Version classiqueVersion mobile
OpenEdition Books

The role of German universities in a system of joint knowledge generation and innovation

 | 
Mirja Meyborg

7. Hypotheses and Methodology regarding the Analysis of the German Universities

Texte intégral

1The basic concept of this work is to picture the role of all German universities regarding joint knowledge generation and innovation. Thus, in order to give insight into this construct, this chapter now illustrates all relevant hypotheses of the PhD thesis and simultaneously indicates the statistical tools that are used to underpin the empirical results. Thereby, this chapter is divided into two parts. First of all, a broad over view of the overall publication and patenting activity of all German universities over the past ten years is illustrated. Thus, it illustrates the German universities as knowledge generators by means of scientific publications, and secondly, it illustrates the role of the German universities in innovation activities through patent applications. In this context, the institutional distribution of the German university cooperation partners is also shown and points to the most important university-university linkages, but also highlights the emerging collaborations between universities and enterprises concerning the scientific publications. Regarding the patent applications, the highly increasing co-applicant networks with enterprises are especially highlighted. Thereby, it makes use of the SNA in order to identify which German universities are nowadays highly involved in close network collaborations and points to those that have developed best over the past ten years in this regard.

  • 44 This broad spatial analysis is only based upon publication data as the data set of the patent appli (...)

2A second cornerstone consists of an extensive spatial distribution measure which illustrates whether distance patterns still matter within the German university knowledge networks as it did years before. First of all, it is shown to what extent proximity still matters by considering all co-authors that are within the 1,000 km corridor. Second, by means of the cluster analysis, all German university co-authors are considered, illustrating how all German university co-authors are spatially distributed, either being regional or supra-regional located. Finally, the spatial analysis makes use of the five established groups of countries, in order to geographically show how the co-authors of the German universities are spread worldwide. As it is done for the first part, this one also highlights the institutional distribution of the German university cooperation partners and points again to the four different university interactions, namely to all linkages to universities, research institutes, enterprises and to themselves, of course with regard to the spatial dimension.44

7.1. Knowledge Generation, Innovation and Collaboration

3This subchapter now illustrates which hypotheses are to be explored and answered in the course of this PhD thesis with regard to the knowledge generation, innovation and collaboration potential of the German universities in general.

4First of all, it is about the publication and patenting activity of the German universities which has highly increased over the past ten years. In addition, it is not only the number of publications and patents that has enjoyed a great increase, but also collaborations for knowledge and innovation have become a lot more important. As mentioned before, nowadays, most countries are faced with a fast expansion of knowledge-based industries and activities whose basic raw material is simply new knowledge (Karlsson and Zhang 2009). Regarding the increase of scientific publications, Godin and Gingras (2000) have illustrated for Canada that the presence of universities in scientific papers has increased from 75% in 1980 to 81. 9% in 1995. Katz and Hicks (1997) have shown that in the UK the percentage papers with at least one university in it have risen from 59.2% in 1981 to 64. 3% in 1994. Further, with regard to the rise of patent applications by universities, a vast majority of nowadays literature confirms that the use of academic knowledge is decisive for technological change, innovation and growth by means of new theoretical insights, techniques, and expertise (Mansfield 1998, Cohen, Nelson and Walsh 2002). Besides, Rosenberg and Nelson (1994) have shown that universities are highly important factors in the development of major innovations. Hence, it is not surprising that the university which is a centre of knowledge and technology generation becomes increasingly essential for innovation and economic growth, too. Last, it is also common sense that knowledge generation and the innovation process itself is not a result of isolated agents either. Thus, collaborating for knowledge has also received broad support in the literature. Archibugi and Coco (2004) have, for example, pointed to the fact that collaborations might help to handle the ever more complex and therefore costly scientific problems.

5In order to proof or disprove above findings regarding the publication, patenting and collaboration behaviour of the German universities, the following hypothesis is derived as follows:

6Hypothesis 1: Scientific publications and patent applications have highly increased over time and collaborations for knowledge and innovation have become much more important.

