SRI International

October 2000

OUTCOMES AND IMPACTS OF THE STATE/INDUSTRY-UNIVERSITY COOPERATIVE RESEARCH CENTERS (S/IUCRC) PROGRAM

SUMMARY REPORT

The views and opinions do not necessarily express those of the National Science Foundation

Report by:

David Roessner

Submitted By:

Science and Technology Policy Program

Prepared for:

The National Science Foundation
Engineering Education and Centers Division

TABLE OF CONTENTS

The main sections of this document can be accessed directly by selecting the appropriate heading below.

EXECUTIVE SUMMARY

Project Background
Selected Results
Conclusions and Observations

PART 1: BACKGROUND

Objectives and Design of the Evaluation
Program Origins, Goals, Requirements and Features

PART 2: RESULTS
PART 3: CONCLUSIONS AND OBSERVATIONS

Outcomes and Impacts of the State/Industry-University
Cooperative Research Centers (S/IUCRC) Program

EXECUTIVE SUMMARY

A. Project Background

NSF supported the present study first to answer the kinds of questions posed by most program evaluations: to what extent has the Program accomplished its goals? What lessons can be learned that might be used to improve the Program’s performance? In addition, NSF wished to identify lessons that could inform the agency’s planning for future NSF-state partnerships. The existence of the NSF Industry/University Cooperative Research (I/UCRC) Program, on which the S/IUCRC was modeled, offered an unusual opportunity to employ comparative analysis. Thus, in addition to outcome and impact data collected about the S/IUCRC Program itself, equivalent data were collected about the I/UCRC Program and used for comparative purposes. A comparative analysis of the two Programs enabled conclusions to be drawn about the independent effects—the "value added"—of the unique features of the S/IUCRC Program. These included explicit technology transfer activities associated with center research, the "non-core" projects sponsored by one or more industrial members that could involve proprietary work, a requirement for states to match NSF financial support, and the possibility of granting exclusive licenses to center members supporting non-core projects.

Most of the outcome and impact data developed for this study were generated from two surveys conducted by SRI: one of the official representatives of all organizations that were formal members of the nine S/IUCRCs initiated in 1991 and 1992 and still receiving NSF funding in 1998, and a second, similar survey of the official representatives of all organizations that are formal members of seventeen I/UCRCs that have existed for a similar period as the S/IUCRCs. The surveys were supplemented by an historical analysis of the origins and state context of the S/IUCRC Program, telephone interviews with State Representatives to S/IUCRCs, and careful review of S/IUCRC annual reports.

B. Selected Results

Center Activities. The profile of participation by member firms¹ in center activities is roughly similar for both types of centers, and is consistent with the profiles of member firm participation in other university-industry cooperative research centers. One difference that stands out is the significantly larger proportion of I/UCRC member firms that participates in sponsored research (61 per cent) relative to the proportion of S/IUCRC member firms that participates in non-core research (43 per cent). This is particularly surprising given that non-core research in S/IUCRCs was designed to be attractive to member firms as a source of solutions to problems, cost savings, and commercially applicable ideas and technology. None of the other potential explanations (given the data available in this study)--the location of member firms, the size of the firm, the length of the firm’s membership in the center--accounts for this difference. This puzzling result indicates that there may be substantial differences between S/IUCRC non-core research and I/UCRC sponsored research, differences that result in greater incentives to I/UCRC member firms.

Outcomes and Impacts on Member Firms. The pattern of participation in center activities by S/IUCRC member firms yields a wide range of specific results for them. By far the most frequently reported result was "obtained access to new ideas or know-how," a result deriving primarily from technical advice to members, core research, and training provided by center faculty (between 90 and 95 per cent of those participating in these activities). Following this was product or process improvement, a result reported by representatives of between 60 and 70 per cent of the member firms that participated in core research, access to center facilities, or receiving technical advice. Only a small proportion—19 per cent or fewer—of representatives of members participating in S/IUCRC activities reported that their firm received the following specific results:

Licensing technology or software developed by the center;
Patenting or copyrighting technology and/or software developed as a result of interacting with the center;
Making unexpected operational changes;
All other direct results.

There is a significant difference between the S/IUCRC and I/UCRC Programs in the extent to which participation in center activities of any kind results in more tangible outcomes related to intellectual property: new product/process development and licensing technology or software developed by the center. For each center activity, S/IUCRC member representatives report these relatively tangible results more frequently than do their I/UCRC counterparts. The higher frequency of results related to intellectual property outcomes reported by representatives of S/IUCRC member firms was not associated primarily with non-core research activities, where effects of the difference in intellectual property provisions of the two Programs should be most evident. Introduction of other factors such as member firm location, firm size, and length of center membership failed to account for the pattern of results observed. Although it seems reasonable to expect that the difference in intellectual property provisions plays some role here, the nature of the research actually undertaken and the actual practices regarding intellectual property protection among all centers, regardless of program, probably better explain this difference in results between the two Programs.

Representatives of member firms of both types of centers value center activities and the specific results that derive from participating in those activities roughly equivalently. As measured by these representatives’ qualitative assessments of the benefits of center membership compared with the costs, a slightly larger proportion of S/IUCRC member representatives estimated that the benefits their firms derive exceeded the costs (62 per cent vs. 50 per cent), but this difference is not statistically significant. A large majority—about three-quarters--of representatives to both types of centers consider the benefits of membership to at least equal the costs.

