PHREE Background Paper Series Docment No. PHREE/92/64 Industry-Universily Collaboration in Developed and Developilng Countries by Linda E. Parker (Consultant) MICROFICHE COPY Report Nn.:11374 Type: (MIS) Title: INDUSTRY UNIVERSITY COLLABORAT Author: PARKER, LINDA Ext.: 0 Room: Dept.: PHREE PAPER SEPTEMBER 1992 Education and Emnp oyment Division Population and Human Resources Department The Worlu 3ank September 1992 This publiaion semas wv as an outla for badk8ound pro*&x on the & ongoing wm* pfvgw ofpoliqy research and ana46i of the Ea and Ernpwn DA'iio th Poutio ad Human Reso*a Dv eme of the Wd Rank ves aprssedae thase of e auo(s), and shold not be audbwd to th World Bank * The Ieniational Bank for Recontucton and Developmentl the World Bank, 1992 ACKNOWLEDGEMEN1S More than anything else, assembling a document like this involves obtaining tidbits of informaton from many people. In ftis regard, I would like to thank especiaDy Tom Eisemon, Jerem Oppenheim, the Science Policy Support Group - Iernatonal Study Group on Academic-Unrsity Relations, and Henry Etzkowitz. Florence Heckman always bad the documents that I needed, no matter how obscure. Tom Eisemon, Carlos Kyboch, Kin Bing Wu, Halsey Beemer, Maurice Boissiere, Vaughn Blanenship, and Arnoldo Pirela made valuable suggesdtons and corrections to various drafts. Fialy, Erik Thuistp provided both the opportnity to write the paper and consistendy practical suggestions for making the project manageable. His thoughtful reading of the first draft and continuous search for interesting models to include injected clarity and variety. rc ABSTRACT Over the past three decades, developed countries have experimented with different ways of using scientific resea and technological development to promote economic gpr*tL One means is through stablishig industy-university research collaboradona In recent years, the number and variety of linkage mechanisms in developed countries has grown rapidly. Developing countries are also increasinl involved in fostering collaboration& At prese, there is no consensus regarding which mens are effective and under what cirum nces. This paper examines a variety of models in use and lessons learnd through trial and error, first for developed countries, then for developing countries Finally, it explores ways to gauge a country's readiness for collaboration. * 7~~~~~~ Abbreviations CIT Center for Technological Innovation EAT East Asia and Japan EC Europea Community EPSCoR Experimental Program to Stimulate Competitive Research ERC Engineering Research Center FWT Fund for World-lass Technology GNP Gross National Product IDB Inter-American Development Bank lTRI Industial Technology Research Institute nlTU Intermediate Technology Transfer Unit I/UCRC Indusuy-University Cooperative Research Centers KAIST Korea Advanced Institute for Science and Technoiogy KOSEF Korea Science and ineering Foundation MITm Ministry for Technology and Industry MC Newly Industrialized Country NSB National Science Board NSF National Science Foundation OECD Organization for European Cooperation and Development OTA Office of Technology Assessment R&D Research and Development RD&E Research Development and Engineering RF Revolving Funds S&T Science and Technology Sl;d Science-Based Industrial Park SME Small- and Medium-Sized Enterprises SRC Science Research Center SIDB Science and Technology Development Board TCC Technology Consultancy Centre UK United Kingdom UNAM National Autonomous University of Meico US United States CONTENT'S Np EX.EC1Yj.=JVtE S1lUMARY .............................. i L BAC1KGROUNnD . ............................................... I LI. Barri.ers to Coaboration ......... ................. *...... 1 L2. Reional Modes ................ .... 3 l2.1l South East Asian Profle . .......................... . 4 .22. Latin Amnercan Profile . ........................... . S 12.3. AfricanProfe .... . ................................ 7 IL CASE STUDIES IN DEVELOPED COUNTRIES ...................... MLI. Research Centers Housed at Universities ....... ................ 9 IL1.1. Regional Programs witbin Countries ...... .................. . 9 IL12. Nationa Progranms . .............................. 10 12. Industrial Extension Services . .............................. 12 UL2.1. Georgia Tech Extension Service ...................... 12 13. Science or Technology Parks .......... ............................. 13 1.3.1. U.S. - Research Triangle Park ........ . . . . ................... .. 13 1l4. Regional Development . ................................... 14 IL4.1. United Kingdom - Industrial Orientation in a University ... 14 I.42. U.S. - permental Proram to Stimulate Coopeaive Research ....................................... 16 11.5. Newly Established Acdvities ....... .............................. 17 I15.1. Japan - Contract Research . ....................... . . 17 I52. Japan - Centers for Cooperative Research ............... 17 11.5.3. Contracting Out Industrial R&D to Univerities .......... 17 M. IESSONS LEARNED IN DEVELOPED COUNTRIES ................... 18 1111. Factors Promoting to Success . ........................ . ..... 19 11.11. Building Partnerships ........... .. ........ .. .. ...... 19 ILl2. Science or Technology Parks . ..................... . 22 -~~~~~~~~~~~~~~~~~~~~~~~~ Pame ne IV. MODELS IN DEVELOPING COUJNTRIES ........... ............... . 24 IV.1. Smal and Informal Afrangements ............ .. . ........ .... . 24 IV.1.1. ¶Turkey - Revolving Funds ........ . . ... 24 IV.1.2. Thailand - S&T Board . ........................... 25 IVZ Consultancy Centers . ....................................... 26 IV.2.1. Ghana - Technology Consultancy Centre ................ 26 IV2.2 Mexico - Center for Technological Innovation ............ 27 rV3. Science Technology, or IndustrialPark ....................... 29 IV3.1. Taiwan - Hsinchu Science-Based IndusaW Plark ......... 29 IV.A Newy Establshed Schemes ................................ 32 IV.4.1. Jordan - SME Vouchers for UniversityR&D ............ 32 IV.42. Romania - Consulting Centres for Private Initiative Development . 32 [V.43. Korea-InterdisciplinaryResearchCenters .............. 33 - AA44 Turkey - Technoparks . . . ..... . 34 V. LESSONS LEARNED IN DEVELOPING COU RIES .................. 36 V.1. Necessary Conditions for Collaboration .......... . . . . ........... . 36 tf.. Factors (Contrabuting to Success ............... .. ........... 37 V: Y 1. Incentives ................................... 39 22 StructuralFactors ... ........................... 40 YV23. Planning andlmplementation ....................... 