Statistical tests for Lorenz dominance and related tests for first order dominance are applied to Taiwan's official income distribution statistics for the period 1976-1992. Inference tests for Lorenz dominance reveal whether significant changes in income inequality have occurred, while first order dominance comparisons show the effects of growth on economic well-being across time. Several conclusions emerge from the study: (1) correcting for sampling errors and considering entire income distributions leads to insights into changes in Taiwan's income distribution that are not revealed by descriptive summary statistics that are relied upon in official reports; (2) Lorenz dominance analysis reveals that, on the whole, income inequality in Taiwan has become significantly more unequal across time; (3) the so called U-shaped inequality experience is not confirmed for the 1976-1992 study period; (4) first order dominance analysis reveals highly significant improvements in the standard living in most year-to-year comparisons; and (5) whether overall social welfare has improved since 1976 depends upon the specific welfare criterion applied and the time period considered. Statistical tests for Lorenz dominance and related tests for first order dominance are applied to Taiwan's official income distribution statistics for the period 1976–1992. Inference tests for Lorenz dominance reveal whether significant changes in income inequality have occurred, while first order dominance comparisons show the effects of growth on economic well-being across time. Several conclusions emerge from the study: (1) correcting for sampling errors and considering entire income distributions leads to insights into changes in Taiwan's income distribution that are not revealed by descriptive summary statistics that are relied upon in official reports; (2) Lorenz dominance analysis reveals that, on the whole, income inequality in Taiwan has become significantly more unequal across time; (3) the so called U-shaped inequality experience is not confirmed for the 1976–1992 study period; (4) first order dominance analysis reveals highly significant improvements in the standard living in most year-to-year comparisons; and (5) whether overall social welfare has improved since 1976 depends upon the specific welfare criterion applied and the time period considered.
