New Income Inequality Measures as Efficient Tools for Causal Analysis and Planning

A new system of inequality indicators is introduced in the paper. It has been elaborated for measuring income inequalities, but can serve as inequality measures for other phenomena as well. The new indicators not only measure the degree of the inequality but are also suitable for measuring its economic motivation. Properties of the new measures, including large sample properties of their estimators, are discussed. The case of the lognormal distribution is of special interest. ONE OF THE methodological problems that have recently arisen in planning consists in forecasting the personal income distribution of the population for some future period. To tackle this problem an experimental simulation model has been worked out by the authors at the Hungarian Planning Office 5,6. The chief aim of this model is to trace, on a sample of households selected at random, the influence of certain economic and demographic phenomena on the income distribution of the population living on wages and salaries. This study deals not with the special problems of these model procedures, e.g. the follow up of changes in wages and salaries, family allowances, pensions, employment, and demographic variables, etc. (being consistent with the plans of national economy). Instead we wish to give some information about our tools of measurement for analyzing the degree and the causes of the inequalities of factual and simulated income distributions. In other words, in connection with this simulation experiment we needed measures which, besides indicating the effect of the various plan variants on income inequality, provided us the opportunity for analyzing the extent to which certain factors, meaningful and important from the economic and planning points of view, contributed to the inequality. As far as we know, none of the various known measures of inequality was used for this latter purpose.