GOSt
GROUP OF ORGANIZATION STUDIES
Academic Entrepreneurship and Scientific Research: Synergy Or Trade-Off? Giancarlo Lauto Copenaghen Business School
[email protected] Massimo Bau’ Università degli Studi di di Udine
[email protected] Cristiana Compagno Università degli Studi di Udine
[email protected]
paper presented at: WOA 2011 – XII Workshop dei Docenti e dei Ricercatori di Organizzazione Aziendale Napoli, 16-18 Giugno 2011
Ref.: Lauto G., Bau’ M., Compagno C. (2011). “Academic Entrepreneurship and Scientific Research: Synergy Or Trade-Off?” in Albertini S., Bergami M., D'Atri A., De Marco M., De Vita P., Ferrara M., Rossignoli C., Salvemini S., Generazioni e RiGenerazioni nei Processi Organizzativi. Napoli. ISBN: 978-88-89677-21-6 http://gost.uniud.it/papers/
Academic entrepreneurship and scientific research: synergy or tradeoff?
ABSTRACT The inclusion of technology transfer among the missions of universities gave rise to the model of the “academic entrepreneur”. But is academic entrepreneurship compatible with the traditional goals of universities, or does it divert scientists and university from their core social function? The literature suggests that complementarities exist between research and patenting. We hold on an emerging stream of studies that adopts a more comprehensive view of technology transfer, and include in our analysis spin-off creation, consulting and combination of consulting and patenting. The empirical study conducted on 249 Italian university researchers indicates that research is more compatible with patenting and less with consulting, and complementarities can be exploited in particularly by senior and star scientists. Moreover, the resources and motivational factors enabling publishing are by and large the same that enable technology transfer.
1. INTRODUCTION In the last decade science policy made an effort to strengthen the relationship between science and industry with the goal of boosting technological innovation and economic competitiveness (Martin, 2001; Pavitt, 2001). Under these stimuli, universities included the support of socioeconomic development at regional and local level among their missions, together with the traditional missions of teaching and research (Shane, 2004; Etzkowitz and Leydesdorff, 2000; Gibbons et al., 1994), thus giving rise to the novel entrepreneurial model of the “academic entrepreneur” (Etzkowitz, 2003). A central issue in the literature on science policy deals with the compatibility between the traditional and the novel missions of academic institutions, and the role of the academic entrepreneur in an Open Science context. This article examines the synergies and trade-offs between traditional and new missions at individual researcher level, asking whether i) an intense orientation towards TT penalizes or improves the production of scientific discoveries, and ii) the configuration of resources supporting TT diverges or overlaps that one supporting the production of scientific publications. This article aims at contributing to the debate on the issue by i) extending the spectrum of TT activities under scrutiny – generally confined to patenting – including consultancy; ii) considering the complementarities of resources and motivational factors to explain scientific performance. The remainder of the article is organized as follows: Section 2 presents the theoretical framework and the empirical evidence on which we build the hypothesis that guide our investigation; Section 3 outlines the research design; we present and discuss the findings in Section 4; Section 5 is devoted to the concluding remarks. 1 !
