当前位置:首页 >> 经管营销 >>

Investigating the factors that diminish the barriers to university–industry


Research Policy 39 (2010) 858–868

Contents lists available at ScienceDirect

Research Policy
journal homepage: www.elsevier.com/locate/respol

Investigating the factors that diminish the barriers to university–industry collaboration
Johan Bruneel a,b , Pablo D’Este b , Ammon Salter a,?
Imperial College Business School, Imperial College London, South Kensington Campus, London, United Kingdom, SW7 2AZ Institute of Innovation and Knowledge Management (INGENIO), Spanish Council for Scienti?c Research (CSIC) - Polytechnic Univ.of Valencia (UPV), Ciudad Politécnica de la Innovación, Valencia 46022, Spain
b a

a r t i c l e

i n f o

a b s t r a c t
Although the literature on university–industry links has begun to uncover the reasons for, and types of, collaboration between universities and businesses, it offers relatively little explanation of ways to reduce the barriers in these collaborations. This paper seeks to unpack the nature of the obstacles to collaborations between universities and industry, exploring in?uence of different mechanisms in lowering barriers related to the orientation of universities and to the transactions involved in working with university partners. Drawing on a large-scale survey and public records, this paper explores the effects of collaboration experience, breadth of interaction, and inter-organizational trust on lowering different types of barriers. The analysis shows that prior experience of collaborative research lowers orientationrelated barriers and that greater levels of trust reduce both types of barriers studied. It also indicates that breadth of interaction diminishes the orientation-related, but increases transaction-related barriers. The paper explores the implications of these ?ndings for policies aimed at facilitating university–industry collaboration. ? 2010 Elsevier B.V. All rights reserved.

Article history: Received 11 March 2009 Received in revised form 3 November 2009 Accepted 26 March 2010 Available online 11 May 2010 Keywords: Universities University–industry collaboration Barriers to collaboration Inter-organizational trust

1. Introduction Collaboration between industry and universities faces signi?cant challenges. While universities are primarily driven to create new knowledge and to educate, private ?rms are focused on capturing valuable knowledge that can be leveraged for competitive advantage (Dasgupta and David, 1994). In addition, universities are becoming increasingly proactive managers of their collaborations with industry, seeking to create valuable Intellectual Property (IP) to foster technology transfer. Accordingly, more and more interactions between university and industry are becoming subject to measurement and management, leading to more formal, contractual exchanges based on codi?ed rules and regulations. Although both these aspects have been acknowledged in the literature on university–industry (U–I) linkages, relatively few studies have investigated the nature of the barriers and the factors that might mitigate them (see also Hall et al., 2001). Given the central importance given by policy to building and supporting U–I links,

? Corresponding author. E-mail addresses: j.bruneel@imperial.ac.uk, johan.bruneel@ugent.be (J. Bruneel), a.salter@imperial.ac.uk (A. Salter). 0048-7333/$ – see front matter ? 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.respol.2010.03.006

the lack of research on the obstacles to collaboration is a serious hindrance to the design of effective policy. In order to advance knowledge in this area, this paper examines two types of barriers: (i) those related to differences in the orientations of industry and universities, what we describe as ‘orientation-related barriers’; and (ii) barriers related to con?icts over IP, and dealing with university administration, what we describe as ‘transaction-related barriers’. This paper explores the mechanisms that can lower the degree to which ?rms encounter these types of barriers through an examination of three important elements that in?uence the ?rm’s perception of these two sets of obstacles to collaboration. First, we explore the impact of the ?rm’s prior experience of working on research projects with universities on the assumption that experience can ease both types of barriers to collaboration. Second, we examine whether the nature of the interaction between the ?rm and its university partner plays a role in the perception of barriers. In this case, we expect that ?rms that articulate their collaborations through multiple channels will perceive barriers as less constraining. We also investigate whether the nature of the interaction – here we contrast education-based with contract-based forms of interaction – that ?rms engage in with university partners positively (or negatively) in?uences the perception of different types of barriers. Finally, we assess how the level of trust in its university partners shapes the ?rm’s perception of the barriers to working with universities (Nooteboom, 2002; McEvily et

J. Bruneel et al. / Research Policy 39 (2010) 858–868

859

al., 2003). Our approach provides a window on some of the mechanisms that may limit the depth and quality of interactions between universities and businesses. The analysis is based on the statistical analysis of a large survey of UK ?rms that have collaborated on publicly funded research projects, combined with data from records of prior involvement in research collaboration with universities. The analysis shows that prior experience of collaborative research lowers orientationrelated barriers and that greater levels of trust reduce both types of barriers studied. We also ?nd that breadth of interaction diminishes the orientation-related, but increases transaction-related barriers. We explore the implications of these ?ndings for research and policy.

2. Barriers associated with U–I collaboration 2.1. Incentives and con?icts between public and private knowledge At the core of the obstacles to U–I collaborations are the different institutional norms governing public and private knowledge (Dasgupta and David, 1994). The university system is rooted in Mertonian norms of science, such as communalism, universalism, disinterestedness and organized scepticism (Merton, 1973). The creation of reliable and public knowledge has been central to the growth of these organizations, leading to support from government for research to expand the pool of economically useful knowledge (Geuna et al., 2003). These institutional norms are fundamental to the way that many academics perceive and perform their work. Indeed, scientists are willing to accept lower wages in order to work within the institutions of science, indicating that many scientists are motivated by intrinsic goals as well as the social objectives of the universities (Stern, 2004; Cohen and Sauermann, 2007). The institutions of science include strong competitive mechanisms and powerful incentive regimes. The priority of establishing reputation through publication is critical to academic success and/or career sustainability. Academics often have to engage in ‘status competitions’ with their peers, based on publication records, institutional af?liations and prizes (Becher, 1989). Many of these competitions take the form of winner-takes-all, in which publishing ?rst or winning the largest research grants precludes others from these same achievements or resources. Given this environment, much of the science system is driven by internal dynamics that are separate from market transactions (Polanyi, 1962; Dasgupta and David, 1994; Stephan, 1996). Peer esteem cannot be bought and must be created by winning favour and reputation among colleagues. Although it might be tempting to see the science system as operating outside the con?nes of market transactions, it is also true that economic and social forces outside the science system itself play a powerful role in shaping scientists and science (Freeman, 1999). Much of the research supported by government is applied, or practically oriented, and focused on solving general social, technical or economic problems using the capabilities of science (Pavitt, 2001). Scientists often hold con?icting and evolving views on the bene?ts of working with industry (Welsh et al., 2008). Moreover, many ?elds of research, such as engineering, by their nature, involve considerable interaction with industrial practice (Rosenberg and Nelson, 1994). In addition, the role of the university as an educator of professionals – doctors, engineers, accountants, lawyers, etc. – means that a large proportion of their staff are focused on ?elds of research that engage with practical problems. For the researchers working in such areas, practical problems provide a powerful stimulus to the development of new ideas (Rosenberg, 2002). Although within these practical-oriented areas of research the norms of science still operate, they do so somewhat differently