When member representative reports of specific results associated with their firms’ participation are analyzed by geographic location, it is clear that in-state location is a major influence on the type and number of specific results realized. Regarding member participation in center activities, a significantly larger proportion of in-state members of S/IUCRCs participates in non-core research than out-of-state members (60 per cent vs. 26 per cent), suggesting that S/IUCRCs are more responsive to in-state members, as the Program’s economic development goal anticipates. (Fifty per cent of S/IUCRC member firms are located in the same state as the centers in which they participate, vs. 41 per cent of I/UCRC members.) Representatives of in-state members of S/IUCRCs report a much wider range of beneficial outcomes than do out-of-state representatives. Further, these in-state member firms’ participation in non-core research was clearly the activity most likely to yield beneficial results relative to out-of-state members; and their participation in this activity resulted in a higher frequency of reports of access to new ideas, improved products or processes, improved technical information for customers or suppliers, and influence on R&D agendas.

Economic Impacts. With respect to member representatives’ qualitative assessments of the benefits and costs they associated with center membership, the differences between the two Programs are striking. In the case of the S/IUCRC Program, representatives of in-state members consider center membership to provide far greater benefits than costs than do representatives of out-of-state members (76 per cent vs. 47 per cent), whereas the reverse is true for I/UCRC member firms, although the difference is not as dramatic (46 per cent vs. 52 per cent). If only the extreme categories are considered, differences in impact by region (at least by this measure) are even more pronounced. Nearly 40 per cent of the in-state S/IUCRC member representatives, compared with 17 per cent of the out-of-state member representatives, regard membership benefits to be much greater than the costs, while for the I/UCRC Program, 25 per cent of the out-of-state member representatives vs. 18 per cent of the in-state representatives consider membership benefits to be much greater than costs.

Regarding member firm participation in specific center activities by region, a considerably greater proportion of in-state S/IUCRC members participates in non-core research than out-of-state members (96 vs. 64 per cent), whereas both in-state and out-of-state members of I/UCRCs participate in sponsored research in about the same proportion (87 vs. 93 per cent). A general pattern of more positive results for in-state members of S/IUCRCs is repeated and reinforced at the level of specific types of outcomes. Representatives of in-state members report a much wider range of types of beneficial outcomes than do out-of-state representatives, but the opposite is largely true for reports from representatives of I/UCRCs.

C. Conclusions and Observations

Based on the analyses conducted in this study, the S/IUCRC Program has been a modest success in the intermediate term as measured against its goals and objectives and compared with the outcomes and impacts of the I/UCRC Program. Centers in the S/IUCRC Program attracted considerably more in-state firms than out-of-state firms and a much greater proportion of small firms than did the I/UCRCs examined in this study. Participation rates in center activities were similar for both Programs, but the frequency and range of beneficial results related to more tangible research outcomes were greater for members of S/IUCRCs than members of I/UCRCs. Some of the S/IUCRC Program’s outcomes and impacts can be attributed to its unique features, some to the more general nature of university-industry cooperative research activities, and some to the characteristics specific to member firms in S/IUCRCs. Others remain unexplained, at least by the data available in this study. For example, the low participation rates of S/IUCRC member firms in non-core research, relative to the participation rates of I/UCRC members in sponsored research, its assumed counterpart, are puzzling and call for further investigation. Outcomes related to intellectual property occurred more frequently for S/IUCRC member firms than for I/UCRC member firms, but could not be clearly identified with the intellectual property provisions attached to S/IUCRC non-core research that are unique to that Program.

The difficulties that State Representatives associated with justifying matching funds for S/IUCRCs suggest ways in which future NSF partnerships with states that involve university-industry collaborative research programs might be designed. First, from the states’ perspective, the leverage that the S/IUCRC matching requirement provides is not high relative to other opportunities. It is seen as one-to-one: one state dollar leverages one federal dollar independent of the match provided collectively by member firms. Second, the typical S/IUCRC budget is not large, about $1 million. States regard this as too small to generate many jobs or result in significant economic impact relative to other university-industry alternatives such as Engineering Research Centers or Science and Technology Centers. Third, small businesses are hard to attract to S/IUCRCs because they are viewed as needing a broad range of support, especially business assistance, which is difficult or impossible to provide within the requirements of the S/IUCRC Program. Finally, a state appears to be better able to incorporate an S/IUCRC into its own priorities and agendas if the center can be folded into an established state program intended to foster university-industry cooperation.

The argument for an NSF-state partnership that targets industry-university cooperation is as valid now as it was when the National Governors Association and NSF first discussed such a program in the late 1980s. The design of a future program, however, must account for the lack of integration in many states between economic development plans and programs with S&T plans and programs, and for variations in the nature of state agencies responsible for support of universities. Given the experience of the S/IUCRC Program, it would seem prudent for a future program whose goal is to strengthen state S&T infrastructures rather than promote regional economic development, with university-industry cooperative research as the means. The form that such cooperative research programs might take would vary considerably, depending on state priorities and needs, and on existing state commitments to such cooperative arrangements.

