Measuring urban polycentrism: a European case study and its implications

New models of urban structures have emerged based on the assumption that metropolitan areas are increasingly decentralized, central business districts becoming less important in terms of employment and new subcentres emerging at the edge of cities. This article presents a method of identifying subcentres in suburban areas. This method is then applied to the question of whether subcentres exist on the periphery of medium-sized European cities and, if so, what kind of activities are located there. This is done by empirically describing and explaining the spatial structure of four urban regions in Belgium. Several spatial analysis techniques are used at different levels of data aggregation. Local autocorrelation is particularly appropriate for detecting clusters of employment, while shift-and-share analyses are useful for looking at the developmental trends. We show that, despite an overall trend to decentralization, polycentrism is still weak.