Technology Transfer Systems in the United States and Germany: Lessons and Perspectives (1997)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

272 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY • a marked increase in the production of European suppliers of computer memories like SGS-Thompson, which became the leading supplier world- wide for EPRON in 1993. • the specification for the 0.35-micron CMOS logic process. For this project, industry, institutes, and universities were joint partners. In the field of equipment and materials, one can see the benefits even for smaller equipment manufacturers. Various companies have strengthened their global competitive position through international cooperation with users, co- producers, and research institutes and through new innovative technologies. They include the following: • ASTI, which introduced its all-plastic ultraclean chemical pump with a favorable global market response; • AST, which is now accepted as one of the leading suppliers of rapid ther- mal processing equipment; • Plasmos, whose share of the worldwide ellipsometer market has increased to about 32 percent; and • Successful sales of GEMATEC’s ELYMAT machine, which is used for mapping on silicon. JESSI cooperation with SEMATECH in the field of minienvironment and mask technology resulted in internationally accepted standards, an increased un- derstanding of U.S. market requirements, and increased European access to U.S. markets. Even though JESSI’s funding ended in 1996, the program has accepted a variety of new projects and decided to continue others. These focus on important application-oriented topics like digital audio broadcasting and also on new devel- opments in integrated circuit technology and equipment for integrated circuit manufacturing. The main achievement of JESSI is that the major suppliers in microelectron- ics and information and communications technologies have been brought together, forming a critical mass for large-sized research projects. TECHNOLOGY TRANSFER FROM UNIVERSITIES Universities HISTORY OF TECHNOLOGY TRANSFER The development of German universities in the nineteenth century was influ- enced by the idealist philosophers as well as the growing industrial sector’s need for well-trained personnel. Philosophers like von Humboldt, Fichte, and Schleiermacher influenced 

TECHNOLOGY TRANSFER IN GERMANY 273 decisively the organization and orientation of the German university system. The idealists, who were involved in the founding of the Berlin University (1809– 1810), viewed research at universities as an important element of teaching. At the outset, however, German academic research was focused primarily on areas such as philosophy, mathematics, and humanities; empirical research had to fight for recognition.11 Nevertheless, by the end of the nineteenth century, German university research had achieved world leadership in several major fields of sci- ence, including medicine, chemistry, and physics. Due to rapid increases in stu- dent enrollments, particularly since 1870, many universities created separate de- partments and institutes with laboratories for natural sciences. Through the mid-1800s, the idealist orientation of German professors and administrators led them to elevate the natural sciences and neglect the “less-dig- nified” engineering sciences. Ultimately, it was the demand of German industry for skilled engineers that led the German states to establish special polytechnical schools outside universities. In the 1870s, the polytechnical schools were el- evated to higher status, becoming technical higher education schools (Technische Hochschulen). Initially, the efforts of these new institutions to achieve academic recognition led them to overemphasize theory and neglect research targeting in- dustrial needs. At the end of the nineteenth century, however, the establishment of engineering schools in the United States induced German technical higher edu- cation schools to begin introducing research laboratories. The main benefit to industry of universities and technical higher education schools was the provision of trained personnel. Even at this early stage of the development of the German academic research system, professors had consultancy arrangements with indus- try. In other words, the first forms of technology transfer appeared. Universities and technical higher education institutes focused on education, whereas the central government and the states established a variety of research institutes in applied areas. A prominent example of the latter is the Imperial Insti- tute for Physics and Technology (Physikalisch-Technische Reichsanstalt), which served as a model for the National Bureau of Standards in the United States. In addition, some smaller research institutes were financed jointly by government and industry. Finally in 1911, the Kaiser Wilhelm Society, the predecessor of the Max Planck Society, was founded, at that time with a strong focus on applied science and nearly totally financed by industry. The increasing engagement of industry in government or industry institutes outside universities was stopped by the economic problems caused by the two world wars. Following World War II, the government and the states assumed important roles in the national innovation system through such institutions as the Max Planck Society, the Fraunhofer Society, and the National Research Centers, today called Helmholtz Centers, which are described in more detail in the follow- ing sections. Increased public investment in R&D, beginning in the 1970s, was motivated by a perception that Germany was lagging technologically compared with the United States. 

274 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY In the educational sector, the main development after World War II was the official recognition of technical higher education schools as equivalent to univer- sities. By integration of other nontechnical disciplines some of them officially became “Technical Universities” or even normal universities. In the 1970s, German universities began to consider seriously their role in technology transfer, and university-industry relationships grew. Between 1970 and 1980, industry support of universities increased by 25 percent, and between 1980 and 1990, such support grew by 44 percent. In addition, universities’ needs for external funds increased with student enrollments and, in recent years, be- cause of a scarcity of public funds. Both factors have created pressure on aca- demic research. Beginning in the 1980s, initiatives on a number of fronts document Ger- many’s increasing interest in technology transfer. Early in the decade, German universities established a special working group of university chancellors to look into the topic (Selmayr, 1987), and the former Ministry of Education and Science (BMBW) initiated a research project called Projekt Wissenschaft (PROWIS) that had a strong technology transfer focus (see the publication list in Allesch et al., 1988). In the mid-1980s, the German Science Council issued a statement on technology transfer (Wissenschaftsrat, 1986). These efforts led to the easing of very strict regulations concerning the budget and personnel structures of univer- sities, recommendations on how to handle technology transfer instruments (e. g., the establishment of external institutes), the establishment of technology transfer units at universities, and a generally more open-minded attitude in universities toward technology transfer. STATISTICS ON GENERAL RESEARCH STRUCTURES This section presents information about the development of research funding at universities and the distribution of money among research fields. It also ana- lyzes the sources of external funds, which are good indicators of the major chan- nels of technology transfer. Data are not available, however, on the four focal areas of this study. A main characteristic of the German research system is the public funding of most universities; students do not have to pay tuition fees. It is the states, not the national government, that are responsible for education and hence for the support of universities. According to the principle of equality of research and teaching, the states assign general budgets to each university without dictating how the money should be used. Since universities are not required to report how much of their general budget they allocate to research and its associated overhead, there are no precise statistical data, only general estimates, on this base of institutional research funding. The latter are based on the assumption that a certain share of the total general budgets is used for research, with the share differing from disci- pline to discipline (Wissenschaftsrat, 1993a). In addition, research funds from 

TECHNOLOGY TRANSFER IN GERMANY 275 10 Total Basic funds 9 External funds 8 7 Billions of 1980 DM 6 5 4 3 2 1 0 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 year FIGURE 3.14 Research funds of German universities in constant 1980 DM. SOURCES: Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie (1996); Bundesministerium für Forschung und Technologie (1993a); Wissenschaftsrat (1993b); calculations of the Fraunhofer Institute for Systems and Innovation Research. third parties (Drittmittel), called external funds (contracts and grants), have to be taken into account. Between 1980 and 1990, the overall research budget of universities increased nominally by 59 percent; in constant 1980 DM, the increase was 21 percent (Fig- ure 3.14 and Table 3.2). The sharp growth that followed in 1991 and 1992 was mainly due to the inclusion of the new states in the former East Germany. At about 12 percent, however, the share of research conducted by former East Ger- man universities is quite small. In 1990, 49 percent of the total research budget of German universities was related to natural sciences and engineering (Figure 3.15). This has to be taken into account when a direct comparison between the total R&D budgets of industry and universities is made, because industrial research focuses primarily on these two areas. If only the natural sciences and engineering are considered, the relative share of universities in the German R&D system is much lower than suggested in the general comparison presented in the “General Structures” section, above. (See in particular Figure 3.2.) In real terms, university institutional research budgets grew by 15 percent and the external funds by 42 percent between 1980 and 1990. Hence, the share of 

