Ranking Lorenz curves, correlations between indices and BINI (Bidimensional INequality Indices)
We establish a general methodology for ordering and comparing Lorenz curves with similar Gini indices. For this purpose, we introduce the notion of raking function, a mapping that quantifies the relative position of a Lorenz curve with respect to others with the same Gini coefficient. Several examples of ranking functions are presented together with their precise meaning and economic interpretation. We also define the index-correlation, a measure of the linear relationship between two indices. As an application, we construct bidimensional and uncorrelated inequality indices that provide extra information to the Gini index.
These two-dimensional indices distinguish and rank distributions with comparable Gini indices and can be included in statistical models avoiding possible multicollinearity problems.
Palabras clave: Gini index Inequality Lorenz curve Stochastic orderings