Innovation networks for regional development - An agent-based simulation approach
Final Report Abstract
The project focused on the integration of industrial dynamics, namely the evolution of innovation networks during an industry life cycle and regional development and growth. Agent-based simulation models (ABMs) allow for such an integration which enormously increases the complexity of the analysis. Accordingly, in the project ABMs are used as the major modelling approach. They are used as a tool to disclose (plausible) causal relationships between firm strategy and the emerging networks. The analysis of network evolution serves as an instrument to test the relevance of attachment and dissolution mechanisms which are implemented in the artificial world of an ABM. ABMs can also be used to develop additional hypotheses with respect to drivers of network evolution and to discover mechanisms which have not been theorized and tested yet. In this sense, the empirical research complements agent-based simulation models and vice versa. An important finding of the modelling exercises is that the impact of regional network relationships depends on technological relationships and the distribution of technological knowledge within different regions. If technological cumulativeness is less important, interregional relationships are less crucial for single actors. If cumulativeness is strong and includes cross-fertilization of different technological fields, interregional relationships are effectively supporting the advancement of the actors’ knowledge bases. Another finding is that following a competence destroying (radical) innovation (typically in early phases of the industry life cycle), collaboration distance increases over time due to codification and diffusion of knowledge. However, collaboration is not beneficial per se if the scope or scale of research activities distract the agents so much that they fail to gain sufficient expertise. In these cases, narrow in-house R&D may be beneficial over broad collaborative R&D. Cooperation transaction costs are to be expected, distorting the beneficial effects of knowledge sharing. The design of the agent-based models of innovation networks and their role for regional development were informed by insights from the various empirical studies of innovation networks: For the networks that have been reconstructed from publicly funded R&D projects in the German automotive industry, e.g. structural as well as individual and dyadic covariates are relevant drivers: The formation of triadic structures could be observed; spatial proximity between firms increases the propensity to cooperate as well as experience; firms with high levels of absorptive capacity tend to be more often involved in networks. Internal R&D has not become obsolete but is a prerequisite to benefit from sources of collective knowledge. Furthermore, location decisions matter because in later stages of the industry life cycle, there is a preference for spatially co-located partners. Consequently, re-location or the opening of a subsidiary in a region where the respective knowledge is bound can be an option for a firm which seeks to gain access to a specific network or specific knowledge-base. Clusters might emerge from this behavior. Besides the geographical distance also the specific impact of technological distances matters: The preference to select partners that have a somewhat similar knowledge-base indicates that new knowledge should be related to the existing knowledge which makes it easier to be understood. Finally, firm size needs to be considered in the simulation models: Typically, large and established firms have more available resources to be informed about subsidies as well as experience to successfully apply for funds, to lobby or to pay consultants. In contrast, small firms are typically the envisaged target of innovation support initiatives. However, the analysis of the automotive network shows that firm size is not a significant factor for the cooperation partner selection.
Publications
- (2012), Innovation Networks, in: Dietrich, M. and Krafft, J. (eds.), Handbook on the Economics and Theory of the Firm, Edward Elgar, Cheltenham, UK
Buchmann, T. and Pyka, A.
- (2013), Cutting Back Models and Simulations, in: Andreas Tolk (ed.), Ontology, Epistemology, and Teleology for Modeling and Simulation - Philosophical Foundations for Intelligent M&S Applications, Springer, Berlin, Heidelberg 2013, 141-156
Deichsel, S. and Pyka, A.
- (2013), Technological Competences and Regional Innovation Networks: Measurement and Visualization with Patent Data, in: Morone, P. (ed.), Knowledge, Innovation and Internationalisation, Routledge, NY
Hartmann, D., Pyka, A. and Ebersberger, B.
- (2013). The European aerospace R&D collaboration network. FZID discussion papers No. 84
Guffarth, D. and Barber, M.J.
- (2014), Against the one-way-street: analyzing knowledge transfer from industry to science, Journal of Technology Transfer, Vol. 39 (2), 219-246
ier, H. and Pyka, A.
(See online at https://doi.org/10.1007/s10961-011-9226-7) - (2014), R&D and Knowledge Dynamics in University-Industry Relationships in Biotech and Pharmaceuticals: an Agent-Based Model, International Journal of Biotechnology, Vol. 13, 137-179
Triulzi, G., Pyka, A. and Scholz, R.
- (2014), Technological progress and effects of (supra) regional innovation and production collaboration. An agent-based model simulation study., Proceedings of the IEEE International Conference on Computational Intelligence for Financial Engineering & Economics (CIFEr), 2014, p 357 - 364
Vermeulen, B. and Pyka, A.
(See online at https://doi.org/10.1109/CIFEr.2014.6924095) - The Effects of Supraregional Innovation and Production Collaboration on Technology Development in a Multiregional World: A Spatial Agent-Based Model Study, Cellular Automata, 2014, pp. 698-707
Vermeulen, B. and Pyka, A.
(See online at https://doi.org/10.1007/978-3-319-11520-7_74) - (2015), The Evolution of Innovation Networks: The Case of a Publicly Funded German Automotive Network, Economics of Innovation and New Technology, 24 (1-2), S. 114-139
Buchmann ,T. and Pyka, A.
(See online at https://doi.org/10.1080/10438599.2014.897860) - Agent-based modeling for decision making in economics under uncertainty, Economics, No. 2015-45, 2015
Vermeulen, B. and Pyka, A.
(See online at https://doi.org/10.5018/economics-ejournal.ja.2016-6) - Product differentiation under bounded rationality.", Proceedings of the 18th Asia Pacific Symposium on Intelligent and Evolutionary Systems, 1:695–709, 2015
Vermeulen, B., La Poutre, J.A., de Kok A.G. and Pyka, A.
(See online at https://doi.org/10.1007/978-3-319-13359-1_53) - Taking the first step - what determines German laser source manufacturers' entry into innovation networks? International Journal of Innovation Management, Vol. 19 (5), 2015
Kudic, M., Pyka, A. and Günther, J.
(See online at https://doi.org/10.1142/S1363919615500504) - Vertical governance change and product differentiation under decreasing component costs, Journal of Economic Dynamics and Control, Vol. 57, 2015, pp. 65-76
Vermeulen, B., Huisman, K.J.M. and de Kok, A.G.
(See online at https://doi.org/10.1016/j.jedc.2015.05.010)
