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

TLDR: In this paper, the effect of labor market dropouts on median earnings by race was investigated. But the restriction on the universe from which published median earnings data by race were discussed explicitly, and the restrict ion most commonly addressed in previous work was found to be less important than an undiscussed restriction.

Abstract: Previous analyses of postwar black/white earnings ratios have found a more rapid rate of increase in the period since 1964 than before. The reason for this acceleration is unresolved. One view is that federal equal-employment activities have increased the relative demand for black labor. An alternative view is that rising relative earnings reflects (1) reductions in relative supply and (2) the "statistical" effect of low earners raising median earnings by withdrawing from the labor market. This study differs from previous work on the subject in two ways. First, the restrictions on the universe from which published median earnings data by race are calculated are discussed explicitly. The restrict ion most commonly addressed in previous work (having positive earnings in the year in question) is found to be less important than an undiscussed restriction (being employed as a wage and salary worker the following March). Second, data on the distribution of earnings are used to determine the effect of labor market dropouts on median earnings, instead of trying to estimate this effect (as well as demand and supply effects) from time series data. This permits comparison of "corrected" and "uncorrected" post-1964 trends. For males, about half of the "uncorrected" trend remains after the relative earnings variable is corrected for labor market withdrawals. For females, between half and four fifths remains.