1—
The Evolving Structure of University-Industry Collaboration in the United States: Three Cases

David C. Mowery

University of California, Berkeley

Introduction

Collaborative relationships between U.S. universities and industry have changed significantly over the past 60 years, as typified by three cases of research collaborations in the chemical sciences. I begin with a summary of the development of the discipline of chemical engineering in the 1920s and 1930s, with a particular emphasis on the role of the Massachusetts Institute of Technology (MIT) and its collaboration with industry. The second case discusses management of patenting and licensing by universities in the postwar period and the development of the Research Corporation, a nonprofit entity created to manage the licensing activities of U.S. research universities. The third case is the Bayh-Dole Act, which sparked another wave of change in university-industry collaboration after its passage in 1980.

One message that I want to impress upon this audience is the lengthy history of university-industry collaboration in the United States. Indeed, both the U.S. research university and the organized pursuit of research and development (R&D) in industry trace their origins back roughly 125 years, and have grown in parallel throughout the twentieth century.1 The historic involvement of publicly funded universities in the United States with agricultural research, much of which was applied in character, and the involvement of these universities with the agricultural users of this research are well-known aspects of U.S. economic history. But throughout this century, the decentralized structure of U.S. higher education and the dependence of public and private universities on local sources of funding also meant that in a broad array of nonagricultural fields, ranging from engineering to physics and chemistry, collaborative research relationships between university faculty and industry were common.2

1  

 D.C. Mowery and N. Rosenberg, Paths of Innovation: Technological Change in 20th Century America (New York: Cambridge University Press, 1998).

2  

 N. Rosenberg and R.R. Nelson, "American Universities and Technical Advance in Industry," Research Policy 23:323–348, 1994.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 7
--> 1— The Evolving Structure of University-Industry Collaboration in the United States: Three Cases David C. Mowery University of California, Berkeley Introduction Collaborative relationships between U.S. universities and industry have changed significantly over the past 60 years, as typified by three cases of research collaborations in the chemical sciences. I begin with a summary of the development of the discipline of chemical engineering in the 1920s and 1930s, with a particular emphasis on the role of the Massachusetts Institute of Technology (MIT) and its collaboration with industry. The second case discusses management of patenting and licensing by universities in the postwar period and the development of the Research Corporation, a nonprofit entity created to manage the licensing activities of U.S. research universities. The third case is the Bayh-Dole Act, which sparked another wave of change in university-industry collaboration after its passage in 1980. One message that I want to impress upon this audience is the lengthy history of university-industry collaboration in the United States. Indeed, both the U.S. research university and the organized pursuit of research and development (R&D) in industry trace their origins back roughly 125 years, and have grown in parallel throughout the twentieth century.1 The historic involvement of publicly funded universities in the United States with agricultural research, much of which was applied in character, and the involvement of these universities with the agricultural users of this research are well-known aspects of U.S. economic history. But throughout this century, the decentralized structure of U.S. higher education and the dependence of public and private universities on local sources of funding also meant that in a broad array of nonagricultural fields, ranging from engineering to physics and chemistry, collaborative research relationships between university faculty and industry were common.2 1    D.C. Mowery and N. Rosenberg, Paths of Innovation: Technological Change in 20th Century America (New York: Cambridge University Press, 1998). 2    N. Rosenberg and R.R. Nelson, "American Universities and Technical Advance in Industry," Research Policy 23:323–348, 1994.

