Choosing measures of inequality for empirical applications

Abstract

This paper is concerned with the distribution of income and the problem of choosing summary measures of inequality for empirical applications. By introducing a simple transformation of the Lorenz curve one is led to three measures of inequality, which jointly prove to represent a fairly good approximation of the inequality in a distribution function and also yield essential information about the shape of the income distribution. The paper also demonstrates that this type of inequality measures have an explicit normative foundation as well as an attractive statistical/geometrical interpretation. Furthermore, it is shown that the measures' sensitivity to transfers depend on the shape of the income distribution. Keywords: Measures of inequality, statistical and normative characterization of inequality.