A NOTE ON THE IMPACT OF TAX ON INCOME REDISTRIBUTION

Much attention has been paid to the problem of income distribution and the policy issue of redistribution in recent years. The OECD study (Sawyer, 1976) is one of the examples of the studies about income distribution that attempt to collect data internationally, particularly for developed countries, with an additional purpose of comparing the degree of income inequality. That study tried to compare inequalities making use of several measures such as Gini, Atkinson, Champerowne, Kuznets, Theil and log variance. The purpose of this note is not to review the OECD study carefully, but to concentrate on the problem of income redistribution policy by making intensive use of the excellent and unique OECD study. More concretely, the note attempts to evaluate tax policy of the OECD countries from income redistribution policy point of view. In general, international comparison of income distribution is of great difficulty, not only because the conceptual framework of income differs from country to country, but also because the method and the source of data are different, as the author of the OECD study emphasizes. Consequently, the empirical result derived from the OECD comparative study must be evaluated with caution.' The note will first present the coefficient of income redistribution for various countries. Secondly, the source for the contribution to the actual income redistribution policy is examined. Concretely, the contribution due to tax progressivity and to average tax rate is estimated numerically. Thirdly, several comments on income redistribution policy are made on the basis of the empirical results. Table 1 presents the coefficients of income redistribution by tax policy for various developed countries. The method of deriving these coefficients is simple: the coefficient C is obtained by dividing the difference between the measure of the pre-tax income inequality (Table 5 of the OECD study) and the measure of the post-tax income inequality (Table 6) by the measure of the pre-tax income inequality. The column headings of the Table identify the authors who proposed the particular formula for measuring the degree of inequality, or the method (log variance). Although the values presented in the table differ considerably not only from formula to formula but also from country to country, the rank from the higher coefficient of income redistribution to the lower coefficient is not significantly different from formula to formula. Table 2 shows the rank correlation cofficients between the Gini coefficient (which is the most widely used formula) and the other formulas. Since all the coefficients show values greater than 0.90, it is reasonable to conclude that the various methods for