7Hypothesis 1 provides a first overview of the overall publication and patenting activity of the German universities and points subsequently to the highly growing collaborations of the German universities. Thereby, the knowledge linkages are illustrated by the increased number of co-authored publications, especially in contrast to single-authored publications. It is further illustrated how the German universities have developed from merely inventors to applicants themselves, pictured through the increasing number of patent applications. In this context, the rising collaboration potential concerning patents is pictured through the increasing number of co-applicants.

8Besides, not only the heterogeneity in the landscape of knowledge generation has highly increased over the past years, but also the universities' former role as knowledge producer has shifted towards a role as knowledge mediator. Thus, universities have highly intensified their collaborations with other universities, public research institutes and enterprises' R&D divisions (Godin and Gingras 2000). In this context, among all possible collaborations, university-university linkages are still and by far the most important ones, but university-industry interactions have developed in a more dynamic way (Stephan 1996, Mansfield and Lee 1996 and Schartinger et al. 2002). Jaffe (1989) has, for example, already discovered by the end of the 1980s that a greater number of patents are generated when business is located in close proximity. Becker (2003) has stated that several studies have found out that joint research with universities increases the probability of firms to be engaged in the development of new products and technologies. Mansfield and Lee (1996) have further illustrated that universities cited by firms tend to be the leading generators of new fundamental knowledge. Finally, with respect to the co-authored publications, a vast majority of nowadays literature confirms that the use of academic knowledge is decisive for innovation and growth by means of new theoretical insights and expertise of a kind that enterprises find difficult to provide themselves with (Mansfield 1998, Cohen, Nelson and Walsh 2002).

9With respect to these concerns, the following hypothesis is developed as follows:

10Hypothesis 2: The role of the German universities has changed from merely knowledge producers towards knowledge mediators, which leads to an increasing importance of universities as a central node for knowledge and innovation transfer.

  • 45 In this context, knowledge institution refers not only to universities but also to research institu (...)

11First of all, an illustration of the institutional distribution of the German university co-authors is presented. Thus, it is shown how the German universities' linkages to other institutional partners (themselves included) have developed over the past ten years, and to what extent they have highly increased not only the important linkages to 'knowledge institutions45, but also to industry partners. In terms of patent applications, it is rather shown how the co-applicants of the German universities have developed over time, pointing to the ever more increasing importance of industry partners in this regard. As each co-author as well as each co-applicant possesses its particular actor's code, it is possible to demonstrate the development of the institutional distribution of all German university cooperation partners.

  • 46 In the case of patent applications, it is proved whether third-party funds have had a significant e (...)
  • 47 For the publications, the years 2000, 2003, 2006 and 2009 has been taken; in the case of the patent (...)

12Further, it is shown whether third-party funds have had a significant effect on the cooperation activity of the German universities with enterprises46In order to evaluate this concern, an ordinary least squares (OLS) regression model is applied. Thus, it is proved whether third party funds (thirdPF), administrative income (admlncome), investment expenditure (Invest) and the number of professors (Prof) are responsible for the increasing rate of their cooperation activity. The model makes use of panel data of four years, and five time periods respectively47, for all 76 German universities and is constituted as follows:

Pubi/Pati = β0 + β1thirdPFi + β2admlncomei + β3Investi + β4Prof + δEliteUni + δTUUni + δMedUni + δSizeUni+ εi

13where δ is the coefficient of the dummy variable and e-, the disruptive factor. As can be seen from the formula, four dummy variables are integrated in order to proof whether significance lies in the dummy variable itself or not.

14Afterwards, by means of SNA, it is illustrated how the German universities are already linked in the chains of contact, i e. how many distinct cooperation partners exist and how the importance of a university as a mediator has developed over the past ten years. For doing this, the values of normalized degree and betweenness centrality are chosen as they well indicate how central each university lies in the knowledge and innovation networks pictured through co-authorships and co-applicants.

