On the decomposition of Gini coefficients by population class and income source: a spreadsheet approach and application

Abstract

There are in the literature many different formulae and methodologies for deriving and decomposing Gini coefficients. Needless to say, different methodologies have their own advantages and limitations. Some Gini formulae are unnecessarily complicated and impractical. Some are biased estimators and/or not decomposable. This paper presents a simple and exact formula and develops a systematic procedure for decomposition by population class and income source in a spreadsheet without using matrix algebra, integration, regression and covariances. The Gini formula and decomposition method can be used for individual data and for evenly or unevenly grouped data. Many complicated concepts regarding the Gini and its components are given an intuitive economic interpretation. A household survey data set for Sichuan, China, is used to demonstrate, step by step, how the methodology is applied in empirical analyses.