from the Mertonian ideal of science. Researchers in these areas are more likely to be engaged on real world problems and interacting with industry, and their status is likely to be co-determined by their reputation among their peers and their standing in industry. This is especially true in the case of engineering (Vincenti, 1990). In contrast to the relatively open nature of the science system, the process of knowledge creation in the private sector is dominated by attempts to appropriate the economic value of what ?rms know in order to gain competitive advantage (Teece, 1986). This ‘private’ knowledge is largely closed, remaining hidden within the ?rm or disclosed in a limited way through patents ?led primarily for the purposes of obtaining temporary monopolies (Allen, 1984; Dasgupta and David, 1994). This is not to say that industry knowledge is completely closed: many forms of knowledge exchange and leakage occur between ?rms working in the same sector. A considerable number of ?rms publish academic and technical papers to signal their competencies or to defend against others’ attempts to control particular areas of technology (Hicks, 1995; Cockburn and Henderson, 1998). They may also participate in open source software projects to help lower the costs of their own development activities (von Hippel and von Krogh, 2003), and there is some evidence that ?rms engage in strategic trading of information with competitors (von Hippel, 1987). Despite these examples of openness, the primary motivation of ?rms’ knowledge creation activities is the appropriation of knowledge for private gain, and openness to external actors is used as a strategic mechanism to gain advantage over competitors (Chesbrough, 2006). Given these two different systems of knowledge production, U–I collaborations are likely to be plagued with con?icts due to a weak attitudinal alignment between partners. Private ?rms often con?ict with university researchers over attitudes towards the topics of research or the timing and form of disclosure of research results. While researchers may be keen to disclose information to gain priority, ?rms may wish to keep secret or appropriate the information. To paraphrase Brown and Duguid (2000), academics wish to create ‘leaky’ knowledge so that their ideas will be acknowledged by their peers while ?rms want the knowledge to be ‘sticky’ so that they can control a resource that is not available to their competitors. University researchers are also likely to choose research topics that are perceived by their peers to be interesting and valuable, while ?rms are likely to choose topics and problems that are perceived as being valuable for the development of new products and services for their customers (Nelson, 2004). This means that the problems that each party may want to explore within a research project may be very different and the types of outputs each partner is interested in may also diverge. 2.2. Con?icts over IP and university administration The growth over the past 30 years of universities as economic actors in their own right, has also been important in shaping the nature of the interaction between universities and ?rms. The rise of the university Technology Transfer Of?ce (TTO) and the increasing attempts by universities to capture formal IP have had a profound impact on the nature of scienti?c efforts (Shane, 2004). These efforts have led to an expansion in university patenting and the creation of a new commercial focus on the part of the universities to create valuable IP and exploit it for ?nancial gain (Henderson et al., 1998; Mowery and Ziedonis, 2002). Support designed to encourage academics to engage with industrial partners can take many different forms and often varies across universities and countries. In the UK, for example, the government has launched a range of initiatives to encourage universities to capture and exploit their IP (Lambert, 2003; Chapple et al., 2005). For some, this focus on commercialization undermines the public commons of science, weakening the institutions of open

860

J. Bruneel et al. / Research Policy 39 (2010) 858–868

science through the imposition of private norms on public activities (Nelson, 2004). For others, the rise of the university as an economic actor creates a new motor for economic development that in the past has been rare?ed and separate (Etzkowitz and Leydesdorff, 2000). Some scholars have attempted to measure the effect of engagement with industry on academic behaviour by examining the impact of patenting on individual researcher’s publication activity (Agrawal and Henderson, 2002; Azoulay et al., 2007; Calderini et al., 2007; Fabrizio and DiMinin, 2008). These studies suggest that there are complementarities between patenting and scienti?c performance, and that those individuals that do the best research are also successful at engaging in real world problems and creating commercial value (Rothaermel et al., 2007). Although these ?ndings are not de?nitive and are liable to differ across research ?elds and across countries, they do suggest that academic engagement in commercial activities or with industry can have complementarities with research performance. Apart from these results at the individual level, we know little about the effects of increasing commercial-orientation in universities on general patterns of U–I collaboration. Evidence from the US since the Bayh-Dole Act suggests that although the level of university patenting has increased since the early 1980s, the quality of these patents has declined over time (Mowery et al., 2001). Moreover, increased university patenting activity began before the Bayh-Dole Act came into force and is highly localized in a few technological ?elds. The pattern is similar in Europe (Geuna and Nesta, 2006). It is unclear whether the changes that have occurred in university patenting activity are a direct consequence of technological changes or of policy. Moreover, we do not know what effect these efforts at commercialization have had on the nature, frequency and types of U–I collaboration. Early research in this area suggests that the increase in university patenting has been accompanied by a slowdown in joint research collaborations (Valentin and Jensen, 2007) and in the pace of private knowledge exploitation across a number of technological areas (Fabrizio, 2007). It is also clear that in some cases, attempts by universities to capture the commercial bene?ts from research have led to signi?cant distributional con?icts between universities and their industrial partners (Florida, 1999; Shane and Somaya, 2007). These distributional con?icts are often accentuated by the unrealistic expectations held by universities about the commercial potential of university research (Clarysse et al., 2007), which can result in their overvaluing IP. These types of con?icts with TTOs and university administration may put a signi?cant strain on collaborations, eventually deterring ?rms from collaborating with universities.

3.1. Experience of collaboration Working with universities on research projects requires not only that ?rms learn to work across organizational boundaries, but also that they have or can build the capabilities to collaborate with partners operating within a different incentive system. Collaborating with a university partner necessitates that ?rms develop operating routines and practices to manage this collaboration. Establishing expectations about when and in what form the results from a joint research project will be published may be controversial, for example. However, once routines and practices have been established, they are likely to be re?ned and reused in subsequent collaborations. For example, problems that emerge in one project can be overcome by careful planning in subsequent projects. Thus, university collaboration is an activity in which ?rms learn from experience and develop richer and more re?ned ways of engaging with the university sector. Indeed, overtime, the experience of collaboration should enable academics and their industrial collaborators to converge in attitudes, learning to share common norms and arrive at a mutual understanding about the nature of the collaboration and the research process. Not all ?rms are interested in making investment in relationships with academics; they tend to fall into the categories of infrequent, intermittent or recurrent partners with universities (Hall et al., 2003; Hertzfeld et al., 2006; Bishop et al., 2009). Frequent and recurrent partners, however, are particularly likely to capitalize on their collaboration experience by transferring the information and knowledge gained through their involvement in multiple and diverse partnerships. Recurrent collaborators are more likely to put in place the necessary routines to reconcile con?icting views on research targets (Gomes et al., 2005), dissemination of results (Hall et al., 2003), and timing of deliverables (Van Dierdonck and Debackere, 1988). These efforts should help to lower the barriers related to research orientation by fostering attitudinal convergence between partners. Collaboration experience could also help to lower transactionrelated barriers to collaboration. Research on inter-organizational alliances shows that collaboration experience is a critical determinant of the success or failure of subsequent alliances (Hagedoorn and Schakenraad, 1994). Although transaction-related barriers involve distribution con?icts that are hard to resolve in any project, past research has found that collaborative experience can help mitigate these con?icts. As Hertzfeld et al. (2006) demonstrate, experienced collaborators use standard protocols as starting points for negotiations on IP ownership, facilitating the setting up of new collaborative agreements. Creating acceptable rules for allocation of patent rights is instrumental to securing the good-will of partners for an effective research collaboration ((Hertzfeld et al., 2006; Jelinek and Markham, 2007). In addition, ?rms that have worked on many projects may have greater experience in negotiating IP contracts with university partners’ TTOs and legal of?ces. In doing so, experience may help to alleviate distribution con?icts through the development and use of standard IP contracts, setting out the allocation of rewards from joint research activities between universities and the ?rm. Moreover, ?rms with experience working on research projects may also be more aware of differences in IP regimes across different universities, which may put them in a favourable position to negotiate with university managers compared to less experienced companies. Given this, it can be expected that experience of collaboration should help to lower transactionrelated barriers. 3.2. Breadth of interaction channels Research shows that ?rms draw bene?ts from universities via rich and varied ways (Gibbons and Johnston, 1974; Cohen et al.,

3. Factors that mitigate the barriers to collaboration Although we know a considerable amount about the factors that lead some ?rms to collaborate or draw knowledge from universities (Meyer-Krahmer and Schmoch, 1998; Tether, 2002; Arundel and Geuna, 2004; Laursen and Salter, 2004), we know little about how the barriers perceived by industry to working with universities may be mitigated. Our current understanding tends to rely on information from non-collaborators, which does not provide insights into how those ?rms that do collaborate with universities overcome these barriers (Mohnen and Hoareau, 2003; Fontana et al., 2006). In this paper, we focus on three potential mechanisms to reduce the obstacles to U–I collaboration: experience of collaboration, breadth of interaction channels, and inter-organizational trust. This section provides a detailed discussion of these mechanisms in helping to lower the barriers to collaboration perceived by ?rms.

