in loss of potential returns. Biotechnology firms must be conscious of the needs and interests of academic collaborators when developing licensing strategies. Royalty reimbursement and technology-sharing arrangements can speed up the process of bringing an advanced technology to market by as much as five years, according to one participant.
Foreign patents: Filing foreign patents, particularly for chemical reagents, is an important but often overlooked component of technology transfer. At least one company representative expressed the view that universities have not been sufficiently diligent in prompt filing of foreign patents.
Interaction with NIH and NIH-funded investigators: NIH funds a significant amount of biomedical research, much of it on the cutting edge of science. Contact with investigators at NIH, and with researchers funded by the agency working in other organizations, has been an important factor in the success of the biotechnology industry.
In-house, state-of-the-art research capabilities: One workshop participant noted that discoveries made by a firm's own researchers can complement advances made by scientists outside the company, and vice versa. It is not possible, the participant noted, to concentrate on external sources of technology and information and to neglect in-house R&D capabilities.
The University Perspective
Universities are facing many new opportunities as a function of increasing ties to industry. At the same time, these relationships challenge the historic mission of university education and research. University-based biotechnology research has been characterized by its close ties to industry and the potential for commercial market applications. New sources of funding, educational opportunities, enhanced possibilities for contributing to the public good, and profit are among the many benefits to universities from their industrial affiliations. There are potential problems, as well, such as conflicts of interest, the loss of top faculty to the private sector, maintaining a proper balance between research and teaching, and new sources of liability.
The passage of the Government Patent Policy Act of 1980 (Public Law 96-517) gave universities new patent rights for inventions developed with federal funds. Prior to this act, only about 4 percent of the more than 30,000 patents held by the federal government were ever licensed. According to one participant at the workshop, patents for inventions developed at universities with federal funding are now licensed approximately 50 percent of the time—an example of how market-based incentives can enhance technology transfer. Workshop participants discussed a number of other factors that can influence the success of university-industry partnerships in biotechnology, including the following:
Speed of licensing: In biotechnology markets, licenses are often issued before a patent. The director of one university technology licensing office noted that, to issue licenses quickly, the institution involves lawyers only in the latter stages of the licensing process. In general, patents are not
as important in biotechnology as in other sectors since products are often made by biological organisms, which are difficult to reproduce.
Role of university scientists in the licensing process: University researchers who direct scientific projects often have insights about the selection of companies for licensing agreements. Participants noted that these researchers should be consulted early in the licensing process.
Product liability insurance: The high cost of liability insurance makes it difficult for universities to license innovations to small companies, particularly in biotechnology. According to several participants at the workshop, some universities support remedial product liability legislation to reduce potential damages in these situations.
Biological materials: Restricting the transfer of biological materials to other scientists was described as counterproductive to the goal of technology transfer.
The Role of NIH in the Transfer of Biotechnology
NIH has played an important role in the development and transfer of biotechnology products and manufacturing processes. The agency invests about $3.5 billion on biotechnology-related R&D, approximately 80 percent of the total spent by the federal government in this area. Approximately one-third of the $3.5 billion is spent on biotechnology-specific research; the remainder supports basic scientific research with wide biomedical applications.
NIH has several in-house units dedicated specifically to technology transfer. One is the Patent Policy Board, which has a number of working subcommittees, including one that reviews CRADA proposals and a second that focuses on royalty distribution. In addition, each of the agency's institutes has a technology development coordinator responsible for monitoring CRADA documentation and acting as a liaison with private firms on technology transfer activities.
NIH representatives at the workshop expressed the view that CRADAs are a productive method for small biotechnology firms to leverage internal R&D resources. The agency has approximately 130 CRADAs in place, of which approximately one-third are with small businesses. NIH also facilitates technology transfer by licensing patented materials and by training postdoctoral students and research fellows. NIH researchers are responsible for the publication of approximately 7,000 technical journal articles each year, as well as presentations at scientific workshops. Both of these activities involve technology transfer objectives.
It was reported that NIH also operates an electronic bulletin board containing lists of technologies available for licensing (identified and sorted by key words and names of researchers interested in collaborating) and that it
will allow texts of patent applications, copies of policy guidelines, and model agreements for cooperative research to be down-loaded to personal computers.
TECHNOLOGY TRANSFER IN A ''MATURE" INDUSTRY: AUTOMOTIVE ELECTRONICS
Universities have played a much smaller role in the technology transfer process in automobile electronics than in biotechnology, according to the director of an academic electronics laboratory. Instead, the automotive industry has relied on in-house R&D capacities and, more recently, worked with the aerospace, computer, semiconductor, and electronics industries.
The director of electrical engineering at one U.S. automobile manufacturer told the workshop that most development work in automotive electronics is performed either in-house or through private sector collaborative R&D projects. There is much less interaction with U.S. universities, he reported. With the exception of a project involving the Department of Energy to develop an electronic car, one participant noted that there has been limited interaction with federal facilities.
Participants also noted that in automotive electronics, the federal government has promoted technology transfer primarily through federal regulatory control. The Clean Air Act of 1970 and the establishment of the Environmental Protection Agency provided the incentives for the automotive industry to use electronics to control tailpipe emissions. Other incentives for developing new technology came with the establishment of corporate average-fuel-economy requirements. In the 1980s, market incentives led to electronic advances in antilock and antiskid braking, digital instrumentation, and intra-and extravehicular communications.
Although advances in aircraft design can in some cases be applied to automotive transportation, a number of barriers exist to this form of technology transfer. For example, under DOD sponsorship, aerospace electronics are developed and manufactured without cost considerations, a critical factor in private sector technology strategies. Similarly, performance requirements for the technologies differ. Differences in design methodology and management style between the two industries have also served to hinder technology transfer. For instance, only recently did the automobile industry start using a true systems engineering approach (standard practice in the defense industry) to develop new vehicles. Rapid technology transfer is also inhibited due to the fact that upstream electronics developers, particularly those developing defense-related technologies, have different cost and performance requirements. Frequently, auto companies must redesign electronics to meet these requirements.
Technology transfer in this sector has also been characterized by in-
creased collaborative R&D partnerships with suppliers. In the future, some workshop participants predicted, there will be increased emphasis on joint ventures, especially with foreign partners. Collaborative agreements with specific, applied technology goals may become more common. In addition, suppliers will be increasingly responsible for the development of new technologies, and acquisitions as a form of technology transfer may increase in frequency, according to participants. One official of a U.S. electronics company indicated that a number of other factors have contributed to successful technology transfer in automotive electronics. Many of these center on market incentives. For instance, many electronics firms have been willing to transfer technology to the automotive industry to help create new markets for their products. Automobile manufacturers view these developments as opportunities to enhance the customer appeal of their products through increased use of electronic components in autos.
