An alternative approach to the calculation and analysis of connectivity in the world city network

The purpose of this paper is to contribute to the empirical 'world city network' (WCN) literature by proposing an analytical framework that considers the significance of the positions of cities in the urban networks created by multinational corporations. To this end, we introduce an approach in which network centrality measures are interpreted against a randomized baseline model that retains the network's original degree distribution. Our framework is applied to Taylor's (2001 Geographical Analysis 33 181-194) specification of world cities being 'interlocked' through the office networks of globalized service firms. We thereby develop a new analytical take on Taylor's specification by using a primary-linkage algorithm that produces a one-mode directed graph based on the initial two-mode city and firm network data. We then randomize the resultant empirical network with a bootstrapping simulation approach, and compare the simulated parameters of this null model with our empirical network parameter, that is, betweenness centrality. We show the potential and limitations of our framework by discussing the position of cities such as Tokyo, Sydney, Melbourne, Almaty, and Karachi, which seem to hold more strategic and valuable positions than is suggested in Taylor's analyses. We discuss the wider significance of our approach by showing that the analytical strategy of interpreting measures against a randomized baseline model can also be adopted in other WCN approaches, including those using other data sources and network-related measures.