In future partnerships with states, should NSF, in collaboration with representatives of states, design programs around ends or means? The S/IUCRC Program was designed around means: industry-university research collaboration was the vehicle for implementing the partnership, and the assumption was that the success of the I/UCRC model made it a suitable guide, with some modification, for a state partnership program. It proved more difficult than expected to modify the I/UCRC model, intended as it was to increase the national competitiveness of industrial firms, in a way that would achieve regional economic development through industry-university cooperative research. The difficulties were of two types. First, the model itself probably was an inappropriate means to achieve the goal of regional economic development. Second, wide variations across states in the organizational and political relationships between science and technology-related programs and strategies on the one hand, and economic development programs and strategies on the other, were incompatible with the inflexible requirement for annual state matching being linked to an economic development goal.

To sum up, the S/IUCRC Program was a valuable experiment in NSF-state partnerships. The Program succeeded in at least two ways. It accomplished, to a modest degree, its goal of promoting state economic development. And it demonstrated that explicit requirements such as state matching, explicit attention to technology transfer, industry-oriented intellectual property provisions, and encouragement of small and medium-sized firms as center members, can contribute to the achievement of this goal, even if applied to a program model that may not have been ideally suited to the context. The findings and conclusions from this study suggest that NSF has the ability to develop future partnerships with the states that will lead to benefits valued highly by both partners.

OUTCOMES AND IMPACTS OF THE STATE/INDUSTRY-UNIVERSITY
COOPERATIVE RESEARCH CENTERS (S/IUCRC) PROGRAM

PART 1: BACKGROUND

A. Objectives and Design of the Evaluation

The NSF State/Industry University Cooperative Research Centers Program, initiated with six centers in 1991 and four more in 1992, drew upon the Foundation’s favorable experience with other programs intended to promote industry-university cooperation in research to design a related program that would incorporate the interests of state governments. In particular, the S/IUCRC Program called for a "new partnership" between the federal government and the states to promote industry-university cooperation, but in a way that would bring explicit benefits to states as well as the nation. In addition to promoting university-based, industrially-relevant research (a goal common to other NSF centers programs), the S/IUCRC Program was to "facilitate diffusion of center research results and technology innovation with a view to impacting economic development. Technology transfer is considered part of the core program." Thus economic development that can be attributed to an S/IUCRC and is confined to the state hosting the center is clearly one desirable, expected impact of the Program. According to the original NSF Program Announcement (NSF 90-87), the S/IUCRCs were to achieve their objectives through programs of fundamental "core" research, industry-sponsored "non-core" projects that could be proprietary, specific technology transfer efforts involving both types of research, and the explicit involvement of small and medium-sized firms.

The existence of the NSF Industry/University Cooperative Research Program since the late 1970s and its rich set of existing center-level data offered an unusual opportunity to employ comparative analysis in the present evaluation, and full advantage of this opportunity was taken. Thus, in addition to outcome and impact data collected about the S/IUCRC Program itself, equivalent data were collected about the I/UCRC Program and used for comparative purposes. The two Programs share important similarities in structure and programmatic activities that make comparative analysis feasible: both types of centers consist of groups of university-based industrial firms and other organizations that provide membership fees; and both engage in fundamental research supported by pooled membership fees, project research often funded by a single member, research facilities that could be accessed by members, technical advice and consulting with center researchers available to members, and training for member organizations’personnel. Equally important, the two Programs differ in ways that highlight the distinctive goals, objectives, and requirements of the S/IUCRC Program. Thus a comparative analysis of the two Programs enables conclusions to be drawn about the independent effects—the "value added"—of the unique features of the S/IUCRC Program – the technology transfer activities, inclusion of small and medium-sized firms, and the local economic impact expectation. As a consequence, SRI’s evaluation of the S/IUCRC Program has two elements: a description and analysis of the outcomes and impacts of the Program itself (which can be compared with the Program’s objectives), and a description and analysis of the Program’s outcomes and impacts relative to those of the I/UCRC Program. These two primary elements of the evaluation were supplemented by impact data obtained from annual reports of the nine S/IUCRCs studied, by an historical and political analysis of the origins and state context for the S/IUCRC Program, and by interviews with the State Representatives to seven of these centers.

Most of the outcome and impact data developed for this study were generated from two surveys conducted by SRI: one of the official representatives of all members (i.e., those organizations that were recognized as formal, fee-paying members to centers as of 1998) of the nine existing S/IUCRCs that were initiated in 1991 and 1992 and still receiving NSF funding in 1998, and a second, similar survey of the official representatives of all firms that were formal members of seventeen I/UCRCs that had existed for at least a similar period as the S/IUCRCs. This research strategy was based on the premise that the major, direct impacts of both Programs are experienced by the firms that are fee-paying members of centers. The primary link from center research, technology transfer, technical assistance and other activities to the regional economy is through member firms’ cost savings, sales increases, employment changes, capital investments, and other changes that have economic consequences for the regional economy and can be attributed to center membership. The official representatives to centers from member firms were assumed to be in the best position to identify the range of impacts that center membership has produced for them, and to estimate the value of those impacts on their firm’s economic performance and competitiveness, and so were the designated respondents to the surveys. As noted above, the surveys were supplemented, in the case of the S/IUCRC analysis, by telephone interviews with state economic development officials and state science and technology policy officials and careful review of S/IUCRC annual reports.