One example of government support for automotive-related technology transfer discussed at the workshop is the Combustion Research Facility funded by DOE. The facility has a staff of approximately 80 government scientists and engineers, and an equal number of visiting staff from academia, industry, and other laboratories. Each year the facility hosts approximately 800 visitors, which facilitates the diffusion of technology. Automotive industry representatives help set the facility's research agenda through involvement in project R&D working groups.
FEDERAL EFFORTS IN TECHNOLOGY TRANSFER
Federal efforts to enhance technology transfer have increased over the past decade.6 Formal links between government facilities and the private sector remain limited, however, in contrast to intrafirm collaborative arrangements.7 One federal initiative discussed at the workshop, the National Center for Manufacturing Sciences (NCMS), involves R&D in support of the machine tool industry. NCMS, a consortium of 120 member companies established under the National Cooperative Research Act of 1984, provides technology outreach to small firms that make up the industry.
Technology transfer can also occur when an organization has specific product development requirements. In cases such as this, it may contract with a federal laboratory to carry out the necessary research. The Department of Energy is also moving to enhance its technology transfer activities, an agency official told the workshop.8 Among the mechanisms DOE is using to transfer technology to the private sector are Superconductivity Pilot Centers that have industry-driven research agendas and require cost sharing. Similar efforts are the Clean Coals Technologies Program, DOE involvement in the Small Business Innovation Research Program, the Energy Conservation Utilization Technology Program, the Advanced Manufac-
turing Initiative, and the Specialty Metals Processing Consortia. In the future, one official reported, DOE will increasingly interact with R&D consortia, state organizations, and other federal agencies in technology transfer activities.
DOE is also developing model agreements for technology transfer, as well as systems for conflict management, so that technical problems do not inhibit technology flow. A DOE official noted that much of the agency's future efforts in technology transfer will be framed by the National Energy Strategy, which among other objectives is targeted at expanding investment in basic research and increasing the number of scientists and engineers engaged in energy research.
The 1988 Omnibus Trade and Competitiveness Act established programs targeted at technology transfer, including the Manufacturing Technology Centers (MTCs) managed under the sponsorship of the National Institute of Standards and Technology (NIST).9 MTCs are designed to assist in modernization of the approximately 100,000 small (50 employees or less) parts manufacturers in the United States. MTCs are located in nonprofit and academic institutions to leverage existing state and local information networks. As of April 1990, three MTCs had been established.
Another NIST initiative is the Advanced Technology Program (ATP), which in 1989–1990 had funding of $10 million; this program of financial awards to industry and industrial consortia aims to speed the commercialization of emerging technologies.10 Workshop participants noted that, although the federal government can play an important role in technology transfer through these and other programs, the government should not determine what specific technologies industry should develop. Similarly, government should not have final authority over how industry and government collaborative ventures operate. These choices need to be made in partnership.
REGIONAL AND STATE TECHNOLOGY TRANSFER
There are important regional economic issues that affect the technology transfer process. One workshop participant argued that the U.S. economy should be viewed as a series of highly concentrated industrial regions. In certain areas of the United States, there is a critical mass of qualified personnel, public R&D support, investment capital, and a technology and manufacturing base to sustain competitive firms. These regions often include distinguished universities and a core of industrial firms that perform advanced R&D.11 Highly skilled scientists and engineers, for example, are often concentrated in these areas. The challenge for these regions, one participant noted, is to maintain and improve the R&D infrastructure to stimulate industrial development. Some areas in the United States are succeeding at this; a strong measure of success is the level of private industrial
R&D funding expended within a region. Where industrial R&D funds are concentrated, one participant argued, there has been strong regional economic growth.
One state technology development official noted that state efforts in R&D and technology transfer often focus on economic development, job creation and training, and technology commercialization activities.12 In contrast, federal efforts in this area have traditionally focused on basic research capabilities and R&D funding. At the state level, technology transfer management is often decentralized. Decisions on resource allocation are made by member firms, which are often required to invest corporate funds in technology projects that the state sponsors. States have adopted many approaches to stimulate technology transfer. Several have established research centers and distributed R&D grants to individuals and firms, similar to the individual investigator grants awarded by NIH and NSF. Others have invested pension funds in venture capital organizations that target technology commercialization.
Ohio's Edison Technology Centers and Pennsylvania's Ben Franklin Partnership Program are two examples of state technology transfer initiatives. Several of the nine Edison Technology Centers operate contract research facilities whereas others are housed at universities. All centers are operated as private, nonprofit organizations. Membership fees, charged on a sliding scale, can be as high as $60,000 per year. These funds are used both to upgrade the infrastructure of the centers and to fund member-directed research. By contrast, the Ben Franklin Partnership Program acts primarily as a technology extension agent, bringing firms together with specialists in corporate management, education, training, and R&D.13 Technical extension services operated by the states, it was argued, can also improve access to new technological information. According to some participants at the workshop, both small and large firms should be targeted by state and local extension services. A state industrial extension service, organized in a fashion similar to the farm cooperative model, would not only help improve technology transfer but also stimulate technology adoption by small and medium-sized manufacturers, participants noted.
SELECTED FEDERAL LAWS AFFECTING TECHNOLOGY TRANSFER
P.L. 94-282, National Science and Technology Policy, Organization, and Priorities Act of 1976: Outlines science and technology policy goals, establishes Office of Science and Technology Policy (OSTP), Federal Science and Technology Survey Committee.
P.L. 96-480, Stevenson-Wydler Technology Innovation Act of 1980: Directs Secretary of Commerce to establish Office of Industrial Technology; establishes Centers of Industrial Technology, and allows each center to assign intellectual property rights, licensing ability; allows secretary to make grants and enter into cooperative agreements to accomplish the above; directs National Science Foundation to provide assistance for establishing centers; allows centers to seek money from other agencies; establishes National Industrial Technology Board; requires federal laboratories to establish an Office of Research and Technology Applications; establishes in the Department of Commerce a Center for the Utilization of Federal Technology; establishes National Technology Medal; requires establishment by secretary and NSF of science and technology personnel exchange program; authorizes funding through fiscal year 1985.
P.L. 96-517, Government Patent Policy Act of 1980: Allows government-owned, government-operated laboratories to grant exclusive licenses to patents.
P.L. 98-462, National Cooperative Research Act of 1984: Allows joint R&D programs; limits damages to single rather than treble in antitrust suits.
P.L. 98-620, Trademark Clarification Act of 1984: Amends P.L. 96-517 to allow contractors to receive patent royalties for use in R&D, awards, or education; permits private companies, regardless of size, to obtain exclusive licenses.
P.L. 99-159, National Science Foundation Authorization Act for FY 1986: Repeals provisions for financial and other conflict-of-interest statements by NSF officials and employees; repeals prohibitions against outside employment and activities; prohibits public disclosure by NSF of certain industrial and business sources of information.
P.L. 99-382, Japanese Technology Literature Act of 1986: Amends Stevenson-Wydler Act to direct DOC to improve availability of Japanese technical literature to U.S. businesses, scientists, and engineers.
