Spatial distribution of serious traffic accidents and its persistence over time

Abstract

This study intended to analyse the spatial distribution of serious road accidents in large cities to test the hypothesis of its over-time persistence. Yearly road accidents with casualties in the Milan municipality from 2015 to 2019 constituted this study data. In order to analyse the accidents spatial distribution, we divided the municipality territory according to three different partitions: one of 88 administrative units, two specially-made grids of respectively 88 squares and 5568 hexagons. We evaluated the over-time persistence of the TA spatial distribution through correlations, Cronbach Alpha, factor loadings and the KMO statistic applied to the accidents densities and Getis-Ord Gi* z scores of the units within the three territorial partitions. Repeated evidence emerged of a high and persistent differential in accidents between the territorial units, suggesting the existence of an infrastructure non-contingent factor underlying road accidents.