276 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY TABLE 3.2 Research Funds of German Universities in billions of DM Nominal values Real values (1980) Indexes (real, 1980) Year Basic External Total Basic External Total Basic External Total 1980 4.82 1.36 6.18 4.82 1.36 6.18 100 100 100 1981 4.92 1.47 6.39 4.69 1.40 6.09 97 103 99 1982 5.08 1.51 6.59 4.68 1.39 6.07 97 102 98 1983 5.26 1.54 6.79 4.71 1.38 6.09 98 101 99 1984 5.31 1.73 7.04 4.69 1.52 6.21 97 112 100 1985 5.51 1.78 7.29 4.75 1.54 6.29 99 113 102 1986 5.86 1.95 7.81 4.95 1.65 6.60 103 121 107 1987 6.19 2.15 8.34 5.11 1.78 6.89 106 131 112 1988 6.53 2.26 8.78 5.31 1.84 7.15 110 135 116 1989 6.83 2.40 9.23 5.38 1.89 7.27 112 139 118 1990 7.29 2.56 9.85 5.53 1.94 7.47 115 142 121 1991 8.64 3.53 12.17 6.27 2.56 8.83 130 188 143 1992 9.33 3.83 12.16 6.50 2.67 9.17 135 196 148 1993 9.59 4.25 13.84 6.41 2.84 9.26 133 208 150 1994 10.06 4.48 14.53 6.53 2.91 9.44 136 213 153 1995 10.31 4.56 14.90 6.56 2.93 9.49 136 215 154 SOURCES: Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie (1996); Bundesministerium für Forschung und Technologie (1993a); Wissenschaftsrat (1993b); calculations of the Fraunhofer Institute for Systems and Innovation Research. external funds within the total budget became more important, increasing from 22 percent of the total in 1980 to 26 percent in 1990. These figures have to be interpreted with care, however, because the apparent growth in external funds is partly due to more complete publication of financial sources (Wissenschaftsrat, 1993b). Furthermore, the method of calculation used by different statistical sources varies, leading to different results (Selmayr, 1989). The following analy- sis of external funds is based primarily on data compiled by the German Science Council (Wissenschaftsrat, 1993b), which seems to be the most consistent source. The external funds come chiefly from semipublic agencies, federal minis- tries, foundations, and industry (Figure 3.16). The major semipublic agency (Förderinstitutionen mit überwiegend staatlicher Finanzierung) is the DFG, which provides about 90 percent of the funds in this category. The DFG is the most important central organization for science promotion and is, to a certain extent, comparable to the National Science Foundation in the United States. The largest part of its budget comes from the central government and the states, each of which usually makes an equal contribution for the support of individual projects (see Meyer-Krahmer, 1990; Wissenschaftsrat, 1993b). The DFG supports all areas of science, including the humanities and social sciences, and is generally, but not exclusively, oriented toward basic research. The coordination of univer- sity research at the federal level is one of its major statutory tasks. In this context, 

TECHNOLOGY TRANSFER IN GERMANY 277 30 25 20 Percent 15 10 5 0 natural engineering medical agricultural humanities sciences sciences sciences and social sciences FIGURE 3.15 Distribution of research funds at universities, according to major areas, 1993. SOURCE: Bundesministerium für Bildung, Wissenschaft, Forschung und Tech- nologie, 1996. an effective instrument is the special research areas (Sonderforschungsbereiche), representing about 25 percent to 30 percent of DFG’s budget (Bundesministerium für Forschung und Technologie, 1993a). The special research areas are tempo- rary institutions at selected universities, established for a period of 12 to 15 years, where scientists from different disciplines cooperate in joint research programs (Deutsche Forschungsgemeinschaft, 1993). Focal programs (Schwerpunkt- verfahren) are another instrument for supporting the supraregional cooperation of scientists of different universities. In the case of external funds from federal ministries, the former Ministry for Research and Technology (Ministerium für Forschung und Technologie [BMFT]), now the BMBF, contributed the largest share, about 86 percent. The BMFT support of university research increased by about 110 percent between 1980 and 1990. In other words, the general increase in external university funds is due largely to the increase in BMFT support. A major reason for this growth was the introduction of collaborative research projects in 1984, whereby several industrial partners as well as university institutes work together (Bundesmin- isterium für Forschung und Technologie, 1993a; Lütz, 1993). The projects of BMFT/BMBF are generally quite application oriented, but they also support many basic research projects, for example in the area of marine science. BMFT fund- ing of university research in 1990 equaled about 67 percent of what the DFG invested in this area. Thus, BMFT/BMBF became a second major force in the external funding of university research. 

278 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY FIGURE 3.16 External research funds of universities, according to major sources, 1980, 1985, 1990. SOURCE: Wissenschaftsrat (1993b). The Ministry of Defense invests relatively little in university R&D. In 1990, the ministry’s contribution represented just 10 percent of all federal funds sup- porting university research. The Volkswagen Foundation (VW-Stiftung) accounts for 70 percent of all foundation financing of research at universities. Although it was founded by a private company, the foundation generally supports basic research projects. According to data of the German Science Council, in 1990, industry contrib- uted 15 percent of all external university R&D funds, an 80-percent increase (in real values) from 1980 and a 115-percent increase from 1970 (Wissenschaftsrat, 1993b). This means that funding from industry has become both absolutely and relatively more prominent, particularly since 1980. Compared with the total re- search budget of universities—including institutional funds—industry support represents a mere 4.4-percent share. The industrial funds can be divided into donations, money for collaborative research, and money for contract research. Most industry support (81 percent in 1990) went to contract research. In addi- tion, 11 percent of industrial funding went for “cooperative research” linked to projects of the AiF (see “Federation of Industrial Research Associations,” be- low). Donations from industry or industrial associations accounted for a modest share, 8 percent, of all industrial funds for universities in 1990.12 Recent data of the BMBF based on a more complete survey than that of the German Science Council provide different figures for the industrial contribution 

TECHNOLOGY TRANSFER IN GERMANY 279 to the university R&D funds.13 According to the BMBF, in 1990, industry sup- port represented 7.7 percent of the total research budget of universities or 25 percent of the universities’ external R&D sources. In 1995, these figures increased to 8.7 percent of the total funds or 28 percent of the external sources (Bundes- ministerium für Bildung, Wissenschaft, Forschung und Technologie, 1996). With- out including the contribution of industry-financed foundations, in 1995, the in- dustry support represents about 7.5 percent of the total funds. To sum up, the BMBF data indicate that in recent years, industrial funding of universities has reached a significant level. Among the international organizations that contribute to university research, the EU is the most important (about 85 percent of total international funding). According to the Science Council, EU funding for universities amounted to DM 23 million in 1990, or 0.8 percent of all external university funds. In contrast to that figure, Reger and Kuhlmann (1995) estimated, using data from the European Commission, that EU funding of German universities came to DM 170 million in 1991. Compared to the average situation, this value may be artificially high, since in 1991 the Second and Third Framework Programs of the EU overlapped. But even if EU contributions came to roughly DM 100 million, this is still a relatively small amount compared with total external funding for German univer- sities (see “Impact of European Research,” above). According to recent BMBF data, the EU funding amounted to about DM 130 million in 1995, or 2.7 percent of all external university funds. This means that the contribution of the EU to university funding has increased considerably in recent years (Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie, 1996). The only other significant external source of academic research funding is project-related funding from the states, which represented about 4 percent of the total external funding in 1990. Finally, it should be noted that not all external funds are linked to research activities: Only 86 percent of those funds were so linked in 1990. This reduced share has already been considered in Figure 3.14 and Table 3.2. For the analysis of technology transfer, it is important to note that the exter- nal funds are not equally distributed across disciplines. For example, in the hu- manities and social sciences, the absolute and relative volume of external funds is rather low; in law, the external funds are about 4 percent of the institutional re- search funds; and in economics, they are about 9 percent (Wissenschaftsrat, 1993a). As Figure 3.17 shows for selected areas, the level of external funding in the natural sciences and engineering is much higher. The greatest amount of external funding, DM 266 million, or 41 percent of the total research funds, is apparent in mechanical engineering. This high proportion of external funds can be taken as a strong indication of considerable industrial funding of technical disciplines; the proportion is much higher than the overall 4.4 percent share of university research funding contributed by industry. In physics and electrical engineering, external funding represents about 29 percent of university research 