P.L. 99-502, Federal Technology Transfer Act of 1986: Amends Stevenson-Wydler Act to authorize government-operated federal laboratories to enter into cooperative R&D agreements with other entities; establishes Federal Laboratory Consortium for Technology Transfer; directs that federal laboratory science and engineering professional duties include technology transfer; requires cash award programs to be established; includes formulas for distribution of royalties from licensing or assignment of inventions (defense program laboratories excluded).
P.L. 100-107, Malcolm Baldrige National Quality Improvement Act of 1987: Amends Stevenson-Wydler Act to establish National Quality Award for U.S. companies.
P.L. 100-418, Omnibus Trade and Competitiveness Act of 1988: Amends
U.S. trade law provisions with respect to (1) U.S. trade agreements; (2) enforcement of antidumping provisions; (3) protection of intellectual property rights (Section 5171); (4) trade adjustment assistance; (5) changes in tariff schedules; (6) export promotion; (7) international debt; and (8) education and training programs to increase U.S. industrial competitiveness. Creates state and local clearinghouse coordinating body at Department of Commerce; changes National Bureau of Standards to National Institute of Standards and Technology; expands mission to include technology transfer activities (Manufacturing Technology Centers, Advanced Technology Program, State Technology Extension Services, and others).
P.L. 100-519, National Institute of Standards and Technology Authorization Act for FY 1989: Amends Stevenson-Wydler Act to establish Technology Administration in the Department of Commerce, headed by new Under Secretary for Technology; consolidated into Technology Administration are the Office of Technology Policy, Office of Commercial Affairs, NIST, National Telecommunications and Information Administration, and National Technical Information Service.
P.L. 100-656, Small Business Competitiveness Demonstration Program Act of 1988: Amends Small Business Act to set forth specified small business eligibility requirements with respect to the Small Business Administration small business and capital ownership development program and the award of government procurement contracts under the small business set-aside program.
P.L. 100-676, Water Resources Development Act of 1988: Authorizes U.S. Army Corps of Engineers laboratories and research centers to enter into CRADAs.
P.L. 101-510, Defense Authorization Act for FY 1991: Begins development and implementation of a National Defense Manufacturing Technology Plan; allows federal laboratories to enter into memoranda of understanding (MOUs) with intermediaries to facilitate cooperative work with small businesses; establishes model programs for national defense laboratories to demonstrate successful relationships between federal, state, or local governments and small businesses.
P.L. 101-189, National Competitiveness Technology Transfer Act of 1989: Extends technology transfer mission to DOE defense program laboratories; allows these facilities to enter into CRADAs.
P.L. 102-245, American Technology Preeminence Act of 1992: Amends Stevenson-Wydler Act to extend the Federal Laboratory Consortium mandate through 1996; allows sharing of intellectual property as a contribution to a cooperative R&D agreement; request DOC judgment on whether to allow federal contribution of funds to a cooperative R&D agreement; allows laboratory directors to make gifts of excess laboratory equipment to schools and nonprofit organizations.
COLLABORATION IN RESEARCH AND DEVELOPMENT: SELECTED EXAMPLES*
John S. Wilson
A research consortium is generally defined as an association of organizations involved in collaborative R&D projects. The goal of most collaborative research ventures is to leverage both scientific and engineering expertise, and financial resources. In practice, research consortia take many different forms, including interfirm collaborations, public-private ventures, and university-based projects.14 The relationships among consortia members, and their financial and other responsibilities, vary. Consortia participants often use different measures to assess the value and effectiveness of their involvement in collaborative R&D.
In March 1991, the project on the Government Role in Civilian Technology and the Academy Industry Program convened a workshop to assess recent experiences with research consortia. Leaders from industry, government, and academia participated in the discussions. The objective of the session was to identify both the characteristics of successful collaborative R&D projects and the obstacles to such efforts.
This paper summarizes the workshop discussions, highlighting key points and issues raised by participants. It does not represent the views or conclusions of speakers or participants at the workshop. This summary does not contain conclusions or recommendations of the Panel on the Government Role in Civilian Technology, Academy Industry Program, National Academy of Sciences, National Academy of Engineering, Institute of Medicine, or National Research Council.
Although the past decade has witnessed an increase in the number and diversity of collaborative R&D projects, such activities have been going on for many years in the United States.15 Development of the computer and the integrated circuit, for example, can be traced to research sponsored and coordinated by the U.S. government during the 1950s and 1960s. The nation's space program was built on collaborative R&D ventures involving government, industry, and universities. The biotechnology industry is largely the product of federally funded research carried out at U.S. universities.
The relatively recent increase in the formation of R&D consortia re-
flects new economic and technological conditions in the global economy.16 Principal among these is the quickening pace at which discoveries and inventions are applied to new commercial products and processes. The global diffusion of new technology and the results of research, combined with the enhanced manufacturing competency of nations, have resulted in intense domestic and international competition in many products and processes. From the innovation phase of technology development to the final commercialization, manufacturing processes and R&D have undergone significant changes over the past several decades. The effectiveness of collaborative R&D ventures should be viewed in light of these changing conditions.
The interactive character of the technology commercialization process is illustrated by the metaphor of athletic competition, as one participant at the workshop noted. Traditionally, the stages leading from invention to marketable product or process improvement have been described as legs in a relay race—sequential and largely discrete stages. Each step in this process involves different personnel, scientific and engineering resources, and facilities. Increasingly common, however, are simultaneous interactions among basic research, applied R&D, and product design and development. In a game, the ball is passed back and forth between team members, just as innovation is advanced by groups working in concert, regularly exchanging designs, prototypes, and manufacturing and marketing plans. In industries with short product life cycles, all elements of product or process development require quick turnaround. Time to market is a critical determinant of success.
Congress and the executive branch have attempted to create an environment in the United States that will accommodate this new dynamic.17 For example, in 1984, Congress passed the National Cooperative Research Act (NCRA), which relaxed antitrust regulations to permit firms in the same industry to collaborate on pre-commercial R&D. Since fiscal year 1985, more than 250 consortia, involving more than 1,000 U.S. businesses, have been formed. Although the number of ventures filing for exemption with the Department of Justice under NCRA has increased, the effect of the law on stimulating collaborative relationships among firms is less clear, some participants noted.
In several technology areas, the federal government has provided increased funding for public-private ventures. The most notable example is the Semiconductor Manufacturing Technology Research Corporation (SEMATECH), a consortium of 14 U.S. semiconductor manufacturers. SEMATECH, authorized for five years in 1987 with a $200 million annual budget, was charged with making the United States a world leader in semiconductor technology. SEMATECH's funding is provided by consortium members and the Department of Defense. Since SEMATECH, the federal government has launched several other joint industry-government R&D programs.