43 V. REAl)NS FOR C014ABORATIO ........N..................... 44 VI.1.. Stages of Development ........................... ... 45 VL2. Indicators of-Readiness ........... ... 46 VL2.1. General Indicators .. . ...... 47 V= Industry Indicators ............... ... 48 VL23. R&D Capacity Iiaoors . . . . . . . . . . . . . . . . . . . . . . . . . . .4 .1 Fi RECS .....*-*-e˘........................................ ˘ .51 EXECUVE SUMMARY Until recently, the important connection between, on the one hand, scientiftc research and tecologcal development and, on the other, industry and economic development bas had litfle impact on countries' science and technology (S&T) policy. Industia countries have been slow to encourage collaboration between the two communities that are most likely to connect S&T with economic development: universities and industes. In some countries, this Is due to the fict that researchc r ,ased In universities or research Institutes traditonaly have had few incentives to colla ..Ate wath researchers in inustry, or there have been strog cultural or legal reasons for the lack of collaboration. On another level, collaboration has not always been necessary. Many tecbnological advmces that have led to fundamental societal changes have been discovered in the absence of a complete scientific rnderstanding of the advance itself. In developed countries collaborative relationships have been considered at best peripheral to the main higher education missions. More frequently, they have been looked upon by students and faculty members as simply undesirable and not appropriate. In ry .^.nt years, however, fincial need on the part of universities has caused a notable change in the deshrability of interaction Industry is also increasingly aware of the cost effectiveness of access to university research facilities, students, and personneL While there are cases in which collaboration has been active for many years, major attention to the possibilities of such interactions came about only ia the 198G. Developing counies are only beginning to explore these relationships. Those with universities that conduct some research and industry that can profit from research-based technological development are in position to engage in mutuaWly beneficia linkages. Some are experimenting with ways to foster joint research activities. At the moment, there is little consensus on which approaches are especially effective under a wide range of circumstancs There are a number of reasons for engaing in industry-universgty collaboraton. One of the most important is exposing students to the needs of industry and giving them exeence in conduc research that has induti application. The experience can motvat graduates to work in industry, which, in turn, can benefit from the knowledge, skills, and techniques students learn while in schooL This diffusion effect is crucia1 in developing countries if they are going to make econo.mic progress thwgh techological advancementm. ii This paper examines a variety of mecbaisms currently in use in developed and developing countries, actors that Influence success, and lessons leaned from experiences in both grps of countries. Given the breadth of the topic, there has been no attempt to be comprehensive. Emphasis i on building research collaborations in the physical sciences and enginerirg. The following topics bave been excluded: agriculture and bealtb, collaborations between research Insdtutes without university connection and industry, acquisition of Imported technology unless it is needed for industiy-university reseawb linkages, intellectual property rights issues, industry-university collaboration that focusses on building businesses, S&T policy or brain drain issues, and technology innovation outside of the context of collaboration. Several conclusions emerge. First, industry-university collaboration cannot be successfl in the absence of a variety of interconnected elements, regardless of the specfic chaacteritcs of any particular collaboration modeL On a fundamental level, there must be universities with appropriate research facilities, faculty members who can perform researc4 industry that is wiing and able to make use of the results of joint activities, and incentives for both parties to collaborate. At a higher level, government policies, progams, and expenditures play crucial roles in determining the extent to which the badc elements we present and the eteent to which collaboration is likely to occur and be successfuL Government involvement can be instrumental in the development of successfdul collaboration; it can also stifle it Models decrbed in this paper illustrate the difficulties that countries have in determining the appropriate amount and type of goveruaent control over the development and functioning of these elements in the face of competing economic, sociaL and cultural influences. The second conlusion is that the concept of industry-university collaboration assumes adoption of at least some of the values and norms associated with Western universities. An institution without professors trained to conduct research, some level of research facilities, graduate programs, or academic freedom is unlikely to be in a position to enwge in fruitful research collaboration with industry. Nonetheless, excessive adherence to Westemn norms and values can discourage coUlaboration with industry, thereby limitdg the role that universities play in economic development. Finally, the eistence of lndustry-unlversity collaboration may be more important than the mechanism used or the utility of the research results. Collaboration is necessary for its symubolic quality: it sends a signal to faculty members of the value of doing work that relates to economic needs. It also provides industry with trairied workers who are familiar with, and ready to work in, the private sector. Taken together, these outcomes can have a remarkable impact on a country or region. L BACKGROUND 1.1. Barters to Collaboraton Developing countries encounter a variety of barners to collaboration. One source is the academic culture often adopted from Western (developed) countries. The goals, value system, methods of operation, and reward system that many developing countries have adopted from Western universities conflict with the needs of developing countries, as well as the culture of industry that developing countries are trying to establish (Blais, 1990; Jones, 1971). Academicians who received research training in developed countries may be even more reluctant than their colleagues in countries to engage in research geared