Corps de l’article

In recent years, an increased attention has been dedicated to the research spin-offs (RSOs) phenomenon (Mustar et al., 2006, 2008; Clarysse et al., 2011; Sternberg, 2014). RSOs are considered as one of the main tools for the external transmission of knowledge and economic valorisation realized in a university, usually referred to as the parent institute. The growing importance of the role of universities in local development - the so-called third mission (Chapman et al., 2011; Philpott et al., 2011; Goldstein, Glaser, 2012; Carree et al., 2012; Treibich et al., 2013; Algieri et al., 2013) – is linked to a large number of initiatives enhancing the academic entrepreneurship both through the creation of internal structures devoted to technology transfer and the creation of RSOs.

The persistence of close links with the parent institute and the importance of the degree of support that a RSO receives from its university for company success have been underlined in the literature[1]. The question surfaces, what about the effective link between RSOs and their parent institute? Are there any differences between companies that aroused more “interest” from their university as opposed to those that did not? The international literature on RSOs has focused on several perspectives and has proposed many taxonomies in recent years. Besides the resource-based view and the business model standpoint, the institutional perspective provides a key focus of these researches: it proposes, in particular, alternate evaluations of the structuring and economic success of RSOs and focuses on the more formal or informal relations with the university (Mustar et al., 2006; Rasmussen, 2006; Wright et al., 2007; Freitas et al., 2013). Along this line, this paper aims at nurturing the debate through the analysis of the results of primary data sources coming from a survey undertaken in RSO founders[2]. Following the management research field with assumptions about stakeholders’ interest, particular attention is paid to a comparison between RSOs, according to the “interest” they fostered from their parent institute.

In order to inhibit the various national institutional structures that might influence the development of RSOs, we decided to focus on a single national basis, namely Italy.

Italy is an interesting case-study often developed in the literature: the Italian industrial districts (Becattini et al., 2003) are well known at worldwide level and the country has always been characterized - and taken into consideration – by a predominance of small and medium businesses as well as family businesses. The innovation ecosystem makes Italy an attention-grabbing case in terms of variety of existing firms and potential distinctiveness. A fertile and industrialised context in the North (Nosella, Grimaldi, 2009) concurs with a strong regional and innovation divide, among the sharpest in the European Union (Iammarino et al., 2009). Notwithstanding the general low level of total early stage entrepreneurial activity registered in recent years, a real entrepreneurial ferment (i.e. the propensity to start new businesses) is observable (GEM, 2014) with a particular increasing number of RSOs foundation (Iacobucci et al., 2011; Netval, 2013). Italy is therefore a country with a relevant potential in terms of technology transfer (Varaldo, Di Minin 2009). Nonetheless, the results are generalizable beyond the Italian case, because European RSOs share most of the same characteristics and problems (Mustar et al., 2008; Visintin, Pittino, 2014).

An extended analysis of the RSO phenomenon has been undertaken thanks to a questionnaire investigation. It started from general descriptive statistics of the answers provided to the several sections of the questionnaire with particular attention to the relationship between RSOs and their parent institute; then, cluster analysis and factor analysis approaches contributed to build some groups and factors useful for understanding the effective nature of this relationship. The cluster analysis revealed the existence of two main kinds of RSOs: “open-oriented” and “autonomous-oriented”. The factor analysis confirmed and improved these results by underlining the importance of three factors that develop more specifically the various dimensions shaping the openness and governance of the companies. The first factor corresponds to “managerial competencies”, therefore, the capacity to support effective company development. The second factor is the “company orientation”, therefore the structure and control over exclusive offering (products or services). Lastly, the third factor refers to “spin-off founders and university choices”, therefore the dependence of the RSO upon its Alma Mater. These findings as well as the descriptive statistics from the questionnaires confirmed the importance of the role played by the RSO parent institute “interest”. Yet, the paper demonstrates that this relation is neither unique nor uniform. Moreover the analysis of the relationship between RSOs and their parent institute underlined that several specificities might be distinguished and, consequently, it calls for different strategies and support from the university towards the various kinds of firms. In particular, the development of managerial competencies and, accordingly, the future economic success seems closely linked to the degree of interest demonstrated by the university. Moreover, when their founders are not keeping an academic position in the university, RSOs are less supported by their parent institute; their founders pursue the economic development of the firm but are not encouraged to fully valorise their dual experience in the mastering of scientific tacit knowledge and in business development.