J. Bruneel et al. / Research Policy 39 (2010) 858–868

861

2002), from joint research collaborations to consultancy work, and informal interactions in meetings and at conferences. While some links require high levels of co-ordination and sustained interaction, offering rich opportunities for knowledge exchange, others are more arm’s-length and rely on impersonal forms of exchange, such as publications. Therefore, research that accounts for only one type of linkage may miss many channels relevant to ?rms’ drawing knowledge from universities and, thus, may provide only a partial understanding of the overall patterns of interaction. Involvement in a variety of channels of collaboration may contribute to better equip the ?rm to manage con?icts over the orientation of research, by strengthening the ?rm’s capacity to balance and align different incentive systems across a diverse set of inter-organizational arrangements (Lawrence et al., 2002; D’Este and Patel, 2007). Moreover, engaging in a broad range of interaction channels creates opportunities for organizational learning by exposing the ?rm to formalized and non-formalized interactions; face-to-face and arm’s-length interactions; and short/targeted and long-term/open-ended interactions. There are substantial synergies between these channels: while casual face-to-face and short-term interactions may not require a formalized-contractual relationship, they are crucial to improving the effectiveness of formal, long-term research agreements (Kogut, 2000). Therefore, engagement in a wider range of interaction channels with universities may enable the convergence of attitudes between the two parties in the exchange, helping to overcome misalignments due to distinct institutional norms. However, interactions with universities facilitated by many different channels may lead to increased transactional con?icts. As many more parts of the university are engaged in the relationship, there is the potential for the ?rm and its collaborations to be subject to increased levels of engagement with the university administration and its many rules and procedures. For example, informal relationships with academic faculty on education-related matters, such as student secondments, are likely to involve very different parts of the university administration than interactions with university research services departments for joint research projects. In addition, transaction-related barriers often arise as a result of concerns in the collaboration about the distribution of bene?ts from outcomes of collaborative projects, and therefore engaging different parts of the university may open the ?rm up to costly negotiations about follow-on rewards from the project with a wide range of different university actors, many of whom may have different incentives and expectations about the relationship. Approaches that have been used in one part of the university to deal with distributional con?icts may not be operative in another part of the university where norms of behaviour differ. Therefore, interacting across different channels can entangle ?rms in messy and labour intensive interactions with the university, which ?rms used to operating along a narrow line of interaction will ?nd problematic. In sum, it can be expected that working across different channels may raise transaction-related barriers, while at the same time the increased breadth of interaction will lower orientation-related barriers. 3.3. Inter-organizational trust U–I research collaboration involves high levels of uncertainty because the research process is beset with many unknowns. Given this, it is almost impossible to specify in advance the followon implications for the disclosure and commercialization of the research. Under such conditions, collaboration partners may seek to take advantage and act opportunistically to appropriate the bene?ts of the collaboration (Williamson, 1993). High levels of trust, on the other hand, can help to reduce the fears that one of the partners

will act opportunistically (Bradbach and Eccles, 1989; Dodgson, 1993). Trust allows partners to be con?dent that their collaborator will treat them fairly and in a consistent way, and will help to resolve any problems that may arise jointly (Rempel and Holmes, 1986; Zaheer et al., 1998). Trust is likely to be especially important in facilitating university–industry links (Santoro and Saparito, 2003), since ?rms and universities are often required to share commercially sensitive information and tacit knowledge. If a collaboration is characterized by low levels of trust, partners are less likely to be forthcoming about the knowledge and information required to make the collaboration successful (Inkpen and Tsang, 2005). Thus, higher trust between partners stimulates rich social and information exchanges and encourages partners to exchange more and valuable knowledge and information (Ring and Van de ven, 1992). Moreover, trust-based relationships facilitate the exchange of dif?cult to codify knowledge and information, which is by de?nition dif?cult to communicate and to trade in markets (Kogut and Zander, 1992). Trust expresses the capacity of ?rms and universities to work together to resolve problems, and demonstrates a willingness to understand and adjust behaviours to align with the needs and expectations of partners (Santoro and Gopalakrishnan, 2001; McEvily et al., 2003). For these reasons, it can be expected that high levels of trust will be associated with lower orientation-related and transaction-related barriers.

4. Data, measures and method 4.1. Data Since our study is designed to capture industry attitudes associated to collaboration with universities, we conducted a survey of ?rms that have been actively engaged in these types of partnerships. To construct the sampling frame for our study, we drew on the records of research projects funded by the Engineering and Physical Sciences Research Council (ESPRC). The EPSRC is the UK’s largest funding council in terms of funding research, and it has a broad remit including engineering disciplines, computer science, mathematics, chemistry and physics. Thus, our study includes a broad range of scienti?c areas and neutralizes the strong lifesciences bias that pervades much of the literature on U–I links. In order to ensure complete records, we surveyed all the private, for-pro?t organizations with formal involvement in EPSRC collaborative projects between 1999 and 2006. After cleaning the records for duplicates, we obtained a sample of 3088 different organizations. The survey was addressed to the lead person named as an industrial collaborator on the EPSRC grant. In the case of companies that participated in multiple ESPRC projects, our approach was to focus on the contact person most frequently named by the ?rm, as this individual is likely to be the key point of contact between the ?rm and its university partners. However, to ensure that our individual level responses were representative of views of their wider organization, we included a top up sample of 343 individuals that were listed as the second most frequent listed as a contact name on the collaborations. This approach left us with a ?nal sample of 3431 individuals. The sampling method in our study does not allow us to explain why ?rms collaborate with universities, which is a topic that has been extensively covered in prior research. Instead, we offer evidence on ?rms’ perceptions of the barriers to collaboration among those that have engaged in cooperation with universities. This allows us to provide information based on actual experience of collaboration rather than the perceptions of non-collaborating ?rms, which inevitably re?ect general attitudes to universities rather than real experience of university collaboration.

862

J. Bruneel et al. / Research Policy 39 (2010) 858–868

The unit of analysis for our survey is the business unit which we de?ned as “an organizational unit producing goods or services which bene?ts from a degree of autonomy in decision making, especially for the allocation of its resources”. This de?nition is consistent with the UK innovation surveys (Stockdale, 2002). We decided to focus on the business unit because some of the ?rms in our sample are large, multi-site organizations. As U–I collaboration is often local in character, collaboration between business units and universities is likely to be decided locally rather than centrally (Cantwell, 1989; Criscuolo, 2005). Our analysis also distinguishes between subsidiaries and independent organizations. In past studies, scholars have tended to focus on projects rather than business units when assessing the nature of the barriers to collaboration (Hall et al., 2001). Our choice to focus on the business unit is also because many of the organizations in our sample are involved in more than one project. For example, more than 10 organizations in our sample have had involvement in more than 40 different EPSRC research projects. Although focusing on project-level interactions can uncover important issues arising from exchanges within a project, the limitation is that it captures information on only one among what may be a portfolio of projects. The barriers that emerge in one project may have been resolved in another project, even with the same university and industry partners. Also, focusing on a single research project can obscure the different relationships likely to be in place between the ?rm and its university partners, such as student training or contract research, and how these broad forms of interaction shape the nature of the barriers faced by ?rms in their engagement with universities in general. By capturing information at the level of the business unit, we go beyond the con?nes of a single project and explore the more general pattern of relations between the ?rm and its university partners. To develop the survey questionnaire, we conducted a number of interviews with industry and with academics as well as reviewing the literature. The survey asked about the barriers to interacting with universities and the frequency of interaction by types of engagement (Cohen et al., 2002). Responses were based on an extensive list of barriers and types of engagement, described below. All factual questions, such as number of times the ?rm engaged in different channels of interaction, referred to the period 2005–2006, whereas attitudinal questions were usually left open-ended. The survey also included a series of questions on respondents’ organizations: size, sector, R&D expenditure, share of staff with higher education degrees, and ownership. Data collection was done in several stages. First, in November 2007, we wrote to the individuals in our sample with an invitation to the individual to go to a website to complete an electronic version of the survey. The invitation included a letter from Professor David Delpy, Chief Executive of the EPSRC, endorsing the study. This ?rst stage elicited 276 responses. To improve the response rate, we telephoned non-respondents to encourage them to respond. This yielded another 176 responses. In the second stage and again to improve the response rate, we conducted another postal survey in February 2008, this time including a paper copy of the questionnaire in order that respondents had the choice of an electronic or paper-based version. This second stage yielded another 188 responses. In the third stage, we used the email addresses collected from the telephone contacts with organizations to send email reminders to non-respondents for whom we had email addresses. This yielded another 13 responses giving us a total of 646 usable responses, representing 600 organizations (four responses were from individuals in the same ?rm but from another business unit). Based on a total survey population of 3088, the response rate was 19%. After removing questionnaires with incomplete information, or from individuals in the same organization (we took the aver-