B. Program Origins, Goals, Requirements and Features

In 1989 the Science and Technology Council of the States, a unit of the National Governors’ Association (NGA), and NSF established a working group to explore an expanded role for states in NSF engineering research programs. This working group’s activity led to creation of the S/IUCRC Program, based on the model of university-industry cooperative research centers established by the NSF I/UCRC Program. The S/IUCRC Program has a number of unique features, relative to the I/UCRC Program, related to its origins as a program of partnership with states. The most important of these features were intended to promote industrially-relevant research with more immediate potential for payoffs to industry. The mechanisms for accomplishing this included explicit technology transfer activities, a program of "non-core" projects sponsored by one or more industrial members that could involve proprietary work, and the possibility of granting exclusive licenses to members supporting non-core projects. A key element of the S/IUCRC Program, originating with the involvement of the NGA in the Program’s design, was the requirement that NSF annual support must be matched by state governments. A related requirement was the appointment of a State Representative to coordinate each center’s activity with the state’s economic development programs and policies. Finally, the S/IUCRC Program requires its centers to "work closely with industry (including small businesses) to facilitate diffusion of center research results and technology innovation with a view to impacting economic development." Table S-1 summarizes the goals, requirements, operating conditions, and practices of the two Programs in a comparative format, and Table S-2 summarizes their formal requirements.

Table S-1.
A Comparison of the Goals, Activities, Requirements, Operating Conditions, and Practices of the S/IUCRC
and I/UCRC Programs

S/IUCRC Program I/UCRC Program
GOAL	Advance state economic development	Strengthen cooperative research relationships between universities and industry
ACTIVITIES	Provide support for a "core" research program of industrially-relevant generic and fundamental research.	Provide support for a research program of industrially-relevant generic and fundamental research.
	Engage in experimentation of technical feasibility to promote technology advancements, technology implementation services, and "non-core" sponsored projects with more immediate potential.	Sponsored research (optional)
	Work closely with industry (including small and medium-sized businesses) to facilitate diffusion of center research results and technology to impact states' economic development through the center's "core" technology transfer program.
FORM OF AWARD	Cooperative agreement	Grant
LENGTH OF AWARD	Initial award of 4 years, with possible renewal for an additional 4 years.	5 year award; centers may apply for a second period of up to 5 years
MEMBERSHIP	Members may be industrial firms or other organizations such as federal laboratories.	Members may be industrial firms or other organizations such as federal laboratories
NSF SUPPORT	$150K, $225K, or $300K annually during award period.	$100 K annually for first 5 years, $50K annually for the second 5 years.
OTHER SUPPORT	NSF annual support must be matched by by the state government and by the center membership fees.	Centers must obtain at least $300K annually in cash membership fees from a minimum of six center members.
	Host universities provide cost sharing through in-kind support (not required).	University cost sharing required.
TREATMENT OF INTELLECTUAL PROPERTY	Core research and technology transfer programs provide non-exclusive, royalty-free licenses or center patents to member companies. Exclusive licenses are permitted for non-core, sponsored projects.	Participating university holds patents on NSF-supported inventions, with participating firms generally receiving royalty-free, nonexclusive rights or options.
		Some centers have by-laws in which non-exclusivity is the default but might allow for non-exclusive royalty-free rights for all members; limited but royalty-bearing licenses for a member that shares in filing costs; exclusive license if only one member is interested; or exclusive license for a non-member if no members are interested.
OTHER REQUIREMENTS		Uniform membership agreements

SOURCRE: http://www.eng.nsf.gov/eec/O-intro.htm; https://www.nsf.gov/pubs/1998/nsf97164/nsf97164.htm: D.O. Gray and S. G. Walters, Managing the Industry/University Cooperative Research Center: A Guide for Directors and Other Stakeholders, Columbus, OH: Baltelle Press, 1998; NSF/EEC staff.

Table S-2
Formal Requirements of the S/IUCRC and I/UCRC Programs

Program Feature S/IUCRC I/UCRC
Fundamental shared, core research	Required	Required
Directed projects sponsored by one or a few firms for its interest	Required	Not required (or desired by NSF)
Scope of impact	State	National
Emphasis on industry input to project selection with evaluator/ center analyst	Required	Required
Industrial support	Required	Required
State support	Required	Not Required
Core technology transfer activities	Required	Not Required

PART 2: RESULTS

The S/IUCRC and I/UCRC Programs are structurally similar programs that encourage university-industry cooperation through the formation of university-based groups of firms pooling some of their resources to support fundamental research. Thus, some of the patterns of involvement in centers by member firms and the results experienced by them should be similar. Differences in these patterns will be due primarily to:

Unique features of the S/IUCRC Program, as described above;
Features of individual centers such as leadership, organizational setting, level of financial and other support, staffing, etc.; and
Characteristics of center member firms and their representatives.

Only the first and third of these types of factors were considered in this evaluation.

