Measurement of Organized Crime in the Italian Provinces with Spatially-Clustered Heterogeneity

In this paper, we propose a novel approach for measuring the presence of Organized Crime across Italian provinces. Recognizing that Organized Crime is a spatially heterogeneous phenomenon, we introduce a composite indicator that accounts for such structural heterogeneity while weighting individual indicators. Our method combines spatially-constrained hierarchical clustering with cluster-specific Principal Component Analysis. Italian provinces are grouped based on a spatially informed dissimilarity matrix, and a separate Principal Component Analysis is performed within each cluster to extract context-sensitive indicator weights. Using publicly available data, we adopt this procedure and new elementary indicators to provide an updated map of the Organized Crime phenomenon diffusion at the provincial scale in Italy.