The bidimensional decomposition of the Gini ratio. A case study: Italy

Abstract

This research is based on the decompositions of the Gini index. The two existing procedures of decomposition are connected: subgroup decomposition and income source decomposition. This bidimensional decomposition enables the computation of some new determinants of inequalities. It is possible to reckon the contribution of each source to the within-group and the between-group components of the overall inequality. This bidimensional decomposition is applied to the Italian consumption in 1989 and 2000.

Acknowledgement

I am greatly indebted to Camilo Dagum, Jacques Silber and Michel Terraza.

Notes

¹ This method yields a gross between-group component that is different from the well-known between-group component. The latter represents the differences in mean, whereas the former has a better specification. It is based on the inequalities between each pair of subpopulation.

² By matching the results of the exclusive methods, a different conclusion can be found compared with the simultaneous technique. Therefore, it is necessary to use carefully the exclusive methods notably when they serve to derive some economic policy implications.