We study the Japan and U.S. patent records of several decades to demonstrate the effect of collaboration on innovation. We find that statistically inventor teams slightly outperform solo inventors while company teams perform equally well as solo comp
anies. By tracking the performance record of individual teams we find that inventor teams performance generally degrades with more repeat collaborations. Though company teams performance displays strongly bursty behavior, long-term collaboration does not significantly help innovation at all. To systematically study the effect of repeat collaboration, we define the repeat collaboration number of a team as the average number of collaborations over all the teammate pairs. We find that mild repeat collaboration improves the performance of Japanese inventor teams and U.S. company teams. Yet, excessive repeat collaboration does not significantly help innovation at both the inventor and company levels in both countries. To control for unobserved heterogeneity, we perform a detailed regression analysis and the results are consistent with our simple observations. The presented results reveal the intricate effect of collaboration on innovation, which may also be observed in other creative projects.
Companies are exposed to rigid competition, so they seek how best to improve the capabilities of their innovations. One strategy is to collaborate with other companies in order to speed up their own innovations. Such inter-company collaborations are
conducted by inventors belonging to the companies. At the same time, the inventors also seem to be affected by past collaborations between companies. Therefore, interdependency of two networks, namely inventor and company networks, exists. This paper discusses a model that replicates two-layer networks extracted from patent data of Japan and the United States in terms of degree distributions. The model replicates two-layer networks with the interdependency. Moreover it is the only model that uses local information, while other models have to use overall information, which is unrealistic. In addition, the proposed model replicates empirical data better than other models.
Many firms these days are opting to specialize rather than generalize as a way of maintaining their competitiveness. Consequently, they cannot rely solely on themselves, but must cooperate by combining their advantages. To obtain the actual condition
for this cooperation, a multi-layered network based on two different types of data was investigated. The first type was transaction data from Japanese firms. The network created from the data included 961,363 firms and 7,808,760 links. The second type of data were from joint-patent applications in Japan. The joint-patent application network included 54,197 nodes and 154,205 links. These two networks were merged into one network. The first anaysis was based on input-output tables and three different tables were compared. The correlation coefficients between tables revealed that transactions were more strongly tied to joint-patent applications than the total amount of money. The total amount of money and transactions have few relationships and these are probably connected to joint-patent applications in different mechanisms. The second analysis was conducted based on the p* model. Choice, multiplicity, reciprocity, multi-reciprocity and transitivity configurations were evaluated. Multiplicity and reciprocity configurations were significant in all the analyzed industries. The results for multiplicity meant that transactions and joint-patent application links were closely related. Multi-reciprocity and transitivity configurations were significant in some of the analyzed industries. It was difficult to find any common characteristics in the industries. Bayesian networks were used in the third analysis. The learned structure revealed that if a transaction link between two firms is known, the categories of firms industries do not affect to the existence of a patent link.
Technological innovation has extensively been studied to make firms sustainable and more competitive. Within this context, the most important recent issue has been the dynamics of collaborative innovation among firms. We therefore investigated a pate
nt network, especially focusing on its spatial characteristics. The results can be summarized as follows. (1) The degree distribution in a patent network follows a power law. A firm can then be connected to many firms via hubs connected to the firm. (2) The neighbors average degree has a null correlation, but the clustering coefficient has a negative correlation. The latter means that there is a hierarchical structure and bridging different modules may shorten the paths between the nodes in them. (3) The distance of links not only indicates the regional accumulations of firms, but the importance of time it takes to travel, which plays a key role in creating links. (4) The ratio of internal links in cities indicates that we have to consider the existing links firms have to facilitate the creation of new links.