The paper is structured as follows. Section 1 focuses on the main RSO perspectives and taxonomies suggested in recent literature as well as on the meaning of university’s “interest”: according to the meaning of interest usually used in organisation studies and strategic management, we aim at investigating - on the specific case of RSOs - the impact of the involvement of specific stakeholders (namely universities) on the success of the firms they are participating to. Data and methodology are presented in section 2. The main results of the empirical investigation are described in section 3: descriptive statistics of the 155 questionnaires received as well as the results of the cluster analysis then a factor analysis approach. Discussion of the results and suggestions for improvement follow.

Theoretical framework: main RSO perspectives and taxonomies in recent literature and the concept of “interest”

In recent years, several theoretical perspectives (Wright et al., 2007) have been used to analyse RSOs: they highlight the complexity of the phenomenon. Consequently, different definitions and many taxonomies have been derived. The type of resources, the business model and the institutional link are used to differentiate between RSOs in order to clarify the heterogeneity and diversity of these firms. Recently, Rasmussen (2011) argued that life-cycle, teleological, dialectic and evolutionary theories (see Table 1 for details) explain different aspects of the RSO venturing formation process. In particular, the dialectic perspective underlines the influence of the university culture on RSO behavior.

Wright et al. (2007) attempted to build spin-off taxonomies to fill the gaps with previous investigations. They classified the scientific production on RSOs (Table 2), distinguishing three main types of companies: the venture capital backed, the prospector and the lifestyle RSOs. This distinction is based on a set of variables identified by the authors and fit into three theoretical perspectives (Table 1). Firstly, the institutional, focused on the scrutiny of formal or informal relations with the parent institute and its strategic choices; then, the business model perspective, looking at the economic design and efficiency of the activities developed by the firm; thirdly, the resource-based view addressing the identification of the key scientific resources, the role they played and how they are coped with.

Shane (2004) focused on needs and sources of finance, featuring the assorted funding sources for RSOs. A first category includes firms needing a minimum amount of finance, because they are financed through personal and family capital. A second category includes those RSOs that need a high level of finance.

If Shane (2004) identified two categories, others identified three or four types of RSOs. Clarysse et al. (2002, 2005) and Degroof and Roberts (2003, 2004) classified companies according to the low selective, supportive and incubator models. According to these different models, the creation of RSOs may vary from several companies of low-medium quality to only a few companies with high level potentialities. The parent institute strategy adopted for the creation of these companies and the availability of adequate infrastructures like science parks and incubators are the main determinants of these results. O’Shea et al. (2005) analysed not only the impact of university policies and the impact of factors such as the availability of adequate infrastructures and venture capital financing, but they also focused on personal characteristics of academics that become managers, with special attention to the desire for independence (Shane, 2004).

Recent focus has been made on network activities, internal communication and adhocracy (Gupte, 2007): it adopted the diffusion of innovation perspective as well. RSOs’ potential of technology transfer calls to take into consideration the relation between “tacit and codified knowledge” (Cowan et al., 2000; Powell, Grodal, 2005; Witt, Zellner, 2007). On the one hand, RSOs develop on the market the potential of the tacit knowledge residing in the heads of scientists; on the other hand, the firm knowledge needs to be sustained by an efficient and effective use of internal resources organised in a business model and linked to university rules. Tables 1 and 2 provide a summary of the characteristics of these main perspectives and taxonomies on RSOs.

The contributions of the literature support the assumption that the parent institute plays a key role in the differences among all these perspectives and taxonomies. The university approach towards these particular kinds of firms determines the prevalence of a low, supportive or incubator model of selection. The company orientation towards a more or less narrow approach with its parent institute may determine the prevalence of a venture capital-backed, a prospector or a lifestyle classification.

Our empirical analysis aims at providing a contribution to the literature on RSO perspectives and taxonomies through a focus on the relationship between these companies and their parent institute “interest”.

The notion of interest is usually associated in the management research field with stakeholders’ interests, which are taken into account in the context of inventory management systems (De Vries, 2013). These interests can be either in conflict or common and they need to be identified and addressed to maximize firm performance (Lansiluoto et al., 2013). Stakeholders may act to protect and achieve their interests (i.e. interest-based view), while Rowley and Moldoveanu (2003) argue that interest intensity is the primary condition driving a stakeholder group to act to influence the focal firm: the stakeholder literature highlights a link between interests and action. In general, the theory and research on interest argues that interest plays a key role in performance optimization, because it may contribute as a mental resource. Interest helps actors accomplish challenging goals: various beneficial outcomes (e.g. heightened attention, adaptive goals, learning) are associated to the presence of interest. Therefore, one can say also that interest is someway involved in the management and optimization of self-regulatory resources (O’Keefe, Linnenbrink-Garcia, 2014).

