Partnerships in Research: The Evolution of Expectations
Christopher T. Hill
George Mason University
In this chapter, I address the evolution of cooperation among companies and between companies and universities in R&D, discuss what participants have expected from this cooperation, and examine the roles government has played in facilitating, responding to, or driving cooperation. My core argument is that we have experienced very substantial changes in how we think about R&D and that our different ways of considering R&D have been accompanied by different expectations about R&D cooperation.
My approach is historical and chronological, considering five eras over the past 75 years. These eras are the pre-World War II era, the postwar era through the late 1960s, the post-oil-embargo era of experimentation, the competitiveness crisis era of the 1980s, and the present era of industrial restructuring. Each of these eras features different approaches to and expectations from cooperative R&D.
The Pre-World War II Era
The pre-World War II era witnessed a transition from industrial technology based largely on the work of inventors like Edison, Ford, and the Wright brothers, to the development of new technology by separately organized R&D departments in large corporations. American universities were only beginning to challenge Europe in fundamental chemical research and the production of doctoral-level chemists and engineers. The government role in R&D was limited. Government arsenals did a modest amount of research in support of the Army and Navy, but it was widely agreed that the federal government should not support fundamental research in universities because its influence would be corrupting. The government did support some applied academic research in agriculture and mining.
Before the war, formal R&D cooperation between universities and industry was rare. To be sure, early leaders like Arthur D. Little at the Massachusetts Institute of Technology (MIT) and Charles F. Burgess at the University of Wisconsin built industrial chemical corporations on the foundations of their academic research, and academic engineering education was quite closely related to industrial practice.
However, an effort by Karl Compton, the president of MIT, to raise $3 million from a number of firms to support academic R&D was unsuccessful. Some attempts were made in the U.S. Congress in the mid-1930s to create a program of government support of industrial research, but these attempts were roundly rejected, not only on ideological grounds but also because influential members of Congress believed that important inventions came from the minds of individual creative men like Edison, and that organized R&D could contribute nothing useful to the furtherance of new technologies.
The Postwar Era
World War II changed many things in American society. The nation's R&D infrastructure, such as it was, was mobilized by Vannevar Bush to support the war effort. The R&D contract was invented as a means for the government to buy something it could not describe in advance—new knowledge. Major new weapons laboratories were created, and universities for the first time received large amounts of money to carry out research for the government. The newly created defense industry, based on mobilization of preexisting civilian industry, was given government funds to conduct and/or manage largescale military R&D.
As the war wound down, R&D leaders sought ways to ensure that federal funds would continue to flow to the universities, weapons laboratories, and defense contractors. Even though the United States emerged from the war as the world's dominant technological nation, there was concern that R&D would not be sustained to help meet the challenges of reenergizing the civilian economy, addressing unmet health needs, and meeting the emerging threat from the Soviet Union. Vannevar Bush led an effort to crystallize the new views of the role and organization of R&D in his seminal report, Science: The Endless Frontier,1 released in 1945.
The Bush report argued that R&D could help address important national problems, and it called for a government-supported National Research Foundation to support R&D for both civilian and military purposes. This report codified the linear model of innovation, in which investments in basic research could be expected to produce useful outcomes. The report and its supporters envisioned a system in which government would support academic research and industry would exploit the results for practical ends.
Bush lost the organizational battle—his comprehensive National Research Foundation was not adopted. The Office of Naval Research, the Atomic Energy Commission, and the National Institutes of Health were all established or expanded before a more limited National Science Foundation (NSF) was established in 1950, largely to support academic fundamental research not covered by the other agencies.
However, Bush's vision of the importance of R&D to national well-being, as well as his ideas about which institutions in society should do what kinds of R&D, did get adopted, not only in government but in industry as well. His report ushered in what I call the ''classical" period or the "golden age" of R&D in America. From about 1950 to 1970, R&D infrastructure and funding grew exponentially, largely on the Bush model. Responding to the Bush model of the centrality of basic research, many major corporations set up fundamental research laboratories, usually isolated from the company's business operations.
A key aspect of the R&D model at the time was corporate self-sufficiency; that is, during the golden age, large corporations sought to develop their own capabilities in each area of technology on which
they depended. To enter a new field, they would hire a group of new employees and invest in the necessary facilities. Ford's River Rouge plant represented the epitome of the self-sufficient approach—raw materials entered the complex at one end and finished cars left at the other.
