Port centrality and the Composite Connectivity Index: Introducing a new concept in assessing the attractiveness of hub ports

A lot of ink has been shed lately on the concept of port connectivity. This is particularly true currently in view of the strength of global shipping alliances (GSA), their ability to jointly ‘manage' the supply of tonnage, and the negative impact such power has had on the frequency of services, the number of companies calling at a port, on containership sizes, and call sizes, i.e., on port connectivity. However, connectivity alone cannot explain the importance of a port as an international hub, its attractiveness to shippers, and its ability to develop new transshipment traffic (no matter how well connected a port is in the Arctic, or in Tierra del Fuego, it will never assume hub-port status). We argue that connectivity needs to be combined with measures of centrality, as these are derived from network theory. We thus introduce the novel concept of composite connectivity: Through an innovative use of two-stage data envelopment analysis (DEA) and complex network theory, we first evaluate the efficiency of ‘basic connectivity' and use this as input in the second stage, which measures the strength of centrality. To do so, we employ such network theory measures as betweenness centrality, closeness centrality, and eigenvector centrality. The "Composite Connectivity Index" — CCI is thus obtained as the ratio of (our measures of) port centrality to port connectivity. The top nine mainland China ports are used as a case-study. Our results (and rankings) conform to the general perception on the international importance of the ports of Shanghai, Shenzhen, and Hong Kong, thus demonstrating the validity of our model. The usefulness of CCI as a decision-support tool for ports with hub aspirations is, we believe, obvious.