A Critical Comment on the Taylor Approach for Measuring World City Interlock Linkages

In the study of economic‐geographic structures, the shifting focus from the national state to the city and its region has highlighted the lack of reliable interurban data sets. The abundance of usable data sets on international structures and flows has no counterpart when studying interurban relations, which makes it hard to draw any extensive conclusions regarding the structure of world city networks. Instead of relying on available data sets, Peter Taylor has developed a method for generating data sets that, it is argued, can be used in research on the structure of the world city network. In this approach, actors are defined as cities with internal attribute service values, values reflecting the presence of different transnational service‐producing corporations in each city. The structural values between each pair of cities are then established by a mathematical formula based on the service value of each firm in each pair of cities. This procedure can be criticized on two accounts. First, although internal attributes on exceptional occasions can be used as a proxy and as a rough estimate for structural values, such studies must have a firm theoretical underpinning in order to be valid from a network‐analytical perspective. If not, such generated structural values become nothing more than a function of internal attributes, thus losing the whole basic idea of social network analysis. Second, the Taylor function used for generating structural values can be questioned. Why should a large presence of TNC offices in a pair of cities imply a larger city interlock link than would be the case between a high‐ranked city and a low‐ranked city, as the city with the larger service value probably serves cities with a lower service value with economic command, control, and support functions?