QED Working Paper Number
1531

This paper combines quantile-based disaggregative statistics and standard error formulas for the statistics to examine changes in the distribution of individuals’ incomes in Canada within a standard statistical inference framework.  Analysis focuses on decile means and income shares, Lorenz curves (as indicators of inequality change) and generalized Lorenz curves (as indicators of change in economic well-being).  The analysis confirms major previous findings as highly statistically significant and reveals much new distributional detail.  Significant and substantial inequality increases occurred over 1990-2015 for both men and women with much stronger increases for men.  As a result, men’s Lorenz curves have fallen sufficiently to change from being above women’s Lorenz curves before 2000 to then lying uniformly and significantly below them.  Generalized Lorenz curves for men are higher over all years than those for women.  However, the two GLC curves have been converging, and the middle 80 percent of the curves are estimated to meet within roughly a generation.  The study illustrates that it is straightforward to undertake distributional analysis within a standard framework of statistical inference.

Author(s)
JEL Codes
Keywords
income distribution change
tests for distributional change
income inequality change
Canadian income distribution
Working Paper