age scores for these),1 we were left with a ?nal sample of 503 organizations. The sample covers a diverse range of ?rms, with representation from organizations of different sizes, across all sectors, including professional services. To check the reliability of our sample, we conducted a number of tests on the respondent population. First, we compared the distribution of respondents and non-respondents by sector and found that the response pattern closely mirrored our sample population.2 Second, we compared early and late respondents: we found no signi?cant differences in terms of structural features, such as size, sector or R&D intensity or attitudes to collaboration (Armstrong and Overton, 1977). Third, we compared electronic and paperbased responses to the survey and found no signi?cant differences. Finally, we compared the two different responses for the 44 organizations where we have multiple respondents and again found no signi?cant differences. These tests increase con?dence that the survey data are reliable. 4.2. Measures and method 4.2.1. Dependent variables As discussed above, barriers to U–I collaboration are based on: (1) differences in incentives and orientation (orientation-related barriers); and (2) con?icts over IP and university administration procedures (transaction-related barriers). To capture the extent to which ?rms indicated that they faced orientation- and transactionrelated barriers in working with universities, we drew on the responses to a question about the general barriers to interaction with universities. Respondents were asked to indicate their level of agreement with 12 statements concerning some likely barriers to interaction with universities. The items were developed based on interviews with industry organizations and the literature on barriers to knowledge exchange between universities and industry. To construct our measure of orientation-related barriers, we focused on the three items directly related to the orientation of university research and researchers. These are: ‘university research is extremely orientated towards pure science’; ‘long-term orientation of university research (concerns over lower sense of urgency of university researchers compared to industry researchers)’; and ‘mutual lack of understanding about expectations and working practices’. Each item is measured on a ?ve point likert scale from ‘strongly disagree’ to ‘do not agree at all’ and is coded 1 if respondents indicate that they ‘agree’ or ‘strongly agree’ with the statement, and 0 otherwise. To calculate the variable orientationrelated barriers, we added these scores so that each organization scored 0 for no barriers and a score of 3 when all orientation-related barriers are perceived as high.3 To construct our measure of transaction-related barriers, we focused on the following four items: ‘industrial liaison of?ces tend to oversell research or have unrealistic expectations’; ‘potential con?icts with university regarding royalty payments from patents or other intellectual property rights and concerns about con?dentiality’; ‘rules and regulations imposed by universities or

1 There are 42 business units with two respondents and two business units with three respondents. Besides the average scores of the multiple respondents, we also ran the models with the scores of each of the respondents. The results from these analyses are consistent with those reported below. 2 Response rates across sectors are not signi?cantly different for the largest populated sectors in our sample frame, ranging from 19% (e.g. machinery and metals) to 25% (e.g. motor equipment and aircraft manufacture; and business services). However, response rates are lower (around 10%) for sectors where the number of ?rms in the sample frame was low (e.g. wholesale and retail trade; ?nancial intermediation; manufacture of food products, among others). 3 As alternative measures, we used average and aggregate scores for both dependent variables. The results from this analysis are consistent with those reported below.

J. Bruneel et al. / Research Policy 39 (2010) 858–868

863

government funding agencies’; and ‘absence or low pro?le of industrial liaison of?ces in the university’ (which was reverse coded). Our transaction-related barriers measure was created using the method as described above for orientation-related barriers. 4.2.2. Explanatory variables We measured collaboration experience as the total length (in months and in log scale) of research experience of working on collaborative projects with universities, funded by the EPSRC, that the organization had accumulated in the period 1991–2004.4 This variable offers a composite measure of experience, taking into account that ?rms may be involved in more than one project and that these projects may span different universities and different periods. The approach offers a more ?ne-grained operationalization than just taking numbers of years the organization has collaborated with universities or total number of its projects. In order to capture the breadth of interaction between businesses and universities, we created a variable measuring the extent to which organizations use different types of interactions with universities during the period 2005–2006. This information was taken from the survey question on the channels of interaction between ?rm and universities. Building on D’Este and Patel (2007), we examined the broad range of channels through which organizations can interact with universities. We focus on ‘joint research projects’, ‘contract research’, ‘consultancy’, ‘training of ?rm employees’; ‘postgraduate training in the company’; ‘recruitment of recent graduates or postgraduates’; and ‘student placements’. To construct the variable, we used a binary code for each channel of interaction, which takes the value of 1 if the ?rm reports having used a given interaction channel, and 0 otherwise. We then simply added up the seven interaction channels to represent the breadth of interaction. Since the nature of U–I interactions may have a distinctive effect on the barriers, we also considered two other measures for breadth of interactions: a variable that captures more informal interactions related to the educational role of universities, including the items ‘training of ?rm employees’, ‘postgraduate training in the company’, ‘recruitment of recent graduates or postgraduates’ and ‘student placements’. We term this variable education-based interactions. We created another variable to capture more formal interactions between industry and universities through contractual relationships, including the items ‘joint research projects’, ‘contract research’, and ‘consultancy’. We call this variable contract-based interactions. Building on Rempel and Holmes (1986) original trust scale and its elaboration in Zaheer et al.’s (1998) for inter-organizational trust, we measured the level of trust through four statements measured on a ?ve-point likert scale. The items include: ‘our university partners may use opportunities that arise to pro?t at our expense’ (reverse coded); ‘based on past experience, we cannot have complete con?dence in our university partners to keep promises made to us’ (reverse coded); ‘we trust our university partners to treat us fairly’; and ‘we trust that con?dential/proprietary information shared with our university partners will be kept strictly con?dential’. As might be expected for such a well-established scale, the Cronbach Alpha was high at .83. 4.2.3. Control variables We also included several other variables that may have an in?uence on the level of barriers that the ?rms face when interacting with universities. First, we control for organization’s level of absorptive capacity. There are several proxies in the literature

used to measure absorptive capacity (Cohen and Levinthal, 1990; Lane et al., 2006). Drawing on Rothwell and Dodgson (1991) and Schmidt (2005), we measure level of absorptive capacity as the percentage of staff with a higher education degree. We chose not to adopt one of the most common operationalizations of absorptive capacity – R&D intensity – because of the high number of service ?rms (over 40%) in our sample. R&D intensity may underestimate the absorptive capacity of service ?rms because these ?rms typically have modest R&D budgets or do not conduct formal R&D, but they may have high levels of absorptive capacity. The variable is categorical and ranges from 1 to 5 according to the percentage of staff with a higher education degree: 1, if the percentage equals or is less than 10%; 2, if the percentage is between 11% and 20%; 3, if the percentage is between 21% and 40%; 4, if the percentage is between 41% and 60%; and 5, if the percentage is between 61% and 100%. We used a categorical approach because we wanted to reduce the burden for respondents.5 Second, we include a measure for ?rm size, i.e. the logarithm of the number of employees, expressed in full-time equivalents, as a control variable (size); large ?rms are likely to have more resources to work with external organizations such as universities (Tether, 2002; Mohnen and Hoareau, 2003). Third, we control for the nature of the ?rm’s organizational structure. This dummy variable identi?es ?rms that are independent rather than being part of a large group (independent). We would expect organizations that belong to a group to have more resources for and more experience of working with universities and, therefore, to face lower barriers to collaborations. Fourth, since we rely on individuals to report information for the organization, it is important to account for differences among these individuals in terms of their educational backgrounds. Individuals with doctoral degrees are likely to be more familiar with university norms compared to individuals with only undergraduate degrees. We also wanted to ensure that our results were not biased by the fact that surveys were addressed to named research collaborators, which might have increased the tendency for respondents to be less sensitive about barriers to universities than individuals not directly involved in the research. In order to control for this, we include a dummy variable that equals 1 if the respondent has a doctoral degree and 0 otherwise (doctoral). Finally, we also include eight dummy variables to account for inter-industry differences in patterns of U–I interaction. 4.2.4. Method of estimation Our dependent variables (i.e. orientation-related barriers and transaction-related barriers) take on non-negative integer values and therefore are count variables. A commonly used method of estimation is the Poisson regression model (or negative binomial regression in the case of overdispersion). However, since our dependent variables are restricted by an upper bound (i.e. the maximum number of barriers is three and four, respectively), Poisson or Negative Binomial distributions are not strictly applicable. An alternative approach is an ordered logistic model; however, this would imply a natural ordering in the level of barriers, which may not apply. Instead, we build on a technique provided by Wooldridge (2002:: 661), who suggests that a dependent variable may be “obtained by dividing a count variable by an upper bound”, and that such an approach allows the application of fractional logit regression (Papke and Wooldridge, 1996). This approach, models E(y | x) as a logistic function, where y is the dependent variable and x is a set of regressors: E(y | x) = exp(xˇ)/[1 + exp(xˇ)]. This model ensures

We took account of records up to 2004 in order to avoid an overlap with the time frame of the questionnaire, which asks ?rms to report information for the period 2005–2006.

