Revisiting employment density as a way to detect metropolitan subcentres: an analysis for Barcelona & Madrid.

The changes on metropolitan areas characterized by the dispersion and concentrated decentralization of both employment and population has lead to a specialized research line addressed to analyze polycentrism. In this line there are two families of subcentre identification: the first based on the analysis of density, and other based on the analysis of mobility flows. Huge efforts have been paid, in the first family, to get robust models in statistical terms. Nevertheless, very little attention has been paid to the very object of analysis: employment density. In this paper a different approach to density calculus is proposed, it allows for prioritize, as subcentres, those municipalities that have features close to the urban paradigm of central cities in Southern Europe: have residents, are attractive in labor terms because retain resident employed population, and at the same time, attract commuters; and impact on the overall density function and have functional relations with other municipalities. The efficacy of the proposed density is tested in the metropolitan areas of Barcelona and Madrid. The analysis suggest that the municipalities prioritized as subcentres using this 'compound density' excels in modifying both the employment and demographic metropolitan density functions, and are more diverse, concentrate more central activities, and strengthen functional relations with other municipalities in comparison to the results of standard density.