This paper analyses the determinants of spatial distribution of foreign technological activity across 96 German regions (1996-2009). We identify foreign inventive activity by applying the ‘cross-border-ownership concept’ to transnational patent applications. The descriptive analysis shows that foreign technological activity more than doubled during the observation period with persistent spatial heterogeneity in Germany. Using a pooled count data model, we estimate the effect of various sources for externalities on the extent of foreign technological activity across regions. Our results show that foreign technological activity is attracted by technologically specialised sectors of regions. In contrast to existing findings this effect applies both to foreign as well as domestic sources of specialisation. We show that the relation between specialisation and foreign technological activity is non-linear and that it is influenced by sectoral heterogeneity. Externalities related to technological diversification attract foreign R&D only into ‘higher order’ regions.

We study how firm innovativeness is related to individual cooperation events and the structure and dynamics of firms’ ego-networks employing a unique panel dataset for the full population of 233 German laser source manufactures between 1990 and 2010. Firm innovativeness is measured by yearly patent applications as well as patent grants with a two year time-lag. Network measures are calculated on the basis of 570 knowledge-related publicly funded R&D alliances. Estimation results from a panel data count model with fixed effects are suggestive of direct innovation effects due to individual cooperation events, but only as long as structural ego-network characteristics are neglected. Innovativeness is robustly related to ego-network size and ego-network brokerage whereas ego-network density reveals some surprising results.

Using a large dataset of 406 subsidized R&D cooperation projects, we provide detailed insights into the relationship between project characteristics and innovation output. Patent applications and publications are used as measures for the innovation output of an R&D project. We find that large-firm involvement is strongly positively related with the number of patent applications, but not with the number of publications. Conversely, university involvement has positive effects on projects’ innovation output in terms of the number of publications but not in terms of patent applications. In general, projects’ funding as measure of projects’ size is an important predictor of the innovation output of R&D cooperation projects. No significant effects are found for the number of partners as (an alternative) measure of projects’ size, for spatial proximity between cooperation partners, for the involvement of a public institute for applied research, and for prior cooperation experiences. We derive conclusions for the design of R&D cooperation support schemes.

Drawing on the systems perspective of innovation processes, this article proposes a conceptual approach for a comprehensive analysis of regional knowledge generation and transfer. Instead of focusing on one single indicator, the approach emphasizes the importance to take multiple channels of knowledge transfer into account. This provides valuable insights into the spatial structure of innovation processes on different levels. We disentangle the innovation process and consider four different layers: i.) publications in peer-reviewed journals, ii.) patent applications, iii.) formal R&D collaboration projects, the iv.) localized input-output relations. Further, we demonstrate the relevance of the „multi-layer approach‟ by applying it empirically to a specific regional innovation system: The Free State of Saxony – a federal state in Germany. We argue that the approach could be a valuable tool to inform policy-makers about knowledge-based regional development strategies.

Systemic innovation theory emphasizes that innovations are the result of an interdependent exchange process between different organizations. This is reflected in the current paradigm in European innovation policy, which aims at the support of collaborative R&D and innovation projects bringing together science and industry. Building on a large data set using project-level evidence on 406 subsidized R&D cooperation projects, the present paper provides detailed insights on the relationship between the innovative success of R&D cooperation projects and project characteristics. Patent applications and publications are used as measures for direct outcomes of R&D projects. We also differentiate between academic-industry projects and pure inter-firm projects. Main results of negative binomial regressions are that large-firm involvement is positively related to pa-tent applications, but not to publications. Conversely, university involvement has positive effects on project outcomes in terms of publications but not in terms of patent applications. In general, projects’ funding is an important predictor of innovative success of R&D cooperation projects. No significant results are found for spatial proximity among cooperation partners and for the engagement of an applied research institute. Results are discussed with respect to the design of R&D cooperation support schemes.

Regarding technological innovativeness, the transformed economy of the former German Democratic Republic (GDR) clearly lags behind the western part of the country. To face this weakness a broad mixture of policy measures was carried out in recent years. Particular attention is drawn to the development of industry concentrations and economic ‘clusters’. However, little is known about the effectiveness of these policy measures regarding how industry concentrations in fact promote innovative performance in Eastern Germany. The present study tries to fill this gap by analysing the relationship between industry concentration in Eastern Germany and regional innovative performance. Our empirical analysis is based upon the number of patent applications of 22 manufacturing industries in 22 Eastern German planning regions. The estimated regression models indicate an inverted-U relationship between the degree of industry concentration and innovative performance. An exceedingly high degree of industry concentration in one region hampers regional innovative output. We discuss policy implications of our findings and give recommendations for future refinement of ‘cluster’-supporting policy schemes in Eastern Germany.

Regarding technological innovativeness, the transformed economy of the former German Democratic Republic (GDR) clearly lags behind the Western part of the country. To face this weakness, a broad mixture of policy measures was carried out in recent years. Particular attention is drawn to the development of industry concentrations and economic ‘clusters’. However, little is known about the effectiveness of these policy measures regarding how industry concentrations in fact promote innovative performance in Eastern Germany. The present study tries to fill this gap by analyzing the relationship between industry concentration in Eastern Germany and regional innovative performance. Our empirical analysis is based upon the number of patent applications of 22 manufacturing industries in 22 Eastern German planning regions. The estimated regression models indicate an inverted U-shaped relationship between the degree of industry concentration and innovative performance. An exceedingly high degree of industry concentration in one region hampers regional innovative output. We discuss policy implications of our findings and give recommendations for future refinement of ‘cluster’-supporting policy schemes in Eastern Germany.