Vecindarios y problemas socialesun acercamiento desde la estadística espacial

Résumé

The Social Disorganization Theory has extensively analysed the relationship between crime and violence and the social environment where it occurs. The hierarchical Bayesian spatial modelling proposes an advanced methodology to study the risk of social problems in the neighbourhoods. This paper uses this approach to analyse three types of social problems in the city of Valencia: drug-related crime, child maltreatment and intimate partner violence against women. Different hierarchical Bayesian spatial models were performed for each of these outcomes, and the influence of the neighbourhood-level variables in the spatial risk of these problems were assessed. The results showed that, regardless of the type of social problem analysed, both in the case of problems that occur in the street, such as drug-related crimes, and in problems that occur behind closed doors, such as child maltreatment and intimate partner violence against women, they show a spatial distribution, i. e., they are not randomly distributed in the city, but there are areas with greater risk than others. In addition, these spatial patterns are related to the neighbourhood characteristics, which would explain the unequal risk in the different areas of the city. Specifically, neighbourhoods with higher concentrated disadvantage and higher immigrant concentration show a higher risk of the different social problems. Risk maps can be very useful to guide local actions, effectively manage the social resources and develop preventive strategies for those neighbourhoods with higher risks.