A1635
Title: On estimating the decomposition of the income inequality measures
Authors: Yuki Kawakubo - Chiba University (Japan) [presenting]
Kazuhiko Kakamu - Nagoya City University (Japan)
Abstract: The problem of simultaneous estimation of the generalized entropy (GE) inequality measures of the population is considered, and the GE of its sub-populations is based on the grouped data and decomposes the GE of the population into the between-group inequality and the within-group inequality. When GE measures are estimated assuming parametric distributions for each population and sub-population, it is impossible to obtain estimates compatible with the decomposition. Therefore, to achieve compatibility with the decomposition, the GE is estimated for each sub-population using a constrained Bayes estimator, using the estimates of the population GE and the between-group inequality as benchmarks. The proposed method is applied to Japanese data and the nationwide GE and the GE of each prefecture are estimated, decomposing the nationwide GE into between-prefecture inequality and within-prefecture inequality. Furthermore, the GE of each prefecture is decomposed into between-municipality inequality and within-municipality inequality.