In the literature on RSOs, a definition of “university interest” (UnivInt) does not exist. Recently, Treibich et al. (2013) have focused on the development of “interactions” between RSOs and their parent institute in terms of duration and involvement in a co-production of knowledge, but going beyond the underlying features and consequences linked to a native interest from the university of origin. And Sternberg (2014) focused on the influence of the regional environment as a success factor of RSOs, while Rasmussen et al. (2014) underlined that the interaction between a RSO and its parent institute is complex and often dialectic and argued that the university department context influences the early RSO development process with the local department level playing a role of moderator of general university “support” towards RSOs. Nonetheless, the influence of the local environment on the development path of RSOs was not linked to the presence or absence of “interest”.

Starting from these assumptions, we adopt in this paper a meaning of interest similar to the utilitarian concept and pragmatic perspective generally used in strategic management theory and research when considering the “stakeholders’ interests”. As a consequence, we associate “university interest” to the positive and useful consequences for RSOs performance and optimization of self-regulatory resources. Actually, the results carried out from 155 RSO founders exhibit that the presence or absence of UnivInt towards a RSO initiative has consequences on the company orientation.

Tableau 1

Summary of the main perspectives on RSO suggested in recent literature

Summary of the main perspectives on RSO suggested in recent literature
Source: Authors’ elaboration

-> Voir la liste des tableaux

Methodology

In order to identify and ascertain the intensity of the relation between a RSO and its parent institute, we aimed at investigating the importance and the effective role of the university involvement in a RSO initiative. The building of a sample of RSOs in the Italian context demonstrated that the explicit differences existing with the university are hardly sufficient enough to discriminate the various situations. Consequently, we decided to supplement them with the entrepreneurs’ perspective through a specific questionnaire. A dedicated section was focused on the relationship between RSOs and the parent institute, meaning whether or not the university showed any interest towards the RSO initiative, how this was perceived by company founders and which consequences aroused from the presence or absence of the parent institute involvement.

Tableau 2

Summary of the main taxonomies on RSO suggested in recent literature

Summary of the main taxonomies on RSO suggested in recent literature
Source: Authors’ elaboration

-> Voir la liste des tableaux

One of the main problems was to identify the actual number of RSOs founded in Italy. In order to define the sample and evaluate the quality of the survey answers, the first step was to look at technology transfer offices and university websites for a list of RSOs and to verify the completeness and updating of this list. Since a large definition of RSOs may be considered, including also companies not participated with the university, this list had to be completed with the Italian science park and incubator tenants list. A final problem was due to the fact that science parks and incubators do not make any difference between RSOs and start-ups, which refers to firms not created by university personnel and therefore not linked to the academic world. Telephone and e-mail contacts with university staff as well as science park and incubator personnel were pivotal in filling this gap and in excluding start-ups from the final list.

The universe of RSOs identified in Italy was 394 firms[3]. A questionnaire was sent to these companies and the response rate was 39.5%[4].

The questionnaire was sent by e-mail to RSO founders between January and June 2008[5]. It was specifically divided in the following sections: a) general characteristics of the RSO; b) funding sources; c) university and RSO relationship; d) incubator/science park and RSO relationship[6]; e) patents; f) industrial partnership[7]; g) geographical location of the RSO.

In order to investigate the consequences of the presence or absence of the parent institute interest towards a RSO initiative, a cluster analysis has been firstly implemented in order to identify the various university strategies at stake vis à vis RSOs and in which extent they might determine specific groups of RSOs. Then we used a factor analysis to fine tune the clusters characterization and to isolate and detect the more influential variables and figures shaping the RSOs development. The various questionnaire questions were clustered into homogeneous groups and it was ascertained whether these groups were different from or similar to one another. A table was contrived with qualitative variables transformed into quantitative ones, following the order of questions provided in the questionnaire. These variables were organized into groups by means of the cluster analysis methods (Everitt et al., 2001). Cluster analysis approach allows for the identification of groups of objects with small within-cluster variation for discriminating variables and high variation between clusters. The list of variables used for the cluster analysis is provided in Annex A. K-means[8], one of the most widely used partitional clustering techniques for cluster analysis (Ayramo, Karkkainen, 2006), was applied. K-means clustering is a prototype-based technique that helps to define a user-specified number of clusters (K), which are represented by their centroids (Tan et al., 2005). We used Euclidean (L2) distance and the sum of the squared error (SSE) metric. One of the main problems is to identify the optimum number of clusters. The use of K-means implies the specification of the number of clusters as an input to the algorithm (Gray, Neuhoff, 1998). In order to choose the more efficient of the grouping divisions, the Calinski/Harabasz pseudo-F index was additionally used (Calinski, Harabasz, 1974; Milligan, Cooper, 1985).