Government R&D funding grew very rapidly during the golden age, especially after the launch of Sputnik in 1957. Soon thereafter, agencies such as NASA and the Defense Advanced Research Projects Agency were established, and funding at NSF grew very rapidly. During the late 1950s and early 1960s, federal funding of civilian R&D grew at rates nearing 35 percent per year. Not unreasonably, the attention of university researchers focused almost entirely on the federal government as a source of funds.
As the 1960s wore on, political and social turbulence began to envelop university campuses, leading to a significant estrangement between academia and the corporate sector, even though the engineering schools by and large were not much engaged in the dialogue about the Vietnam War or civil rights. Science schools were, and for them it was a very tough time.
As a result of these trends, R&D cooperation among different institutions was very limited during the golden age. It was limited because industry thought it should be self-sufficient, because government was funding the universities so generously that most faculty wanted little to do with industrial funding, and because the campuses were not congenial to an industrial presence. During this era, what industry wanted from the universities was its graduates and its published papers. Meanwhile, the government was pursuing its own objectives: hands-off basic research or highly focused mission research, largely in the field of national defense. So universities, industry, and government were proceeding along three independent tracks with relatively little interaction.
During this period, there was a limited amount of cooperative industrial R&D under rather special circumstances. For example, the chemical and petroleum industry agreed to support development of the thermophysical and chemical properties of materials at several universities. Some regulated industries also supported joint R&D in such institutions as the Institute of Gas Technology at the Illinois Institute of Technology and in the Electric Power Research Institute. A few traditional sectors whose firms had little or no indigenous R&D capability also supported joint R&D efforts, as in the textile and pulp and paper industries.
In the mid-1960s, the U.S. Department of Commerce attempted to introduce a "Civilian Industrial Technology Program" to provide federal funds to nonprofit organizations to perform R&D for lagging industries such as textiles, clothing, shoes, and home construction. This proposal was defeated in Congress, at least in part because the industries that were supposed to benefit were radically opposed to it.
I would like to mention an unusual government-university-industry contract in the mid-1960s in which the Advanced Research Project Agency supported Monsanto and Washington University to establish a Materials Research Laboratory in the field of polymeric composite materials. I served as a member of the laboratory faculty, where we often could not tell whether our colleagues were acting as Monsanto employees or as graduate students at the university, and whether some of the faculty were acting as faculty or as Monsanto employees. This ambiguity made some at the university nervous—today this would be quite ordinary, but at the time it was very unusual.
The laboratory made some important contributions to composite materials technology. More important, however, in my view is that it taught Monsanto and Washington University how to work together and enabled them to build a trusting relationship. It was key, I believe, to the very large investment that Monsanto later made in life sciences research at Washington University, which in turn was the key to enabling Monsanto to make its path-breaking transformation from an industrial chemicals firm to a life sciences-based firm.
The Post-Oil-Embargo Era of Experimentation
In the decade after the 1973–1974 oil embargo, the consensus reflected in the Bush report dissolved. Companies became disillusioned by the failure of fundamental research to deliver a constant stream of breakthroughs. Major corporate R&D laboratories were closed, and R&D was decentralized, Because of the economic recession of that era, federal R&D funding began to shrink, and universities began to look for other sources of research support.
In addition, the end of the Vietnam War opened the way to a rapprochement between the universities and industry, and it became possible to work together again. The energy crisis stimulated many universities to engage in research into practical industrial topics such as energy production and conservation. Paradoxically, in the 1960s universities had to learn to become "relevant," and helping with energy and environmental technology became a new focus for their societal engagement.
Furthermore, the challenge from Japan began to emerge during the end of this period, even as firms believed themselves to be under constant challenge to meet new environmental, safety, health, and energy conservation standards.
All these factors led to a new interest in cooperation in research. One bit of evidence for this change was that the Carter administration's Domestic Policy Review on Industrial Innovation conducted from 1978 through 1980 recommended that the federal government support generic applied research for industry and that the antitrust laws be modified to facilitate horizontal cooperation in R&D among competing firms.