The Federal Coordinating Council for Science, Engineering, and Technology, in consultation with university and industry representatives, recently proposed a new government-industry collaborative venture: a High Performance Computing and Communications Program. The program would combine the efforts of federal agencies, universities, and U.S. businesses to extend the U.S. leadership in advanced computing and networking. It also would be targeted at accelerating the development and application of technologies for commercial, educational, and environmental uses.
Universities, which conduct about half of all basic research in the United States, have become increasingly involved in collaborative R&D activities. One-to-one relationships between faculty members and industry scientists are the most common form of these efforts. During the 1980s, there was also an increase in more complex, formal, university-based R&D partnerships with industry. The National Science Foundation has also sponsored university-based R&D centers (the Engineering Research Centers), which serve as focal points for collaborative projects between academia and industry.
Promise and Outlook for Consortia
In the changing environment for technological and economic competition, consortia may complement other methods of strengthening U.S.-industry research and development. For private firms, the advantages of collaboration include reducing the risk and costs of R&D work, eliminating duplication of effort, leveraging internal resources, and gaining access to technology and expertise not available in-house.18 Collaborative R&D projects are often centered on long-term, applied research. Several workshop speakers suggested that collaborative projects should focus on research horizons of two to seven years, the shorter time frame for industrial R&D, and the longer for basic research.
Consortia have been proposed as a mechanism to link public and private sector activities for promoting national economic interests. To date, however, most R&D consortia should be viewed as experimental, participants noted. Their value as strategic tools for altering the dynamics of industrial development and technological advance is unclear. Proprietary research conducted in industrial laboratories remains the primary focus of industrial R&D activities in the United States and most other nations. In Japan, for example, most research is performed within individual firms, and of the one-third of projects classified as collaborative R&D, most involve firms that do not compete in the same product markets.19
Several workshop participants noted that consortia face many of the same pressures and obstacles that confront in-house research programs. They pointed out, for example, that collaborative ventures must produce quick
results, even though one rationale for forming partnerships is to lengthen research horizons. Moreover, participants noted that the transfer of research results to member firms is more complicated than simply passing an innovation to an industrial sponsor. They also stressed the importance of sustained interactions between consortium researchers and those in industry who are involved in process and product development. One participant suggested that describing consortia as R&D centers is counterproductive because it implies a separate island of activity rather than a continuous innovation cycle.
PRIVATE SECTOR COLLABORATION
The United States' experience with collaborative R&D extends to the early post-World War II era, when U.S. industry first became involved in research—focused on supercomputers, aircraft, and semiconductors, for example—to develop technologies for use in military hardware. Collaboration between industries associated with aerospace efforts of the National Aeronautics and Space Administration in the 1950s spurred joint R&D efforts in the commercial aircraft industry. Prior to the 1970s, collaborative industrial R&D efforts that were not directly supported by the government occurred among companies in vertical sectors. Such was the case with automobile manufacturers and petrochemical firms collaborating to develop ceramics for use in auto bodies, for example.
Today, joint R&D consortia involving manufacturers in horizontal business sectors—such as semiconductors, chemicals, advanced materials, and telecommunications—are on the increase. The Semiconductor Research Corporation, Microelectronics and Computer Technology Corporation (MCC), and Software Productivity Consortium, are examples of collaborative efforts that bring together companies in similar product markets.
Prior to 1984, U.S. antitrust law limited horizontal R&D collaboration. Collaboration among firms in the same industry, therefore, was most often carried out under the sponsorship of trade associations. Trade association activities included work to establish industry-wide technical standards, for example. Regulated industries, such as the gas, electrical power, and telecommunications sectors, were permitted to engage in collaborative ventures because firms in these industries were not in direct competition.
Private Sector Perspectives
Representatives of several firms at the workshop described collaborative R&D ventures, relatively uncommon during the early 1980s, as an integral part of today's corporate research efforts. One research director reported that his company considers joint R&D projects so important that it
has made collaboration a core competency objective of the firms' competitive strategy. To that end, the firm has developed a process model for planning, guiding, and evaluating collaborative projects. Under this framework, the firm cooperates with suppliers, participates in multicompany consortia, and is active in ''centers of excellence'' that it sponsors on university campuses. The method used by the firm to assess its success in R&D collaboration is patterned after the total quality control model employed in manufacturing processes. Among other features, it allows the company to measure how its performance compares with that of other firms involved in collaborative research. Workshop discussions revealed several common motivations for collaborative research. The importance attached to each anticipated benefit, outlined below, varied among firms.
Acquiring needed expertise: Firms with diverse product lines in rapidly evolving industries often require the R&D expertise of many scientific and engineering disciplines. Even companies with large R&D budgets, however, are not likely to have all the necessary personnel in-house. Collaborative ventures give firms access to outside sources of talent, which cost the company less in terms of both human and financial resources than R&D work performed in-house.
Leveraging research investments and building "critical mass": Cost sharing enables firms to both intensify efforts in specific R&D areas and expand their research agendas. For large projects that are too expensive and too risky to undertake alone, consortia may be the only alternative to forgoing promising research inquiries.
Building technical competence, monitoring technological progress : R&D collaboration can help a company build research competency in areas of technology development where its competitors currently have the lead. Similarly, collaborative activities permit firms to monitor progress in fields important to their future technological advancement.
Acquiring intellectual property: Even in areas where a firm leads the competition, collaborative research may, through the generation of ideas for new products, strengthen the company's market position. By adding to its portfolio of patented intellectual property in a particular area, a firm can enhance its competitive advantage.
Improving research efficiency: Noting the problems associated with the "not-invented-here syndrome," several corporate research directors stated that research needs, in some instances, can be addressed more efficiently by working with outside organizations. If fundamental proprietary advantages are not sacrificed, collaboration may be the best or only way to accomplish specific research goals.
Facilitating the development of standards: In the computer and
microelectronics industries, the need for interoperable hardware and software facilitates R&D collaboration. Hardware manufacturers and software publishers must work together to develop industry-accepted standards for their products.
Elements of Effective Collaboration
Before a company establishes collaborative R&D ventures, one research director at the workshop maintained, it must have a clear vision of where it wants to be two, five, and ten years in the future. It must have short-and long-term R&D strategies. Firms should also fully understand their relative strengths in areas viewed as critical to current and future success. On the basis of that understanding, a firm can determine how best to allocate its research funds and whether accomplishing specific objectives will require work with outside organizations.
Workshop participants suggested that the most successful collaborations are based on a close working relationship among researchers. They stressed, however, that collaborations must not be viewed strictly as activities involving research personnel. Collaborative projects should have "champions" at each level within a company or operating division. Close and continuing contact between scientists and the managers or technical personnel who will eventually apply research results to products or processes is essential to technology transfer.
One objective of collaborative research is to advance the state of the art in a particular area of science or engineering. To do this, several participants stressed, collaborative R&D ventures must have clearly defined goals. One company research director suggested identifying broad areas of interest and then selecting a few projects to focus on. If the focus of a collaborative project is wide, priorities may be obscured and the effectiveness of the effort may be diminished.
