Gender, price, and quantity effects in U.S. earnings inequality: revisiting counterfactual density estimates

I decompose changes in the U.S. household earnings distribution from 1975 to 2018 to examine the labor market processes underlying its evolution over time. I model the distribution of earnings as a function of price effects (wages) and quantity effects (work hours and household employment), each of which are specified separately for men and women, and then apply a semi-parametric density estimation technique to infer their contributions to distributional measures. Results indicate that changes to the male wage distribution explain much of the growth in earnings inequality, but that its contribution varied greatly over time, with peak contributions in the mid 1990s; changes in female work hours have actually mitigated inequality growth, particularly by raising earnings in the lower and mid portions of the distribution, with more consistent effects over time. These results demonstrate the relevance of work hours in addition to wage rates in explaining earnings inequality growth, and the importance of gender differences therein.