OCR for page 7
--> Old-Style Collaboration I label the first case "old-style collaboration," although this type of collaboration continues in many sectors of U.S. industry and in many U.S. universities. There are several key characteristics of this collaboration, many of which are illustrated by the collaboration between MIT and Standard Oil of New Jersey that contributed to the development of the discipline of chemical engineering in the United States before 1940. This academic discipline was developed with major contributions from MIT, the University of Wisconsin, and the University of Illinois. But many of the innovative collaborative relationships that underpinned the growth of chemical engineering was centered at MIT. Arthur D. Little, Warren Lewis, and other faculty encouraged the development of these collaborative relationships, involving research and teaching, the exchange of students in cooperative education, the foundation of the school of chemical engineering practice, and a nearly parallel growth of organized research in engineering at MIT and in industry. The development of chemical engineering research, teaching, and practice was influenced by the symbiotic relationship between Standard Oil of New Jersey and MIT faculty who worked to codify, advance, and disseminate the key tenets of the emergent discipline. Much of the collaboration during this period involved joint development of these new practices in both the academic and the industrial laboratories combined with relatively widespread dissemination, particularly through teaching and textbooks. The Standard Oil refinery in Baton Rouge, Louisiana, also played a key role as an unofficial external laboratory and employer of a great many of the graduates and a number of the faculty at MIT in the school of chemical engineering. The MIT-Standard Oil collaboration culminated in the development of fluidized bed catalysis in 1941. Research conducted at MIT complemented research done in the Baton Rouge refinery and produced an important patent that was assigned to Standard Oil. Although intellectual property and formal patents clearly were an important component and an important output of this research activity, the university' s direct role in managing, licensing, and seeking the assignment of the intellectual property was quite different from what we observe today. Indeed, it contrasts with the policy that evolved at MIT during the decade after this breakthrough in catalysis. The key to this style of collaboration was personnel exchange—primarily from MIT to industry—through faculty consulting, faculty rotations to and from industry, and placement of graduates. Personnel exchange was a very important component of technology transfer, bringing expertise from MIT to industry and transferring practical knowledge from industry back into academia, where it was refined and codified, supporting the development of a broader engineering discipline. As in many other areas of engineering or scientific research, access by faculty to industrial facilities was important, as the scale and type of equipment in industry often were not available in universities. The industrial collaborators obtained the ownership of or were assigned the intellectual property resulting from collaboration, and a great deal, although not all, of the results of the research by academics in the industrial context were published. Many of these characteristics still apply to much of the collaboration operating between U.S. universities and industrial firms today. Nevertheless, many of these collaborations contrast with others that center on patent licensing. There are some important contrasts to keep in mind between this "old style" and what we see emerging since 1980. A number of features of the post-1980 relationship between U.S. universities and industry are illustrated by the origins, growth, and decline of the Research Corporation during the 1940–1980 period.

OCR for page 7
--> The Research Corporation3 Origin and History The second case of collaboration, that of the Research Corporation, is interesting in its own right because of its origins and its unusual function within the developing academic and industrial R&D system in the United States. It also provides insight into the characteristics of technology licensing—particularly the licensing of academic technology—by an organization that has been in place for a much longer period of time than most individual campus technology transfer or patent licensing offices. The Research Corporation was founded in 19124 on the basis of a technology invented at the University of California at Berkeley, where Frederick Cottrell, a professor at the university, developed antiprecipitant pollution technology for reducing particulate emissions. His work was motivated in large part by a desire to mitigate the pollution generated by the activities of local firms in the San Francisco area. Cottrell developed a substantial portfolio of patents that he licensed, using the proceeds from the licensing activity to support scientific research through the extension of grants to other researchers. The ultimate goals of this patent licensing operation thus were philanthropic, in some contrast to more recent university licensing activities. For the first several decades of its existence, the Research Corporation focused on expanding and marketing its patent portfolio in antipollution technology. As a result, the Research Corporation developed a reputation and expertise in patent management and licensing, accumulating knowledge of patent application procedures, management of patent litigation, and management of patent licensing. In addition, it received a number of donations of patents from academic inventors. The Williams-Waterman patent for vitamin B1 in the early 1930s was one such patent given to the Research Corporation by the inventors so that the corporation could manage the licensing and use the income to support other areas of scientific research. The Research Corporation expanded its patent management and licensing activities through an agreement with MIT in 1937. Karl Compton, president of MIT, wished to extract more income from patents obtained by MIT faculty. Based on the Research Corporation's reputation, as well as a personal relationship between Compton and the then president of the Research Corporation, an agreement was negotiated between MIT and the Research Corporation. Compton's decision to contract with a third party to manage MIT's patent portfolio was influenced by MIT professor Vannevar Bush, who was an industrial inventor of some repute with a substantial number of patents to his name. Bush believed that MIT should not be involved directly in management of patents, and believed that MIT particularly should avoid direct involvement in management of licensing contracts because of the risk that such activities would be criticized by politicians or industrial firms. The agreement committed the Research Corporation to use its best efforts to obtain patents on MIT inventions, license the patents, and, not insignificantly, to pursue infringers. The Research Corporation's work with MIT grew moderately until about 1940, when MIT's expanding wartime research made patenting and licensing less important. Nevertheless, the Research Corporation negotiated numerous agreements with other U.S. universities after World War II. The 3    D.C. Mowery and B. Sampat, "Patenting and Licensing of University Inventions: Lessons from the History of Research Corporation," paper presented at the conference on R&D and Economic Growth in the 20th Century, Haas School of Business, University of California, Berkeley, March 27, 1999. 4    "The Research Corporation, An Experiment in Public Administration of Patent Rights," Journal of Industrial and Engineering Chemistry 4 (12), 1912.

