The Grid of Sweden - A Micro-unit Analysis of Vulnerable Neighborhoods

Abstract: Through a national collection, the Swedish Police identify and classify vulnerable neighborhoods. Areas are assessed through police perceptions regarding high concentrations of certain problems and criminal activity, such as public acts of violence with risk of harming third parties, open drug markets and organised crime structures. The purpose of this study has been to see whether it is possible to statistically discover these neighborhoods based on socioeconomic and demographic data. Initially, in a national comparison, areas that are defined as vulnerable neighborhoods by the national collection, was compared with other areas in the country. This was done based on a statistical grid consisting of squares with the dimension of 250 x 250 meters, with each square holding information about socio-demographic data. The main aim has been to identify a statistical model that more objectively can identify squares that are vulnerable or not, compared to the police's more subjective assessment. Result from logistic regression analyses implies that vulnerable neighborhoods from the national collection show greater odds at having high concentrations of residents with foreign background, higher unemployment rates and more households with single parents. Lastly, the best fitted regression model for explaining these areas by the means of pseudo R2-value, were used to calculate a prediction value for each square. This value was then analysed using a GIS-software, to discover any areas that in the national collection was classified as vulnerable, but according to the model no longer met the criteria, and then vice versa. The overall result indicate that it is possible to discover areas with higher concentrations of certain characteristics seen in vulnerable neighborhoods, using spatial analyses and logistic regressions of micro-places, to more objectively classify these areas. By aggregating crime data, the result of this study can in the future mean a more effective implementation for police authorities.