Variance Estimation for Measures of Income Inequality and Polarization - The Estimating Equations Approach

Milorad S. Kovacevic and David A. Binder

Abstract:
The estimating equations technique for variance estimation is demonstrated on a variety of income inequality and polarization measures when data are obtained in a complex survey. This method, based on the Taylor linearization, is computationally nonintensive and easy to implement. Six different measures are considered. An example based on data from the Canadian Survey of Consumer Finance is given.

Keywords:
Coefficient of variation; exponential measure; Gini index; Lorenz curve ordinates; polarization curve ordinates; polarization index.