280 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY budgets. In absolute terms, external support for physics R&D is second only to that for mechanical engineering. This outcome is quite remarkable because of the generally basic orientation of physics research; unfortunately, more detailed sta- tistics on the industrial funds in physics are not available. External funds amount to about 25 percent of the total research budget in computer science, chemistry, and biology. Thus, all focal areas of this report have above-average levels of external funding and probably relatively high contributions from industrial sources. These budget statistics, however, can lead to an underestimate of research activities supported by external funds. In the German system, external sponsors pay only direct personnel costs and, to a limited extent, costs for facilities. They generally do not pay any overhead costs (e.g., for buildings, administration, cen- tral services). This is true for public and semipublic as well as private sponsors. Therefore, research with external funding has to be cofinanced, or matched, by infrastructure funds of about the same amount. Universities must take these in- frastructure funds from their base-institutional support. These infrastructure funds can be considered cross-subsidies to external funds. To sum up, the share of research supported by external funds is about twice as high in terms of time and personnel than it is in terms of budget. This relationship is indicated in Figure 3.17. For example, in terms of time and personnel, the real share of research supported by external sponsors is equivalent to 82 percent of university research funds in mechanical engineering, 58 percent in physics and electrical engineer- ing, and 50 percent in computer science, chemistry, and biology.14 The German delegation is of the opinion that the shares of external support in those areas have increased since 1990 and in many technical institutes have reached 100 percent. Many of these institutes often have the opportunity to acquire additional external funds but cannot take advantage of them because of insufficient infrastructure funds; this insufficiency generally is manifested by a lack of space (see also Hochschulrektorenkonferenz, 1996). Against this background, the overall figure of 4.4 percent of industrial funds within the research budget of universities, according to the data of the German Science Council, has to be adjusted to at least 8 percent in terms of research personnel and time. In other words, the industrial share can be considered equiva- lent to the U.S. share of 6.9 percent, because in the United States, industrial sup- port generally covers the full cost of research, including overhead. If the more realistic share of 7.5 percent of industrial funds, according to the BMBF data, are taken, the German level including related infrastructure funds is even substan- tially above 10 percent. On the basis of available statistics, it is quite difficult to assess the growth rates of external funding for specific disciplines because the disaggregated fig- ures for 1980 and even 1985 are quite incomplete. According to a survey of the Science Council (Wissenschaftsrat, 1993a) and the German delegation’s own es- timates, computer science shows the highest increase in external funding, and the 

TECHNOLOGY TRANSFER IN GERMANY 281 Mechanical engineering Electrical engineering Physics External funds Infrastructure funds Computer science Institutional funds Chemistry Biology 0 100 200 300 400 500 600 700 Million DM FIGURE 3.17 Relation of external, related infrastructure, and institutional base R&D funds of universities in selected areas in 1990 in current DM. SOURCE: Wissenschaftsrat (1993a). increase in electrical engineering seems to be considerable, too. It is, however, not possible to say to what extent the increase in electrical engineering is related to (traditional) energy technology or (modern) electronics. The external funding in mechanical engineering shows only a moderate growth rate, probably because the rates at the beginning of the 1980s were already quite high. In order to provide at least a rough estimate of industrial funding in different disciplines, the data for the University of Karlsruhe, one of the largest technical schools in Germany, are presented in Figure 3.18. The school of mechanical engineering receives the largest volume of industrial funds, but the growth rate in the 1980s was quite modest. These findings support the general results for exter- nal university funds. Electrical engineering, computer science, chemistry, and biological sciences (including geography)15 occupy the next positions, whereas physics is quite low on the scale. This can be taken as an indication that the high general level of external funds in physics does not necessarily reflect a high share of industrial funds. The high absolute level of industrial funding for computer science is due to specialization in this area at the University of Karlsruhe. The highest growth rates can be observed for computer science, electrical engineer- ing, and biological sciences (including geography), which confirms the upward trend found in the general data for computer science and electrical engineering for the total amount of external funding. 

282 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY FIGURE 3.18 External funds from industry at the University of Karlsruhe, for selected areas, 1980 and 1990. SOURCE: Universität Karlsruhe (1995). A special finding at the University of Karlsruhe is the extremely high level of industrial funds in the school of civil engineering. (The level is twice as high as that in mechanical engineering.) However, the majority of these funds are raised as a consequence of the activities of two departments, which act as official certi- fying institutions in the field of building materials. Also, according to the chan- cellor of the university, many construction companies expect universities to con- duct the bulk of needed civil engineering research. The general statistics of the Science Council for “other engineering,” however, indicate that the dominance of civil engineering as a recipient of industrial funds cannot be generalized to Ger- many as a whole. To sum up, the major external sources of funds linked to technology transfer between universities and industry are collaborative research funded by BMBF and, at a much lower level, contract research paid for by industry. These activi- ties are concentrated in the natural sciences and engineering, especially the latter, leading to distinctly higher rates of industry funding than the average rates sug- gest. The funds for collaborative research as well as for research contracts in- creased considerably during the 1980s; the greatest growth was in the areas of electrical engineering and computer science. ADMINISTRATIVE STRUCTURES The different means of technology transfer at German universities are largely determined by the public status of these institutions. This obliges the universities 

TECHNOLOGY TRANSFER IN GERMANY 283 to manage their budgets according to rules established by public law. As public institutions, universities have to follow strict guidelines for budgetary planning and must balance revenues with expenditures. They are not allowed to make a profit. Funds must be clearly linked to projects or well-defined tasks and plan- ning must be done on an annual basis. This system leads to a certain rigidity and can hamper industry-oriented projects, where greater flexibility is needed. The universities have recognized this problem and—especially in the 1980s—intro- duced more flexible ways of managing external funds (see Selmayr, 1987). For example, universities are now allowed to make profits from externally funded projects and to use them to fill in the financial gaps between different contract projects. However, some problems still exist, like the inflexible handling of travel and related expenses for invited visitors. Furthermore, the administrative proce- dure for achieving more flexible solutions is quite complex. Most universities, however, have sufficiently experienced administrative staff to cope with these problems. As a consequence of their public status, universities are not allowed to engage in entrepreneurial activities. In particular, they cannot work with new technology-based firms. Furthermore, full professors are generally permanent civil servants (Beamte), and the other scientific staff are salaried public employees (Angestellte). Profes- sors and the scientific staff are employed full-time over the whole year. Projects financed by external sources are taken on in addition to teaching and “regular” research and do not lead to additional income. The external money becomes part of the university budget and the related research activities are considered regular activities (Dienstaufgaben). The major advantage for the professor in participat- ing in such research is the possibility of obtaining additional staff and equipment and enlarging the scope of his or her R&D activities. However, there is a second way to carry out projects for private clients. Pro- fessors have permission to take on secondary activities (Nebentätigkeiten), as long as their regular work is not restricted in a decisive way. Most states assume that the upper limit for these secondary activities is about one-fifth of the total work time (Hartl and Hentschel, 1989). Professors can retain the money from their secondary activities for their private use; hence, there is a strong financial incentive for this type of external activity. Secondary activities are subject to official approval, but in the case of professors, this is usually routine. In the case of scientific assistants, approval of secondary activities is rarely granted, how- ever. It is almost impossible for professors and scientific staff to engage in sec- ondary activites together. Another important aspect for technology transfer is the organizational struc- ture of German universities. The major bodies in charge of the distribution of institutional funds for teaching and research are the various schools (Fakultät, Fachbereich). The schools comprise a number of chairs (Lehrstühle) responsible for different areas of teaching. In the natural sciences and engineering, several chairs often establish joint university institutes. The institutes are the most inter- 