to local needs if they have accepted the Western values of basic research and participate in international scientific networks. The result can be pareularly strong biases against involvement with the practica problems faced by industry (Jones, 1971). Mexico provides other examples of how values and conditions can inhibit collaboration. First, the social image of science and technology does not place researchers in an important sociml position. Second, there is basicaly no industral research and development (R&D) activity in the country and many good researchers stay abroad after completing scientific training to take advantage of greater opportunities. Third, until recently, university researchers' salaries have been too low for them to be able to conduct research actively. Taken together, there has been little reason or opportunity for performing research, much less collaborating with industry (Sober6n and Rodriquez, 1991). In developing countries, professors are often government employees on 12 month contracts As a practical matter, regulations often prohibit them from earning money for monducting research for industry as a consultant. Thu', even when industry is interested in making use of professors' research experise, it may be illegal for research services to be remunerated. This arrangement provides little or no flexibility for conducting research. Besides the lack of national tradition of collaboration or awareness of its value, another obstacle is the perception that collaboration will threaten traditional academic values. Some faculty members and university administrators are afraid that industrial collaboration will endanger their institutions' basic research and graduate training missions (Bollag, 1990; Fairweather, 1990). Similarly, university researchers are sometimes concerned that engaging in industry-sponsored or applied research will be of no benefit, and possibly hurt, their career (NSB, 1982). The career constraint problem stems froim the academic 2 reward structure not placing as much value on research with Industrial or practical relevance. Another peroeived threat to academic values is that industrial influences will restrict academic freedom, especialy when intellectual property rights conflict with traditional dissemination of kaowledge (Berman, 1990: NSB, 1982; Van Dierdonck and others, 1990). The reward structure in universities in many developing countries is based on promotion aiteria that are easily counted, e.g, publications. While this may not send signals that encourage collaboration with industry, It is a means of rewarding performance in a meritocratic fashion when a country is multicultual. The following table gives a sense of the range of differences between academic and industrial research: ZINalAm=ects Unhe9sm Induay Focus of the R&D Basic research; Applied research; curiosit-oriented experimeatal development Basic rationale Advance knowledge lncrease efficiency Aim New ideas Profits Characteristics Idea-centered Practical; produc-centered Framework Open Closed, confidential Evaluation By peers By the boss Schedule Open-ended Tght predetermined Recognition Scientific honors Saiy increases Source: Blais (1990), p. 13. Barriers are not one-sided. As has been the case in Mexico, industries in developing countries often engage in lite if any research. Reasons for this include (Dahlman and Brimble, 1990): * no immediate need; * little or no incentive for firms to compete; 3 * sophisticated technology is imported as turnkey package deals; * firms in intemational joint ventures rely on R&D of parent firms; and * small firms cannot afford to pay for R&D. The size of a firm affects the ease with which collaboration can develop. University researchers may not be interested in cooperating with smal firms because they tend to be interested in problems that are not enough of a challenge to academic research. Additionally, small firms often lack adequate in-house research organizations or personnel to build a linkage and funds to pay for university research. Small high-tech firms, however, are an exception (NSB, 1982). If industries see little reason to invest in R&D and universities perform basic research with minimal relevance to the needs of the productive sector, the likelihood of collaboration is low (Dahlman and Brimble, 1990). Finally, industry-university collaboration is not inherently natural for either party. One notable stumbling block is the 'we haven't done this before" syndrome. A variation from industris perspective is that, until collaboration with academia is tried, no one can see what good "'outsiders'" can be to the firm (McHenry, 1990, p. 40). Nonetheless, resistance tends to lessen once both parties try worldng together and learn how to operate in each others environment (Van Dierdonck and others, 1990). 12. Regional Models The paths that some clusters of developing countries have taken to build higher education systems and R&D capacity have decidedly regional characteristics. This section constructs profiles of R&D and higher education development in selected regions. History, especially the nature of contacts with and influences of developed or metropolitan countries, plays a profound role. Regional approaches are reflected in: (1) when and how higher education and academic research are strengthened; (2) the focus and role of academic research; (3) the role of govermment in higher education and R&D; and (4) barriers to building capacity. As with all profiles, they are generalizations. Exceptions certainly exist; nonetheless, it is useful to examine the key elements of each profile prior to looking at siic collaborative mechanisms. 