OCR for page 7
--> Figure 1.1 Invention administration agreements: 1946–1982. number of administration agreements between the Research Corporation and individual U.S. universities grew from roughly 5 in 1946 to more than 280 by the early 1980s (see Figure 1.1). The Research Corporation sought cost savings and efficiencies through its centralized management of a substantial and diverse patent portfolio. However, these cost savings proved to be elusive; centralized management for such a large, diverse patent portfolio was not very cost effective. Difficulties with Centralized Patent Management The anticipated cost savings associated with centralized management of patenting and licensing were not realized for several reasons. First, the expertise developed through experience in patenting and licensing in one area of technology did not always improve performance in these activities in other technology fields. The Research Corporation focused its efforts and became most proficient in patenting and licensing in the biomedical area, as demonstrated by a number of important vitamin patents and some early drug patents. Indeed, MIT became frustrated with the Research Corporation's lack of expertise in patenting and licensing electronics inventions, a field in which MIT accumulated numerous patents from its wartime R&D. In 1947, MIT terminated its exclusive relationship with the Research Corporation as manager of its patent portfolio and used other organizations—such as large patent law firms—to supplement its relationship with the Research Corporation. Adding to the costs of the licensing process was the need for frequent interaction between the Research Corporation representatives and client universities. The universities with whom the Research Corporation developed patent licensing agreements insisted on a great deal of "handholding." Geographic proximity was essential to the Research Corporation's relationships with a large number of geographically disbursed institutions. The need for geographic proximity had not been foreseen by the

OCR for page 7
--> Research Corporation, which increased its costs because considerable travel expenses and a large number of branch offices were necessary, and it made it very difficult to reap the efficiencies associated with this centralized management. Finally, conflicts developed from the Research Corporation's contractual agreement with many of its institutional clients to extract as much income as possible from patent licensing. In some cases, the client universities found that the Research Corporation's focus on income maximization impaired the universities' relationships with industrial firms that supported research or were important philanthropic academic donors. The MIT patent on magnetic memory for computer technology was one such case. In this case, the efforts of the Research Corporation to prosecute the patent and to litigate the patent against alleged infringement by IBM produced quite a bit of tension with a major corporate supporter of the university. MIT took back this patent from the Research Corporation and licensed it on more generous terms to IBM. This conflict between the goal of maximizing income and a broader set of institutional goals of the universities remains significant in the operation of a number of contemporary technology transfer and patent licensing offices in the United States. The overall costs of the Research Corporation and the cost per patent rose substantially from 1960 to 1982 (see Figure 1.2). As a result, the Research Corporation encountered cost pressures that in turn caused a decline in its net income (see Figure 1.3). The mid-1970s in particular brought recurrent deficits. These deficits appeared prior to the Bayh-Dole Act, which led a number of universities to enter technology licensing independently of the Research Corporation. Another important aspect of the Research Corporation' s activities, which applies to most contemporary licensing offices, was the dominance of its overall income by a very small number of inventions, Figure 1.2 Costs of the patent development division, 1957–1982.

OCR for page 7
--> Figure 1.3 Net income of the Research Corporation, 1961–1982. and the dominance of those inventions by biomedical and life sciences patents. From 1960 to 1975, 75–90 percent of the Research Corporation's income was provided by patents on only four inventions: nystatin, cortisone, reserpine, and hybrid corn. Along with the growth of costs and the associated emergence of operating income deficits, the expiration of these ''home-run" patents in the mid-1970s intensified financial pressures on the Research Corporation. In summary, the history of the Research Corporation suggests that centralized management of patenting for multiple institutions is difficult and is not always as cost efficient as frequently suggested. This is a consequence of the limited spillovers that often exist of management expertise across technology categories. This tension between centralization and decentralization characterizes a number of contemporary university licensing offices. In addition, the conflict between the income-maximizing technology broker and the interests of institutional clients cannot be neglected. And finally, any analysis should recognize the concentration of licensing income among a very small number of inventions. Bayh-Dole5 Expanding Patenting and Licensing Activity The third of these cases is the so-called Bayh-Dole era. The Bayh-Dole Act was passed in 1980 and facilitated the licensing of inventions based on patents from federally funded research by U.S. Universities 5    D.C. Mowery, R.R. Nelson, B. Sampat, and A.A. Ziedonis, "The Effects of the Bayh-Dole Act on U.S. University Research and Technology Transfer: An Analysis of Data from Columbia University, University of California and Stanford University," forthcoming in Industrializing Knowledge, L. Branscomb and R. Florida, eds. (MIT Press, Cambridge, Mass., 1999).