Finally, a factor analysis was performed on the main variables that influenced the differences between the two Clusters identified by the cluster analysis in order to make clear the specific variables differentiating the RSOs participating to each group and the characteristics and the consequences of the university involvement (i.e. presence or absence of UnivInt). Principal component extraction factors was the method chosen and the Kaiser-Meyer-Olkin was the measure of sampling adequacy. This last one was quite low, but acceptable (see Annex B). To determine the number of components the latent root criterion (Eigenvalue > 1.0) was applied: a three-dimensional solution explaining 47% of the variance was obtained.

Survey results

This section analyses first of all the contents of the 155 questionnaires fulfilled by Italian RSO founders. The main result confirms that, in our sample, various groups exist and are characterized by different situations. A specific focus is more precisely on the two sub-groups of RSOs that fostered more interest from their parent institute (100 questionnaires) and the ones without their parent institute interest (42 questionnaires)[9].

Secondly, the data available have been treated through a cluster analysis approach in order to verify whether some specific arrangements of RSOs with university interest emerge or not and how they are shaped. Effectively, this process enabled us to characterize, at first, two sub-groups of RSOs. Finally, a factor analysis was run on the main significant variables identified by the cluster analysis: this final step was required in order to ultimately identify the specificities and orientation of RSOs with and without the parent institute interest.

The parent institute “interest” towards a RSO initiative: a preliminary outlook

In order to provide an answer to our investigation about the consequences of the presence or absence of university interest on the firm development and efficiency, we firstly focused on the link between a RSO and the parent institute: “interest” towards the spin-off initiative, the time required for administrative procedures, the main difficulties and impediments encountered, the backing and resources expected from the parent institute. Actually, section C of the questionnaire included questions that aimed at highlighting how many companies benefited from receiving positive attention from their university and at understanding the main RSOs’ reasons for looking or not for their parent institute involvement. We complemented this first evidence with additional answers from the questionnaire: this approach enabled to identify the general characteristics and orientation of RSOs with and without UnivInt.

In a nutshell, the results highlighted the emerging attention of Italian universities towards the RSO phenomenon with a high percentage of interest (65% of the questionnaires) and a low level of absence of UnivInt towards a spin-off initiative (27%). The main reasons for searching for a lasting relationship with the parent institute were linked to the benefits deriving from university contacts and the use of university resources in terms of infrastructures and personnel. Otherwise, the companies without any relationship with their parent institute underlined the absence of real benefits, meaning a sort of prejudice linked to a lack of business culture on the part of the university. A positive interest led to the parent institute support for the creation of the RSO, in terms of participation in the capital of the company and/or of other forms of support and promotion (i.e. aid from technology transfer offices, the possibility to use resources and laboratories, the use of the university logo, the possibility to be hosted in the university structures and the employment of the university’s PhD candidates in the company). In general, the various forms of support and promotion provided by a parent institute are set in the specific rules for spin-off creation issued by the university itself (Salvador, 2009). As a consequence, a lasting relationship is created between the RSO and its parent institute: this may provide mutually beneficial consequences in terms of RSO performance and university prestige. In the case of lack of interest on the part of the parent institute, the RSO project is created with the autonomy and independence of the founders. The main characteristics of the two kinds of companies - RSOs with and without their parent institute interest - are demonstrated in the following section as well as the reasons for company creation, the solutions to the lack of managerial competences, and the relationship with banks. The results highlighted that most of the companies with patents and the companies that benefited from regional, national and European grants are RSOs with UnivInt. Nonetheless, a strongly higher level of turnover of RSOs with interest compared to RSOs without their parent institute interest was not registered. Interestingly, companies without UnivInt have a deeper independence attitude compared to the ones with interest. Notwithstanding, answers suggest that they suffer more from a management knowledge gap and from difficulties with banks.