4

5 We also conducted an additional analysis where we substituted R&D intensity for percentage of staff with higher education degrees. Again, the results were consistent, suggesting our main results did not change according to the control used for absorptive capacity.

864

J. Bruneel et al. / Research Policy 39 (2010) 858–868

Table 1 Type of barriers to university interaction for SMEs and large ?rms (% of ?rms that indicated that they agree or strongly agree with the item in the questionnaire). SME % 31 69 34 50 57 58 27 Large ?rms % 36 59 34 49 54 53 24

Type Orientation-related barriers

Barrier University research is extremely orientated towards pure science Long-term orientation of university research (concerns over lower sense of urgency of university researchers compared to industry researchers), Mutual lack of understanding about expectations and working practices Industrial liaison of?ces tend to oversell research or have unrealistic expectations, Potential con?icts with university regarding royalty payments from patents or other intellectual property rights and concerns about con?dentiality, Rules and regulations imposed by universities or government funding agencies, Absence or low pro?le of industrial liaison of?ces in the university (reverse coded)

Transaction-related barriers

Table 2 Number of orientation-related barriers to university interaction, by sector. Number of ?rms Industrial sector Chemical Chemical-related Machinery and metal Electronics and instruments Transport Utilities and construction Business services Other 37 23 53 70 15 46 182 77 Number of orientation-related barriers (% of ?rms) 0 11 26 8 27 27 22 27 29 1 41 43 36 29 7 26 32 26 2 41 30 43 31 47 33 31 31 3 8 0 13 13 20 20 9 14

that the predicted values of y are in (0, 1) and that the effect of any xj on E(y | x) diminishes as xˇ → ∞. The method is non-linear and can be estimated using quasi-maximum likelihood, and partial effects ? may be “evaluated at the ˇj and interesting values of x” Wooldridge (2002: 661). 5. Results To understand the nature of university barriers to collaboration, we ?rst explore the different obstacles that organizations face. Table 1 lists the seven barriers to university interaction and the percentage of small and medium sized enterprises (SMEs) and large ?rms that indicated agreement with the different statements on these barriers to interaction with universities.6 Overall, we ?nd that differences between SMEs and large companies are relatively small, with both types of ?rms indicating that orientation-related barriers are lower than transaction-related one. As might be expected, SMEs perceive the barriers related to the long-term orientation of universities and to the rules and regulations imposed by universities or government funding agencies as more important than larger ?rms. Tables 2 and 3 present the variation between sectors on the importance of orientation-related and transaction-related barriers, respectively. The picture from these two tables is quite mixed. Table 2 shows that orientation-related barriers are highest in the machinery and metals, transport (which includes transport equipment and aircraft manufacture), and utilities and construction industries, while ?rms in chemical-related industries have the lowest proportion of ?rms reporting a high number of orientation-related barriers. This might be expected given that the chemical-related industries include many science-based ?rms, for example pharmaceuticals ?rms. However, Table 3 shows that ?rms operating in the chemical and transport industries, both relatively

high-tech sectors, report the highest number of transaction-related barriers, while those in utilities and construction have the lowest proportion of ?rms reporting a high number of transaction-related barriers. Thus, there is no clear-cut divide between high-tech and low-tech sectors with regards to the perception on how important barriers are. Table 4 reports the descriptive statistics and correlation matrix of the independent variables in the model. Overall, the level of correlation between the main variables is low, suggesting that multicollinearity is not a concern. In the ?rst stage of analysis (Models 1a and 1b – Table 5), we enter only the control variables. It can be seen that absorptive capacity (percentage of higher educated staff) is negatively associated with orientation-related barriers. Further, larger ?rms perceive higher transaction-related barriers, and individuals with doctoral degrees are more inclined to perceive higher transactionrelated barriers to interactions with universities. In Models 2a and 2b, we introduce our key explanatory variables. First, we suggested that organizations with more collaboration experience will perceive fewer orientation-related barriers than their less experienced counterparts. The coef?cient of prior collaboration experience is negatively associated with barriers related to differences in orientation (?.05; p ≤ .01), but not to barriers related to transactions. To complement these results we undertook several more analyses exploring different measures of experience, such as number of projects, number of university partners and research project size. The results (which are not shown for reasons of space) are additional with those reported above. These results suggests that routines learnt through conducting joint research with universities, lower the barriers related to the basic and long-term nature of university research, helping to overcome attitudinal differences between the partners on research methods and targets. However, experience of working with universities does not lower the perceived barriers related to university administrative procedures and con?icts over IP. Such con?icts may involve disputes on the distribution of rewards arising from a research project and, if appropriate IP-arrangements are not put in place, these con?icts may persist even in the face of repeated and multiple collaborations. Therefore,

Table 3 Number of transaction-related barriers to university interaction, by sector. Number of ?rms Industrial sector Chemical Chemical-related Machinery and metal Electronics and instruments Transport Utilities and construction Business services Other 37 23 53 70 15 46 182 77 Number of transaction-related barriers (% of ?rms) 0 5 9 17 10 0 24 12 17 1 19 26 21 19 47 43 25 29 2 30 35 32 34 13 13 36 29 3 30 30 25 29 27 20 22 17 4 16 0 6 9 13 0 5 9

6 Following the EU de?nition, SMEs are organizations that are autonomous, that employ less than 250 people, and whose annual turnover does not exceed D 50 million or whose annual balance sheet total does not exceed D 43 million.

J. Bruneel et al. / Research Policy 39 (2010) 858–868 Table 4 Descriptive statistics and correlation matrix of independent variables. Variable 1. Orientation-related barriers 2. Transaction-related barriers 3. Collab. experiencea 4. Breadth of interaction 5. Education-based interaction 6. Contract-based interaction 7. Inter-organizational trust 8. Absorptive capacity 9. Size 10. Independent 11. Doctoral Mean 1.33 1.84 157.6 3.64 2.13 1.51 3.65 3.46 4.46 .53 .48 S.D. .96 1.12 491.6 2.04 1.40 1.03 .63 1.54 2.33 .50 .50 Min 0 0 0 0 0 0 1.25 1 0 0 0 Max 3 4 8869.8 7 4 3 5 5 11.51 1 1 1 .21 ?.06 ?.03 .05 .00 ?.31 ?.10 ?.01 .00 ?.00 2 3 4 5 6 7 8 9

865

10

.01 .18 .13 .17 ?.21 .07 ?.03 .02 .20

.15 .15 .09 .03 ?.09 .25 ?.18 .03

.89 .78 ?.04 .06 .52 ?.27 .25

.40 ?.02 .03 .49 ?.26 .19

?.06 .08 .37 ?.19 .23

?.03 .09 ?.15 ?.01

?.28 .10 .16

.44 .09

.07

Note: Coef?cients with an absolute value above .09 are signi?cant at the .05 level, two-tailed. Industry dummies are not reported. a For prior collaboration experience, natural logarithm is used in correlations and analysis but actual values are reported in descriptive statistics.

experience plays only a partial role in mitigating the barriers to U–I collaboration. Next, we suggested that breadth of interaction is likely to be associated with lower orientation-related barriers, but with higher transaction-related barriers. The results con?rm this swing in the relationship between breadth of interaction and orientationrelated and transaction-related barriers: while the coef?cient is signi?cant and negative in the case of orientation-related barriers (?.06; p ≤ .01), it is signi?cant and positive for transaction-related barriers (.12; p ≤ .001). These ?ndings cast light on how broader U–I ties can have both a positive and a negative effect on the barriers to collaboration. The fact that breadth of engagement is negatively associated to orientation-related barriers suggests that collaborations involving multiple channels allow ?rms to cope better with the problems associated with divergent priorities and time horizons in the research. It also indicates that ?rms’ willingness to invest across many areas of engagement enables the building of routines for long-term and mutually bene?cial exchanges. However, working with universities across many different channels is also likely to involve negotiation with more university actors, including different administrative departments and possibly the TTO. As a result, broad patterns of engagement might mean involvement in numerous and lengthy interactions with university administrators, who are likely to be highly risk averse and may be responding to differing agendas and mandates. Thus, broad engagement appears to raise transaction-related barriers to collaboration.