OCR for page 7
--> and other nonprofit organizations. It codified and rationalized federal policy, which previously had been a patchwork of agency-specific policies and agreements negotiated with individual universities. The Bayh-Dole Act did not make licensing possible that had previously been impossible, but it stated that Congress, and thus the federal government, believed that this licensing was a legitimate use of federal funds and mandated a common government-wide policy. The act produced rapid expansion of U.S. universities' patenting and licensing activities. The number of licensing offices maintained by U.S. universities has grown from roughly 25 in 1979 to well over 200 today. The number of patents assigned to U.S. universities has more than quadrupled during that period of time. The research summarized in Mowery et al. (1999) examines data from Stanford University, Columbia University, and the University of California (UC) system before and after Bayh-Dole. This includes two universities, the UC system and Stanford University, that were active in patenting and licensing before Bayh-Dole, and Columbia University, which has become a major licensor of technology since the passage of Bayh-Dole. There was increased patenting activity after Bayh-Dole at both UC and Stanford, but there was also an increase in university patenting at both institutions substantially before the passage of the Bayh-Dole Act. The increased patenting activity was concentrated in the biomedical area. During the 1975–1990 period, the share of biomedical patents in UC patents increased from 25 percent to 65 percent. And the major increase in the biomedical portion of UC inventive output took place well before the passage of Bayh-Dole. We find a similar timing and composition of expanded inventive output at Stanford (see Figure 1.4). More than Bayh-Dole underpins these developments, something that must be kept in mind in any evaluation of this law's effects. The increase in biomedical invention disclosures and patenting activity prior to 1980 appears to be a consequence of the rapid growth of federal funding in biomedical research, notably under the auspices of the National Institutes of Health, and especially the war on cancer that began in the early 1970s. These events triggered a wave of scientific advances in molecular biology and led to the development of the biotechnology industry. For example, the application for the Cohen-Boyer patent predates Bayh-Dole. Changes in federal policy toward intellectual property rights in the late 1970s and 1980s also enhanced the legal strength and economic value of patents in the biomedical and biotechnology areas. Table 1.1 depicts trends from 1970 to 1995 in gross income from licensing (in constant 1992 dollars) at UC, Stanford, and Columbia. The UC earned slightly more than $1 million in FY 1970, and almost $60 million in FY 1995. Stanford University's earnings increased from $180,000 in FY 1970, the first year of operation of Stanford's Office of Technology Licensing, to almost $36 million in FY 1995. Columbia University, a more recent entrant into the patenting arena, did not even have a technology licensing office until 1982. Columbia earned $540,000 in FY 1985, and this grew to nearly $32 million in constant dollars by the mid-1990s. Equally important is the domination of these income flows at all three universities by a small number of inventions. The share of gross income from the top five inventions at the UC was almost 80 percent in 1970, although by 1995, this figure had dropped to 66 percent. The similar share at Stanford was 69 percent of gross income in 1975, which increased to 85 percent in 1995. Columbia's income has been even more concentrated, with more than 90 percent of gross income for FY 1985–1995 flowing from the five most lucrative licensed inventions. These "top five" inventions at each university also are derived largely from biomedical technologies. At UC, because of the early importance of agricultural inventions, biomedical inventions account for only 34 percent of its licensing income from the five most lucrative inventions in 1970. But by the end of the period, all of the top five UC inventions were biomedical. For Stanford, the biomedical inventions accounted for 87 percent of the earnings of the top five inventions in FY 1975, a figure that

