Abstract

This study suggests the use of a geographically-varying Gini coefficient to better understand local dynamics between income inequality and socioeconomic attributes. We estimate local Gini coefficients by disaggregating the traditional Lorenz curve approach at small census areas using spatial microsimulation techniques. As our case study, we chose the city of Santiago, Chile. Despite being ranked as a high-income country by the World Bank, Chile is the most unequal OECD country and the seventh worst in the world. Results reveal interesting spatial patterns that relate income inequality to income and educational level. For instance, the most unequal areas of the city correspond, perhaps against all odds, to the most well-off neighbourhoods having, in addition, the highest educational level. This result suggests that from a certain level of high income in areas with high-income inequality, the spatial cohesion of neighbourhoods may be mainly caused by the educational level of households. Results also reveal that the least unequal areas are mostly observed across neighbourhoods with low educational levels and low per capita incomes.