45th Congress of the European Regional Science Association: "Land Use and Water Management in a Sustainable Network Society", 23-27 August 2005, Amsterdam, The Netherlands
Usually, analyses of knowledge spillovers, if not relying on aggregated data, are based either on surveys conducted with enterprises or on surveys conducted with research establishments. Comparative case studies on micro level that include both groups are rather the exception. Therefore the knowledge transfer mostly can be illustrated just for one of these groups. Moreover knowledge and information rarely are differentiated. The set of data used in this paper allows to overcome these weaknesses. Based on 23 innovation networks located in the eastern part of Germany, the knowledge and information transfer between almost 700 participants, which interacted during a period of 5 years, can be observed. Following the pattern of regional systems of innovation (RIS) within the dataset the distinction of certain groups of participants is arranged (e.g. manufacturing enterprises, service enterprises, universities, non-university research establishments). Their uniform and common reference system - the respective regional innovation network – can be seen as additional quality of the data. The first part of the paper focuses on the determinants of knowledge spillovers within these innovation networks. It is analyzed, in what respect the co-operation experiences and in particular the network experience of the participants have a relevance regarding the knowledge transfer. Beyond that it is examined whether network characteristics (e.g. the coherence of the network on the whole; strength of ties in detail) affect the knowledge transfer. It is also examined whether intensive contacts affect only the transfer of knowledge, or whether the intensity of contact equally shape the information flow. Finally it is analysed, if division of labour is connected with the range of knowledge transfer. In the second part of the paper empirical results are presented that demonstrate the central role played by public research institutions in the process of knowledge transfer. The results indicate that universities are adding most information and most knowledge within the networked process of innovation. The winners of knowledge exchange – considering absolute as well as relative profits – are the manufacturing enterprises. Further the results confirm the assumption that public research holds an "antenna function” (boundary spanning function) for the enterprises due to its integration into the international science community.