OCR for page 7
--> Figure 1.4 Biomedical disclosures from 1976–1989 at the University of California (top panel) and  Stanford University (bottom panel). increased to 97 percent in 1995. At Columbia, the biomedical inventions' share of the income from the top five inventions grew from 81 percent in FY 1985 to 91 percent in FY 1995. Just as was true of the Research Corporation's licensing income, these three research university systems derive the majority of their licensing income from biomedical inventions. The prominent role of biomedical inventions in licensing income can be attributed to a number of factors, among the most important of which is the strength and value of intellectual property in this technology. Biomedical

OCR for page 7
--> TABLE 1.1 Selected Data on University of California, Stanford University, and Columbia University Licensing Income, FY 1970 to 1995 (all earnings are given in thousands of 1992 dollars) University and Selected Data 1970 1975a 1980 1985 1990 1995 University of California Gross income 1,140.4 1,470.7 2,113.9 3,914.3 13,240.4 58,556.0 Gross income of top five earners 899.9 1,074.8 1,083.0 1,855.0 7,229.8 38,665.6 Share of gross income from top five earners 79% 73% 51% 47% 55% 66% Share of income from top five earners associated with biomedical inventions 34% 19% 54% 40% 91% 100% Share of income from top five earners associated with agricultural inventions 57% 70% 46% 60% 9% 0% Stanford University Gross income 180.4 842.6 1,084.4 4,890.9 14,757.5 35,833.1 Gross income of top five earners   579.3 937.7 3,360.9 11,202.7 30,285.4 Share of gross income from top five earners   69% 86% 69% 76% 85% Share of income from top five earners associated with biomedical inventions   87% 40% 64% 84% 97% Columbia University Gross income       542.0 6,903.5 31,790.3 Gross income of top five earners       535.6 6,366.7 29,935.8 Share of gross income from top five earners       99% 92% 94% Share of income from top five earners associated with biomedical inventions       81% 87% 91% a Figures for Stanford University are for FY 1976 instead of 1975.

OCR for page 7
--> patents are profitable because intellectual property protection is strong and the economic significance of an individual patent is substantial. That is, a patent on a molecule for a pharmaceutical company is typically both stronger and more valuable than a patent on a design technique for an integrated circuit. The data on the growing number of university licensing offices in the United States indicate substantial entry into patenting and licensing by universities as a result of Bayh-Dole. But for many of the universities that have entered the pursuit of licensing, the lack of home runs and a relatively modest flow of biomedical inventions in particular mean that their licensing activities are unprofitable. Although profit is by no means the only or necessarily the most important motive for establishing a licensing office, the probability that a given university will realize a net profit on its licensing activities appears to be modest at best. A second issue raised by Bayh-Dole involves its effects on collaboration between U.S. universities and industry. The emphasis on intellectual property for both the university and the potential industrial partners in collaborations that has been sparked by Bayh-Dole can either support or limit such collaboration, and these effects will differ among industries and among segments within industries. In an industry as large and as diverse as the chemical industry, for example, some firms find that the emphasis on patenting and licensing facilitates collaboration. Other firms in other segments of the industry, such as heavy chemicals, find that this emphasis on defining at the outset the ownership of intellectual property flowing from collaboration is an impediment to collaboration. Assessing Bayh-Dole Are the effects of Bayh-Dole on U.S. universities and industry desirable? We believe they are mixed. We find little evidence that the incentives created by Bayh-Dole for patenting, licensing, and associated transfer of technology have shifted the academic research agenda toward applied research, a concern voiced by some observers. But this is a very tentative and preliminary conclusion. In some spheres, however, such as biotechnology research tools, including research materials or genetic material, the invention is itself an input to the scientific research process. If universities expand their patenting and licensing of these research tools, research results formerly placed in the public domain will be taxed by their inventors. The Cohen-Boyer patent, for example, embodies knowledge that might have been published and widely exploited by industry without a patent. Cohen-Boyer was licensed on a nonexclusive basis, so it was widely employed. But it is not clear that the transfer and application of this knowledge was actually facilitated by patenting and the presence of the university as a licensor. Moreover, in the burgeoning area of research tools, a proliferation of patents and restrictive licensing agreements could limit the use of important inputs into scientific research. There is also a broader issue of whether the establishment of private property rights over publicly funded research is itself desirable. Conclusions The three cases discussed here support several conclusions about the origins and structure of university-industry collaboration. First, the form and structure of collaboration have changed over time. In the early twentieth century, there was a complementary relationship between the emerging industrial research activities of private firms and the university-based R&D. Gradually, MIT and other U.S. universities sought greater control of the management and the income associated with university developed intellectual property. By the 1980s, the structure of many university-industry collaborations