Returns to Cooperative Research
Collaborative projects are judged on the basis of their accomplishments. A new product or more advanced technical application for existing product lines, manufacturing process technology improvements, and enhanced scientific understanding of physical or chemical processes all reflect successful collaboration. Not every project will "succeed" according to these criteria, one computer R&D laboratory director noted. Other measures of progress, such as a better understanding of the topic being investigated, are also important. Periodic and end-of-the-project evaluations can yield insights into improving future collaborative endeavors, he noted.
Perspectives of Consortium Managers
R&D consortia offer a number of potential benefits to member firms, according to one consortium official involved in semiconductor, computer, and software systems R&D. Efficiencies result from sharing the costs and risks of research and from reducing R&D duplication. Consortia can also monitor and fund university research, as well as serve as vehicles for integrating and transferring the results of that research. Collaborative R&D arrangements may promote the development of standards for specific technologies. In the semiconductor and information technology industries, for example, collaboration can advance the development of standards and protocols for hardware, software, and communication networks. One such industry-wide activity, undertaken by SEMATECH, qualifies and certifies vendors that supply materials and equipment to member firms.20
Outside Pressures and Changing Circumstances
To be successful, collaborative research ventures must be structured and operated in ways that are consistent with the realities of corporate R&D, participants noted. For example, although consortia are often viewed as mechanisms to facilitate long-term research, most companies invest in short-term R&D, an emphasis that may influence how they judge the value of investments in collaborative R&D projects. Rather than focusing exclusively on long-term projects, it was suggested consortia should accommodate a mixture of projects with varying time horizons. Plans for R&D projects should be reviewed periodically and revised to reflect circumstances, such as changes in membership. Consortium managers and researchers should recognize that their organization represents only one of several options available to companies pursuing their corporate research goals. Finally, it was noted that consortia management needs to understand that breakthrough research advances may be less important than continuous, incremental improvements in processing, product quality, marketing, and customer service. Understanding these elements of success can lead to productive redirection of consortia activities.
Enhancing Technology Transfer
A plan for transferring the results of research conducted by consortia must be established, in as detailed a manner as practical, at the start of the collaborative venture, according to several workshop participants. Participants also suggested that R&D sponsors maintain regular contact with researchers to facilitate the exchange of ideas, prevent surprises, and foster
company ownership of the research products. It was stressed that the results of collaborative research can rarely be used immediately by sponsors. Typically, additional investment and effort are required to commercialize a technology or to apply a new processing technique effectively. Firms not familiar with the way research results are transferred to marketable products are unlikely to make this important allocation of resources.
In one innovative approach to technology transfer, the Microelectronics and Computer Technology Corporation, in Austin, Texas, is creating start-up companies to commercialize the products of consortium research. MCC owns a small stake in the new firms, and consortium members that sponsor research leading to a new product are charged preferential rates if they license the technology. One workshop participant suggested that this approach may foster better vertical integration between suppliers and member companies.
Sharing Technology and Information
There have been problems associated with technology and information sharing in most collaborative ventures. For example, some members of R&D projects have been reluctant to delegate highly qualified staff to collaborative projects. Others noted that firms in collaborative R&D projects face difficulties in sharing technology and technical information with other members. Workshop participants observed that in most collaborative ventures, participating companies must share intellectual property with other firms. There is a clear risk of losing proprietary technologies to competitors; however, firms should agree to a set of rights and responsibilities at the outset of projects. Clear intellectual property rights guidelines cannot, however, eliminate all risks. In the case of a firm sharing technology with a key supplier, for example, there is the danger that the supplier will sell this newly gained knowledge to other customers, who in turn compete with the innovating firm.
Based on one research director's experience, however, there is little evidence that collaborative R&D projects lead to the loss of key technologies or proprietary information. In the computer and microelectronics industries, he reported, product life cycles are usually so short that, even if a new technology is revealed to other firms, the competitor is still not likely to be the first to market. When firms are successful in taking advantage of intellectual property developed by another company, any advantage is usually short-lived. The exception, he noted, is for breakthrough technologies that create entirely new business opportunities and markets.
Number of Participating Firms
An executive from one firm suggested that as the number of sponsoring firms grows, the value of the research diminishes for each sponsor. Conversely, he noted, if a particular area of R&D is deemed critical to a company's success, the firm will be more likely to perform the work in-house. Therefore a small number of participants in a venture might be more effective than large numbers of firms, it was suggested. Research activities conducted under this framework might be more closely aligned with the sponsors' interests, and therefore the motivation for close involvement would be stronger. Compared to small (two-to four-member) collaborative projects, partnerships with many sponsors have more difficulty defining common interests and goals. Industry representatives and consortium managers reported, however, that there are no specific rules for determining the optimal number of collaborators for specific projects. One participant noted that individual firms must determine whether a partnership is required to accomplish a specific R&D objective. If such an arrangement is necessary, companies should then determine the specific type of collaboration that will best serve the firm's long-term R&D needs.
Collaboration Beyond Research
Technological innovation is only one contributor to a firm's competitive performance. Several workshop participants maintained that consortia should broaden their agendas.21 For example, one participant suggested that collaborative R&D efforts include pre-competitive areas and initiatives to promote vertical integration between suppliers and customers. Some participants argued that U.S. antitrust laws block collaborative activities in areas aside from R&D. Revisions that clarify legal limits on collaboration, specifically for joint production ventures, would be helpful, they suggested.
U.S. universities are an important source of knowledge that supports the technology development process. Although universities account for only 9 percent of public and private research funding in the United States and employ just 10 percent of the nation's scientists and engineers, they conduct approximately 50 percent of the basic research performed in the United States.22 Since the end of World War II, university research, funded largely by the federal government, has generated new ideas and scientific knowledge that has assisted in the development of industrial products and processes. For more than two decades, university research and the applied R&D and commercialization activities undertaken by industry were able to sustain U.S. leadership in many areas of technology.
In the 1970s, global competition in many industries intensified, and the time to commercialize the results of basic research began to decrease. Foreign competitors became increasingly proficient at converting ideas (some of which were the result of research conducted in U.S. universities) into marketable products. During the 1980s, U.S. firms sought to strengthen relationships with academic researchers, and the number of collaborative ventures between business and universities increased.
Perspectives on University-Industry Collaboration
Industries vary in their motivations for forming research partnerships with universities. Industry perceptions of potential returns, an important influence on whether to collaborate, are driven by a number of factors. These include the specific technical field in which the collaboration occurs, the stage of scientific and technological development within the industry, and whether the research efforts involve basic scientific inquiry or applied R&D. Workshop participants from computer and microelectronics firms, for example, noted that universities are often promising sites for multidisciplinary research ventures involving several firms and teams of university scientists. In contrast, executives from pharmaceutical and biotechnology firms reported that their partnerships with universities were almost exclusively one-to-one collaborations between individual researchers.
