Statistical Discrimination, Affirmative Action, and Mismatch

This paper analyzes the economic consequences of a±rmative action in the presence of statistical discrimination. In the model, workers with di®ering abilities have com- parative advantages in jobs with di®ering complexities. Employers, having a biased belief on the ability of minority workers, require higher credentials when promoting them to more productive jobs, which discourages their human capital investment. When a±rmative action policy is enforced, some under-quali ̄ed minority workers are promoted to di±cult jobs. Those workers, as well as some majority workers who are over-quali ̄ed for, but have to take, easy jobs lose because their comparative advan- tages are not utilized. This ine±ciency due to mismatch is not necessarily outweighed by the long term gain brought about by the policy, if groups di®er substantially in their human capital investment costs. Appropriately reinterpreted, the model explains why drop-out rates and the returns to college education di®er between blacks who attend black and non-black colleges.

Download Paper

Paper Number
95-06
Year
1995
Authored by