OCR for page 7
--> was precisely what Vannevar Bush had sought to avoid, as universities were directly involved in patenting and licensing research results. It remains an open question as to whether Bush' s concerns over the potential exposure of universities to political criticism resulting from their direct involvement in patent management were well founded. These changes in the structure of the relationships between U.S. universities and private firms have had other consequences. The complementary relationships that developed in some areas of industry and technology have been replaced or supplemented by a more directly competitive relationship between industrial and academic research in some areas. Recent agreements such as that between Novartis and the University of California, Berkeley (see Chapter 6 by Todd La Porte), align a group of faculty in a university with a single firm, affecting the competitive position of that firm vis-à-vis others. In the area of research tools, universities often compete with and rely on biotechnology firms in developing and licensing research tools and genetic material. In many cases, universities need the research material from the biotechnology companies, but also compete to license these research tools to pharmaceutical firms, among others. There is now a very complex mix of competition and cooperation. Consider Genentech, a firm that came out of the University of California, San Francisco, and an early licensee of the Cohen-Boyer patent. It is now entangled in a very interesting web of litigation with the UC. Biotechnology is unique in many respects and does not describe the entire landscape of university-industry collaboration, but at least in these instances, the relationship is starting to look very different from the "old-style" collaboration and has taken on elements of competition in place of cooperation. The effects of the emergent competitive relationship between university and industry research are very unclear for political support for the continued public funding of the university research. The lengthy and rich history of university collaboration in the United States is an outgrowth of some unusual structural characteristics of the U.S. system of higher education. By comparison with those of Japan, Germany, France, and the United Kingdom, the U.S. higher education system has been much larger throughout this century. Almost any international comparative analysis shows that the number of students and the number of institutions are much greater than in these other industrial economies. The U.S. system also is characterized by a more diverse mix of institutions, including research universities, liberal arts colleges, and public and private institutions. These diverse institutions are not managed nationally or centrally, but compete fiercely with one another for prestige, for students, for faculty, and for resources. This is a very different structure from that of most other industrial economies. These differences have direct implications for the effects of foreign emulation of the Bayh-Dole model. A number of foreign governments, such as that of Japan, have passed legislation that creates a legal basis for so-called "technology licensing organizations." But the radical differences between the Japanese and the U.S. university systems mean that the effects of such legislation may be quite different from those of Bayh-Dole, keeping in mind that Bayh-Dole alone does not explain the upsurge in patenting and licensing activity in the United States. The other structural characteristics of the U.S. university system, rather than Bayh-Dole alone, have contributed powerfully to the longstanding collaboration between U.S. universities and industry that dates back almost a century. I have focused here on a relatively narrow slice of the channels through which universities in the United States transfer technology, interact with, and collaborate with industry. But it is important to recognize the number, breadth, and diversity of these channels. They include publication, training of students, and faculty consulting, in addition to these more conventionally defined channels of technology transfer. One of the remaining challenges and potential risks associated with the Bayh-Dole Act is the danger that an excessive focus on intellectual property and licensing and patenting as the primary channel for technology transfer will have a chilling effect on other channels. This is a risk, and not yet