Regardless of the industry sector involved, research directors reported that they rely on collaborative research relationships with universities to generate innovative ideas and understanding that might assist long-term corporate R&D strategies.
In contrast to microelectronics firms, companies in the biotechnology and pharmaceutical industries are more likely to view university research as the source of discoveries leading to new products.23 Firms in biotechnology are likely to allocate larger shares of their R&D budgets to support collaborative ventures with universities than are firms in the more established pharmaceutical industry. In absolute terms, however, large pharmaceutical manufacturers allocate significantly larger amounts of funds to university-based research programs.
Pharmaceutical and Biotechnology Industries
The experiences of major pharmaceutical firms and biotechnology companies, discussed at the workshop, illustrate the variety of strategies and expectations that guide firms' interactions with universities. Pharmaceutical firms often pursue a strategy in which partnerships with universities include many different R&D projects, thereby improving prospects for major discoveries.
Between late 1988 and early 1991, one company formed an average of one new collaboration per month to support its drug discovery efforts. These collaborative ventures were formed in addition to grants to academic investigators and departments, contracts for clinical evaluation of new drugs, and other types of relationships with universities. Collaborators were selected on the basis of careful review, which includes an assessment of how well the goals of the university partners matched the targets specified in the pharmaceutical company's drug discovery portfolio. In biotechnology firms, relationships with universities are often influenced by the industry's strong dependence on basic research conducted by academic and government investigators.
Today, many established and start-up biotechnology firms continue to view universities as a basic research arm, while they devote nearly all of their in-house efforts to applied research. One biotechnology company, for example, funds more than 100 collaborative ventures with universities in the United States, Western Europe, Japan, and Canada. The company also supports academic researchers in areas closely aligned with its product development goals. These projects, however, constitute a small part of its collaborative research arrangements with universities. Company resources are devoted to identifying and licensing the products of university research that have commercial potential. Moreover, it was noted that almost all first-generation biotechnology products, such as human growth hormone and insulin, can be traced to university research.
Through collaborative R&D, firms maintain access to technology developments in academic laboratories. Many therapeutic products were developed by these broad-based research relationships from collaborative ventures with universities, participants noted. In contrast, less product development for pharmaceutical firms has resulted.
Characteristics of Effective Collaboration
Successful collaborative ventures involving industry and universities usually begin with close interaction between personnel in each institution. Several workshop participants indicated that laboratory scientists are more likely to identify the benefits of joint research than company managers and university administrators. Scientists also are more likely to devise specific research plans, participants asserted. One participant advised against negotiating financial and legal details of R&D ventures until scientists have established a tentative research agenda.
Industrial R&D managers have an important role to play in ensuring successful collaboration. Ideally, managers should nurture relationships with academic researchers and their institutions, offer direction for research projects as required, and have their performance evaluated accordingly, one
participant noted. Several university and industry representatives strongly urged that those involved in collaborative research set goals at the outset. In contrast to the situation even a few years ago, reported one former university administrator, there is now widespread agreement that academic and industry research partners must agree on technical or scientific tasks to be accomplished. Prospective R&D partners also should attempt to anticipate problems that could undermine the collaborative relationship. Differing views on intellectual property rights and the publication of research results were two frequently mentioned sources of friction.
Issues and Concerns
Workshop participants were in general agreement that collaboration between U.S. firms and universities has the potential to enhance the competitive performance of participating companies. These arrangements can also benefit academic institutions, their faculties, and students. Some participants argued that foreign firms have been more successful than domestic companies in realizing benefits from collaborative R&D involving U.S. universities. Considerable disagreement was expressed during workshop discussions concerning the responsibilities that partners in university-industry ventures should undertake.
Impediments to Negotiations
Companies' experiences in forging collaborative relationships with universities varied widely. Several participants from industry reported that negotiations with university administrators are often protracted and agreement is more difficult to reach than with prospective private sector collaborators. Representatives of small firms, who reported few problems, suggested that the delays encountered by large firms are related to problems with corporate, rather than university, bureaucracies.
The question of intellectual property rights to the discoveries and inventions produced by collaborative ventures elicited considerable debate. An executive of a computer manufacturing firm suggested that when universities enter into intellectual property agreements with businesses, their bargaining position is influenced by past experiences with drug companies. The result can be a "one-size-fits-all" policy for dealing with intellectual property that is not appropriate for all types of R&D. Often, in industries other than pharmaceuticals, patents are cross-licensed between firms and serve as a negotiating point to ensure that each company has access to the other's technology.24 In contrast, a patent for a new drug confers monopoly power on the owner and therefore is more valuable, warranting royalty payments that might be deemed excessive in other industries.
Other business executives maintained that universities often do not appreciate the considerable expense that companies incur as they go through the many steps involved in converting a discovery into a marketable product. Typically, they said, these costs exceed by at least ten-fold the amount spent on initial research. Moreover, industry executives pointed out, not all discoveries are equivalent. In the case of pharmaceuticals, for example, when a new drug is discovered by the academic partner in a university-industry collaboration, the company is usually granted royalty payments. On the other hand, if company-funded collaborative research yields a new tool or processing method, the sponsoring firm is likely to expect a royalty-free license, although it might allow the university to license the tool or method to other companies.
Not all participants from industry found university procedures for handling intellectual property issues difficult to manage. They reported that university negotiating positions involving technology licensing had not been a significant barrier to collaboration. One university representative maintained that many academic institutions differentiate between types of discoveries and inventions. Most also recognize that the value of patent protection varies among industries, he argued. Several academic officials noted that universities are working to develop approaches for negotiating intellectual property rights that take institutional differences into account. Conflicts over intellectual property rights have, however, presented barriers to successful university-industry cooperative ventures.
Access for Small Firms
Small firms often have limited resources, research capacity, or experience with collaborative ventures. Some participants at the workshop noted that these problems are associated with small firms' difficulties in forming R&D partnerships with universities. As a consequence, one speaker asserted, only large companies have access to discoveries and inventions arising from university-based research. While acknowledging the need to be alert to this concern, participants from pharmaceutical and biotechnology companies reported that small firms appear to have access to university research and its products. One participant noted that, in some cases, small firms succeed in licensing a university-owned invention but then lack the resources to carry out the work necessary to bring the product to market.
Commitment of U.S. Firms
Some industry participants questioned the commitment that domestic firms make to collaborative ventures with universities. It was noted that foreign firms appear to benefit to a greater degree than U.S. companies
from partnerships with U.S. academic institutions. One research director for a major U.S. multinational firm maintained that American companies typically do not bring the same level of dedication, diligence, and interest to university-industry collaborations as do their foreign counterparts.