2. BACKGROUND The aforementioned missions of universities translate into three dimensions of performance at the individual scientist-level. Scientists are invested of the functions of preserving the stock of knowledge by transmitting it to the new generations, and of expanding that knowledge. In Open Science, curiosity-driven scientists’ research activities are coordinated through the adhesion to the collective institutions of the scientific community. Research priorities, legitimacy of intellectual approaches, resource allocation, and scientists’ material compensation are governed through a selfregulating collegiate mechanism based on the publication system. This system rewards the scientists who first disclose findings that contribute to the advance of the field serving as an input to the subsequent investigations (Dasgupta and David, 1994; Merton, 1957; Polanyi, 1962; Whitley, 1984). In addition to the priorities autonomously defined in Open Science, scientists are exposed to the demands of science-external stakeholders. Scientists are required to produce results relevant for the socio-economic community they belong to, and, most importantly, for industrial sponsors which increasingly substitute declining public funding. Enhancing the process of innovation, engagement in technology transfer (TT) responds to both societal and industrial expectations. Moreover, commercialisation of research outcomes opens additional sources of income for scientists and universities. Besides financial resources, interaction with industry offers scientists learning opportunities, opens up new avenues for research, extends the network of potential collaborations, and allows access to complementary research equipment (Meyer-Krahmer and Schmoch, 1998;Rosenberg and Nelson, 1994; Stephan et al., 2007). On the other side, it threatens the Open Science mechanisms for regulation of knowledge production, since industry-funded research rewards secrecy disclosure causing delays in publication (Calderini et al., 2007; Fabrizio and Di Minin, 2008; Geuna and Nesta, 2006; Lee, 2000; Mowery et al., 2001; Murray and Stern, 2007) hindering diffusion in further studies or leading to a privatization of the “scientific commons”. These aspects are particularly significant in fields such as genetics or biochemistry in which the ethical issue about publish or patent must considers both the societal impact and the interests of private funders (Macer, 2002). Moreover, involvement of industry may incentivise scientists to neglect more fundamental lines of research in favour of more contingent, problem driven issues (Lee, 1996). Successful commercialization of research outcomes often implies the transfer of the tacit knowledge requiring the personal involvement of the inventor (Agrawal, 2006; Thursby et al., 2007a; Thursby et al., 2007b); this phenomenon is even more pronounced in the creation of spinoff companies, requiring scientists to build complementary managerial competences to run the firm. In other words, TT diverts limited time and resources available from research to activities in which scientists are less efficient (e.g. Fabrizio and Di Minin, 2008; Lowe and Gonzales-Brambila, 2007; Stephan et al., 2007). A recent review (Larsen, 2011) analyzed several empirical studies that have already addressed our first research question relative to the compatibility between the missions of research and TT. From this literature, substantial consensus emerges about the positive relationship between the TT (patenting in particular) and research productivity; evidence about the effect on research impact and the focalization towards basic research is instead less unambiguous. Fewer studies focussed on creation of spin-off companies and provide contrasted evidence (Buenstorf, 2009; Lowe and Gonzales-Brambila, 2007). The studies that investigated the effect of patenting on research performance in the Italian setting (Balconi and Laboranti, 2006; Balconi et al., 2004; Breschi et al., 2007; Calderini, Franzoni and Vezzulli, 2007; Baldini et al. 2007), substantially confirm the general findings emerged in the literature. 2 !
To the best of authors’ knowledge, only a recent article (Landry et al., 2010) adopted a multidimensional view of research performance, extending the analysis of TT activities to informal collaboration, consulting, and spin-off formation. That study suggests that complementarities exist between different forms of TT and also with research. We thus expect Hypothesis-1 that a positive relationship between TT and research publication exists. Sharing such a multidimensional vision of research performance and following (Compagno et al., 2010a), we articulate TT into four activities: creation of spin-off; patenting; consulting; conjoint patenting and consulting. We will test Hypothesis 1 with reference to each of these TT activities. To address our second research question relative to the nature of the factors supporting scientists’ productivity, accordingly with (Landry et al., 2006; Landry et al., 2007), we interpret individual researchers’ performance through the lenses of the Resource-Based View of the Firm (Barney, 1991; Wernerfelt, 1984). This framework suggests that the heterogeneity and uniqueness of the resource endowment explains the difference in scientists’ performance – the “competitive advantage”. The approach considers an array of financial, technological, cognitive, organizational, social and human resources. Previous studies (Compagno et al., 2007, 2008, 2010a, 2010b) (Compagno et al., 2007; Compagno et al., 2008; Compagno, Lauto and Baù 2010a; Compagno et al., 2010b) extended this framework to consider