General characteristics of RSOs: year of creation, form of society, capital, turnover

Considering the overall structure of our prototypical sample, the industry sectors showed a strong prevalence of the ICT sector (33%) followed by the biopharmaceutical one (25%). This result is coherent according to the influence exerted by the Internet revolution (Benghozi et al., 2009) and it is in line with evidence found in the literature (Mustar, 1997; Shane, 2004; Gupte, 2007; Clarysse et al., 2007).

Looking at the year of creation, the RSOs of our sample are relatively young firms: most of the companies analysed were established between 2003 and 2007. A comparison between RSOs with and without interest from their parent institute revealed that the number of companies with UnivInt increased sharply between 2003 and 2007. Nonetheless, also the number of RSOs without interest from the parent institute increased between 2005 and 2007. This result is in line with the growing number of RSOs in Italy (Netval, 2013) and the increasing issuing of university rules for these companies (Salvador, 2009).

These firms are most of all limited liability companies; some of the few companies in the form of joint-stock companies benefited from the interest of a university. Two thirds of the companies are service oriented while only one third is product oriented: looking at companies with parent institute interest, it emerged that 38% are product companies, while 62% are service companies. On the other hand, RSOs without UnivInt are 24% product companies and 76% service companies. Less than 20% of the founders left the university position to work full time in the spin-off firm. Furthermore, few companies hold patents: 63% are RSOs with UnivInt, while 37% are RSOs without UnivInt.

Initial and present capital is low (between 10,000 and 20,000 euro) and reveals low increases in the capital registered. A comparison between RSOs with UnivInt and RSOs without it, revealed the absence of strong differences both in the initial as well as in the present capital of the company.

The questionnaire results confirmed the importance of personal and family capital as a source of financing (Roberts, 1991) as well as the availability of public funds. Companies that benefited from regional, national and European grants, and won regional competitions are most of all RSOs with UnivInt.

A comparison of the company turnover in the year 2007 revealed a subtle difference between the two samples of RSOs: in general, RSOs supplemented by UnivInt did not register a strongly higher level of turnover compared to RSOs without their parent institute interest.

The most important reasons for looking for the creation of a relationship between the company and its parent institute were linked to the opportunity to maintain contacts with the parent institute and the possibility to use university resources such as students and laboratories. The most cited disadvantage was bureaucracy. In case the university interest led to university participation in the capital of the company (84%), administrative procedures required in general from three to six months. In case of lack of any relationship with the university, the main reasons were linked to lack of real advantages and a slack of dynamism by the university.

The managerial knowledge gap, its roots and consequences

Main motivations for company creation have been the willingness to use research results and to work independently. Questionnaire results highlighted that companies with UnivInt have chosen as main reasons for company creation “use research results” (43%) and “move from idea to market” (27%). Alternatively, RSOs without UnivInt have mainly chosen “desire for independence” (31%), “lack of job in university” (20%), “use research results” (21%).

The literature on this particular kind of firms has highlighted a lack of managerial and business competencies (Shane, Stuart, 2002; Lockett et al., 2003; Heirman, Clarysse, 2004; Wright et al., 2004; Shane, 2004; Iacobucci et al., 2011). The existence of a lack of business competence is underlined by the fact that only 21.29% of the sample chose “no lacks” as first answer. The solutions adopted for the knowledge gap revealed by the present survey were mostly “self-training” and “aid by the incubator”. All but few of the respondents that chose “no lacks” are companies with UnivInt as well as most of the companies that chose “aid from industrial partner” as the main solution. On the contrary, aid from an “external manager” and “self-training” have been chosen more by RSOs without UnivInt (17% and 38% respectively).

The knowledge gap has one of its main consequences in the relationship with venture capitalists and banks. Only 11% of the sample is participated by a venture capital. While more than 50% of the sample affirms to have no troubles with banks, the main difficulties are characterized by lack of competence by spin-offs (15%) or by banks (15%). Two thirds of respondents that chose “no difficulties” are companies with UnivInt, while only one third are companies without UnivInt.

Two clusters structuring RSOs with and without UnivInt

The comparison between RSOs with interest and the ones without it highlighted interesting but somewhat contradictory findings. More specifically, even if most of the companies with patents and those companies that benefited from various forms of grants are RSOs with UnivInt, they did not exhibit a strongly higher level of turnover compared to RSO without their parent institute interest. Interestingly, these last ones suffered more from a management knowledge gap and from difficulties with banks.