4 LIL Sauth Eat AA= Profle There are three basic elements to the South East Asian model. First, constituent countries are guided by the Japanese approach to economic development. Second, there are actually two groups of countries within the South East Asian group: the Newly Industrialized Countries (NICs) and another group consisting of countries that have made less progress than the NICs. The NICs include Taiwan, Hong Kong, South Korea, and Singapore, while the otL i e-oup contains Thai]- ' Malaysia, the Philippines, and Indonesia. Third, the N12> A v; ichieved rapid technological development despite engaging in relatively little basic rese'r- and establishing relatively few industry-university linkages. The Japanese approach is characterized by movement from a "'catch-up'" phase, which involved the absorption of technologies from developed countries, to a phase in which the Japanese developed autonomous technologies. The role of R&D was to solve individual problems with applications. Basic science had clearly lower priority. Beginning in the late 1980s, the country's priorities shifted. This change reflected an awareness of the importance of basic research, and that the co"ntry had neglected it while in pursuit of technological advancement (Okamoto, 1991). Consistent with the Japanese approach, basic research capacity in all South East Asian countries is limited. On a less formal basis, academic researchers engage in consultancies with industry (BioTechnology International, 1990). It appears that the NICs chose to build technological capacity, at least up to a certain point, pror to building basic research capabilities and increasing productvity. Iitially, emphasis was on topics with indigenous relevance and importing and exploiting existing technologies. In time, the emphasis broadened to include some level of basic research and developing indigenous technologies (Ranis, 1990). To address the deficiencies of academic research, these countnes have built government-industry research centers. In some cases researchers are academics. In others, the centers and researchers have no ties to universities (BioTechnology International, 1990; Sigurdson and Anderson, 1991). Regardless of where they work, many receive their research training abroad. Again following the Japanese approach, South East Asian countries have for years relied heavily on other countries for provision of the trainng that is essential for technology adaptation (Thulstrup, 1992). Once NICs set about to increase research productivity, progress is rapid. According to Coward (1990), Taiwan and South Korea increased the number of papers in international journals at least 50% from 1985 to 1988. The increases were not due to excessively small 5 numbers in 1985. In fact, the countries had the hghest numbers of articles in 1985 and experienced the largest percentage increases in the region Despite these advances, it does not appear that South East Asia is working consistently toward developing a top quality scientific comunity. The preference seems to be for a scientifically literate citizenry and foused expansion of basic research capacity to feed into existing industrial emphases and for training graduate students (Ranis, 1990). From a bibliometric standpoint, China is considered a NIC (Coward, 1990). However, unlike other South East Asian countries, it has followed the same corse as nhdia and provided vigorous support for basic research. Other South East Asian countries - Thailand, Malaysia, Philippines, and Indonesia - - also increased research productivity, but the numbers and percentages were noticeably smaller than those for the NICs. The difference is so large that Thailand, the most productive among the second group in 1985 and 1988, accounted for less in both years than the smallest and least productive NIC Singapore (Coward, 1990). The latter, however, does not take into consideration the fact that Singapore has more scientists per 10,000 people than does Thailnd. While research productiity increases among NICs are impressive, it is unclear how the match between industry needs and university research will develop. Nonetheless, the role of university-based research in the MCs differs from that in developing countries in other regions, as well as in many developed countnes. Despite building the capacity of academic science, NICs are unlikely for some years to become significant forces pushing out the frontiers of science. Their primary focus will remain creating institutions and incentive systems that link applied research with technological advancement for the sake of economic exansion (Rosenber&, 1990). I.2.2. Lain Ameicn Proftle Most Latin American countries have centralized R&D systems. Many have been in existence for decades. Political and economic difficulties have lead to erratic support for government-dominated R&D, emigration of researchers, and, in some cases, solicitation of research support from industry (Ailes and others, 1988). Practically all of Latin American science is supported by national governments, but it has not been a consistent priority over time. R&D eWenditures as a proportion of Gross 6 National Product (GNP) are low compared with developing countries in other regions of the world. Inflation, large external debt, a small scientific community, and declining opportunities to obtain higher education are hurting efforts to expand R&D capability. In some cases, government support for both R&D and higher education has been reduced (Ailes and others, 1988; Lavados, 1991; Sober6n and Rodriguez, 1991). The five Latin American countries producing the largest numbers of research papers in intemnationally recognized journals (i.e., Brazil, Argentina, Mexico, Chile, and Venezuela) typically emphasize basic bioscience research. They also tend to have a publication profile that resembles that of developing countries that specialize in a few fields rather than the broader profile that characterizes NICs (Ailes and others, 1988). The exception is Brazil. While most productive, it resembles NICs most closely in terms of relative numbers of publications and breadth of fields in which articles are publ;shed (Coward, 1990; IDB, 1988). Brail's performance is due to several decades of consistent investment in R&D and strong efforts to prepare university-level research personnel (IDB, 1988). In general, given the number of world-class researchers in Latin America, the number of publications in internationally recognized journals from the region is notably low (Coward, 1990; IDB, 1988). While these researchers were producing articles at a rate comparable to that common in the NICs in 1985, growth rates among the latter far outpaced those of the Latin American countries by (Coward, 1990). This may be a reflection of the degradation in government support of R&D in the face of the economic crises during the 1980s in many Latin American countries. Latin America has no tradition of industry-university collaboration. R&D has been carried out almost exclusively in universities or research institutes, not in industry. Even then, it has been relatively constrained. While some universities have a long history of research, most concentrate on urdergraduate teaching. In the best of circumstances, research institutes go only as far as the pilot scale level with a new idea. Some consider this type of work to be beneath the level of professional researchers. Similarly, firms have no pilot plants; they prefer to buy foreign technologies, rather than develop their own. Hence, cultural norms prevent transferring research results to firms that could use them. This represents a particular loss in biotechnology, since some countries in the region are especially strong in this area of research (IDB, 1988). In recent years, the principal means of developing industry-university relations appears to have been the science or technology park. Brazil and Argentina have established a 7 number and plan others, while Mexico is developing one (Ailes and others, 1988; Blanpied, 1989; Rudin, 1990). For these and other collaborative efforts to work, some academic researchers must shed their dsdain for practical application and industry must learn to look for solutions to their production problems in their country's R&D infrtrucure (IDB, 1998; Waissbluth and others, 1988). If these changes do not occur, it is not clear from the region's bistory that it will be able to find alternative means to develop useful industry-university collaborations. 