Center Activities. As a requirement of membership in either type of center, member firms provide fees that are pooled to support a program of fundamental, generic research. Virtually by definition, most member firms would be expected to participate at some minimal level in core research (S/IUCRCs) and fundamental research (I/UCRCs). Participation in all other center activities is discretionary, although the program of non-core research is required of S/IUCRCs, whereas its assumed counterpart, sponsored research, is not required of I/UCRCs. If these two activities are, in fact, reasonably similar, then it might be expected that member participation in non-core research in S/IUCRCs will occur relatively more frequently than participation in sponsored research in I/UCRCs. But in at least one obvious respect, stemming from the unique intellectual property provisions of the S/IUCRC non-core program, the two activities are not comparable. Specifically, while I/UCRCs are not encouraged to develop intellectual property protection for products of their research, the opportunity to grant exclusive licenses for the products of non-core research was expected to provide an incentive for S/IUCRC member firms that does not exist for members of I/UCRCs.

Nearly all members of the nine S/IUCRCs studied participate in core research—at minimum they pay their dues, participate in research project selection, and are sent information about research results. Forty-three per cent of the members participate in non-core research, about 60 per cent use center facilities and receive technical advice or consulting from center personnel, and just under a third received training from their center. In the case of the I/UCRC Program, essentially all members of the I/UCRCs studied participate in fundamental research as a consequence of their membership fees. Sixty per cent support sponsored research; 42 per cent use center facilities; 66 per cent receive technical advice or consulting; and about a quarter receive training. Although the profile of participation in center activities is roughly similar for both types of centers, one difference that stands out is the significantly larger proportion of I/UCRC member firms that participate in sponsored research (61 per cent) relative to the proportion of S/IUCRC member firms that participate in non-core research (43 per cent). This is particularly surprising given that non-core research was designed to be attractive to member firms as a source of solutions to problems, cost savings, and commercially applicable ideas and technology. None of the other potential explanations (given the data available in this study)--the location of member firms, the size of the firm, the length of the firm’s membership in the center--accounts for this difference. This puzzling result is thus due to a factor not included in the analysis, and may indicate that there are substantial differences between S/IUCRC non-core research and I/UCRC sponsored research, differences that offer greater incentives to I/UCRC member firms.

Outcomes and Impacts on Member Firms. Two formal requirements of the S/IUCRC Program not shared by the I/UCRC Program are state financial support and the inclusion of technology transfer activities as part of the program of research. Both of these requirements reflect the S/IUCRC Program’s goal of state economic development, and predict that, relative to I/UCRC member firms, S/IUCRC member firms might be more likely to be located in the same state as the center with which they are involved (as a consequence of the center’s recruiting strategies and explicit attention to technology transfer). The proportion of small firms choosing to become members of S/IUCRCs would also be expected to be greater than in the case of I/UCRCs, since other research has suggested that large firms are more likely to join university-based groups of firms than are small firms, for reasons related to the kinds of benefits firms more frequently experience from such involvement (longer term, less tangible benefits such as access to students, ideas, expertise, and know-how rather than shorter term problem solving and new technology). In particular, the S/IUCRC Program’s differential emphasis on technology transfer and on the involvement of small and medium-sized firms should produce this result. Despite these differences, however, both Programs share the common core of fundamental research, so it would be surprising if the profile of results obtained by member firms of these two types of centers differed widely from the profile of results recorded in studies of other industry-university cooperative research activities: the most frequently experienced results have longer term and relatively intangible benefits for the involved firms; students are highly valued; specific, tangible results such as new products and processes, licensed technology or software, and patentable technology are relatively infrequent outcomes (but highly valued when they do occur).

The pattern of participation in center activities by S/IUCRC member firms yields a wide range of specific results for them. By far the most frequently reported result was "obtained access to new ideas or know-how," a result deriving primarily from technical advice, core research, and training (between 90 and 95 per cent of those participating in these activities). Following this was product or process improvement, a result reported by between 60 and 70 per cent of the member firms that participated in core research, access to facilities, or receiving technical advice. Only a small proportion—19 per cent or fewer—of representatives of members participating in center activities reported that their firm received the following specific results:

Licensing technology or software developed by the center;
Patenting or copyrighting technology and/or software developed as a result of interacting with the center;
Making unexpected operational changes;
All other direct results.

In the case of the I/UCRC Program, by far the largest proportion of member representatives (82-95 per cent) reported that, as a result of their firm’s participation in center activities, they obtained access to new ideas or know-how. About 30 per cent of the representatives reported that their firm developed a new product or process, with project research as a the primary route to this result, although other activities except training led to new products or processes at a just slightly lower rate. Less than 9 per cent reported licensing technology or software developed by the center, or patenting or copyrighting technology or software developed as a result of interacting with the center. Between 10 and 20 per cent reported making unexpected operational changes.

There is a significant difference between the two Programs in the extent to which participation in center activities of any kind results in more tangible outcomes related to intellectual property: new product/process development and licensing technology or software developed by the center. For each activity, S/IUCRC member representatives report these relatively tangible results more frequently than do their I/UCRC counterparts. One possible explanation for this is the difference in the guidelines for intellectual property protection provided by each type of center to its member firms. However, the higher frequency of results related to intellectual property outcomes reported by representatives of S/IUCRC member firms was not confined to non-core research activities, where effects of the difference in intellectual property provisions of the two Programs should be most evident. Introduction of other factors such as member firm location, firm size, and length of center membership factors failed to account for the pattern of results observed. Although it seems reasonable to expect that the difference in intellectual property provisions plays a role here, the nature of the research actually undertaken and the actual practices regarding intellectual property protection in each type of center probably better explain this difference in results between the two Programs.

