Modelling crime rates using economic, social and spatial variables in urban areas

The definition of crime has varied across time and geography, and it is hard to come up with a globally definition. However, understanding the sources of crime is more crucial in crime prevention and creating livable environments. This research aims to analyze the variables affecting burglary crime rates in urban settings using data for 50 randomly selected streets for the years 1998-2006 in Izmir, Turkey. Four major crime factors are described in the literature: (1) social and cultural, (2) economic, (3) demographic, (4) spatial factors. The final model includes eight independent variables to explain the burglary crime rates: (1) street lenght, (2) wall lenght, (3) presence of iron fence, (4) whether the street is on a main road, (5) presence of mixed landuse, (6) presence of natural surveillance, (7) presence of bus stops, and (8) economic statu. We found that burglary crime rates increase with measure of the street, measure of the wall, measure of the iron fence and decrease with finding on highway, existence of some units except residential units, natural surveillance, existence of bus stop, economic statu. The model is able to explain %98 of the variation in burglary crimes (R2=0,983). All variables but one variables were found significant at the 0,10 level. Keywords: Crime, city, economic factors, social factors, spatial factors