Elasticidades da pobreza: novas fórmulas de cálculo e avaliação de diferentes pressupostos

Abstract

The paper presents the expressions to compute the Foster, Greer and Thorbecke poverty indices assuming a log-normal income distribution and shows that these expressions produce good estimates of those indices for Brazil. Next, the paper derives expressions for the elasticities of those poverty measures with respect to the mean income and to inequality indices. Finally, using panel data for the 27 Brazilian states (or federation units) from 1992 to 2004, it is shown that the elasticities of the poverty measures with respect to inequality obtained with the log-normal method fit the data better than the elasticities estimated using Kakwani’s (1993) assumption about the pattern of the Lorenz curve change.