Human capital and black-white earnings gaps, 1966–2019

Abstract

This paper estimates the contribution of human capital, measured using educational attainment and standardized test scores, to Black-white gaps in labor market outcomes in three separate samples of both men and women spanning the period from 1966 through 2019. There are three main findings. First, the magnitude of the reduction in the earnings gap that occurs after controlling for education and test scores has become much larger over time, suggesting a growing contribution of these traits to Black-white earnings disparities. Second, growth in the explanatory power of education and test scores has been primarily due to increases in the association between these traits and the likelihood of non-work, with no clear increases for hourly wages or other intensive earnings margins. Third, in most cases these trends are due to changes in the returns to the human capital traits, rather than changing racial gaps in the human capital traits themselves. These main findings apply qualitatively to both men and women. However, the magnitude of attenuation is much larger for women than for men, and conditional racial gaps in labor market outcome among women are often close to zero. Evidence suggests that this parity between Black and white women conditional on education and test scores is not primarily due to sample selection. Overall, the paper’s findings highlight how structural developments in the US labor market, such as increasing returns to skill and changing labor force participation rates by skill level and gender, have had disparate impacts across racial groups despite being race-neutral on their face.