Finally, as expected, the coef?cient of inter-organizational trust is negative and signi?cant in Models 2a (?.75, p ≤ .001) and 2b (?.37, p ≤ .001), indicating that high trust in university partners is associated with lower barriers. It is interesting that trust reduces both orientation-related and transaction-related barriers. This may be because trust relies on strong bonds of mutual understanding and adjustment and, therefore, helps ?rms to manage their different expectations of the research and to lower the considerable transaction costs of working with university partners. Table 6 presents the results for the in?uence of education-based and contract-based interactions on orientation-related barriers and transaction-related barriers respectively. There is a strong negative association between education-based interactions and orientationrelated barriers (?.08, p ≤ .01), but not between contract-based interactions and orientation-related barriers (?.03). Also, we ?nd that both types of interaction have a strong positive in?uence on the number of transaction-related barriers: .11, p ≤ .05 for educationbased interactions and .13, p ≤ .001 for contract-based interactions. The effects of the other explanatory and control variables do not change. These results show that those interactions that involve informal and frequent face-to-face contacts contribute signi?cantly to attenuating the orientation-related barriers, while broader interactions (both education and contract-based) increase the extent of transaction-related barriers. These ?ndings support the above results suggesting that the differential effect of breadth of interactions on perceived barriers increases transaction-related barriers,

Table 5 Fractional logit regression estimates of orientation-related and transaction-related barriers to interaction. Orientation-related barriers Model 1a Control variables Absorptive capacity Size Independent Doctoral Industry dummies Explanatory variables Collaboration experience Breadth of interaction Inter-organizational trust Narrow trust (three-items) Constant Log pseudolikelihood df (residual) No observations ?.08* ?.03 .02 .08 Yes Model 2a ?.07** .01 ?.13+ .13 Yes ?.05** ?.06** ?.75*** .075 ?133.20 491 503 2.86*** ?101.11 488 503 ?.48* ?52.58 491 503 Transaction-related barriers Model 1b .03 ?.02 .02 .42*** Yes Model 2b .01 ?.08*** ?.00 .35*** Yes .00 .12*** ?.37*** .83*** ?32.60 488 503 Model 2c .01 ?.08*** .01 .35*** Yes .00 .12*** ?.29*** .51*** ?36.04 488 503

One-tailed. Unstandardized coef?cients are reported. * p ≤ .05. ** p ≤ .01. *** p ≤ .001. + p ≤ .10.

866

J. Bruneel et al. / Research Policy 39 (2010) 858–868

Table 6 Fractional logit regression estimates of orientation-related and transaction-related barriers to interaction involving education-based and contract-based interaction. Orientation-related barriers Model 3a Control variables Absorptive capacity Size Independent Doctoral Industry dummies Explanatory variables Collaboration experience Education-based interaction Contract-based interaction Inter-organizational trust Narrow trust (three items) Constant Log pseudolikelihood df (residual) No observations ?.08*** ?.01 ?.13+ .13 Yes ?.05** ?.09** ?.74*** 2.83*** ?101.01 488 503 Model 4a ?.08** ?.01 ?.12+ .11 Yes ?.05** ?.05 ?.73*** 2.80*** ?102.27 488 503 Model 5a ?.07** .01 ?.13+ .13 Yes ?.05** ?.08** ?.03 ?.74*** 2.85*** ?100.92 487 503 Transaction-related barriers Model 3b .02 ?.06** ?.01 .38*** Yes .00 .13** ?.39*** .92*** ?35.20 488 503 Model 4b .02 ?.05*** ?.02 .37*** Yes .00 .16*** ?.38*** .87*** ?35.63 488 503 Model 5b .01 ?.08*** ?.00 .35*** Yes .00 .11* .13*** ?.37*** .82*** ?32.58 487 503 Model 5c .01 ?.08*** .01 .35*** Yes .00 .11* .13*** ?.29*** .51*** ?.36.01 487 503

One-tailed. Unstandardized coef?cients are reported. * p ≤ .05. ** p ≤ .01. *** p ≤ .001. + p ≤ .10.

but lowers orientation-related barriers. These ?ndings highlight the importance of education-based interactions for breaking down orientation-barriers. The dimension of trust comes out as the strongest mechanism lowering barriers relating to both orientation and transactions. However, the fact both transaction-related barriers and trust address con?dentiality and potential distributional con?icts in the way that they have been operationalized may raise a concern about endogeneity.7 We therefore used an alternative operationalization of inter-organizational trust, labelled “narrow trust”, that omits the item ‘we trust that con?dential/proprietary information shared with our university partners will be kept strictly con?dential’. We reran the analysis with transaction-related barriers as dependent variable and “narrow trust” as explanatory variable; the results of this additional analysis are reported in models 2c (Table 5) and 5c (Table 6). The results show that the effect of “narrow trust” is negative and signi?cant associated with transaction-related barriers across the different models. Also the effects of the other variables are the same. These ?ndings demonstrate that the attenuating effect of trust on transaction-related barriers is robust. 6. Conclusions and implications Although it has been widely understood that there are substantial barriers to successful collaboration and knowledge exchange between universities and ?rms, few studies have attempted to measure and map these perceived barriers or investigate what may attenuate them. From our analysis, it is clear that many types of barriers plague collaboration between industry and universities – from orientation of the university and its researchers, to attitudes and behaviour of university administration and the TTO. Although the ‘classic’ barrier to U–I collaboration – the university’s longterm orientation – remains substantial, other factors are important in constraining collaboration, especially those related to IP and administrative procedures. Some authors argue that IP-related barriers have become more prevalent in U–I interactions as a consequence of policies designed to encourage universities to increase the commercialization of

7

We thank an anonymous reviewer for bringing this to our attention.

research and to adopt a more aggressive strategy towards negotiations over IP (Siegel et al., 2003; Hertzfeld et al., 2006). While our study does not address these aspects directly, it does show that transaction-related barriers are much more dif?cult to mitigate than orientation-related barriers. In particular, while collaboration experience and breadth of interactions equip ?rms to handle (and potentially overcome) barriers related to con?icts of interest in research priorities, they do not help ?rms to handle IP-related barriers. In this respect, we show that transaction-related barriers are particularly sensitive to government policy and higher education governance. For instance, changes in the system of governance of U–I collaboration in the UK to favour the involvement of an increasing number of parties both within the university (e.g. university research contracts of?ce, TTO, the department) and the ?rm (e.g. the IP of?ce, research labs and ?rm headquarters). This trend is likely to exacerbate IP-related barriers since multiple collaborations can increase both the costs and time required to build new collaborations. This fact may require greater attention be given to investing in ex-ante IP agreements to avoid follow-on con?icts over the distribution of rewards from a project. However, the bene?ts of putting in place such agreements must be set aside the costs of negotiation to determine whether such upfront negotiations are necessary or convenient for either the university or the ?rm for a given project. At the same time, older and more informal systems of exchange and collaboration are coming under increasing scrutiny from university administrators. Such efforts to bring exchanges and interaction ‘in from the cold’ could have the effect of raising transaction-related barriers, especially if these efforts are organized around the requirements of central university rules and regulations. Thus, increasing attention to the management of U–I links through government policy efforts and university administration could increase the barriers to such interactions. It would be unfortunate if the efforts to manage (and potentially support) these interactions result in increasing the barriers to collaboration. The challenge for policy, then, is to ?nd straightforward, simple mechanisms for management and monitoring of U–I interactions. Achieving this will require attention to the costs and bene?ts of management and monitoring efforts, and the careful weighing of the value of monitoring against negatively perceived intrusion.