Consequently, these results fostered a deeper analysis in order to characterize these distinctive features and understand whether some specific organisations emerge or not in such structure variety. To this aim, we chose to apply a cluster analysis in order to test the existence of groups with similar features amongst the sample of 155 RSOs and verify whether or not the UnivInt someway influenced the groupings. Actually, confirming our assumptions, the analysis made clear the emergence of two main clusters of RSOs, exhibiting specific characteristics. Interestingly, the main specificities of the two clusters not only reflect the similarities and the differences that emerged in the descriptive comparison between RSOs with and without UnivInt but they also deepen the emerged features and enable to better understand the specificities of RSOs with and without UnivInt.

The cluster analysis results

The results of the questionnaire respondents division into several groups demonstrate that the best partition is the one that differentiates the RSOs into two groups (cf. results for Calinski/Harabasz pseudo-F index). No correlation problems emerged (see Annex C).

Table 3 describes the analysis of variance (ANOVA) results for the grouping into two clusters. ANOVA highlights the variables that most contributed towards the identification of the two clusters: the reasons for company creation (var06&var07) and relationships with banks (var19) were the most influential.

More precisely, Table 4 provides a description of the mean values of all the variables in the two groups.

Clusters description

According to the K-means partitional clustering method applied, Cluster 1 comprises 65 questionnaire respondents and Cluster 2 comprises 73 questionnaire elements. One might underline that the size of these two clusters is quite similar. Looking at Tables 3 and 4, the key features of the two groups can be identified. Some specific variables do in fact characterize each group more than others. Both in Cluster 1 and in Cluster 2 the main reasons for the company creation, the solution adopted for lack of managerial competencies and the relationship with banks are the key characteristics (variables 06, 07, 10 and 19).

First of all, some similarities are identifiable. Both the two Clusters are characterized by a similar distribution of RSOs created in the last five years or before; limited liability companies (srl) and joint-stock companies (SpA) are equally distributed in the two clusters; in both of the two Clusters there is a prevalence of low capital companies as well as few increases in capital; the variables linked to the location (var25-27) have underlined a general prevalence of indifference as a verdict on company location as well as an overall predominance of no aid from the Region, with a subtle prevalence of aid in Cluster 2. Slight differences have been observed in the influence of var22 (hospitality in a science park-incubator).

Yet, some differences exist: Cluster 1 shows a predominance of companies located in the North and in the Centre of Italy and a prevalence of product companies with a number of shareholders not exceeding five persons. Cluster 1 comprises all the motivations for company creation about going from an idea to the market and personal prestige. The solutions to lack of managerial competencies (var10) highlighted a prevalence of no lacks or aid from the industrial partner. Linked to these characteristics, the relationship with banks (var19) underlined a lack of difficulties in Cluster 1. Concerning the market (var11), a prevalence of importance of the international market is observable. A predominance of regional competitions (Start-cup) and venture capital financing is identifiable as main source of financing. Cluster 1 highlighted also a higher presence of patents and industrial partnerships compared to Cluster 2.

Finally, looking at Cluster 2, it includes RSOs located in the North and in the South of the country and it is characterized by a prevalence of limited partnerships (sas) as well as service companies with a number of shareholders over five. Nearly all the few founders that left – or did not have – the university position are included in Cluster 2 that is also characterized by less interest from the parent institute (var21). Looking at the main motivations for company creation, Cluster 2 includes all the choices for company creation concerning lack of job in the university and desire for independence, as well as self-training, aid from the incubator and an external manager as the main solutions to lack of managerial competencies. The relationship with banks highlighted the presence of difficulties due to lack of competencies by spin-offs and by banks. About the market, the national market is predominant in Cluster 2 and the distribution of the sources of financing highlighted a prevalence of personal and family capital, bank loans, regional, national and European grants.

Tableau 3

ANOVA results for the grouping into two clusters[10]

ANOVA results for the grouping into two clusters10

-> Voir la liste des tableaux

Tableau 4

Summary statistics: mean values of variables by cluster

Summary statistics: mean values of variables by cluster

-> Voir la liste des tableaux

To sum up, the cluster analysis identified two groups. These two clusters mostly confirm and share many characteristics quite similar to the ones that emerged in the descriptive comparison between RSOs with and without UnivInt. Therefore, according to the above description, we suggest to refer to them as “Open-oriented RSOs” (Cluster 1) and “Autonomous-oriented RSOs” (Cluster 2). Figure 1 provides a summary of the main specificities of these two clusters.

