Statistical Discrimination and Gender Wage Gap: A Model

In this paper, I extend the model in Coate and Loury (CL) (1993) to show how statistical discrimination by employers can help create gender wage gap for men and women with equal earning potentials. Given that employers do not perfectly observe a worker’s skill type and partly rely on the average skills level of his (her) peers for inference purpose, employers’ differential treatment of male and female workers can create different skill-investment incentives for them, which in turn justify employers’ discrimination in the first place. The second result of this paper which is not possible within the original CL framework is that I point to the possibility that there exist circumstances under which the gender wage gap can not be eliminated without the formerly advantaged sex being negatively affected.