284 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY esting partners for industry because they have a sufficient critical mass and inte- grate professors from several disciplines. The additional staff financed on the basis of external funds is generally linked to institutes, not to chairs. Many insti- tutes employ 50 or more people and are the main users of external funds, espe- cially industrial funds. In the case of the University of Karlsruhe, the external funds are raised chiefly by such institutes, with a few large ones generating the greatest share of university income that comes from industry. This is true for other technology-oriented universities like those in Stuttgart, Aachen, Darmstadt, and Hannover. These large university institutes have a variety of links to industry (e.g., seminars for industrial experts, supportive associations that include indus- trial members). TRANSFER CHANNELS At German universities, the major channels of technology transfer are col- laborative research with industrial partners funded through BMBF projects and contract research for industrial clients. This statement is based on statistics for external university funds (see “Statistics on General Research Structures,” above), interviews with university professors conducted in the context of this report, and a special survey on the four focal areas (see “Technology Transfer in the Four Focal Areas,” below). From the perspective of the universities, a special advantage of contract re- search for industrial clients is the possibility of using the funds more flexibly than is possible with institutional funds or funds from public projects. Thus, paradoxi- cally, the rigidity of the public status of universities is actually a direct incentive for getting money from industry. Universities are not obliged to price contract research services on the basis of total costs. They can—but need not—exclude general overhead costs already covered by institutional funds. In consequence, they can offer these services at relatively low prices, which is a special advantage for industrial clients (see Püttner and Mittag, 1989, and “Statistics on General Research Structures,” above). Of course, other motives, such as the attainment of new research results, also play an important role. These will be discussed in detail in the section “Technology Transfer in the Four Focal Areas.” According to interviews and a survey, university professors view collabora- tive research with industrial partners as a very positive channel of technology transfer. It provides interesting insights into industrial research results and needs, and the resulting scientific independence is greater than in the case of contract research for industry. However, collaborative research seems to be less effective than is often assumed. First, the involved companies primarily expect to receive public funds and are often less interested in university-industry relations. Sec- ond, each partner linked in a common project generally works in its own labora- tory; the outcomes are presented solely at a few meetings. Third, the model of 

TECHNOLOGY TRANSFER IN GERMANY 285 cooperation involving several industrial enterprises and one or more university institutes leads to an orientation toward precompetitive research; hence, technol- ogy transfer mainly occurs in the early stages of the innovation cycle. Rarely are collaborative research projects followed by contract research projects, according to those interviewed for this report. In contrast to the American situation, in Germany, grants from industry to universities without clearly determined tasks and deliverables are exceptions. The main reason for this situation is obviously the German tradition of university- industry relations. In addition, the tax regime is not very favorable: Grants can be deducted from taxable income (tax deductions) as normal donations, or special expenses (Sonderausgaben), with upper limits. Therefore, companies contribute to the base funds of university institutes only in few cases of special common interest. One instance where industry contributes to university base funds is in the definition of focus projects in the area of biotechnology, in which research insti- tutions, industry, and the BMBF cooperate. For this purpose, eight so-called “gene centers” (Genzentren)16 were established for a limited period of about 10 years (Bundesministerium für Forschung und Technologie, 1993a). In these cen- ters, university institutes, Fraunhofer institutes, and Max Planck institutes coop- erate. The BMBF pays the major part of the project costs, and some large enter- prises contribute to the base funds as well as to the project costs. In the area of chemistry, the relations between industry and university are particularly close and have existed for many decades (see Herrmann, 1995). Chemical companies often have permanent consultancy contracts with professors or university institutes. Furthermore, the chemical industry has established a special fund (Fonds der Chemischen Industrie), which was set up as early as 1950, with the purpose of supporting university research. The university scien- tists personally get financial assistance for research purposes, but without clearly defined projects. Thus, the aid helps to enlarge the scope of university research, especially in basic activities. For the period from 1995 through 1997, the fund plans to grant a total of DM 21.7 million for these purposes (Fonds der Chem- ischen Industrie, 1995). All in all, the chemical industry provides a good example of how to improve stable long-term relations and cooperation with universities. Collaborative and contract research represent only a part of a broad range of technology transfer mechanisms. Universities present the results of their research in scientific articles, at fairs, and at conferences. Especially at conferences, uni- versity researchers meet with people from industry to discuss the applicability of research to industry. Furthermore, academic and industrial researchers often meet informally or discuss their problems in telephone conversations. Knowledge ex- change is broadly supported by a variety of scientific associations with academic as well as industrial members, which organize conferences and publish journals (Schimank, 1988b). As to the different channels of technology transfer, Allesch et al. (1988) asked university professors about the different forms of their “con- 

286 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY tacts with industry” (Praxiskontakte). The professors reported that informal consultancy and the provision of personnel (e.g., graduates) are by far the most frequent type of contact with industry. Of course, the intensity and length of these informal contacts are hardly comparable to what happens with formal con- tract or collaborative research, but the significance of informal channels of tech- nology transfer should not be overlooked. Consultancy by university professors for private clients is in general carried out as a secondary activity, because professors can obtain income in addition to their regular salary. In addition to consultancy and expert evaluations, professors also can conduct research for private companies as a secondary activity. If the professor uses staff or equipment from the university, he or she has to pay the related costs. According to Allesch et al. (1988), professors carry out 40 percent of their contracts with industry as secondary activities. (The present situation is discussed in the section “Technology Transfer in the Four Focal Areas.”) When contract research is a secondary activity, the legal requirements are quite complex and sometimes represent a barrier to cooperation with private en- terprises (Püttner and Mittag, 1989). The major constraint in this regard seems to be the regulation that contracts from third parties cannot be split into regular and secondary activities (Püttner and Mittag, 1989). Thus, it is not possible that in such a project, the professor carries out his or her part as secondary activity and the scientific staff is engaged therein as a regular activity. Since the scientific staff generally is only allowed to have regular activities, the professor must do his or her part as a regular activity, too. The professor may then no longer be inter- ested in the additional work because he or she receives no additional payment as in secondary activities. In the context of consultancy for industry, the role of polytechnical schools/ technical colleges (Fachhochschulen) has to be mentioned. In contrast to univer- sities, polytechnical schools are less science oriented. Their primary missions are education and the development of practice-oriented capabilities. Their research activities are rather limited; therefore, they are not explicitly mentioned in the statistical discussion above. Nevertheless, professors at polytechnical schools do considerable consulting for industry, especially regarding the solutions for prob- lems that emerge from the daily business on the companies’ shop floors. Further- more, students at polytechnical schools are obliged to prepare their master’s theses on subjects relating to industrial enterprises. In the state of Baden- Württemberg, the Steinbeis Foundation has established an effective network with polytechnical institutes for supporting technology transfer (see “Technology Transfer to Small and Medium-Sized Enterprises,” above). The major form of transfer of personnel to industry includes the provision of graduates and the permanent transfer of scientific staff from university to indus- try. In contrast, the temporary transfer of professors or scientific staff remains a rare event. Due to the requirements of public law, the short-term transfer of per- sonnel is linked to a variety of conditions and is not easy to put through (Püttner 