1.23. AIi*an P)oJTe The status of R&D in Africa is intimately linked with each countrys colonial experience and its residue. During the colonial era, the establshment of governmental and quasi-governmental research institutes generally preceded the creation of universities by several decades. In Anglophone countries, the institutes were established by colonial governments and commodity growers (Eisemon and Davis, 1991). In Francophone Africa, the institutes were established by the French government and remained administrativel responsible to France after decolonization, as well as dependent upon France for financial and research support For many years, other Western countries, both directly and through and international organizations, have provided a large portion of R&D support to Africa for international research centers. As a result, individual African nations have had little control of their own scientific activity (Eisemon and others, 1985). Even now, many research insttutes in Francophone countries depend upon France for their scientific staff and financing (Eisemon and Davis, 1991). A few African universities weie established well before decolonializtion. Being patterned after institutions in the colonizing countries, they encouraged research by faculty members from the beginning (Eisemon and Davis, 1991). Modern universities were established when independence from colonial powers appeared likely. As with many developing countries, most of the domestic support for higher education has come from national govermments. Degree programs resemble those of colonizing countries, and undergraduate teaching takes precedence over training graduate students (Eisemon and Davis, 1991). While universities have been the focal point for research, African countries have traditionally had notably small scientific communities (Eisemon and Davis, 1992). The scientific culture is low, and the cultural climate is not conducive to growing S&T capacity. Science teacbing is poor starting in primary school, and few opportunities exist to diffuse the culture of science into society. Governments rarely approaah universities wMth 8 scientific or technological problems. Governiment offiialWare not adequately trained to recognize the scientific or technological nature of the p'roblems. Further, due to vested interests within ministries, they may not wish to seek solutions to problems (Ayiku, 1991). At present, most countries in Sub-Saharan Afica face a three-way dilemmna: long- term shrinking of government support for higher education and research, pressures for mass undergraduate education, and a weakening of academic research capacity. Under these conditions, it is unlkely that the countries will be able to improve the quality of either undergraduate or graduate education or research significantly (Heyneman and Etienne, 1988). With govermment contributions for academic and non-academic R&D dropping, support in African countries comes increasingly from foreign sources. This contrasts with large increases in governmental expenditures for R&D in many South East Asian countries since the late 1970s (Eisemon and Davis, 1992). There is litde opporunity for industry-university collaboration. Industries are weak and are rarely in position to make use of research findings from universities. Research infratructure is lackdng, scientific communites are small and isolated, and there is a dearth of demand for scientific or research expertise (Muskin, 1992). Even in agiculture, lack of government support for collaboration and few incentives for both conducting research of use in the productive sector and operating extension services have resulted in few linkages (Seymour, 1991). However, an innovative collaborative approach that involves a university in Kenya, a university-based research center im Slovenia, and industry is being prepared by K^ornliauser (1992). Nigeria provides a special example of the situation in African countries. It has by far the highest most research activty. While university research has been conducted for years, little of the results have been commercialized. At the university end, the incentive structure for researchers places significant emphasis on research output and publications. Articles must appear in learned, preferably foreign, journals. Publicaton in local journals, of which there are few, is not rewarded, neither is unpublished work Any technology-relatcd work not intended for publication is considered by many to be unworthy of recogniton. Thus, it is not surprisng that there is little transfer of indigeLous research results to production processes and commercialized products (Ogbimi, 1990). 