Looking at the relationship more systematically, Table 7.4 gives a correlation matrix for the relation between top income shares and broader measures of income inequality. Usingdatafrom theLIS,the WIID,andthe WTIDover thepast30 years,the tableshows Pearson correlations between three top income shares (the top percentile, the top decile,and the lower nine percentiles in the top decile) and the Gini coefficient, the Atkinson index using two inequality aversion parameters, and the income ratios between the 90th percentile and 10th percentile (P90/P10) and the median (P90/P50). The correlations are the lowest for the WIID Gini coefficients, 0.25 and 0.42 for two of the top share measures. When using the LIS data, correlations are markedly higher, between 0.53 and 0.57 for the top percentile and between 0.64 and 0.74 for the two other income shares.45