Methods to recognize housing hardship hot spots for developing urban regeneration policies

The issue of housing in Italy remains of great interest to scholars, policy makers and social partners. Although 80% of Italian families own their homes, both the quality of these houses and difficulty in meeting the needs of those weaker sections of society below the so-called “poverty line” present significant contemporary issues. Such hardship is generally concentrated in densely populated urban areas. In light of this situation, there is an urgent need to review the strategies adopted in dealing with the housing crisis in large metropolitan areas: access to owned or rented housing, the absence of services functional to residence and the overcrowding of homes of vulnerable groups are just some of the issues that create situations of housing hardship, and thus urban poverty. The present case study arises from the need to identify risk areas (hot spots) of poverty, characterized by situations of housing hardship, towards which urban regeneration policies should ideally be directed. From these considerations it becomes necessary to define and develop characteristic indicators of hardship, able to estimate poverty in defined areas (Montrone, Perchinunno, Torre, 2007). The presence of a wide range of definitions on the issue of poverty creates the necessity to develop more than a single indicator but, rather, a group of indicators in order to define the living conditions of the different subjects, departing from a dichotomous logic and instead moving towards “fuzzy” classifications in which each unit can simultaneously belong and not belong to the category of poor under certain conditions. Finally, through the use of spatial clustering methods, the aggregation of territorially contiguous spatial units may be identified through the imposition of constraints on the various component units of each cluster (Patil and Taillie 2004; Kulldorff and Nagarwalla 1995). The SaTScan (Kulldorff 1997) and the DBSCAN (Density Based Spatial Clustering of Application with Noise) methods are of particular methodological interest to the present work and are applied to different case studies in order to monitor behaviour and compliance to the context in question.