motivational factors affecting scientific performance, such as the “occupational hazard” and the “adherence to the values Open Science”. The former refers to the level of the researcher’s tolerance to undermine his/her reputation in the academia in consequence of an involvement in extra-academic activities. Consistently with the incentive structure of Open-Science, we reconnect scholars’ motivation to the model defined by (McClelland, 1961) as “Need for achievement” that identifies recognition and prestige within the community of practice as the element driving motivation processes. Activities diverting scientists’ time and energy from research – such as TT – may potentially compromise the development of scientific careers and thus compromise the recognition of a scientist in his/her community. The latter motivational factor refers to the commitment of the researcher to the Mertonian norms of communalism, universalism, disinterestedness and organized scepticism (Merton, 1942), that might conflict with the goals of TT. From previous research we derive a set of hypothesis on the direction of the effect of these resources and motivational factors on research productivity. Financial resources provided by industry (Blumenthal et al., 1996; Gulbrandsen and Smeby, 2005), the EU, and extra-academic sources in general (Geuna and Nesta, 2006) improve scientific productivity. Hypothesis 2 holds that funding from non-academic stakeholders has a positive impact on scientific production. University-industry co-authorship and contract research with industrial and no-profit organizations are positively related with research productivity (Van Looy et al., 2004) as well. Hypothesis-3 expects that the extension of social capital with extra-academic stakeholders has a positive impact on scientific production since it enables these forms of collaborations. The research team is the most immediate organisational resource a scientist may benefit from. Relying on a stable group of direct collaborators, scientists may achieve economies of specialization through a division of the intellectual labour and generate heterogeneity of skills and competences (Wuchty et al., 2007). We thus formulate Hypothesis-4 that the size of the research team has a positive impact on scientific production. The effects of human capital on scientific production are ambiguous (Carayol and Matt, 2004): from one side, the expected returns of investments in human capital decrease with the remaining activity period, and incentives for producing research decline once promotion is obtained. On the other side, scientific human capital enables a clearer vision of the dynamics in the field, the identification of the most promising avenues, the development of rich professional networks and a better the exploitation of external resources. Moreover, according to Feldman and colleagues (2001), once a researcher achieves important academic milestones – such as academic status, tenure, reputation – she tries to obtain 3 !
financial return on his human capital through TT activities. For this reason we do not formulate any hypotheses regarding human capital. !
Turning to the motivational factors, theory and empirical evidence suggest that researchers driven by the “need for achievement” would focalize their intellectual resources in order to to excel in research activities. Those aspects are strongly related to the academic lifecycle models (Stefan and Levin, 1996). Shane (2004) suggests that academic researcher establish more structured forms of TT in the later stages of their career, adopting a career oriented behaviour in early stages. Audretsch (2000), for example found that – in comparison to industrial scientists who start firms – university scientists are older and more experienced, a fact that the author attributes to the academic reward system. Even more, Shane and Khurana (2003) found that investors and external stakeholders evaluate more positively the high status of researchers favouring the possibility to obtain the resources that they need. In other words, two aspects seems to be significant. First, the motivational factor would more than compensate the net benefit arising from the engagement in TT. Second, to reach a higher status in the academic career favour the research option instead of the TT option. This choice reduces the professional risk within academia, postponing the undertake of TT activities to a later stage, when the status favour sponsors’ judgement (Merton, 1973; Latour, 1987). For this reason, we expect Hypothesis-5 that the aversion to professional risk has a positive impact on scientific production. Researchers who define their professional identity in close adherence to the values of the scientific community (Figueiredo Moutinho et al., 2007; Lam, 2010) and who obtain satisfaction from doing research reach higher levels of productivity (Chen et al., 2006; De Witte and Rogge, 2010) improve the quality of their job. We thus expect Hypothesis-6 that the adherence to the values of Open Science has a positive impact on scientific production. Previous studies addressing the role of resource and motivational factors as drivers of TT and spinoff creation in the Italian context (Compagno et al., 2007, 2008, 2010a, 2010b) show that notable differences exist among different forms of university-industry interaction but, in general, external financial resources, extension of social capital outside academia, professional hazard are positively associated to the engagement in these activities. These studies also find that the adherence to the values of science has no effect in explaining the involvement in university-industry interaction.