J. Bruneel et al. / Research Policy 39 (2010) 858–868

867

An important ?nding from this study is that inter-organizational trust is one of the strongest mechanisms for lowering the barriers to interaction between universities and industry. It suggests that the traditional system of informal reciprocity and exchange, which dominated U–I exchanges in the postwar era, should be an important part of attempts to support and build U–I collaborations. Building trust between academics and industrial practitioners requires long-term investment in interactions, based on mutual understanding about different incentive systems and goals. It also necessitates a focus on face-to-face contacts between industry and academia, initiated through personal referrals and sustained by repeated interactions, involving a wide range of interaction channels and overlapping personal and professional relationships.

Acknowledgements The author names are ordered alphabetically. This research was conducted as part of the Advanced Institute of Management’s Innovation and Productivity Grand Challenge, supported by the UK’s Economic and Social Research Council and Engineering and Physical Sciences Research Council. Pablo D’este and Ammon Salter respectively acknowledge ?nancial support from Generalitat Valenciana through the project GV/2009/048,and from the UK Innovation Research Centre (UK?IRC), funded by BIS, TSB, ESRC and NESTA. We are grateful to Kate Bishop for her efforts on the survey. The paper has bene?ted from comments from Oliver Alexy, Keld Laursen, Markus Perkmann and Bruce Tether. We are indebted to the EPSRC for its generous support for the research. The authors are solely responsible for any errors or statements in this paper.

6.1. Future research and limitations Understanding the perceived barriers to U–I collaboration is important because it uncovers the problems and challenges that have emerged in the U–I collaboration process. Much of the research on U–I links relies on secondary information on the problems and challenges involved in collaboration. There is very little information on those ?rms actually involved in these collaborations. The present study looks at the mechanisms that may mitigate the barriers to U–I collaboration, and may help to set in place policies that will alleviate the problems before they undermine what might be rewarding sets of collaborations. Although we draw information from multiple sources and over time, our study focuses on one period, which makes it dif?cult to draw inferences about the direction of causality. Future research should explore the barriers over time, and examine the factors that lower or raise the barriers to collaboration. It may be that policy interventions, such as new university IP policies or changes in university funding regimes, will have a signi?cant impact on the perceived barriers. In addition, although we have suggested that problems of coordination within the university may give rise to the increases of transaction-related barriers, our study does not explicitly measure the efforts of university to coordinate its activities and therefore we cannot know whether this explanation is the correct one. Like many studies in this area, we have treated the university as a single unit, yet universities are rich, complex organizations, rife with diversity and even con?ict. It is possible that in the future university systems may be more tightly integrated to enable more effective engagement with industry across different channels and therefore lower transaction-related barriers for industry. Future research should also examine the impact of barriers on the outcomes of collaboration. Although it is assumed that these barriers hinder effective knowledge exchange, we do not have evidence on how the perceived barriers shape subsequent collaborations. For example, it would be useful to know whether a bad experience of university collaboration deters the ?rm from future collaboration with a university. We also do not know how the use of ex-ante IP agreements may help lower downstream con?icts between universities and industry partners or whether such agreements themselves create barriers for successful collaborations. This too is a critical area for further research. In this paper, we focused on a sample of ?rms that had been involved in formal research projects, but many ?rms never get this far in their interactions with universities. For most ?rms, U–I interaction involves a long process of learning through small steps, such as enabling student placements through to more extensive engagement. We know little about ?rms’ progression from informal, infrequent interactions to long-term, sustained collaboration with universities. An understanding of this progress may offer great potential for effective policy measures to support U–I collaboration. References
Agrawal, A., Henderson, R., 2002. Putting patent in context: exploring knowledge transfer at MIT. Management Science 48, 44–60. Allen, J.T., 1984. Managing the Flow of Technology: Technology Transfer and the Dissemination of Technological Information within the R&D Organization. MIT Press, Cambridge (MA). Armstrong, J.S., Overton, T.S., 1977. Estimating non-response bias in mail surveys. Journal of Marketing Research 14, 396–402. Arundel, A., Geuna, A., 2004. Proximity and the use of public science by innovative European ?rms. Economics of Innovation and New Technologies 13, 559–580. Azoulay, P., Ding, W., Stuart, T., 2007. The determinants of faculty patenting behaviors: demographics or opportunities. Journal of Economic Behavior & Organization 63, 599–623. Becher, T., 1989. Academic Tribes and Territories – Intellectual Enquiry and the Cultures of Disciplines. The Society for Research into Higher Education & The Open University, Milton Keynes. Bishop, K., D’Este, P., Neely, A., 2009. Gaining from interactions with universities: multiple methods for nurturing absorptive capacity. Research Policy, Forthcoming. Bradbach, J.L., Eccles, R.G., 1989. Price, authority, and trust: from ideal types to plural forms. Annual Review of Sociology 15, 97–118. Brown, J.S., Duguid, P., 2000. The Social Life of Information. Harvard Business School Press, Boston, Massachusetts. Calderini, M., Franzoni, C., Vezzulli, A., 2007. If star scientists do not patent: the effect of productivity, basicness and impact on the decision to patent in the academic world. Research Policy 36, 303–319. Cantwell, J., 1989. Technological Innovation and Multinational Corporations. Blackwell, Oxford. Chapple, W., Lockett, A., Siegel, D., Wright, M., 2005. Assessing the relative performance of U.K. university technology transfer of?ces: parametric and nonparametric evidence. Research Policy 34, 369–384. Chesbrough, H., 2006. Open Business Models. Harvard Business School Press, Cambridge, Mass. Clarysse, B., Wright, M., Lockett, A., Mustar, P., Knockaert, M., 2007. Academic spinoffs, formal technology transfer and capital raising. Industrial and Corporate Change 16, 609–640. Cockburn, I.M., Henderson, R.M., 1998. Absorptive capacity, coauthoring behavior and the organization of research in drug discovery. Journal of Industrial Economics XLVI, 157–181. Cohen, W.M., Levinthal, D.A., 1990. Absorptive capacity: a new perspective of learning and innovation. Administrative Science Quarterly 35, 128–152. Cohen, W.M., Nelson, R.R., Walsh, J., 2002. Links and impacts: the in?uence of public research on industrial R&D. Management Science 48, 1–23. Cohen, W.M., Sauermann, H., 2007. Schumpeter’s prophecy and individual incentives as a driver of innovation. In: Brusoni, S., Malerba, F. (Eds.), Perspectives on the Economics of Innovation. Cambridge University Press, London. Criscuolo, P., 2005. On the road again: researcher mobility inside the R&D network. Research Policy 34, 1350–1365. D’Este, P., Patel, P., 2007. University–industry linkages in the UK: what are the factors underlying the variety of interactions with industry? Research Policy 36, 1295–1313. Dasgupta, P., David, P., 1994. Towards a new economics of science. Research Policy 23, 487–522. Dodgson, M., 1993. Learning, trust and technological collaboration. Human Relations 46, 77–95. Etzkowitz, H., Leydesdorff, L., 2000. The dynamics of innovation: from national systems and “Mode 2” to triple Helix of university-industry-government relation. Research Policy 29, 109–123. Fabrizio, K., DiMinin, A., 2008. Commercializing the laboratory: the relationship between faculty patenting and publishing. Research Policy 37, 914–931. Florida, R., 1999. The role of the University: leveraging talent, not technology. Issues on Science and Technology XV, 67–73.

