Wage Differentials Are Larger Than You Think
NBER Working Paper No. 168
This paper will employ a method (devised in Lazear  ) to estimate the unobserved component of wages. The size of this component will be calculated for non-whites and whites separately and then compared. Since, as it turns out, the component is larger for whites than non-whites, observed wage differentials understate true differentials. Furthermore, comparison of the period between1966-1969 with the 1972-1974 period reveals that this unobserved differential increased substantially over time. The results of this study suggest that although the pecuniary non-white -- white differential has narrowed substantially between 1966 and 1974 for young men, the on-the-job training differential has increased by almost the exact same amount. This implies that in real wealth terms there has not been any narrowing of the differential at all. This will become more apparent in later years as those non-whites who were hired into skilled jobs today fail to be promoted or obtain higher paying jobs elsewhere at the same rate as their white counterparts.
Published: "The Narrowing of Black-White Wage Differentials Is Illusory" The American Economic Review. Vol.69, Number4, pages 553-564, September 1979.