9 II. CASE STUDIES IN DEVELOPED COUNTRIES University/industry collaboration can take many forms, be initiated in a number of ways, and take place on different scales. This section presents a variety of mechanisms that have been in existence for a number of years and about which there is some operational information. It is arranged by the scale and complexity of the activities. ILL. Research Centers Housed at Universities Il.A. RPzonal Iiogv5m wit Cowa Research centers have existed at U.S. universities for over a century and even longer in Europe. Most have external support, e.g., from private foundations or the U.S. Government, and there was a rise in the number receiving support in the 1980s. One reason was that U.S. state governments developed strategies to enhance state-level economic development by supporting a variety of R&D programs, including research centers located at universities. The focus of these centers has ranged from applied research to advanced technology development. In most cases, they have university, state government, and industry support. The high point for development of regional technology programs in the U.S. was the nid to late 1980s, during which time efforts were made to catalog programs across the country (Minnesota Department of Trade and Economic Development, 1988) and case studies were performed (Peters and Wheeler, 1988). Starting in the late 1980s, U.S. states began to experience economic downturns. Reductions in revenues have put a number of centers programs in jeopardy. It is possible to exract some lessons from this boom-and-bust experience. David Osborne (1990) has examined a variety of programs and has identified a number of lessons to be learned. One of the biggest mistakes is for government planners to neglect a detailed study of their state's economy on the regional and industrial levels Many state offidals moved too quickly into program planning before making an assessment of their state's strengths and weaknesses. Programs often did not match needs. Much money and effort went into activities that were not needed, while areas needing assistance, e.g., public education and job training, were practically ignored. 10 Osborne (1990, p. 56) also found that programs were often not structured to become self-sustaining. To do so, the government's role becomes one of "'wholesaling"', meaning that a government 'would attempt to use public resources and leverage to change private- sector behavior ... to nudge businesses to broaden their research relationships with academia" Osborne feels that the most popular center model, the industry-university research center, "appears to be seriously flawed" (p. 57). Academic priorities, particularly basic research and the training of graduate students, are given higher priority than the needs of cocperating businesses and govermment. Collaborating firms are not intimately involved in setting the research agenda. The results of this include little technology transfer, little if any impact on local businesses, but a significant numnber of publications. For the time being, it may not be possible to determine the overall impact and effectiveness of regional technology-based programs. Wbile waiting long enough for programs to have had a chance to work is important, the situation is more complex. At the apogee of development of such programs in the US, Peters and Wheeler (1988) concluded that existing analytical methods were inadequate to provide answers to the main question of whether a particular program or technology strategy affected economic development of a region. 11.2. Naioa Pkqmms A. US,_- Agmeng &eahenters During the 1980s, the U.S. National Science Foundation (NSF) established a number of programs that support development of university-based research centers. One of these programs, the Eneerig Research Centers (ERC) program, supports iuterdsiplinay research that is performed collaboratively by universities and industry. The goal of the program is to bring engineering and scientific disciplines together to address fundamental research issues crucial to the next generation of technological advances using an engineering systems perspective. To accomplish this, each center must have actve participation and long-term commitments from industry and other user organizations (NSF, 1988). The program has been a model for China's and Korea's ERC programs (see discussion of Korea's program below). One of the reasons is that the U.S. ERC program incorporates a new approach to training undergraduate and graduate students. Specifically, the program aims to teach students to be comfortable using a cross-disciplinary team approach to problem solving. Students participate actively in center research at the host 11 institution as well as in industry laboratories, and take new courses developed direcdy from the content of the center's research. ERCs are supposed to develop a "new" type of engiueer who is familiar with the cross-disciplinary, integrated view of technology from research to product (NSF, 1988). B. U.S. - IndustyUniverity Cooperative Research Centers NSF has operated one . program of centers for nearly 20 years. The Industry/University Cooperative Research Centers (I/UCRCs) differ from the ERCs in goals and scale. I/UCRC goals are (NSF, 1989): * to develop industry, state, and other support for industry-university interaction on industrially relevant fundamental research topics; * promote university research to provide a knowledge base for industrial and technological advancement while training students; and i promote research centers that become self-sustaining with industry, state, and other funding within a five-year period. While ERCs receive on average US$2 million annually from NSF and additional funds from industry, most fully operational I/UCRCs receive from US$50,000 to $100,000 from NSF and require a total of US$300,000 from at least six firms in order to have a sufficient research base. The industrial relevance and eventual self-sustaining characteristics of the I/UCRCs are notewortby (NSF, 1989; NSF, 1988). Two recent developments in the I/UCRC program are relevant. First, the program was modified for the 1991 competition to make state participation mandatory. Previously, state funds could have been included in the funding of a center, but they were not required. The change made it easy to incorporate the new I/UCRCs into existing state S&T strategies. It was also a response to growing criticism by states with their own programs that the growing number of Federal research centers programs that encouraged state matching funds often forced state officials to allocate funds for Federal centers that contributed little or nothing to a particular state's S&T strategy. Second, NSF has entered into agreements with a few European countries to replicate the I/UCRCs and pair the replicas with existing NSF centers with similar research direction for the sake of expanded collaboration (Schwarkopf, 1992). 