Analysis of ownership network of European companies using gravity models

Abstract

Social network analysis is increasingly applied to modeling regional relationships. However, in this scenario, we cannot ignore the geographical economic and technological nature of the relationships. In this study, the tools of social network analysis and the gravity model are combined. Our study is based on the Amadeus database of European organizations, which includes 24 million companies. The ownership of parent subsidiaries was modeled using economic, technological, and geographic factors. Ownership was aggregated to the NUTS 3 regional level, to which average corporate profitability indicators, the GDP per capita characterizing the economic environment, and the number of patents, which is a proxy of the technological environment, were assigned to NUTS 3 regions. The formation of the ownership network between 2010 and 2018 was characterized using this dataset. As the proposed model accurately describes the formation of ownership relationships marked with edges, it is possible to estimate network properties, such as modularity and centrality.