A new spatial concentration index: A transportation cost approach impartido por Dr. Mauro Ferrante, University of Palermo.
The topic of measuring spatial concentration has received great attention in the last decades. Beyond a major interest in measuring spatial concentration of economic activities (Ellison and Glaser, 1997; Krugman, 1991; Liu, 2014), many other disciplines are concerned with phenomena distributed across space (see for example, Freemana et al., 1996). Although most of the problems affecting the currently used indices for measuring spatial concentration are well documented in the literature, only few papers made the attempt to overcome the raised issues, and the proposed solutions are generally ad-hoc solutions. These approaches generally try to combine some known concentration indices (e.g. Gini index, Ellison and Glaser index) with spatial autocorrelation indices (Moran’s I), in the attempt of catching these two facets of spatially concentrated phenomena, namely the level of the observed phenomenon in the considered units, and the spatial distributions of the units (Arbia, Piras, 2009; Guimarães, et al., 2011). Starting from these premises, this paper proposes a new index for measuring spatial concentration based on the solution of the transportation problem. The proposed approach allows to take into account for the level of concentration of the phenomenon of interest according to its spatial distribution. After reviewing the main indices currently used for measuring spatial concentration, by highlighting their main criticisms, a new index for measuring spatial concentration is proposed. An empirical application on data derived from per-capita GDP for selected European Countries is reported and results derived from the proposed index are compared with those deriving from the currently used indices.
