Spatial heterogeneity in composite indicator: A methodological proposal

Most of the existing literature on the construction of composite indicator (CI) in the Benefit of the Doubt (BoD) framework fails to consider the exogenous external (tangible or intangible) factors, which might have influences on individual components of the CI. The present research proposes an original method to estimate the CI by considering the external factors directly into the robust BoD model. More specifically, a particular focus has been devoted to the “spatial conditioning” in which much of the underlying observed differences in CI scores may be attributed to a multitude of territorial conditional variables. The proposed method is illustrated using two data sets: one containing simulated data and the other one relative to real data covering the social supply delivered by the Italian municipalities. Results stress the importance of properly taking into account the local non-independence of the units in order to better separate the performances of individual units by the contribution of their territories.