12 112. Industrial Extendon Services 11.2.1. Gge2ia Tech EAkwion Seavice The concept of extension services is well-established in the US. The Morill Acts of 1862 and 1890 provided Federal land for states to use for the creation of higher education institutions with curricula in which agriculture and tht "mechanic arts (engineering) were equal to science and classical studies. These institutions became known as land-grant colleges and universities. Two additional pieces of legislation, in 1887 and 1914, provided that the land-grant institutions establish agricultural experiment stations and extension services. Under this arrangement training, experiment stations, and extension services dealing with agriculture flourished. For engineering, only the training component was developed (Jones and others, 1990). The following examplt is a significant not only because of its success, but also because by virtue of its existence it is an exception. In 1960, the Georgia Institute of Technology (Georgia Tech) established an Industrial Extension Division. After over 30 years of operation, it is not only among the oldest, but also most successful examples of industrial extension services in the U.S. A network of 12 regional offices around the state of Georgia are linked to a central office in Atlanta, which in turn, is the point of contact with the Georgia Tech Research Institute and Georgia Tech's engineering faculty. The extension agents, who have faculty appointments with the Research Institute, serve a wide range of industrial clients with a plethora of technical needs ranging from production line appilication to sophisticated computer-assisted design. Most projects undertaken by the extension service are short-term and "involve manu processes, fcility and materials planning, methods improvement, or cost contror (Jones and others, 1990, p. 14). The service is designed for small, technologically unsophisticated firms that need assistance in achieving greater economic and competitive equity. The agents are assigned to geographic regions, and therefore generalists. When the request is more demanding in scope than a single agent can handle, researchers from the Research Institute are available if a particular experdse is required. Rarely do extension services offer research services (Jones and others, 1990). Tbe Industi Extension Division is expensive, labor-intensive, and difficult to operate. TIpically, two agents serve as many as 500 companies in a given region. The Division pays for up to SIve days of their time per project. Any additional time is paid by 13 the company in a contract negotiated with the agents. State suppot is low-, agents and research staff must spend much ime lookng for and conducting contract research. In 1988 the entire extension system conducted $85 million worth of contrc research and received only $3 million in state and other funding (Jones and others, 1990). Georgia Tech's operation is the model for the Technology Consultancy Centre in Ghana, which Georgia Tech personnel helped to develop (discussed below) (Behrman and Fischer, 1990). 11.3. Science or Technology Parks Developed countries havn -xperimented with large indusuy-university collaborative arrangements called technoparks, science parks, or technology parks. Because of a wide variety of applications of these terms, science tIhnolngy pajk will refer to a project in which high-technology firms establish operations on a large parcel of land on, or adjacent to, a nimversty or university-affilated research institute for the sake of collaborating with the host institution, as well as with other participating firns. While the general concept has eed for several decades, it became especially popular during the 1980s. 113.1. US - R&earh Tianle Paro Among the oldest science parks in developed countries is Research Triangle Park in North Carolina. Established in the late 1950s, the Park exists within a triangle foraned by the Unersity of North Carolina at Chapel Hill, North Carolina State University at Raleigh, and Duke University, in Durham. The Triangle region had been hard hit by the declining textile indutry. Initiative for the Park came from the state's governor. From the start, the Park bas had managed, rather than spontaneous, development. Initial development was slow, possibly because, unlike other early parks, It did not contaL a premier research university. The Park began to make significant progress in 1965 as a result of the announcement by IBM of its intention to establish a major R&D facility at the Park. This led to the creation of 9,000 jobs and encouraged other large technology intensive firms to establish R&D operations at the Park. IBMs initial move amounted to a vote of confidence that was vital to the success of the Park (Monck and others, 1988). This Park is different from other parks in the U.S. First, as mentioned above, the assodated universities were not among the top ranks of research universities at the time that the Park began. Second, the physical climate is less attractive than that of other parks. 14 Third, the Park's development has always been managed. (Monck and others, 1988). Finally, initiative to establsh it came from the state (OTA, 1984). The first three are noteworthy because they go against traditional thinling about the characteristics that are important for attracting companies to join a park Nonetheless, there are two significant similhrities with other parks. First, initiation of the parks and early support came from strong, determined individuals. Second, had the parks been evaluated too early, they might have been deemed failures (OTA, 1984). 11.4. Regional Development 1A.41 Unied ingdom - Indusriad OrXtion i a Univy The University of Salford hosts one of England's most diversified programs of industry-university linkages. As with many of the other programs, it is designed to reverse regional economic decline brought on by a decaying industrial base. The immediate simulus for creation of the program was a reduction in the University's public appropriation in excess of 40% starting in 1981 and spread over three years. The University's new Vice- chancellor responded by developing a strategy designed not only to generate income but also to establish a distinctive niche for the institution in the UK higher education system. The history of the University's development has been documented by Segal Quince Wicksteed (1985 and 1988). The objectve was to develop a different ype of institution that would have a strong industrial orientation. To achieve this objective, the University set up a variety of mec3anisms to link the institution with industry. In the first five years, the University set in motion a wide-ranging plan that included: * obtaining industrial sponsorship for four endowed chairs as in the German model; * appointing industrial leaders to sit on the University's research committee so that industry would play an active role in determining the overall research direction; * developing a new mechanism to promote the capabilities of the University around the world; * developing a collaborative project with large firms nearby to provide education and training in information technologr, and * establishing a technology park next to the University. 