Innovation networks: the effects of collaboration in the oil and gas industries

Abstract

Collaboration is a premise of systemic thinking that underlies the theory of social networks. However, multivariate network researchers have not yet developed a generic model that explains this relationship. This qualitative study of the relationship between network metrics and the innovation performance of patents by Sinopec Limited, Royal Dutch Shell plc, BP plc, and Total S.A. aims to determine whether collaboration affects patent productivity given network metrics such as centrality, structural holes, and hierarchy. Social actors are direct agents of the network’s structure because they change, break, and build new ties. The behavior of the social group can be evaluated by analyzing the characteristics of their ties. Therefore, network analysis is an appropriate tool for the study of social actions at the structural, individual, and relational levels. The sample used in this study consists of 1,811 patents related to the C10 area of knowledge with regard to oil, fuel gases, and lubricants. The innovation networks for these patents are composed of 3,213 researchers. In the sample, Sinopec has 1,902 researchers, or 59%, followed by Shell (640, or 20%), BP (382, or 12%), and Total (289, or 9%). The study shows that Sinopec has higher patent productivity than do the other companies, with a descriptive level p-value of less than 0.05 and a 95% confidence interval. In contrast, BP, Shell, and Total cannot be differentiated based on patent productivity. The regression models in this study provided an explanatory power of 59.03% for Total, 42.43% for BP, 39.10% for Sinopec, and 14.14% for Shell. Although these models are limited to those specific companies, their statistically significant results explain patent productivity, confirming the hypotheses and providing a basis for future research. They are valid and can contribute to management activities, although they cannot predict the productivity of an innovation network.