Sulla interpretazione dei parametri dei modelli analitici per la distribuzione del reddito personale

Abstract

For continuos and non-negative random variables, Zenga specifies a partial ordering using ratio between population quantiles of income distribution as a fruitful device to obtain a partial ordering of Lorenz's and Zenga's concentration curves respectively. The purpose of this paper is to use Zenga's contribution to investigate the concentration curves behaviour when one inequality parameter changes while the others remain constant. An important feature of this ratio between population quantiles is that it gives a straightforward information about the Lorenz' and Zenga's concentration curves behaviour without requiring the numerical knowledge or estimation of their mathematical forms. The quantiles ratio is used to analyse the concentration behaviour resulting from changes of each inequality parameter at a time for the following income distribution models lognormal, Singh and Maddala, Dagum Type I, and the generalized-Pareto specified by Stoppa.