Interpretation—redistribution requires the transfer of wealth from high to low incomes.
André Decoster and Guy Van Camp 98. Decoster has a PhD in economics from KU Leuven. Van Camp works for KU Leuven’s Center for Economic Studies, December 1998. “The Unit of Analysis in Microsimulation Models for Personel Income Taxes: Fiscal Unit or Household?”, Center for Economic Studies Discussion Paper Series DPS 98.33. https://econpapers.repec.org/paper/eteceswps/ces9833.htm
If the main objective of the evaluation of a tax reform is to give empirical content to the redistributive effects of different tax systems before and after the reforms, a formal expression of the concept of redistribution is required. Redistribution is defined here as the shift of income which occurs in the post-tax distribution from high to low incomes. Hence, it can be measured by comparing the pre-tax Lorenz curve with the post-tax concentration curve. Our measure of redistributive effect is therefore defined as: ΠRS = 2∫01 [LX-T(p) – LX(p)]dp (2) where the superscript of ΠRS refers to Reynolds and Smolensky (1977) who applied this measure to the US tax system, and LX(p) = ∫0y xf(x)dx/μ 0 ≤ p ≤ 1 and p = F(y) (3) LX-T(p) = ∫0y ( x-t(x))f(x)dx/μ(1 – t) 0 ≤ p ≤ 1 = p and p = F(y) (4) are, respectively, the Lorenz curve of pre-tax income (x) and the concentration curve with respect to pre-tax income of after-tax income (x – t(x)), with: p: the population shares of the pre-tax distribution f(x): the density function of x F(x): the distribution function of x μ: mean income (pre tax) t(x): the tax liability corresponding with taxable income x t: the average tax rate. Defining areas by: GX = 1 − 2∫01 LX(p)dp (5) CX-T = 1 − 2∫01 LX-T(p)dp (6) it is easily seen that the measure of redistribution ΠRS, can be rewritten as: ΠRS = GX – CX-T (7) where GX is of course the Gini coefficient of pre-tax income and CX-T is the area between the diagonal and the concentration curve of post-tax income with respect to pre-tax income. The difference between the Lorenz curve of post-tax income, and the concentration curve of post-tax income is important, if there is reranking. Only in the absence of reranking, CX-T equals the Gini coefficient of the post-tax income distribution, and ΠRS can be interpreted as the reduction of the Gini coefficient.