Document Type


Publication Date



labor markets, worker pay, wages, wage discrimination


Labor market discrimination is defined as a failure to receive compensation equivalent to workers' productivity. In an efficient labor market, a worker's productivity attributes-labor force experience, education, tenure, etc.-and innate ability will be duly rewarded regardless of race, gender, ethnicity, or other individual characteristics. Workers receive pay commensurate with individual productivity. Employers do discriminate in pay and employment between individual productivity characteristics and ability. This is not labor market discrimination. It reflects market efficiency, since workers with productivity attributes and abilities that are highly demanded by employers receive the highest wage rates. It is also well known that females and minorities consistently have lower earnings than white male workers. This implies that females and minorities either have lower productivity than white males or there is labor market discrimination. It is generally accepted that females and minorities as a group have lower mean education and labor force experience as well as other measurable productivity attributes. To the degree that women and minorities are prevented from acquiring the high paying productivity attributes, the outcome reflects pre-labor market discrimination rather than labor market discrimination. After adjusting for measurable productivity difference between racial and gender groups, about 13 % of the 30% difference between white and black male earnings is attributed to differences in productivity characteristics and about 17 % of the earnings differential is potentially discrimination. When comparing white women to white men, about 25% of the 33% difference in male and female pay is due to potential discrimination while 8 % of the pay differential is due to productivity differences. The discrimination component in the wage decomposition may or may not measure market discrimination. The difference in the rate of return for productivity variables (i.e., regression parameter estimates) may reflect differences in quality. For example, blacks may receive less education due to lower quality schools even though they may have the same number of years of education as whites. Women may have different quality labor force experience because they may plan to exit the labor force to raise children. The difference between constant terms in the wage decomposition, or the unexplained residual, may estimate labor market discrimination or it may reflect unobservable productivity characteristic differences between the groups. Again, the discrimination component of the wage decomposition may measure labor market discrimination, but it also may reflect pre-market discrimination or decisions with respect to specific groups to invest in different productivity factors.


Department of Economics, South Dakota State University

Series Number


Number of Pages


Included in

Economics Commons