Geographic proximity is clearly a factor in the levels of member participation in center programs and, consequently, in the benefits derived from participation. When member representative reports of specific results associated with their firms’ participation are analyzed by geographic location, it is clear that in-state location is a major influence on the type and number of specific results realized. Regarding member participation in center activities, a significantly larger proportion of in-state members of S/IUCRCs participates in non-core research than out-of-state members (60 per cent vs. 26 per cent), suggesting that S/IUCRCs are more responsive in their research program to in-state members, as the Program’s economic development goal anticipates. Representatives of in-state members of S/IUCRCs report a much wider range of beneficial outcomes than do out-of-state representatives. Further, these in-state member firms’ participation in non-core research was clearly the activity most likely to yield beneficial results relative to out-of-state members; and their participation in this activity resulted in a higher frequency of reports of access to new ideas, improved products or processes, improved technical information for customers or suppliers, and influence on R&D agendas.

Although less prominent than regional economic development as a program goal, the S/IUCRC Program encouraged centers to recruit and retain small and medium-sized firms as members. The mean percentage of small firms represented across the nine centers studied was 41.2; looking only at the in-state members, 45 per cent are small businesses. Small firms participate in center activities in about the same proportion as larger firms and, although the differences are not large, non-core research and technical advice are slightly more highly valued than core research and access to facilities by small and medium-sized firms (representatives of only three small firms assigned a value to the training they received, and the mean value was very high). Representatives of small firms assessed their overall ratio of benefits to costs in about the same way that representatives of larger firms did. Eleven per cent of the I/UCRC member firms are small businesses. A somewhat larger proportion of these small firms are in-state than out-of-state: 14 per cent vs. 10 per cent. Across center activities, slightly fewer small firms seek training than do larger firms, and slightly more participate in project research. Representatives of small firms assess the comparative costs and benefits of center membership in the same way that representatives of larger firms do.

Economic Impacts. Despite SRI’s efforts to use a variety of measures of regional economic impact, lack of reliable data limited this study to reports of various types of benefits by representatives of member firms, classified by region—whether or not the member firm was in the same state as the center in which it participated. First, with respect to member representatives’ qualitative assessments of the benefits and costs they associated with center membership, the differences between the two Programs are striking. In the case of the S/IUCRC Program, representatives of in-state members consider center membership to provide far greater benefits than costs than do representatives of out-of-state members (76 per cent vs. 47 per cent), whereas the reverse is true for I/UCRC member firms, although the difference is not as dramatic (46 per cent vs. 52 per cent). If only the extreme categories are considered, differences in impact by region (at least by this measure) are even more pronounced. Nearly 40 per cent of the in-state S/IUCRC member representatives, compared with 17 per cent of the out-of-state member representatives, regard membership benefits to be much greater than the costs, while for the I/UCRC Program, 25 per cent of the out-of-state member representatives vs. 18 per cent of the in-state representatives consider membership benefits to be much greater than costs. Only 6 per cent of the in-state S/IUCRC member representatives regard the benefits to be much less than the costs, while 18 per cent of the in-state I/UCRC representatives consider the benefits to be much less than the costs; 20 per cent of the out-of-state member representatives of S/IUCRCs and 12 per cent of the out-of-state representatives of I/UCRCs had this negative response.

Regarding member firm participation in specific programmatic activities by region, a considerably greater proportion of in-state S/IUCRC members participates in non-core research than out-of-state members (96 vs. 64 per cent), whereas both in-state and out-of-state members of I/UCRCs participate in sponsored research in about the same proportion (87 vs 93 per cent). Geographic proximity matters (positively) to members of S/IUCRC in obtaining technical advice or consulting (71 per cent vs. 53 per cent) but much less so for members of I/UCRCs (71 per cent vs. 64 per cent). This general pattern of more positive results for in-state members of S/IUCRCs is repeated and reinforced at the level of specific types of outcomes. In the case of the S/IUCRC Program, in-state members report a much wider range of types of beneficial outcomes than do out-of-state members, but the opposite is largely true for members of I/UCRCs.

Another measure of the economic impact of centers is the number of center graduates or students hired by member firms, and the geographic pattern of those hires. Neither the S/IUCRC nor the I/UCRC Program has an explicit educational component, so that the hiring of center students and graduates is not an expected outcome of either Program. However, several studies of the impact on industrial firms of participation in university-industry cooperative research (e.g., SRI’s study of the impact on industry of participation in ERCs) consistently show that companies value human capital—the staff with whom they interact and the students and graduates they hire—more than many other products of university research. By this measure, the S/IUCRC program does not do as well as does the I/UCRC Program. Proportionately fewer members of S/IUCRCs hire center graduates or students (17 per cent vs. 33 per cent); and in-state members are no more likely than out-of-state members to hire students or graduates (although all members who hired them value them highly). In contrast, 43 per cent of the I/UCRC member firms that hired center students or graduates were in-state, while 28 per cent were out-of-state. This result is not readily explainable but may be a consequence of the higher concentration of small firms—small firms that are not apparently hiring or can attract center graduates—in the S/IUCRC program.