Figure 1

-> Voir la liste des figures

Three influential factors

The cluster analysis highlighted the presence of two main Clusters. Then, it turned out necessary to complement the identification of the two groups, precisely featuring the orientation of RSOs with and without their UnivInt and the effective meaning of more and less open-oriented RSOs. The factor analysis made clearer the specific components of the variables distinguishing the two clusters. More precisely, it contributed to highlight three main Factors, identified as follows (Table 5).

Factor 1 includes the items pertaining to the solutions adopted for lack of managerial competencies and the relationship with banks: therefore it can be labelled “Managerial competencies”. Factor 2 groups the items pertaining to product or service orientation and the presence of patents. Therefore, Factor 2 can be labelled “Company orientation”. Finally, Factor 3 includes items that show a negative link between company location and the choice of some shareholders of leaving the parent institute on the one hand and the university interest towards the RSO initiative on the other hand. Questionnaire results showed that most of the RSOs are located in the North of the country as well as the few founders that left the university position. It revealed that if company founders maintain the university position there is more probability to attract interest from their parent institute. As a consequence, Factor 3 can be labelled “Founders and university choice”.

Tableau 5

Factor analysis results

Factor analysis results

-> Voir la liste des tableaux

To sum up, these results enable to identify the orientation of RSOs with and without UnivInt. The importance of “managerial competencies” highlighted by Factor 1 confirms the peculiarities of the two clusters: absence of lack of managerial competencies and absence of difficulties with banks were prevalent in Cluster 1, while lack of business knowledge and problems with banks because of lack of competencies characterized Cluster 2. Looking at “company orientation” (Factor 2), product companies (Cluster 1) are more used to hold patents compared to service companies (Cluster 2). Finally, about founders’ “choice” (Factor 3), if RSO founders do not leave the university position there is more probability to have a greater interest by the parent institute (Cluster 1).

Discussion and concluding remarks

Recently, a great attention has been devoted to the third mission of universities (Chapman et al., 2011; Philpott et al., 2011; Goldstein, Glaser, 2012; Carree et al., 2012; Treibich et al., 2013; Algieri et al., 2013). Looking specifically at the RSO phenomenon, two main dimensions have been underlined in the literature: the persistence of close links with the parent institute and the importance of the degree of support that a RSO receives from its university for company success (Westhead, Storey, 1995; Mustar, 1997; Steffensen et al., 1999; Chiesa, Piccaluga, 2000; Shane, 2004; Rothaermel, Thursby, 2005; Treibich et al., 2013; Rasmussen et al., 2014). Our survey confirmed the importance of this support, but it also highlighted some specificities of RSOs with and without their UnivInt. According to our results, RSOs that aroused interest from their university are more product oriented and have more patents compared to RSOs without their UnivInt. Furthermore, RSOs with UnivInt have less difficulty with banks and fewer knowledge gaps compared to RSOs without UnivInt. Notwithstanding, no significant differences were observed in the capital as well as in the turnover.

Our research highlighted further interesting findings. The two groups of RSOs identified by the cluster analysis exhibit some clear similarities but also several specificities. A first Cluster, named Open-oriented RSOs, includes more product companies with patents, internationally oriented, located in the North and the Centre of Italy, with few shareholders with university position and linked to their parent institute interest. These RSOs have been founded for moving from an idea oriented strategy to the market and – in some few cases – for founders’ personal prestige. They have more probabilities to be financed from venture capitalists and have few difficulties with banks as well as no lack of managerial competencies as an industrial partner might sometimes provide them. A second Cluster, named Autonomous-oriented RSOs, consist of more service companies without patents, nationally oriented, located in the North and South of Italy, with many shareholders without a university position and with less interest from the parent institute. These companies have been founded because of lack of jobs in the university and for the desire of independence. They are financed considerably more by personal and family capital, bank loans and regional, national or European grants. These RSOs face difficulties with banks because of lack of competencies as well as lack of managerial competencies filled by self-training, aid from incubator or from an external manager.

Finally, three main significant factors confirm the importance of managerial competencies, the product or service attitude and the link between RSO founder’s position in university, the parent institute interest and the company location.