OCR for page 7
--> a reality, but is something that needs to be kept in mind when managing these activities within universities and when explaining these activities and relationships to other constituencies, both political and industrial. Discussion Thomas Manuel, Council for Chemical Research, Inc.: You mentioned a common belief that a very small fraction of university patent offices are profitable. Are there valid data that attest to that? It seems to be anecdotal truth, but it's hard to find the numbers. David Mowery: Surprisingly, these data do not exist. The surveys published by the Association of University Technology Managers, for example, report gross income, not net income. There is one piece that was published by two academics from Portland State University, and whose accuracy I tend to discount, that suggests that the proportion of profitable offices could be as high as 30–40 percent. This is probably a high estimate. But that is the only estimate I have seen. My guess is that the true proportion is probably closer to 20 percent. Francis Via, General Electric: Thank you for a wonderful overview, especially the quantitative data of trends we all believe in. As a result of these data, have you seen cases in which universities are changing their role relative to interacting with particular industries, such as the steel industry or the chemical industry? Being associated with the chemical industry, I have concerns that cooperative programs are going overseas as a result of the drive of intellectual property rights or becoming global, and we are becoming global, but to a greater extent than would be justified otherwise. And many times the expectations of universities are associated with the biotechnology. Once we did some research in this area for the Industrial Research Institute, in which we looked at the 50 top income producers for all universities in 1994. And of the top 50, only one was associated with chemistry—separations technology. The proliferation of this information is very important to us. But are you seeing changes in universities interacting with various industries in that there may be greater flexibility with one industry that is more interested in the other approaches to technology transfer rather than licensing? David Mowery: This is a good question, and I do not have an easy answer to it. Some universities have expressed concern that a focus on income, on licensing, and licensing with a view to maximizing income will have a negative effect on relationships with firms in other sectors such as nonbiotechnology. But there is a high-level administrative pronouncement to this effect, combined with a set of internal norms for evaluation and for rewarding the technology managers that continue to focus heavily on income because, of course, it's what can be counted. It's what you can see. This broader array of objectives is in many cases difficult to measure, difficult to quantify, operates over a longer time horizon and, therefore, it's much more difficult to get it into the managerial consciousness of the people in the technology licensing office. But it's a risk, and you are exactly right. We do see the growth of "foreign competition" in the demand for their services, for example, the ability of Procter & Gamble to establish relationships with universities and the People's Republic of China without as much concern over intellectual property rights. U.S. universities are aware of this and are so scared of it that they cannot even talk about it. So it's an issue that is not very clearly recognized or articulated.

OCR for page 7
--> John Tao, Air Products and Chemicals, Inc.: I agree with a lot of what you said in your presentation. In your answer to Thomas Manuel, I agree with your number of 20 percent. In the survey by the Association of University Technology Managers, at least the 1997 numbers, if you use the table that reports the number of professionals in the technology transfer function, apply a number including overhead, and weigh that against the gross income, most of the technology transfer offices are not profitable. Laren Tolbert, Georgia Institute of Technology: There is another number that is even more difficult to arrive at, and that is the lost opportunity. That number would weigh all this down even further. There are any number of cases I could point to in which grants were not awarded and agreements were not signed because it was never possible to reach an intellectual property agreement. This is another issue that needs to be thrown into the mix. Robert Lochhead, University of Southern Mississippi: The focus on biomedical innovation intrigues me, and it would be interesting to look at the deeper reason for it. One of the reasons might be that the culture of biomedical researchers differs from that of other researchers. Is this driven by more-focused government funding? Is the time ripe, for example, for understanding genetics the same way that understanding the atom and molecules drove chemistry 100 years ago? Or is it that the biomedical field is just getting more of its share of bright people today because drivers for the biomedical segment are so common? David Mowery: Many of the points you suggest are among the drivers. Federal funding for biomedical research has increased in real terms at a rate of about 7 percent per year for approximately the past 30 years. That cannot be said about many other areas of federal funding for academic research. But there are major scientific advances that, as you suggest, have revolutionized the field and, very importantly, this is an area in which the scientific frontier is arguably closer to commercially attractive applications than many other areas of research. And the funding and intellectual excitement attract very bright people. And finally, there is the combination of strengthening of intellectual property rights that occurred in the United States during the 1980s, in particular, and operated with particular force precisely in the life sciences and biomedical area, and this has facilitated licensing. Licensing contracts in some respects are more straightforward to draw up. The value of an individual patent is higher in many cases and the strength of that patent is greater. So all of these factors have elevated the significance of the biomedical area within this licensing. And, as has been suggested several times, the danger is that university technology licensing managers or university presidents think that the whole world is as profitable as biotechnology and that the research and the licensing goals do not come into conflict. That, I think, is the challenge here because the money looks good when you see numbers like those for Stanford, Columbia, and the UC system. Ashok Dhingra, Dupont: I enjoyed your insight into partnerships. My question is in regard to Procter & Gamble undertaking R&D in China and the globalization of their businesses. R&D is being globalized. General Electric, United Technology, Dupont, and many other companies are now moving into Asia Pacific to start these partnerships for many reasons, not only for capturing some of the core competencies, but also helping develop the local markets, especially the technology support that may be needed to grow the markets there. Do you have any study or any data that show the effectiveness of these partnerships in China, India, or Singapore versus the partnerships that are done in the United States?

OCR for page 7
--> David Mowery: No. I wish I did. It is a very interesting topic to pursue because my impression is that U.S. universities have their heads in the sand on this issue. Yet it's a reality. It would be very interesting to look at this in more detail.