Channels of R&D Spillovers: An Empirical Investigation

R&D spillovers are, potentially, a major source of endogenous growth in various recent "new growthJ theory" models. The purpose of this paper is to investigate empirically the effectiveness of various channels of R&D spillovers. The analysis is based on a survey conducted among 358 Swiss R&D executives representing 127 different lines of business, mainly in the manufacturing sector. The results can be summarized as follows: 1. Undertaking independent R&D was perceived by the R&D executives questioned as the most effective channel of R&D spillovers at the intra-industry level. This was followed by reverse engineering for product innovations and the utilization of publications and information from technical meetings for process innovations. 2. Learning methods that rely on interpersonal communication were judged as moderately effective in the following order of importance: 1. publications and technical meetings, 2. conversations with employees from innovating firms, and 3. hiring away employees from innovating firms. Especially the last method is not valued as effective in the Swiss context. 3. Learning methods related to the patent system - licensing technology and patent disclosures in the patent office were seen as moderately effective or not effective at all 4. The effectiveness of the various channels of R&D spillovers varies from one industry to another. 5. Finally results of the methods of multivariate statistical analysis (correlation, principal components and cluster analysis) suggested that the various channels of R&D spillovers could be reduced to subgroups, so that patterns of learning of competitive technology could be established