The first key role is played by choices and the second role is given by managerialcompetencies. Choices and managerial competencies influence the company’s orientation. These three components taken together are the pillars of the cluster analysis grouping and should be taken into account for RSOs policy strategy. Going into details, choices mean reasons for company creation and founder personal choices: the cluster analysis highlighted a positive link between moving from idea to market and RSO founder with a university position on the one hand, and the parent institute interest on the other hand (Cluster 1). Otherwise, the cluster analysis highlighted a link between lack of job in the university and desire for independence with less interest from the parent institute (Cluster 2). Therefore, it seems that RSOs founded by university staff have more probability to be judged as interesting initiatives by their university and consequently receive support from their parent institute. It would be useful to understand why: further research could highlight whether this interest is linked to university prestige or to a desire for more revenues. This has consequences on company managerial competencies. In fact, a knowledge gap is observable more in RSOs not founded by university personnel (Cluster 2). Again, possible reasons of this finding can be identified in the more easiness for RSOs supported by their parent institute to fill the knowledge gap through aid from university structures as well as meetings, tutorship, consultancy services, networking with colleagues and not only through self-training. The lack of managerial and business competencies has consequences on the company’s orientation - product or service with or without patents - meaning that entrepreneurial competencies have a pivotal importance. Notwithstanding, the overall performance of Italian RSOs seems to be lower compared to other start-ups (Salvador, 2011a). It is also true that spin-off founders with a university position have to balance a trade-off between the time devoted to teaching and research and the daily management of the firm. This could be one of the possible reasons for the absence of marked differences in turnover between RSOs more or less open-oriented.

Therefore, from these considerations it can be assumed that spin-off founders without a university position should be more encouraged and aided in fostering the potentialities of their codified and tacit knowledge. This means first of all that they need to be helped in filling the knowledge gap of managerial and business competencies, that is one of the main weaknesses of RSOs underlined in the literature (Shane, 2004; Wright et al., 2007, 2004; Mustar et al., 2008; Clarysse et al., 2011; Iacobucci et al., 2011). In order to fill this knowledge gap, entrepreneurship education could be a crucial factor in fostering not only entrepreneurial attitude but also entrepreneurial competence and growth of this particular kind of firms (Gorman et al., 1997; Pittaway, Cope, 2007; Fayolle, 2008; Bureau et al., 2012). RSOs have some peculiarities that need to be valorized in order to become successful. To this aim, key actors that could help RSOs are not only universities but also business schools (Wright et al., 2007) in partnership with science parks, incubators and technology transfer offices. The “brand name” of science parks and incubators could be an important signal of credibility (Salvador, 2011) and the proactive role of technology transfer offices should not be underestimated (Clarysse et al., 2007; Jain, George, 2007). Therefore, if the cluster grouping and the factor analysis results are reasonable, and if the overall performance of these companies is not brilliant (Mustar et al., 2008; van Geenhuizen, Soetanto, 2009; Bathelt et al., 2010; Visintin, Pittino, 2014), as a consequence it is important to focus more on autonomous-oriented RSOs. This focus could help in understanding why in general RSOs do not show a rapid growth with consequent suggestions for improvement. This means first of all that autonomous-oriented RSOs should receive more interest from their parent institute – and encouraged to fulfil this aim even if they have an independence attitude – and should have the possibility to fill the knowledge gap. Secondly, the time devoted to the spin-off initiative by founders with a university position should be strictly balanced: neither the research and teaching duties nor the company management have to be neglected to the detriment of one over the other. In other words, personal prestige and self-celebration should not overcome the enlargement of the scientific perspective and wealth creation. Last but not least, business schools in partnership with science parks, incubators and technology transfer offices should have the chance and should be encouraged to play a more active role not only in filling the entrepreneurship education gap but also in fostering the potentialities of the codified and tacit knowledge held by this particular kind of firms for company creation and employment opportunities.

Nevertheless, this empirical investigation is not without limitations. First of all, the population of RSOs did not cover the universe. Secondly, the study is limited to the Italian context and does not attempt at providing a cross analysis with other European countries. Thirdly, the analysis relies on data covered within a given time period. Nonetheless, these limitations and potential biases do not affect negatively the significance of the findings, because the sample was more than representative, Italy is an interesting case-study about RSOs and the survey covers a period of RSOs growing at EU level. Furthermore, the originality of this empirical investigation is given by the provision of original data taken directly from “effective” RSOs through a questionnaire investigation. Given the lack of reliable official data, this analysis provided primary data and highlighted interesting findings and served to foster a better understanding of the perceptions of Italian RSOs that might stimulate further research. Future research along this line and with a focus on other EU countries could highlight further consequences of the presence or absence of “interest” of the parent institute for a RSO. The focus on the presence or absence of the parent institute interest and its main reasons and consequences may be a key step in order to shed more light and to improve future policy strategies for a particular kind of firm like a RSO.