Geographical Distribution of Crime: A Spatial Econometric Analysis

Abstract

Since long time social sciences have focused their attention on the causes of crime activities and this attention evolved over time. The sociological school emphasised the role of neighbourhood in delinquency activities, their stability over time and the existence of a negative relation between crimes and business centres (Shaw and McKay, 1942). This perspective stimulated the analysis of the existing nexus between crime activities and their geographical proximity. The economics approach moved differently. Since late 60s the economic analyses devoted their attention to the detection of mechanisms affecting the choice and behaviour of criminals (Becker, 1968; Stigler, 1970; Ehrlich, 1973). But nowadays many analyses have moved further detecting the socio-economic causes (i.e. income inequality and unemployment) that affect crime (Merlo, 2004). Recently, also thanks to the technological advance – i.e. more user-friendly software and the possibility to easily mange GIS applications – some scholars belonging to different disciplines moved in this direction and the geographical analyses of crime started spreading very fast (Anselin et al. 2000). Respective to Italian crime, after an extensive literature on qualitative analysis, in the last few years scholars from different research fields have showed an increasing interest on the quantitative evaluation of crime; viz. the assessment and the measuring of criminality and its principal determinants (Marselli and Vannini, 1997; Marselli and Vannini, 1999; Buonanno and Leonida, 2003). In fact in Italy the crime phenomenon is characterized by some stylized facts: high spatial and time variability; presence of criminal organizations deep-seated in local territorial areas; and high percentage of crimes connected to economic reasons (Marselli and Vannini, 2000). The goal of this paper is twofold. Firstly we want to detect the spatial diffusion of crime activities (distinguishing between criminal activities per se and administrative and civil criminal activities) in Italian provinces. Secondly using spatial econometric techniques, we want to identify an economic model of crime that depicts the spatial interactions among criminality rates and their principal socio-economic determinants. Finally, using network analysis techniques, we try to detect if geographical spillovers are more relevant than other aspects (i.e. social networks, institutions or environmental aspects) in shaping crime in Italian provinces.