As part of this study, SRI reviewed selected state science and technology (S&T) and economic development plans and strategies. The review revealed a lack of consistency and integration between the two in many states. In the small number of states in which both kinds of planning occurred formally, often the groups and agencies involved were different, and the resulting plans had different emphases. Typically a structural and political gap exists between state economic and S&T strategies and activities. Only a few states have successfully integrated S&T quickly, smoothly, or permanently into existing agencies responsible for economic development; in other states the loci of political support and stakeholders involved in S&T differ from those involved in economic development.

SRI’s interviews with the State Representatives to S/IUCRCs revealed in some detail the tensions implied by the inconsistencies and gaps just described. First, from the states’ perspective, the leverage that the S/IUCRC matching requirement provides is not high relative to other opportunities. It is seen as one-to-one: one state dollar leverages one federal dollar. Second, the typical S/IUCRC budget is not large, about $1 million. States regard this as too small to generate many jobs or result in significant economic impact relative to other university-industry programs such as ERCs or Science and Technology Centers. Third, small businesses are hard to attract to such centers because they are viewed as needing a broad range of support, especially business assistance, which is difficult or impossible to provide within the requirements of the S/IUCRC Program. A state appears to be better able to incorporate an S/IUCRC into its own priorities and agendas if the center can be folded into a broader state program intended to foster university-industry cooperation.

PART 3: CONCLUSIONS AND OBSERVATIONS

To increase the payoffs for states from university-industry cooperative research, the federal government can encourage state-based firms to join university-industry research centers, encourage the conduct of shorter-term, industrially-relevant research, facilitate the transfer of results from university research to industry, and encourage intellectual property guidelines that offer incentives for industry to collaborate with university researchers. The S/IUCRC Program was designed to accomplish all of these results. Unique among federally-supported, university-industry collaborative research programs is the S/IUCRC Program’s requirement for state matching of federal (NSF) and industrial funds for centers. This requirement implies a more systematic integration of this Program into state S&T and economic development plans than is the case for other state-supported industry-university cooperative research programs. At least initially, the S/IUCRC Program must compete for state funds with other federal programs that require matching. The formal matching requirement, which calls for new money in state agency budgets rather than a reallocation of university funds or passthroughs, can result in tension between economic development agencies or S&T agencies and state Boards of Regents.

SRI’s interviews with State Representatives to S/IUCRCs provided additional information about the complex setting in which state-supported university-industry collaborative research programs exist. Over the more than seven years that the nine S/IUCRCs that SRI studied have been in existence, State Representatives initially designated to oversee the centers have remained largely stable in both organizational location and identity. About half the Representatives (and thus the oversight activity and initial budget justifications for matching funds) are located in state or nonprofit agencies responsible for economic development, and about half are located in some form of state S&T agency. Representatives report that matching funds typically were approved initially as part of a larger agency budget request to the state legislature (e.g., the Ohio Edison Program, the New York State Centers for Advanced Technology within the NY State Office of Science, Technology and Academic Research). Subsequent budget requests proved somewhat easier to justify, since they often occurred within the request of a larger, more well-established entity. Yet justification for S/IUCRC matching funds was not straightforward, either initially or subsequently.

The difficulties that State Representatives associate with justifying matching funds for S/IUCRCs are revealing. They reflect the tensions discussed briefly above, and provide suggestions for ways in which future NSF partnerships with states that involve university-industry collaborative research programs might be designed. First, from the states’ perspective, the leverage that the S/IUCRC matching requirement provides is not high relative to other opportunities. It is seen as one-to-one: one state dollar leverages one federal dollar. Second, the typical S/IUCRC budget is not large, about $1 million. States regard this as too small to generate many jobs or result in significant economic impact relative to other university-industry programs such as ERCs or Science and Technology Centers. Third, small and medium-sized businesses are hard to attract to S/IUCRCs because they are viewed as needing a broad range of support, especially business assistance, which is difficult or impossible to provide within the requirements of the S/IUCRC Program. A state appears to be better able to incorporate an S/IUCRC into its own priorities and agendas if the center can be folded into a broader state program intended to foster university-industry cooperation.

Based on the analyses conducted in this study, the S/IUCRC Program has been a modest success as measured against its goals and objectives and compared with the outcomes and impacts of the I/UCRC Program that served as its model. The S/IUCRC Program attracted considerably more in-state firms than out-of-state firms and a much greater proportion of small firms. Participation rates in center activities were similar for both Programs, but the frequency and range of beneficial results related to more tangible research outcomes were greater for members of S/IUCRCs than members of I/UCRCs. Some of the S/IUCRC Program’s outcomes and impacts can be attributed to its unique features, some to the more general nature of university-industry cooperative research activities, and some to the characteristics of center member firms. Others remain unexplained, at least by the data available in this study. For example, the low participation rates of S/IUCRC member firms in non-core research, relative to the participation rates of I/UCRC members in sponsored research, its assumed counterpart, are puzzling and call for further investigation. Outcomes related to intellectual property occurred more frequently for S/IUCRC member firms than for I/UCRC member firms, but could not be clearly identified with the unique intellectual property provisions attached to S/IUCRC non-core research.