3. RESEARCH DESIGN The first step of our analytical strategy addresses the nexus between publishing and TT with through a descriptive analysis; then, we identify the resources and motivational factors enabling research productivity with inferential statistics. We take into exam the population of researchers working in Italian Universities, considering both tenured and untenured (doctoral students and research assistants) researchers given the involvement of the latter in the processes of TT. While previous studies focus on Science and Engineering, we consider all disciplines to provide a wide-shot picture of TT. To account for the differences across disciplines in the opportunities of TT, we take care of including field controls in the regression models. The selected sample comprises 5.269 researchers belonging to 41 Italian Universities We gathered the information on research and TT activities, resource endowment and motivations through a questionnaire sent by e-mail between September and October 2007. We received 249 answers from scholars affiliated to 28 universities (response rate of 4.7%). The study refers to 4 !
activities undertaken by the researchers during the previous five years; the period is regarded as the onset of the diffusion of the entrepreneurial culture in the Italian academic context (Piccaluga and Balderi, 2006). The distribution of the respondents according to academic status and gender mirrors that of the academic population as of 31st December 2006 (Italian Ministry of University and Research – MIUR), while the distribution by field shows the predominance of the Engineering, Computer Science and Chemistry areas that account for 37% of the sample, probably in consequence of the higher interest for TT in these fields. The dependent variable in our study is “research productivity”, measured by the number of selfreported publications. We attribute weight 5 to work published in international outlets. In the descriptive analysis, we compare the productivity between researchers who attempted to establish a spin-off company and the remaining of the population, and among the three profiles of TT and researchers who do not engage in TT. The “Complex” profile includes researchers who protected IPR and have been involved “often” or “very often” in consultancy. The “Technological” profile includes exclusively researchers who protected IPR, the “Consultant” profile exclusively those who worked intensively in consultancy. “Disengaged” researchers have been involved in neither activity. We normalized the research productivity on the mean and standard deviation of each discipline, to account for the cross-field differences in productivity: a positive normalized productivity indicates that a given researcher’s production is superior to the mean of his/her discipline. In the inferential analysis, we adopt a negative binomial regression model since our dependent variable represents skewed count data. To account for the correlation of observations, we clustered robust standard errors based on the 28 university affiliations. We included in the model three controls: for the gender (Sax et al., 2002), for the importance of sources funding internal to the academic system, and for the involvement in teaching. Since teaching absorbs resources otherwise dedicated to research, but also offers intellectual complementarities (Mitchell and Rebne, 1995) we modelled a non-linear relationship including a linear and a squared normalized term. The explanatory variables are summarized in Table 1.
Table 1. Descriptive statistics !
!"#$%&'(&)*!
!"'"*0"*(-2,%&,78"#! "#$#%&'(!)&*+,'-./.-0! 9*0"'"*0"*(-2,%&,78"#:)*(&*;);#! CD-#&2%4!3,2+.28! O&./%-#!3,2+.28! CH!3,2+.28! OOI3,2+.28!
! 1*,2-!*3!$#43!&#)*&-#+!),54.'%-.*2$6!7#.8(-!9!3*&!.2-#&2%-.*2%4!),54.'%-.*2! ! ! EF)*&-%2'#!*3!)&./%-#!3.2%2'.28G!F.D#+!%2+!-(#!CH!I!J.K#&-!$'%4#!! ! ! ! ! ?6>L!=M@! N6B>!=B6@! B!=:LM@!
+,6-./4,%"-5,63! ! !=>6L;@!
B!=B:M@!
9!=BM@!
96A;!=:6>!=BM@!
!=>69M@!
N69
B!=! :9Z!BLM! BAZ!AM! Z!BNM!
69@! N!=LM@! ! ! ! !
B:!=BM@! B>!=9M@! ! ! ! !
! ! P-%-,$! ! ! ! Q.#4+! ! ! ! ! ! ]#2+#&!
1*2$,4-%2-! W.$#28%8#+! Q,44!)&*3#$$*&! Y$$*'.%-#!)&*3#$$*&! Y$$.$-%2-!)&*3#$$*&! H2-#2,&#+! C28.2##&.28