Vulnerabilidade à pobreza dos municípios brasileiros entre 2000 e 2010: uma abordagem multidimensional

Abstract

The poor today may not be the poor of tomorrow. In this context, the concept of vulnerability is introduced as the probability of well-being loss. This study investigates factors capable of preventing the incidence and the deepening of poverty, and acting towards overcoming poverty as well. In order to consider the several aspects related to poverty, this study of vulnerability to poverty creates a Multidimensional Poverty Index (MPI), based on the Capabilities Approach and data from Demographic Census of 2000 and 2010. Differently from what is commonly used in the vulnerability to povertys literature, the focus is on the municipality instead of the individual or the family. The analysis of MPI shows an improvement of the multidimensional poverty picture in all national territory between 2000 and 2010. The Ordered Probit model is applied to the investigation of municipalitys vulnerability to poverty. The results confirm the idea that human capital contributes positively to the vulnerability reduction and so it does the better quality of life. Economies that are more capable of producing more business opportunities to the locals also have positive effects on the reduction of that problematic. On the other hand, the size of the municipalitys population contributes in a way of increasing its vulnerability to poverty. The municipalities considered as vulnerable were defined as those that show higher probability of descending movement in the MPIs distribution than the probability of ascending movement and the probability of stability along the MPIs distribution. Following the national tendency of improvement in the index made in this study, a reduction in the number of vulnerable municipalities is observed all over the Brazilian regions. With the use of the Exploratory Spatial Data Analysis, we confirm the hypothesis that poverty and vulnerability to poverty treated separately in this study are not spatially random.