An executive from a foreign-based multinational firm maintained that foreign firms do not enjoy preferential access to U.S. university-based research. Rather, he reported, European and Japanese firms more fully recognize that they have an obligation when they enter into relationships with universities. That is, the foreign firms realize that they must bring innovative ideas to the collaboration and that it is their responsibility to take and use discoveries or inventions that result from it.
One speaker suggested that U.S. firms may be more diligent partners in the growing number of relationships they are forming with research institutions in other countries. This may result, in part, from the greater effort required by U.S. companies to cultivate ties with unfamiliar overseas partners.
Conflict of Interest and Public Perceptions
Real and perceived conflicts of interest can present obstacles to collaborative relationships between universities and industry.25 Potential conflicts are common, according to one university representative, and the challenge for both parties is to manage them appropriately. One set of concerns involves the level of financial compensation for products resulting from research conducted at universities. Some participants argued that research supported by the government should mean that the results of such research are publicly available. Conflict may also arise when university faculty serve as consultants to industry, an activity viewed by some as in conflict with education goals.
Moreover, the question of whether foreign firms should have access to the results of research conducted at U.S. universities continues to be debated. This concern reflects the larger issue of foreign participation in U.S.-based collaborative ventures and, conversely, the participation of U.S. firms in foreign R&D ventures.
The government accounts for nearly half of the total R&D spending in this country. It is the nation's largest single employer of scientists and research engineers and provides funding for most basic research conducted by universities. In addition, the more than 700 federal laboratories employ nearly one-sixth of the scientists in the United States. Defense R&D constitutes a large percentage of the total federal R&D budget, accounting for more than 60 percent of such expenditures during the last decade.
Some policy analysts and business leaders have suggested that U.S.-economic interests would be better served if the government assumed a broader and more active role in civil technology development. One industry executive at the workshop noted that many of the ''critical technologies'' identified by various groups will probably not contribute to the nation's defense. Yet progress in many of these key areas, he argued, is crucial to the performance of many U.S. industries, to national economic performance, and to improvement in the U.S. standard of living.
Since the late 1970s, the federal government has provided incentives to encourage the transfer technology from federal laboratories to the private sector; it established SEMATECH, created NSF Engineering Research Centers, and established the Advanced Technology Program at the National Institute of Standards and Technology.
The ATP is centered on the development of pre-competitive, generic technology with significant commercial promise.26 In fiscal year 1991, ATP awarded some $9 million to support 11 industry-led R&D projects in such areas as x-ray lithography for semiconductor manufacturing, high-temperature superconductivity, flat-panel display manufacturing, and optical recording.27
Federal Laboratory-Industry Collaboration
There is considerable debate about the utility of federal initiatives, the most productive modes of collaboration, and the need for a national policy for technology transfer.28 Participants at the workshop discussed the fact that technology transfer requires sustained, market-driven technology programs. As one laboratory director explained, early technology transfer efforts were guided by the mistaken assumption that federal scientists and engineers had created a vast wealth of technology with commercial applications. Measures to improve private sector access to the laboratories, it was believed, would yield a stream of off-the-shelf technologies that could easily be converted to commercial applications. That has not been the case, several participants noted. Most of the technology at the federal laboratories, according to one participant, was developed to support the missions of federal agencies rather than to address commercial needs. In recent years, Congress has passed measures to create incentives for laboratory-industry collaboration and technology transfer. For example, federal scientists and engineers can now collect royalties on patented inventions they create.29 Procedures for licensing federal technology have been simplified, and industry sponsors of collaborative research conducted at national laboratories can own or secure exclusive licensing rights to inventions arising from these joint efforts. Nonetheless, the missions of many of the federal laboratories do not reflect the commercial technology needs of industry. One
participant noted that the NIST is the only federal facility with the explicit mission of assisting U.S. industry.
Some federal agencies that conduct research are taking steps to foster collaborative relationships with industry. The Department of Energy, for example, has explicitly included technology transfer in the mission statement of each of its laboratories.30 DOE laboratories are now evaluated on the basis of how successfully they work with industry. In addition, the department's decisions on investments in capital and equipment are guided by its desire to make the laboratories better partners with industry. Another participant noted, however, that mission statements can be changed easily. Other, more difficult changes—in facilities, staff, and location—will be needed if federal laboratories are to fulfill a role in technology development and transfer, they noted.
SEMATECH's goal is to develop a domestic capacity for world-class levels of semiconductor manufacturing by 1993. The federal government supplies half of the consortium's $200 million annual budget; SEMATECH's 14 member companies provide the remaining funds. The five-year experiment conducts in-house research and funds outside R&D projects. In recent years, SEMATECH has spent a growing portion of its budget on R&D projects conducted by a group of 140 equipment and materials manufacturers that make up the consortium's fifteenth member, SEMI/SEMATECH. These outside R&D projects now comprise half of SEMATECH's research budget. SEMATECH provides about 10 percent of the funding for equipment improvement projects and about 30 percent of the support for joint development projects focusing on new equipment. Collaborating companies provide the rest. The consortium also funds research at its 11 university-based centers of excellence and at several national laboratories. In-house research is carried out by the consortium's 700-person staff, one-third of whom are on loan from corporate members.
In addition to its research program, SEMATECH is working to improve relationships between semiconductor manufacturers and domestic suppliers. For example, SEMATECH's corporate members conduct qualification tests of new equipment at a single site. Representatives from other firms come to the test site, where they evaluate equipment performance and provide feedback to the supplier.31
Workshop participants noted that the consortium appears to be achieving its technical goals on schedule. By mid-1993, the end of its original
five-year charter, SEMATECH is expected to reach its goal of producing memory chips with 0.35-micron circuitry at its pilot fabrication plant in Austin, Texas, using equipment from domestic U.S. suppliers. One speaker predicted that with SEMATECH's contributions, U.S. semiconductor manufacturers and their suppliers will reach parity with the Japanese in equipment and some processes by 1993. Some participants argued that the United States should have firms capable of producing world-class quality equipment for each stage in the semiconductor manufacturing process: lithography, furnaces and implantation, etching, planarization, and deposition.
Some participants at the workshop reported that semiconductor manufacturers are benefiting from the consortium's programs. For example, two firms have used technical information from SEMATECH's pilot fabrication facility to guide planning of new manufacturing plants. Other participants have used SEMATECH's technical expertise to guide the purchase of new equipment. Workshop participants discussed plans for SEMATECH after 1993, when its initial five-year authorization ends. Some suggested that current SEMATECH programs could be extended to new areas of manufacturing. One such possibility, a participant noted, is "clean sheet" designs for factories built to manufacture the next generation of high-density memory and logic chips. SEMATECH might also be used to compare the advantages of costly, large-scale production facilities with the benefits that might be achieved from small fabrication plants.