15 Unlike other examples of technology parks, the one a Salford is not considered the focal point of the isitution's industry.university linkages. There are also applied research Institutes with their own ties to industiy, such as the Advanced ManActuring Technolog Centre, plus other mechanisms not mentioned above. This multi-faceted approach to linkages has several interesting organizational consequences. First, university administrators have direct oversight over all acivities. They direct and influence, if not control, as much as they caa. Second, the administrators want to know as much as pob&ble about what individual faculty members are doing. The latter must register their outside work with the University authorities. Third, academics do not always comply with what some perceive to be unnecessary and burdensome bureaucratic requirements; they conduct their industry work 'uderground'. At some institutions, as much as 50%io of the coBaborative work is done covertly. The control that makes thbis approach necessary likely discourages those who do not want to bother with registering their outside work but have no desire to resort to covert arrangements from collaborating altogether. In the long run, no gains from this, except possibly those trying to control the academics. Another result of the desire of University officials to control what they can is that faculty members are no longer allowed to form their own companies. However, University officials have allowed at least one faculty member form a company as a division of the University's industrial liaison company. It remains to be seen if the control aspect restricts the effectiveness of the overall program so that it never reaches its potential. Time will also tell if a university that shifts so sharply toward an industrial, short-term perspective makes a significant sacrifice in not continuing to create new knowledge that industry wil need in the future. A university that performs little or no research cannot support the kind of graduate program that produces the kind of graduates that industry wants and needs. It is far too early to determin- if the University is successful in reshain itself to have an industrial orientation. Some of the services and programs may be more successful than others. Further, if the scheme succeeds, one consequence may be that the changes within the institution are so significant that the University ceases to be a university in the traditional sense. 16 11.4<. US . ExpemA1 NqSM to SfimuWoe Cooemli0e Ran* In some developed countries, the pattern of economic and technological development across a country has been uneven. Some regions become more advanced than others. There are a number of ways to identify the less developed regions, including share of national R&D support, research capacity of universities, production of baccalaureate and graduate degrees in science and engineering fields, and proportion of technology-intensive companies to other types of firms. No single indicator is perfect, but developing countries tend to be weak in a most or all of these areas. Uneven development can create a perception of the country being divided into two camps - the "haves" and the "have nots". This occurred in the US. The national government's response was to create the Experimental Program to Stimulate Competitive Research (EPSCoR). Establhshed in 1980 by the National Science Foundation, the program has grown from five eligible states to eighteen plus the Commonwealth of Puerto Rico (NSF, 1990). The program provides US$1.5 million per year to each state that qualifies for an award through a competitive process. Each award supports infitructure improvement and research enhancement. The infrastructure component is intended to make permanent improvements in the state's research environment based on what is appropriate for the state. It includes human resource development through involvement in the research component and financial commitments from other sources that will become institutionalized at the end Of the grant period. In this way, infrastructure improvement becomes self-sustaning (NSF, 1990). The research component involves either group projects or the creation of a research center. While the program as a whole is geared toward bringing participating researchers up to a level where they are competitive for Federal research suppot, the centers also focus on involving students in research and creating linkages with other academic institutions, government laboratories, and induistiy. The linkages emphasize the exchange of personnel and facility access, as well as the provision of leveraging funds (NSF, 1990). Cooperating fis need not be located in awarded states. The European Community (EC) Commission is currently studying the EPSCoR model in connection with the deveiopment of the ECs new SRIE program for Less Developed Regions. 17 II.S. Newly Established Activities 11.5.1. Jaa - Conrct Rhward, For decades, Japan has focused on applied R&D. A higher proportion of Japan's R&D activities are privately supported than in other developed countries. Until recently, the country has not viewed basic research or industry-university relationships as important for its economic health. The government is presently changing its R&D priorities somewhat. More basic research is being supported and new initiatives provide incentives for private industry to support a greater portion of the country's R&D activities and, at the same time, promote mdustry-university collaboration (Rosenberg, 1990). The Japanese government now allows universities to conduct research contracteNd by private industrial firms. The system also allows universities to employ industrial scientists and engineers on a contractual basis to conduct research. In 1988, the national universities earned 3 million yen from such contracts (Sigurdson and Anderson,. 1991). 3.52. J^pxm - Ceteuw for Coqpeffiv Raesh In 1987, the Ministry for Technology and Industry (MlII) established a series of centers for collaborative research between the national universites and prite industry. Initially, three universities developed centers. Five more universities did so in 1988. More centers have been established since then (Sigurdson and Anderson, 1991). Given the comparative newness of industry-university collaboration in Japan and the rise in importance of research institutes and research corporations, the role of the universities in Japanese S&T is unclear (Sigurdson and Anderson, 1991). IL.53. C