In the late 1970s, MIT established a research center in the field of polymer processing with financial support by government and with financial support and active participation in the research by industry. This activity set the stage for much of the government support of industry-university cooperative research in the ensuing decade, including the highly successful Engineering Research Centers program of NSF.
The Bayh-Dole and Stevenson-Wydler acts of 1980 created new incentives for universities and federal laboratories to work with industry on R&D and technology transfer. Stevenson-Wydler also authorized NSF's experimental program in University-Industry Cooperative Research Centers and gave the Department of Commerce the authority (never exercised) to fund Cooperative Technology Centers to help industry with research. The National Cooperative Research Act of 1984 provided a "safe haven" for industrial research consortia to help them avoid challenge under the antitrust statutes.
William Norris, CEO of Control Data Corporation, became a convert to the cause of cooperative industrial research and was instrumental in the creation of the Microelectronics and Computer Technology Corporation in Austin, Texas, in which a number of companies conducted joint R&D related to computer technology. He based his concept heavily on his reading of the role of government-supported cooperative R&D in Japan. Similar efforts were initiated leading to SEMATECH in the semiconductor chip industry and to the National Center for Manufacturing Sciences in the discrete parts manufacturing sector. Each of these efforts began with representatives of firms in the same industry cooperating to address common problems and opportunities, with the expectation that each would commercialize the results independently. This kind of R&D cooperation among competitors is sometimes referred to as "horizontal" cooperation.
Thus, by the early 1980s, cooperation in R&D among firms, as well as between universities, industry, and federal laboratories, had begun to be viewed as quite normal and reasonable, although the number of participants was still relatively limited.
The Competitiveness Challenge
The fourth era, a period during which industrial R&D was heavily shaped by the "competitiveness crisis," overlaps to some extent with the prior period of experimentation with new forms of R&D cooperation. Many causes can be adduced for the poor performance of the U.S. economy during the early 1980s. Some blame the monetary measures taken in 1979 by the Federal Reserve Board to break the endemic inflation of the 1970s for the resulting high interest rates, high unemployment, rapidly growing negative trade balance, and loss of manufacturing jobs during the early 1980s. Others argue that the need to take strong macroeconomic action was itself a reflection of lagging U.S. industrial performance, where ineffective R&D, sluggish commercialization, low rates of investment in modernization, and slow productivity growth all contributed first to the "stagflation" of the late 1970s and then to the poor performance in the early 1980s. Still others argue that "unfair" Japanese trade practices contributed to the worsening trade balance.
One popular interpretation of the competitiveness crisis was that U.S. firms were not able to transform the results of R&D into marketable products quickly enough, especially as compared with Japan. Following this diagnosis, the prescription was to find quicker and more effective ways for industry to exploit the results of research at universities and in the federal laboratories, as well as the results from the companies' own fundamental research labs.
One response to this perception was the passage of the Federal Technology Transfer Act in 1986. This act directed the federal laboratories not only to transfer technology to industry, but also to engage in "cooperative research and development agreements" with firms and consortia. Later, the Omnibus Trade and Competitiveness Act of 1988 created the Advanced Technology Program (ATP) at the National Institute of Standards and Technology under which the federal government would help finance precompetitive R&D at firms and consortia. ATP ushered in the era of cost-shared industry-government collaboration in R&D that reached its high point with the Technology Reinvestment Project (TRP), a short-lived program to fund dual-use technology on a cost-shared basis. TRP was embraced publicly by President Clinton in March 1993 and soon thereafter became a target of partisan attack. Although the TRP only lasted about two years, it helped shape our view that government, industry, and universities should collaborate on R&D for mutual gain.
Thus, by the end of this era around 1993, cooperation had become widely accepted as important, but the dominant model remained horizontal cooperation on R&D on shared problems with common objectives.
The Era of Industrial Restructuring
We are now firmly engaged in the era of industrial restructuring, which began in earnest in 1992 or 1993. The past half dozen years have witnessed a remarkable recovery of the American economy. Productivity is up, unemployment is at historic lows, inflation is in check, the stock market is up, and interest rates are low. The information technology revolution is transforming our way of life, how we do business, and how we govern ourselves. New ways of managing businesses and new approaches to quality performance have restored American firms to their old positions of world leadership.
