A new class of  inequality measures based on a ratio of L-statistics

Summary
The choice of a suitable weight function is a feasible and desirable direction for inequality comparison. This paper proposes a new class of indices based on a ratio of two L-statistics in which the weight function in the numerator involves the expected values of order statistics from a random variable describing the normative attitude to the measurement of inequality. Also, the article describes and discusses a graphical technique, based on the QQ plot, which is useful to select the most
appropriate theoretical frequency distribution. A large-scale Monte Carlo study has been conducted to assess the sampling and asymptotic properties of some of a number of inequality indices. The simulations show that the newclass is able to provide indices that compete well with the traditional measures of income inequality and industrial concentration.

keywords
L-statistics, expected values of order statistics, inequality measures, simulations

METRON - International Journal of Statistics
2004, vol. LXII, n. 1, pp. 137-160

Back to main publications	Home page	Adobe pdf