Government-Sponsored R&D Collaboration in Other Countries
Government-led collaborative R&D efforts in other nations have received significant attention.32 Collaborative programs in Europe were discussed at the workshop. The changing nature of Japan's collaborative R&D programs was also discussed, as programs that initially focused on applied R&D projects increasingly emphasize basic science and engineering.33
The European Community
The long-term research programs jointly sponsored by the governments and businesses of the European Community (EC) are an expression of Europe's emphasis on new technology.34 According to one participant, these R&D projects also reflect the recognition that technology development requires substantial economic support, that the scale of the required efforts often outstrips the capacities of individual companies and nations, and that the globalization of technology shortens technical advantages. Many European firms view collaborative R&D as a way to achieve a competitive edge. This widely held perception distinguishes European companies from their American counterparts, which are more likely to pursue individual R&D initiatives exclusively with internal resources.
Many of the barriers to collaborative R&D that European nations are working to overcome, such as cultural and language differences and competing national interests, do not inhibit U.S. collaborative ventures. While acknowledging that transnational collaborations can lead to administrative confusion and the waste of resources, several participants asserted that European countries have built a viable framework for collaborative R&D. EC nations have forged a consensus on the technological areas most vital to future industrial competitiveness and are allocating their R&D support accordingly.
Government-led cooperative R&D in Japan has evolved through three stages since 1950. During the 1950s and 1960s, the country's efforts focused on creating the science and engineering foundation for economic development. Initially, government support for R&D was directed to individual firms, which contributed between 50 and 70 percent of project funds.
To stimulate research efforts, some participants noted, the government created a network of trade organizations, the Engineering Research Associations. With government support and guidance, the associations organized collaborations among small and midsize companies to address common technical barriers. Industrial trade associations remain an important feature of collaborative R&D in Japan. During the 1970s, the focus of government-led collaborations changed. Efforts were directed at elevating the performance of Japanese companies in technology generation. Collaborative projects were designed to refine advanced technologies. The specific goals of each collaborative venture were developed by the government in close consultation with the trade associations.
The third stage of Japan's government-led collaborative R&D efforts came as Japanese companies began to dominate world markets for many high-technology consumer products. This current phase of Japanese government-industry research focuses on the development of next-generation technologies. One recent initiative is the government-funded Key Technology Center (KTC) program, which has provided support for more than 60 projects since 1985.
The Key Technology Center (KTC) Program
In contrast to earlier initiatives focused on next-generation technology, the KTC program involves high-risk R&D. The program functions as a venture capital fund, providing seed money in the form of government-owned equity shares to support industrial collaborations, which typically involving 10 or fewer companies. (Participation of foreign-owned subsidiaries in KTC programs is not prohibited.) Participants are expected to
produce commercially useful technology and to operate profitably over a seven-to ten-year period.
The Optoelectronic Technology Research Corporation (OTRC) is an example of this new approach to cooperative R&D. Founded in 1986, the center has 13 member companies and an annual budget of about $7.5 million, 70 percent of which is supplied by the Japanese government. OTRC's research agenda is the product of a planning process involving member companies, government, and universities. The goal of the center's research program is to develop the knowledge and tools to design and manufacture integrated circuits that combine electronic and photonic technologies.35 Optoelectronic integrated circuits are expected to be important components of future communications and computing equipment. R&D work at OTRC is divided between the center's laboratory in Tsukuba Science City, outside Tokyo, where 20 scientists focus on process-related issues, and the in-house laboratories of member firms, where collectively some 30 researchers work on various devices and applications.
Technology transfer between member firms is promoted by the dissemination of reprints of articles and other presentations by center scientists, an annual workshop for company representatives, and semiannual panel discussions in which government and university scientists also participate. To date, the center's R&D work has not produced any inventions that have resulted in patents. Should patented technologies result from the research, decisions on licensing the inventions to nonmember firms will be made on a "case-by-case basis." Rights to patented inventions, however, are assigned to the member firms, not the government.
Issues for Consideration
Several workshop participants suggested criteria for evaluating the merits of proposed collaborative ventures and defining the roles of prospective partners. Two criteria that might be used to select areas of collaboration are (1) high levels of technical and financial risk and (2) a chance of high returns for successful ventures. Without the incentive of profit, individual businesses will not undertake R&D on potentially risky projects.
There are other mechanisms by which technologies developed through government-supported collaborations can lead to product refinement and commercialization. A clear lesson of the past, participants noted, is the need to pull technology from R&D ventures. Historically, an industry representative noted, government-funded projects have been operated under the assumption that technology will flow to industry, which will then commercialize the product or process application. As was noted often throughout the workshop, however, this expectation leads to less than successful technology transfer outcomes.
One participant noted that the government will need a way to assess the "value added" by each prospective partner in a collaboration. The personnel and technical contributions of collaborating firms must enhance prospects for achieving project goals and, ultimately, the technological advantage that motivated the cooperative effort. Moreover, steps must be taken to ensure the participation of small companies, an important source of new ideas and innovations. Government loans or loan guarantees might be used to encourage innovative small companies to participate in cooperative projects. In addition, it was suggested, carefully crafted tax incentives might encourage large U.S. companies to assist smaller businesses as they work to develop new technologies and products. Finally, many participants recommended that collaborative projects should be required to meet a series of technical milestones and should be monitored periodically for progress.
Participation of Foreign Companies
As participants noted, it is increasingly difficult to distinguish between U.S. and foreign companies. Just as U.S.-based firms have subsidiaries overseas, foreign corporations have operations in the United States that employ U.S. workers and provide federal and state tax revenue. Some public and private sector representatives at the workshop suggested that if a foreign-owned U.S. subsidiary pays taxes in the United States, it should be eligible to participate in government-supported R&D collaborations. It was also noted that foreign participation in collaborative research can benefit U.S. companies. In many key areas, workshop participants said, foreign firms are at the forefront of technological know-how. To be competitive, U.S. companies must draw on these repositories of expertise.
APPENDIX B Legislative Request for the Study
OMNIBUS TRADE AND COMPETITIVENESS ACT (P.L. 100-418)
"(2) The Committee shall render to the Secretary and the Congress such additional reports on specific policy matters as it deems appropriate.
"(c) NATIONAL ACADEMIES OF SCIENCES AND ENGINEERING STUDY OF GOVERNMENT-INDUSTRY COOPERATION IN CIVILIAN TECHNOLOGY.—
"(1) Within 90 days after the date of enactment of this Act, the Secretary of Commerce shall enter into contracts with the National Academies of Sciences and Engineering for a thorough review of the various types of arrangements under which the private sector in the United States and the Federal Government cooperate in civilian research and technology transfer, including activities to create or apply generic, nonproprietary technologies. The purpose of the review is to provide the Secretary and Congress with objective information regarding the uses, strengths, and limitations of the various types of cooperative technology arrangements that have been used in the United States. The review is to provide both an analysis of the ways in which these arrangements can help improve the technological performance and international competitiveness of the United States industry, and also to provide the Academies' recommendations regarding ways to im-