The Social Network Analysis of Techno-Economic Systems: Comparing Results Based on Binary and Weighted Networks

Abstract

The network analysis of a technological system combines the interindustry transactions with a matrix of sectoral innovative efforts as measured by R&D investment intensity. The matrixes of interindustry transactions of R&D-embodied products (innovations) are weighted matrixes where the interindustry flows measure the intensity of the innovation diffusion. In the past, studies using this approach in innovation studies have transformed weighted matrixes into binary matrixes of zero and one element where the flows less than a selected threshold value were considered to be zero and the flows greater than the threshold value were counted as one. Such matrix transformation leads to the loss of a great deal of information. In the present study, using degree and clustering coefficients for both binary direct as well as weighted direct techno-economic networks of the manufacturing sector of the German economy, we show that the binary directed network analysis is incapable of refined ranking of interindustry innovation transactions. The total degree index based on the weighted network of the German techno-economic system assigns a unique ranking to each sector, and clustering coefficients show that at least 75% of sectors in the network of Germany have two links with the other industries. However, the same indices based on the binary network are incapable of such refined ranking.