Regional and demographic determinants of poverty in Brazil

Abstract

Why are some areas poor? One possible reason could be that they exhibit high concentrations of households with the “wrong” characteristics, such as a low level of education, larger families, etc. If households were free to migrate, any social policy aimed at improving educational levels in these areas, for example, would end up stimulating migration. As a result, the inequality observed after the adjustment process would indicate a scenario of no geographical differences in living standards (income, living conditions, etc.). It would be completely explained by the different sets of personal characteristics across regions. In this case, there would be no place for targeting a specific region in the implementation of policy measures: there would be no need for “regional” policies. On the other hand, moving can be costly and risky for poor people, because of transportation costs and other personal (emotional, noneconomic) factors. As Ravallion (1993) points out, it could be difficult for policymakers to target household characteristics, indicating that geographic targeting could be needed, even after factoring in migration. Park et al. (2002) evaluated the effectiveness of the regional targeting of a large-scale, Chinese poverty alleviation program, and found out that political factors affected targeting and that leakage grew as a result of increased coverage. Bird and Shepherd (2003) tested social and political exclusion variables and typical geographical variables in explaning rural poverty in Zimbabwe, and showed that proximity to urban areas is an important factor.