Black-White Earnings Ratios Since the Civil Rights Act of 1964: The Importance of Labor Market Dropouts

Abstract

Published median earnings of wage and salary workers by race and sex are based on distributions that exclude those without earnings and those not employed as wage and salary workers at the time of the survey. It has been argued that the changing importance of these exclusions, rather than any improvement in the distribution of offered wages for blacks, accounts for the apparent increase in black-white relative median earnings. This paper presents a new method for correcting the published medians for such censoring. The apparent improvement in relative earnings is reduced but not eliminated by the correction.