What lessons for NSF does this study suggest for future partnerships with the states? NSF’s prior experience with formal intergovernmental partnerships is limited primarily to the series of programs initiated by the Intergovernmental Science and Public Technology Program in 1967 and extending until 1982, and the EPSCoR Program, begun in 1978 and continuing today in expanded form. The latter Program’s survival can be attributed in large part to its use of the merit review process for proposal review, its focus on research universities as the award recipients, the fact that research is a primary output from the Program’s awardees, and its focus on states that have been relatively less successful in attracting NSF research money. All this is consistent with NSF’s mission, and it is likely that any future programs involving partnerships with states would have to incorporate these elements.

In future partnerships with states, should NSF, presumably in collaboration with representatives of states, design programs around ends or means? The S/IUCRC Program was designed around means: industry-university research collaboration was the vehicle for implementing the partnership, and the assumption was that the success of the I/UCRC model made it a suitable guide, with some modification, for a state partnership program. The goal of state economic development came second and seemed entirely appropriate at the time of the NSF/NGA meetings from which the S/IUCRC Program’s features emerged. It proved more difficult than expected to modify the I/UCRC model, intended to increase the national competitiveness of industrial firms, in a way that would achieve regional economic development through industry-university cooperative research. The difficulties were of two types. First, the model itself probably was inappropriate for the goal of regional economic development. Second, wide variations across states in the organizational and political relationships between science and technology-related programs and strategies on the one hand, and economic development programs and strategies on the other, were incompatible with the inflexible requirement for annual state matching linked to an economic development goal.

Two decades of state experience with using science and technology as a lever for economic development has led to mixed results, widely varying strategies across the states, highly uneven levels of political and financial support over time, and no clear lessons—other than the obvious one that state political timetables and criteria do not mesh well with the timetables for payoffs from university research, or with the intangible forms that those payoffs typically take. If there is any single ingredient that seems necessary, if not sufficient, for a successful state program of S&T-leveraged economic development, it is continuing political commitment from successive administrations. The design of the overall strategy seems less important than the level and stability of commitment. Some states appear to have accepted the theory that investment in S&T infrastructures pays off, thus integrating current economic development theory into concrete state programs intended to promote economic development. In other states, economic development plans remain isolated politically and organizationally from S&T plans, and in still other states there is only limited commitment to S&T. Some states have made university-industry cooperation an integral part of their S&T and/or economic development plans (e.g., New York), while others have not. The S/IUCRCs studied faced quite different environments when it came to justification for matching funds. In some cases the Center Director developed the rationale and made the case before the legislature; in others, the organization charged with providing the matching funds in the state (e.g., the New York State Office of Science, Technology, and Academic Research; the Ohio Department of Development) included the request in a larger budget request, shielding it somewhat from close scrutiny.

States such as Ohio and New York, which have made a considerable and stable commitment to the support of university-based research centers whose goal is state economic development, provided a more hospitable atmosphere for S/IUCRCs. The Edison Centers in Ohio and the Centers for Advanced Technology (CAT) in New York offered existing homes for the S/IUCRC budget and a relatively well-accepted rationale for state investment in university-industry collaborative research. Beyond this, state-supported programs of university-based centers such as CAT can provide the greater flexibility and broader range of program elements that are more consistent with the requirements of regional economic development than were the features of the S/IUCRCs. Despite the Program’s relative success in attracting small businesses and state-based firms as members, it could not, alone, provide the business assistance and spin-off assistance so often needed by firms and entrepreneurs approaching the university for help.

The EPSCoR Program did not face the uncertainties and variable political environment across states that the S/IUCRC Program did. With no explicit connection to state economic development via its formal matching requirements, NSF/EPSCoR could work with the state agency in each state most closely associated with supporting research universities. The political and organizational links to economic development either had already been made or were forged as part of larger state plans for developing the state’s university-based research capabilities. Although a strong university research system is a means to larger state goals, it has increasingly been accepted by states as virtually an end in itself, requiring little further justification. Thus there is an advantage to partnering with states with programs that either do not prescribe the means, or that emphasize means that have achieved the status of ends.

Does an "EPSCoR" for university-industry cooperative research make sense for NSF? There is little doubt that a rationale for federal support of state efforts to create or strengthen university-industry cooperation in research exists, nor is there much doubt that such cooperation pays off both regionally and nationally. The argument for an NSF-state partnership that targets industry-university cooperation is as valid now as it was when the NGA and NSF first discussed such a program. The design of such a future program, however, must account for the lack of integration in many states between economic development plans and programs and S&T plans and programs, and for variations in the nature of state agencies responsible for support of universities. Given the experience of the S/IUCRC Program, it would seem prudent to design such a program as one whose goal is to strengthen state S&T infrastructures rather than promote regional economic development, with university-industry cooperative research as the means. The form that such cooperative research programs might take would vary considerably, depending on state priorities and needs, and on existing state commitments to such cooperative arrangements. A winning proposal to New York or Ohio would look quite different from one from, say, Oklahoma, where universities only recently have been encouraged by the state to engage in cooperative research with industry.

SRI International

TABLE OF CONTENTS

Outcomes and Impacts of the State/Industry-University Cooperative Research Centers (S/IUCRC) Program

Outcomes and Impacts of the State/Industry-University
Cooperative Research Centers (S/IUCRC) Program