TECHNOLOGY TRANSFER IN GERMANY 287 and Mittag, 1989). In recent years, university administrations have become more open to this kind of technology transfer; but up to now, only minor changes can be observed. German universities, especially technical faculties, have a long-standing tra- dition of appointing high-level researchers from industry as professors (Wissen- schaftsrat, 1986). This leads to practice-oriented education and close research ties between universities and industry. In some cases, industrial firms endow professorships for a limited time period. As previously discussed, technology transfer from universities to industry was a major issue in science policy in the early 1980s. A highly visible outcome of the belated initiatives is the formation of technology transfer units at all uni- versities with schools of engineering or natural sciences. According to Kuhlmann (1991), these units serve primarily a so-called window function. In other words, they provide information to industry on the research capabilities of the university. Technology transfer units serve also a catalyst function by bringing industrial clients and university institutes or individual professors together. The units often help companies find the appropriate institute or professor to address specific prob- lems. Other, less dominant functions are the systematic monitoring of industrial needs, the negotiation of contracts with industrial partners, and the supply of services (e.g., business consultancy). To sum up, the transfer units have only a limited supportive function and cannot replace the transfer activites of the univer- sity professors discussed above (see, for example, the criticism of Reinhard and Schmalholz, 1996). The transfer units, being responsible for the university de- partments altogether, can only assist firms in finding appropriate professors. How- ever, the latter can present their specific research capacities more precisely than a general unit can, and have to build up the actual industry contacts. Nevertheless, the transfer units play a decisive role in making contact with SMEs, which are their major clients. Professors often do not have sufficient time to actively ad- dress this heterogeneous group and generally work with big companies. All in all, the transfer units have become an indispensable instrument of technology transfer. AN-INSTITUTES AND OTHER EXTERNAL INSTITUTIONS Professors can establish, as secondary activities, private institutes, as long as the legal limitations on work time are observed (Hartl and Hentschel, 1989; Tettinger, 1992). These institutes vary widely, from being completely indepen- dent to having close links to a university. A major advantage of external insti- tutes is that they make it easier to carry out applied research and development, which in general goes beyond the scope of the usual research activities of univer- sities. Further advantages are the simpler administrative procedures concerning contracts and the employment of scientific staff. To a certain extent, external institutes can help to enlarge the personnel and equipment capacity of universi- 

288 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY ties. Since the mid-1980s, most universities have supported the establishment of external institutes. In this context, it is worth noting again the quite inflexible regulations con- cerning employment of professors and university staff. These rules hamper both the temporary transfer of personnel from universities to industry and the estab- lishment of part-time employment contracts. The latter could in theory be a means of accommodating parallel activities of scientific staff at universities and external institutes. In practice, this instrument is rarely used. A special type of external institute is the so-called An-Institute. An-Insti- tutes are legally defined as independent bodies in order to achieve sufficient ad- ministrative flexibility (Krüger, 1995; Tettinger, 1992). They may have a com- pletely private or semipublic status. In most cases, they are nonprofit institutions and thus pay reduced taxes. Important common characteristics of all An-Insti- tutes are that they are officially acknowledged by a university and operate under a cooperation agreement. Some states (Länder) have official rules and regula- tions for An-Institutes. The main goals of An-Institutes are to • foster technology transfer and application-oriented research and develop- ment; • perform research in areas that are the focus of university research; and • perform research that does not fit into the administrative structures of universities. To summarize, An-Institutes are “mediators” between universities and in- dustry. Because of their legal independence, they have short decision paths and can react to market demands and opportunities in a flexible way. Furthermore, they can establish a business-oriented budgeting and accounting system. For example, they can freely use their budgets for special remunerations of their staff, for public relations activities, or for the further professional training of their re- searchers. For interested companies, especially SMEs, the research areas and competences of An-Institutes are more transparent than those of large universi- ties with a variety of faculties and internal institutes. This is a special advantage that helps An-Institutes integrate themselves into regional commercial activities. At the same time, An-Institutes have close relations to universities and thus good access to basic research. In most cases, the directors of An-Institutes are also regular (part-time) professors at universities and are engaged in teaching. Hence, they can employ the brightest students. Some critics fear that university research activities are being shifted to the An-Institutes and that universities are losing external funds from industry. In reality, universities generally profit from the industrially oriented activities of An-Institutes and acquire additional funds through the cooperation agreements. The various An-Institutes differ not only in their legal status, but also in the 

TECHNOLOGY TRANSFER IN GERMANY 289 scope of their research. Some An-Institutes have narrow markets linked to a special industry, for example VLSI design for the microelectronics industry (Institut für Mikroelektronik [IMS], Stuttgart). Others have broad markets, for example, software systems for the manufacturing industry (Forschungszentrum Informatik [FZI], Karlsruhe) or office technology for all industries (Oldenburger Forschungs- und Entwicklungsinstitut für Informatik-Werkzeuge und -Systeme [OFFIS], Oldenburg). The institutes with broad markets normally have multiple directors. As a general rule, An-Institutes carry out research in strategic areas, such as information technology and microelectronics. The various legal statuses of An-Institutes correspond to their diverse budget structures. In some states, like Baden-Württemberg, the An-Institutes receive one-third of their institutional funds from the state, one-third from contract re- search for industrial clients, and one-third from projects for public clients, such as the BMBF, the European Commission, the states, and so on. In this regard, the model of the An-Institutes is comparable to that of Fraunhofer institutes (see “The Fraunhofer Society,” below). However, many An-Institutes receive no pub- lic contribution to their institutional base and so depend almost totally on private and public contracts. In some cases, industrial partners provide some institu- tional funds. The main problem for An-Institutes is survival in a market that is dominated by competitors from large institutions with superior organization and connections (e. g., Fraunhofer institutes), more generous basic funding (e. g., national research centers), or hidden overheads (e. g., universities). Therefore, only An-Institutes with a special competence profile, close linkages to industrial partners, and dy- namic structures have a potential for long-term survival. The activities of private institutes and particularly An-Institutes represent a considerable portion of technology transfer. According to a recent official sur- vey, the R&D-related expenditures of An-Institutes amounted to DM 580 million in 1994, equal to 4 percent of the R&D expenditures of universities and about 50 percent of those of the Fraunhofer Society (Bundesministerium für Bildung, Wissenschaft, Forschung und Technologie, 1996).17 All in all, they have devel- oped into an important institutional sector that supports technology transfer. Technology centers and science parks are a further source of technology transfer. These facilities aim to establish technology-oriented enterprises in the vicinity of universities (Eberhardt, 1989; Wissenschaftsrat, 1986). Collaboration between universities and such enterprises may include the use by firms of the expertise of universities or the paid use of university equipment. From the per- spective of the German authorities, a clear legal distinction between universities and private companies is necessary to avoid any dependence of university re- search on the private sector. For instance, universities are not allowed to hold shares in industrial enterprises in technology parks. The companies in technol- ogy centers are often already well-established firms, which use the special facili- ties at universities. But the centers also support the establishment of spin-off 

