|
|
|
| Author: | J.T.W. Alleblas |
Abstract:
This lecture has investigated whether there are possibilities of involving negative incomes in the determination of income dispersion.
The Gini coefficient was chosen from a number of criteria examined.
When this coefficient was applied to distributions with negative income, the coefficient proved capable of rising above 1. This would imply that distributions with negative incomes may in principle have a greater dispersion than distributions with positive incomes only, which may be presumed to be incorrect.
This phenomenon was examined and may be explained by the fact that the equation makes no allowance for the make-up of the maximum dispersion in the case of distributions with negative incomes.
For this reason the result will have to undergo a correction corresponding to the ratio of the areas of the maximum dispersion in the case of distributions with negative incomes and the maximum dispersion in the case of distributions with positive incomes only.
The corrected Gini coefficient was checked by actually calculating the areas by means of regression analysis.
The results of these calculations are such that the coefficient calculated in our way seems to be a good means of showing the dispersion of income distributions with negative incomes.
|
Download Adobe Acrobat Reader (free software to read PDF files) |
|