7.2. Proximity Patterns in Times of Globalisation

15The question to what extent these collaborations benefit from the proximity of partners has been a major subject of t h e scientific debate in the field of economic geography and knowledge and innovation economics for decades. Empirical evidence on this issue is manifold. A vast majority of that literature still suggests that the diffusion of knowledge highly clusters geographically, i.e. they have pointed to the fact that knowledge spillovers are rather found within a short distance. Audretsch (1998), for example, has highlighted that the diffusion of knowledge from the university that creates that knowledge to any other institutional partners tends to be spatially restricted. Besides, there are several studies that have further identified localized knowledge spillovers based on patent data, innovation counts or co-authored publications (Jaffe et al. 1993, Anselin et al. 1997 and McKelvey et al. 2003). Last, Zucker et al. (2001) have shown that biotechnology firms are strongly influenced by the location of successful giants in academic research institutions. In contrast, another strand of literature questions the importance of geographical distance and refers to the development of modern ICT that facilitates cooperation among distant partners and enables them to efficiently interact without face to face contacts. In this context, it is video-conferencing or the possibility to share ideas via computer-mediated technology that makes the knowledge generating process less dependent from spatial proximity (Hancock and Dunham 2001 and Wainfan and Davis 2004). Further, empirical findings have shown in the case of university-industry interaction that global cooperation is more common than regional collaborations (Audretsch and Stephan 1996, McKelvey et al. 2003). Arundel and Geuna (2004) have further illustrated that firms interested in codified knowledge, eg. in terms of patents or publications, are less likely to perceive geographic distance a barrier. The same applied to university-university linkages as Archibugi and Coco (2004) have found a dramatic increase in the internationally co-authored publications which has been also facilitated by the strong diffusion of information and communication technologies.

16Being aware of above findings, t h e following hypothesis is set up:

17Hypothesis 3: Spatial Proximity has lost in importance regarding close network collaborations concerning knowledge spillovers between the German universities and other institutional cooperation partners.

18First of all, it is shown to what extent proximity still matters by considering all German university co-authors that are within the 1,000 km corridor as described above. In this context, it is also illustrated how the institutional cooperation partners of the German universities are spatially distributed. For doing this, Spearman's rank correlation coefficient is used to shed light on the distance pattern. Spearman's rank correlation coefficient is a non-parametric measure of correlation, using ranks to calculate the correlation and looks as follows:

19It shows the direction of association between any two variables x and y. If x tends to increase when y increases, the coefficient of Spearman's rank correlation is positive. The other way round, if x is likely to decrease when y increases, the Spearman correlation coefficient is negative (Bosch 1996). Second, by means of the cluster analysis, it is illustrated how all German university co-authors are spatially distributed, either being regionally located or supra-regionally located. Last, an overview of the spatial distribution depending on where the coauthors are coming from is demonstrated in order to show how they are worldwide presented.

20To sum up, chapter seven provides a detailed overview and impression of the development path of the publication and patenting activity of all German universities as well as of their cooperation potential, especially with regard to the institutional distribution of their cooperation partners. This chapter is rounded up with first reflections regarding the spatial distribution of the German university co-authors in general, bringing up first statements regarding nowadays importance of spatial proximity.

Notes

44 This broad spatial analysis is only based upon publication data as the data set of the patent applications is still too low to identify specific behavioural patterns in this regard.

45 In this context, knowledge institution refers not only to universities but also to research institutes as both are highly engaged in knowledge-intensive tasks.

46 In the case of patent applications, it is proved whether third-party funds have had a significant effect on their overall patenting activity.

47 For the publications, the years 2000, 2003, 2006 and 2009 has been taken; in the case of the patents, four time periods have been developed, namely 1999-2000, 2001-2002, 2003-2004, 2005-2006 and 2007-2008.

Table des illustrations

URL http://books.openedition.org/ksp/docannexe/image/237/img-1.jpg
Fichier image/jpeg, 16k

Acheter