290 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY companies. In the case of the technology center in Karlsruhe, 11.5 percent of all companies are spin-offs from universities, and 20 percent are spin-offs from other research institutions.18 TECHNOLOGY TRANSFER IN THE FOUR FOCAL AREAS Results of a German Survey Because available statistics and surveys do not contain any specific informa- tion about technology transfer in the four focal areas of this report, FhG-ISI con- ducted its own survey of German university institutes (not including An-Insti- tutes and other external institutes). The survey included institutes in the focal areas of production technology, microelectronics, software, and biotechnology19 and was conducted in May 1995. The addresses of presumably relevant institutes were determined with the help of a manual on universities and research institu- tions wherein the research area of each institute is briefly described (Vademecum, 1993). In all, 783 questionnaires were sent out, and 332 questionnaires with valid responses were sent back (Table 3.3). This response rate of 42 percent has to be considered very high, particularly since the description of the institutions in the manual often was quite poor, so that the selection of really appropriate institu- tions was difficult. The high response might have been due to the user-friendly design of the questionnaire, which had a limited number of questions, the major- ity of which could be answered by simple multiple choice. The first group of questions concerned the volume and composition of exter- nal funds.20 A striking result was the very high proportion of external research funds in all of the focal areas. The average share of external funds was 62 percent (Table 3.4), which is considerably higher than the average figures cited in official statistics for superordinate areas (e.g., in 1990, the figures were 41 percent for mechanical engineering and 33 percent for electrical engineering; see Figure 3.17). TABLE 3.3 Response Rate of Survey Sent to German Universities, by Focal Area Area Questionnaires Sent Out Questionnaires Sent Back Response Rate Production technology 185 97 52% Microelectronics 155 60 39% Software 175 68 39% Biotechnology 268 107 40% Total 783 332 42% 

TECHNOLOGY TRANSFER IN GERMANY 291 TABLE 3.4 Percent Share of University External Funds in Four Focal Areas, 1995 Share of External Share of Industrial Share of Secondary Funds Within Funds Within Activities Within Area Total Budget Total Budget Industry Contracts Production technology 68 25 11 Microelectronics 63 18 10 Software 43 13 16 Biotechnology 69 12 25 Total 62 17 15 Two reasons may help explain these differences: • Due to the increasing number of students, the relative share of research funds from institutional sources has diminished since 1990, and the uni- versities have become more active in the acquisition of external funds. • Since the questionnaire was clearly oriented toward technology transfer, primarily institutions with a high level of external funds answered (re- spondent bias). However, the main reason for the disparity is probably different. Expert interviews revealed that many professors are not aware of the real cost structures and do not sufficiently take into account the contribution of institutional funds to universities’ overhead costs (see the related discussion in “Statistics on General Research Structures”). Only some respondents answered in terms of money. In consequence, the results presented in Table 3.4 are a little bit lower than the real values in terms of personnel and time. Among the focus areas, the high share of external funds in production tech- nology is closely related to a high share of industrial income. It seems to be easier in biotechnology research than other technology areas to acquire external funding through BMBF, EU, DFG, and other sources. The average level of industry-related research within the total research ac- tivities is (as explained above) probably a little higher than 17 percent.21 In any case, it is far above the average level of about 8 percent for universities alto- gether. The industrial budget does not include collaborative projects with indus- trial partners funded by public sources (e.g., BMBF, EU), so the actual rate of industry-related activities is even higher. Production technology, microelectronics, software, and biotechnology are ranked first through fourth, respectively, in terms of the percentage of industrial funds that make up their total budgets. This ranking results because the focal areas are at different stages of their technology cycles, reflected by different de- grees of concentration on basic research. For example, in production technology, 

292 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY TABLE 3.5 Orientation of University R&D Activities, by Percent, 1995 Area Basic Research Applied Research Experimental Development Production technology 29 53 18 Microelectronics 41 47 12 Software 50 38 12 Biotechnology 66 27 7 Total 47 41 12 29 percent of R&D activities can be labeled as basic research; the amount of basic research is much higher in biotechnology (66 percent; Table 3.5).22 Thus, biotechnology is still at an early stage of development, whereas production tech- nology has already matured. It is interesting to note that in all areas, the univer- sities do not restrict their activities to basic and applied research but devote some effort to experimental development work. A rather interesting result shown in Table 3.4 is the relatively low level of secondary activities within industrial contracts (average 15 percent) compared with the results of Allesch et al. (1988), who found an average share of 40 per- cent. A partial explanation is that the relative share of regular activities has in- creased since 1984, when Allesch et al. conducted their survey. The greater flex- ibility of university administration has made the integration of industrial contracts into regular work easier (Wissenschaftsrat, 1993a). Second, Allesch et al. fo- cused on individual professors, whereas the present questionnaire included whole research teams. Thus, with respect to professors, the level of secondary activities is more important than Table 3.4 suggests. Secondary activities are still a rel- evant incentive for technology transfer. Survey respondents also were asked to assess the importance of different channels of technology transfer. As it is not possible to measure and compare the various channels using common quantitative units, the respondents could choose from among the statements “very important,” “important,” “somewhat impor- tant,” and “not important.” These assessments were meant to reflect the specific importance of the industrial contacts for the institution, not general opinions. For the analysis of the results, the statements were arranged in an ordinal scale from 1 (not important) to 4 (very important). In Table 3.6, the assessments of the differ- ent channels and the overall mean scores are recorded. The respondents regarded collaborative research as the most important trans- fer channel, with a mean score of 3.2. Despite the various points of criticism raised in accompanying interviews, this type of technology transfer, which is primarily supported by BMBF programs, seems to be very effective. Informal channels (e. g., telephone conversations or informal meetings; see also Rappa and Debackere, 1992) are second in importance, with a score of 3.0. Thus, the 

TECHNOLOGY TRANSFER IN GERMANY 293 TABLE 3.6 Channels of University Technology Transfer by Percent and Mean Score Very Somewhat Not Mean Important Important Important Important Score Cooperative research 53 25 12 10 3.2 Contract research 35 25 18 22 2.7 Consultancy 21 31 36 12 2.6 Informal contacts 34 39 20 7 3.0 Industry-related committees 10 23 32 35 2.1 Workshops, conferences 24 35 28 13 2.7 Organization of seminars 14 30 32 25 2.3 Exchange of scientists 16 25 30 29 2.3 Provision of personnel for industry 27 31 24 17 2.7 Exchange of pubications 10 25 36 28 2.2 Industrial participation in master’s and doctoral theses 29 33 20 17 2.7 establishment of appropriate conditions for the arrangement of informal meetings is important (e.g., the availability of travel funds and meeting rooms). With scores of 2.7, 2.6, 2.7, 2.7, and 2.7, respectively, contract research, consultancy, work- shops and conferences, provision of personnel for industry, and industrial partici- pation in master’s theses and doctoral dissertations are quite important, too. In contrast, participation in industry-related committees and the organization of seminars for people in industry generally are viewed as only somewhat impor- tant. The same applies to the exchange of publications, which is a major instru- ment for information exchange in academia but obviously is less important for industry contacts. The exchange of scientists was given a low score, a result that confirms the outcome of other studies. This low score means that the temporary exchange of scientists is rarely used. But according to the interviews with profes- sors, when used, the exchange of scientists has been very effective. The results, disaggregated according to the four focal areas, are similar to those for the total sample, but not completely uniform. For instance, contract research has a high score in the application-oriented area of production technol- ogy, and a low score in biotechnology with its distinct focus on basic reasearch. A detailed discussion of these differences, however, lies beyond the scope of this study. Not suprisingly, university researchers saw the availability of additional funds as the most important advantage of industry contacts (Table 3.7). How- ever, the opportunity to confer with industry had almost the same impact. Thus, technology transfer does not only flow from universities to industry, but aca- 

294 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY TABLE 3.7 Benefits to University Researchers from Contacts with Industry, by Percent and Mean Score (percent total sample), 1995 Very Somewhat Not Mean Important Important Important Important Score Additional R&D funds 66 22 8 5 3.5 Flexibility of industrial funds 51 25 14 10 3.2 Additional facilities 31 30 26 13 2.8 Opportunity to confer with industry 54 33 11 3 3.4 References for acquisition of public funds 22 33 27 18 2.6 demic researchers receive new intellectual input from industry as well. This find- ing was confirmed by interviews, in which university scientists emphasized the relevance of information from industry for their research and for improved, prac- tice-oriented teaching. As already explained in the context of administrative struc- tures, the flexibility of industrial funds compared with public funds is a major incentive for German universities to undertake contract research for industry. As to the barriers to industry contacts (Table 3.8), university researchers regard only the short-term orientation of their industrial partners as relevant (mean score of 2.9). All of the other reasons were “somewhat important” or even “not important” (scores between 1.8 and 2.3). The low score for administrative barri- ers confirms interview results indicating that today’s university administrations cope better with the problems of industrial contracts than in the 1980s (Selmayr, 1986). University researchers’ assessment of a limited indigenous industrial base as a barrier (mean score of 2.3) showed interesting differences in the four focal areas. (This internal differentiation is not indicated in the tables.) In biotechnol- ogy, the mean score was 2.6 (still an “important” barrier); in microelectronics, the TABLE 3.8 Barriers to Industry Contacts, by Percent and Mean Score, 1995 Very Somewhat Not Mean Important Important Important Important Score Less interesting topics 8 23 34 35 2.0 Industry’s short-term orientation 35 32 21 12 2.9 Restrictions of publications 10 26 39 25 2.2 Administrative problems 7 17 38 38 1.9 Unfair contracts 4 14 38 44 1.8 Limited industrial base in Germany 20 28 17 35 2.3 

TECHNOLOGY TRANSFER IN GERMANY 295 TABLE 3.9 Reasons for Industry Interest in University Research, by Percent and Mean Score, 1995 Very Somewhat Not Mean Important Important Important Important Score Observation of scientific development 40 42 16 3 3.2 Solution of technical problems 39 36 20 6 3.1 Personnel recruitment 26 43 25 5 2.9 mean score was 2.5; in software, it was 2.4; and in production technology, it was 1.9 (“somewhat important”). These results reflect the strong focus of German industry on all areas of mechanical engineering and a lower level of specializa- tion in information technology, microelectronics, and biotechnology. The university researchers were asked to describe what they believe to be the reason for industry’s interest in their research (Table 3.9). It is interesting that they ranked “observation of scientific development” even higher than “solutions to technical problems.” This ranking confirms once again researchers’ belief in a scientific dialogue between universities and industry on mid- and long-term ques- tions and an acknowledgment that industry needs solutions for its immediate tech- nical problems. In addition, the provision of qualified personnel—a basic func- tion of universities—plays an important role. As mentioned above, the relative importance to universities of different trans- fer channels, industry contacts, and barriers to working with industry are gener- ally the same in all selected areas. Nevertheless, differences in the absolute val- ues of the scores can be observed (Table 3.10). To demonstrate this effect, the mean scores of all responses to a group of questions were combined and then averaged. In the case of transfer channels, those working in production technol- ogy generally saw the different channels more positively than did those in bio- technology. The similarity of this result to the differences in industrial funding in these four areas is obvious. The same phenomenon emerges with respect to the TABLE 3.10 Average Mean Scores in Major Question Groups Area Channels Benefits Barriers Production technology 2.8 3.3 2.1 Microelectronics 2.8 3.2 2.2 Software 2.6 3.0 2.3 Biotechnology 2.3 2.9 2.2 Total 2.6 3.1 2.2 

296 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY benefits of industrial contacts, but to a lesser extent. In the group of questions dealing with barriers to industrial contacts, the differences between the areas are negligible although the scores are generally low. All in all, the contacts between universities and industry in the selected areas are above average, and universities are more engaged in technology transfer to industry than generally assumed. Of course, the differences between the ana- lyzed institutions are large, and technology transfer could be improved in many cases. Nevertheless, the potential for a further increase in technology transfer seems to be limited—at least in the selected focal areas. It is important to take the different stages in the technology life cycle into account. In biotechnology, for instance, a great increase in applied research and a corresponding reduction in basic research would be detrimental to the quality of research given the present stage of the area’s technology life cycle. Comparison with the American Situation The results presented above give interesting insights into how technology transfer occurs at German universities. It is informative to compare this with the situation at American universities. A direct comparison is not possible, because an equivalent U.S. survey does not exist. But Cohen et al. (1994) conducted a survey of the University-Industry Research Centers (UIRCs), which are in many respects comparable to German university institutes. For the purpose of the present study, Cohen et al. (1995) prepared a special analysis for the four focal areas. UIRCs are research centers at U.S. universities that get base funds from the federal government, mostly the National Science Foundation, on the precondition that they also raise money from industry. In most cases, the industrial funds are base funds, too, and are not linked to contracts with clearly determined deliver- ables. The funding companies, however, are involved in the general planning of research activities and have early access to research results. With respect to the four focal areas, Cohen and his colleagues received input from 411 UIRCs (Table 3.11), a magnitude of response comparable to the Ger- TABLE 3.11 Responses to the Survey of UIRCs, 1990 Area Number of UIRCs Production technology 109 Microelectronics 64 Software 129 Biotechnology 109 Total 411 SOURCE: Cohen et al. (1995, Table 3.1). 

TECHNOLOGY TRANSFER IN GERMANY 297 TABLE 3.12 Industrial Contributions to UIRCs, Percent Share by Area, 1990 Area Share Production technology 41 Microelectronics 30 Software 33 Biotechnology 21 Total 31 SOURCE: Cohen et al. (1995, Table 2). man survey (Table 3.3). Only in software was the German absolute response rate distinctly lower, but the remaining sample is still sufficiently large. A revealing outcome of the U.S. survey is the share of the industrial contri- bution to the funds of the UIRCs (Table 3.12). Like in Germany, U.S. centers devoted to production technology receive the highest share, those for biotechnol- ogy the lowest, and the area of microelectronics falls in the middle. The share of U.S. industrial funds for software R&D are comparable to, or even a little higher than, that for microelectronics, whereas in Germany funding for microelectronics research is near the level of funding for biotechnology. This difference may be due to closer university-industry relations in U.S. software development. The level of industry contributions to the UIRCs is generally higher than the average level of industry contributions to German universities, because the spe- cial mission of UIRCs is to improve technology transfer.23 In contrast, the Ger- man survey sample covered all types of university institutes and also included institutes with few industrial relationships. The U.S. survey, like the German one, asked respondents about the distribution of their R&D activities in basic research, applied research, and experimental development (Table 3.13). The dif- ferences between the four focal areas are less distinct in the United States than they are in Germany (Table 3.5), but the level of basic research in production technology is lowest in both Germany and the United States. The distribution of the three types of R&D activity in the United States is comparable to that in Germany for production technology and microelectronics. But the U.S. orientation toward basic research is clearly less pronounced in soft- ware and biotechnology. Of course, such comparisons are of limited usefulness, because the German and American interpretations of the different R&D types might be different. The higher U.S. level of applied R&D in software, however, correlates to the higher share of industrial contributions in this area. In the case of biotechnology, the difference between Germany and the United States is so large that it cannot be explained by a methodological bias. To summarize, the applica- tion orientation in German academic R&D is apparent in production technology 

298 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY TABLE 3.13 Orientation of R&D Activities at UIRCs, Percent Share, 1990 Area Basic Research Applied Research Experimental Development Production technology 32 46 22 Microelectronics 44 42 14 Software 38 44 18 Biotechnology 44 41 15 Total 44 43 18 SOURCE: Cohen et al. (1995, Table 3.3). and microelectronics and just as it is at U.S. UIRCs (Table 3.13). In contrast, the German university research in software and microelectronics appears to have a distinctly basic orientation. Unfortunately, the available U.S. data do not shed light on the extent to which other academic research outside UIRCs is oriented toward more basic activities. Like the German survey, the UIRC survey asked about the relevance of dif- ferent transfer channels (Table 3.14). Because of the different structures of Ger- man university institutes and American UIRCs, responses to the UIRC survey do not always have a counterpart in the German survey. Nevertheless, some com- parisons can be made. The U.S. scores, however, seem to be generally higher than the German ones (Table 3.6). This is due to different ways of analyzing the questionnaires. According to the German approach, all questionnaires are in- cluded as long as the respondents assessed the importance of some channels of technology transfer. The channels not marked by respondents were considered to be “not important.” According to the U.S. approach, however, only questions with a definite answer were included. If the German questionnaires are dealt with according to the U.S. method, the scores of German respondents rise and become comparable to the U.S. figures (Table 3.14). The only distinct difference con- cerns the temporary work of UIRC/university personnel in industry laboratories, where the U.S. score is clearly higher; in other words, the movement of personnel is less often a mode of technology transfer in Germany. The approach of the UIRC survey to assessing the benefits of industry con- tacts was different than that of the German survey, and the U.S. data are com- bined rather than separated out by the four focal areas (Table 3.15). The U.S. questionnaire asked whether or not the UIRCs see a benefit, without further dif- ferentiation, so that only the percentages of positive answers are available. The outcome, however, indicates that the U.S. and German respondents gave similar rankings to the value of “R&D funds,” “opportunity to confer with industry,” and “equipment.” In other words, the U.S. survey, like the Germany survey, revealed the importance of dialogue with industry for the advancement of academic research. As to the barriers to industry contacts, the U.S. survey asked only about restrictions on publication. Thirty-nine percent of UIRCs reported that partici- 

TECHNOLOGY TRANSFER IN GERMANY 299 TABLE 3.14 Channels of U.S. UIRC and German University Technology Transfer, Mean Score in the Four Focal Areas U.S. Mean Score (1990) German Mean Score (1996) Collaborative R&D projects 3.4 3.5 Seminars, workshops, symposiums 2.9 3.0 Research papers, technical reports 2.8 2.6 Telephone conversations 2.9 — UIRC personnel in industry labs 3.3 2.8 Industry personnel in UIRC 3.5 — Informal meetings with industry people 3.3 3.2 Delivery of prototypes or designs 3.4 — NOTE: German mean scores are calculated according to the method used in the U.S. survey. SOURCES: Cohen et al. (1995, Tables 18 to 22); survey by the Fraunhofer Institute for Systems and Innovation Research. pating companies can require information to be deleted from research papers be- fore they are submitted for publication; 58 percent said that companies can delay the publication of research findings, and 34 percent indicated that companies are able to both delay publication and have information deleted. The data do not indicate the actual frequency of these interventions. In Germany, the problem of publication restriction exists, too, but is generally less important (see Table 3.8). However, the Geramn and U.S. data sets are not really comparable due to the different types of questions asked. TABLE 3.15 Benefits of Industry Contacts at UIRCs, by Percent, and at German Universities, by Mean Score Percent Share of UIRCs German Mean Score R&D funds 91 3.5 Opportunity to confer with industry 70 3.4 Equipment 68 2.8 Information on industry needs 56 — Operational funds 49 — Access to industrial facilities 45 — Practical experience for students 38 — Research direction 36 — Industry personnel loaned to academic programs 22 — Other 6 — None of the above 1 — SOURCES: Cohen et al. (1994, Table 3.29); survey by the Fraunhofer Institute for Systems and Innovation Research. 

300 TECHNOLOGY TRANSFER SYSTEMS IN THE UNITED STATES AND GERMANY All in all, the results of the U.S. survey confirm the German outcome. They emphasize the importance of collaborative research and informal contacts for technology transfer and highlight shortcomings of the German system with re- spect to the difficulty of temporarily moving academic researchers into industrial laboratories. The data seem to indicate that German universities have less of an orientation toward applied research in software technology and biotechnology than their U.S. counterparts, a result that might be due to a lack of complete data for all types of research units of U.S. universities (i.e., not only UIRCs). PATENTS AND PATENT STATISTICS Intellectual property rights, especially patents, play an important role in tech- nology transfer. The particular situation at German universities is characterized by the privilege of professors to exploit for their own benefit inventions created during their work on institutional base funds at the university (Verwertungs- privileg). The consequences of this policy for technology transfer are contradic- tory. On the one hand, the private holding of patents can be an incentive, if the invention is generated within the framework of existing ties to industry. In this case, the patent is licensed or transferred directly to the industrial partner, leading to a generally moderate extra income for the professor. On the other hand, if no industrial partner is directly available, the professor has to pay the patent applica- tion fees at his or her own risk. Therefore, many inventions at universities are not patented. Later on, as a result, companies may not be interested in investing in further development because the basic idea has not been protected. If the research is funded by external sources, especially the BMBF, the uni- versity, not the professor, is responsible for patent protection. Due to the increas- ing relevance of external funding, the significance of the exploitation privilege is diminishing. However, the incentives for patenting by the universities themselves are low due to various factors. Among the most important are that most universi- ties have neither funds nor infrastructure to support patenting and licensing ac- tivities; inventions resulting from federally funded academic research generally can only be licensed on a nonexclusive basis to industrial partners; and a portion of any licensing income earned from developments with federal government funds must go back to the funding agency. In recent years, the University of Karlsruhe and the University of Dresden established patent and licensing offices comparable to those at American univer- sities. These offices offer professors advice on patent affairs and, if the invention seems to be marketable, provide financing for the patent application and search for potential licensees. (For more details, see Schmoch et al., 1996a.) Some federal states plan to start similar programs, with the aim of better supporting inventors at universities. The states do not wish to abolish professors’ exploita- tion privilege, but rather to offer institutional support. It is not possible to directly track German academic patents. However, the 

TECHNOLOGY TRANSFER IN GERMANY 301 German database PATDPA allows one to search for the title “Professor” among inventors or applicants. Such a search turns up not only university-related pat- ents, but also inventions by former professors now working in industry. Thus, the search sample is somewhat too broad and does not include inventions by scien- tific assistants at universities. Nonetheless, it can be assumed that the largest part of the search sample adequately reflects university patents.24 From 1980 to 1990, the number of patent applications registered for professors jumped by 46 percent (Figure 3.19). This rate of increase is comparable to that for external university funds (42 percent), lower than that for industrial funding (80 percent), but higher than that for overall university research budgets (21 percent). Obviously, the number of patents is linked primarily to the share of external funds. Remarkably, 54 percent of university patents are applied and owned by companies (Becher et al., 1996). These patents are obviously sold directly by the professors who have taken advantage of the exploitation privilege. It is interesting to note that the number of patent applications by German professors in 1992 was about 1,000, whereas the number of patent applications originating in American universities was about 2,500 (Association of University Technology Managers, 1993; Schmoch et al., 1996a). Despite the absolute dif- ference in patent activity between the two countries, the relative number in Ger- many in relation to the gross domestic product seems to be quite high. However, it has to be taken into account that U.S. universities reported about 8,000 inven- 1400 Professo (total) 1200 Professo Number of Applications (private 1000 800 600 N 400 200 0 8 0 8 1 8 2 8 3 8 48 5 8 6 8 7 8 8 8 9 9 0 9 1 9 2 9 3 Application y FIGURE 3.19 Patent applications to the German Patent Office by German university professors. NOTE: private = application by the professor; total = includes applications by firms or other institutions. SOURCE: Schmoch et al. (1996a).