These new developments have had major impacts on R&D and especially on cooperation. By virtue of the determination to emphasize core competencies and outsource everything else, firms have largely abandoned the idea of self-sufficiency. Strategic alliances have become the norm, nowhere more so than in research and technology development. Firms enter into a plethora of agreements for joint R&D, not so much with competitors as with firms, universities, and government laboratories that can bring
skills and capabilities to the enterprise that it would be too expensive or take too long to develop inhouse. R&D collaboration has shifted from a horizontal to a "vertical" approach in which firms and others along the supply chain work together to bring new technologies to the market faster and with higher performance.
In the vertical model of R&D cooperation, universities are no longer seen as places at which to support long-range, blue-sky research but rather as essential suppliers to corporations of key research activities that are necessary to accomplish the company's main business. As suppliers, universities find themselves working under tight contracts, with deliverables and due dates, and with partners having expectations that intellectual property agreements will serve the needs of the corporation. As a result, the management of intellectual property in university-industry cooperative arrangements has become increasingly contentious. Furthermore, there is some indication that firms are beginning to focus their cooperative arrangements on a small number of universities that are viewed as institutional strategic partners. Such arrangements go beyond agreements with individual faculty members. In some research that my colleagues and I have been doing, we have found that purchasing managers are helping manage R&D cooperation—further evidence that R&D partners are being viewed as suppliers.
R&D cooperation has become the norm in universities, industry, and federal laboratories. R&D leaders are deeply engaged in cooperation, and the old single-investigator model of a research career is in decline. A well-functioning technology transfer operation is now essential for universities and laboratories, and the job of external technology manager is of increasing importance in firms. Today's industrial R&D organizations follow a cooperative model that is the antithesis of Bush's model with its outmoded focus on separation of roles, self-reliance in industrial R&D, and "basic research is best." In the new environment, there has been little demand for change in federal R&D policy since 1992, other than calls for greater federal spending on R&D and for a more thoughtful allocation of existing federal funds among national security, biomedical, and all other national objectives.
Ashok Dhingra, Dupont: In regard to the Monsanto-Washington University collaboration, how would you rate that partnership as to its relevance to the commercial success of the projects that they had at that time? The Japanese took over the carbon fiber technology, which was being developed under this program. And then Dupont took the leadership for the organic fiber.
From your perspective, what were the successes of what I think was the very first lasting partnership between university and industry? How did society benefit from that long funding and high-quality research?
Christopher Hill: I assume that you are speaking specifically about the composite materials, not the biomedical collaboration. The important work in the composites area probably had to do with the development and elaboration of the field of the micromechanics of composites that was done on the mechanical engineering side of that partnership.
It was interesting because it was both chemical engineering and mechanical engineering mechanics on the university side. On the chemical side a lot of the work on naturally occurring composites, that is to say the two-phase polymer systems that John Cardose and his students did, has been important. The work that I was involved in was much more limited. I didn't make much of a contribution. We did some
of the first controlled rheology of high concentrations of fiber-filled suspensions, but this was all normal science and not terribly interesting.
Hank Whalen, PQ Corporation: You mentioned that there has been no change in science policy since approximately 1992. But Vernon Ehlers, vice chair of the House Science Committee, did a study on national technology policy, the summary of which came out last September. It' s my understanding that the published copy of the final report will be issued shortly. This report may not change anything, but it is at least an attempt.
Christopher Hill: Vernon Ehlers, a member of Congress from Michigan, was given the formal task by then-Speaker Gingrich to come up with a new Vannevar Bush report. I was given that charge once, and I have been part of other groups that were given that charge, and many others have tried to do it.
I argue that you cannot do a new Bush report until the world has decided that there is something new to say. Bush did not invent his new world out of whole cloth. And everyone who has said, "Let's get ten smart people around a table for six months and invent a new world of research," always fails. It is just too hard to do. So did Ehlers and his colleagues come up with a new Bush report? The answer is no.
My major point at a recent symposium on Ehlers' report at the American Association for the Advancement of Science was that they managed to make a nice statement on U.S. science policy circa 1987. Ehlers' report suggests that we could all cooperate. That would be good. But we've been cooperating since the early to mid-1980s. Second, the Ehlers report is almost completely oblivious to the emergence of the Internet. All it says about the Internet is that it's a neat way for scientists to exchange data.