868

J. Bruneel et al. / Research Policy 39 (2010) 858–868 Papke, L.E., Wooldridge, J.M., 1996. Econometric methods for fractional response variables with an application to 401(k) plan participation rates. Journal of Applied Econometrics 11, 619–632. Pavitt, K., 2001. Public policies to support basic research: what can the rest of the world learn from US theory and practice? (And What They Should Not Learn). Industrial and Corporate Change 10, 761–779. Polanyi, M., 1962. The republic of science: its political and economic theory. Minerva 1, 54–73. Rempel, J.K., Holmes, J.G., 1986. How Do I Trust Thee? Psychology Today 20 (2), 28–34. Ring, P.S., Van de ven, A.H., 1992. Structuring cooperative relationships between organizations. Strategic Management Journal 13, 483–498. Rosenberg, N., 2002. America’s University/Industry Interfaces 1945–2000. Working paper, Department of Economics, Stanford University. Rosenberg, N., Nelson, R.R., 1994. American universities and technical advance in industry. Research Policy 23, 323–348. Rothaermel, F.T., Agung, S.H., Jiang, L., 2007. University entrepreneurship: a taxonomy of the literature. Industrial and Corporate Change 16, 691–791. Rothwell, R., Dodgson, M., 1991. External linkages and innovation in small and medium-sized enterprises. R&D Management 21, 125–137. Santoro, M., Gopalakrishnan, S., 2001. Relationship dynamics between university research centers and industrial ?rms: their impact on technology transfer activities. Journal of Technology Transfer 26, 163–171. Santoro, M., Saparito, P., 2003. The ?rm’s trust in its university partner as a key mediator in advancing knowledge and new technologies. IEEE Transactions in Engineering Management 50, 362–373. Schmidt, T., 2005.Absorptive capacity: one size ?ts all? A Firm Analysis of Absorptive Capacity for Different Kinds of Knowledge, Mannheim. Shane, S., 2004. Academic Entrepreneurship: University Spinoffs and Wealth Creation. Edward Elgar, MA. Shane, S., Somaya, D., 2007. The effects of patent litigation on university licensing efforts. Journal of Economic Behavior & Organization 63, 739–755. Siegel, D.S., Waldman, D.A., Link, A.N., 2003. Assessing the impact of organizational practices on the productivity of university technology transfer of?ces: an exploratory study. Research Policy 32, 27–48. Stephan, P., 1996. The economics of science. Journal of Economic Literature XXXIV, 1199–1235. Stern, S., 2004. Do scientists pay to be scientists. Management Science 50, 835–853. Stockdale, B., 2002. UK Innovation Survey. Department of Trade and Industry, London, http://www.dti.gov.uk/iese/ecotrends.pdf. Teece, D., 1986. Pro?ting from technological innovation: implications for integration collaboration, licensing and public policy. Research Policy 15, 285– 305. Tether, B.S., 2002. Who co-operates for innovation, and why: an empirical analysis. Research Policy 31, 947–967. Valentin, F., Jensen, R.L., 2007. Effects on academia-industry collaboration of extending university property rights. Journal of Technology Transfer 32, 251– 276. Van Dierdonck, R., Debackere, K., 1988. Academic entrepreneurship at Belgian universities. R&D Management 18, 341–353. Vincenti, W., 1990. What Engineers Know and How They Know It. John Hopkins Press, Baltimore. von Hippel, E., 1987. Cooperation between rivals: informal know-how trading. Research Policy 16, 291–302. von Hippel, E., von Krogh, G., 2003. Open source software and the “PrivateCollective” innovation model. Organization Science 14, 208–223. Welsh, R., Glenna, L., Lacy, W., Biscotti, B., 2008. Close enough but not too far: assessing the effects of university–industry research relationships and the rise of academic capitalism. Research Policy 37, 1854–1864. Williamson, O., 1993. Opportunism and its critics. Managerial and Decision Economics 14, 97–107. Wooldridge, J.M., 2002. Econometric Analysis of Cross Section and Panel Data. The MIT Press, Cambridge Massachusetts. Zaheer, A., McEvily, B., Perrone, V., 1998. Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science 9, 141–159.

Fontana, R., Geuna, A., Matt, M., 2006. Factors affecting university–industry R&D projects: the importance of searching, screening and signalling. Research Policy 35, 309–323. Freeman, C., 1999. The social function of science. In: Swann, B., Aprahamian, F. (Eds.), J.D. Bernal: A Life in Science and Politics. Verso, London, New York. Geuna, A., Nesta, L.J.J., 2006. University patenting and its effects on academic research: the emerging European evidence. Research Policy 35, 790–807. Geuna, A., Steinmeuller, W., Salter, A.J. (Eds.), 2003. Science and Innovation: Changing Rationales for the Public Funding of Research. Edward Elgar, Cheltenham, UK. Gibbons, M., Johnston, R., 1974. The role of science in technological innovation. Research Policy 3, 220–242. Gomes, J.F.S., Hurmelinna, P., Amaral, V., Blomqvist, K., 2005. Managing relationships of the republic of science and the kingdom of industry. Journal of Workplace Learning 17, 88–98. Hagedoorn, J., Schakenraad, J., 1994. The effect of strategic technology alliances on company performance. Strategic Management Journal 15, 291–311. Hall, B.H., Link, A.N., Scott, J.T., 2001. Barriers inhibiting industry from partnering with universities: evidence from the advanced technology program. Journal of Technology Transfer 26, 87–98. Hall, B.H., Link, A.N., Scott, J.T., 2003. Universities as research partners. Review of Economics and Statistics 85, 485–491. Henderson, R., Jaffe, A., Trajtenberg, M., 1998. Universities as a source of commercial technology: a detailed analysis of university patenting, 1965–1988. The Review of Economics and Statistics LXXX, 119–127. Hertzfeld, H.R., Link, A.N., Vonortas, N.S., 2006. Intellectual property protection mechanisms in research partnerships. Research Policy 35, 825–838. Hicks, D., 1995. Published papers, tacit competencies and corporate management of the public/private character of knowledge. Industrial and Corporate Change 4, 401–424. Inkpen, A.C., Tsang, E.W.K., 2005. Social capital, networks, and knowledge transfer. Academy of Management Review 30, 146–165. Jelinek, M., Markham, S., 2007. Industry–university IP relations: integrating perspectives and policy solutions. IEEE Transactions in Engineering Management 52, 257–267. Kogut, B., 2000. The network as knowledge: generative rules and the emergence of structure. Strategic Management Journal 21, 405–425. Kogut, B., Zander, U., 1992. Knowledge of the ?rm, combinative capabilities, and the replication of technology. Organization Science 3, 383–397. Lambert, R., 2003. Lambert Review of Business–University Collaboration. HM Treasury, London. Lane, P., Koka, B., Pathak, S., 2006. The rei?ciation of absorptive capacity: a critical review and rejuvenation of the construct. Academy of Management Review 31, 833–863. Laursen, K., Salter, A., 2004. Searching low and high: what types of ?rms use universities as a source of innovation. Research Policy 33, 1201–1215. Lawrence, T.B., Hardy, C., Phillips, N., 2002. Institutional effects of interorganizational collaboration: the emergence of proto-institutions. Academy of Management Journal 45, 281–290. McEvily, B., Perrone, V., Zaheer, A., 2003. Trust as an organizing principle. Organization Science 14, 91–103. Merton, R., 1973. The Sociology of Science: Theoretical and Empirical Investigations. University of Chicago Press, Chicago; London. Meyer-Krahmer, F., Schmoch, U., 1998. Science-based technologies: university–industry interactions in four ?elds. Research Policy 27, 835–851. Mohnen, P., Hoareau, C., 2003. What type of enterprise Forges close links with universities and government labs? Evidence from CIS 2. Managerial & Decision Economics 24, 133–146. Mowery, D.C., Nelson, R.R., Sampat, B.N., Ziedonis, A.A., 2001. The growth of patenting and licensing by U.S. universities: an assessment of the effects of the Bayh-Dole act of 1980. Research Policy 30, 99–119. Mowery, D.C., Ziedonis, A.A., 2002. Academic patent quality and quantity before and after the Bayh-Dole act in the United States. Research Policy 31, 399–418. Nelson, R.R., 2004. The market economy, and the scienti?c commons. Research Policy 33, 455–471. Nooteboom, B., 2002. Trust: Forms, Foundations, Functions, Failures and Figures. Edward Elgar, Cheltenham.


赞助商链接
相关文章:
...and affiliated firm innovation The case of Taiwa...
Barriers and drivers for... 8页 免费 Factors affecting...(i) investigating the options of standardization ...CBM industry is expected to prosper in the ...
更多相关标签: